|Home | About | Journals | Submit | Contact Us | Français|
Rickettsia parkeri, a member of the spotted fever group Rickettsia, is the causative agent of American boutonneuse fever in humans. Despite the increased recognition of human cases, limited information is available regarding the infection of invertebrate and vertebrate hosts for this emerging tick-borne disease. Toward the development of a viable transmission model and to further characterize the pathology associated with R. parkeri infection, inbred mouse strains (A/J, BALB/c, C3H/HeJ, and C3H/HeN) were intravenously and intradermally inoculated with 105 low-passage-number R. parkeri (Portsmouth strain), and infection, gross pathology, and histopathology were scored. Additionally, a quantitative real-time PCR (qPCR) was performed to estimate rickettsial load in heart, lung, spleen, and liver tissues of infected mice at 19 days postinoculation. Of the A/J, BALB/c, and C3H/HeN mice, none displayed universal pathology consistent with sustained infection. Compared to age-matched control mice, the intravenously inoculated C3H/HeJ mice exhibited marked facial edema and marked splenomegaly upon gross examination, while the intradermally inoculated mice developed characteristic eschar-like lesions. The C3H/HeJ mice also exhibited the greatest concentrations of rickettsial DNA from heart, lung, liver, and spleen samples when examined by qPCR. The similarity of the pathology of human disease and sustained infection suggests that the C3H/HeJ strain of mice is a promising candidate for subsequent experiments to examine the tick transmission, dissemination, and pathology of R. parkeri rickettsiosis.
During the past decade there has been a dramatic increase in human infection with tick-borne obligate intracellular bacteria belonging to the genera Anaplasma, Ehrlichia, and Rickettsia, collectively referred to as the tick-borne rickettsial diseases (TBRD) (4). Within the genus Rickettsia, there are numerous pathogenic species grouped together based on molecular and antigenic characteristics as the spotted fever group (SFG) Rickettsia. The clinical signs and pathology associated with these agents in the SFG Rickettsia are variable. For example, infection with Rickettsia rickettsii, the agent of Rocky Mountain spotted fever, is typically associated with a rash, fever, and a high mortality rate (13); in contrast, Rickettsia parkeri rickettsiosis (also referred to as American boutonneuse fever or Tidewater spotted fever) causes a milder disease with eschar, rash, regional lymphadenopathy, and low-grade fever with no reported fatalities (13, 22). The contributing factors responsible for the variation in disease manifestation among SFG species are currently unknown. The recognition of R. parkeri has increased since the first confirmed human case of disease reported in 2004 (15), with more than 20 cases emerging around the southeastern United States (13, 14, 25). The sympatric distribution of pathogenic SFG Rickettsia in the United States (Western Hemisphere) warrants the differentiation of these infections to better comprehend the epidemiological landscape of TBRDs.
Animal models for rickettsial disease have been used to characterize pathology, test vaccine efficacy, and examine transmission parameters. Guinea pigs have been the prototypical animal model for SFG Rickettsia; however, these models frequently use unnatural routes of inoculation, and under certain conditions, guinea pigs can have active infection without clinical signs (12). Inbred mouse strains also serve as models of disease for SFG rickettsioses, but the antigenic diversity of SFG Rickettsia (16) causes variability in the susceptibility of mice to rickettsial pathogens (5). Susceptibility to a particular species of SFG Rickettsia may depend on the strain of mouse used in the study. As an example, R. akari, the agent of rickettsialpox, is uniformly lethal for Mai:(S) and BALB/c mice, while many other strains, including DBA/1J, DBA/2J, AKR/J, and c57BL/6J, are relatively nonsusceptible (19). Compounding factors include differences in mouse strains, which play a role in their susceptibility to SFG rickettsioses. Therefore, while one strain of mice may be susceptible to an individual agent within the SFG, it is possible that the same strain of mice will exhibit a varied susceptibility to different species of SFG Rickettsia. Previous studies highlight these variations in susceptibility. For example, C3H/HeJ mice appear to be the most susceptible strain for R. conorii, a species closely related to R. parkeri (8), while BALB/c has been reported to be the most susceptible strain of mouse for R. australis, and C3H/HeN mice have been found to be the most susceptible strain for R. typhi (5, 6, 9, 24). While the differences in clinical outcomes may be due to the mouse strain or the route of inoculation, clinical distinction between R. rickettsii infections and other typically nonfatal, eschar-associated disease-causing agents, such as R. conorii and R. parkeri, has not been reported in an animal model.
The variable susceptibility of mouse strains to different SFG Rickettsia is also complicated by the manner in which the disease manifests in the model organism. A model that results in a fatal outcome is not needed, and may be inappropriate, for a relatively mild infection such as R. parkeri. While mouse models have proven essential in the characterization of rickettsial pathogenesis and provide insight into the pathobiology of rickettsioses, the variability in host susceptibility requires a multistrain analysis to identify a vertebrate host that approximates most closely the pathology of the infection in human hosts. Toward the development of a model for emerging eschar-associated rickettsial infection, we examined the infection dynamics of R. parkeri in four strains of mice. Initial susceptibility studies were expanded upon using a susceptible strain of mice to recapitulate the pathology of R. parkeri eschar-associated rickettsiosis.
Four mouse strains, A/J, BALB/c, C3H/HeJ, and C3H/HeN, were selected based on previous studies of strain susceptibility to rickettsiae (5, 6, 24). Strains A/J and C3H/HeJ were obtained from the Jackson Laboratory (ME). C3H/HeN mice were obtained from Charles River Laboratories (MA). BALB/c mice were obtained from the Louisiana State University Division of Laboratory Animal Medicine. All mice were 6-week-old males in good body condition. All animal research was performed under the approval of the IACUC at Louisiana State University.
Semipurified rickettsiae were recovered from R. parkeri (Portsmouth strain , passage 4)-infected Vero cells (5 days postinoculation [dpi]) via needle (27 gauge) lysis of host cells and low- and high-speed centrifugation (20, 21). The absolute concentration of rickettsiae was determined utilizing the Live/Dead BacLight bacterial viability kit (Invitrogen), a bacterial counting chamber, and a fluorescence microscope (11). The Rickettsia organisms were resuspended in sucrose-phosphate-glutamic acid buffer (SPG) (7) to a desired inoculation dose of 5.5 × 106 rickettsiae/200 μl. This dose was determined based upon published studies using other Rickettsia spp. in various mouse strains (1, 6) and our pilot studies utilizing various concentrations of R. parkeri. Uninfected Vero cell culture was prepared utilizing the same techniques as that described above with the exception of the bacterial counting. The final lysed Vero cell suspension was diluted with the same volume of SPG as the rickettsial suspension.
Animal infection assays were completed in a series of experiments. Four strains of mice were tested for susceptibility to R. parkeri, followed by the intravenous (i.v.) inoculation of the susceptible strain for a prolonged observation period, and finally the intradermal (i.d.) inoculation of the susceptible strain. Initially, all strains of mice were challenged with R. parkeri Portsmouth by intravenous (tail vein) inoculation. Each strain of mouse was divided into three groups: age-matched, untouched control (n = 1), SPG-injected control (n = 3), and R. parkeri Portsmouth injected (n = 6). Mice in the pilot group were observed for 7 days postinoculation (dpi) for any overt physical signs of disease. At 7 dpi, mice were sacrificed and whole blood was collected via cardiocentesis. Samples of heart, lung, liver, and spleen were snap-frozen in liquid nitrogen for nucleic acid extraction, while additional samples of these same tissues were placed in 10% neutral buffered formalin for histopathology. In the long-term intravenously inoculated group of C3H/HeJ mice, cheek bleeding was performed every other day to every third day for complete blood counts and nucleic acid extraction. The mice were sacrificed at 19 dpi and tissues collected as described for the pilot group. For the intradermal inoculation group, intradermal injections of R. parkeri were administered in the skin at the nape of the neck and in the proximal third of the tail and sacrificed 27 dpi. For these intradermal injections of Rickettsia, the volume of inoculum was reduced to 50 μl but with the same absolute Rickettsia count of 5.5 × 106 rickettsiae.
Blood films were made within 8 h of collection to avoid aging artifacts of the blood and stained with Diff-Quik (Siemens). The smears were examined in a blinded manner by a board-certified veterinary clinical pathologist. The complete blood counts were determined by performing total nucleated cell counts with blood diluted 1:100 in 2% acetic acid and counted in a hemacytometer. Differential leukocyte counts were performed by a veterinary clinical pathologist and were used to derive absolute differential leukocyte concentrations based upon the total nucleated cell count.
Tissues for histopathology were fixed overnight in 10% neutral buffered formalin. All sampled tissues were routinely processed and embedded in paraffin, and 3- to 4-μm sections were cut for hematoxylin and eosin (H&E) staining. The sections were examined in a randomized manner by a board-certified veterinary anatomic pathologist. Selected tissues were examined by immunohistochemistry for evidence of infection with R. parkeri using an immunoalkaline phosphate technique with a polyclonal anti-R. rickettsii antibody, diluted 1/500, as described previously (13).
Frozen tissue samples for genomic DNA (gDNA) extraction were processed using the Qiagen DNeasy blood and tissue kit, with some modifications. Approximately 10 mg of tissue was placed in a 2 ml Safe-Lock microcentrifuge tube (Eppendorf) to which two sterile 5-mm stainless steal beads (Qiagen) were added. Twenty μl of proteinase K (Qiagen) and 180 μl ATL buffer (Qiagen) were then added to each tube, and samples were placed in a TissueLyser (Qiagen) for two cycles of 30 s at 30 hertz. Tubes were centrifuged at 7,500 × g for 5 min and then incubated for approximately 16 h in a 56°C water bath. After incubation, extraction was completed according to the manufacturer's instructions. Extracted DNA was stored at −20°C until it was used for quantitative real-time PCR (qPCR). The Rickettsia primers and species-specific fluorescent-labeled probe for R. parkeri (RaRp 17.181F, 5′-GCATTACTTGGAGCAGTTCT-3′; RaRp 17.289, 5′-CGCCATTCTACGTTACTACC-3′; Rp 17kDa probe, 5′-6-carboxyfluorescein-TGCAGAGCTTACCTCACAGAGAGCTT-3′) and the mouse primers and fluorescent-labeled probe (Cfd1461, CAGTTTCTTGCTGGCTATTGG-3′; Cfd1570, 5′-CCACGTAACCACACCTTCG-3′; Cfd Probe, 5′-5HEX (hexachlorofluorescein)-TGTAGGGCAGGGCCACTAGTGACAT-3′) were designed from the Rickettsia 17-kDa gene sequence and the mouse cfd sequence using Primer3 software. The 17-kDa antigen gene encodes a common rickettsial surface antigen protein, while the mouse cfd gene encodes the complement factor D protein, which is common to most mammals. To quantify a portion of R. parkeri 17-kDa gene in mouse tissues, serial dilutions of a plasmid containing single-copy portions of the R. parkeri 17-kDa and mouse cfd genes were amplified along with the unknown samples. Briefly, qPCR components and the template that included 2× LightCycler 480 probe master mix (Roche); 75 nM each primer; 200 nM each probe; DNase/RNase-free water; and 5 μl of gDNA template (samples), water (negative control), or serial 10-fold dilutions (3.5 × 108 to 3.5 × 103 copies) of pCR4-TOPO-Rp17kDa+MmCfd were premixed in 35-μl volumes in 96-well plates and aliquoted in triplicate 10-μl reaction mixtures on 384-well plates (18). Quantitative PCR then was performed using a LightCycler 480 system II (Roche). The analysis of amplification was conducted with LightCycler 480 software.
Data are expressed as means ± standard errors of the means (SEM), and the significant differences were determined using one-way analysis of variance (ANOVA) with Graphpad Prism and SAS. P values of <0.05 were considered significant.
At the time of necropsy, no gross abnormalities were observed in any of the A/J or BALB/c mice in age-matched control, SPG-injected control, and experimental groups. In contrast, 100% of the intravenously inoculated C3H/HeJ mice exhibited marked facial edema (6/6), while the control mice of this strain did not exhibit any discernible facial edema, and neither did any of the other strains regardless of inoculation group. The edema was defined as marked based on swelling being severe enough to prevent the affected mice from opening their eyes. Upon necropsy, 100% of the infected C3H/HeJ group also displayed marked splenomegaly of 2 to 3 times that of the untouched control, while 100% of the infected C3H/HeN group displayed mild to moderate splenomegaly of 1.5 times that of the untouched control. The degrees of splenomegaly were determined by side-by-side visual comparisons of control and treatment groups. The SPG-injected control groups for both C3H strains did not exhibit any enlargement of the spleen.
The liver, spleen, and heart were examined from all groups. None of the A/J or C3H/HeN mice showed any inflammation within the heart regardless of group (Table 1); however, mild inflammation (defined as 2 or fewer foci of 25 or fewer inflammatory cells) was observed in the heart tissue of SPG-injected C3H/HeJ, R. parkeri-injected C3H/HeJ, and all BALB/c groups. The untouched control C3H/HeJ, SPG-injected C3H/HeJ, untouched control C3H/HeN, R. parkeri-injected C3H/HeN, SPG-injected A/J, R. parkeri-injected A/J, and SPG-injected BALB/c groups all exhibited mild myocardial/epicardial mineralization (defined as 2 or fewer foci of mineralization). The R. parkeri-injected C3H/HeJ and R. parkeri-injected BALB/c groups both showed moderate myocardial/epicardial mineralization (defined as between 2 and 6 foci of mineralization within a section). The untouched control BALB/c mouse exhibited marked epicardial mineralization (more than 6 foci of mineralization). Because of the inflammation and myocardial/epicardial mineralization observed in all groups of the BALB/c mice, these findings were not considered to demonstrate an abnormal process associated with the inoculation with R. parkeri but rather a normal strain variation typical for all of the BALB/c mice in our study. The SPG-injected C3H/HeN and the untouched control A/J groups displayed no myocardial or epicardial mineralization. Myocardial necrosis and degeneration were not observed in any of the mice except for the untouched control BALB/c mouse (mild myocardial degeneration and necrosis defined as two or fewer foci with three or fewer cells displaying necrosis and degeneration). The spleens from all groups were free from obvious inflammation histologically. The untouched control C3H/HeJ mouse, all C3H/HeN mice, SPG-injected group A/J, R. parkeri-injected group A/J, and all BALB/c mice had mild neutrophil loads (defined as less than 20% of the cells consisting of neutrophils and two or fewer focal accumulations of neutrophils) present within the spleen, in contrast to the SPG-injected C3H/HeJ, R. parkeri-injected C3H/HeJ, and untouched control A/J groups, which had moderate neutrophil loads (defined as greater than 20% of the cells consisting of neutrophils or between two and six focal accumulations of neutrophils). Mild inflammation (defined as less than 5 foci of fewer than 25 inflammatory cells) was observed in the livers of the untouched control C3H/HeN mouse, R. parkeri-injected A/J mice, and R. parkeri-injected BALB/c mice, while the R. parkeri-injected C3H/HeN mice exhibited moderate hepatic inflammation (defined as 5 to 10 foci of fewer than 25 inflammatory cells or fewer than 5 foci of more than 25 inflammatory cells).
No overt hematologic abnormalities were observed during the examination of the blood films. The volume of blood collected was insufficient for genomic DNA extraction, therefore qPCR was not performed on these samples.
To determine the bacterial burden within the sampled organs, qPCR for the rickettsial 17-kDa antigen gene was used to quantify R. parkeri in the heart, lung, liver, and spleen at 7 dpi. C3H/HeJ mice had significantly greater numbers of Rickettsia in 75% of tissues examined. Rickettsia was not detected in any of the tissues collected from the A/J strain of mice; however, rickettsial DNA was detected in all tissues analyzed from the R. parkeri-injected C3H/HeJ mice (heart mean, 11,600 copies per 5 μl extracted DNA [cped] [range, 0 to 32,600 cped]; lung mean, 15,400 cped [range, 3,530 to 23,600 cped]; liver mean, 8,200 cped [range, 622 to 24,500 cped]; spleen mean, 1,970 cped [range, 0 to 7,000 cped]). In contrast, rickettsial DNA was detected in only 50% of the tissues of the BALB/c mice (lung mean, 173 cped [range, 0 to 1,630 cped]; liver mean, 62,400 cped [range, 7,010 to 158,000 cped]), and 50% of the tissues of the C3H/HeN mice (heart mean, 1,540 cped [range, 0 to 11,500 cped]; lung mean, 192 cped [range 0 to 2,270 cped]) (Fig. 1). The only tissue in which the rickettsial DNA was not greatest in the C3H/HeJ mice was the liver, where the BALB/c mice contained the highest rickettsial burden.
To evaluate whether the lack of histopathology in the C3H/HeJ mice was due to insufficient time for the progression of disease, mice were tail vein inoculated with R. parkeri and sacrificed at 19 days postinoculation. At 17 dpi, 50% (2/4) of the R. parkeri-inoculated mice presented with necrotic lesions on the tail (Fig. 2A), which measured 3 and 5 mm in diameter. No other gross lesions were observed. The age-matched control mice and the SPG-inoculated control mice exhibited no gross lesions for the duration of the experiment. A moderate to marked mononuclear vasculitis comprising lymphocytes, plasma cells, mast cells, and macrophages was present in the tails of each of the 4 R. parkeri-inoculated mice (Fig. 2C). Moderate to marked fibrinoid vascular necrosis was present as well as mild to moderate perivascular edema. In contrast, no vasculitis or necrosis was observed in either the untouched control or SPG-inoculated control mice. Immunohistochemistry for SFG Rickettsia revealed positive staining in endothelial cells and macrophages within the tail lesions, indicating the presence of intracellular Rickettsia (Fig. 2B).
Blood samples were collected every second to every third day during the 19-day intravenous inoculation study. The total leukocyte concentrations in the R. parkeri-inoculated mice increased 6 dpi and exceeded the values of the uninfected mice. The leukocyte concentration then fell below the values of the uninfected mice at 14 dpi. The neutrophil count in the experimental group was significantly higher than that of the uninfected group at 6 and 10 dpi, whereas lymphocyte concentrations were significantly lower than those of the control group at all time points except 10 dpi (Fig. 3). Monocyte concentrations showed a pattern similar to that of the neutrophils, with a peak concentration at 10 dpi and then returning to control levels. Eosinophil concentrations were significantly lower in the experimental group; however, they did not mirror the changes over time of the other cell types. These changes were consistent among mice within the R. parkeri-inoculated group. When examined by qPCR, very low concentrations of rickettsial DNA were detected in the blood of 50% (2/4) of R. parkeri-inoculated mice, one of which also had a tail lesion, at 19 dpi. Rickettsial DNA was not detected in any of the other blood samples at any time point.
Eight mice were intradermally inoculated with R. parkeri to determine the propensity for the development of eschars in C3H/HeJ mice. Two of the eight intradermally inoculated mice developed crustings at the site of tail inoculation, which were approximately 3 mm in diameter at their maximum and began developing between 8 and 10 dpi. The crust on one mouse resolved by 17 dpi, while the other had decreased in size by 18 dpi, at which point this mouse was sacrificed. A third mouse, while not forming crusts, developed a 4- by 1-mm linear depigmentation at the site of tail inoculation 4 dpi. This lesion remained unchanged for the duration of the experiment. In contrast, at 7 dpi two additional mice developed circumferential swelling of the tails around the site of inoculation, which extended cranially and caudally for approximately two intervertebral spaces. The swelling was resolved in both mice by 10 to 12 dpi. Further, a linear depigmentation of approximately 3 by 1 mm was present on the tail at 20 dpi of one of the mice that had previously had circumferential swelling. All of these mice displayed vasculitis of various severities at the site of tail inoculation. No gross or histopathologic lesions were observed at the site of Rickettsia inoculation on the nape of the neck, and necropsy revealed no gross lesions on any of the internal organs. At 19 dpi, rickettsial DNA was recovered by PCR from multiple mice at the site of inoculation regardless of the manifestation of gross pathology, while the inoculation site in the skin at the nape of the neck was negative for all mice.
Four strains of mice with known susceptibility to various SFG Rickettsia strains were examined for susceptibility to infection, and pathological parameters were assessed. Of the four strains used, R. parkeri DNA was detectable in all tissues examined only from the C3H/HeJ mice, suggesting susceptibility to sustained infection. Additionally, although most models for rickettsioses have been based upon abnormal routes of infection (intravenous and intraperitoneal), the more natural route of intradermal inoculation of the C3H/HeJ mice resulted in several of the mice developing eschar lesions characteristic of human infection, which is a novel finding in mice utilized for SFG Rickettsia modeling. Therefore, although animal models of rickettsial infection appear multivariate, this study provides an experimental model of R. parkeri rickettsiosis which can be utilized to elucidate the mechanisms of infection and pathogenesis.
Based on the temporality of the experiments and the quantities of Rickettsia recovered by qPCR, the resulting values are indicative of rickettsial replication. In the mouse strains which showed no gross pathology, for example, the rickettsial concentrations were much lower than those in the C3H/HeJ mice and were actually undetectable in most of the tissues. While impaired clearance may be an explanation for the greater concentrations of rickettsial DNA recovered from the C3H/HeJ mice, the extended time course of the consequent studies would have been expected to eliminate this factor, as the rickettsial DNA should have degraded much faster in the absence of the homeostatic environment of the living pathogen. To further support this, research enumerating the growth at the lesion must be undertaken, along with studies which consider the effect of tick feeding on rickettsial replication at the site of inoculation.
Animal models have proven vital to expanding our knowledge of the pathogenesis of infectious diseases. Within Rickettsia, murine models have elucidated the importance of TLR4 signaling and subsequent dendritic cell activation to the clearance of pathogenic Rickettsia (9, 10, 17). Mouse models have also revealed that R. prowazekii lesions are independent of humoral immune response (1). These discoveries highlight the importance of an accurate animal model, and the identification of the C3H/HeJ strain as a viable model for R. parkeri rickettsiosis provides a platform for future revelations.
This animal model may also prove beneficial in ways that some other model systems for rickettsial diseases do not. The C3H/HeN strain of mouse is used for modeling R. conorii infection; however, this system results in a fatal disease, which lacks consistency with the human illness (9). In many ways this model remains more than adequate for studying the pathogenesis of this rickettsiosis, but the terminal nature of the infection precludes the use of long-term studies into the progression of and eventual recuperation from disease. In contrast, the C3H/HeJ model of R. parkeri rickettsiosis has not yet proven fatal even in prolonged studies, and it shows promise as a model of progression and resolution of eschars.
TLR4 has been shown to be of great importance in the immune response to rickettsial infection (9, 10, 17). Mice with competent TLR4 signaling are more resistant to rickettsial infection (9). This is primarily because of the stimulation of dendritic cells and subsequent production of gamma interferon (IFN-γ), followed by the expansion of the population of activated natural killer cells (10). This provides a probable explanation for the observed susceptibility of C3H/HeJ mice to R. parkeri. The lack of stimulation of the Th1 response secondary to TLR4 signaling also leads to the expansion of T regulatory cells, which may result in the suppression of proinflammatory immune responses (9). These deficiencies in the immune regulation of the C3H/HeJ mice in response to rickettsial infection warrant further study, including comparisons to the TLR4-competent C3H/HeN strain, to elucidate the mechanisms of rickettsial recognition and clearance by the vertebrate host.
The facial edema, which developed in the intravenously inoculated C3H/HeJ strain, was an unexpected finding. Gross edema is not a common finding in R. parkeri rickettsiosis; however, edema is frequently observed histopathologically in the immediately adjacent tissues surrounding inflamed vessels. The mechanism of this edema is related to the inflammation and necrosis of the infected vessels. The distribution of the edema is intriguing, as the head and neck are areas commonly associated with tick infestation. Although not tested in the current study, it is possible that a distribution pattern of the heavier infection of the vessels of the head and neck increases the likelihood of uptake by and infection of the tick vectors.
Eschars are common to most SFG Rickettsia and form at the site of tick or mite inoculation of Rickettsia as a result of local dermal and epidermal necrosis with marked vasculitis (23). The eschars that developed in this murine model are especially important, as this is a prominent clinical sign of R. parkeri rickettsiosis as well as two of its closest relatives, R. conorii and R. africae (8, 22). The ability to recapitulate these lesions in a murine model provides a basis for comparison to the effects of these closely related SFG rickettsiae. Eschars from patients with Mediterranean spotted fever express high mRNA levels of tumor necrosis factor (TNF), IFN-γ, interleukin-10 (IL-10), RANTES, indoleamine-2,3-dioxygenase, and inducible nitric oxide synthase (3). The lesions from R. parkeri-infected C3H/HeJ mice can be used to determine if a similar pattern of immunological response develops for this rickettsiosis, while another potential source of future research into eschars is evaluating the differences in pathogenesis at the site of tick inoculation between eschar-associated rickettsioses and those that do not typically cause eschars. Although the non-eschar-associated agent of Rocky Mountain spotted fever, R. rickettsii, is the most commonly implicated etiologic agent of rickettsial disease in the United States, there are at least seven eschar-associated rickettsioses encountered by clinicians in the United States, including R. parkeri rickettsiosis, rickettsialpox, 364D rickettsiosis, cat flea-associated rickettsiosis, African tick bite fever, Mediterranean spotted fever, and R. massiliae rickettsiosis (2). It would be very interesting to establish the differences in pathogenesis between Rickettsia associated with these distinct disease manifestations.
The histopathology of the eschars is similar to that reported in humans, with epidermal necrosis and perivascular dermatitis being the prominent features (2, 13, 25). The inflammatory infiltrates associated with this murine model eschar are also analogous to the inflammation described in human patients with R. parkeri rickettsiosis (13), suggesting a similar immunological response, but this will require further investigation to confirm. Rickettsiae were present within the vascular lesions as assessed by immunohistochemistry. As described for patients with R. parkeri rickettsiosis, the SFG Rickettsia strains were found predominantly in the cytoplasm of mononuclear cells (macrophages) and endothelial cells (13). These similarities in necrosis at the site of inoculation, inflammatory pattern, and infiltrate and distribution of Rickettsia support the concept that the infection process is similar for the murine model of infection and the human infection at least at the site of inoculation, which further supports the use of the C3H/HeJ mouse as a model for R. parkeri rickettsiosis. The absence of these eschar-like lesions at the intradermal inoculation site at the nape of the neck suggests that additional factors are necessary to stimulate infection at this site. It is likely that the tick vector plays a significant role in the ability of R. parkeri to infect its vertebrate host, and future studies are necessary to elucidate this process.
The presence of rickettsiae within the hepatic tissue of the BALB/c strain of mice is interesting when the lack of inflammation observed histologically is considered. Possible explanations for this discrepancy include differences in sampling sites utilized for the nucleic acid extraction and histopathology and contamination during sample acquisition, nucleic acid extraction, or qPCR preparation. The latter seem unlikely for various reasons. Contamination during sample acquisition should have resulted in similar findings in other organs from the BALB/c mice, while the nucleic acid extraction and qPCR preparations were repeated with similar results, suggesting that either no contamination occurred or that very similar contaminants were present at two separate time points. Sampling error appears to be the most probable cause for the discrepancy and is a commonly encountered issue in diagnostic settings. It is probable that rickettsial lesions were present in different foci of the hepatic tissue, including that sampled for nucleic acid extraction, but the same was not true for the sections evaluated by histopathology.
With the increase in TBRD recognized in humans, further study into the transmission of the responsible organisms and the pathogenesis of the diseases has become of paramount importance. R. parkeri serves as an elegant example of this, as the first confirmed case of human disease was reported less than a decade ago. The relatively recent recognition of this organism as a pathogen has deeper implications, considering research into its biology remains in its infancy compared to that for other rickettsiae. To elevate the knowledge base regarding R. parkeri and its disease, a commonly available inbred mouse strain with strong potential for use as a model for research into the pathogenesis of R. parkeri rickettsiosis has been identified in the current study. The C3H/HeJ inbred mouse strain was found to be susceptible to infection with R. parkeri, while typical gross and histopathological features of human disease developed at the site of intradermal inoculation, and also in concert with findings in humans with R. parkeri rickettsiosis, the infection has thus far proven nonfatal. This work identifies a new tool for investigators in subsequent studies of the pathogenesis of R. parkeri rickettsiosis, which can now include the detailed characterization of the immune response to the bacteria along with incorporating the effects of the tick vector on the establishment and maintenance of infection.
We thank Michael T. Kearney for his assistance with statistical analysis and members of the Macaluso laboratory for their technical assistance.
This research was supported by the National Institutes of Health (grant AI077784).
This work was part of B. J. Grasperge's doctoral dissertation.
Published ahead of print 5 March 2012