Whole Genome Sequencing of Nontuberculous Mycobacterium (NTM) Isolates from Sputum and Environmental Specimens in Co-Habiting Patients with NTM Pulmonary Disease CURRENT STATUS:

Background : Nontuberculous mycobacterium (NTM) species are ubiquitous microorganisms. NTM pulmonary disease (NTM-PD) is caused not by human-to-human transmission but by independent environmental acquisition. However, recent studies using next-generation sequencing (NGS) have reported trans-continental spread of Mycobacterium abscessus among patients with cystic fibrosis. Results : We investigated NTM genomes through NGS to examine transmission patterns in three pairs of co-habiting NTM-PD patients who were suspected of patient-to-patient transmission. Three pairs of patients with NTM-PD co-habiting for at least 15 years were enrolled: a mother and a daughter with M. avium PD, a couple with M. intracellulare PD, and a second couple, one of whom was infected with M. intracellulare PD and the other of whom was infected with M. abscessus subsp. massiliense PD. Whole genome sequencing was performed using NTM colonies isolated from patients and environmental specimens. Genetic distances were estimated based on single nucleotide polymorphisms (SNPs) in the NTM genomes. Comparing SNPs in the consensus regions, the minimum pairwise SNP distances of NTM isolates derived from the two pairs of patients infected with the same NTM species were over 10,000. In phylogenetic analysis, the NTM isolates from patients with M. avium PD clustered with isolates from different environmental sources. Conclusions : In conclusion, considering the genetic distances between NTM strains, the likelihood of patient-to-patient transmission in pairs of co-habiting NTM-PD patients without overt immune deficiency is minimal.


Background
The prevalence of nontuberculous mycobacterial pulmonary disease (NTM-PD) is increasing in developed countries (1)(2)(3)(4). Several explanations for this epidemiological change have been proposed, including awareness and improved detection of NTM-PD, increased populations with risk factors such as bronchiectasis or use of immunosuppressants, and disinfection of drinking water in urban areas resulting in selective advantages for NTM (1).
The development of next-generation sequencing (NGS) technology enables the identification of massive numbers of single nucleotide polymorphisms (SNPs) by whole genome sequencing (WGS).
Since NTM is a ubiquitous microorganism, it is generally assumed that all patients with NTM-PD acquire NTM from their environment, not from other infected individuals. However, recent NGS studies showed that this might not be the case, at least for patients with cystic fibrosis (8,9). Bryant and colleagues collected Mycobacterium abscessus isolates from patients with cystic fibrosis, performed WGS, and analysed phylogenetic relationships among these isolates. Surprisingly, they found strong evidence supporting human-to-human transmission of M. abscessus among patients with cystic fibrosis, and identified some M. abscessus isolates that were widespread globally (9). Although this finding remains controversial, these reports have raised concerns that NTM might be transmitted not only among immunosuppressed individuals such as cystic fibrosis patients but also among immunocompetent individuals (10). This would be especially important for hospital infection control, since isolation practices for NTM-PD patients without cystic fibrosis are not as strict as those for pulmonary tuberculosis patients generally.
Recently, we diagnosed three pairs of NTM-PD patients with no immunodeficiency who had been cohabiting for at least 15 years. We investigated the genomes of NTM isolates derived from the patients and environmental samples in their houses to understand the source of infection using WGS.

Patient Characteristics
Three pairs of patients with NTM-PD who had been co-habiting for at least 15 years were enrolled (Table 1). A mother and a daughter (Patients A and B) with M. avium PD had lived in an apartment in an urban area (HOME-1) for 15 years. A couple (Patients C and D) with M. intracellulare PD had lived in a house in a rural area (HOME-2) for 30 years. A second couple (Patients E and F) with M.
intracellulare PD and M. abscessus subsp. massiliense PD had lived in an apartment in an urban area (HOME-3) for 30 years. No patients were suspected of any immunodeficiency disorders. They were HIV-negative, were not taking any immunosuppressants and had no history of recurrent infection of any organs.
NTM Isolation and Sequencing 4 Among 12 environmental specimens from HOME-1, seven specimens from either the kitchen or the bathroom were culture-positive. Subsequently, 18 morphologically distinct isolates were purified (Supplementary Table 1, Additional File 1). However, none of the 15 environmental specimens from HOME-2 yielded any NTM isolates and only one of 15 specimens from HOME-3 yielded NTM isolates.  avium and used for further analysis. Pairwise SNP distances between every pair of isolates at those positions were calculated and clusters on the histogram of SNP distances were observed ( Figure 1).
The SNP distance between the isolates from Patient A and B was 15,513. By contrast, the SNP distances among all replicates (between the isolates from patient A and the kitchen in HOME-1, and between the isolates from patient B and the bathroom) were less than 150 and these isolates clustered together closely in the phylogenetic tree ( Figure 1). Using high-confidence SNPs from 14 M.
avium subsp. avium isolates, phylogenetic analysis was performed ( Figure 2). The isolate from Patient A and its replicates clustered with the specimens from the kitchen (scale on surface of kitchen faucet), while the isolate from Patient B clustered with the isolates from the bathroom (hot water from bathroom faucet, hot water from showerhead) and the kitchen (cold water from kitchen faucet).
In HOME-2, the SNP distance between the M. intracellulare isolates from Patients C and D was 25,136, even larger than the SNP distances from Patient E in HOME-3. The SNP distance between In our study, a phylogenetic tree based on WGS was consistent with these previous reports and demonstrated the environmental sources of two M. avium-PD patients living together (HOME-1).
Patient A seemed to have acquired NTM from the kitchen, while Patient B acquired NTM from the bathroom or kitchen.
Although we thoroughly collected environmental specimens from water and biofilms on the faucets and showerhead, we could not culture any NTMs from HOME-2, and identified only one NTM from HOME-3. A previous study reported that NTM species could be isolated from 22 of 37 (59%) households of NTM-PD patients. In addition, a positive correlation was observed between the number of samples collected per house and the number of NTM-positive samples (15). We collected 15 water samples from each of the two houses; this number is equal to or larger than that of most studies, suggesting that there may be sources of NTM-PD other than water. The patients living in HOME-2 (Patient C and D) were farmers and their home was a high soil environment that included a yard and hay ( Table 1). Given that previous studies reported that a majority of M. intracellulare was isolated from soil samples, especially in high soil environments (14,16,17), the source of M. intracellulare for these patients could be soil instead of water.

7
In HOME-3, three morphologically distinct colonies were cultured from one environmental specimen and sequenced separately. However, we failed to identify the species because no isolates matched to NTM reference genomes. Considering that the sequencing reads of NTM isolates from patients E and F aligned to NTM reference genomes, the environmental specimen isolates might not be closely related to NTM in HOME-3. A previous study showed that, among 19 patients with M. avium-PD or M.
intracellulare-PD living in low soil environments, no patients had genetically identical isolates compared with soil sample isolates from their houses (17). Our results suggested that a patient with M. intracellulare-PD (Patient E) in HOME-3 (low soil) may have been exposed to the NTM outside the house. The lag time between the acquisition of NTM from the environment and diagnosis of NTM-PD also makes it more difficult to identify the source of NTM (14).

Patient Enrolment
Pairs of adult patients living together and diagnosed with NTM-PD at Seoul National University Hospital satisfying the following inclusion criteria were enrolled: age >18 years; typical respiratory symptoms such as chronic cough, sputum, or haemoptysis; findings suggestive of NTM-PD on computed tomography; identification of NTM in ≥2 sputum cultures or in ≥1 bronchoscopic specimen; living in the same household prior to diagnosis with NTM-PD; and consented to collection of environmental samples in their home. This study was approved by an institutional review board (IRB Number: 1804-064-936) and registered at ClinicalTrial.gov (NCT03532438).

8
The most recently-isolated NTM from participant sputum or bronchial wash specimens from participants was collected. In addition, we visited patients' homes to collect environmental specimens. One litre of water was aseptically collected into sterile containers directly from all faucets and showers in bathrooms and kitchens. Swabs were taken from the inside of faucets and showerheads in bathrooms and kitchens. As previously described (14,18), each water specimen was passed through a 0.45 µm filter. Filters were rinsed with 10 mL of sterile distilled water and transferred to Middlebrook 7H11 plates and Mycobacterial Growth Indicator Tubes (MGITs). Swabs were washed in sterile distilled water and then processed in the same manner as the water samples.
All culture-positive MGITs were transferred to new Middlebrook 7H11 plates. If culture-positive Middlebrook 7H11 plates had more than two morphologically distinct colonies, each distinct colony was separately transferred to a 3% Ogawa media plate, incubated, and purified as a single colony. If the colonies were homogenous and discrete on Middlebrook 7H11 plates, we collected the isolates without a further purification step.

DNA Preparation and Sequencing
All biomass from Middlebrook 7H11 plates or Ogawa media plates taken from sweeps of colonies was mixed with 425-600 µm glass beads, vortexed for 5 min, incubated at 80°C for 10 min and then centrifuged. DNA was extracted from the supernatant using a QIAamp DNA mini kit (Qiagen inc, Hilden, Germany) as previously described (8  Supplementary Files