Abstract
Lactobacilli are part of the microbiota and are also used as probiotics. However, in recent years they have been associated with invasive infections, especially bacteremia. Lactobacillus spp. are usually susceptible to penicillins, macrolides, and carbapenems, but Lactobacillus rhamnosus is intrinsically resistant to glycopeptides. The aim of this study was to determine the antimicrobial susceptibility profile and resistance mechanism of a clinical isolate of L. rhamnosus isolated from 10 sets of blood cultures of the same patient. The isolate was identified by Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry (Bruker Daltonics; BD, Bremen, Germany) and 16S rRNA gene sequencing. In vitro susceptibilities to penicillin, ampicillin, imipenem, vancomycin, erythromycin, clindamycin, and linezolid were determined with gradient test strips (bioMérieux, France) on Mueller-Hinton agar plates supplemented with 5% defibrinated horse blood and 20 mg/L β-NAD. The isolate was resistant to vancomycin and imipenem. Polymerase chain reaction test was positive for blaOXA-48 and the presence of this carbapenemase was confirmed by gene sequencing. Although plasmid analysis suggested that the blaOXA-48 is chromosomal in this isolate, it is still an alarming finding for potential transmission of antibiotic resistance genes to other bacteria in the gut. To our knowledge, this is the first report of the presence of blaOXA-48 in a Lactobacillus spp. and has utmost importance as these bacteria are used as probiotics. The isolation of these bacteria from sterile body sites should not go unnoticed.
Introduction
Lactobacillus species are Gram-positive, facultatively anaerobic or microaerophilic nonspore-forming, acid-tolerant, and catalase-negative rods, which are part of the human microbiota (Goldstein et al., 2015) They have been utilized in the production of dairy products such as cheese, yoghurt, and more recently, probiotics (Korhonen et al., 2010; Goldstein et al., 2015). They may also cause various infections in patients with certain underlying conditions, such as malignancies or gastrointestinal disorders (Magnus, 2003; Cannon et al., 2005; Salminen et al., 2006; Korhonen et al., 2010; Goldstein et al., 2015) Lactobacillus rhamnosus and Lactobacillus casei have been reported as the most common clinical isolates (Goldstein et al., 2015). L. rhamnosus is intrinsically resistant to glycopeptides and is generally susceptible to penicillins and carbapenems (Salminen et al., 2006; Korhonen et al., 2010; Goldstein et al., 2015). In vitro studies on the antimicrobial susceptibility of Lactobacillus spp. are scarce, despite the fact that these organisms are gaining importance as clinical pathogens. Moreover, their use in the food industry and probiotics might not be as safe due to this anticipated character (Georgieva et al., 2015). Several studies on Lactobacillus spp. isolated from poultry, dairy products, and probiotics show that antibiotic resistance genes are carried by these organisms, which is a disturbing finding as gastrointestinal tract may act as a reservoir for antibiotic resistance genes (Schjorring and Krogfelt, 2011; Georgieva et al., 2015; Dec et al., 2017; Li et al., 2019). The presence of resistance genes for aminoglycosides, tetracyclines, chloramphenicol, and macrolides have been reported in Lactobacillus spp. isolated from food and humans, but the genes responsible for carbapenem resistance have not been determined yet (Danielsen et al., 2007; Klare et al., 2007; Gueimonde et al., 2013; Dec et al., 2017; Stsepetova et al., 2017; Li et al., 2019). In the present study, we investigated in vitro antibiotic susceptibility of a clinical L. rhamnosus isolate and found an enzyme responsible for resistance to imipenem.
Materials and Methods
Identification
A Gram-positive rod was isolated from 10 sets of blood cultures of a febrile patient considered to have a probable diagnosis of endocarditis, hospitalized in the cardiovascular ward. Blood cultures (BD BACTEC™ FX; Becton Dickinson) were sent to the central microbiology laboratory in a time span of 8 days from July 26, 2018 to August 2, 2018. Each blood set consisted of one aerobic and one anaerobic bottle. Six sets (12 bottles) of blood cultures were sent to the central microbiology laboratory on July 27, 2018 and 4 sets (8 bottles) of blood cultures were sent to the central microbiology laboratory on August 2, 2018.
Isolates were subcultured on blood, MacConkey, and chocolate agar and aerobically incubated at 35°C for 48 h. All colonies were identified as Lactobacillus spp. by their colony morphology, gram staining, and catalase tests. Ten isolates were further identified by Matrix-Assisted Laser Desorption/Ionization-Time-of-Flight Mass Spectrometry (MALDI-TOF) (Bruker Daltonics; BD, Bremen, Germany).
Molecular identification
Bacterial DNA was isolated by commercial kits (Qiagen) and the DNA quantity and quality measurements were done using spectrophotometry (NanoDrop). Molecular identification was done by amplifying the 16S rRNA gene region that is around 1500 bp using primers F-CCGAATTCGTCGACAACAGAGTTTGATCCTGGCTCAG and R-CCCGGGATCCAAGCTTACGGCTACCTTGTTACGACTT and polymerase chain reaction (PCR) amplicons were sequenced in pair-end fashion through ABI 3500 sequencer. The sequences were assigned to the closest species by BLASTn assignment against NCBI 16S rRNA database.
Antimicrobial susceptibility tests
In vitro susceptibilities to penicillin, ampicillin, imipenem, vancomycin, erythromycin, clindamycin, and linezolid were determined with gradient test strips (bioMérieux, France) on Mueller-Hinton agar plates supplemented with 5% defibrinated horse blood and 20 mg/L β-NAD (MH-F; Becton Dickinson) (EUCAST, 2018). Streptococcus pneumoniae ATCC 49619 strain was included as the quality control. Antibiotic susceptibility test was repeated three times. As all 10 isolates had exactly the same pattern of antimicrobial susceptibility, molecular investigations were carried out with one of the isolates.
Investigation of carbapenemase genes
PCR was conducted to determine the following carbapenem resistance genes: KPC-type, NDM-1, IMP-type VIM-type, SPM-type, AIM-type, and OXA-48 (Poirela et al., 2011). To validate the existence of blaOXA-48, the amplicons were pair-end-sequenced using the ABI 3500 system and the sequences were aligned against NCBI nr (nonredundant) protein database by BLASTX (version 2.8.1).
To investigate whether blaOXA-48 genes are harbored in the previously reported whole genome sequencing assemblies, the carbapenemase blaOXA-48 sequence was searched in silico in the NCBI RefSeq database (Klindworth et al., 2013). One hundred fifty-eight whole-genome assemblies that were the total collection of L. rhamnosus assemblies published were downloaded from the RefSeq database. The blaOXA-48 amino acid sequence gathered from uniprot database was searched against the L. rhamnosus assemblies using tblastn (version 2.8.1) program with protein/DNA translation alignment parameters (substitution matrix: BLOSUM62, gap costs: existence-11, extension-1).
Two different plasmid isolation kits (Thermo Fisher PureLink and Qiagen QIAprep Spin Miniprep) were separately used to reveal the existence of plasmid in the isolates. Twenty microliters of DNA extracts from each isolation were subject to size separation by electrophoresis in 0.8% agarose gel (Sigma Aldrich) with 1 × Tris-Boric acid–EDTA buffer at 110 V for 2 h followed by 0.7 μg/mL ethidium bromide staining. Supercoiled DNA ladder (Invitrogen) was used for the plasmid length determination.
Results
Bacterial identification
According to the conventional biochemical tests and MALDI-TOF analysis, all 10 isolates were identified as L. rhamnosus and this result was further confirmed with 16S rRNA gene sequencing with an alignment identity score of 99%.
Antimicrobial susceptibility test results
Applying the gradient strip method, the following minimum inhibitory concentration (MIC) values were obtained: vancomycin: 256 mg/L, penicillin: 0.75 mg/L, ampicillin: 1.5 mg/L, imipenem: 4 mg/L, erythromycin: 0.032 mg/L, clindamycin: 0.125 mg/L, and linezolid: 0.5 mg/L. There are not any published clinical breakpoints or epidemiological cutoffs established for L. rhamnosus in EUCAST standards (EUCAST, 2018). Breakpoints for some antibiotics against Lactobacillus spp. can be found in Clinical and Laboratory Standards Institute (CLSI) documents. According to CLSI, vancomycin ≤2 mg/L, ampicillin and penicillin ≤8 mg/L, imipenem ≤0.5 mg/L, erythromycin ≤0.5 mg/L, clindamycin ≤0.5 mg/L, and linezolid ≤4 mg/L are categorized as susceptible. According to these breakpoints, our isolate is resistant to vancomycin and imipenem.
Molecular tests
Molecular analysis of the seven different genes responsible for carbapenem resistance revealed that, the isolate was positive only for blaOXA-48. Aligning the sequenced amplicons against the NCBI database provided an in silico support to this claim.
The in silico investigation of blaOXA-48 gene in the available whole-genome sequencing assemblies of L. rhamnosus strains showed that no previously sequenced genomes harbored blaOXA-48 gene homologs. This search includes the plasmids carried by the investigated strains as the assemblies contain both chromosomal and plasmid DNA.
The procedure followed to determine the presence of plasmids revealed that the L. rhamnosus strain does not contain any plasmids.
Discussion
The awareness of the clinical significance of L. rhamnosus is increasing, and Lactobacillus species isolated from sterile body parts are being reported more frequently (Z'Graggen et al., 2005; Chazan et al., 2008; Robin et al., 2010; Naqvi et al., 2018). These bacteria were ignored as contaminants in the past but their significance is now more evident (Magnus, 2003; Salminen et al., 2004). Although Lactobacilli are considered to be a part of the human microbiota and some of them are employed as probiotics, their safety is debatable as they have been reported in various infections especially bacteremia and endocarditis and their detection from blood cultures are clinically significant (Magnus, 2003; Salminen et al., 2004, 2006; Cannon et al., 2005). Isolation of the same L. rhamnosus isolate from 10 sets of blood cultures of the same patient prompted us to investigate it in more detail.
Preliminary identification was performed by colony morphology, Gram staining, negative catalase tests, and MALDI-TOF identifying the isolate as L. rhamnosus. This identification was confirmed by the molecular analysis of the 16S rRNA gene (Stackebrandt and Goebel, 1994). L. rhamnosus is in the L. casei group being closely related to L. casei and Lactobacillus paracasei phenotypically and genotypically. This might cause difficulties in the identification at species level, even with the 16S rRNA gene sequencing, which is considered to be the current convention (Huang et al., 2018). We employed MALDI-TOF analysis in parallel with sequencing.
Antimicrobial susceptibility tests for the isolate were performed on MH-F medium as suggested for the fastidious organisms by EUCAST (2018). However, clinical breakpoints for Lactobacillus spp. do not exist in EUCAST standards, therefore, the results were categorized according to CLSI criteria (CLSI, 2016). An MIC value of 256 mg/L for vancomycin in our isolate is in accordance with other reports on L. rhamnosus indicating intrinsic resistance to glycopeptides (Magnus, 2003; Salminen et al., 2006). However, MIC of imipenem was observed to be quite high and the strain was categorized as resistant according to CLSI (2016). There are several in vitro studies on the susceptibility of L. rhamnosus to carbapenems; Arpi et al. (2003) reported a clinical L. rhamnosus isolate, which had an MIC of >32 mg/L for imipenem; Danielsen et al. (2007) reported blood culture L. rhamnosus isolates, which had MICs of 4 to >32 mg/L. Vanichanan et al. (2016) reported carbapenem-resistant Lactobacillus spp. with an MIC >32 mg/L from a renal transplant recipient with a history of probiotic consumption. The mechanism of carbapenem resistance has not been investigated in those studies. Korhonen et al. (2010) investigated several β-lactamase genes in L. rhamnosus isolates, including blaKPC and blaVIM with microarray hybridizations, but did not detect any of them.
We investigated bla genes for KPC, NDM-1, IMP-type, VIM-type, SPM-type, AIM-type, and OXA-48 using PCR. Results for blaOXA-48 were positive, and this finding was confirmed by gene sequencing followed by sequence alignments. In addition, considering the possibility of acquiring this gene by any recent horizontal transfer events that could be mediated by plasmids, the presence of plasmids were investigated and no plasmids carried by the strain were detected. Therefore, it is likely that blaOXA-48 gene is located in the chromosome of the isolate. However, further investigations are required to strengthen the evidence for location of the detected blaOXA-48 gene in this isolate.
OXA-48 is a molecular Class D carbapenemase and is the most prevalent carbapenemase in Gram-negative isolates in Turkey (Albiger et al., 2015; Grundmann et al., 2017). However, Gram-positive clinical isolates harboring blaOXA-48 is a very significant finding, as blaOXA-48 has not been reported in any Gram-positive organisms at present. Toth et al. (2016) reported and analyzed Class D β-lactamases from various Gram-positive organisms isolated from soil. They point out that Class D enzymes can be present in gram-positive organisms but they have a different substrate-binding mode than Gram-negative bacteria (Toth et al., 2016). This may explain our isolate being susceptible to penicillin and ampicillin but resistant to imipenem. The molecular basis of this hypothesis is yet to be defined. A weakness of the current study is the lack of a detailed plasmid isolation and characterization. Therefore, advanced molecular methodologies should be adopted to reveal the existence or nonexistence of blaOXA-48-harboring plasmids.
Conclusions
To our knowledge, this is the first report on the presence of blaOXA-48 carbapenemase in a Lactobacillus spp. Although our results suggest that its genetic location is in the chromosome, its potential to be transferred to other gut bacteria should be kept in mind. We did not investigate the original reservoir for this gene in all isolates included, however, as OXA-48 carbapenemases are the most frequent enzymes causing carbapenem resistance in Enterobacteriaceae in Turkey, our future aim is to investigate various Lactobacillus spp. from humans as well as probiotics and perform detailed studies on the activity of the enzyme.
Footnotes
Disclosure Statement
No competing financial interests exist.