During this study in 2002, we screened almost all patients who were suspected clinically to have a tick-borne rickettsiosis and who were hospitalized in the infectious diseases department of Khabarovsk Municipal Hospital No.10. Five to 20 days before onset of the disease, most patients reported tick bites or tick exposure. Some had typical eschars on the skin without a history of a tick bite. The illness was acute, febrile, and usually involved a macular or maculopapular rash. All patients were successfully treated with a 14-day course of doxycycline. Patients with clinically evident features of tick-borne encephalitis (neurologic signs) or Lyme disease (erythema migrans) were excluded from the study. In total, samples from 65 patients were investigated, 64 blood samples and 17 skin biopsy results. Paired sera from most of the patients were tested with conventional serologic methods. The first serum sample was taken immediately after admission to the hospital, usually 1–9 days after the onset of the disease. The second serum sample was taken 4–18 days later.

Buffy coats were separated from fresh blood specimens immediately after collection at patient’s admission. DNA was extracted from the buffy coat and skin biopsies samples with QIAmp DNA Blood Mini Kit and DNeasy Tissue Kit (Qiagen, Tokyo, Japan) according to manufacturer’s instructions.

Oligonucleotide primers used in this study and annealing temperatures are shown in Table 1. We used the nested polymerase chain reaction (PCR) method for primary detection of the rickettsial DNA in human samples. The gltA (citrate synthase) gene was chosen as the target for amplification because of its genus specificity and conservativeness. The primer CS1d (16), which is used traditionally to amplify the full-length gltA gene, is complementary to the ultimate 5′-end of the open reading frame for this gene of R. prowazekii. Since we expected to find DNA of the spotted fever group rickettsiae, we modified the primer to be completely homologous to the corresponding portion of the gene of R. conorii (Table 1). R. conorii has been chosen as a template to design primer CS2d as the only spotted fever group rickettsia having a completely sequenced genome at the time of investigation. Comparison of these two primers indicated that the newly designed primer (CS2d) was at least 100 times more sensitive than CS1d in the amplification of the serially diluted DNA of R. montanensis, which was used as a positive control (data not shown). Primers CS2d and CSendR amplify the full-length of the gltA gene. Primers CS877f and CS1258r were used in the nested PCR assay. To amplify the full-length of the gltA gene, we used PCR followed by hemi-nested PCR using the product of the first PCR as a template. Two conservative regions of the outer membrane protein A (ompA) gene were amplified (base positions 91–680 and 3,608–6,789 with respect to the sequence published for R. rickettsii) by using previously described primers (17). Initial amplification of the 3,182-bp longer region was achieved with primers 190-3588f and 190-5044r and with primers 190-4859f and 190-6808r. Products of this reaction were used later in the nested and hemi-nested PCR reactions with corresponding primers. The ompB gene was amplified from clinical samples and DNA by using standard primers (18), except for the ultimate forward primer 120-M59, which was found to produce nonspecific amplicons with the human DNA in clinical samples. Instead, for clinical specimens, we used a newly designed primer, 120-M36. As with the ompA, the DNA from clinical samples was amplified in two steps. In the first step, we amplified two halves of the gene, with a small overlapping region with primers 120-M36 with 120-2988 and 120-2788 with 120-4879. The products of these reactions were used as templates for each of the specific nested or heminested reactions for seven regions. All primers were purchased from Eurogentec (4102 Seraing, Belgium).

Amplification has been carried out by routine methods in a final volume of 50 μL with appropriate negative controls (10,17). In all reactions, as a positive control, we used a mixture of DNA of the following microorganisms: Anaplasma phagocytophilum, Neorickettsia sennetsu, Wolbachia pipientis, Francisella tularensis, Bartonella henselae, Borrelia garinii, Coxiella burnetii, and R. montanensis. We found that the cocktail of DNA of tick-borne or phylogenetically close to tick-borne bacteria is convenient to use in PCR with potentially polyinfected human samples.

We checked all samples for other possible tick-borne bacterial and pathogens with the following primers under conditions suggested in published references: SL primers for ospA gene of Borrelia (19); BhCS.781p and BhCS.1137n primers for Bartonella citrate synthase (20); HE1, HE3, and HE4 primers for 16S rDNA of Ehrlichia chaffeensis (21); GE3A, GE10r, GE9f, and GE2 primers for 16S rRNA gene of A. phagocytophilum (22); and P3708 and p4257 primers for the gene encoding p44 protein of A. phagocytophilum (23) [data not shown]. No positive results have been obtained among patients described here.

The PCR products were purified for DNA sequencing with the QIAquick PCR purification kit (Qiagen) and then directly sequenced by using PCR primers. Sequencing reactions were carried out with a D-rhodamine terminator cycle DNA sequencing kit (Applied Biosystems, Foster City, CA) according to the manufacturer’s instructions. Sequencing reaction products were resolved by electrophoresis with an ABI Prism 377 sequencer (Applied Biosystems). The results obtained were processed into sequence data with AutoAssembler software (Applied Biosystems). The sequences of the gltA, both regions of the ompA gene, and the ompB genes were aligned by using the software Genetix-Win 5.1 (Software Development Co., Ltd., Japan). Sequences in the ompB genes of R. heilongjiangensis and R. hulinensis were not available in the GenBank database so we amplified and sequenced the gene. The sequences used for comparison were obtained from the GenBank database, aligned, and then corrected manually to preserve codon alignment and conserved motifs. Sites with ambiguous alignments were removed before phylogenetic analysis. The phylogenetic tree was calculated by using neighbor-joining method with MEGA2 Version 2.1 software (available from: http://megasoftware.net). Internal node support was verified by using the bootstrap method with 100 replicates.

Two serologic tests were performed in Khabarovsk Plague Control Station: immunofluorescence studies with a combined antigen consisted of two local strains of B. garinii and one local strain of B. afzelii, and an enzyme-linked immunosorbent assay detected antibodies against tick-borne encephalitis virus (24). For further investigations, sera were dried on blotting papers as described previously (25) and transported to Marseilles, where microimmunofluorescence testing (26) was performed by using in-house prepared antigens of R. heilongjiangensis (strain 054, ATCC VR-1524), R. hulinensis (strain HL-93, ATCC VR-1527), R. sibirica (strain 246, ATCC VR-151), R. conorii (strain Moroccan ATCC VR-141), C. burnetii (strain Nine Mile, ATCC VR-615), Orientia tsutsugamushi (strains Gilliam, Karp, Kato, and Kawazaki), E. chaffeensis (strain Arkansas), A. phagocytophilum (strain Webster), Bartonella henselae (Houston-1, ATCC-49882), and B. quintana (strain Oklahoma). Antigens were applied by pen point to 18-well microscope slides, dried for 30 min, and fixed. Appropriate positive- and negative-control serum samples were tested on each slide together with twofold dilutions of patients’ serum samples made in 3% nonfat dry milk in phosphate-buffered saline (PBS). Slides were incubated in a moist chamber for 30 min at 37°C, washed twice in PBS and once in distilled water (10 min each); reactive antibodies showed fluorescein isothiocyanate–conjugated goat anti-human γ chain and μ-chain immunoglobulins (BioMérieux, Marcy l’Etoile, France). After the conjugate was added, slides were incubated for 30 min at 37°C, washed in two PBS for 10 min and for 5 min in distilled water, and mounted in buffered glycerol. Endpoints for each antigen were the lowest concentrations of serum that definitely conferred fluorescence on bacteria.

Nucleotide sequences obtained during this study were deposited in GenBank under the following numbers: AY260451 for ompB gene of R. heilongjiangensis, strain 054; AY260452 for ompB gene for R. hulinensis; AY280712 for ompB gene; AY280711 for previously tandemly repeated region portion of ompA gene; AY280710 for another portion; and AY280709 for gltA gene of Rickettsia spp. found in this study.