Enrichment Blood Culture Isolation of Bartonella henselae from Horses with Chronic Circulatory, Musculoskeletal and/or Neurologic Deficits

Background: Because published case-based evidence is limited and only one experimental equine infection study has been reported to date, bartonellosis, a recently recognized equine disease, is rarely considered as a differential diagnosis by equine clinicians. Microbiological documentation of Bartonella spp. infections has been difficult to achieve in the laboratory; therefore, elucidation of the potential clinical and pathological aspects of chronic bartonellosis in horses has not been thoroughly investigated.


Advances in Biotechnology & Microbiology
cats, dogs and humans [1][2][3]. However, equine bartonellosis remains largely undefined as a disease of horses. These highly adaptive, intravascular and endotheliotropic bacteria are transmitted among domestic and wild animals throughout the world, primarily by blood-feeding arthropod vectors, including bedbugs, biting flies, fleas, lice, mites, sand flies, ticks and potentially spiders [4][5][6]. In addition, bites, scratches, needle sticks, and for some species, transplacental, have all been implicated as modes of disease transmission [1][2][3][4]. Persistent intravascular infection with different Bartonella spp. has been reported in numerous pet, production and wildlife animals [1][2][3]. Following transmission in nature, duration of bacteremia in horses is unknown, whereas following experimental intra dermal inoculation of four horses with culture-grown Bartonella henselae, three sero-converted and developed bacteremia and mild cutaneous disease manifestations [7]. It was not until 2008 that B. henselae was isolated or detected for the first time by polymerase chain reaction (PCR) amplification in the blood of two adult horses from North Carolina (USA), one with vasculitis and the other with chronic intermittent shifting leg lameness [8]. In 2009, B. henselae infection was documented in an aborted equine fetus in Indiana (USA) [9], and in 2011, B. henselae DNA was amplified from the bone marrow and spleen of a horse that succumbed to hemolytic anemia in Germany [10]. More recently, B. henselae was isolated using the BAPGM enrichment blood culture approach from a Thoroughbred foal with granulomatous hepatitis, pathology that had been previously reported in dogs and humans [11].
Serology, PCR amplification of organism-specific DNA sequences from blood, and conventional blood culture isolation techniques have not proven useful for confirming a diagnosis of canine, equine or human Bartonella bacteremia [12][13][14][15]. Relatively recent advances in microbiological culture technology have enabled more sensitive documentation of B. henselae bacteremia in foals and in horses with presumptive vasculitis and chronic arthropathy [13]. The purpose of this study was to summarize historical, clinical and laboratory findings in horses with laboratory confirmation of Bartonella spp. bacteremia using Bartonella ePCR™ (Galaxy Diagnostics, Inc. Morrisville, NC USA) platform.

Materials and Methods
Horses included in this study had a history of chronic symptoms in conjunction with poor skeletal muscle exercise tolerance, microcirculatory deficits, and/or neurologic dysfunction. Chronicity was based upon owner reported symptoms extending over a period of at least two years. Poor exercise tolerance was defined as delayed forward phase of stride and/or muscle weakness (with or without muscle spasms) and/or shifting leg lameness. Microcirculatory involvement was defined in terms of sequelae indicative of poor arterial oxygenation of soft tissue(s) and/or chronic inflammation in a wide range of tissues and/or evidence of impaired venous and lymphatic drainage. Neurologic dysfunction was defined as unexplained cranial nerve deficits and/or symptoms consistent with neurotransmitter imbalances. Additional historical abnormalities involving endocrine system impairment, a finding reported in other mammalian species with confirmed bartonellosis, were considered as adjunct criteria for Bartonella bacteremia testing.

Historical, Physical Examination and Blood Specimen Collection
Study population: Each owner provided historical and demographic data for their horses' travel and arthropod exposure history. Physical examinations were concurrently performed by the respective primary care veterinarians in conjunction with blood specimen collection from each horse on three alternate days. Ethylene diamine tetraacetic acid (EDTA)anti-coagulated blood and serum samples from horses were shipped to Galaxy Diagnostics, Inc. (Research Triangle Park, NC, USA) from Australia and the United States for Bartonella ePCR™ testing.

Bartonella IFA serology
As previously described by Cherry NA et al. [14], IFA was performed to document detection of Bartonella antibodies using B. henselae (San Antonio-2, and Houston-1 strains), B.vinsonii subsp. berkhoffii (genotypes I and II), B. koehlerae and B. quintana organisms grown in mammalian cells as antigens. Serum was available from 10 of the 22 horses described in this study. Sera (kindly provided by Dr. Bruno Chomel, University of California, Davis) from three horses experimentally-infected with B. henselae (SA2 strain, isolate designation NCSU 2008-EO-1) were used as IFA positive controls. In our assay, endpoint titers for these three horse sera were 1:512, 1:512, and 1:2048. The starting dilution was 1:16, with endpoint titers defined as the last dilution at which brightly-stained organisms could be detected by fluorescence. A cut-off titer defining sero-reactivity was arbitrarily defined as 1:64.

Bartonella ePCR™
Bartonella ePCR™, an adaptation of a previously-described approach that combines PCR amplification of Bartonella spp. DNA from blood before and after inoculation into BAPGM, was performed [12,13,15,16]. For each blood sample, 2mL of EDTA blood was spiked into 10mL BAPGM and incubated at 35-37 °C with 5% CO2. DNA was extracted directly from 200µl of EDTA blood and from 200µl of BAPGM enrichment blood cultures after three incubation time points (8,14 or 15, and 21 or 22 days) using an automated QIA Symphony Workstation and DSP DNA blood mini kit (Qiagen, Valencia, CA). As extraction controls, a positive culture control consisting of BAPGM spiked with B. henselae and a negative culture control containing only BAPGM were processed simultaneously with each sample set to ensure extraction was completed successfully and that extraction buffers and reagents were not contaminated with Bartonella DNA. Bartonella DNA was amplified using conventional Bartonella genus primers targeting the 16S-23S intergenic spacer region
In addition, Bartonella genus qualitative real-time PCR (qPCR) was performed targeting the Bartonella 16S-23S ITS gene using oligonucleotides (5' CTTCAGATGATGATCCCAAGCCTTYTGGC3') and (5' GCCCTCCGGGRTAAAYCGGAAACC 3') as forward and reverse primers, respectively. Amplification was performed in a 25µl final volume reaction containing 12.5µL of MyTaq Premix (Bioline USA, Inc., Taunton, MA) for conventional PCR or SensiFAST™ SYBR No-ROX kit (Bioline USA, Inc., Taunton, MA) for qPCR, 0.3µL of 100µmol/L of each forward and reverse primer (IDT®DNA Technology, Coralville, IA), 6.9µL molecular grade water, and 5µL DNA from each sample tested. Amplification was performed using an Eppendorf Mastercycler EPgradient® (Eppendorf, Hauppauge, NY) for conventional PCR and the Bio-Rad CFX96 Touch TM Real Time PCR Detection System (Bio-Rad, Inc., Hercules, CA) for qPCR with 0.01 and 0.001 pg/µL B. henselae SA2 genomic DNA as PCR positive controls. Negative controls included BAPGM and molecular-grade water. For conventional PCR, thermocycling conditions consisted of initial denaturation at 95 °C for 2 minutes, followed by 55 cycles with denaturation at 94 °C for 15 seconds (s), annealing at 68 °C for 15s, and extension at 72 °C for 18s. PCR products were analyzed by 2% agarose gel electrophoresis. For qPCR, thermo cycling conditions consisted of initial denaturation at 95 °C for 2 minutes, followed by 45 cycles with denaturation at 94 °C for 10s, annealing at 66 °C for 10s, and extension at 72 °C for 10s. Positive amplicons were sequenced to identify Bartonella sp. and ITS genotype. Bacterial species and genotype were defined by comparing similarities with other sequences deposited in the GenBank database using the Basic Local Alignment Search Tool (BLAST version 2.0).

Advances in Biotechnology & Microbiology Discussion
Bartonella henselae bacteremia was documented by BAPGM enrichment blood culture in nine horses with chronic, intermittent, relapsing, and at times progressive symptoms that spanned a several year time frame. Importantly, all of these horses had overlapping symptoms involving the musculoskeletal, nervous, and endocrine and circulatory systems. The extent to which B. henselae bacteremia contributed to individual or the collective symptoms reported in this case series could not be determined. Previously, Merrell and Falkow described the challenges associated with establishing disease causation for stealth pathogens, including members of the genus Bartonella [18]. As vector-transmitted, intracellular, organisms, the evolutionary lifestyle of this genus is characterized by erythrocyte and endothelial cell tropisms. Based upon in vitro studies, B. henselae can also invade monocytes, dendritic cells, CD34+ progenitor cells, pericytes, and synoviocytes (unpublished data) [19][20][21][22]. Thus, once bacteremic infection is established, the organism is able to circulate throughout the microvasculature of the host, and presumably localize in microcapillaries within various tissues, thereby involving organs throughout the body. Chronic micro vascular infection, in conjunction with interactions associated with the host immune response, presumably results in a diverse and fluctuating spectrum of symptoms that may span years in duration. Satisfying Koch's postulates has represented the historical benchmark for establishing infectious disease causation; however, the original postulates have substantial limitations when attempting to support disease causation associated with stealth pathogens. Because infections with various Bartonella species induce similar or identical pathology in different animal hosts, including humans, the Postulate of Comparative Infectious Disease Causation was proposed as an additional fifth Koch's postulate [23].
This postulate states that microbiological documentation of the same infectious agent(s) in three different animal species with identical pathology supports stealth pathogen infectious disease causation. This postulate has medical utility when assessments are based upon common histopathological and microbiological findings in three different animal hosts. It is also possible that the comparative documentation of similar, subtle and non-specific symptoms among different companion animal species may further facilitate our understanding of the biological consequences of long standing B. henselae bacteremia in horses and other non-reservoir-adapted animals. Because animals share the same environment and because their behaviors are generally carefully monitored by their owners, symptoms and behavioral patterns observed in pets may have comparative medical implications for their human counterparts.
Diagnostic confirmation of B. henselae bacteremia in horses remains challenging. For reasons that remain unclear, serology in naturally-infected horses is insensitive. Similar to previous reports [13], only one horse in this study was B. henselae seroreactive, despite a history of chronic symptoms and concurrent diagnostic documentation of bacteremia. When using BAPGM enrichment blood culture to test dogs, cats and humans, shorter incubation periods are often sufficient for growth and detection of Bartonella in bacteremic individuals, whereas incubation for 21 or 22 days was necessary to document bacteremia in all but two of these nine horses. This observation could reflect the possibility that bacteremic horses maintain lower levels of bacteremia compared to most infected cats, dogs or humans, as comparable amounts of blood are inoculated into the liquid medium, regardless of the animal species being tested. When testing humans for evidence of Bartonella spp. bacteremia using BAPGM enrichment blood culture, obtaining specimens from three time points within a seven day period increased diagnostic sensitivity [24]. This approach also appears to have microbiological utility when testing horses, as no horse was PCR positive in all three blood culture specimens. Thus, until a more sensitive microbiological method becomes available, the Bartonella ePCR™ triple draw approach is recommended when assessing Bartonella spp. bacteremia in horses.
There are several limitations inherent in this study. The duration and types of symptoms varied widely among horses that were tested, thereby decreasing the clinical utility of the reported findings. Case controlled studies will be necessary to define specific symptoms or symptom patterns that would support the need for Bartonella spp. diagnostic testing. Clinical evaluations were performed by different equine clinicians, in different countries, all dealing with a broad range of economic, logistical and technical limitations related to specimen collection and the provided medical care. For all horses in this study, there had been extended, prior diagnostic and therapeutic efforts to manage their symptoms. Depending upon geographic location and clinician preference, some horses were tested for other infections contributing evidence of exposure to Borrelia and Toxoplasma spp. Testing performed by the authors was directed solely at determining if a horse was Bartonella spp. bacteremic. Thus, other infectious or non-infectious diseases were not systematically evaluated. Historically, use of the BAPGM blood culture platform has documented bacteremic coinfections with B. henselae and other bacteria in dogs [25] and humans [26]. Thus, there appears to be a complex pathobiome of blood that potentially changes with age, breed (genetics), nutrition, toxin exposure and interactions with a spectrum of bacteria, viruses and protozoa in the environment. Subcultures were not performed on samples from horses in this study to assess co-infections. Based upon evolving comparative clinical, microbiological and pathological data, tick borne pathogens, including Bartonella species, may constitute an important component of the pathobiome of blood in horses and other animals [27].
Within the context of differential diagnoses, the parameters described within this group of bacteremic horses could be incorporated by equine practitioners into a wide range of routine clinical evaluations such as lameness exams and pre-purchase exams. As part of a complete physical examination the examiner might review any and all presenting complaints and if necessary engage the horse owner in a series of specific questions for the express purpose of collecting complete historical data including history of exposure to biting insect or arthropod vectors and travel history. For horses presenting with history of chronic and/or sporadic exercise intolerance, fluctuating degrees of unsoundness, deficits in neurotransmission and/or chronic circulation deficits and/or relapsing symptoms such as those reported herein, testing for bartonellosis should be considered. The extent to which Bartonella bacteremia might contribute to behavioral abnormalities is beyond the scope of this report, however histories of neurotic, argumentative, unruly and/ or aggressive behavior patterns as seen in other species with chronic bartonellosis [28] lend support to a higher index of suspicion when considering testing options.
We conclude that the extent to which bartonellosis causes chronic disease manifestations in horses deserves additional research consideration. In-depth, sequential "lifetime" studies that span years in duration may be required to establish whether some or all of the symptoms described in the horses in this case series are causally associated with bartonellosis. It is possible that persistent intravascular infection with Bartonella spp. contributes to complex disease expression in horses, as seemingly occurs in humans and other animal species.