Infectious Disease Reports 2019; volume 11:8132

Janibacter with most reliable diagnostic tool. However, evidence of genomic sometimes a diagnosis cannot be made Correspondence: Ying Bai, Division of using blood culture because of poor labora- Vector-Borne Diseases, Centers for Disease polymorphism isolated tory infrastructure, presence of uncommon Control and Prevention, 3156 Rampart Road, from resected heart valve pathogens, or pretreatment of the patient Fort Collins, Colorado 80521 USA. in a patient with aortic stenosis with antibiotics prior to sample collection. Tel. 1-970-266-3555. Up to 35% of all infective endocarditis E-mail: [email protected] cases remain blood culture negative either Lile Malania,1 Ying Bai,2 Key words: aortic stenosis; ; genome 3 4 due to slow growth of the bacteria or inap- polymorphism; . Kamil Khanipov, Marika Tsereteli, 2 5 1 propriate media. Mikheil Metreveli, David Tsereteli, Calcific aortic stenosis is also a major 1 1 Acknowledgements: The authors would like Ketevan Sidamonidze, Paata Imnadze, problem in many countries. In the United to thank Natalia Abazashvili, Tamriko 3 2 Yuriy Fofanov, Michael Kosoy States, the incidence rate of aortic stenosis Giorgadze, and Nazibrola Chitadze for their 1National Center for Disease Control has been reported as 27.1 per 10,000.3 The assistance with laboratory tests; Guram and Public Health, Tbilisi, Georgia; mechanism of calcification remains unclear. Katstadze, and Neli Chakvetadze for their par- ticipation in investigation of the human case; 2Division of Vector-Borne Diseases, The hypothesis is that low grade chronic or recurrent bacterial endocarditis with specif- Levent Albayrak and George Golovko for the Centers for Disease Control and assistance with genome analysis; and ic calcifiable bacteria is a cause of calcifica- Prevention, Fort Collins, CO, USA; 4 Christina Nelson for careful revision of the 3Department of Pharmacology and tion of the aortic valves. Along those lines, manuscript, thoughtful comments, and con- an association between Chlamydia pneumo- Toxicology, University of Texas Medical structive suggestions. nia and Mycoplasma pneumonia and calci- Branch, Galveston, TX, USA; 4Tbilisi fication in aortic stenosis has been Contributions: LM, YB, and KK contributed State Medical University, Tbilisi, reported.5,6 More recently, a case of equally. LM, MK, conception and design; LM, Georgia; 5Department of Cardiology, Bartonella endocarditis on a bioprosthetic MT, MM, DT, data collecting; LM, KS, bacte- High Technology Medical Center, aortic valve that caused rapidly progressive riologicalonly analysis; YB, KS, sequence analy- University Clinic, Tbilisi, Georgia aortic stenosis was reported.7 ses; KH, YF, genome analysis; LM, YB, MK, Due to recent descriptions of Bartonella manuscript writing; MK, PI, manuscript reviewing and final approval. infections in the country of Georgia8 and the importance of identifying sources of bacte- Abstract useConflict of interests: the authors declare no The authors report isolation and identi- rial endocarditis, a collaboration was estab- potential conflict of interests. fication of two strains of bacteria belonging lished between hospitals in Tbilisi (the cap- to the genus Janibacter from a human ital of Georgia) and the National Center for Funding: the work was supported by the patient with aortic stenosis from a rural area Disease Control and Public Health International Science and Technology Center Project (ISTC G-1683-p). of the country of Georgia. The microorgan- (NCDC&PH), where culturing of fastidious isms were isolated from aortic heart valve. microorganisms from endocarditis cases Received for publication: 1 April 2019. Two isolates with slightly distinct colony has been performed. In this paper, we report isolation and identification of two bacterial Revision received: 10 July 2019. morphologies were harvested after sub-cul- Accepted for publication: 11 July 2019. turing from an original agar plate. strains, both belonging to the genus Preliminary identification of the isolates is Janibacter, from heart valve tissues collect- This work is licensed under a Creative based on amplification and sequencing of a ed during an investigation of a human case, Commons Attribution-NonCommercial 4.0 fragment of 16SrRNA. Whole genome which originally was suspected as endo- International License (CC BY-NC 4.0). sequencing was performed using the carditis, but later diagnosed as a case of aor- ©Copyright: the Author(s), 2019 Illumina MiSeq instrument. Both isolates tic stenosis. Detailed characterization of the Licensee PAGEPress, Italy were identified as undistinguished strains of full genomes of the both isolates was per- formed with identification of a significant Infectious Disease Reports 2019; 11:8132 the genus Janibacter. Characterization of doi:10.4081/idr.2019.8132 whole genome sequences of eachNon-commercial culture difference in gene composition between has revealed a 15% difference in gene pro- cultures of these two Janibacter. file between the cultures and confirmed that both strains belong to the genus Janibacter blood were used to inoculate chocolate agar with the closest match to J. terrae. Genomic plates. The inoculated plates were incubated comparison of cultures of Janibacter Materials and Methods at 35oC with 5% carbon dioxide for 30 days. obtained from human cases and from envi- Bacterial culture Each individual colony from agar was sep- ronmental sources presents a promising Surgically removed cardiac valve mate- arately streaked onto secondary agar plates. direction for evaluating a role of these bac- rial and whole blood were collected from a The subcultures were harvested separately teria as human pathogens. patient with suspected endocarditis. The by scraping bacterial growth after sufficient clinical part of the work was approved by growth was observed. DNA was heat the Ethical Committee of the High extracted at 95oC for 10 min from whole Technology Medical Center, Tbilisi, bacterial cells. The prepared DNA was Introduction Georgia. The material was frozen, trans- stored at -80oC. Bacteria are a common source of infec- ferred to the National Center for Diseases tive endocarditis with inflammation of the Control and Public Health, Tbilisi, Georgia, PCR amplification and sequencing endothelia cells and valves of the heart, usu- and stored at -80oC until it was plated. The Amplification of the 16S rRNA frag- ally caused by bacteria.1 Blood culture is the homogenized cardiac tissue and whole ments from isolated DNA of both cultures

[page 22] [Infectious Disease Reports 2019; 11:8132] Article was performed by conventional broad- sion, the patient’s fevers became more per- slightly different morphology were range PCR assay using 63F and 1187R sistent and she was evaluated in an outpa- observed on the agar plate inoculated with primers in a C1000 Touch Thermal Cycler tient clinic. She was prescribed empiric heart tissues. Both colonies had circular (Bio-Rad, Hercules, CA). The PCR prod- antibiotics of unknown type, but did not shape and light cream color with a slight ucts were analyzed for the presence of improve. She was admitted to the intensive difference in size. No hemolysis on choco- amplicons of the appropriate size by elec- care unit in May 2014 due to periodic late agar was observed. These two colonies trophoresis in 1.5% agar gels containing squeezing pain in the chest that radiated were separately sub-cultured onto new agar GelGreen stain (Biotium, Hayward, CA). under the left shoulder blade with the origi- plates. Microscopy indicated Gram-positive Positive and negative controls were includ- nal diagnosis of endocarditis. The patient coccobacilli. When bacterial mass on both ed in each PCR assay. PCR products with lived in a rural area near the town of Kareli plates was sufficient, colonies were harvest- appropriate sizes of amplicons were puri- in the region of Shida Kartli, Georgia ed on brain heart infusion medium and fied using a QIAquick PCR Purification Kit (South Caucasus). She was engaged in agri- stored at –70oC until the strains were (Qiagen, Valencia, CA) according to manu- cultural work and reported frequent con- shipped to CDC-Fort Collins for additional facturer’s instructions, then sequenced in tacts with farm animals (pigs and cattle) and characterization. both directions using ABI 3130 Genetic pets at home. By Mega BLAST search, the ribosomal Analyzer (Applied Biosystems, Foster City, Clinical signs upon hospital admission amplicons from both isolates were identi- o CA). included fever of 38 C, ischemic disease, fied as a Janibacter sp. based on the frag- angina, and arterial hypertension. No rash ment amplification and sequence analysis Genome analysis or lymphadenopathy was reported. with the highest match to . Whole genome sequencing (WGS) was Coronary angiographyshowed unaltered The 16S rRNA sequences from both iso- performed using the Illumina MiSeq instru- coronary vessels. Echocardiography detect- lates were undistinguished. ment. Sequencing of the two samples result- ed critical stenosis of the aortic valve with Sequencing of each isolate genome ed in 8 million 51 nucleotide long sequenc- high gradient and critical opening. On aus- resulted in 2 and 6 million sequencing reads ing reads. The sequencing reads were fil- cultation there was rough systolic murmur 51 nucleotides long in size trimmed to 32 tered to exclude: (a) nucleotides below the at every auditory point, with epicenter on basesonly long non-overlapping subsequences quality threshold of 0.05 (using modified the aorta. Heart rate – 76 beats/min, pulse – with a minimum threshold quality of 10. To Richard Mott algorithm); (b) unknown arrhythmic, average filling and tension. T/A determine the presence and identity of – 130/60 mmHg. The patient’s WBC was microbial contaminants, sequencing data nucleotides; and (c) sequencing library 3 adapters. The reads were then assembled 5.60 10 /uL (normal range: 3.4 – 9.6 was tested against a collection of 13,774 3 use allowing for up to 5% mismatches using 10 /uL), neutrophil count 42% (normal complete bacterial, viral, plasmid, phage, CLC Genomics Workbench De Novo range: 40% – 60%), and lymphocyte count and fungal reference genome sequences, Assembly tool (http://www.clcbio.com) and 48% (normal range: 20% – 40%), platelets which included all reference genomes avail- – 193 103/uL(normal range: 157 – 371 able from the NCBI Reference Sequence the original sequenced reads were mapped 3 back to the assembled contigs (Table 1). 10 /uLfor females); and ESR – 10 mm/h (collection as of November 2013). Rapid Annotation using Subsystem (normal range: 1 – 20 mm/h for females). family had the high- Chest X-ray and abdominal ultrasound did Technology tool kit (RASTtk) was used to est number of identified genes from the not show any abnormalities. The case was database. A more detailed mapping to every cut open reading frames from the assembled 9 diagnosed as aortic stenosis, and the patient genome entry in NCBI under the contigs on the PATRIC web interface. underwent heart valve replacement. As Intrasporangiaceae family was performed. Annotated features were clustered together standard post-surgical prophylaxis, the For references, which did not contain a fully in protein space using CD-HIT with the fol- patient was treated with Ceftriaxone 1 gram assembled genome, the contigs were con- lowing parameters: 70% global identity and two times per day after the surgery. catenated into one long sequence that was 70% alignment coverage.10 Figure 1 shows No bacterial growth was observed on treated as a genome. Lowest Common a Venn diagram of the feature clustering agar plates inoculated with blood of the Ancestry analysis of full length 16S between the two sequenced isolates and the patient. On the seventh day after inoculation sequences assembled from whole genome closest reference genome (Janibacter terrae of agar plates with homogenized cardiac tis- sequencing reads showed a 99.9% match to NBRC 107854 = JCM 12887).Non-commercial Blastn was sues two small bacterial colonies were the genus Janibacter.13 The closest match used to compare the features annotated with observed. Two slow-growing colonies with by genome analysis was to Janibacter ter- the closest reference, Janibacter terrae NBRC 107854 = JCM 12887 (GCA_001598955.1).11 The results were visualized using Blast Ring Image 12 Table 1. Assembly statistics genomes of two strains of Janibacter (TB1 and TB2) Generator (Figure 2). Strain TB1 Strain TB2 N75 20,309 22,383 Results N50 35,574 48,911 N25 56,767 84,095 A 60-year-old previously healthy Minimum Contig Length 205 210 female experienced symptoms of headache, dizziness and syncope episodes, general Maximum Contig Length 118,653 165,607 weakness, exertional fatigue, and shortness Average Contig Length 15,154 22,752 of breath for 2 years. She also reported low- Number of Contigs 219 142 grade episodic fevers in the preceding 6 Total Bases Assembled 3,318,728 3,230,833 months. One month prior to hospital admis-

[Infectious Disease Reports 2019; 11:8132] [page 23] Article rae.14 The similarity between the both 2005 at least nine cases of human infections bacteria as a potential source of infection; strains and J. terrae was calculated as a associated with these microorganisms have however, we wish to keep a conservative Jaccard distance measured by proportion been reported. Among these cases, one case approach for a validation of this claim. between a number of shared genes by aver- was associated with J. melonis,16 one case Cardiac valve tissue is considered sterile; age number of genes between two com- with J. hoylei,17 two cases with an unde- nevertheless, though a contamination of pared genomes. According to the calcula- scribed species of Janibacter,18,19 and five cardiac tissue during the surgery or cultur- tion, the strain TB1 was 84.4% similar to J. cases with J. terrae.20,21 The cases associat- ing procedure is very unlikely, we cannot terrae, while the strain TB2 was 87.9% sim- ed with J. terrae are especially interesting rule out such a possibility. More important- ilar to J. terrae. Comparison of the genomes since the isolates that we report here from ly, we cannot prove a causal connection of TB1 and TB2 has revealed some poly- the cardiac tissues were closely related by between this bacterium and pathogenic morphic differences. The gene clustering the genomic analysis to this species of manifestations in the patient’s heart because results confirmed an approximately 15% Janibacter. Fernandez- Natal et al.20 report- of very limited information available about difference in the gene composition between ed cases of infection by J. terrae in four these bacteria. Therefore, an etiological role the two stains (Figure 1). febrile patients with severe underlying con- of this bacterium as a source of aortic steno- ditions in Spain. All four patients were sis requires further investigations. treated with antibiotics, two of them with Nevertheless, the presented findings sug- favorable outcome, and two severely gest that clinicians should be aware of a Discussion immunocompromised patients died. More possible risk of infection by bacteria of the 21 Two discoveries can be highlighted recently, Wan et al. reported a case of genus Janibacter with the ability to cause from the reported results: 1) both bacterial bilateral psoas abscess caused by J. terrae. invasive cardiac disease. cultures isolated from a human patient with Although this is the first report of bacte- Having limited data on potential risk aortic stenosis were identified as a species ria of the genus Janibacter isolated from a factors, it is hard to speculate about routes of genus Janibacter that has never been pre- cardiac valve, these microorganisms were of transmission of the bacteria to the viously reported as a source of a cardiac dis- broadly described as “coryneform” bacteria patient. The patient’s report of frequent ease in humans or animals; and 2) complete with some representatives being reported as exposureonly to farm animals, however, may 22 genome comparison indicated genomic dif- sources of endocarditis. Coryneform bac- suggest a possible zoonotic source of the ference between bacterial cultures obtained teria, defined as non-sporing Gram-positive infection. In Germany, Janibacter DNA from the same cardiac tissues. rods with an irregular outline, were histori- was identified from a biofilter treating air The report of the isolation of two close- cally dismissed as a part of the normaluseemissions from a livestock facility.24 ly related bacterial strains belonging to the human skin flora when recovered from The study has demonstrated additional genus of Janibacter from the case of aortic human patients, but have increasingly been perspectives based on the ability of the stenosis is novel not only for this region, but implemented as causes of significant infec- computational analysis of the whole bacter- presents the first known case worldwide. tions in clinical practice.23 ial genomes to identify and characterize fas- The bacteria of genus Janibacter were orig- Identification of a Janibacter strain in tidious microorganisms discovered in clini- inally found in the environment,15 but since cardiac tissue can lead to evaluation of this cal specimens. Although 16SrRNA analysis

Non-commercial

Figure 1. Venn diagram comparison of gene composition Figure 2. Ring image displaying the similarity of the two isolates between strains TB1 and TB2 cultured from valve tissues of a from valve tissues of a patient with aortic stenosis with Janibacter patient with aortic stenosis, and Janibacter terrae. terrae (strain NBRC 107854 = JCM 12887).

[page 24] [Infectious Disease Reports 2019; 11:8132] Article identified Janibacter species in the culture fibrotic and calcified areas of degenera- diaminopimelic acid in the cell wall. Int colonies, the whole genome sequencing tive aortic stenosis. Int J Cardiol 2006; J Syst Bacteriol 1997;47:529-34. allowed to more precisely differentiate 108:43-7. 16. Elsayed S, Zhang K. Bacteremia caused these strains from other Janibacter species 7. Schnitzer K, Or Z, Sawaed S, et al. by Janibacter melonis. J Clin Microbiol and to evaluate differences between Rapidly Progressive Bioprosthetic aor- 2005;43:3537-9. colonies. The latter finding highlighted the tic valve stenosis due to Bartonella 17. Lim YK, Kweon OJ, Kim HR, et al. fact that the bacterial population occupying species endocarditis. Ann Thorac Surg First case of bacteremia caused by cardiac tissues from one patient was poly- 2017;104:57-9. Janibacter hoylei. APMIS 2017;125: morphic. Genomic comparison of cultures 8. Kandelaki G, Malania L, Bai Y, et al. 665-8. of Janibacter obtained from human cases Human lymphadenopathy caused by 18. Loubinoux J, Rio B, Mihaila L, et al. and from environmental sources presents a ratborne Bartonella, Tbilisi, Georgia. Bacteremia caused by an undescribed promising direction for evaluating a role of Emerg Inf Dis 2016;22:544-6. species of Janibacter. J Clin Microbiol these bacteria as human pathogens. 9. Wattam AR, Abraham D, Dalay O, et al. 2005;43:3564-6. PATRIC, the bacterial bioinformatics 19. Worodria W, Anderson J, Cattamanchi database and analysis resource. Nucl A, et al. The role of speciation in posi- Acids Res 2014;42:581-91. tive Lowenstein-Jensen culture isolates References 10. Fu L, Niu B, Zhu Z, et al. CD-HIT: from a high tuberculosis burden coun- 1. Brouqui P, Raoult D. Endocarditis due accelerated for clustering the next-gen- try. PLoS One 2011;6:e27017. to rare and fastidious bacteria. Clin eration sequencing data. Bioinformatics 20. Fernández-Natal MI, Sáez-Nieto JA, Microbiol Rev 2001;14:177-207. 2012;28:3150-2. Medina-Pascual MJ, et al. First report 2. Subedia S, Jenningsa Z, Chen SC. 11. Camacho C, Coulouris G, Avagyan V, et of bacteremia by Janibacter terrae in Laboratory approach to the diagnosis of al. BLAST+: architecture and applica- humans. Infection 2015;43:103-6. culture-negative infective endocarditis. tions. BMC Bioinformatics 2018;10: 21. Wan WY, Mughal A, Ward K. Bilateral Heart Lung Circulation 2017;8:763-71. 421. psoas abscess caused by Janibacter ter- 3. Gersony WM, Hayes CJ, Driscoll DJ, et 12. Alikhan NF, Petty NK, Ben Zakour NL, onlyrae, an unusual condition and organism al. Bacterial endocarditis in patients et al. BLAST Ring Image Generator resulting in laboratory and management with aortic stenosis, pulmonary steno- (BRIG): simple genome conundrums. Acta Clin Belg 2016;19:1- sis, or ventricular septal defect. comparisons. BMC Genomics 2011;12: 4. Circulation 1993;87:1121-6. 402. use 22. Graevenitz A. Importance of 4. Cohen DJ, Malave D, Ghidoni JJ, et al. 13. Pruesse E, Peplies J, Glöckner FO. Coryneform Bacteria in Infective Role of oral bacterial flora in calcific SINA: accurate high-throughput multi- Endocarditis. Infect Dis Rep 2015;7: aortic stenosis: an animal model. Ann ple sequence alignment of ribosomal 6103. Thorac Surg 2004;77:537-43. RNA genes. Bioinformatics 2012;28: 23. Bernard K. 2012. The genus 5. Higuchi ML, Higuchi-Dos-Santos MH, 1823-9. Corynebacterium and other medically Pierri H, et al. Mycoplasma pneumoni- 14. Yoon JH, Lee KC, Kang SS, et al. relevant coryneform-like bacteria. J ae and Chlamydia pneumoniae in calci- Janibacter terrae sp. nov., a bacterium Clin Microbiol 2012;50:3152-8. fied nodules of aortic stenotic valves. isolated from soil around a wastewater 24. Kristiansen A, Lindholst S, Feilberg A, Rev Inst Med Trop Sao Paulo 2002;44: treatment plant. Int J Syst Evol et al. Butyric acid- and dimethyl disul- 209-12. Microbiol 2000;50:1821-7. fide-assimilating microorganisms in a 6. Pierri H, Higuchi-dos-Santos MH, 15. Martin K, Schumann P, Rainey FA, et biofilter treating air emissions from a Higuchi Mde L, et al. Density of al. Janibacter limosus gen. nov., sp. livestock facility. Appl Environ Chlamydia pneumoniae is increased in nov., a new actinomycete with meso- Microbiol 2011;77:8595-604. Non-commercial

[Infectious Disease Reports 2019; 11:8132] [page 25]