Background
Corynebacterium genus consists of Gram-positive aerobic or anaerobic facultatively pleomorphic rods with a high G + C content DNA. Some species are part of human skin or mucosa [
1,
2].
Corynebacterium striatum has been increasingly associated with severe infections in both immunocompetent and immunocompromised hosts [
3,
4]. However,
C. striatum isolates have been included among the etiologic agents of bacteremia with or without central venous catheter (CVC) in place [
5,
6], endocarditis [
7], breast abscesses [
8], septic arthritis [
2], osteomyelitis [
4] and several other invasive diseases. In addition, studies have evidenced
C. striatum as an emerging multidrug-resistant (MDR) pathogen related to nosocomial outbreaks in several countries [
9‐
16].
The main risk factors for acquisition MDR
C. striatum infections highlighted by Verroken et al. [
13] were: prolonged hospital stay, advanced stage of chronic obstructive pulmonary disease, recent administration of antibiotics and exposure to an invasive diagnostic procedure. Besides, empirical antibiotic therapy may select MDR Gram-positive skin flora that can become the etiologic agent of nosocomial invasive diseases [
17]. The emergence of MDR
C. striatum and its involvement in nosocomial infections require appropriate interpretive criteria to the selection of the adequate antibiotic therapy [
16].
Most reports of nosocomial infections and outbreaks caused by
C. striatum mainly encompassed the respiratory tract [
10,
13,
18]. On the other hand, few studies have investigated bloodstream and catheter-related infections by
C. striatum [
15,
19‐
21]. In a Brazilian tertiary care hospital located at Rio de Janeiro metropolitan area, a nosocomial outbreak caused by MDR
C. striatum mostly isolated from tracheal aspirates samples was initially verified in 2009 [
12]. Subsequently, cases of bloodstream and catheter-related infections caused by
C. striatum isolates were noticed in the same hospital. In the present study, we aimed to investigate the clonal relationship, antimicrobial susceptibility profiles, ability of biofilm formation, molecular detection of resistance genes to amynoglycosides, quinolones, compounds of the MLSB group (macrolides and lincosamides) and chloramphenicol of these
C. striatum invasive isolates.
Discussion
Antimicrobial resistance has a major impact on human health [
36]. At present, reports on the emergence and spread of multiresistant bacterial species are important to support the progress of resistance control policies. Our data show bloodstream and catheter-related infections caused by different clones of MDR
C. striatum in Brazil. In accordance to Chen and co-workers [
37], our findings emphasize that
C. striatum from blood and catheter segments should not be considered only as contaminant, since in our study most of the isolates were found in pure cultures (82%) or in significant numbers.
In Brazil the isolation of
C. striatum from hospitalized patients with signs and symptoms of infection was observed in some studies [
38‐
40]. In a previous study, we documented four PFGE profiles during a nosocomial outbreak caused by
C. striatum in Rio de Janeiro, Brazil. MDR clones related to the profiles of PFGE I and II were predominant. In that opportunity only two isolates of PFGE I and II profiles were isolated from blood samples [
12]. Due to subsequent increased number of cases of bloodstream and catheter-related infections caused by
C. striatum isolates in HUPE, current investigation of the clonal relationship of these
C. striatum isolates revealed the permanence of the MDR PFGE profiles I and II in the nosocomial environment as invasive clones. However, the PFGE I profile was found predominant among patients with hematogenic infections. In addition, other new MDR PFGE profiles (V to IX) emerged as etiologic agents of bloodstream and catheter-related infections. Interestingly, one non-multiresistant clone (PFGE profile X) was also related to a case of catheter-related infection in the newborn.
PFGE is a valuable tool to investigate the clonal relatedness of microbial strains during nosocomial outbreaks. Several nosocomial infections and outbreaks studies have employed this methodology for typing of
C. striatum isolates [
13,
15,
16,
18]. PFGE is a stable and reproducible genotyping method, however, it is time consuming and standardizations for inter-laboratory comparisons do not exist for
C. striatum isolates genotyping. Thus, the method is only applicable to compare isolates for regional epidemiology surveillance. Gomila and co-workers described the development of a multilocus sequence typing (MLST) scheme for
C. striatum. [
9]. However, the proposed MLST scheme has not been adopted by scientific community, perhaps due to the limited number of genes (ITS1 region,
gyrA and
rpoB) that comprise it. PFGE has limitations but it is the tool available so far for the discrimination of this bacterial species. Future researches are required to evaluate genotyping methods that provide useful data for global surveillance of infections caused by
C. striatum.
Antimicrobial susceptibility testing remains rarely performed on
Corynebacterium spp. in many laboratories [
41]. The method of susceptibility by disk-diffusion is widely used by microbiology laboratories in Brazil and in other countries [
12,
42‐
44]. Moreover, CLSI guidelines do not provide breakpoints for disk-diffusion while BrCAST, based in EUCAST document, provides breakpoints for corynebacteria susceptibility testing only for some antibiotics [
28,
29], excluding various antimicrobials, such as cephalosporins, carbapenems and lipopeptides, thus many researchers often use staphylococcal breakpoints [
12,
16,
45]. Susceptibility testing should be performed on clinically significant
C. striatum isolates. However, studies suggest that according to the severity of the infection, empiric treatment should be carried out with vancomycin and linezolid because of low levels of susceptibility to other antimicrobials [
13,
46]. In some studies, therapy with an association of at least two of the following antimicrobial agents has been reported: vancomycin, rifampin, linezolid and daptomycin [
47,
48].
Resistance to imipenem by
C. striatum isolates has been related in some countries, such as Japan [
49], Spain [
9,
10] and Italy [
11]. Most of our
C. striatum isolates showed resistance to imipenem (76,2%). Combination therapy that includes imipenem for the treatment of MDR
C. striatum infections in Brazil should be more prudent.
Effective treatment of corynebacterial infections with daptomycin has been reported in the literature [
48,
50]. In this study, all
C. striatum isolates showed susceptibility to daptomycin. However, resistance to daptomycin in
C. striatum has been documented and may occur during therapy in patients with invasive infections. Consequently, some authors recommend caution in daptomycin monotherapy for treatment of these infections [
41,
51].
Some mechanisms of antimicrobial resistance have been reported in
Corynebacterium species. Studies of the sequences encoding the A subunit of the gyrase enzyme in strains of
C. striatum, Corynebacterium amycolatum and
Corynebacterium macginley have shown that resistance to fluoroquinolones is associated with mutations of a spontaneous nature in this gene and depends on the number of mutations and the type of amino acid that was exchanged [
16,
32,
52,
53]. Twenty-two isolates were resistant to the quinolones tested. The combinations of amino acids Val/Asn and Tyr/Asp in positions 87 and 91 of QRDR region
gyrA gene, respectively, found in three of our isolates have not been described in the literature for
Corynebacterium species until the present moment, conferring resistance to ciprofloxacin and moxifloxacin.
The
ermX gene (
erythromycin ribosome methylation) encoding the rRNA methylase enzyme leads to simultaneous resistance to macrolides, lincosamides and streptogramins B (MLSB) [
45,
54]. This gene was found on chromosomes, plasmids and transposons of corynebacteria. We found this gene in 22 isolates, except in the MDS 2376 isolate, which may indicate that the
ermX gene can be involved in the clindamycin and erythromycin resistance phenotype of our isolates.
Antimicrobial susceptibility studies have shown
C. striatum isolates resistant to aminoglycosides. Consequently, the use of aminoglycosides as second-line complementary antimicrobial for treatment of
C. striatum infections should be cautious [
55,
56]. The mechanisms of aminoglycoside resistance most common are the aminoglycoside modifying enzymes (AME’s) classified in 3 classes:
aac (
acyl-coenzyme A-dependent acetyltransferase),
ant (
nucleoside triphosphate-dependent nucleotidyl transferases) and
aph (
nucleoside triphosphate-dependent phosphotransferases). These enzymes are often disseminated by various mobile genetic elements and many aminoglycosides can be inactivated by more than one enzyme [
57]. In this study, the
aph gene was found in 22 isolates, except in the MDS 2376 isolate, but one
aph-positive isolate showed susceptibility to gentamicin, confirmed by MIC E-test strip (MIC = 0.12 mg/L, according to BrCAST). Future analyzes of the region where the gene is located and other mechanisms of resistance to aminoglycosides in all
aph-positive isolates should be made.
The
cmx gene is responsible for coding the efflux protein to chloramphenicol and has already been found in transposons, plasmids and genomes of
Corynebacterium species [
58‐
60]. All isolates
cmx-positive were resistant to chloramphenicol. The
cmx gene sequence showed similarity above 99% with the sequences of
cmx gene found in chromosomes, plasmid and genomic island of
Pseudomonas aeruginosa and other
Corynebacterium species deposited in GenBank/NCBI.
Biofilm is a structure that facilitates several bacterial processes influencing virulence and resistance to antimicrobials such as adhesion capacity, metabolite exchange, cellular communication, protection to antimicrobials, protection against host immune attacks. Consequently, the formation of bacterial biofilms leads to an increase in healthcare costs and extend hospitalization [
61,
62]. Previously, Souza and co-workers verified that
C. striatum PFGE profiles I to IV formed biofilm on hydrophilic and hydrophobic surfaces.
C. striatum PFGE profile I, predominant isolated from nosocomial outbreak, showed the greatest ability to adhere to all surfaces produced much more biofilm than the other profiles [
35]. In Japan, Qin and co-workers observed that all 6
C. striatum isolates identified as predominant PFGE profile had high ability to produce biofilm in glass cover-slips after 72 h post-incubation [
15]. In the present study, biofilm formation and survival on four abiotic surfaces (glass, metal, polyurethane and silicone) were demonstrated 48 h post-incubation of bacterial cells representative of MDR
C. striatum PFGE profiles I and II isolated from patients with bloodstream infections. Similar to
C. striatum PFGE profile I isolated from patients undergoing endotracheal intubation procedures, PFGE profile I isolated from bloodstream and catheter-related infections also showed a higher ability to adhere to and to survive on abiotic surfaces of medical devices including those used in invasive procedures. MDR
C. striatum viable cells were able to multiply and to produce mature biofilms on both types of catheter surfaces.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.