Analysis of Seroreactivity against Cell Culture–Derived Bartonella spp. Antigens in Dogs
- Presented as a poster at the 7th International Conference on Bartonella as Animal and Human Pathogens, April 25–28, 2012, Raleigh, NC.
Little is known about the specificity of Bartonella spp. immunofluorescent antibody (IFA) assays in dogs. Bacteremia in sick dogs most often has been associated with Bartonella henselae (Bh), Bartonella vinsonii subspecies berkhoffii (Bvb), and Bartonella koehlerae (Bk). Clarification of the diagnostic utility of IFA serology when testing against these organisms is needed.
To evaluate the specificity of Bartonella IFA assays utilizing 6 cell culture–grown antigen preparations.
Archived sera from SPF dogs (n = 29) and from dogs experimentally infected with Bvb (n = 10) and Bh (n = 3).
Antibodies (Abs) to Bvb genotypes I, II, and III, Bh serotype I, strains H-1 and SA2, and to Bk were determined by IFA testing.
Serum from naïve SPF dogs shown to be negative for Bartonella bacteremia did not react with any of the 6 Bartonella antigens by IFA testing. Dogs experimentally infected with Bvb genotype I developed Abs against homologous antigens, with no cross-reactivity to heterologous Bvb genotypes, Bh H-1, SA2 strains, or to Bk. Dogs experimentally infected with Bh serotype I developed Abs against Bh H-1, but not to Bh SA2 strain with no cross-reactive Abs to Bvb genotypes I–III or to Bk.
Conclusions and Clinical Importance
Bartonella spp. Ab responses during acute experimental infections are species and type specific.
Bartonella vinsonii subspecies berkhoffii
- Bartonella henselae strains
Houston-1= H-1, San Antonio-2 = SA2, and California-1 = CAL-1
Bartonella alpha-Proteobacteria growth medium
immunofluorescent antibody assay
intergenic transcribed spacer region
enzyme-linked immunosorbent assay
polymerase chain reaction
- NCSU-CVM VBDDL
North Carolina State University-College of Veterinary Medicine Vector Borne Disease Diagnostic Laboratory
specific pathogen free purpose-bred dogs
Bartonella spp. constitute an environmentally diverse genus of vector-transmitted bacteria that appear to be the cause of an expanding spectrum of disease manifestations, and as such are considered important emerging pathogens in dogs worldwide.[1-3] Using a sensitive detection platform that combines Bartonella alpha-Proteobacteria growth medium (BAPGM) enrichment culture with polymerase chain reaction (PCR), Bh, Bvb, and Bk represent the most frequent species infecting dogs in North America. When the BAPGM enrichment blood culture/PCR was used by Bai and colleagues to test dogs from Thailand, infection with additional rodent reservoir-adapted Bartonella spp. was identified. Four Bvb genotypes have been implicated in bacteremic infections in cats, dogs, horses, humans, and a red wolf (Canis lupus rufus).[5-7] Two serotypes of Bh including strains Houston-1 (serotype I), San Antonio-2 (serotype I), and California-1 (serotype II) have been detected in cats, dogs, and humans. Bvb, Bh, and Bk also have been implicated in canine and human cases of endocarditis.
Serosurveys involving convenience samples of canine sera using laboratory strains of Bartonella, primarily Bh (H-1), Bvb (genotype I), and Bartonella clarridgeae (Bcl), have identified a range of seroprevalence rates from 1 to 35% in dogs.[10-13] Among 1872 healthy working dogs tested against Bvb by ELISA, 162 (8.7%) were seroreactive. A study that compared healthy (n = 99) and sick (n = 301) dogs by IFA using Bh and Bvb antigens showed that 10.1% of healthy dogs were seroreactive against Bh, whereas only 1% reacted with Bvb. The sick dog group had seroprevalence rates of 27.2 and 4.7% to Bh and Bvb, respectively. Of 1920 dogs tested in a hospital setting and presumed sick, 69 (3.6%) were seroreactive to Bvb by IFA, with higher prevalence associated with outdoor lifestyle conditions. A third study, again using ELISA assays with various Bartonella spp. antigens and a population of 3,417 sick dogs, demonstrated an overall Bartonella spp. seroreactivity in 102 (3.0%), with breed and outdoor lifestyle associated with seroreactivity. The antigens used indicated possible differences in specificity because 36/102 (35.3%) were reactive against Bh, 34/102 (33.3%) against Bcl, and only 2/102 (2%) against Bvb. Cross-reactivity or exposure to multiple species was suggested by the 29.4% that showed reactivity to >1 antigen in that study. Without the aid of molecular or culture methods, definitive identification of more than 1 etiologic agent is difficult to achieve.
The purpose of this study was to evaluate Bartonella IFA serologic specificity by testing serum obtained from naïve SPF dogs and from dogs experimentally infected with Bvb or Bh against an expanded panel of Bartonella spp. antigens grown in tissue culture. In 1 previous study that tested dog sera by Western immunoblotting, no differences in seroreactivity were documented when agar culture–grown Bvb was compared with Bvb grown in DH82 cells. The decision to grow diagnostic antigens in cell lines or to use agar-grown bacteria has been based largely upon laboratory preference or historical use patterns. For obligate intracellular pathogens, such as Rickettsia and Ehrlichia spp., the use of cell lines is mandatory to achieve organism growth while the cellular background provides an added benefit of providing a structural context for bacterial visualization. A number of mammalian cell lines (Vero, Hep-2, HeLa, and DH82) have been utilized successfully for the cultivation of Bartonella antigens for IFA testing.[15, 16] In 2 previous studies, no serologic cross-reactivity was observed between Bvb genotype I and Bh when serum from dogs experimentally infected with Bvb or Bh was used in IFA tests utilizing tissue culture–grown antigens.[11, 14] However, as a rapidly expanding number of Bartonella spp. have been found to infect dogs,[1-3] additional studies to characterize canine humoral immune responses are warranted to better understand potential serodiagnostic benefits and limitations.
Materials and Methods
Sources of Sera
Serum samples used in this study had been stored at −80°C. Dogs that were experimentally infected were in studies approved by the appropriate Institutional Animal Care and Use Committees (North Carolina State University or Colorado State University). Overall, serum samples came from the following sources:
Archived serum samples from these 29 dogs were tested as a component of the present study to represent a set of negative control sera. These included sera from 13 SPF dogs that were Bartonella seronegative (<1 : 16) and agar culture negative before inoculation in 2 experimental Bartonella infection studies conducted at NCSU between 1995 and 1998[3, 17] and sera from 16 SPF retired breeder beagles, also Bartonella seronegative as well as PCR and BAPGM enrichment culture negative before use in experimental Bartonella exposure studies conducted at CSU (Lappin, MR, unpublished data).
Bvb Experimental Infection
Archived serum samples collected from 10 dogs in the 1995 and 1998 experimental infection studies were chosen for testing against 6 Bartonella antigens. These 10 serum samples from purpose-bred dogs had been collected on days 28 or 34 after IV inoculation with 108–109 colony-forming units of agar-grown Bvb genotype I.[3, 17] Based on blood agar culture using the lysis centrifugation method or Bartonella spp. 16-23S intergenic transcribed spacer region (ITS) PCR of DNA extracted from whole blood during the previous studies, bacteremia was confirmed in dogs selected from the 1995 and 1998 studies with evidence of persistence of infection of at least 1-month duration as shown by PCR or culture for all 10 dogs. Infection status also was assessed in these 2 previous experimental studies[3, 17] by documentation of Bvb seroconversion in all dogs within the first 2 weeks after inoculation with reciprocal IgG antibody titers against Bvb I antigen rising from <16 to reciprocal titers ranging from 64 to 4,096.
Bh Experimental Infection Study
Cats inoculated with the CSU-1 strain of B. henselae (serotype I) were infested with Ctenocephalides felis fleas that then were allowed to migrate through mesh enclosure barriers to 2 adjacent cages housing dogs. Sera from 3 of the 16 SPF dogs in this study were selected based upon documentation of Bh DNA in peripheral blood using Bartonella ITS PCR of BAPGM-enriched blood cultures. Serum samples from these dogs were used for further testing against 6 Bartonella antigens in the present study. Evidence of infection also had been assessed during the CSU study (unpublished data) by documentation of reciprocal IgG antibody titers to Bh H1 antigen rising in all 3 dogs from <16 to reciprocal titers of 64 after 5 weeks of flea exposure.
Bartonella IFA Serologic Testing
Antibody titers to Bvb, Bh, and Bk antigens were determined in all 42 sera by IFA assays in the Intracellular Pathogens Research Laboratory, the research component of the NCSU-CVM VBDDL. Bartonella strains were isolated from naturally infected cats or dogs with species characterizations made by PCR amplification and DNA sequencing. Canine isolates of Bvb genotype I (NCSU 93CO-01, ATCC type strain #51672), Bvb genotype II (NCSU 95CO-08), and Bvb genotype III (NCSU 06CO-01) and feline isolates of Bh H-1 strain (NCSU 93FO-23), Bh SA2 strain (NCSU 95FO-099), both considered serotype I8, and Bk (NCSU 09FO-01) colonies were passed from agar plate–grown cultures into Bartonella-permissive cell lines, DH82 cells (a canine monocytoid cell line) for Bh strains H-1 and SA2, Bvb I and Bk and Vero cells (a mammalian fibroblast cell line) for Bvb II and III to obtain antigens for IFA testing. For each antigen, heavily infected cell cultures were spotted onto 30-well Teflon-coated slides (Cel-Line1 ), air dried, acetone-fixed, and stored frozen. Serum samples diluted in phosphate-buffered saline containing normal goat serum, Tween-20, and powdered nonfat dry milk to block nonspecific antigen binding sites were screened at dilutions of 1 : 16 to 1 : 64. All sera that were reactive at a titer of 1 : 64 were further tested with 2-fold dilutions out to 1 : 8,192. Fluorescein-conjugated goat anti-dog IgG (ICN1) was used to visualize bacteria within cells using a fluorescent microscope.2 To avoid confusion with possible nonspecific binding found at low dilutions, a cutoff of 1 : 64 was used to define a seroreactive titer.
Sera from 29 blood culture–negative SPF dogs were not reactive (<1 : 16) to any of the 6 Bartonella antigens tested.
Bvb Experimental Infection Study
Infection previously had been confirmed in all 10 Bvb genotype I-inoculated dogs by bacterial isolation on agar or PCR from postmortem tissues. During the course of the earlier study, Bvb genotype I IFA titers progressively ranged from 1 : 64 to 1 : 4,096. In the serum samples examined for this study from PID 28 or 34, IgG titers against the homologous Bvb antigen were between 1 : 256 and 1 : 1,024, whereas titers against Bvb genotypes II and III, Bh, and Bk were consistently negative at 1 : 16.
Bh Experimental Infection Study
Three dogs experimentally infected with Bh by flea exposure and proven Bh-bacteremic by culture in BAPGM with detection of Bartonella DNA by PCR developed Ab response against Bh H-1 (titers of 1 : 128), but not to Bh SA2. In addition, there was no seroreactivity against Bvb genotypes I, II, and III or to Bk antigens.
Based upon the results obtained from this study, naïve SPF dogs were not seroreactive to the panel of Bartonella spp. antigens. Serum samples from experimentally infected dogs (seronegative and PCR-negative) with Bvb genotype I or Bh serotype I developed a specific antibody response to the infecting organism. For example, dogs experimentally infected with Bvb genotype I reacted to the homologous antigen with no cross-reactivity to heterologous Bvb genotypes, Bh or Bk antigens. Serum samples obtained from dogs experimentally infected with Bh by fleas reacted to Bh H-1, but not to Bh SA2, Bk, or Bvb genotypes I, II, or III. If infection with other Bartonella spp. results in similar antigenic specificity in dogs, serology will have major diagnostic limitations, unless large panels of antigens are used to test each patient's serum. There are now over 30 named or Candidatus Bartonella spp. and at least 14 Bartonella spp. have been associated with human infections whereas 8 Bartonella species have been associated with infections in dogs.
In a 1997 study conducted at NCSU, Bvb antibody reactivity was tested by IFA in SPF dogs experimentally infected with Ehrlichia canis (n = 10) and Rickettsia rickettsii (n = 10). Neither of these organisms induced cross-reactions against Bvb I antigens. In a subsequent study, there was no cross-reactivity against Bh antigens in sera from 6 dogs experimentally infected with R. rickettsii and 7 dogs infected with E.canis. Although Bartonella, Ehrlichia, and Rickettsia all are genera within the class alpha Proteobacteria, cross-reactivity across genera was not detected in these 2 experimental infection studies. In human Bh and B. quintana endocarditis patients, serologic cross-reactivity has been reported to Chlamydophila psittaci and Coxiella burnetii. Based upon the results of this study and previously published results, it seems more likely that reactivity to multiple genera is the result of prior exposure to, or concurrent infection with, other organisms rather than cross-reactivity among phylogenetically related genera. As some dogs and human patients are seroreactive to multiple Bartonella sp. antigens, chronic infection with a specific Bartonella sp. could result in repeated exposure of the host to bacterial antigens and potentially different antigens over time, which might result in cross-reactivity among Bartonella species or strains. In addition, sequential or concurrent transmission of >1 vector-borne Bartonella sp. may further complicate the clinical and microbiologic interpretation of serology results in naturally exposed dogs.
Failure to detect antibodies by IFA testing in dogs infected with Bartonella spp. has been previously documented. For reasons that remain unclear, while using the same IFA assays as described in this study, seroreactivity was not detected in the majority of Bh-infected dogs (75% seronegative) or in approximately half of Bvb-infected dogs tested in a previous NCSU study involving 61 actively infected (based upon enrichment culture PCR and DNA sequencing) sick dogs. Serotypes and genotypes of the Bh- and Bvb-infected dogs, respectively, were not specified in the previous 61-dog study; however, the same 3 Bvb genotypes were used for IFA testing in that study as were used in the current study. Inasmuch as Bh H-1 and Bh SA2 (both serotype I strains) are the predominant 16S-23 ITS strain types identified in dogs tested by the NCSU-VBDDL, a serotype II isolate, ie, the Marseille serotype, was not included in the previous or the current study. Therefore, some IFA-negative serology results may have been attributable to the test strain selected by the laboratory for IFA testing. Thus, IFA serology appears to have low diagnostic sensitivity, and negative IFA results should be interpreted with caution in seroepidemiologic studies, when testing individual dogs, or when screening healthy dogs as potential blood donors. Similarly, a subset of 69 of 92 Bartonella-infected humans did not have IFA antibodies to the infecting Bartonella species or genotype in another NCSU study using the same antigen sources described in this study.
We conclude that dogs develop a Bartonella species– or Bvb genotype–specific antibody response after experimental IV inoculation or flea-associated infection with a defined Bartonella sp. As illustrated by a recent study that examined Bartonella infections in a multi-dog and human household, there is greater ecologic and serodiagnostic complexity associated with exposure to, and infection with, ≥1 Bartonella spp. than is currently appreciated in veterinary or human medicine. Also, with added sensitivity in PCR and culture modalities, the evidence of blood-borne infections caused by a diversity of Bartonella spp. as illustrated in pet dogs in the United States or stray dogs in Thailand is increasing. Because of the escalating number of Bartonella spp. that are being described around the world, the use of IFA to diagnose acute Bartonella infections may be problematic unless patient sera are tested against an extensive panel of antigens.
Conflict of Interest: In conjunction with Dr Sushama Sontakke and North Carolina State University, Dr Breitschwerdt holds U.S. Patent No. 7,115,385; Media and Methods for cultivation of microorganisms, which was issued October 3, 2006. He is the chief scientific officer for Galaxy Diagnostics, a company that provides diagnostic testing for the detection of Bartonella species infection in animals and human patients. Dr Ricardo Maggi has led research efforts to optimize the BAPGM platform and is the Scientific Technical Advisor for Galaxy Diagnostics. All other authors have no potential conflicts.
Thermo Fisher Scientific, Waltham, MA
Carl Zeiss Microscopy, LLC, Thornwood, NY