Prevalence of Rickettsia species antibodies and Rickettsia species DNA in the blood of cats with and without fever

Abstract

Rickettsia species antibodies have been detected in some cats but it is unknown whether infected cats develop clinical signs. The prevalence of Rickettsia species deoxyribonucleic acid (DNA) in blood from clinically ill cats has not been determined. The objective of this study was to determine if cats with fever (body temperature ≥102.5°F [39.2°C]) were more likely to have evidence of rickettsial infection than healthy, age-matched, control cats with a body temperature<102.5°F. Rickettsia species polymerase chain reaction (PCR) assays were performed to detect rickettsial DNA extracted from blood (71 paired samples), indirect immunofluorescence assays (IFA) were performed to detect serum antibodies against Rickettsia felis (90 paired samples) and Rickettsia rickettsii (91 paired samples), and the results between pairs were compared. All samples were negative for Rickettsia species DNA. More cats with fever were seropositive for R felis or R rickettsii than control cats, but results were not statistically significant. Results of this pilot study failed to show an association between Rickettsia species DNA or Rickettsia species antibodies and fever.

Fever is a common finding in sick cats. Infection, immune-mediated diseases, and neoplasia are common causes of fever, with infectious disease thought to be the most common cause in cats. ¹ A large number of viral, bacterial, rickettsial, and protozoal organisms have been associated with fever in cats. For some agents, the diagnosis may be easy to achieve because there are concurrent clinical syndromes that direct the diagnostic workup or the organism can be identified with widely available tests. Examples include feline leukemia virus, feline immunodeficiency virus, panleukopenia virus, and respiratory viruses.

Recently, fever in people has been attributed to infection with ‘stealth’ organisms. ² These agents evade the immune system, cause subtle clinical signs, and often are not easily detected by traditional diagnostic methods. Examples in people that can also be associated with disease in cats include Ehrlichia species, Anaplasma phagocytophilum, and Bartonella species. ^3–5 Advances in serological and molecular diagnostic tests have aided in the diagnosis of these organisms, however, the tests are generally specific and so a particular agent must be suspected to trigger the veterinarian to choose the diagnostic test of choice. We believe there are previously unrecognized organisms of this type that infect cats and induce fever. Rickettsia species are possible causes but have not be studied extensively in cats.

Rickettsia felis deoxyribonucleic acid (DNA) has been amplified from clinically ill humans in Mexico, France, Brazil, Germany, Texas, California, and other places. ^6–9 Fever was one manifestation of illness in these people. The cat flea, Ctenocephalides felis, has been identified as a host and biological vector of R felis. ^7,10,11 Several infected humans have a history of flea exposure, suggesting that C felis may be capable of transmitting the organism to humans. ¹⁰ Cats are commonly infested with C felis, and R felis DNA was amplified from 62.4% of fleas collected from cats in the United States. ¹² Antibodies against R felis, R. rickettsii, and R. typhi have been detected in cats in the United States, indicating that cats are exposed to rickettsial organisms. ^8,13,14 Findings of these previous studies led to our hypothesis that Rickettsia species may be associated with fever in cats.

In a previous study, R felis DNA was not amplified from the blood of cats known to be infested with R felis infected fleas. ¹² However, those cats were generally healthy other than their flea infestations. To the authors' knowledge, the prevalence of Rickettsia species DNA or R felis antibodies has not been reported in a population of clinically ill cats. The objective of this pilot study was to determine the prevalence of Rickettsia species DNA in blood, as well as the prevalence of R felis and R rickettsii antibodies in serum, of cats with and without fever.

Materials and Methods

Cats

Samples selected for this study were initially collected for a study of Bartonella species IgG (immunoglobulin G) and DNA in the blood of cats with and without fever. ¹⁵ Samples were requested from participating veterinarians by announcements made at continuing education meetings throughout the United States and via email and phone consultations with one of the authors (MRL). The test group included cats with an elevated body temperature of ≥102.5°F (39.2°C), for which a cause could not be determined on physical examination, and for which the veterinarian felt was related to fever, rather than stress or environmental factors. The control group consisted of cats with a body temperature<102.5°F (39.2°C) and no detectable medical problems, which were examined in the same clinic on the same day. Information requested for all cats included state of origin and age. States were classified as low flea-risk (Alaska, Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah, Wyoming) or high flea-risk (all other states) based on a previous study. ¹⁶ Upon receipt at Colorado State University, DNA was extracted from blood stored in ethylenediaminetetraacetic acid (EDTA) and assayed for Bartonella species DNA by polymerase chain reaction (PCR) and sera were assayed for Bartonella species IgG by enzyme-linked immunosorbent assay (ELISA). The DNA extracts and sera were saved at −20°C or −70°C until used in this study. ¹⁵

Immunofluorescence assays (IFA) slides

Slides coated with R felis infected cells, for use with indirect IFA, were obtained from The University of São Paulo, Department of Veterinary Medicine and Animal Health. R felis was cultivated in C6/36 cells as previously described, except that infected cells were incubated at 28°C. Cells were harvested when nearly 100% of the cells were infected. ¹⁷ Infected cells were centrifuged at 4000×g for 10 min. The pellet was washed in 0.1 M phosphate buffered saline (PBS, pH 7.4), centrifuged again, and re-suspended in PBS containing 1% bovine calf serum and 0.1% sodium azide. Ten microliters of R felis infected cell suspension were applied on to each well of 12-well antigen slides (IFA slides, Precision Lab Products, LLC Middleton, WI). Cell suspensions on the slides were air-dried, then fixed in acetone for 10 min. Slides were stored at–80°C until shipped overnight with cold packs to Colorado State University. Slides coated with R rickettsii-infected cells were obtained from a commercial supplier (Rickettsia rickettsii IFA slides, Prototek International, St Paul, MN).

IFA assays

The negative control used in all assays was sera collected and pooled from laboratory-reared, specific pathogen-free cats housed without potential for exposure to arthropods For the R felis assay, known positive serum samples from an infected cat and an infected dog were used as positive controls. The control cat serum was determined to be positive based on a reaction with the R felis antigen at a titer of 1:128, as well as reactions to R rickettsii and Rickettsia parkeri at titers>1:128. Canine serum that was positive for R rickettsii antibodies was used as a positive control in the R rickettsii assay. IFA slides control sera were stored at −20°C until used.

Slides and sera were removed from the freezer and allowed to come to room temperature. Suspect samples and the negative control sample were diluted 1:20 with 0.075% Tween in PBS-Tw+1% bovine serum albumin (BSA) solution. The feline R felis positive control sample and the canine R rickettsii positive control sample were diluted at 1:10 in 0.075% PBS-Tw+1% BSA solution. Aliquots (25 μl) of the appropriate positive and negative controls and suspect sera were used to coat individual IFA wells with care taken to control cross-contamination. Slides were incubated in a pre-warmed humidity chamber at 37°C for 30 min and then washed briefly using PBS-Tw. Slides were washed on a rocker in PBS for 5 min followed by an additional 5 min in deionized water (dH₂O).

Fluorescein-conjugated goat anti-cat heavy chain IgG (Kirkegard and Perry Laboratories, Gaithersburg, MD) and fluorescein-conjugated goat anti-dog heavy chain IgG (Kirkegard and Perry Laboratories) (R rickettsii positive control wells only) were diluted at 1:40 in 0.25% Evan's blue and 25 μl was pipetted on to appropriate wells of the IFA slides for the respective Rickettsia species. Slides were incubated at 37°C for 30 min in the humidity chamber and then washed again as described previously. After the final wash, two drops of mounting fluid (nine parts glycerol to one part PBS) were applied, and a cover slip was placed on each slide.

Slides were read by two examiners independently on a fluorescent microscope. Discordant samples were assayed a second time, and if still discordant, were read by a third examiner. Discordant samples were considered positive if, after the second assay, at least two of the three examiners reported it as positive. All samples that were positive at 1:20 were evaluated serially in two-fold dilutions to the end point. The reported titer was the last dilution to give positive fluorescence.

Rickettsia species PCR assay

The DNA extracts were thawed at room temperature and Rickettsia species PCR assays were performed using previously described protocols. ^12,18 Oligonucleotide primers for the citrate synthase gene (gltA) and outer membrane protein B (ompB) gene were utilized. The sensitivities of the gltA and ompB PCR assays were 1.10 fg and 0.110 fg, respectively. On each run for both PCR assays, a negative control, consisting of PCR reagents and PCR water as template (no DNA template added), and positive controls consisting of purified R felis DNA and R rickettsii DNA, were included.

Statistical analyses

To determine the association between R felis or R rickettsii antibodies and fever, the distribution of the paired sample results were evaluated by logistic regression to determine odds ratios and 95% confidence intervals. Significance was defined as P<0.05.

Results

Cases were included if sufficient sample volume existed for performance of individual assays (Rickettsia species PCR assays, R felis IFA, and R rickettsii IFA) on the sample from both the febrile cat and the paired control cat. There were 71, 90, and 91 paired samples available for testing by Rickettsia species PCR assays, R felis IFA, and R rickettsii IFA, respectively. Sample pairs came from the following states; Alabama (one), Arkansas (one), California (four), Colorado (25), Connecticut (one), Florida (one), Georgia (seven), Illinois (three), Indiana (two), Kansas (three), Louisiana (two), Massachusetts (five), Michigan (two), Minnesota (four), Mississippi (two), Missouri (two), Nebraska (one), New Jersey (six), New Hampshire (two), North Carolina (three), Ohio (one), Pennsylvania (one), South Carolina (two), Tennessee (two), Texas (one), Utah (one), Virginia (one), and Wisconsin (one). The state of origin was not recorded for three sample pairs. Of the represented states, only Colorado and Utah were considered low flea-risk, so at least 65 of 91 sample pairs (71.4%) were from high flea-risk states.

It was previously determined that many of the cats assessed in this study had Bartonella species IgG in serum or Bartonella species DNA amplified from blood, suggesting that many of the cats had been exposed to fleas. ¹⁵ However, Rickettsii species DNA was not amplified from any of the blood samples by use of either PCR assay.

Antibodies against R felis were detected in serum from five cats with fever (5.6%: titers; 1:20, 1:20, 1:40, 1:40, 1:80) and two control cats (2.2%: titers; 1:20, 1:20), but were not concurrently detected in the serum of any paired sample. A greater number of cats with fever were R felis seropositive compared to control cats, but results were not statistically significant (odds ratio=2.5; 95% CI=0.485, 12.886; P=0.2734). The R felis seropositive cats were from Georgia (two), North Carolina (one), California (one), Ohio (one), Nebraska (one), and New Jersey (one). These states are considered to have a high flea-risk. Of the R felis seropositive cats, three cats with fever and both control cats were positive for Bartonella species antibodies, and one cat with fever and one control cat were PCR assay positive for Bartonella henselae DNA.

Antibodies against R rickettsii were detected in serum from six cats with fever (6.6%: titers; 1:20, 1:20, 1:40, 1:40, 1:40, 1:80) and two control cats (2.2%: titers; 1:20, 1:20), but were not concurrently detected in the serum of any paired sample. A greater number of cats with fever were R rickettsii seropositive compared to control cats, but results were not statistically significant (odds ratio=3.0, 95% CI=0.606, 14.86, P=0.1785). The R rickettsii seropositive cats were from Georgia (two), Colorado (two), Arkansas (one), California (one), Connecticut (one), and New Jersey (one). Of the R rickettsii seropositive cats, one cat with fever and one control cat were positive for Bartonella species antibodies. Of the seven R rickettsii seropositive cats with samples available for Bartonella PCR assay, one cat with fever was PCR assay positive for B henselae DNA.

One control cat (R felis titer=1:20; R rickettsii titer=1:20) and one cat with fever (R felis titer=1:40; R rickettsii titer=1:40) were antibody positive for both organisms.

Discussion

The cases included in this study represented 28 states in the United States, 26 of which are considered to have a high flea-risk. Bartonella species IgG antibodies and Bartonella species DNA were detected in>30% of the samples, indicating that many of cats included in this study were exposed to fleas. However, despite a previously reported prevalence rate of 62.4% for R felis infection in C felis collected from cats in the United States, the results of this study did not show an association between fever in cats and Rickettsia species DNA in blood or serologic evidence of exposure to R felis. ¹² These findings indicate that R felis is not a common cause of fever in cats or merely reflects the sample population tested.

The prevalence of R felis antibodies in cats described here (seven of 180 cats tested; 3.9%) is lower than the seroprevalence reported in two of three previous studies performed in the United States. In one study, the authors refer to unpublished data that showed a R felis seroprevalence rate of 8% in febrile cats in the northeastern United States. ⁸ In another study, the R felis seroprevalence rate amongst pet, shelter, and feral cats in California and Wisconsin was 11.1%. ¹³ The third study evaluated for associations between R typhi antibodies and clinical disease in cats in North Carolina; the authors detected Rickettsia species antibodies in nine of 436 cats (2.1%) that were likely a species other than R typhi and so suspected to be R felis. ¹⁴

It is possible that studies with higher prevalence rates are related to higher flea burdens in the study populations, as the northeastern United States, California, and Wisconsin are all considered to have a high flea-risk. ¹³ In addition, in the study with 11.1% R felis seroprevalence rate, 150 of 170 cats (88.2%) were shelter cats or feral cats. These cats may be more likely to have flea infestations, and may be more susceptible to infection due to concurrent diseases, stress, or sub-optimal housing when compared to client-owned cats. In the current study, all of the samples were from client-owned cats and 26 pairs were from Colorado or Utah which may have contributed to the lower number of R felis seropositive cats reported in this study.

While a statistically significant association between fever in cats and antibodies against R felis was not found in this study, numerical differences were noted between groups. When discordant samples for R felis antibodies within pairs were examined, there were more discordant pairs which included a seropositive cat with fever and a seronegative cat without fever than pairs which included a seronegative cat with fever and a seropositive cat without fever. In addition, the highest titers reported for R felis antibodies in this study (1:80 for both) were detected in the serum of febrile cats. These results suggest that R felis may be associated with fever in some cats and that a larger study should be performed.

Similar to the R felis results, there was no statistically significant association between fever in cats and presence of R rickettsii antibodies in serum, and the prevalence of R rickettsii antibodies was low (eight of 182 cats; 4.4%). These findings indicate that R rickettsii is not a common cause of fever in cats or may relate to the population of cats tested. None of the cats in this study were known to be exposed to known R rickettsii vectors and the majority of states of origin are not considered endemic for R rickettsii infection. An R rickettsii seroprevalence rate of 17.2% was reported in the study of pet, shelter, and feral cats in California and Wisconsin. ¹³ As discussed for R felis, inclusion of a large number of shelter and feral cats may have contributed to the higher R rickettsii seroprevalence rate reported in that study. A future study should be performed comparing R rickettsii antibody and Rickettsia species PCR assay results between cats with and without clinical illness in a state that is endemic for R rickettsii, such as North Carolina or Georgia.

Failure to amplify Rickettsia species DNA from the blood of these cats may simply reflect lack of exposure as discussed for R felis and R rickettsii serology results. However, in a previous study, when R felis infected fleas were allowed to feed on cats, 5/16 cats were transiently PCR positive for R felis DNA in blood. ¹⁹ Thus, it is possible that the negative PCR test results in the cats of the study described here merely related to the timing of sample collection; the microbial DNA may have been cleared from the blood by a rapid and effective immune response by the time the cats were tested. The finding that some cats were R felis and R rickettsii seropositive but PCR negative supports this hypothesis. In future studies, it would be optimal to collect samples from clinically ill cats more than one time. Lastly, it is also possible that the organisms were present in some cats but sequestered in other tissues like the skin or spleen.

In summary, we failed to find an association between R felis and R rickettsii and fever in cats of this study but sample sizes and distribution of cases may have influenced the results. Larger study populations of sick and healthy cats are needed to further assess for associations between Rickettsia species and clinical diseases in cats.