The problem of Mycobacterium abscessus complex: multi-drug resistance, bacteriophage susceptibility and potential healthcare transmission

Abstract

Objectives: Mycobacterium abscessus complex is responsible for 2.6-13.0% of all non-tuberculous mycobacterial pulmonary infections and these are notoriously difficult to treat due to the complex regimens required, drug resistance and adverse effects. Hence, bacteriophages have been considered in clinical practice as an additional treatment option. Here, we evaluated antibiotic and phage susceptibility profiles of M. abscessus clinical isolates. Whole-genome sequencing (WGS) revealed the phylogenetic relationships, dominant circulating clones (DCCs), the likelihood of patient-to-patient transmission and the presence of prophages.

Methods: Antibiotic susceptibility testing was performed using CLSI breakpoints (n = 95), and plaque assays were used for phage susceptibility testing (subset of n = 88, 35 rough and 53 smooth morphology). WGS was completed using the Illumina platform and analysed using Snippy/snp-dists and Discovery and Extraction of Phages Tool (DEPhT).

Results: Amikacin and Tigecycline were the most active drugs (with 2 strains resistant to amikacin, and one strain with Tigecycline MIC of 4 μg/mL). Most strains were resistant to all other drugs tested, with Linezolid and Imipenem showing the least resistance, at 38% (36/95) and 55% (52/95), respectively. Rough colony morphotype strains were more phage-susceptible than smooth strains (77%-27/35 versus 48%-25/53 in the plaque assays, but smooth strains are not killed efficiently by those phages in liquid infection assay). We have also identified 100 resident prophages, some of which were propagated lytically. DCC1 (20%-18/90) and DCC4 (22%-20/90) were observed to be the major clones and WGS identified 6 events of possible patient-to-patient transmission.

Discussion: Many strains of M. abscessus complex are intrinsically resistant to available antibiotics and bacteriophages represent an alternative therapeutic option, but only for strains with rough morphology. Further studies are needed to elucidate the role of hospital-borne M. abscessus transmission.

Keywords: Bacteriophages; Multi-drug resistance; Mycobacterium abscessus complex; Next-generation sequencing; Prophages.

Fig. 1.

Phylogeny and phage characteristics of M. abscessus strains. The phylogeny of 90 M. abscessus strains with reference strains is shown and each is assigned to dominant circulating clones (DCC1–DDC7) as defined previously [7]. Strains chosen to represent previously characterized dominant circulating clones are the following: ATCC19977 for DCC1, GD54 for DCC2, G220 and JHN_AB_0023_1 for DCC3, 976 for DCC4, 1100 for DCC5, A47 for DCC6 and FLAC047 for DCC7 as performed in previous studies [24]. Strain subspecies designations are shown (massiliense, bolletii, abscessus). The colony morphotype of each strain is indicated (R = rough, S = smooth), together with their susceptibilities to infection by 7 M. smegmatis phages and 4 lytically grown phages derived from M. abscessus prophages and described previously [17] (coloured circles). Filled circles indicate either no phage infection (black) or an efficiency of plating relative to M. smegmatis of greater than 10⁻². Some strains did not propagate well and their susceptibilities are not shown. The prophage content of each strain is shown by coloured squares with filled squares noting the presence of a prophage grouped within clusters (MabA, MabB, etc., or singleton) as indicated above. The AMI and CLA resistance profiles are indicated at the extreme right. Purple indicates resistance to Amikacin while green indicates resistance to Clarithromycin. Light purple and light green indicate intermediate resistance to Amikacin and Clarithromycin, respectively. AMI, Amikacin; CLA, Clarithromycin; M. abscessus, Mycobacterium abscessus; M. smegmatis, Mycobacterium smegmatis.

Fig. 2.

Summary of M. abscessus phage susceptibilities. The proportions of rough morphotype (a) or smooth morphotype strains (b) susceptible to either one or two more phages, or resistant to all phages tested are shown. The susceptibilities for just the 7 M. smegmatis phages tested (orange) or all 11 phages shown in Table 2 (blue) are shown. M. abscessus, Mycobacterium abscessus; M. smegmatis, Mycobacterium smegmatis.

Fig. 3.

Characterization of M. abscessus prophages. (a) Gene content distance tree of identified prophages generated by phamnexus (available on GitHub at https://github.com/chg60/phamnexus) and the UPGMA algorithm available in Splitstree 4. The tree was illustrated with FigTree, where branches are labelled and coloured according to their designated clusters, similar to Fig. 1. ProphiT9875-1 shares little gene content with any previously identified phages although the genome shares some similarity to MabD cluster phages. (b) Illustration of the regions within prophages of T1615 and T11960, which encode a type II restriction-modification system. (c) Example solid media plates of screens for lytically growing spontaneously released phages. Lawns of M. abscessus strains T7193B and T9907 are shown growing on solid media, and 96 culture supernatants were spotted onto the lawns at positions indicated by the grids. Several supernatants show lytic infection (red circles), which were then purified and amplified. Strain T9907 is more typical in showing sensitivity to few or no supernatants; T719B is susceptible to a relatively high proportion of released phages and is also sensitive to five of the M. smegmatis phages (Fig. 1). (d) Examples of strain sensitivities to lytically growing prophages. Phage lysates were diluted and spotted onto lawns of M. abscessus GD123 and GD254, which illustrate the variations in phage infection profiles. Phage phiGD02 is a previously isolated control phage. M. abscessus, Mycobacterium abscessus; M. smegmatis, Mycobacterium smegmatis.