A Variable-Length PCR Target Protein of Ehrlichia chaffeensis Contains Major Species-Specific Antibody Epitopes in Acidic Serine-Rich Tandem Repeats

Culture and purification of ehrlichiae. E. chaffeensis (Arkansas strain) and E. canis (Jake strain) were propagated and purified by size exclusion chromatography as previously described (17, 31). The fractions containing bacteria were frozen and utilized as antigen and DNA sources.

Preparation of E. chaffeensis genomic DNA and antigens.Genomic DNA and antigens were purified from E. chaffeensis (Arkansas strain) as previously described (18).

PCR amplification of E. chaffeensis VLPT gene fragments.Oligonucleotide primers for the amplification of the E. chaffeensis VLPT gene fragments were designed manually or by using Primer Select (Lasergene v5.08; DNAStar, Madison, WI) according to the sequence in GenBank (accession number AF121232) and then were synthesized (Sigma-Genosys, Woodlands, TX) (Table 1). Seven gene fragments corresponding to the four single VLPT TRs (VLPT-R4, VLPT-R3, VLPT-R2, and VLPT-R1), the C terminus of VLPT (VLPT-C), the combination of repeats R3 and R2 (VLPT-R32), and the nearly full-length VLPT (VLPT-R4321-C) containing multiple repeats (R4, R3, R2, and R1) and the C terminus of the E. chaffeensis VLPT gene were amplified using a PCR HotMaster mix (Eppendorf, Westbury, NY), with E. chaffeensis (Arkansas strain) genomic DNA as the template (Tables 1 and 2). The thermal cycling profile was as follows: 95°C for 4 min; 35 cycles of 94°C for 30 s, the annealing temperature (3°C less than the lowest primer melting temperature) for 30 s, and 72°C for the appropriate extension time (30 s/500 base pairs); a 72°C extension for 7 min; and a 4°C hold.

Expression and purification of recombinant E. chaffeensis VLPT proteins.The amplified PCR products were cloned directly into the pBAD/Thio-TOPO (Invitrogen, Carlsbad, CA) or pTriEx-6 3C/LIC (Novagen, Madison, WI) expression vector. Escherichia coli cells (TOP10; Invitrogen) were transformed with the plasmids containing the E. chaffeensis VLPT gene fragments, and positive transformants were screened by PCR for the presence of the insert and proper orientation and were sequenced with an ABI Prism 377XL DNA sequencer (Applied Biosystems, Foster City, CA) at the University of Texas Medical Branch Protein Chemistry Core Laboratory. Recombinant protein expression was performed for 4 h after induction with 0.2% arabinose (pBAD/Thio-TOPO) or 0.5 mM isopropyl-β-d-thiogalactopyranoside (IPTG) (pTriEx-6 3C/LIC). Recombinant proteins were purified under native conditions, using HisSelect columns (for pBAD/Thio-TOPO; Sigma, St. Louis, MO) or StrepTactin Superflow columns (for pTriEx-6 3C/LIC; Novagen), and quantitated with the bicinchoninic acid protein assay (Pierce, Rockford, IL) according to the manufacturer's instructions.

E. chaffeensis VLPT synthetic peptides.Five synthetic peptides, corresponding to the N-terminal fragment (VLPT-N; 17 amino acids) and four individual TR units (R4, R3, R2, and R1; 30 amino acids each) of the E. chaffeensis VLPT protein, as well as seven overlapping peptides corresponding to the different regions of R3 (R3-1 to R3-7) and a 20-amino-acid N-terminal peptide of R4 (R4-N), were synthesized (Bio-Synthesis, Lewisville, TX) (Table 3). The lyophilized powder was resuspended in molecular biology-grade water (1 mg/ml).

Antisera.A convalescent-phase anti-E. chaffeensis dog serum was derived from an experimentally infected dog (no. 2251). Sera from HME patients were a kind gift from Focus Technologies (Cypress, CA). Rabbit anti-VLPT-R3 antiserum was generated against the synthetic E. chaffeensis VLPT-R3 keyhole limpet hemocyanin-conjugated peptide by a commercial vendor (Bio-Synthesis).

Gel electrophoresis and Western immunoblotting.Purified E. chaffeensis or E. canis whole-cell lysates or recombinant proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose, and Western immunoblotting was performed as previously described (15), except that primary dog and human sera were diluted 1:100 and rabbit anti-VLPT-R3 antiserum was diluted 1:2,000.

Carbohydrate detection.Glycan detection on the recombinant protein VLPT was performed with a digoxigenin glycan detection kit (Roche, Indianapolis, IN), as previously described (21).

ELISA.Enzyme-linked immunosorbent assay (ELISA) plates (MaxiSorp; Nunc, Roskilde, Denmark) were coated (0.5 μg/well; 50 μl) with recombinant proteins or synthetic peptides in phosphate-buffered saline (pH 7.4). Proteins and peptides were adsorbed to the ELISA plates overnight at 4°C with gentle agitation, subsequently washed thrice with 200 μl Tris-buffered saline containing 0.2% Tween 20 (TBST) and blocked with 100 μl 3% bovine serum albumin (BSA) in TBST for 1 h at room temperature with agitation, and washed again. Convalescent-phase anti-E. chaffeensis dog or human sera diluted (1:100) in 3% BSA-TBST were added to each well (50 μl) and incubated at room temperature for 1 h with gentle agitation. The plates were washed four times, and 50 μl alkaline phosphatase-labeled goat anti-dog or -human immunoglobulin G (heavy plus light chains) secondary antibody (Kirkegaard & Perry Laboratories, Gaithersburg, MD) diluted (1:5,000) in 3% BSA-TBST was added and incubated for 1 h at room temperature. The plates were washed four times, and substrate (100 μl of BluePhos; Kirkegaard & Perry Laboratories) was added to each well. The plates were incubated for 30 min in the dark with agitation, color development was read on a microplate reader (VersaMax; Molecular Devices, Sunnyvale, CA) at A₆₅₀, and data were analyzed by SoftmaxPro v4.0 (Molecular Devices). Optical density (OD) readings represent the means for three wells (± standard deviations), with the OD of the buffer-only wells subtracted.

Immunoelectron microscopy.Immunogold electron microscopy was performed on E. chaffeensis-infected DH82 cells as previously described (8), except that primary rabbit anti-VLPT-R3 peptide serum was diluted 1:10,000. Uninfected DH82 cells were used as a negative control.

Mass spectrometry.Mass spectrometry was performed using a matrix-assisted laser desorption ionization-time-of-flight (MALDI-TOF) mass spectrometer (Voyager-DE STR; Applied Biosystems) at the University of Texas Medical Branch Mass Spectrometry Core Laboratory.

Analysis of secreted VLPT protein. E. chaffeensis-infected DH82 cell culture supernatants (1 ml) were collected every day without disturbing the cell monolayer and were centrifuged at high speed (10,000 × g for 5 min) to pellet cells and bacteria. Supernatants were subsequently concentrated 10-fold (Centricon Ultra centrifugal filter with a 10-kDa cutoff; Millipore, Billerica, MA) for gel electrophoresis and Western immunoblotting using an anti-VLPT-R3-specific polyclonal antibody.

Sequence analysis.The E. chaffeensis VLPT protein was evaluated for potential O-linked glycosylation and phosphorylation with the computational algorithms YinOYang v1.2 (http://www.cbs.dtu.dk/services/YinOYang ) (13) and NetPhos v2.0 (http://www.cbs.dtu.dk/services/NetPhos ) (3). Potential signal sequences or nonclassical secretion was identified with the computational algorithms SignalP 3.0 (http://www.cbs.dtu.dk/services/SignalP ) (2) and SecretomeP 2.0 (http://www.cbs.dtu.dk/services/SecretomeP ) (1) trained on gram-negative bacteria. Nucleic acid and amino acid alignments were performed with MegAlign (Lasergene v5.08; DNAStar). E. chaffeensis VLPT epitopes were examined for homology to other Ehrlichia sp. proteins (including VLPT orthologs) by using the protein-protein basic local alignment search tool (BLAST [http://www.ncbi.nlm.nih.gov/BLAST ]).

E. chaffeensis VLPT protein composition and characteristics.Serine (33 residues; 16.7%), leucine (22 residues; 11.1%), glutamate (20 residues; 10.1%), and aspartate (17 residues; 8.6%) were the most frequently occurring amino acids in the E. chaffeensis VLPT protein, accounting for 46.5% of the entire amino acid content (Fig. 1A). Moreover, in the repeat region of the VLPT protein, the occurrences of these four residues became more frequent, with a composition of 20%, 12.5%, 13.3%, and 10%, respectively, which accounted for 55.8% of the entire repeat region amino acid content. Four cysteine residues associated with disulfide bonds were present in the carboxyl-terminal domain of the protein, but this domain was dominated by tyrosine residues (19.7%). Due to the large proportion of the strongly acidic residues glutamate and aspartate, the VLPT protein was highly acidic (pI 3.8). The N-terminal region (17 amino acids) and the largest domain, the TR region (120 amino acids), were highly acidic (pI 3.2 and 3.8, respectively), and the carboxyl-terminal domain (61 amino acids) was the least acidic domain (pI 4.7).

• Open in new tab
• Download powerpoint

FIG. 1.

(A) Amino acid sequence of VLPT protein showing all domains and locations of four TRs (numbers of amino acids are given in parentheses). (B) Phylogenetic tree showing the relationships of the four E. chaffeensis VLPT repeats. The scale represents the amino acid percent identity.

The TRs of VLPT were nonidentical, but R3 and R2 had the highest amino acid identity (83%) and R4 and R1 had 53% and 49% identity with R3, respectively (Fig. 1B). A BLAST search with amino acid sequences from VLPT-R3 and VLPT-R4 found no homology between the VLPT repeats and other known ehrlichial proteins or proteins from organisms in closely related genera.

Identification of native VLPT protein.Western blotting identified a native protein with a molecular mass of ∼32 kDa (∼6.2 kDa larger than the predicted mass of 25.8 kDa) and five less prominent proteins (22 to 30 kDa) in E. chaffeensis whole-cell lysates and supernatants from E. chaffeensis-infected cells that reacted with monospecific rabbit antiserum against the synthetic VLPT-R3 peptide (Fig. 2A). Furthermore, anti-VLPT-R3 antibody did not react with E. canis whole-cell lysate (Fig. 2A). A protein of similar size (∼32 kDa) in E. chaffeensis whole-cell lysates and supernatants from infected cultures also reacted with anti-E. chaffeensis dog serum (Fig. 2B). Preimmunization rabbit serum or dog serum controls did not recognize E. chaffeensis whole-cell lysates or supernatants (data not shown).

• Open in new tab
• Download powerpoint

FIG. 2.

Identification of native VLPT in E. chaffeensis whole-cell lysates (lane 1), supernatants derived from E. chaffeensis-infected cells (lane 2), and E. canis whole-cell lysates (lane 3) incubated with anti-VLPT-R3 peptide antibody (A) and anti-E. chaffeensis dog serum (B). Preimmunization rabbit serum or dog serum controls did not recognize E. chaffeensis whole-cell lysates or supernatants (data not shown).

Epitope-containing regions of VLPT.To determine the major epitope-containing regions of VLPT, recombinant proteins corresponding to distinct VLPT domains (R4321-C, R32, R1, R2, R3, R4, and C) (Fig. 3) were expressed. All of the recombinant VLPT proteins expressed in pBAD/Thio-TOPO exhibited molecular masses that were substantially larger by SDS-PAGE than those predicted by amino acid sequence, except for VLPT-R3 and VLPT-C (Table 2). VLPT-R32 expressed in an alternative vector, pTriEx6 3C/LIC, with a substantially smaller N-terminal fusion protein (2.4 kDa compared to 13.1 kDa for pBAD/Thio) also exhibited an increased molecular mass (3.7 kDa larger than predicted), but it was smaller than that of VLPT-R32 expressed in pBAD/Thio (5.2 kDa larger than predicted). The increased molecular masses exhibited by recombinant VLPT proteins suggested that the protein was posttranslationally modified; moreover, several predicted glycosylation sites were identified on VLPT by a computational algorithm. However, carbohydrate was not detected on any recombinant E. chaffeensis VLPT polypeptides (data not shown). To further confirm the actual molecular mass of one recombinant VLPT protein, the mass of the two-repeat-containing VLPT-R32 expressed in pTriEx6 3C/LIC was determined by MALDI-TOF. The mass of VLPT-R32 was 9,206 Da, which was slightly smaller than the mass (9,325 Da) predicted by the amino acid sequence (excluding the 2.4-kDa expression tag), demonstrating that posttranslational modifications were not present.

• Open in new tab
• Download powerpoint

FIG. 3.

Schematic of synthetic and recombinant peptides used to map the VLPT epitopes.

By Western immunoblotting, the large TR-containing proteins R4321-C and R32 and individual repeat units R2 and R3 reacted strongly with anti-E. chaffeensis dog serum, but recombinant fragments representing the carboxy-terminal domain (C) and repeat units R1 and R4 were not reactive with anti-E. chaffeensis dog serum (Fig. 4A). The reactivity of VLPT (synthetic polypeptides and corresponding recombinant proteins) with anti-E. chaffeensis dog serum was also examined by ELISA to identify potential conformational epitopes (Fig. 4B). VLPT-N (synthetic), VLPT-C (recombinant), and VLPT-R1 (synthetic and recombinant) polypeptides did not react with anti-E. chaffeensis dog serum. Similar to results by Western immunoblotting, VLPT-R3 and -R2 peptides (synthetic and recombinant) reacted strongly with anti-E. chaffeensis dog serum; however, VLPT-R3 (synthetic) was more immunoreactive than VLPT-R2 (synthetic). Conversely, recombinant VLPT-R4, which was not immunoreactive by Western immunoblotting, reacted (synthetic and recombinant) strongly with anti-E. chaffeensis dog serum by ELISA, indicating that a conformational epitope was present in VLPT-R4 (Fig. 4A and B).

• Open in new tab
• Download powerpoint

FIG. 4.

Immunoreactivities of synthetic and recombinant peptides of E. chaffeensis VLPT with anti-E. chaffeensis dog (no. 2251) serum. (A) SDS-PAGE and total protein staining of purified recombinant peptides (top) and corresponding Western immunoblot probed with anti-E. chaffeensis dog serum (bottom). M, Precision protein standard (Bio-Rad); Ctrl, purified recombinant thioredoxin. (B) Immunoreactivities by ELISA of small recombinant and corresponding synthetic VLPT polypeptides (N [synthetic only], R1, R2, R3, and R4) and large VLPT protein fragments (recombinant only; C, R4321-C, and R32). The OD readings represent the means for three wells (± standard deviations), with the OD of the buffer-only wells subtracted.

Identification of major VLPT immunodeterminants.To further localize the major epitopes of the E. chaffeensis VLPT protein, seven overlapping peptides (designated R3-1 to R3-7) corresponding to the different locations within VLPT-R3 were incubated with anti-E. chaffeensis dog serum (Fig. 3 and 5A). Peptides R3-6 and R3-7 (C-terminal region) were not immunoreactive, but R3-2, R3-3, R3-4, and R3-5, corresponding to the N-terminal region, were found to react similarly and strongly with anti-E. chaffeensis dog serum by ELISA (Fig. 4B), indicating that the N-terminal region (23 amino acids; SDLHGSFSVELFDPFKEAVQLGN) of VLPT-R3 contained a major antibody epitope. Peptide R3-3 (14 amino acids; HGSFSVELFDPFKE) was the smallest peptide that reacted strongly with anti-E. chaffeensis dog serum (Fig. 5A and B). Peptides R3-1 and R3-6, which differed by three (C-terminal) and five (N-terminal) amino acids, respectively, were not reactive (Fig. 5A and B).

• Open in new tab
• Download powerpoint

FIG. 5.

(A) Sequence and orientation of overlapping peptides (seven peptides) representing VLPT-R3. (B) Immunoreactivities of overlapping VLPT-R3 peptides by ELISA with anti-E. chaffeensis dog serum.

To examine and compare the immunodeterminant in VLPT-R4, a 20-amino-acid peptide (HEPSHLELPSLSEEVIQLES) corresponding to R3-5 in VLPT-R3 (Fig. 4A) was not immunoreactive with either dog serum or patient sera (data not shown), indicating that the third epitope of VLPT in R4 was molecularly distinct, consistent with the divergence noted in the amino acid sequences of R4 compared to R3 and R2 (Fig. 1B).

Immunoreactivity of VLPT-R3 peptides with HME patient sera.Three HME patient sera (patient no. 1, 4, and 12) that had detectable E. chaffeensis antibodies by immunofluorescence assay were used to examine the immunoreactivity of VLPT-R4, -R3, and -R2 (synthetic and recombinant) by ELISA (Fig. 6A to C, respectively). Consistent with the immunoreactivity exhibited with anti-E. chaffeensis dog serum, VLPT-R3 and -R2 also exhibited the strongest immunoreactivity with HME patient sera, and two patients (no. 1 and 12) exhibited a strong antibody response to VLPT-R4 (Fig. 6A to C).

• Open in new tab
• Download powerpoint

FIG. 6.

(A to C) Immunoreactivities of synthetic and recombinant E. chaffeensis VLPT repeats (R2, R3, and R4) with three HME patient sera by ELISA. Ctrl, purified recombinant thioredoxin. (D) Synthetic E. chaffeensis VLPT-R3 reacting in ELISA with 14 HME patient sera (lanes 1 to 14), anti-E. chaffeensis dog serum (lane 15), and normal human serum (lane 16). The normal human serum did not recognize other peptides and proteins as well (data not shown).

The immunoreactivities of the three HME patient sera with the seven overlapping synthetic peptides (R3-1 to R3-7) from VLPT-R3 were determined by ELISA (Fig. 6A to C). Peptides R3-4 (17 amino acids) and R3-5 (20 amino acids), which contained similar amino acid sequences (Fig. 5A), reacted strongly and consistently with all HME patient sera tested (Fig. 6A to C). Comparing the overlapping peptides, the minimum peptide sequence critical for this immunodeterminant was 17 amino acids (peptide R3-4). Antibodies from HME patients and a dog experimentally infected with E. chaffeensis reacted similarly to VLPT-R3 (Fig. 4B and 6A to C). However, antibodies in human sera were directed primarily against peptides R3-4 and R3-5 within VLPT-R3 (Fig. 6A to C). Normal human serum did not recognize these peptides and proteins (data not shown).

The reactivity of VLPT-R3 with a larger panel of HME patient sera (from 14 patients) that had detectable E. chaffeensis antibodies was determined. All patient sera reacted with VLPT-R3 (synthetic) (Fig. 6D), indicating that this epitope is consistently recognized by humans, and the reactivity of antibodies in patient sera with this epitope completely correlated with immunofluorescence assay results. Normal human serum did not recognize VLPT-R3 (Fig. 6D, lane 16).

Temporal secretion of E. chaffeensis VLPT.VLPT was detected in supernatants from infected cells as early as 1 day postinfection and increased in quantity through 6 days postinfection (Fig. 7). The VLPT protein was not observed in uninfected DH82 cell culture supernatant.

• Open in new tab
• Download powerpoint

FIG. 7.

Western immunoblot of DH82 cell culture supernatant (0 to 6 days postinfection; lanes 1 to 7, respectively) of E. chaffeensis probed with VLPT-R3 peptide antibody. M, Precision protein standard (Bio-Rad).

Cellular and extracellular localization of VLPT.Several characterized ehrlichial proteins are differentially expressed on dense-core ehrlichiae (gp120, gp36, and gp47). However, like its ortholog gp19 of E. canis, the E. chaffeensis VLPT protein was observed on the membranes of morulae and the surfaces of both reticulate and dense-core ehrlichiae but was also detected on the morula fibrillar matrix by immunoelectron microscopy (Fig. 8A). Anti-VLPT-R3 antibody did not react with uninfected DH82 cells (Fig. 8B).

• Open in new tab
• Download powerpoint

FIG. 8.

Electron photomicrograph of an ultrathin section of E. chaffeensis-infected DH82 cells demonstrating E. chaffeensis VLPT localization in both reticulate and dense-core ehrlichiae (A) and a corresponding ultrathin section containing uninfected DH82 cells (negative control) (B). Cells in both panels were incubated with rabbit VLPT-R3 peptide antibody (1:10,000). Bar = 1 μm.