Phenotypic and molecular characterisation of a novel species, Mycobacterium hubeiense sp., isolated from the sputum of a patient with secondary tuberculosis in Hubei of China.

A new fast-growing mycobacterium, designated strain QGD101T, was isolated from the sputum of an 84-year-old man suspected of tuberculosis in Wuhan Medical Treatment Center, Hubei, China. This strain was a gram-staining-negative, aerobic, non-spore-forming and catalase-positive bacterium, which was further identified as the NTM by PNB and TCH tests. The moxifloxacin and levofloxacin exhibited strong suppressing function against QGD101T with MIC values of 0.06 and 0.125 µg/ml after drug susceptibility testing of six main antimicrobial agents on mycobacteria. Based on the sequence analysis of 16S rRNA, rpoB, hsp65 and 16S-23S rRNA internal transcribed spacer, the strain QGD101T could not be identified to a species level. Mycobacterium moriokaense ATCC43059T that shared the highest 16S rRNA gene sequence similarity (98%) with strain QGD101T was actually different in genomes average nucleotide identity (78.74%). In addition, the major cellular fatty acids of QGD101T were determined as C18:1ω9c, C16:0 and C18:2ω6c. The DNA G + C content was 64.9% measured by high performance liquid chromatography. Therefore, the phenotypic and genotypic characterisation of this strain led us to the conclusion that it represents a novel species of mycobacteria, for which the name Mycobacterium hubeiense sp. nov. (type strain QGD101T = CCTCCAA 2017003T = KCTC39927T) was proposed. Thus, the results of this study are very significant for the clinical diagnosis of tuberculosis and future personalised medicine.

Introduction

Non-tuberculous mycobacteria (NTM) are a large family of acid-fast bacteria, which can be opportunistic pathogens and cause pulmonary disease resembling tuberculosis, lymphadenitis, skin disease or disseminated disease [1-4]. Until now, more than 160 NTM species have been found in List of Prokaryotic names with Standing in Nomenclature (http://www.bacterio.net/mycobacterium.html). Members of NTM are able to be distinguished from photochromogens, scotochromogens and nonchromogens. NTM can also be classified based on the rate of growth [5-7]. NTM are widely distributed in the environment, particularly in wet soil, marshland, streams, rivers and estuaries [8, 9].

In clinical practice, NTM can cause the TB-like clinical presentations. Importantly, most NTM (except M. kansasii) are inherently resistant to or only partially susceptible to the standard anti-tubercular drugs. Thus, the diagnosis of NTM infection is critical for choosing effective treatment [10]. In this study, a clinical strain QGD101T was isolated from the sputum sample of a patient suspected of tuberculosis during July 2012 in Wuhan Medical Treatment Center, Hubei, China (30°67′N 114°29′E). By the image findings and medical history, the case was primarily diagnosed as secondary suspected tuberculosis. The patient had the pathological characteristics such as pulmonary fibrosis cavity, hemoptysis, etc., and the commonly used anti-tuberculosis drugs isoniazid and rifampicin had no obvious effect on conventional treatment. The pathogen culture for mycobacteria was positive. By the use of molecular identification and evolutionary analysis, this mycobacterium was identified as a new species. The results of this study have a great significance for clinical diagnosis of emerging diseases and the replenishment of new mycobacterial resources.

Materials and methods

Ethics statement

Ethical approval is granted by the Ethics Committee of Wuhan Polytechnic University (ID Number: 20120720006). The sputum sample was obtained from a patient suspected of tuberculosis in Wuhan Medical Treatment Center in Hubei Province, on 20 July 2012.

Microbiological analysis

The early morning sputum specimens were collected from an 84-year-old man diagnosed as secondary tuberculosis and were processed with the standard protocol [11]. After decontamination, each sample was cultured onto Löwenstein–Jensen (L–J) media at 37 °C for 8 weeks. The cultures were inspected weekly and growth was examined by visual inspection for colonies. Negative slides were then confirmed by Ziehl–Neelsen staining.

Biochemical and physiological-testing

Biochemical and physiological-testing, including iron uptake, urease testing, nitrate reduction and pyrazinamidase testing [12], were performed for QGD101T, Mycobacterium barrassiae CIP 108545T and Mycobacterium moriokaense ATCC 43059T. Catalase activity was determined using bubble production H2O2 solution [13].. Other biochemical testing was carried out using VITEK 2 ANC and VITEK 2 GP testing kits according to the protocols of the manufacturer (bioMerieux). Growth characteristics at 28 °C, 37 °C and 42 °C were respectively assessed after 5 days of incubation in L–J medium.

The drug susceptibility testing (DST) for QGD101T

According to the resazurin microplate assay [14], the commonly used antimicrobial agents such as rifampin (RIF), ethambutol (EMB), streptomycin, levofloxacin (LVX), kanamycin (KAN), azithromycin (AZI) and moxifloxacin (MOX) were respectively used to determine the MIC value of strain QGD101T.

HPLC analysis

For quantitative analysis of the cellular fatty acid composition, strain QGD101T was cultured at 37 °C for 5 days. The DNA G + C content was determined by means of reversed-phase high performance liquid chromatography, which was according to the protocol described by Mesbah et al. [15].

DNA extraction, PCR amplification and sequencing

Genomic DNA of M. tuberculosis isolate QGD101T was extracted by the classical phenol–chloroform method and stored at −20 °C. Oligonucleotide primers were designed using Primer 5.0 (PREMIER Biosoft, Palo Alto, CA, USA) and Oligo 6 software (Molecular Biology Insights, Inc., Cascade, CO, USA). Amplification of the 16S rRNA gene and 16S-23S rRNA internal transcribed spacer (ITS) sequence was performed as previously described [10]. The hsp65 gene was amplified using primers Tb11 (5′-ACCAACGATGGTGTGTCCAT-3′) and Tb12 (5′-CTTGTCGAACCGCATACCCT-3′) targeting positions 396–836 of the gene sequence of M. tuberculosis [16]. The same experimental protocol was used for amplification with the exception that annealing was performed at 57 °C for 1 min. Amplification of rpoB was done with primers MycoF (5′-GGCAAGGTCACCCCGAAGGG-3′; base positions 2573 to 2592) and MycoR (5′-AGCGGCTGCTGGGTGATCATC-3′; base positions 3316 to 3337) in two conserved regions flanking the most variable rpoB region [17]. Direct sequencing was performed on amplified products with the 3730xl DNA Analyser. Similarity values and percent divergence of the 16S rRNA gene were calculated using MegAlign of Lasergene (DNA-Star, Madison, WI, USA).

Sequence analysis and phylogenetic classification

Molecular typing and species identification were performed using BLAST search (http://www.ncbi.nlm.nih.gov/BLAST/). Multiple-sequence alignment was performed in MEGA version 7.0.18. Sequences were then trimmed to start and finish at the same nucleotide position for phylogenetic analysis. The sequences of the 16S rRNA, rpoB, hsp65 genes and ITS sequence of the isolated strain were compared with available sequences from GenBank using the BLAST program to determine their approximate phylogenetic affiliation. The 16S rRNA phylogenetic trees based on mycobacterial species were constructed using neighbour-joining algorithm following Kimura 2-parameter correction and maximum likelihood methods in MEGA 7.0.18. The significance of the branching order was estimated by the bootstrap method based on 1000 replications.

Genomic analysis

The complete genome sequence of strain QGD101T was determined using the Illumina Miseq. The reads were assembled de novo using SOAPdenovo version 2.04 and the gaps of the assembled sequence were closed using GapCloser version 1.12, and the average nucleotide identity (ANI) value was calculated using JSpeciesWS [18].

Results

Genotype identification analysis

The strain QGD101T shares 99% identity with M. barrassiae CIP108545T and M. moriokaense ATCC43059T in the 16S rRNA gene. It shares 96% identity in the rpoB gene with M. barrassiae CIP108545 and M. moriokaense CIP105393. The ITS sequence analysis showed that this strain shares 85% identity with M. chubuense NBB4. The hsp65 gene sequence analysis also showed that this strain shares 92% identity with M. moriokaense.

Results of rpoB gene identification are more approximate to the 16S rRNA gene phylogenetic classification. The neighbour-joining tree (Fig. 1) indicated that strain QGD101T was a member of the genus Mycobacterium and formed a subcluster with M. barrassiae CIP108545T and M. moriokaense ATCC43059T with a 63% bootstrap value. Also, the rpoB NJ tree showed that QGD101T was closely related to M. barrassiae CIP108545T and M. moriokaense CIP 105393T, both with an 84% bootstrap value (Fig. 2).

Fig. 1.

Neighbour-joining trees based on 16S rRNA gene sequences of strain QGD101T and some type strains. The tree was rooted with N. farcinica ATCC3318T. Percentages indicated at nodes represent bootstrap levels supported by 1000 resampled datasets; values <50% are not shown. The significance of the branching order was estimated by the bootstrap method calculated 1000 replications. Bar, 0.01 substitutions per nucleotide position.

Fig. 2.

Neighbour-joining trees based on rpoB sequences of strain QGD101T and some type strains. Percentages indicated at nodes represent bootstrap levels supported by 1000 resampled datasets; values <50% are not shown. The significance of the branching order was estimated by the bootstrap method calculated 1000 replications. Bar, 0.1 substitutions per nucleotide position.

A GenBank accession number for the 16S rRNA, hsp65 and rpoB gene sequence and 16S-23S rRNA intergenic spacer sequence of strain QGD101T are KR995241, KT007146, KT007149 and KR995223, respectively.

Comparative analysis of genomic sequences

The genomic analysis of QGD101T was obtained by genome sequencing, and the whole sequence of QGD101T has been upload on Genbank. The GC content in the QGD101T gene region is 66.99%, while GC content in the intergenetic region is 62.49%. To evaluate the similarity between genome sequences, ANI values were analysed between the strain and reported genomes of closest species (Fig. 3). The strain QGD101T showed 78.75% ANI values compared to the closest reference strain M. moriokaense ATCC43059T (NZ_MVIB01000100). These DNA relatedness and ANI results indicate that strain QGD101T represents a novel genomic species that is distinct from its closest relatives.

Fig. 3.

Correlation plot based on QGD101T and reference strains ANI values.

This Whole Genome Shotgun project has been deposited at DDBJ/ENA/GenBank under the accession PIZU00000000. The version described in this paper is version PIZU01000000.

Phenotypic identification analysis

The phenotypic and biochemical characteristics of strain QGD101T are presented in Table 1. The strain QGD101T can grow on media containing either TCH or PNB, indicated that strain QGD101T is a fast-growing NTM.

Table 1.

VITEK 2 GP and VITEK 2 ANC test kits results

VITEK 2 GP test	M. hubeiense	VITEK 2 ANC test	M. hubeiense
D-amygdalin	−	ELLMAN	−
Phosphatidylinositol Phospholipase C	−	Phenylalanine arylamidase	+
D-xylose	−	D-cellobiose	−
Arginine Dihydrolase 1	−	Arbutin	−
β-Galactosidase	−	5-Bromo-4-chloro-3-indoxyl-β-glucoside	−
α-glucosidase	−	5-Bromo-4-chloro-3-indoxyl-β-glucuronide	−
Ala-Phe-Pro arylamidase	−	β-Galactopyranose indoxyl	−
Cyclodextrin	−	α-Arabinosidase	−
L-aspartate arylamidase	−	5-Bromo-4-chloro-3-indoxyl-α-galactoside	−
β-Galactopyranose	−	β-Mannosidase	−
α-Mannosidase	−	Arginine GP	−
Phosphatase	−	Pyruvate	(–)
Leucine arylamidase	+	Maltotriose	−
L-proline arylamidase	−	Esculin hydrolyse	−
β-Glucuronidase	−	β-D-fucosidase	−
α-Galactosidase	−	5-Bromo-4-chloro-3-indoxyl-β-N-acetyl-glucosamide	−
L-pyrrolidonyl-arylamidase	−	5-Bromo-4-chloro-3-indoxyl-α-mannoside	−
β-Glucuronidase	−	α-L-fucosidase	−
Alanine arylamidase	+	L-arabinose	−
Tyrosine arylamidase	+	d-Ribose 2	−
D-sorbitol	−	Phenylphosphonate	−
Urease	−	α-L-arabinofuranosidase	−
Polymixin B resistance	−	GRAM	+
D-galactose	−	MORPH	−
D-ribose	−	AERO	+
L-lactate alkalinisation	−
Lactose	−
N-acetyl-D-glucosamine	−
D-maltose	−
Bacitracin resistance	−
Novobiocin resistance	−
6.5% NaCl	−
D-mannitol	−
D-mannose	−
Methyl-B-D-glucopyranoside	−
Pullulan	−
D-raffinose	−
O/129 resistance	−
Salicin	−
Sucrose	−
D-trehalose	+
Arginine dihydrolase 2	−
Optochin resistance	−

+, positive; −, negative.

The differential characteristics between strain QGD101T and its reference species in the genus Mycobacterium are shown in Table 2. The strain QGD101T could also grow on sodium glutamate glucose agar but not on MacConkey agar. M. barrassiae, M. moriokaense and strain QGD101T were all able to grow at 42 °C, however, none of them could produce pigment. All three strains showed positive activities for urease, 5% NaCl tolerance, iron uptake, nitrate reductase and negative activities for arylsulfatase, akaline phosphatase and esculin. There are also different characteristics among the three species. For example, M. barrassiae and M. moriokaense, but not strain QGD101T, showed positive activities for the pyrrolidonyl arylamidase test. For the α-glucosidase test, only M. moriokaense showed positive activities. For their carbon source, M. barrassiae was able to assimilate both d-sorbitol and l-arabinose except d-mannitol. M. barrassiae was able to use l-arabinose only. However, none of the three chemicals could be used by strain QGD101T as its sore carbon source. These differences clearly demonstrated that the strain QGD101T is different from M. moriokaense and M. barrassiae, other biochemical characteristics of strain QGD101T are shown in Table 2.

Table 2.

Differential characteristics of strain QGD101T and closely related species

Characteristic	M. barrassiae	M. moriokaense	Strain QGD101T
Growth at 42 °C	+	+	+
Pigmentation	－	－	－
Arylsulfatase	－	－	－
Catalase	+	+	+
Urease	+	+	+
5% NaCl tolerance	+	+	+
Iron uptake	+	+	+
Nitrate reductase	+	+	+
Pyrrolidonyl arylamidase	±	±	－
Alkaline phosphatase	－	－	－
β-Galactosidase	－	+	－
α-Glucosidase	－	+	－
Esculin	－	－	－
Utilisation of the following as a sole carbon source：
D-mannitol	－	－	－
D-sorbitol	+	－	－
L-arabinose	+	+	－

+, positive; −, negative.

Note: Stains: M. barrassiae CIP108545T, M. moriokaense ATCC43059T. All data from the present study.

The major fatty acids (>10% of the total fatty acids) disclosed by gas-liquid chromatography analysis detected in strain QGD101T were C18:1ω9c (36.82%), C16:0 (29.9%) and C18:2ω6,9c (11.31%), for which the Sherlock Microbial Identification System (MIS) showed no match with QGD101T (Fig. 4). And the DNA G + C content value of strain QGD101T was 64.9% (65.25% ± 0.35%), which was consistent with the GC values sequenced from the genomic analysis.

Fig. 4.

Fatty acid chromatogram of QGD101T.

Drug resistance of strain QGD101T

The DST of strain QGD101T showed that the MIC values of different drugs were: RFB (0.5 µg/ml), EMB (0.5 µg/ml), MOX (0.06 µg/ml), LVX (0.125 µg/ml), RIF (8 µg/ml), KAN (8 µg/ml) and AZI (128 µg/ml). According to the drug susceptibility criteria in the Susceptibility Test of Mycobacteria, Nocardia and Other Aerobic Actinomycetes [18], this result indicated that strain QGD101T was sensitive to drugs RFB, EMB, MOX and LVX, but resistant to RIF, KAN and AZI.

Discussion

In view of the development of environment and climate change, more mycobacteria are discovered and the membership list of genus Mycobacterium is ever expanding [19]. Therefore, discovering novel species of NTM and identification of clinical isolates are of great significance for the clinical practice. Especially, accurate differentiation of NTM from the M. tuberculosis complex is needed for effective patient management, because many NTM strains are resistant to the anti-mycobacterial agents [20]. This study suggested that the differential identification of mycobacteria is very significant to initiate personal medication in clinical practice, especially for the TB suspect.

In this study, a mycobacterium was isolated from a suspected TB patient. In order to accurately identify this bacteria, biochemical analysis, molecular sequence analysis, G + C content determination and cell wall fatty acid composition were performed. In most cases, 16S rRNA gene sequence can provide species-specific signature sequences useful for the identification of bacteria [21]. Further, the combination of multiple genes, 16S rRNA, hsp65 and rpoB analysis is preferred [10 22 23]. The rpoB gene encodes the β-subunit of RNA polymerase containing conserved sequence regions flanking highly variable regions. It has been developed as a suitable tool for the accurate taxonomic identification of Mycobacterium [22]. The hsp65 gene is a highly conserved housekeeping gene coding for a 65 kDa protein containing epitopes that are unique as well as epitopes that are common to various species of mycobacteria [24].

The M. moriokaense group was composed of M. barrassiae and M. moriokaense, sharing 99% identity in the 16S rRNA gene [25]. The strain QGD101T shares 99% identity and 98% query cover with M. barrassiae CIP108545T and M. moriokaense ATCC43059T in the 16S rRNA gene. The phylogenetic tree of 16S rRNA suggested the close relationship of strain QGD101T to M. moriokaense and M. barrassiae. By genomic analysis, QGD101T showed 78.75% ANI values relative to the closest reference strain M. moriokaense ATCC43059T (NZ_MVIB01000100). These DNA relatedness and ANI results indicate that strain QGD101T represents a novel genomic species that is distinct from its closest relatives.

The strain QGD101T could grow on media containing either TCH or PNB indicating that QGD101T is NTM. The differential characteristics between strain QGD101T and the reference species of Mycobacterium were obvious, especially the ability to assimilate D-sorbitol, L-arabinose and the activity to β-galactosidase, α-glucosidase, suggesting that the strain QGD101T represented a new species in the genus Mycobacterium.

To identify the drug resistance of this species, we chose MOX for patients. After 1 week of medication, the patient stopped coughing up blood, imaging examination showed that the lung shadow was weakened. After 2 months of treatment, the symptoms disappeared completely. This indicates that MOX can effectively inhibit the growth of the pathogen (QGD101T) in the patient, providing hope for the recovery of other patients infected with the strain.

In general, the physiological, biochemical characteristics and the genotype analysis demonstrated the clear differences between QGD101T and the type strains of its closest phylogenetic neighbours. Therefore, the genome analysis indicated that strain QGD101T represented a novel genomic species, and for which the name Mycobacterium hubeiense sp. nov. (type strain QGD101T = CCTCCAA 2017003T = KCTC 39927T) was proposed.