Legionella quinlivanii strain isolated from a human: A case report and whole genome sequencing analysis

Abstract

We describe a strain of Legionella quinlivanii isolated from a bronchoalveolar lavage specimen from an 83-year-old patient in the province of Québec. Identification was done using 16S rRNA sequencing. The strain could replicate efficiently in human THP-1 macrophages and maintained a low level of cytotoxicity. Upon analyzing the whole genome sequencing data, the icm/dot secretion system was present, but the strain lacked some effector genes known to express proteins toxic to cells. The pathogenicity of this Legionella species should be investigated further.

Legionella quinlivanii was first described in 1989 after isolation from water in the evaporative air conditioning system of a bus in Australia (1). Since its discovery, this bacterium has been recovered from environmental water in the United Kingdom and Greece, and potting soil in Switzerland (2–4). Clinical cases of pneumonia attributed to this species have not been reported. Here, we report the isolation of L. quinlivanii from a bronchoalveolar lavage (BAL) specimen from a patient in the province of Québec, Canada.

Case Presentation

In 2015, an 83-year-old woman was assessed for a recent history of mild hemoptysis and increased sputum production. She had a history of severe but stable chronic obstructive pulmonary disease (Gold III). She also had a history of breast cancer treated with chemotherapy and radiotherapy in 2006. Her daily medication included inhaled fluticasone/salmeterol combination, albuterol, and tiotropium. She also took levothyroxine, simvastatin, paroxetine, hydrochlorothiazide, and nifedipine daily. She had not required antibiotics or systemic corticosteroids treatment in the past year. On physical examination, the patient was afebrile, and lung auscultation was unremarkable. No significant exposure in her environment with a risk of Legionella acquisition could be identified, and she had no recent travel history. A chest X-ray and subsequent chest computed tomography (CT) scan revealed three lung nodules (two in the upper left lobe and one in the lower right lobe) and bronchiectasis. Of note, a chest X-ray from 2013 did not reveal lung nodules. Bronchoscopy with bronchoalveolar lavage was performed a few days later and grew Legionella sp showing moderate growth (estimate of 10³ UFC/mL), along with Moraxella catarrhalis and Aspergillus fumigatus. Neither Legionella urinary antigen detection nor nucleic acid amplification test detection on BAL was performed. The patient was treated with azithromycin for 2 weeks without any clinical or radiological improvement. Further diagnostic work-up, including a PET-scan and a lung biopsy, later revealed that the patient had stage IV squamous cell lung carcinoma.

The isolate of Legionella sp was sent to the provincial reference laboratory (Laboratoire de santé publique du Québec [LSPQ]) for definitive identification. The bacteria were identified as L. quinlivanii (identifier ID143958) using 16S rRNA sequencing (5). The strain could not grow on BCYE agar without L-cysteine. The identification was further confirmed by the National Microbiology Laboratory in Winnipeg by both 16s rRNA and mip gene sequencing (6).

Whole genome sequencing was performed on the strain (accession number PRJNA375793). Genomic DNA was purified with the MagAttract DNA Mini M48 Kit using the BioRobot M48 (Qiagen, Toronto, ON). A sequencing library was prepared using the Nextera XT DNA Library Preparation Kit (Illumina, Inc., San Diego, CA). Sequencing was performed on an Illumina MiSeq platform with the MiSeq Reagent Kit v3 (600 cycles, paired ends). The quality of the raw sequence data was checked using FastQC (Babraham Bioinformatics, http://www.bioinformatics.babraham.ac.uk/projects/fastqc). The sequence reads were de novo assembled into contigs using SPAdes v3.8.0 (Center for Algorithmic Biotechnology, St. Petersburg State University, St. Petersburg, RU, http://cab.spbu.ru/software/spades/) and filtered using an in-house Python script in order to keep contigs with coverage of >5× and length of >1,000 bp. A total of 23 contigs, ranging in size from 1,284 to 497,485 bp, with an N50 of 340,063 bp (for a total length of 3,684,187 bp with an average G+C content of 43%) was generated. The mean depth of coverage was 63×. Contigs were annotated using the NCBI Prokaryotic Genome Automatic Annotation Pipeline v2.10 (PGAAP) (7). A total of 3,092 protein-coding sequences, 40 tRNAs, 5 rRNAs (5S, 16S, and 23S), and 31 pseudogenes were predicted. The icm/dot secretion system region I of the isolate is located on contig 8, ranging from positions 33,122 to 41,357, and region II is on contig 7 between positions 81,830 to 104,137. Region I and region II of ID143958 strain share 95% and 94% homology, respectively, with the L. quinlivanii reference strain ATCC 43830 and have the same gene organization (8).

We evaluated the intracellular multiplication in human cultured macrophages (THP-1) and cytotoxicity toward THP-1 cells, as previously described (9–11). The L. pneumophila JR32 strain, derived from Philadelphia-1 and the isogenic icm/dot deficient dotA mutant, was used as a positive and negative control, respectively, for both assays. As shown in Figure 1A, the L. quinlivanii strain replicated efficiently (at the level of the positive control) in human THP-1 macrophages, compared with the negative control. As shown in Figure 1B, our strain maintained a low level of cytotoxicity, similar to the dotA mutant control.

Figure 1:

Interaction of Legionella quinlivanii isolate ID143958 with macrophages

Human macrophage cell line THP-1 cells were infected with L. pneumophila JR32 (black line, positive control) and the isogenic, intracellular growth deficient dotA mutant (green line, negative control), and the L. quinlivanii isolate ID143958 (red line). Intracellular growth was measured by CFU counts (A) after infection at MOI of 0.1. Cytotoxicity (B) was evaluated with the MTT assays. Cells were infected with a range of MOIs (0.1–1,000), and the assay was performed 72 hours post-infection.

CFU = Colony-forming units; MOI = Multiplicity of infection

Discussion

Since its first identification in 1976, more than 60 species of Legionella have been identified, about 25 of which have been associated with human disease (12). The ability of the remaining species to cause disease is currently unknown. Diseases caused by this opportunistic pathogen include the pneumonic form, Legionnaires’ disease, and the flu-like form, Pontiac fever. The conditions under which an individual develops either legionnaires’ disease or Pontiac fever are not fully understood but may depend on the health status of the individual, the degree of exposure to the organism, and/or the strain-specific virulence (9).

L. quinlivanii isolates have been reported only from environmental sources in some parts of the world. During an outbreak investigation in Adelaide, Australia, L. quinlivanii was identified in 2.7% of the 632 Legionella isolates recovered from a variety of sources of water samples during 1986–87 (13). In 1991, Birtles et al proposed a second serogroup type for two strains of L. quinlivanii isolated in the United Kingdom from two unrelated environmental sources (3). L. quinlivanii isolate was also recovered from commercial potting soil in Switzerland during an investigation in 2006–07 (4). Finally, in Greece, during standard hotel cooling tower system management, an isolate of L. quinlivanii was recovered in 2009 (2). To date, there is only one report of detection of antibody titres from a patient in a study in France (14). In their study, Berger et al analyzed 210 patients with pneumonia in an intensive care unit in order to elucidate the role of amoeba-associated microorganisms as etiologic agents of pneumonia. They reported one patient who seroconverted for L. quinlivanii. Despite this report, the pathogenicity of L. quinlivanii is still considered unknown, and no human isolation was reported before this case (12).

Lawrence et al recently demonstrated that an environmental strain of L. quinlivanii has the ability to infect human macrophages (9). We demonstrated the same capability for the human strain in this study, underlying its ability to replicate in human cells. We also performed a cytotoxicity assay, and our strain did not demonstrate a major effect on THP-1 cells. The only strain analyzed by Lawrence et al demonstrated a low level of cytotoxicity, and they concluded that no infection remained by 72 hours of analysis (9). These findings could explain, in part, the fact that the patient did not have any clear symptoms associated with the presence of the L. quinlivanii strain in her lungs. However, cytotoxicity toward macrophages does not capture the entire complexity of this phenotype in the lungs, where the immune system’s fight also causes damage (15). This patient had bronchiectasis and, therefore, could have been colonized with the bacteria. The icm/dot deficient dotA mutant is known to be avirulent, failing to evade fusion with the lysosome (16). As our strain is able to replicate into human cells and possess the icm/dot system, the strain could lack essential effectors that can cause cytotoxicity to the cells. Some effector genes known to express proteins that are toxic for cells are absent of our strain genome, such as the Sid family (17,18), AnkX (19), Lgt1, Lgt2, VpDA, and Lpg0645 (20).

We report here a strain of L. quinlivanii isolated from human BAL. The pathogenicity of this Legionella species should be investigated further. The availability of whole genome sequence data could improve the knowledge about this emerging Legionella species.

Funding Statement

This study was conducted with Laboratoire de santé publique du Québec (LSPQ) internal funding. Dr Faucher is holding a team research project grant (188813) from Fonds de recherche du Québec–Nature et technologies, and a Discovery Grant from Natural Sciences and Engineering Research Council of Canada (418289-2012).

Competing Interests:

Dr Leduc reports grants and personal fees from Biomérieux, outside the submitted work.

Ethics Approval:

N/A

Informed Consent:

N/A

Registry and the Registration No. of the Study/Trial:

N/A

Animal Studies:

N/A

Funding:

This study was conducted with Laboratoire de santé publique du Québec (LSPQ) internal funding. Dr Faucher is holding a team research project grant (188813) from Fonds de recherche du Québec–Nature et technologies, and a Discovery Grant from Natural Sciences and Engineering Research Council of Canada (418289-2012).

Peer Review:

This article has been peer reviewed.