The Role of Pseudomonas aeruginosa RNA Methyltransferases in Antibiotic Resistance

Methyltransferases play a fundamental role in aminoglycoside resistance of Gram-negative bacteria, and some of its mechanisms were described in the past years, especially in Escherichia coli ; however, it remains unsolved for other resistant bacteria such as Pseudomonas aeruginosa . Despite hurdles to determine resistance acquisition, high-throughput approaches (genomics, transcriptomics, and proteomics) have allowed data mining and analysis in a systemic way. Likewise, bioinformatics modelling of homologous genes or proteins has permitted to eluci-date the emerging resistance in this pathogen. P. aeruginosa is a bacterial resistance treat since practically all known resistance mechanisms can be described using this model, particularly RNA methyltransferases. The RNA methyltransferases perform methylation or demethylation of ribosomal RNA to allow or restrict the antibiotic resistance development. The Kgm and Kam methyltransferases families are found in P. aeruginosa and confer resistance to several aminoglycosides. Loss of native methylations may also confer a resistant phenotype. The P. aeruginosa RsmG has high structural homology with Thermus aquaticus protein. Today, molecular data will promote a new paradigm on antibiotic therapy for treatment against P. aeruginosa . This chapter provides an overview of what role P. aeruginosa’ s methyltransferases play in antibiotic resistance, induced by methylation or demethylation in the ribosome.


Introduction
Nowadays, aminoglycoside antibiotic regimen remains as a prevailing therapy for the treatment of Pseudomonas aeruginosa pathogen, predominantly for respiratory complications in cystic fibrosis patients. However, P. aeruginosa strains are emergent multidrug-resistant. Bacteria develop resistance to aminoglycosides by producing aminoglycoside-modifying enzymes such as acetyltransferase, phosphorylase, and adenyltransferase [1]. The bacterial ribosome is a primary antibiotic Ribonucleotides methylated where the methyl moiety is located either in 16S or 23S ribonucleotides: m 6 A [66,67], m 6 2 A [68], m 5 C [67], m 4 Cm [69], m 2 G [66], m 7 G [70], m 3 U [67,71], m 5 U [72], and Ψ [68]. Molecules are designed using the figures on the next web page, https://mods.rna.albany.edu/mods/modifications/search/. The structure shows clearly the methyl (CH 3   region of 16S and 23S rRNAs and it is known to have 10 methylated nucleosides in 16S rRNA and 14 methylated nucleotides in 23S rRNA [4]. Different methylation sites have been identified within the 16S rRNA which yield different aminoglycoside resistance phenotypes [5]. One type group of 16S rRNA methylases is produced by istamycin producer Streptomyces tenjimariensis, which methylates m 1 A1408 residue. Another group of 16S rRNA methylases is synthesized by gentamicin producer Micromonospora purpurea that methylates residue G1405; nonetheless, in P. aeruginosa the modification takes place in helix 44, with a secondary target in 23S rRNA helix 69 [6]. They bind specifically to the aminoacyl site (A-site) of 16S rRNA within the prokaryotic 30S ribosomal subunits and interfere with protein synthesis [7].
Probably RNA methylation began prior to DNA methylation in the early forms of life evolved on Earth [8], allowing to hypothesize perhaps that methylases appeared before polymerases. Ribonucleotide can be methylated by methyltransferases or demethylated by demethyltransferases. The structure of the bacterial ribosome has a molecular mass of 2.5 mega Daltons. In E. coli, the 50S subunit is composed of 23S rRNA (2904 nt), 5S rRNA (120 nt), and 33 ribosomal proteins, while the 30S subunit is composed of 16S rRNA (1542 nt) and 21 ribosomal proteins (S1-S21) [9].
The 16S rRNA resistance methyltransferases modify only intact 30S subunits, but the molecular details of their target recognition mechanisms are not quite elucidated yet. Such studies are becoming all the more necessary [10]. RNA methylation has been observed in different types of RNA species, viz., mRNA, rRNA, tRNA, snoRNA, snRNA, miRNA, and tmRNAs. Specific RNA methyltransferases are synthesized by cells to label these RNA species according to their needs and prevailing environmental conditions surrounding the cells, and this molecular labeling system is a constituent of epigenetics. New molecular structures provide crucial new insights that may provide a starting point for strategies to suppress these emerging causes of pathogenic bacterial resistance to aminoglycosides [11]. Nonetheless, bacteria develop resistance to aminoglycosides by producing aminoglycoside-modifying enzymes such as acetyltransferase, phosphorylase, and adenyltransferase. These enzymes, however, cannot confer a broad aminoglycoside resistance spectrum due to its substrate specificity [1].

Antibiotic resistance
P. aeruginosa has a large genome among Gamma proteobacteria, which allows it to improve many resistance mechanisms in a versatile way, for example, by transmissible plasmids or integrons. P. aeruginosa derepresses the chromosomal AmpC cephalosporinase [12,13]; it also acquires genes for AmpC enzymes, class A carbenicillinases or β-lactamases, class D oxacillinases, and class B carbapenem-hydrolyzing enzymes [14], as it occurs in other bacteria like E. coli and K. pneumoniae. Other mechanisms include modifying the structure of topoisomerases II and IV to become quinolone resistant [15], decreasing outer membrane permeability by the partial or total failure of OprD proteins [12], overexpressing the active efflux systems with broad substrate patterns [16,17], or synthesizing aminoglycoside-modifying enzymes as adenylyltransferases, acetyltransferases, and phosphoryltransferases [18].
The range of antibiotic resistance in P. aeruginosa is wide, and it represents a major difficulty for health care by its unsuccessful treatment, as a consequence of its low intrinsic antibiotic susceptibility, an effect of the interaction between multidrug Bacterial multidrug resistance (MDR) is an important concern in P. aeruginosa since this microorganism is capable of mixing several mechanisms, transposons, plasmids, and chromosomally encoded genes, such as methyltransferases or pumps [22]. Methyltransferase genes are spread in bacterial genome ready to trigger antibiotic resistance. Table 1 compares the reported methyltransferase proteins worldwide, being annotated for P. aeruginosa in UniProt database [23].
One mechanism of adaptation which facilitates natural selection in bacteria is the hypermutation of some genes or chromosomal regions. Previous work in patients with P. aeruginosa showed that hypermutation causes a problematic effect during a chronic respiratory infection (CRI) [24], where P. aeruginosa was up to 6.5-fold higher in mutator backgrounds. Other elements associated with high antimicrobial resistance are integrons. These elements were found among isolates from Iran patients with P. aeruginosa, which the int1 integron was prevalent [25].
Integrons linked to transposons, plasmids, and chromosome are responsible for bacterial antibiotic resistance [26,27]. Integrons are composed of three elements: (1) the integrin-associated promoter (Pc), which is required for transcription and expression of gene cassettes (genetic elements that encode antibiotic resistance genes) within the integrin; (2) the intI gene, in which coding for the integrase IntI is crucial for site-specific recombination; and (3) the adjacent recombination site attI, which is recognized by integrase. On the other hand, P. aeruginosa carrying transposon Tn1696 is an element that encodes the CmlA gene, an exporter of the major facilitator (MF) superfamily which provides antibiotic resistance, specifically against chloramphenicol [28]. P. aeruginosa has a broad spectrum in cephalosporin resistance mechanism, mediated by the extended-spectrum β-lactamases (ESBLs). High prevalence of multidrug resistance in burn patients and production of oxa-10, per-1, and veb-1 genes by P. aeruginosa isolates confirm the presence of antibiotic-degrading enzymes [29].
The Pathosystems Resource Integration Center (PATRIC) is a massive database that integrates genomic data and analysis tools to support biomedical research on bacterial infectious diseases. The platform provides an interface for biologists to discover data and information and conduct comprehensive comparative genomics and other analyses in a one-step source. PATRIC database provides complete genome information and data regarding susceptibility or resistance [30] to several antibiotics; including aminoglycosides, polymyxin B, colistin, ceftazidime, piperacillin, imipenem, ciprofloxacin, levofloxacin, and meropenem in P. aeruginosa. We report in Figure 2 antibiotic resistance or susceptibility from different P. aeruginosa strains as well as in Figure 3 with those antibiotics mentioned.

Resistance to antibiotics through rRNA methylation
Kgm and Kam families are two different groups of SAM-dependent RNA methyltransferases, which modify nucleotides of 16S rRNAs in the specific drug-binding site to confer self-resistance in aminoglycoside-producing bacteria [31]. The Kgm and Kam families have been distinguished based on their nucleotide targets, G1405 and A1408, respectively. The kgmB and armA genes (Kgm family kanamycin gentamicin methyltransferase) methylate m 7 G1405(N7) position that confers a high level of resistance against gentamicin, kanamycin, and tobramycin. The addition of a methyl group in this position interferes directly with the binding to the antibiotic, inducing a steric hindrance between the modified base and the structure of the antibiotic, causing electrostatic repulsions derived from the positive charge in the modified base [32]. On the other hand, the kamA and npmA genes (Kam family kanamycin-apramycin methyltransferase) methylate m 1 A1408(N1) position conferring a high level of resistance to kanamycin, apramycin, and neomycin [5].
Another interesting non-aminoglycoside resistance related to RNA methylation is the macrolide-lincosamide-streptogramin-B (MLSB) antibiotics, which is strongly associated with the expression of the methyltransferase of ErmC RNA that causes the dimethylation of the N-6 atom of adenine and interacts with the nucleotide 2058 in the 23S rRNA. Such antibiotics bind to overlap sites within the 50S ribosomal subunit tunnel near the peptidyl transferase center, either by inhibiting the catalysis directly at the peptidyl transferase site or by acting as a physical barrier to the extension of the peptide chain inside the tunnel [33]. Many more erm-type methyltransferase genes have been identified in a wide range of Gram-positive and Gram-negative bacteria. Among them, the ermB, ermF, and ermA genes are transferred by transposons, and the ermC gene transferred by plasmids. The family of Erm methyltransferases that mediate the mono-or dimethylation of A2058 consists of approximately 40 different classes of methylases [34,35].

Resistance or susceptibility to antibiotics through rRNA demethylation
The rsmG gene encodes a 16S rRNA mRNA which methylates the N7 of nucleotide G527 within the 530 loop of 16S rRNA; one of the main examples is the loss of native methylations that confers a resistance phenotype to streptomycin. Streptomycin interacts with the rRNA in the adduced region (loop 530), and the loss of methylation correlates with a low level of resistance. Although this resistance is at a low dose of antibiotic, the mutation of rsmG apparently has a mutator effect which promotes the appearance of a high number of mutants resistant to high doses of streptomycin [36]. Another interesting aspect is that not only methylation generates resistance; cases have been reported where demethylation also promotes resistance. The first and best characterized example is ksgA gene (RsmA protein), which encodes the native methyltransferase KsgA or RsmA, responsible for the N6 dimethylation of A1518 and A1519 in the 3′-terminal fork of the 16S rRNA in the 30S rRNA. It was the first resistance to aminoglycosides (kasugamycin) associated with demethylation in the 16S rRNA [37]. It was found that adenine methylated by MTase is far from the binding site of kasugamycin, so this demethylation should lead to a conformational rearrangement which would be associated with the acquisition of antibiotic resistance [38].
Another research showed that preventing adenine methylation from occurring, resistance to kasugamycin can be induced; the base U793 fills the site usually occupied by the methylated adenines and the adjacent bases, A792 and A794, [39]. The phenomena mentioned above give place to a conformational change, causing the union site of Ksg to be blocked by the U793. Accordingly, it can be assured that this structural change in the helix 24 causes resistance to Ksg [39].
Likewise, it was found that the tlyA gene in Mycobacterium tuberculosis encodes the MTase 2'-O-ribose TlyA responsible for the C1409 methylations in the 16S rRNA and C1920 in the 23S rRNA. The loss of such methylations confers resistance to capreomycin and viomycin, two antibiotics which bind at the interface of the ribosome subunit and are used to help define their binding site. Another example of the absence of methylations in the 23S rRNA is the lack of methylation in U2584 (E. coli numbering), which causes resistance to sparsomycin in 23S rRNA Halobacterium salinarum [40].
Recent findings regarding intrinsic resistance refer to the Ψ at position 2504 of the 23S rRNA in E. coli, where inactivation of the rluC gene confers significant resistance to clindamycin, linezolid, and tiamulin [41]. The cfr gene was originally discovered in an isolate of a multiresistant plasmid during a follow-up study of chloramphenicol resistance in Staphylococcus spp. isolates. The molecular characterization of the resistance led to the gene encoding a methyltransferase that methylated the nucleotide A2503 in the 23S rRNA. In E. coli and S. aureus, there is a natural methylation of A2503 mediated by the methyltransferase encoded by the yfgB gene (rlmN). The lack of natural methylation in A2503 confers a slight increase in susceptibility to tiamulin, hygromycin A, sparsomycin, and linezolid [42].
RsmH also called S-adenosyl-L-methionine (cytosine 1402 -N 4 )-methyltransferase methylates the N 4 -of cytosine 1402 [54] ( Table 1). This enzyme catalyzes the following chemical reaction: S-adenosyl-L-methionine + cytosine 1402 in 16S rRNA → S-adenosyl-L-homocysteine + N 4 -methylcytosine 1402 in 16S rRNA (see reaction in UniProt, KEGG, or MetaCyc). Experiments performed with gene knockout of rsmH and rsmI have shown in E. coli BW25113 strain that ΔrsmH and ΔrsmI increase in doubling times by 15 and 12%, respectively; however, ΔrsmH/ΔrsmI increases in doubling time by 29% compared with a wild type cultured at 37°C, indicating that gene knockout caused a slight but significant change in phenotype about cellular growth properties in the absence of both rsmH and rsmI [54]. As well as E. coli, P. aeruginosa conserves rsmH and rsmI genes in both aminoglycosideresistant and susceptible strains; therefore, it is important to study the mutations also in its strains.
Other interesting P. aeruginosa methyltransferases associated with aminoglycoside resistance are m 5 C1404, m 1 A1408, and m 7 G1405 [6]. Among the last group mentioned, there are some well-studied methyltransferases, such as ArmA, RmtA, RmtB, RmtC, RmtD, RmtF, and RmtG ( Table 2). This group is characterized for providing resistance to 4,6-disubstituted 2-deoxystreptamine (2-DOS) aminoglycosides [6]. For example, ArmA was found in Klebsiella pneumoniae [56]; as for P. aeruginosa, among 100 Korean multidrug-resistant isolates, 14 carried this enzyme [57]. The armA gene encodes for 16S RNA methyltransferase that methylates guanine (1405)-N 7 . The same gene in P. aeruginosa ( Table 2) presents variable occurrence as it is part of the accessory genome. A multiple alignment, using the listed P. aeruginosa ArmA proteins (16S rRNA (guanine (1405)-N(7))-methyltransferase)) revealed identical homology for this marker. The location of m 7 G1405 methyltransferase genes across the prokaryotic genome is variable, as it has been found in several studies mentioned above. The rmtA gene in P. aeruginosa carries a mobile element Tn5041 [58,59] identified previously in Enterobacteriaceae [56], while the rmtB gene identified in Serratia marcescens is located in the flanking of Tn3-like region responsible for multiple antimicrobial resistance [59,60], and both methyltransferases and mobile elements are present in P. aeruginosa ( Table 2). The rmtC and rmtF genes ( Table 2) might have been acquired from plasmids as part of mobile genetic elements and finally integrated and stabilized on the chromosome [61]. The rmtG gene ( Table 2) is likely located in the chromosome [62]. The m 1 A1408 methyltransferases are present in pan-aminoglycoside-resistant strains, which were identified by Wachino et al. [63] and provide resistance by these classes of methyltransferases to both 4,5-disubstituted 2-DOS and 4,6-disubstituted 2-DOS aminoglycosides as well as NpmA case. These two classes of methyltransferases are

suit between RsmI proteins from P. aeruginosa (red) and E. coli (blue).
Although the alignment displays a structural shifting, the overall topology of the active site is maintained. The protruding residues from each protein depict the active site, harboring SAM ( fluorescent green). The structure of E. coli has already been solved by Zhao et al. [49], without obtaining, until now, the crystal structure of RsmI from P. aeruginosa.  very important for antibiotic resistance, thanks to the similarity of these enzymes with those homologs found in aminoglycoside-producing actinomycetes [6]. Those new genes and proteins will be better studied in expression and structure, to be related to epidemiological data. Looking at Table 2, it seems that the expression of methyltransferase might be related to geographical prescription. However, this hypothesis does not seem well founded: microbiological, molecular, and epidemiological understanding of RNA methyltransferases in P. aeruginosa will allow the rational use of aminoglycosides and maybe will be not replaced for new antibiotics.

Final considerations
Antibiotic resistance is a serious concern for public health and environment. To comprehend the molecular interaction of the methyltransferase in aminoglycoside resistance will be a more efficient way to rationalize its use and consumption. It will be better to clarify the panorama of the rational use of the aminoglycosides to diminish the rapid development of resistance before considering its replacement, since P. aeruginosa is still susceptible to them, and, moreover, currently it is known why other Gammaproteobacteria are resistant to them. Why are methylation and demethylation a feedback of the antibiotic environmental pressure in bacteria? Bhujbalrao and Anand [64] suggest us some insights using KsgA, exploring the factors which govern the resistance to antibiotics. They observed within loop1 and loop12 of rRNA switched chimera efficiently methylated mini-RNA substrates in vitro, showing that these structural elements suffice for local orientation of the rRNA. In addition, in vivo they notice that the head domain plays a more critical role in leading the enzyme to the select ribosomal region and serves as a sensor of the global environment.
As Kim et al. [65] discuss in letter to editor (Dr. Hur), investigating with P. aeruginosa and aminoglycoside resistance proposes that "less aminoglycoside consumption correlates with less resistance levels"; therefore, we consider that is a requisite for an antibiotic cycling strategy at the global level; also they discussed the rates of amikacin or gentamicin-resistant declining trends, according to the data from KONSAR Korean program in 2011 either for P. aeruginosa, K. pneumoniae, or Acinetobacter spp. [64]. With the knowledge about aminoglycoside resistance molecular mechanisms comparing to the rational prescription cited in Korea, for example, we hypothesize that low methylation rate in the nucleotide substrate of RsmH or RsmI is close to the anchor point of gentamicin in 16S RNA, indicating a possible association with gentamicin or aminoglycosides resistance [5].
RsmG, RsmH, and RsmI methyltransferases belong to the core genome (constitutive genome), while ArmA is part of the accessory genome with identical protein sequences among close species in Proteobacteria. Nowadays, the enzymatic activity has been well described; however the antibiotic resistance remains unsolved, perhaps as a consequence of broad usage of aminoglycoside in hospital environment, allowing the development of resistant bacteria. In the future, probably the treatment of P. aeruginosa will take into account the genetic trait of each isolate, strain, or species with the set of resistance genes, and surely methyltransferases will be included routinely in clinical care and high throughput or genomic medicine therapies.
© 2019 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.