Mab_3083c Is a Homologue of RNase J and Plays a Role in Colony Morphotype, Aggregation, and Sliding Motility of Mycobacterium abscessus

Mycobacterium abscessus is an opportunistic pathogen causing human diseases, especially in immunocompromised patients. M. abscessus strains with a rough morphotype are more virulent than those with a smooth morphotype. Morphotype switch may occur during a clinical infection. To investigate the genes involved in colony morphotype switching, we performed transposon mutagenesis in a rough clinical strain of M. abscessus. A morphotype switching mutant (smooth) named mab_3083c::Tn was obtained. This mutant was found to have a lower aggregative ability and a higher sliding motility than the wild type strain. However, its glycopeptidolipid (GPL) content remained the same as those of the wild type. Complementation of the mutant with a functional mab_3083c gene reverted its morphotype back to rough, indicating that mab_3083c is associated with colony morphology of M. abscessus. Bioinformatic analyses showed that mab_3083c has a 75.4% identity in amino acid sequence with the well-characterized ribonuclease J (RNase J) of M. smegmatis (RNase JMsmeg). Complementation of the mutant with the RNase J gene of M. smegmatis also switched its colony morphology from smooth back to rough. These results suggest that Mab_3083c is a homologue of RNase J and involved in regulating M. abscessus colony morphotype switching.


Introduction
Mycobacterium abscessus is a rapid-growing nontuberculous Mycobacterium (NTM) and is often present in soil and water [1,2]. It has been found to infect lung, skin, and soft tissue, especially in immunocompromised patients [3][4][5][6] and is the second most common bacterial species isolated from cystic fibrosis patients [7,8]. The prevalence of M. abscessus infection is increasing in hospitals in Taiwan [9,10]. Because M. abscessus is highly resistant to multiple antibiotics, treatment of M. abscessus infections is often difficult [11,12]. It has been shown that M. abscessus is resistant to first-line anti-tuberculosis agents such as rifampin and ethambutol [13,14]. Currently, treatment for M. abscessus infection consists of intravenous amikacin combined with oral macrolide such as clarithromycin, as suggested by the American Thoracic Society and the Infectious Diseases Society of America [4]. However, the cure rate of M. abscessus pulmonary disease is only 30-50%, much lower than other NTM diseases [15]. Fatal M. abscessus infections have also been reported, especially after lung transplantation [16,17]. Several reports have indicated that M. abscessus can infect both immunodeficient and immunocompetent patients [2,6,18]. M. abscessus exhibits two different colony morphotypes, smooth and rough [19], with different properties such as cell wall content, hydrophobicity, sliding motility, and aggregation ability [20][21][22]. The rough strain is more virulent than the smooth strain, and the change from smooth to rough morphotype may occur during infection [23,24]. Studies have shown that a difference between the smooth and rough variants is the amount of glycopeptidolipid (GPL), which is a major surface lipid component of M. abscessus, M. smegmatis, M. chelonae, and M. avium complex [25,26]. Single nucleotide polymorphisms or multiple indels with GPL locus were identified by several comparative genomics analysis in the isogenic smooth and rough M. abscessus pairs [26]. Deficiency in GPL is associated with the rough morphotype and increased invasiveness [21]. Rough morphotype variants are also more potent than smooth variants in stimulating human macrophages through TLR2 to produce cytokines [27]. Cells of rough strains have been shown to persist and multiply in mouse and human monocytes, while those of smooth strains are rapidly cleared [24,26].
In order to investigate the mechanism of colony morphotype switch, we attempted to identify genes involved in colony morphology and virulence of M. abscessus. In our previous study [28], a transposon mutagenesis was performed in the rough M. abscessus cs1c-R strain. A mutant with a smooth morphotype designated mab_3083c::Tn was isolated. Results of bioinformatic analyses suggest that mab_3083c is a homologue of ribonuclease J. RNase J is an enzyme present in many bacteria and plays an important role in all aspects of mRNA metabolism including ribosomal RNA maturation, mRNA decay and stability [2]. In this study, we investigated the role of the mab_3083c gene, homologue of RNase J, in colony morphology switching and surface-associated properties of M. abscessus.

Sliding Motility Assay
Sliding motility was examined as previously described [30]. M. abscessus cells were inoculated on 7H9 with 0.3% agar. After an incubation at 37 • C for 3 days, the sliding distance was measured.

Aggregation Assay
The aggregation assay was performed as previously described with modifications [22]. M. abscessus cells with an optical density of OD 600 = 0.1 were cultured in a tube containing 5 mL of 7H9 medium with 10% OADC. After an incubation at 37 • C with continuous rotation for 2 days, the cultures were allowed to stand still at room temperature for 10 min. The upper portion of each culture containing dispersed cells was then removed to determine the OD 600 value. The aggregated cells at the bottom of each tube were completely suspended by vortexing with glass beads of 4.5 mm in diameter (Biospec, Bartlesville, OK, USA), and the cell suspension was measured for its OD 600 value. The aggregative index of each culture was calculated as the ratio of OD 600 value of aggregated cells to that of dispersed cells [22].

Intracellular Survival Assay
This assay was performed as previously described [28]. The human acute monocytic leukemia cell line THP-1 was acquired from ATCC. THP-1 cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (GIBCO, Gland island, NY, USA) and differentiated into macrophage by treatment with phorbol-12-myristate-13-acetate (500 ng/mL) (Sigma, St. Louis, MO, USA) at 37 • C in a CO 2 incubator overnight in 24 well plates (SPL Lifesciences) with a total amount of 500 µL, 2 × 10 5 cells per well. The culture medium was refreshed, and the culture was incubated for 2 more days. The cells were then infected with M. abscessus at a multiplicity of infection (MOI) of 1 for 2, 24, 48, or 72 h. The infected cells were washed to remove extracellular bacteria. The washed cells were then lysed by incubation with 1% Triton X-100 for 30 min, and the cell lysate was plated on 7H11 agar plates to determine the number of intracellular mycobacteria [32,33].

Sequence Alignment
According to the result of Kyoto Encyclopedia of Genes and Genomes (KEGG, http://www.genome.jp/kegg/ (accessed on 23 March 2021)) analysis, the functional role of Mab_3083c was recommended as RNase J. The following amino acid sequence alignments were performed by clustalW2 and NCBI program.

Statistical Analysis
The experimental data are presented as mean ± standard deviation. Statistical analysis was carried out by Student's t-test, one-way ANOVA, and Tukey's test for multiple comparisons.

Characterization of the Mab_3083c::Tn Mutant
To investigate factors associated with colony morphotype switching of M. abscessus, a transposon mutagenesis was performed on the rough strain cs1c-R (wild type). A mutant named mab_3083c::Tn with the colony morphotype switched from rough to smooth was obtained ( Figure 1A, middle panel). To confirm that this morphotype switching was due to the defect in mab_3083c, the plasmid pYUB-mab_3083c containing the wild type mab_3083c of M. abscessus was introduced into the mab_3083c::Tn mutant. The colony morphology of this complemented strain (designated mab_3083c::Tn/mab_3083c) was found to revert back to rough ( Figure 1A, right panel), indicating that mab_3083c is associated with the rough morphotype of M. abscessus.
To determine the insertion site of Tn5, genomic DNA was isolated from the mab_3083c::Tn mutant and then digested with BamHI. The resulting DNA fragments were circularized by ligation and used as template for PCR with primers KAN-2 FP-1 and KAN-2 RP-1 that anneal at the two ends of Tn5 ( Figure 1B). The PCR products thus obtained were sequenced. Results of this experiment showed that the Tn5 was inserted into mab_3083c (GenBank accession no. NC_010397) between nucleotides 365 and 366 downstream from the initiation codon ( Figure 1B).
To confirm the presence of mab_3083c in the genome of the complemented strain (mab_3083c::Tn/mab_3083c), PCR was performed with primers ( Figure 1B, dotted arrow) designed to amplify a portion (862 bp) of mab_3083c ( Figure 1B,C). As the control, the same PCR was performed on genomic DNA of the mab_3083c::Tn mutant, and a 2083-bp fragment ( Figure 1C) was determined, indicating that the 1221-bp Tn5 was inserted into the 862-bp fragment of mab_3083c ( Figure 1B). Both the 2083-bp ( Figure 1C) and 862-bp fragments ( Figure 1C) were generated from the complemented strain (mab_3083c::Tn/mab_3083c), suggesting that some copies of the introduced mab_3083c on pYUB-mab_3083c were integrated into the chromosome of mab_3083c::Tn/mab_3083c.

No Association of GPL with Colony Morphotype Switching of the mab_3083c::Tn Mutant
A previous study showed that the strain with a smooth colony morphotype contained higher amounts of GPL than the rough strain [21]. To investigate the GPL profiles of wild type (cs1c-R), mab_3083c::Tn, and complemented (mab_3083c::Tn/mab_3083c) strains, lipids of these strains were extracted and analyzed by TLC. The lipid profiles of three strains were no significant difference (Figure 2A,B).

No Association of GPL with Colony Morphotype Switching of the mab_3083c::Tn Mutant
A previous study showed that the strain with a smooth colony morphotype contained higher amounts of GPL than the rough strain [21]. To investigate the GPL profiles of wild type (cs1c-R), mab_3083c::Tn, and complemented (mab_3083c::Tn/mab_3083c) strains, lipids of these strains were extracted and analyzed by TLC. The lipid profiles of three strains were no significant difference (Figure 2A,B).

No Association of Mab_3083c with Intracellular Survival and Susceptibility to Hyd Peroxide and Lysozyme of M. abscessus
It has been shown that cells of the rough strain of M. abscessus can persist in monocytes, while those of the smooth strain are rapidly cleared [20]. To inv whether mab_3083c is associated with the intracellular survival of M. abscessus, TH were respectively infected with the wild type (cs1c-R) and the mab_3083c::Tn mut MOI of 1, and the CFU of intracellular bacteria was counted at 2, 24, 48, and 7 infection. As shown in Figure 3A To investigate the susceptibility of the wild type and the mab_3083c::Tn m oxidative stress and lysozyme, the bacteria were incubated with 10-30 mM hydro oxide for 2 h or 0.5-2.5 mg/mL lysozyme for 24 h. Bacteria incubated with broth hydrogen peroxide or lysozyme were used as controls. The susceptibility was ex as the percentage of CFU of bacteria with treatment divided by that without tr

No Association of Mab_3083c with Intracellular Survival and Susceptibility to Hydrogen Peroxide and Lysozyme of M. abscessus
It has been shown that cells of the rough strain of M. abscessus can persist in human monocytes, while those of the smooth strain are rapidly cleared [20]. To investigate whether mab_3083c is associated with the intracellular survival of M. abscessus, THP-1 cells were respectively infected with the wild type (cs1c-R) and the mab_3083c::Tn mutant at an MOI of 1, and the CFU of intracellular bacteria was counted at 2, 24, 48, and 72 h post infection. As shown in Figure 3A H2O2, and 1.4 ± 0.4, and 0.9 ± 0.3 to 30 mM H2O2, respectively ( Figure 3B). The percentages of susceptibility of the wild type and the mutant were 13.0 ± 1.4 and 8.6 ± 0.7 to 0.5 mg/mL lysozyme, 3.9 ± 0.6 and 1.0 ± 0.5 to 2.5 mg/mL lysozyme, respectively ( Figure 3C). Since there was no significant difference in H2O2 and lysozyme susceptibility between the wild type and the mutant, it is unlikely that mab_3083c regulates the intracellular survival of M. abscessus.  To investigate the susceptibility of the wild type and the mab_3083c::Tn mutant to oxidative stress and lysozyme, the bacteria were incubated with 10-30 mM hydrogen peroxide for 2 h or 0.5-2.5 mg/mL lysozyme for 24 h. Bacteria incubated with broth without hydrogen peroxide or lysozyme were used as controls. The susceptibility was expressed as the percentage of CFU of bacteria with treatment divided by that without treatment. The results showed that the percentages of susceptibility of the wild type and the mutant were 52.3 ± 8.7 and 46.8 ± 9.4 to 10 mM H 2 O 2 , 7.4 ± 0.8 and 5.6 ± 0.7 to 20 mM H 2 O 2 , and 1.4 ± 0.4, and 0.9 ± 0.3 to 30 mM H 2 O 2 , respectively ( Figure 3B). The percentages of susceptibility of the wild type and the mutant were 13.0 ± 1.4 and 8.6 ± 0.7 to 0.5 mg/mL lysozyme, 3.9 ± 0.6 and 1.0 ± 0.5 to 2.5 mg/mL lysozyme, respectively ( Figure 3C). Since there was no significant difference in H 2 O 2 and lysozyme susceptibility between the wild type and the mutant, it is unlikely that mab_3083c regulates the intracellular survival of M. abscessus.

Higher Sliding Motility and Less Aggregation Capability of the Mab_3083c::Tn Mutant
According to previous studies, smooth strains of M. abscessus displayed higher sliding motility and less aggregation capability [22,30]. To test whether the mab_3083c::Tn mutant have similar properties, its motility was examined on 7H9 medium with 0.3% agar. The sliding distance was measured and plotted ( Figure 4A,B). The wild type strain was found to have a low motility (2.83 ± 0.38 mm), while the mab_3083c::Tn mutant was highly motile (8.56 ± 0.18 mm). The mab_3083c complemented strain (mab_3083c::Tn/mab_3083c) was similar to the wild type with a low sliding phenotype (1.44 ± 0.29 mm). of M. abscessus to H2O2. Approximately 3 × 10 7 bacteria were inoculated into 7H9 broth containing 10, 20, or 30 mM of H2O2 and incubated at 37 °C for 2 h. The CFU of the treated cells was then counted, and the susceptibility to H2O2 was determined. (C) Susceptibility of M. abscessus to lysozyme. Approximately 3 × 10 7 bacteria were inoculated into 7H9 broth containing 0.5 or 2.5 mg/mL of lysozyme and incubated at 37 °C for 24 h. The CFU of the treated cells was counted, and the susceptibility to lysozyme was determined.

Higher Sliding Motility and Less Aggregation Capability of the Mab_3083c::Tn Mutant
According to previous studies, smooth strains of M. abscessus displayed higher sliding motility and less aggregation capability [22,30]. To test whether the mab_3083c::Tn mutant have similar properties, its motility was examined on 7H9 medium with 0.3% agar. The sliding distance was measured and plotted ( Figure 4A,B). The wild type strain was found to have a low motility (2.83 ± 0.38 mm), while the mab_3083c::Tn mutant was highly motile (8.56 ± 0.18 mm). The mab_3083c complemented strain (mab_3083c::Tn/mab_3083c) was similar to the wild type with a low sliding phenotype (1.44 ± 0.29 mm).
To test whether mab_3083c affects the aggregation capability of M. abscessus, the variants were grown in 7H9 media with 10% OADC and 0.05% Tween 80 for two days. Cultures of the mab_3083c::Tn mutant were found to be turbid, whereas those of the wild type and the complemented (mab_3083c::Tn/mab_3083c) strains had bacterial aggregates at the bottom and clear supernatants. To quantify the aggregation ability, the aggregation index of each culture was calculated and plotted. As shown in Figure 4C,D, the mab_3083c::Tn mutant (1.70 ± 0.03) displayed a lower aggregation ability than the wild type (4.49 ± 0.62), and the mab_3083c complemented strain (9.49 ± 0.69) regained the aggregative phenotype. These results suggest that mab_3083c is associated with a low sliding and a high aggregation ability of M. abscessus.  To test whether mab_3083c affects the aggregation capability of M. abscessus, the variants were grown in 7H9 media with 10% OADC and 0.05% Tween 80 for two days. Cultures of the mab_3083c::Tn mutant were found to be turbid, whereas those of the wild type and the complemented (mab_3083c::Tn/mab_3083c) strains had bacterial aggregates at the bot-tom and clear supernatants. To quantify the aggregation ability, the aggregation index of each culture was calculated and plotted. As shown in Figure 4C,D, the mab_3083c::Tn mutant (1.70 ± 0.03) displayed a lower aggregation ability than the wild type (4.49 ± 0.62), and the mab_3083c complemented strain (9.49 ± 0.69) regained the aggregative phenotype. These results suggest that mab_3083c is associated with a low sliding and a high aggregation ability of M. abscessus.

Identification of Mab_3083c as a Homologue of Ribonuclease J
Amino acid sequence alignment between Mab_3083c and RNase J of M. smegmatis mc 2 155 (msmeg_2685) indicated that they share a 75.4% identity ( Figure 5A). To confirm that Mab_3083c possesses RNase J activity, the msmeg_2685 gene, which encodes RNase J of M. smegmatis, was introduced into the mab_3083c::Tn mutant by electroporation of pYUB-msmeg_2685. The resulting strain was designated mab_3083c::Tn/msmeg_2685. This msmeg_2685 complementation was found to revert the colony morphology of the mab_3083c::Tn mutant back to the rough morphotype ( Figure 5B), the same as complementation with mab_3083c ( Figures 1A and 5B).

Identification of Mab_3083c as a Homologue of Ribonuclease J
Amino acid sequence alignment between Mab_3083c and RNase J of M. smegmatis mc 2 155 (msmeg_2685) indicated that they share a 75.4% identity ( Figure 5A). To confirm that Mab_3083c possesses RNase J activity, the msmeg_2685 gene, which encodes RNase J of M. smegmatis, was introduced into the mab_3083c::Tn mutant by electroporation of pYUB-msmeg_2685. The resulting strain was designated mab_3083c::Tn/msmeg_2685. This msmeg_2685 complementation was found to revert the colony morphology of the mab_3083c::Tn mutant back to the rough morphotype ( Figure 5B), the same as complementation with mab_3083c ( Figures 1A and 5B).
It has been shown that D85K and H86A mutations generated at the active site of M. smegmatis RNase J resulted in loss of its RNase activity [34]. To prove that Mab_3083c has RNase J activity, similar mutations, D89K and H90A, were generated in mab_3083c on pYUB-mab_3083c. The resulting plasmid was named pYUB-mab_3083cD89K, H90A and introduced into the mab_3083c::Tn mutant to perform the complementation experiment. The new variant designated mab_3083c::Tn/mab_3083cD89K, H90A was found to have a smooth morphotype ( Figure 5B) and higher sliding ability ( Figure 5C), the same as the mab_3083c::Tn mutant, indicating that the mutated mab_3083c gene (mab_3083cD89K, H90A) failed to complement the mab_3083c::Tn mutation. Taken together, these results provided evidence that Mab_3083c functions as an RNase J.  It has been shown that D85K and H86A mutations generated at the active site of M. smegmatis RNase J resulted in loss of its RNase activity [34]. To prove that Mab_3083c has RNase J activity, similar mutations, D89K and H90A, were generated in mab_3083c on pYUB-mab_3083c. The resulting plasmid was named pYUB-mab_3083c D89K, H90A and introduced into the mab_3083c::Tn mutant to perform the complementation experiment. The new variant designated mab_3083c::Tn/mab_3083c D89K, H90A was found to have a smooth morphotype ( Figure 5B) and higher sliding ability ( Figure 5C), the same as the mab_3083c::Tn mutant, indicating that the mutated mab_3083c gene (mab_3083c D89K, H90A ) failed to complement the mab_3083c::Tn mutation. Taken together, these results provided evidence that Mab_3083c functions as an RNase J.

Discussion
GPLs are a member of glycolipids produced by many mycobacteria genus involved in morphotype switching, biofilm formation, sliding motility, pathogenicity, and immunomodulation [26,[35][36][37]. Notably, many studies demonstrated that absence of GPL in Mycobacterium promotes bacterial aggregation and loses the ability to produce biofilm and sliding ability [26]. In this study, morphotype switching strain (mab_3083c::Tn mutant) was selected. Interestingly, no significant differences were found between wild type and mutant in GPL profiles, intracellular survivability, susceptibility to hydrogen peroxide and lysozyme. According to bioinformatics analysis, the role of mab_3083C was recommended as RNase J. Phenotype alterations regulated by RNase J have been reported in Bacillus subtilis [38] and Streptomyces venezuelae [39]. RNase J1 knockout in Bacillus subtilis affected its spore formation and maturation as well as alteration in cell appearance with disordered peptidoglycan layer and long spiral filaments [38]. The RNase J mutant of Streptomyces venezuelae also has defects in cell development, sporulation, and ribosome assembly. In our study, the RNase J mutant of M. abscessus (mab_3083c::Tn) was viable, suggesting that RNase J is not essential for its growth, unlike Streptococcus pyogenes in which both RNase J1 and RNase J2 are essential [40].
As some ribonucleases, RNase J regulates mRNA stability or ribosomal RNA maturation [34,[41][42][43]. In Gram-positive bacteria, RNase J is functionally equivalent to RNase E, which plays an important role in mRNA decay in Gram-negative bacteria [44]. According to the structure of Thermus thermophilus RNase J [45], RNase J possesses three domains: βlactamase domain, β-CASP domain, and C-terminal domain. D85 and H86 in RNase J Msmeg are located at the active site of the β-lactamase domain, and D85K and H86A mutations of M. smegmatis RNase J result in the loss of its exonuclease activity [34].
Some studies have described the role of RNase J in mRNA turnover. For example, the RNase J1/J2 double mutant of Bacillus subtilis was found to have increased levels of approximately 300 mRNA transcripts and decreased levels of equal numbers of mRNA transcripts [46]. Studies have also shown that RNase J can affect the stability of specific mRNAs [44,46]. It is possible that the RNase J (mab_3083c) of M. abscessus affects the stability of some mRNAs related to its colony morphology, sliding motility, and aggregation ability. Further studies are warranted to identify the genes encoding these mRNAs.

Conclusions
Most reports on the morphotypic switch of M. abscessus were associated with GPL locus. Our study demonstrated that the non-GPL locus gene mab_3083c also plays a role in this switch. The disruption of mab_3083c led to colony morphology switching from rough to smooth, increased sliding motility, and reduced aggregation ability of M. abscessus. Complementation of the mab_3083c::Tn mutant with RNase J Msmeg (msmeg_2685) switched its colony morphology from smooth back to rough. These results suggest that mab_3083c is the RNase J gene of M. abscessus.