ORIGINAL RESEARCH

Comparative bioinformatics analysis of antimicrobial resistance gene pool in the genomes of representatives of genus Corynebacterium

Kulshan TA, Bugaeva IO, Soboleva EF, Allyanova MS, Popov DA, Shvidenko IG
About authors

Razumovsky Saratov State Medical University, Saratov, Russia

Correspondence should be addressed: Tatiana А. Kulshan
B. Kazachiya, 112, Saratov, 410012, Russia; ur.xednay@nahsluk.anajtat

About paper

Author contribution: Kulshan TА — study planning, literature review, dealing with molecular genetic data (selection of genomes, genome annotation, comparative analysis of gyrA amino acid sequences), data analysis, manuscript writing; Bugaeva IO — study planning, data analysis, interpretation of findings, manuscript writing; Soboleva EF — literature review, data analysis, manuscript writing; Allyanova MS — literature review, analysis of the antimicrobial resistance gene pool in the genomes of corynebacterial strains, dealing with the PATRIC online service; Popov DA — literature review, search for gyrA amino acid sequences in the genomes of corynebacterial strains, comparative analysis of amino acid sequences; Shvidenko IG — advising during manuscript writing, data analysis.

Received: 2023-10-20 Accepted: 2023-12-03 Published online: 2023-12-19
|

Today, antimicrobial multidrug resistance of bacterial infectious agents poses a serious threat to global public health. Irrational use of antimicrobials for treatment of humans, in the livestock sector and agriculture is a determinant of widespread resistance to drugs among bacteria [13].

Selective pressure of antimicrobials on the bacterial population contributes to realization of various resistance mechanisms emerging due to acquisition of genetic determinants of resistance or spontaneous mutations [1, 46]. Assessment of evolutionary transformation of bacterial genomes associated with antibiotic resistance contributes to optimization of treatment strategies and preventive measures. 

Currently, the greater role played by normal flora members, specifically by members of the genus Corynebacterium, in infectious diseases can be associated with the spread of genes responsible for antimicrobial resistance across bacterial genomes. The increasingly frequent isolation of Corynebacterium as pathogens, especially in immunocompromised individuals, is indicative of the greater role in the development of infectious complications in patients played by Corynebacterium [2].

The following Corynebacterium species are of special importance for development of infections: C. amycolatum (skin and soft tissue infections, bacteremia, endocarditis, genital infections), C. urealyticum (acute and chronic urinary tract infections, urolithiasis), C. striatum (true bacteremia, bacterial colonization of prostheses, catheters, breathing tubes, etc.), C. jeikeium (bacteremia, endocarditis, pneumonia, skin and soft tissue infections), C. aurimucosum (acute and chronic joint infections, diabetic foot ulcer infection), C. genitalium (urinary tract infections) [2, 3, 5, 714]. Multidrug resistance of some Corynebacterium species to β-lactams, macrolides, aminoglycosides, quinolones, tetracyclines and rifampicins, lincosamides, etc., should be noted [1, 4, 1214].

However, the data on the Corynebacterium drug resistance are contradictory, that is why our study was aimed to conduct bioinformatics analysis of the pool of antimicrobial resistance genes in the published genomes of some representatives of the genus Corynebacterium.

METHODS

The study involved data on the whole genome nucleotide sequences of 22 strains of six Corynebacterium species (C. amycolatum, C. urealyticum, C. striatum, C. jeikeium, C. aurimucosum, C. genitalium) readily available from NCBI GenBank, isolated in different countries over the years (tab. 1).

Bioinformatics analysis of whole genome sequences aimed at the search for antibiotic resistance genes in the specified genomes was performed using PATRIC (Pathosystems Resource Integration Center), Comprehensive Antibiotic Resistance Database (CARD), and Database of AntibioticResistant Organisms (NDARO) [15].

Amino acid sequences of gyrA gene were acquired from Genbank. The UGENE (Unipro UGENE) 48.1 software package was used for analysis of gyrA amino acid sequences [16]. Amino acid sequence alignment was performed using the MUSCLE tool integrated into UGENE.

RESULTS

Bioinformatics analysis showed that the genomes provided comprised various combinations of antimicrobial resistance genes. A total of 25 different genes encoding resistance to drugs exhibiting antimicrobial activity were determined (tab. 2).

It should be noted that the following genes were significantly less often found in the genomes of studied isolates (tab. 3):

  1. tetO (tetW) (encodes resistance to tetracyclines) — was not found in genomes of 19 strains (86.4%);
  2. aph (3')-I, aph (6)-Ic (encode resistance to aminoglycosides) — were not found in genomes of 14 strains (63.6%);
  3. ermX (encodes resistance to macrolides, lincosamides, streptogramins) — was not found in genomes of 13 strains (59%);
  4. lsu (rplF) (encodes resistance to fusidic acid) — was not found in genomes of 12 strains (54.5%);
  5. cmx (encodes resistance to chloramphenicol) — was not found in genomes of eight strains (36,4,3%);
  6. ispC (dxr) (encodes resistance to fosfomycin) — was not found in genomes of seven strains (32%);
  7. gibB (encodes resistance to aminoglycosides), oxyR (encodes resistance to изониазиду), fabG (encodes resistance to triclosan) — were not found in genome of one strain (4,5%) (Corynebacterium striatum 824M, Corynebacterium striatum 1197, Corynebacterium striatum 708C, respectively).

However, resistance to aminoglycosides, fusidic acid, fosfomycins was encoded by several genes. In this regard, the lack of one gene in the genome can not indicate the isolate sensitivity to these antimicrobial substances.

All other genes provided in tab. 2 were found in 100% of genomes of 22 Corynebacterium strains. 

C. striatum 2308 was the strain containing 24 identified antimicrobial resistance genes out of 25. Only the tetO (tetW) gene was not found in its genome. According to the literature, this strain was isolated in 2011 from the blood culture of a man, who was treated at the hospital in Rio de Janeiro. Based on phenotypic characteristics, it showed sensitivity to tetracycline (MIC 1 mg/L), linezolid (MIC 0.25 mg/L) and vancomycin (MIC 0.5 mg/L) only [12]. The data of bioinformatics analysis we have obtained confirm the phenotypic study results [12]: no tetO (tetW) gene (tetracycline resistance), no genes encoding resistance to oxazolidones (linezolid) and glycopeptides (vancomycin). It is worth noting that no linezolid and vancomycin resistance genes were found in any of the studied strains. However, the authors point out that this strain showed phenotypic resistance to erythromycin (MIC > 256 mg/L) and clindamycin (MIC > 256 mg/L), as well as to gentamicin (aminoglycoside) (MIC 256 mg/L) [12]. Such phenotypic effects may result from the presence of genes ermX and aph (3')-I, aph (6)-Ic.

Corynebacterium amycolatum ICIS 9 extracted from vagina of a healthy woman in 2017 in Russia turned out to be one more strain with the genome showing the lack of gene ispC (dxr) (fosfomycin resistance) only. However, fosfomycin resistance is also encoded by the murA gene, which was found in the genome. The authors of the paper considered Corynebacterium amycolatum ICIS 9 as a potential probiotic agent for treatment of vaginal dysbiosis [911]. The Corynebacterium amycolatum ICIS 9 phenotypic resistance to antimicrobials (amikacin, gentamicin (aminoglycosides), amoxicillin (β-lactams), clarithromycin (macrolide), chloramphenicol, ciprofloxacin (fluoroquinolone) and tetracycline) was determined [911]. Indeed, our bioinformatics study showed that the genome of this isolate comprised genes encoding resistance to penicillins, aminoglycosides, macrolides, chloramphenicol, fluoroquinolones, and tetracyclines (tab. 2).

As for strains, the genomes of which lack a significant number of antimicrobial resistance genes (6–10 genes), these included the following: C. amycolatum ICIS 99, C. amycolatum ICIS 53, C. amycolatum SB-1, C. amycolatum 1189, C. striatum 824M, C. striatum 708 (tab. 3).

The C. striatum 708 strain extracted from synovial fluid of a patient in the UK (BioSample: SAMN34403526) comprised the lowest number of antimicrobial resistance genes (19 genes).

Currently, many causes of antimicrobial resistance of microorganisms are distinguished. This phenomenon results not only from the presence of genetic determinants associated with antimicrobial resistance, but also with various mutations in these genes. It has been found that mutations in the short regions of genes gyrA and gyrB (quinolone resistancedetermining regions (QRDR)) encoding А and В subunits of DNA gyrase result in quinolone/fluoroquinolone resistance [9].

In Corynebacterium, quinolone/fluoroquinolone resistance results from spontaneous mutations in the gene encoding gyrase A subunit [12, 13]. It has been found that mutations associated with amino acid changes in positions 87, 88 and 91 increase the minimum inhibitory concentrations (MICs) of quinolones/fluoroquinolones. Thus, amino acid substitutions in position 87, Ser (S) to Arg (R), Phe (F), Val (V), in position 88, Ala (A) to Pro (P), in position 91, Asp (D) to Tyr (Y), Gly (G), Ala (A), increased the ciprofloxacin, levofloxacin and moxifloxacin MICs [12, 13]. In this regard we considered it necessary to conduct a molecular genetic analysis of the gene amino acid sequence in 22 studied strains. GyrA of Corynebacterium glutamicum ATCC 13032 (GenBank ID: NP599264) was used as a reference when performing comparative analysis and determining the amino acid position number [13].

According to the literature data, the C. striatum ATCC 6940, C. jeikeium ATCC 43734 and C. urealyticum DSM 7109 isolates showed quinolone/fluoroquinolone resistance [13]. The gyrA amino acid sequences of these strains were used as controls.

The analysis showed that in the C. striatum ATCC 6940, C. jeikeium ATCC 43734, C. amycolatum     1189, C. aurimucosum UMB7769, C. striatum 1115, C. urealyticum 994, C. urealyticum 996, C. urealyticum DSM 7109, C.  jeikeium K411, C. amycolatum SB-1, C. genitalium ATCC 33030 strains, position 87 was occupied by Ser (S), while position 91 was occupied by Asp (D). According to the literature, such gene structure ensured the strains’ sensitivity to quinolones/ fluoroquinolones, despite the presence of resistance genes [12, 13].

Ser (S) replaced with Arg (R) in position 87 was reported in C. amycolatum ICIS 53, C. amycolatum ICIS 99. As for C. amycolatum VH6958 strain, Ala (A) replaced with Pro (P) in position 88 was reported in addition to Ser (S) replaced with Arg (R) in position 87. It is worth paying attention to the C. amycolatum BER245 strain, for which Asp (D) replaced with Tyr (Y) in position 91 was reported along with Ser (S) replaced with Arg (R) in position 87. Such mutations dramatically increased the MICs of quinolones/fluoroquinolones [12, 13].

C. urealyticum VH3073 had two unique substitutions: 87 — Ser (S)/ Val (V) and 91 — Asp (D)/ Tyr (Y). C. striatum 2308, C. striatum 708C, C. striatum 824M had only one amino acid substitution: 87 — Ser (S)/ Val (V). Moreover, strains carrying unique substitutions were identified: 87 — Ser (S)/ Ile (I), 91 — Asp (D)/ Ala (A) — C. amycolatum ICIS 9; 87 — Ser (S)/ Ile (I), 91 — Asp (D)/ Gly (G) — C. jeikeium 574; 87 — Ser (S)/ Phe (F), 91 — Asp (D)/ Gly (G) — C. striatum 1197 (figure). The evolutionary significance of these substitutions is to be determined in further studies.

Thus, 11 isolates have position 87 occupied by Ser (S), in 4 strains it is occupied by Val (V), in 4 strains by Arg (R), in 2 strains by Ile (I), in one strain by Phe (F). A total of 21 strains have position 88 occupied by Ala (A), while in one isolate it is occupied by Pro (P). A total of 17 isolates have position 91 occupied by Asp (D), in 2 strains it is occupied by Tyr (Y), in 2 strains by Gly (G), in 1 strain by Ala (A).

To summarize, it is worth noting that double mutations in gyrA described in the literature as mutations causing a dramatic increase in MICs of quinolones/fluoroquinolones were found in: C. amycolatum VH6958 isolated in 2016 in Spain (BioSample: SAMN18038700) — Ser (S) replaced with (R) in position 87, Ala (A) replaced with Pro (P) in position 88; C. amycolatum BER245 isolated in 2011 in Brazil from the patient with otitis — Ser (S) replaced with Arg (R) in position 87, Asp (D) replaced with Tyr (Y) in position 91; C. urealyticum VH3073 isolated in 2017 in Spain from the patient’s urine (BioSample: SAMN12621417) — Ser (S)/ Val (V) substitution in position 87, Asp (D)/ Tyr (Y) in position 91.

One mutation was found in two strains (C. amycolatum ICIS 53, C. amycolatum ICIS 99): Ser (S) replaced with Arg (R).

DISCUSSION

The spread of antimicrobial drug resistance genes by horizontal transfer causes the increase in the number of resistant microorganisms, including opportunistic pathogens. It is worth noting that Corynebacterium strains, such as C. amycolatum ICIS 53, C. amycolatum ICIS 9, C. amycolatum ICIS 99 isolates extracted from vaginal discharge of healthy women we have studied, had a large enough pool of antimicrobial resistance genes [9, 11]. In this regard, it is necessary to continuously monitor antimicrobial resistance of bacteria in order to develop effective measures against their growing resistance to antimicrobial drugs. The databases containing information about antibiotic resistance of bacteria will make it possible to compare the results obtained using different methods and estimate the abundance of antimicrobial resistance genes.

Our findings allowed us to single out a core set of antimicrobial resistance genes comprised in the Corynebacterium genomes. These data can be used as potential estimates of the use of antimicrobials for treatment of patients. However, molecular genetic testing should be combined with other methods based on phenotypic assessment of sensitivity to drugs, since the data on phenotypic and genotypic resistance are not always correlated.

Antimicrobial resistance can be associated with various mutations. In particular, quinolone/fluoroquinolone resistance is realized mainly through acquisition of point mutations in the sequence of gyrA gene encoding DNA gyrase A subunit, while overexpressed efflux pump can also contribute to acquisition of quinolone resistance [12, 13]. In C. amycolatum, alteration in the GyrA position 87 ensured resistance to all the tested quinolones/fluoroquinolones [12, 13]. Such substitutions were also observed in the genomes of C. amycolatum strains we had assessed. Furthermore, some Corynebacterium strains carried several mutations in the gyrA amino acid sequence, which increased the MICs of quinolones/fluoroquinolones [12, 13]. Investigation of genetic variability through mutation is important for the study of evolutionary transformation of bacterial genomes and can be used to develop rapid molecular diagnostic tests.

CONCLUSIONS

A growing etiological significance of Corynebacterium for infectious diseases, especially as hospital-acquired pathogens among immunocompromized patients having a history of the long-term hospital stay, several courses of antibiotic therapy and treatment with the use of invasive medical devices, determines the need to constantly monitor pathogens. Antimicrobial resistance of bacteria is a major concern: 1) it was found that there was a large pool of antimicrobial resistance genes (25 genes) forming various combinations in the Corynebacterium genomes. The presence of gene was correlated to the isolate capability of being resistant to antimicrobial drugs. This represented an important evolutionary effect of the impact of antibiotics on the population structure of microorganisms. It should be noted that antimicrobial resistance is most often encoded by several genes. Variability of antimicrobial resistance determinants emphasizes the need for continuous monitoring of the Corynebacterium resistance profiles; 2) mutations were detected in the gyrA amino acid sequences of the studied strains (positions 87, 88, 91), which were considered to be associated with quinolone/fluoroquinolone resistance.

The goal of the study was achieved. The limited data on Corynebacterium, including molecular genetic data, hamper comparative analysis. Expansion of the range of strains, including ones represented in various databases, will contribute to better understanding of the genome structure, phenotypic characteristics, while identification of the range of antimicrobial resistance genes will expand the knowledge about the directions of antibiotic therapy.

КОММЕНТАРИИ (0)