|Home | About | Journals | Submit | Contact Us | Français|
Obligate vacuolar pathogens produce proteins that localize to the host cell-derived membranes of the vacuoles in which they reside, yielding unique organelles that are optimally suited for pathogen survival. Anaplasma phagocytophilum is an obligate vacuolar bacterium that infects neutrophils and causes the emerging and potentially fatal disease human granulocytic anaplasmosis. Here we identified APH_1387 as the first A. phagocytophilum-derived protein that associates with the A. phagocytophilum-occupied vacuolar membrane (AVM). APH_1387, also referred to as P100, is a 61.4-kDa acidic protein that migrates with an apparent molecular weight of 115 kDa on SDS-PAGE gels. It carries 3 tandem direct repeats that comprise 58% of the protein. Each APH_1387 repeat carries a bilobed hydrophobic alpha-helix domain, which is a structural characteristic that is consistent with the structure of chlamydia-derived proteins that traverse inclusion membranes. APH_1387 is not detectable on the surfaces of A. phagocytophilum dense core organisms bound at the HL-60 cell surface, but abundant APH_1387 is detected on the surfaces of intravacuolar reticulate cell and dense core organisms. APH_1387 accumulates on the AVM throughout infection. It associates with the AVM in human HL-60, THP-1, and HMEC-1 cells and tick ISE6 cells. APH_1387 is expressed and localizes to the AVM in neutrophils recovered from A. phagocytophilum-infected mice. This paper presents the first direct evidence that A. phagocytophilum actively modifies its host cell-derived vacuole.
Obligate vacuolar pathogens remodel the host cell-derived compartments in which they reside into unique organelles called pathogen-occupied vacuoles (PVs). PVs are developmentally arrested and sequestered outside the normal endocytic continuum and are optimal niches for intracellular survival (29). The PV membrane (PVM) provides a crucial interface between the host and the pathogen. Pathogen-encoded proteins that localize to the PVM play critical pathobiological roles, which include providing structural integrity to the PVM, hijacking vesicular traffic, and intercepting host signal transduction pathways. Therefore, identification and study of pathogen-derived PVM proteins are crucial for understanding the survival strategies of intravacuolar pathogens.
Anaplasma phagocytophilum is an Ixodes sp. tick-transmitted obligate vacuolar bacterium that infects granulocytes and causes the emerging and potentially fatal disease human granulocytic anaplasmosis (HGA) (14, 47). A. phagocytophilum is a member of the family Anaplasmataceae, which contains other tick-transmissible pathogens that infect peripheral white and red blood cells. Other Anaplasmataceae pathogens include Ehrlichia chaffeensis, the causative agent of human monocytic ehrlichiosis, and Anaplasma marginale, which infects bovine erythrocytes. HGA is an important cause of morbidity and is the second most common tick-transmitted disease in the United States. The clinical manifestations of HGA range from subclinical infection to severe disease, including death. The nonspecific symptoms include fever, chills, headache, malaise, and myalgia, and the severe complications include prolonged fever, shock, leucopenia, thrombocytopenia, high levels of C-reactive protein and hepatic transaminases, pneumonitis, acute renal failure, and hemorrhages. Fatal opportunistic infections have resulted when antibiotic therapy was not provided in a timely fashion.
Neutrophils are the primary effector cells of microbial killing and use oxidative and proteolytic mechanisms to destroy engulfed pathogens. Yet A. phagocytophilum establishes a vacuolar safe haven within neutrophils, as well as mammalian myeloid and endothelial cell lines and tick embryonic cell lines (21, 34, 47, 53). An A. phagocytophilum-occupied vacuole (ApV) lacks early endosomal markers and most late endosomal markers, does not acidify, is not sensitive to brefeldin A treatment, and avoids fusion with lysosomes and NADPH oxidase-carrying secretory vesicles and specific granules (10, 28, 33, 51). Tetracycline treatment promotes lysosomal fusion of the ApV (23), which confirms that bacterial protein synthesis is required to prevent ApV maturation and suggests that bacterial proteins incorporated into the ApV membrane (AVM) are likely to be important mediators of its altered fusogenic properties.
Because PVM alteration by pathogen-derived proteins is a paradigm of intracellular pathogenesis and there have been no previous reports of A. phagocytophilum-encoded proteins that localize to the AVM, we sought to identify and characterize such proteins. Since A. phagocytophilum is an obligate intracellular bacterium, we hypothesized that bacterial proteins induced in host cells represent AVM candidate proteins. An analogous strategy was used to identify the first three chlamydial inclusion membrane proteins (Inc proteins) (4, 42, 43). Storey et al. identified 3 such A. phagocytophilum proteins, P100, P130, and P160, which had apparent molecular masses of 100, 130, and 160 kDa, by screening an A. phagocytophilum genomic expression library with convalescent dog serum (45). Recombinant forms of these proteins were recognized by HGA patient antisera. P160, which has since been renamed AnkA because it carries a series of ankyrin repeats (13), is the first A. phagocytophilum type IV effector that was identified and does not localize to the AVM (27, 31, 40). P100 and P130 correspond to APH_1387 and APH_0032, respectively, in the annotated A. phagocytophilum proteome (25). Both of these proteins are acidic, carry tandem direct repeats, and contain segments that putatively traverse the AVM. Because it is induced upon infection and because it shares secondary structure characteristics with chlamydial Inc proteins, we performed detailed studies of APH_1387 and identified it as the first A. phagocytophilum-encoded protein that localizes to the AVM. This paper presents the first direct evidence that A. phagocytophilum actively modifies the host cell-derived organelle in which it resides and is an important step in deciphering this pathogen's intracellular survival strategies.
The human promyelocytic cell lines HL-60 (ATCC CCL-240; American Type Culture Collection [ATCC], Manassas, VA) and THP-1 (ATCC TIB-202) and A. phagocytophilum strain NCH-1-infected HL-60 and THP-1 cells were cultivated in Iscove's modified Dulbecco's medium supplemented with 10% fetal bovine serum (FBS) (IMDM-10) (Invitrogen, Carlsbad, CA) at 37°C in a humidified incubator in the presence of 5% CO2. In some experiments, A. phagocytophilum strain HZ, which was a gift from Ralph Horwitz of New York Medical College (Valhalla, NY) and Yasuko Rikihisa of Ohio State University (Columbus, OH), was used. The human microvascular endothelial cell line HMEC-1 (1) was obtained from the Centers for Disease Control and Prevention (Atlanta, GA). Uninfected and A. phagocytophilum-infected HMEC-1 cells were propagated in MCDM 131 (Mediatech, Herndon, VA) supplemented with 10 ng/ml epidermal growth factor (Becton Dickinson, Franklin Lakes, NJ), 1.0 μg/ml hydrocortisone (Sigma, St. Louis, MO), and 10% FBS in the presence of 5% CO2 in a humidified incubator.
The entire APH_1387 sequence was analyzed using several algorithms in order to obtain clues that would help predict its secondary structure and its ability to associate with the AVM. TMPred (www.ch.embnet.org/software/TMPRED_form.html) was used to determine if APH_1387 has a transmembrane domain. Protean, which is part of the Lasergene software package (version 8.02; DNASTAR, Madison, WI), was used to assess whether APH_1387 has regions of alpha helices and beta strands, alpha amphipathic sequences, and hydrophobicity, using the Garnier-Osguthorpe-Robson, Eisenberg, and Kyte-Doolittle algorithms, respectively (19, 22, 30). BLASTP (blast.ncbi.nlm.nih.gov/Blast.cgi) was used to identify protein sequences to which APH_1387 exhibits homology. The SWISS-MODEL (swissmodel.expasy.org), 3Djigsaw (bmm.cancerresearchuk.org/~3djigsaw), ESyPred3D (www.fundp.ac.be/sciences/biologie/urbm/bioinfo/esypred), and Geno3d (geno3d-pbil.ibcp.fr/cgi-bin/geno3d_automat.pl?page=/GENO3D/geno3d_home.html) algorithms were used in attempts to predict the tertiary structure of APH_1387.
The pMal-c2x vector (New England Biolabs, Ipswich, MA) was made into a Gateway destination vector using the Gateway vector conversion system (Invitrogen, Carlsbad, CA) by following the manufacturer's protocol; this yielded pMal-c2x/DEST, which was transformed into One Shot ccdB Survival competent Escherichia coli cells (Invitrogen). pMal-c2x/DEST was isolated from overnight cultures of transformants using a Qiaprep Spin miniprep kit (QIAGEN, Valencia, CA). The APH_1387 gene was amplified using primers 5′-CACCATGTATGGTATAGATATAGAGCTAAG-3′ and 5′-CTAATAACTTAGAACATCTTCATCG-3′ and Platinum Pfx DNA polymerase (Invitrogen); the underlined nucleotides correspond to nucleotides in a Gateway-compatible sequence. After we confirmed that the amplicon was the expected size by agarose gel electrophoresis, the amplicon was purified using a QIAquick PCR purification column (QIAGEN) and cloned without ligation into pENTR/D-Topo (Invitrogen) by following the manufacturer's instructions, which yielded the entry plasmid pENTR-APH_1387. The entry plasmid was transformed into chemically competent E. coli One Shot cells (Invitrogen) and subsequently isolated from an overnight culture. The insert sequence and cloning junctions were verified.
One hundred fifty nanograms of pENTR-APH_1387 was incubated with 150 ng of pMal-c2x/DEST and LR Clonase II (Invitrogen) at 25°C for 1 h to facilitate recombination of the APH_1387 insert downstream of the gene encoding maltose-binding protein (MBP) of pMal-c2x/DEST, which yielded pMal-c2x-APH_1387. Proteinase K was added to a final concentration of 0.18 μg/μl, and then the preparation was incubated at 37°C for 10 min, after which the LR recombination reaction mixtures were transformed into E. coli DH5α cells (Novagen, Madison, WI). Cultures of the transformants were grown in Luria-Bertani medium containing 100 μg ml−1 ampicillin at 37°C with shaking at 250 rpm. When the cultures were in the mid-logarithmic phase of growth (optical density at 600 nm [OD600], 0.4), expression of MBP-APH_1387 was induced at 37°C by adding isopropyl-β-d-thiogalactopyranoside to a final concentration of 0.5 mM, and the cultures were grown for another 3 h. Portions (80 ml) of the induced bacterial suspensions were harvested by centrifugation at 4,000 × g for 10 min at 4°C. Each pellet was resuspended in 5 ml column buffer (20 mM Tris-HCl [pH 7.4], 200 mM NaCl, 1 mM EDTA) and frozen overnight at −20°C. Frozen pellets were thawed in cold water and placed in an ice-water bath, in which they were sonicated. Soluble crude extracts were recovered by centrifugation at 9,000 × g for 20 min at 4°C. The crude extracts were loaded onto columns containing amylose resin (New England Biolabs) at a flow rate of 1 ml per min. The columns were washed with 12 column volumes of column buffer. MBP-APH_1387 was eluted in column buffer containing 10 mM maltose, and 3-ml MBP-APH_1387 fractions were collected and were visualized by Coomassie brilliant blue staining following resolution by SDS-PAGE. Desired fractions were pooled and concentrated by centrifugation through an Amicon Ultra-4 centrifugal filter with a 50-kDa cutoff (Millipore, Bedford, MA) to minimize the content of undesired breakdown products. Concentrated protein preparations were quantified using the Bradford assay (7).
MBP-APH_1387 was submitted to Animal Pharm Services (Healdsburg, CA) for production of rabbit polyclonal antiserum. Immunoreactivity against MBP-APH_1387 and native A. phagocytophilum APH_1387 was confirmed by Western blotting, as was the lack of recognition of APH_1387 by preimmune serum. Anti-MBP-APH_1387 is referred to below as anti-APH_1387.
To isolate host cell-free A. phagocytophilum dense core (DC) and reticulate cell (RC) organisms, infected (≥90%) HL-60 cells were mechanically disrupted by repeated passage through a 27-gauge needle. To isolate DC organisms, infected HL-60 cells were subjected to eight 8-s bursts on ice interspersed with 8-s rest periods using a Misonix S4000 ultrasonic processor (Farmingdale, NY) with an amplitude setting of 30, which destroyed host cells and fragile RC organisms but not hardy DC organisms, as confirmed by transmission electron microscopy (data not shown). Bacteria were separated from unbroken host cells and debris by differential centrifugation as described previously (48).
Whole-cell lysates were fractionated by SDS-PAGE, transferred to nitrocellulose, and screened using anti-APH_1387 followed by horseradish peroxidase-conjugated goat anti-rabbit IgG, as described previously (11). Actin and A. phagocytophilum Msp2 (P44) were detected using anti-human actin monoclonal antibody (MAb) (Sigma, St. Louis, MO) and anti-Msp2 (P44) MAb 20B4 (a gift from J. Stephen Dumler of Johns Hopkins University, Baltimore, MD) (37, 44), respectively. To quantitatively assess the increase in APH_1387 band signal intensity over the course of infection of HL-60 cells, densitometry was performed using ImageJ (National Institutes of Health, United States) (rsb.info.nih.gov/ij), and the ratios of the densitometric values for APH_1387 and actin at different time points were plotted.
Host cell-free A. phagocytophilum bacteria were isolated and synchronous infections were established as described previously (48). At the appropriate time points after addition of bacteria, aliquots were removed and processed for Western blot analysis as described above or for immunofluorescence or immunoelectron microscopy as described below.
C3H/HeN scid mice were infected with NCH-1 exactly as previously described (6). On day 8 postinfection, whole blood was collected by cardiocentesis in EDTA and centrifuged, and the leukocyte-rich buffy coat was removed and cytospun (Cytospin 4; Thermo Shandon, Pittsburgh, PA) onto glass slides at 113 × g for 5 min. The slides were fixed in ice-cold methanol and stored at −20°C until they were used. Ulrike Munderloh (University of Minnesota, Minneapolis, MN) kindly provided slides of paraformaldehyde-fixed A. phagocytophilum-infected ISE6 cells. A. phagocytophilum-infected HL-60, THP-1, or HMEC-1 cells were cytospun onto glass slides at 70 × g for 2 min, which was followed by fixation in 4% (vol/vol) paraformaldehyde in phosphate-buffered saline (PBS) for 1 h. All slides were placed in ice-cold methanol for 30 s to facilitate permeabilization and stored at −20°C until they were needed. Slides were screened using rabbit anti-APH_1387 and mouse anti-Msp2 (P44) MAb 20B4, both at a dilution of 1:500, followed by Alexa Fluor 594-conjugated goat anti-rabbit IgG and Alexa Fluor 488-conjugated goat anti-mouse IgG (Invitrogen), respectively, at dilutions of 1:1,000, as described previously (9). Slides were mounted with the Prolong gold antifade reagent (Invitrogen) and examined to determine the presence of ApVs and surface-bound bacteria with a TCS-SP2 AOBS confocal laser scanning microscope (Leica Microsystems, Bannockburn, IL) at the Virginia Commonwealth University Department of Neurobiology and Anatomy Microscopy Core Facility.
HL-60 cells were incubated with host cell-free A. phagocytophilum for 40 min at 37°C. The cells were washed three times with PBS and centrifuged at 300 × g for 5 min to remove unbound bacteria. At appropriate time points 9 × 106 cells were washed with cacodylate buffer and fixed in 1 ml of 3% paraformaldehyde-0.05% glutaraldehyde in 0.1 M cacodylate buffer for 45 min at 4°C. The cells were washed four times by resuspending them in 1 ml cacodylate buffer and incubating them at room temperature for 15 min. Next, the Virginia Commonwealth University Department of Neurobiology and Anatomy Microscopy Core Facility transferred washed cells onto 200-mesh Formvar-coated nickel grids. The grids were washed twice with PBS, blocked with 0.1% bovine serum albumin (BSA)-PBS for 1 h, and then stained sequentially for 1 h each with anti-APH_1387 (1:200) and goat anti-rabbit IgG (1:10) conjugated to 6-nm gold particles (Electron Microscopy Sciences, Hatfield, PA). Grids were viewed and images were recorded using a JEM-1230 transmission electron microscope (JOEL, Tokyo, Japan) equipped with a Gatan UltraScan 4000SP 4K × 4K charge-coupled device camera.
The Student's t test (paired), performed using the Prism 4.0 software package (GraphPad, San Diego, CA), was used to assess statistical significance. A P value of <0.05 was considered statistically significant.
APH_1387 is an acidic protein (pI 3.67) that contains 578 amino acids and has a predicted molecular mass of 61.4 kDa. It has 3 tandemly arranged direct repeats consisting of 93 amino acids (amino acids 180 to 272), 122 amino acids (amino acids 304 to 425), and 130 amino acids (amino acids 428 to 557) that together comprise 58% of the protein (Fig. (Fig.1A).1A). BLASTP searches using the entire APH_1387 sequence, the tandem repeat region (amino acids 180 to 557), or individual repeats identified no proteins that exhibited homology with APH_1387. A BLASTP search using the amino-terminal nonrepeat region (amino acids 1 to 179) identified a 93-amino-acid stretch exhibiting 32.3% identity to AM_924 and AMF_1226 of the A. marginale St. Maries and Florida strains, respectively (8, 16). APH_1387 is predicted to consist largely of amphipathic alpha helices (Fig. (Fig.1B).1B). Chlamydial Inc proteins possess transmembrane and/or hydrophobic domains that facilitate insertion into the PVM (3, 4, 41). Likewise, many PVM proteins possess hydrophilic domains that extend into the host cytoplasm to interact with host proteins (41). Although APH_1387 lacks a typical transmembrane domain, a 24-amino-acid sequence (AQVPVVAEAELPGVEAAEAIVPSL) that is part of each of the 3 direct repeats constitutes a bilobed hydrophobic domain (amino acids 200 to 223, 324 to 347, and 448 to 471) (Fig. (Fig.1B).1B). Bilobed hydrophobic domains are highly conserved among all chlamydial Inc proteins and are hypothesized to facilitate insertion of these proteins into the chlamydial inclusion membrane (41). Between the 3 hydrophobic domains are alpha-helical stretches that are largely hydrophilic. This periodicity of hydrophobic and hydrophilic domains could conceivably enable APH_1387 to traverse the AVM multiple times. Attempts to predict a tertiary structure for APH_1387 using SWISS-MODEL, ESyPred3D, 3Djigsaw, and GENO3D were unsuccessful because none of these programs was able to identify similar sequences with known structures on which to model the APH_1387 structure. Sequencing of the PCR products obtained for the entire APH_1387 coding sequence of A. phagocytophilum strains NCH-1, HZ, and HGE-1 revealed that this gene's sequence is the same in these 3 strains and the USG3 strain, the strain in which it was originally sequenced (45) (data not shown).
We initiated our characterization of APH_1387 by expressing it as an MBP fusion protein in E. coli and using the recombinant protein to raise rabbit polyclonal antiserum. To test the efficacy of anti-APH_1387 and to determine if APH_1387 is expressed during A. phagocytophilum infection of promyelocytic HL-60 cells, we performed Western blot analyses. Anti-APH_1387 recognized APH_1387 as a single band at ~115 kDa in an A. phagocytophilum-infected cell lysate but not in an uninfected HL-60 cell lysate (Fig. (Fig.2).2). A. phagocytophilum has a biphasic developmental cycle in which an infectious DC organism binds, enters, and changes into a replicative RC organism that subsequently divides by binary fission (48). The numerous RC organisms, which are noninfectious, revert to DC organisms before they are released to infect naïve host cells. We confirmed that a host cell-free population of A. phagocytophilum DC and RC organisms can be recovered following syringe lysis, which does not damage the fragile RC organisms, while a pure DC organism population can be recovered following sonication, which destroys the RC organisms (data not shown). To assess whether RC or DC organisms express more APH_1387, we screened Western-blotted lysates of host cell-free A. phagocytophilum organisms recovered following syringe lysis (DC and RC organisms) or sonication (only DC organisms) that were normalized using the levels of Msp2 (P44), which is a constitutively expressed outer membrane protein (24) (Fig. (Fig.2).2). The intensity of the APH_1387 band was considerably higher for the lysate derived from DC and RC organisms than for the lysate generated from DC organisms. After longer exposures, additional anti-APH_1387-reactive bands at ~61 and 90 kDa were detected in lysates derived from the DC and RC organisms, as well as infected HL-60 cells (data not shown).
We next screened A. phagocytophilum-infected HL-60 cells with anti-APH_1387 in conjunction with a MAb against Msp2 (P44) (37, 44) at 0.7, 8.5, and 24 h postinfection and visualized the cells by LSCM. At 0.7 h postinfection, very little or no APH_1387 was detected on Msp2 (P44)-positive organisms that were bound to the HL-60 cell surface (Fig. 3A to D). By 8.5 h, ApVs containing individual A. phagocytophilum bacteria were detected. At this time point, the AVM was positive for APH_1387 (Fig. 3E to H). At 24 h, the AVM was strikingly distinguished from enclosed bacteria by exclusive staining for APH_1387 (Fig. 3I to L). A. phagocytophilum bacteria within ApVs that were positive for both Msp2 (P44) and APH_1387 were green [corresponding to Msp2 (P44) staining] spheroid organisms, each of which was surrounded by a red ring (corresponding to APH_1387 staining) or a yellow ring [corresponding to APH_1387 and Msp2 (P44) staining] (Fig. 3J to L). At 24 h, 94% ± 3.5% of the ApVs were positive for APH_1387. The AVM was negative for Msp2 (P44) at all time points.
In addition to HL-60 cells, A. phagocytophilum infects and resides in ApVs in the human monocytic cell line THP-1, the human microvascular endothelial cell line HMEC-1, and the Ixodes scapularis embryonic cell line ISE6 (21, 34, 47, 53). To determine if APH_1387 is expressed and localizes to the AVM in each of these cell lines, A. phagocytophilum-infected THP-1, HMEC-1, and ISE6 cells were examined by LSCM at 24 h postinfection. As observed for HL-60 cells, virtually all ApVs in each cell line were positive for APH_1387 (Fig. 3M to X). Indeed, at 24 h postinfection 93% of ApVs were positive for APH_1387 in THP-1 cells (data not shown). Notably, the plasma membranes of heavily infected THP-1 cells (Fig. 3O and P), ISE6 cells (Fig. 3W and X), and HL-60 cells (data not shown) and occasionally uninfected cells that were adjacent to the heavily infected cells were also positive for APH_1387.
We next assessed the temporal expression and AVM localization patterns of APH_1387 over the course of a synchronous A. phagocytophilum infection of HL-60 cells. Host cell-free bacteria were added to HL-60 cells and allowed to bind for 40 min, which was followed by removal of unbound organisms. We and other workers have determined that it takes 4 h for bound A. phagocytophilum DC organisms to internalize in nascent vacuoles (5, 10, 28). Beginning at 8 h after addition of bacteria, which corresponded to ~4 h after entry, APH_1387 was weakly detected as a 115-kDa band by immunoblot analysis (Fig. 4A and B). The intensity of this band increased continually during the time course. Beginning at 18 h, several additional bands were also detected, and the primary bands were at ~61 and 90 kDa. As we have shown previously (12), the intensity of the Msp2 (P44) band, which is used as an infection control, increases throughout the course of an infection. APH_1387 was detected on >80% of the ApVs at 12 h and on >90% of the ApVs at all later time points by LSCM (Fig. (Fig.4C4C).
To more closely examine when APH_1387 is expressed and associates with the AVM during the course of infection, we screened synchronously infected HL-60 cells over a 48-h period by using immunoelectron microscopy. APH_1387-negative spheroid DC organisms were bound to the HL-60 cell surface at 0.7 h (Fig. (Fig.5A),5A), and some organisms had internalized into nascent vacuoles by 4 h (Fig. (Fig.5B).5B). By 8 h, which corresponded to ~4 h after entry, the internalized bacteria had changed into elongated, pleomorphic RC organisms and had begun to replicate. The surfaces of the replicating RC organisms were positive for APH_1387, as were portions of the AVM (Fig. (Fig.5C).5C). More APH_1387 was detected on the AVM and on intravacuolar A. phagocytophilum bacteria as the infection progressed (Fig. 5D to G). At 24 h, individual HL-60 cells contained numerous ApVs, each of which harbored several DC organisms that had rough outer membranes or contained multiple mature RC organisms that were changing into DC organisms (Fig. (Fig.5F),5F), which is consistent with our previous report on the replication kinetics of A. phagocytophilum in HL-60 cells (48). At this time point, APH_1387 heavily decorated the AVM, as well as the surfaces of intravacuolar DC and RC organisms. APH_1387 labeling of the AVM was most pronounced for mature ApVs that contained several A. phagocytophilum bacteria at 48 h (Fig. 5G and H).
Next, we investigated APH_1387 expression in vivo. MBP-APH_1387, but not MBP alone, was recognized by antiserum from an HGA patient (Fig. (Fig.6A),6A), which confirmed a previous report by Storey and colleagues (45). Likewise, Msp2 (P44)- and APH_1387-positive ApVs were detected by LSCM in murine buffy coats isolated from A. phagocytophilum-infected mice at 8 days postinfection (Fig. 6B to E).
APH_1387 is the first identified A. phagocytophilum-derived protein that localizes to the AVM. It is expressed very early following A. phagocytophilum entry into a nascent host cell-derived vacuole, is expressed continually throughout the intracellular part of the bacterium's life cycle, and heavily decorates the AVM. APH_1387 potentially traverses the AVM 3 or fewer times by means of the 3 bilobed hydrophobic alpha-helical domains that are predicted to be in each direct repeat. These 24-amino-acid hydrophobic domains could easily traverse the AVM, as their length exceeds the minimum length for transmembrane helices (20 amino acids) (38). The bilobed hydrophobicity motif of APH_1387 is a feature that is shared with all chlamydial Inc proteins (41). In contrast to APH_1387, however, each chlamydial Inc protein has a single bilobed hydrophobic motif that is 40 to 60 amino acids long. Portions of pathogen-derived PVM proteins are typically exposed on the PVM's cytoplasmic face, where they project into the cytoplasm to interact with host cell proteins (41). While anti-APH_1387-mediated immunogold labeling demonstrated that APH_1387 localizes to the AVM, we cannot accurately determine whether one or more portions of APH_1387 are presented on the AVM's cytoplasmic face.
APH_1387 is expressed and modifies the AVM in vivo in neutrophils during murine infection and in vitro during intracellular residence in human HL-60, THP-1, HMEC-1, and tick ISE6 cells, and it is expressed throughout the course of intracellular development. The examination of A. phagocytophilum whole-genome transcription profile data that was performed by Nelson and colleagues for infected HL-60, HMEC-1, and ISE6 cells revealed that APH_1387 is transcribed in each cell line (35), which corroborates the findings of our protein expression analyses. Thus, APH_1387 is conceivably important for A. phagocytophilum survival in all eukaryotic host cells that this bacterium infects. This hypothesis is further supported by the high degree of sequence conservation in at least 4 geographically diverse A. phagocytophilum strains. Notably, anti-APH_1387 stains the plasma membranes of heavily infected host cells and uninfected host cells that are adjacent to heavily infected cells, which suggests that APH_1387 may associate with the host cell plasma membrane.
Tandem repeat proteins of pathogenic bacteria have been implicated in hijacking host signaling pathways, adhesion, immune evasion, and other host-pathogen interactions (15, 20, 26, 27, 31, 50, 52, 56). The genomes of A. phagocytophilum and E. chaffeensis each encode multiple acidic tandem repeat-carrying proteins. Some examples that have been studied or identified are P47, P120, and a variable-length PCR target protein of E. chaffeensis and A. phagocytophilum APH_1387 and APH_0032 (P130) (18, 32, 45, 54). Like APH_1387, P47 and P120 have been observed to be associated with the PVM (18, 39). Unlike APH_1387, which is found on intravacuolar RC and DC organisms, P47 and P120 are preferentially expressed by DC ehrlichiae (18, 39). A protein alignment of APH_1387 and P120 (encoded by ECH_0039 in the annotated E. chaffeensis genome) (55) that was performed by Storey and colleagues revealed that two short repeat segments that help comprise each APH_1387 tandem repeat region share sequence similarity with portions of the tandem repeat region of P120 (45). E. chaffeensis P47 is an acidic (pI 4.2) tandem repeat-carrying protein that interacts with host molecules involved in cell signaling, transcriptional regulation, and vesicle trafficking (50). The target host molecules with which APH_1387 presumably interacts and/or the pathobiological role of APH_1387 is likely unique to A. phagocytophilum intracellular survival, as APH_1387 exhibits very little or no homology with any previously described protein. Alternatively, APH_1387 may be important for adding structural integrity to the AVM and may have no target ligand. The only other known Anaplasmataceae pathogen-derived inclusion membrane protein that has been identified is Ehrlichia canis AF21920, which is a slightly basic (pI 8.49) 24.0-kDa protein with an unknown function that lacks tandem repeats but does carry 5 nonlobed hydrophobic domains that are predicted to traverse the PVM (46).
Upon electrophoresis APH_1387 migrates primarily as a 115-kDa band and also produces less prominent bands; the primary less prominent bands are a 90-kDa band and a band at a predicted molecular mass of 61.4 kDa. While other workers have attributed this aberrant migration to glycosylation (17), our assessments of both native and recombinant APH_1387 indicate that neither protein is glycosylated (unpublished data). All Anaplasmataceae acidic tandem repeat proteins analyzed to date migrate at molecular weights considerably higher than their predicted molecular weights. Alternatively, the acidic nature of APH_1387 may prevent it from amply binding SDS, which would retard proper SDS-PAGE resolution, as has been observed for other acidic proteins and was recently reported for E. chaffeensis P47 by Wakeel and colleagues (A. Wakeel, X. Zhang, and J. W. McBride, presented at the Twenty-Third Meeting of the American Society for Rickettsiology).
Large amounts of APH_1387 associated with the surfaces of intravacuolar A. phagocytophilum RC and DC organisms are detected by indirect immunofluorescence and immunoelectron microscopy. Yet APH_1387 was not detected on the surfaces of glutaraldehyde-fixed host cell-free A. phagocytophilum bacteria that were recovered following mechanical disruption of infected HL-60 cells, washed with PBS, and added to naïve HL-60 cells, and it was only weakly detected on methanol-permeabilized bacteria by indirect immunofluorescence analysis. Accordingly, we hypothesize that APH_1387 is not integrated into the A. phagocytophilum outer membrane but instead is loosely associated with it and is easily washed away during the isolation procedure. We further hypothesize that this is because APH_1387 is transiently associated with the surfaces of intravacuolar bacteria while it is being secreted and that it subsequently integrates into the AVM. It is not known how APH_1387 is secreted. A. phagocytophilum encodes a type IV secretion system that is homologous to the system of Agrobacterium tumefaciens (36), and the A. phagocytophilum effector, AnkA, can be heterologously secreted by A. tumefaciens (31). The C termini of type IV substrates have a net positive charge compared to the rest of the protein sequence (49). This is not the case for the APH_1387 C terminus, which has a net negative charge (pI 3.34). Moreover, type IV secretion system-mediated delivery of effectors is contact dependent (2, 49). Thus, APH_1387 is probably not delivered to the AVM via type IV secretion. Also, A. phagocytophilum lacks genes encoding type II and type III secretion systems (25).
Multiple pathogen-encoded PVM proteins have been found for a diverse array of obligate vacuolar pathogens. Indeed, ≥45 and ≥65 Inc proteins have been confirmed and/or are predicted for Chlamydia pneumoniae and Chlamydia trachomatis, respectively (41). Thus, it is likely that APH_1387 is just one of a multitude of AVM proteins that are waiting to be identified and functionally characterized. Dissecting the roles of A. phagocytophilum-derived AVM proteins, confirming their route of delivery, and identifying their eukaryotic host cell ligands are areas that are ripe for investigation and are essential for fully comprehending the intracellular survival strategies of this unusual bacterial pathogen that effectively resides in the primary effector cell for killing microbes.
We thank J. Stephen Dumler of Johns Hopkins University for providing MAb 20B4; Erol Fikrig of Yale University for providing HGA patient antiserum; Ulrike Munderloh for providing A. phagocytophilum-infected ISE6 cells; Yasuko Rikihisa and Ralph Horwitz for providing A. phagocytophilum strain HZ; Naomi Walker and Dexter Reneer for technical assistance; and Jere W. McBride and Richard T. Marconi for invaluable discussions.
This work was supported by NIH grants DK065039 and AI072683 and by funds from the National Research Fund for Tick-Borne Diseases. The Virginia Commonwealth University Department of Neurobiology and Anatomy Microscopy Facility is supported in part by funds from NIH-NINDS Center core grant 5P30NS047463.
Editor: A. Camilli
Published ahead of print on 8 March 2010.