• Open Access

Phylogenetic Diversity of Bacteria Isolated from Sick Dogs Using the BAPGM Enrichment Culture Platform


  • A.C. Davenport,

    1. Intracellular Pathogens Research Laboratory, Center for Comparative Medicine and Translational Research, College of Veterinary Medicine, North Carolina State University, Raleigh, NC
    Search for more papers by this author
  • P.E. Mascarelli,

    1. Intracellular Pathogens Research Laboratory, Center for Comparative Medicine and Translational Research, College of Veterinary Medicine, North Carolina State University, Raleigh, NC
    Search for more papers by this author
  • R.G. Maggi,

    1. Intracellular Pathogens Research Laboratory, Center for Comparative Medicine and Translational Research, College of Veterinary Medicine, North Carolina State University, Raleigh, NC
    Search for more papers by this author
  • E.B. Breitschwerdt

    Corresponding author
    • Intracellular Pathogens Research Laboratory, Center for Comparative Medicine and Translational Research, College of Veterinary Medicine, North Carolina State University, Raleigh, NC
    Search for more papers by this author

  • Data from this study was presented as an abstract at the 2012 American College of Veterinary Internal Medicine Forum, New Orleans, LA.

Corresponding author: E.B. Breitschwerdt, DVM, Department of Clinical Sciences, College of Veterinary Medicine, North Carolina State University, 1060 William Moore Drive, Raleigh, NC 27607; e-mail: ed_breitschwerdt@ncsu.edu.



Bartonella alpha-Proteobacteria growth medium (BAPGM) enrichment culture has proven useful for documenting Bartonella species infection and has facilitated growth of other fastidious bacteria from human samples.


To report non-Bartonella bacterial isolates obtained from canine samples cultured using BAPGM enrichment culture.


Between 2004 and 2008, 695 specimens from 513 dogs were tested by the NCSU-IPRL using the BAPGM enrichment culture. Over the same period of time, blood samples from 270 dogs were cultured by the NCSU-CML using Bactec-Plus Aerobic/F media.


BAPGM isolates were characterized using Bartonella genus primers and 16S rDNA primers followed by DNA sequencing. NCSU medical records were retrospectively reviewed. Blood culture results from the NCSU-CML were compared with BAPGM blood culture results.


Seventy-nine non-Bartonella isolates were obtained from 69/513 dogs. The most commonly isolated phylum was Proteobacteria (48.1%) with alpha-Proteobacteria being the most commonly isolated class. Staphylococcus and Sphingomonas were the most commonly isolated genera. The majority of the remaining isolates were bacteria that are rarely isolated from canine samples. Comparison of NCSU-CML and IPRL (BAPGM) blood culture isolates showed alpha-Proteobacteria were isolated more often from BAPGM.

Conclusions and Clinical Importance

Use of insect cell culture enrichment medium, such as BAPGM, appears to enhance the growth of alpha-Proteobacteria, but also results in isolation of non-alpha-Proteobacteria from sick dogs. Future studies are needed to elucidate the utility of BAPGM and other “nonconventional” growth media and methods for isolation of fastidious organisms and to determine if these organisms play a causal role in disease development.


Bartonella alpha-Proteobacteria growth medium


ethylene diaminetetraacetic acid


fluorescent in situ hybridization


immune-mediated hemolytic anemia


immune-mediated neutropenia


North Carolina State University, College of Veterinary Medicine, Intracellular Pathogens Research Laboratory


immune-mediated thrombocytopenia


North Carolina State University-Veterinary Health Complex Clinical Microbiology Laboratory


North Carolina State University, College of Veterinary Medicine, Veterinary Health Complex


polymerase chain reaction

Historically, culture has been the gold standard for microbiological confirmation of bacterial infections, but standard culture methods often fail to isolate highly fastidious, intracellular, or uncultivable microorganisms that require specific growth conditions or cell culture approaches to achieve successful isolation. Based upon phylogenetic analysis of DNA sequences deposited in GenBank, these are estimated to be >36,000 unique bacterial species inhabiting earth, but the majority of these organisms have never been successfully isolated using contemporary bacteriologic methods.[1, 2] Clinical and microbiological utilization of molecular-based diagnostic tests such as PCR, in conjunction with conventional microbiological approaches, support a much more complex role for normal bacterial flora and for fastidious, intracellular, or uncultivable bacteria in the spectrum of animal and human diseases.

In both human and veterinary medicine, culturing bacterial organisms from diagnostic specimens has largely relied upon the use of standard growth media and formulations that are primarily derived from mammalian blood, serum or tissue extracts, to which additional nutrients or substrates are added.[3] Our laboratory, has reported the development of a novel liquid bacterial growth medium, Bartonella alpha-Proteobacteria growth medium (BAPGM), which is based upon a biochemical formulation that promotes the growth of insect cells in culture and is optimized to achieve more successful isolation of Bartonella spp.[4] As members of the genus Bartonella are primarily transmitted by arthropod vectors, this tropism provided the original rationale for the development of an “insect-based” growth medium to improve isolation of these fastidious bacteria. Utilization of BAPGM enrichment culture, which incorporates PCR targeting Bartonella genes, before and after enrichment blood culture has improved the diagnostic sensitivity for confirmation of Bartonella infection in animal and human patients.[5-7] Historically, a variety of specialized media and techniques has been employed by microbiologists to improve the isolation of other fastidious organism, one notable example being Mycobacterium, a genus with slow dividing times and specialized growth requirements for which a variety of egg-based, agar-based and selective media has been developed that have dramatically improved the isolation of this genus, but some species (Mleprae) remain uncultivable.[8]

Although BAPGM initially was optimized to enhance the growth of Bartonella species, non-Bartonella isolates have been obtained from veterinary and human patient samples after BAPGM enrichment culture. Examples have included isolation of Mkansasii from a dog with intractable pleural effusion[9] and the isolation of bacteria from pericardial effusion samples from horses with fatal, fibrinous pericarditis—a disease historically believed to represent a sterile inflammatory process.[10] In 2007, Cadenas et al reported the use of 16S rDNA PCR and DNA sequencing to identify non-Bartonella BAPGM subculture isolates obtained from human blood, fluid, and tissue samples. The purpose of this study was to determine the identity of non-Bartonella bacteria isolated by means of BAPGM culture from sick dogs, using 16S rDNA PCR and sequencing. BAPGM culture isolates obtained from the blood of these dogs were compared with isolates obtained from a cohort of sick dogs that had conventional blood cultures performed over the same time interval.

Materials and Methods

Dogs, Specimens, and Isolate Sources

Between July 2004 and December 2008, 695 diagnostic specimens from 513 dogs were submitted to the North Carolina State University College of Veterinary Medicine Intracellular Pathogens Research Laboratory (NCSU-CVM IPRL) for culture using the BAPGM platform. Bartonella sp. infection was documented in 61 of these 513 dogs, as previously described.[11] Seventy-nine Bartonella spp. PCR negative (non-Bartonella) isolates were derived from 69 sick dogs, of which 14 were coinfected with a Bartonella sp. All non-Bartonella isolates were included in this study, regardless of whether the dog was infected with a Bartonella sp. Twenty-five of the 69 sick dogs having non-Bartonella bacteria isolated by BAPGM culture were examined at the Veterinary Health Complex at North Carolina State University (NCSU-VHC); the remaining dogs (44) were examined by veterinarians in private or specialty practices (non-NCSU-VHC cases) who submitted blood or other specimens for BAPGM enrichment cultures. Standardized or complete medical records for the non-VHC cases were not consistently available. Therefore, information reported for these dogs was limited to signalment, sample source, and isolate identification. Detailed medical records for the 25 NCSU-VHC patients were reviewed retrospectively.

Dogs that had blood samples submitted for culture to the NCSU Clinical Microbiology Laboratory (NCSU-CML) between 2004 and 2009 were identified using the NCSU-VHC database. These blood cultures were collected in a manner recommended by NCSU-VHC protocol, which entails aseptic collection of 1–10 mL of blood that is, based on the volume, placed into pediatric or adult Bactec-Plus Aerobic/F media1 before transporting it to the NCSU-CML. For most cases, 1–2 additional aseptically collected blood samples were obtained from different venous sites at later time points (30–60 minutes after the previous blood sample was collected), but this information was not always recorded in the medical records. Bactec + Aerobic/F media is an all-purpose soybean casein-based broth with resins that facilitates growth of aerobes and facultative anaerobes. Cultures were incubated at 37°C and subcultured onto Columbia blood agar, MacConkey agar, and chocolate agar as indicated by the detection of CO2 production, which indicates microbial growth in the bottles. If no growth was detected, the culture vials were discarded after 5 days. Plates were incubated for 48–72 hours and, if organism growth was visualized, colonies were identified based on morphology and metabolic activity using commercially available kits.2 Medical records were reviewed to obtain culture results.

Specimen Processing Using the BAPGM Platform

Blood and fluid specimens were inoculated into BAPGM enrichment medium using aseptic technique in a biosafety hood (2 mL of blood in 10 mL of medium or 1 : 5 ratio if sample volume was <2 mL). For tissue specimens, 0.5–1.0 g of tissue (depending on specimen size) was excised from the interior of the biopsy specimen, sliced several times with a sterile scalpel blade and inoculated into 5.0 mL of BAPGM. All specimens were incubated for a minimum of 7 days (maximum enrichment time, 21 days) under the following conditions: 35°C, 5% CO2, in a water-saturated atmosphere. After enrichment culture in a liquid BAPGM flask, subculture specimens were inoculated onto 10% sheep blood TSAII agar plates, placed inside plastic bags (to avoid potential cross-plate contamination and to minimize dehydration of the agar medium), and maintained for up to 3 weeks, under the same incubation conditions. Whenever individual colonies were of adequate size to be selected with a sterile loop, a single colony was preferentially selected for DNA extraction. For most isolates, however, the colonies were very small (pinpoint), so pooled colonies were collected by scraping the surface of the agar plate with a sterile loop. All isolates were stored in sucrose-phosphate-glutamate buffer (218 mM sucrose, 3.8 mM KH2PO4, 7.2 mM K2HPO4, 4.9 mM l-glutamate, pH 7.2) until processed for this study. An uninoculated BAPGM enrichment medium flask (negative control) was simultaneously processed with each sample or each sample batch. Liquid BAPGM negative control cultures were checked for turbidity, and the uninoculated medium remained clear with no development of turbidity throughout this study. Also, no bacterial growth was observed on any of the blood agar plates that were subinoculated from the BAPGM negative control cultures, which was an indication that no contamination resulted from setting up the agar plate cultures. Although colonies were never visualized on negative control subculture agar plates, each plate was routinely scraped with a sterile loop, resuspended in SPG buffer, and processed by DNA extraction and PCR in the same manner as clinical specimen isolates. DNA extraction was performed with BioRobot M48, using Qiagen MagAttrack DNA M48 kit.3

PCR Testing, Cloning, and Sequencing

Using previously described PCR amplification and DNA sequencing methods, patient blood, BAPGM enrichment culture specimens, and subculture isolates were screened for evidence of Bartonella sp. infection targeting the 16S–23S intergenic spacer (ITS) region before inclusion in this study.[5] Also, as described in a previous study from our laboratory, the PCR premix and RNase-free water provided in each DNA extraction kit, was treated with DNase I in order to minimize the potential amplification of contaminant bacterial DNA in these reagents.[12] Subsequently, PCR targeting the 16S rRNA gene was performed using modified universal primers 515f: 5′-GTGCCAGCAGCCGCGGTAA-3′ and 1391R: 5′-GACGGGCGGTGWGTRCA-3′ as forward and reverse primers, respectively. Amplification and product analysis was performed as previously described[12] using an Eppendorf Mastercycler ep,4 and Bhenselae (Houston I strain) genomic DNA was used as a PCR positive control. Negative controls included liquid media (BAPGM), SPG buffer, and molecular-grade water. No 16S rDNA amplicons were obtained in any negative control lane. Cloning and sequencing were performed as previously described.[12] To identify bacteria at genus, species, and strain levels, sequence analysis and alignment with GenBank5 sequences was performed using AlignX and ContigExpress software.6


Specimen Sources, Bartonella Testing, and BAPGM Subculture Isolates

After BAPGM enrichment, subculture isolates were obtained from 69 of 513 dogs tested during the study period (Fig 1). A total of 93 unique bacterial DNA sequences (isolates) were obtained from the 69 dogs. Seventy-nine of the 93 DNA sequences were non-Bartonella genera. A single isolate was obtained from 49 of the 69 dogs whereas polymicrobial infection was documented in 20 dogs. Two or 3 organisms were isolated from 16 and 4 dogs, respectively. Fourteen dogs were coinfected with a Bartonella spp., of which 5 were infected with Bvinsonii subsp. berkhoffii, 1 with Bhenselae, and 3 with both Bvinsonii subsp. berkhoffii and Bhenselae. The Bartonella PCR amplicons were not successfully sequenced for the remaining 5 dogs, therefore the infecting species were not determined.

Figure 1.

Flow diagram of the distribution of BAPGM subculture isolates obtained from 69 dogs included in this study.

Specimen sources for the 79 non-Bartonella subculture isolates included aseptically collected blood (n = 56), serum (n = 4), lymph node aspirates or biopsies (n = 5), effusions (n = 5; n = 3 pleural, n = 2 pericardial), rib biopsy specimen (n = 1), joint fluid (n = 1), aqueous humor (n = 1), bone marrow aspirate (n = 1), and aortic valve tissue (n = 1). Bacterial colonies were identified as early as 24 hours or at any incubation time point up to 4 weeks after subinoculation from the liquid BAPGM enrichment culture onto a blood agar plate. Duplicate specimens (ie the same source) processed from 5 dogs yielded identical bacterial isolates and were counted as 1 specimen, and the isolate was reported only once for these 5 dogs. Two dogs had multiple specimens that yielded inconsistent results: 2 lymph node aspirates (lymph node sites not noted) from a single dog yieldeddifferent organisms (a Bacillus sp. and Staphylococcus sp.) and 3 BAPGM enrichment blood cultures from 1 dog resulted in isolation of Bvinsonii subsp. berkhoffii from 2 specimens and a Staphylococcus sp. from the third.

Phylogenetic Characterization of Isolates

All 79 non-Bartonella isolates were members of 4 bacterial phyla: 38 isolates were Proteobacteria (Class alpha, gamma, and beta-Proteobacteria), 23 isolates were Firmicutes (Class Bacilli, Class Clostridia), 14 isolates were Actinobacteria (Class Actinobacteria), and 3 isolates were Bacteroidetes (Class Bacteroidetes). Based upon available 16S rDNA sequences reported in GenBank, 1 bacterial organism could not be identified at the phylum level and 7 bacteria could not be identified at the genus level. The 14 Bartonella isolates, members of the Class alpha-Proteobacteria, were not included in these cumulative numbers. Proteobacteria (38/79, 48.1%) was the most commonly isolated phylum. Organisms within the Class alpha-Proteobacteria were more frequently isolated (n = 21) than gamma (n = 9), and beta (n = 8)-Proteobacteria. The most frequently isolated non-Bartonella genera were Staphylococcus (Phylum Firmicutes) and Sphingomonas (Phylum Proteobacteria, Class alpha-Proteobacteria). The distribution of all non-Bartonella isolates identified in this study is outlined in Table 1.

Table 1. 16S rDNA sequence identity of 79 non-Bartonella bacteria isolated from 69 sick dogs by means of BAPGM enrichment culture
PhylumGenera SequencedNumber of Isolates
Proteobacteria (n = 38)
α Sphingomonas 13
Ochrobactrum 3
Methylobacterium 2
Brevundimonas, Sphingopyxis, Sphingobium1 each
β Burkholderia 3
Achromobacter, Acidovorax, Ralstonia, organisms not ID'd to genus level: Burkholderiales, beta-Proteobacteria1 each
γ Pseudomonas 3
Moraxella 3
Cibimonas, Serratia, organisms not ID'd to genus level: gamma-Proteobacteria1 each
Firmicutes (n = 23) Staphylococcus 15
Lactobacillus 2
Streptococcus 2
Bacillus 2
Aneurinibacillus, Eubacterium1 each
Actinobacteria (n = 14) Corynebacterium 5
Arthrobacter 4
Nocardia, Actinomyces, Nocardiodes, organisms not ID'd to genus level: Proprionibacteriaceae, Uncultured Actinobacteria1 each
Bacteroidetes (n = 3) Chryseobacterium 2
Uncultured Bacteroidetes1
Unclassified Bacterium1

Retrospective Review of NCSU-VHC Cases

Twenty-five dogs having non-Bartonella bacteria cultured by means of BAPGM enrichment were patients of the NCSU-VHC and had medical records available for review. These dogs had highly variable clinical presentations and samples typically were submitted for BAPGM testing when bartonellosis or another infectious disease was suspected. Because serological and molecular vector-borne disease diagnostic panels are commonly submitted for NCSU patients with immune-mediated diseases, BAPGM testing of these patients was likely overrepresented. Retrospective review of the clinical presentations of these 25 dogs identified several (n = 7) dogs in which an immune-mediated disease (polyarthritis [n = 4], immune-medicated cytopenias [n = 3]) was suspected, an effusive disease was documented (n = 5), or endocarditis was confirmed by echocardiography or necropsy (n = 4). Sphingomonas was isolated from 3 of the 4 dogs diagnosed with polyarthritis. Information on these individual dogs is available as a supplement to this report.

Thirty-one unique bacterial sequences were obtained from the 25 dogs examined at the NCSU-VHC. Although the VHC cases were predominantly tertiary referrals, the phylogenetic distribution of BAPGM isolates was similar to isolates from non-VHC cases. The majority of these isolates were members of the Phylum Proteobacteria (17/31, 54.8%), of which the majority were in the Class alpha-Proteobacteria (11/17, 65%). The most commonly isolated non-Bartonella bacteria was Sphingomonas (n = 6), followed by Staphylococcus spp. (n = 5). Six VHC dogs were coinfected with a Bartonella spp.

Thirteen of the 25 dogs having medical records available for review had specimens concurrently submitted to the NCSU-CML for culture by standard microbiologic methods. Eight dogs had specimens collected at the same time and from the same site for submission to both the CML and IPRL. Three of these 8 dogs had the same organism isolated by both laboratories, whereas divergent isolation results were obtained for 5 dogs (Table 2).

Table 2. Direct comparison of isolates from samples using the BAPGM platform and conventional culture techniques
Results Agree (Same Organism Isolated): 3/8Results Disagree with Disparate Results: 1/8Results Disagree because of No Growth in CML: 4/8
Group G Streptococcus—mitral valve endocarditis (blood)Arthrobacter (BAPGM), Ecoli (CML)—Chylothorax (pleural effusion)Sphingobium—aortic valve endocarditis (blood)
Serratia—aortic valve endocarditis (blood) Burkholderiales—Polyarthritis (synovial fluid)
Nocardia—rib osteomyelitis (rib) Sphingomonas—polyarthritis (blood)
  Sphingomonas—pericarditis (pleural effusion)

Comparison of Blood Culture Isolates Obtained Using BAPGM to Those Obtained from Conventional Blood Cultures from a Cohort of Dogs over the Same Time Period

Fifty-seven of the 79 non-Bartonella isolates obtained in this study were isolated from aseptically obtained blood samples. Of the 57 BAPGM enrichment blood culture isolates, the phylogenetic distribution was as follows: 56.1% Proteobacteria, 26.3% Firmicutes, 12.2% Actinobacteria, and 3.5% Bacteroidetes. Consistent with the distribution noted for all specimen sources, the majority of Proteobacteria were in the alpha-Proteobacteria class.

Because direct comparison of cultures using the different techniques and media employed in the CML and IPRL was only available in a small number of cases, BAPGM blood culture isolates were compared with blood culture isolates obtained from dogs that were patients at the NCSU-VHC. Over the same time period as this study, bacterial isolates were obtained from 43 of 270 dogs that had blood samples submitted for culture to the NCSU-CML. Forty-nine isolates were obtained from the 43 dogs. Isolates had the following phylogenetic distribution: 59.2% Firmicutes, 32.6% Proteobacteria (all from Class gamma-Proteobacteria), and 8.2% Actinobacteria. The majority of positive cultures yielded growth of a single isolate, but there were 6 polymicrobial infections. No organisms from the Class alpha-Proteobacteria were isolated from blood cultures at the NCSU-CML during the period of this study.


With the advent of new and more sensitive microbiological techniques, evolving evidence suggests that bacterial contributions to the pathogenesis of many historically “non-infectious” disease processes may be more common than is currently appreciated. It is estimated that fewer than 5% of known bacterial organisms have been isolated using contemporary bacteriologic methods, but isolation remains the clinical gold standard for the diagnosis of bacterial infections.[1, 13] Utilization of molecular-based diagnostic tests such as PCR and FISH, in conjunction with conventional microbiological approaches, is generating data that supports a much more complex role for normal bacterial flora and for fastidious, intracellular, or uncultivable bacteria in the spectrum of animal and human diseases such as histiocytic ulcerative colitis in Boxer dogs, caused by invasive Esherichia coli, and Whipple's disease, a malabsorptive gastrointestinal disease of humans caused by the bacteria Trophyrema whippelii.[14, 15] The results of this study indicate that using an insect-based culture medium such as BAPGM for enrichment culture, followed by subculture isolation and 16S rDNA PCR amplification will assist in the identification of bacterial infection in some patients. Although future studies are required to understand the relationship between specific organisms and clinical manifestations potentially associated with these infections, this study provides the 1st reported isolation of many of these bacteria from dogs. Characterizing their potential role in disease may prove to be of immense importance as clinicians and microbiologist strive to understand the complex interrelationships among microorganisms, patient immune status, and certain disease states.

The use of BAPGM resulted in more frequent isolation of organisms within the Class alpha-Proteobacteria compared with blood samples cultured by conventional techniques in our clinical microbiology laboratory. Although a limited number of dog samples were simultaneously tested by both BAPGM and conventional techniques, alpha-Proteobacteria were rarely isolated by conventional techniques. This difference most likely is related to the special growth requirements of organisms within the alpha-Proteobacteria class of bacteria and the long division times of some organisms (ie, Bhenselae has a 22-hour dividing time), which could result in false-negative cultures at the time the plates or culture bottles are routinely discarded. Additionally, some bacteria may exist in a cell wall-deficient or relatively inert state within the patient. Therefore, prolonged incubation in a liquid environment might facilitate reconstitution of the bacteria, thereby allowing for successful subculture onto a solid medium.

The non-Bartonella organisms most frequently isolated using BAPGM in this study were Staphylococcus and Sphingomonas. Sphingomonas, a gram-negative organism within the Class alpha-Proteobacteria, is a well-known environmental bacterium, commonly found in soil, water, water pipes, and coral reefs. Recently, Sphingomonas has been identified as an opportunistic pathogen, particularly in human hospital intensive care settings. There are numerous case reports and reviews identifying Spaucimobilis as a cause of nosocomial bacteremia and sepsis in people.[16, 17] Spaucimobilis is known to form a biofilm and often is isolated from patients with indwelling devices, those receiving dialysis or in association with mechanical ventilation.[17] Sphingomonas also has been identified as a contaminant of IV infusions, as occurred in 2007 when an outbreak of bacteremia was linked to contaminated fentanyl.[18] Although bacteremia and sepsis are the most commonly reported clinical manifestations of Sphingomonas infection, Sphingomonas has been isolated from human patients with peritonitis, urinary tract infection, pneumonia, meningitis, splenic abscessation, arthritis, osteomyelitis, and cholecystitis.[17] To our knowledge, infection with this organism has not been published previously in the veterinary literature. In this study, Sphingomonas was isolated from 3 of the 4 dogs diagnosed with polyarthritis, suggesting a possible association between Sphingomonas infection and polyarthritis in dogs, though additional studies are required to investigate this possible relationship. Based on this study, BAPGM enrichment culture facilitated the isolation of Sphingomonas sp., which may prove to be a clinically relevant bacterium in dogs. Staphylococcus was the most frequently isolated non-Bartonella bacterium using BAPGM in this study. Staphylococcus is recognized as easy to culture and identify in a microbiology laboratory, but a subset of Staphylococci has proven to be difficult to cultivate because of special growth requirements and slow doubling times (so-called small colony variants). Small colony variant Staphylococcus infections are a well-documented cause of recurrent and persistent infections in humans.[19] In this study, some Staphylococci grew quickly (colonies identified within 24 hours), but visualization of colonies often required days to weeks after enrichment and subculture. Some of these isolates may have been fastidious forms of Staphylococci, such as small colony variants, which may have been missed using techniques with shorter incubation times, although direct comparison by simultaneous culture in our CML was not available in any of these cases.

The pathogenicity of many of the organisms isolated in this study remains unknown. Some of the isolates are well-recognized pathogens (eg, Streptococcus, Staphylococcus, Pseudomonas, Nocardia, Burkholderia) whereas other organisms (eg, Sphingobium, Sphingopyxis, Aneuribacillus, Cibimonas, Nocardioides) have no reported association with disease in human or veterinary patients. Some of the isolated organisms, including Arthrobacter spp., Moraxella osloensis, Methylobacterium, Acidovorax, Chryseobacterium meningosepticum, Ochrobactrum, Achromobacteria, Lactobacillus, and Serratia marcescens, have only been reported as human pathogens relatively recently—most often as nosocomial or opportunistic infections.[20-28] Only a few of these latter organisms have previously been reported as a cause of disease in dogs. For example, Achromobacteria was isolated from a dog with an infected total hip implant.[29] Corynebacterium auriscanis has been linked to a variety of infections, such as deep pyoderma, vaginitis, and cysts in dogs.[30] Smarcescens has been reported as a cause of bacteremia, nosocomial infections, and more recently endocarditis in dogs.[11, 31, 32] Moraxella, Lactobacillus, Corynebacterium, and Arthrobacter spp. have been reported as normal bacterial flora in the canine mouth[33, 34] and dogs with inflammatory bowel disease have been shown to have altered duodenal microbiota with an increased proportion of several organisms in the intestines, including Pseudomonas, Brevundimonas, and Achromobacter.[35]

In 2007, Cadenas et al reported the 16S rDNA characterization of 34 BAPGM isolates that were cultured from blood or tissue specimens obtained from 130 human patients suspected of having bartonellosis. The isolates were equally represented in composition by 3 bacterial phyla (Firmicutes, Proteobacteria, and Actinobacteria). The most common isolates included Staphylococcus spp. (6 isolates), Arthrobacter (5 isolates), Bacillus (4 isolates), Methylobacteria spp. (3 isolates), and Sphingomonas (2 isolates). There were 7 polymicrobial non-Bartonella infections and 9 Bartonella infections from which another bacteria was isolated concurrently.

There were several limitations to this study and to the clinical use of BAPGM enrichment culture for diagnosis of non-Bartonella bacterial infections in dogs. The long incubation times employed with BAPGM culture, although advantageous for isolation of fastidious organisms, also result in a long turnaround time, often taking several weeks until results are available. Despite instructions to use aseptic sample technique, it is possible that some of the bacterial isolates characterized in this study are the result of contamination at the time of sampling. We acknowledge that an enrichment culture approach with prolonged incubation times may enhance the isolation of nonpathogenic commensal organisms, normal flora, contaminants, or environmental bacteria. Laboratory contamination is considered an unlikely source of the isolates described in this study. An uninoculated BAPGM culture control was incubated alongside each sample set and in no case resulted in a positive culture. Another limitation of this study was that detailed medical records were only available for a set of patients (VHC cases) and subculture isolation results were not used in the clinical management of these patients. Additionally, the population of dogs for which samples were submitted for BAPGM culture included dogs that either were suspected of having bartonellosis or were unusual cases in which an underlying infectious etiology was considered a possibility, therefore the study population was biased. Finally, the pathogenicity of many of the isolates identified in this study remains unknown and we isolated a very diverse population of organisms. Therefore, it was not possible, nor was it our purpose, to propose a causal relationship between a particular organism and a specific clinical presentation. Additional studies are required to investigate the pathogenicity and disease phenotype associated with many of the organisms reported in this study.

Although most commonly recognized bacterial pathogens are effectively isolated with standard microbiologic practices and growth media, most organisms in nature have not been cultivated. As such, the majority of these environmental bacteria have not been linked to diseases in mammals, potentially because these organisms do not cause disease or potentially because historical and microbiological approaches have failed to isolate and thereby implicate these bacteria as a cause or cofactor in disease progression. We conclude that future use of special isolation media optimized to grow fastidious organisms, when combined with molecular diagnostic assays such as 16s rDNA PCR, will increase the identification of bacterial organisms in patient sample submissions. Although BAPGM is not ideal for the growth of all bacteria, the recent microbiological use of BAPGM provides an example of how the use of a nonconventional medium can facilitate the identification of unusual and infrequently isolated bacteria from sick patients.


This research was supported in part by a grant from the American College of Veterinary Internal Medicine Foundation, Bayer Animal Health, and by the State of North Carolina. The authors thank veterinarians from throughout the United States who requested testing of the dogs reported in this study and Tonya Lee for editorial assistance.

Conflict of Interest Declaration: Authors disclose no conflict of interest.


  1. 1

    BD Diagnostics, Sparks, MD

  2. 2

    VITEK; Biomerieux, Marcy l'Etoile, France

  3. 3

    Qiagen; GmbH, Hilden, Germany

  4. 4

    Eppendorf AG, Hamburg, Germany

  5. 5

    National Library of Medicine, Bethesda, MD

  6. 6

    Vector NTS suite 10; InformMax Inc, Fredrick, MD