Molecular Characterization, Drug Resistance and Virulence Analysis of Constitutive and Inducible Clindamycin Resistance Staphylococcus aureus Strains Recovered from Clinical Samples, Tehran – Iran

Introduction

Staphylococcus aureus is an opportunistic pathogen causing various hospital and community-acquired infections ranging from pyogenic skin infections to life-threatening diseases.¹ According to the evidence, specific virulence factors such as i) cell surface components including collagen-binding protein, clumping factor, fibronectin-binding protein, and elastin-binding protein, ii) secreted factors, e.g. staphylokinase, toxic shock syndrome toxin (TST), hemolysin, exfoliative toxins, staphylococcal enterotoxins (SEs), lipase and panton-valentine leukocidin (PVL), play a pivotal role in pathogenesis and also are related to severity of the infection.^1,2 Recent studies have demonstrated an evolved resistance to various antibiotic types among S. aureus, which has raised real concerns.^3–6 Due to emerging simultaneous resistance to several antibacterial agents, the choice of chemotherapeutic options and treatment of serious infections caused by S. aureus has become problematic.^7–9 Although various antibiotics such as vancomycin, linezolid, and quinupristin-dalfopristin may be considered as drugs of choice, S. aureus strains with reduced susceptibility and resistance to these agents also emerged. Therefore, the use of macrolide-lincosamide-streptogramin B (MLSB) antibiotics as an alternative approach to treating such infections was taken into consideration.^3,4,6 However, some previous studies reported a constitutive and inducible resistance to clindamycin in S. aureus strains due to the indiscriminate use of MLS_B antibiotics.^4,6,10 Therefore, particular attention should be paid to the detection of constitutive and inducible clindamycin resistance genotypes and phenotypes to prevent clinical therapeutic failure in patients with S. aureus infections.¹¹ With respect to the limited data in this context, the present study was designed to describe the phenotypic and genotypic resistance pattern and the presence of the virulence factors. Coagulase (coa) typing and staphylococcal cassette chromosome mec (SCCmec) typing were used to characterize the genotype of the constitutive and inducible clindamycin resistance isolates.

Materials and Methods

Study Design, Isolation of Bacteria and Ethics Statement

This descriptive cross-sectional study was conducted on 204 nosocomial S. aureus strains recovered from clinical samples such as pus (36.3%), wound (31.9%), blood (13.7%), sputum (7.3%), cerebrospinal fluid (5.9%), and urine (4.9%) during the period of 1 year, from August 2018 to July 2019. The Ethics Committee of the Shahid Beheshti University of Medical Sciences in Tehran, Iran certified the protocol of this project (IR. SBMU. MSP.REC. 1396.700). At first, S. aureus isolates were phenotypically identified using standard microbiological and biochemical techniques and then were subjected to polymerase chain reaction (PCR) assay for the presence of nuc gene for definitive confirmation.¹² A total of 86 nosocomial S. aureus isolates were included in the study based on resistance to erythromycin and resistance and/or susceptibility to clindamycin in accordance with standard clinical and laboratory standard institute (CLSI) guidelines. Isolates with resistance to both clindamycin and erythromycin-resistant were considered to be constitutive resistance phenotype (cMLS_B). Isolates with resistance to erythromycin but susceptible to clindamycin were tested by the D test. Inducible resistance phenotype (iMLS_B) was defined for isolates showing resistance to erythromycin and susceptible to clindamycin with a D-shaped zone around the clindamycin disk, flattened from the side of erythromycin disk. Isolates with both inhibition zones showing a circular shape (D test negative) were classified as the MS phenotype (CLSI 2019).

Antimicrobial Activities

The disk diffusion method using cefoxitin (30 μg) disk in Mueller-Hinton agar (Merck, Germany) according to the CLSI was applied for the screening of methicillin resistance isolates. In addition, PCR assay was used for the detection of mecA gene.¹²

The Kirby–Bauer disk diffusion method was used to determine the susceptibility of the isolates against penicillin, ceftriaxone, amikacin, gentamicin, tobramycin, kanamycin, tetracycline, linezolid, teicoplanin, ciprofloxacin, rifampicin, quinupristin-dalfopristin, and trimethoprim-sulfamethoxazole (Mast Co., UK) based on the CLSI recommendation (CLSI 2019). The minimum inhibitory concentration (MIC) value for vancomycin, mupirocin, tigecycline, and fusidic acid was determined using the broth microdilution method. Results for fusidic acid and tigecycline were interpreted according to the European Committee for antimicrobial susceptibility testing (EUCAST) breakpoints (http://www.eucast.org). Low-level and high-level mupirocin resistance (LLMUPR, HLMUPR) were defined if MIC values of 8–256 µg/mL and ≥512 µg/mL were obtained. S. aureus strains ATCC 25923, ATCC 43300 and ATCC 29213 were used as reference strains.

DNA Extraction

In this study, the DNA of each strain was extracted using the phenol-chloroform extraction method with the modification of adding lysostaphin (Sigma-Aldrich, United States) for bacterial lysis. DNA concentration and purity were investigated using a NanoDrop 2000 spectrophotometer (spectrophotometer, Thermo Scientific, Wilmington, DE, USA).

Amplification of Resistance-Related Genes

The resistance encoding genes vanA, vanB, mupB, mupA, fusA, fusB, fusC, mecC, msr(A), msr(B), erm(A), erm(B), erm(C), tet(M), ant (4΄)-Ia, aac (6΄)-Ie/aph (2˝), aph (3΄)-IIIa and virulence determinants including exfoliative toxin (eta, and etb), Panton-Valentine leukotoxin (pvl), and toxic shock syndrome toxin (tst) genes were detected by PCR.^12,13 In order to detect target genes, PCR amplified products were resolved using electrophoresis in 1.8% (w/v) agarose gel, stained with ethidium bromide (0.5 mg/mL) and visualized under UV light using a gel documentation system (Bio-Rad Laboratories, USA).

SCCmec Typing

For methicillin-resistant S. aureus (MRSA) strains, determination of SCCmec types was performed using multiplex PCR assay, as previously described by Nezhad et al S. aureus ATCC 10442 (type I), S. aureus N315 (type II), S. aureus 85/2082 (type III), S. aureus MW2 (type IV), S. aureus WIS 173 (type V) and S. aureus HDE288 (type VI) were recruited as control strains.

Staphylocoagulase (SC) Typing

Multiplex PCR assay with four-set primers (A-D), was used for determination of SC types (I–X) as previously described by Hirose et al Set A contained primers for identification of SC types I, II, III, IVa, IVb, Va, and VI while set B contained primers for identifying SC types VII, VIII, and X. Set C was used to identify SC types IX and Vb. SC types IVa and IVb were distinguished using set four primers (Set D).¹⁴

Results

Out of 204 S. aureus tested isolates, the overall prevalence of cMLS_B, iMLS_B, and MS phenotypes were 47 (23%), 29 (14.2%), and 10 (4.9%), respectively. Out of 86 tested isolates, 25 isolates were obtained from hospital H1 (29.1%), 20 isolates from hospital H2 (23.3%), 18 isolates from hospital H3 (20.9%), and 23 isolates from hospital H4 (26.7%). The under-study isolates were recovered from wound (28/86, 32.6%), pus (21/86, 24.4%), blood (19/86, 22.1%), sputum (7/86, 8.1%), CSF (6/86, 7%), and urine (5/86, 5.8%). According to our analysis, the rate of invasive and non-invasive S. aureus was found to be 29.1% and 70.9%, respectively. All the invasive S. aureus isolates were methicillin-resistant with iMLS_B (40%; 10/25), MS (40%; 10/25) and cMLS_B (20%; 5/25) phenotypes. A total of 86 S. aureus strains included in present study, 47 were isolated from female patients (54.7%) and 39 were recovered male patients (45.3%) with a median age of 41.4 years, ranging from 15 to 59 years. Of these examined 86 S. aureus strains, 75.6% (65/86) and 24.4% (21/86) were MRSA and methicillin-sensitive S. aureus (MSSA), respectively. The cMLS_B phenotype was observed in both MRSA (26, 30.2%) and MSSA (21, 24.4%) strains, whereas phenotypes of iMLS_B and MS were found only in MRSA strains. The most common S. aureus isolates with iMLS_B phenotype in the present study were isolated from wound infection (19.8%, 17/86) while cMLS_B phenotypes were from pus (22.1%, 19/86) and MS phenotypes were from blood infection (9.3%, 8/86). Distribution of different MLS_B phenotypes in S. aureus strains isolated from various clinical specimens is shown in Table 1. Based on data obtained from the disk diffusion test, all isolated S. aureus strains were found to be susceptible to linezolid, teicoplanin, and vancomycin. As summarized in Table 2, resistance to all antibacterial agents (but no amikacin and fusidic acid) was more common among MRSA isolates than among MSSA isolates. All the four of the fusidic acid-resistant isolates were MSSA isolates with cMLS_B phenotype, based on the results of the micro-broth dilution method. The rate of resistance to mupirocin was found to be 9.3%. Six isolates indicating high-level resistance to mupirocin (HLMUPR) belonged to MRSA strains with iMLS_B (5 isolates) and MS (one isolate) phenotypes and two isolates showing low-level resistance to mupirocin (LLMUPR) were identified as MRSA strains with iMLS_B phenotype. Two tigecycline resistant isolates belonged to two MRSA strains with iMLS_B and MS phenotypes.

Table 1 Distribution of iMLSB, cMLSB and MS Phenotypes in 86 Nosocomial S. aureus Strains Isolated from Clinical Sources

Table 2 Antimicrobial Resistance Pattern of Nosocomial MRSA and MSSA Isolates with Inducible and Constitutive Phenotype

Our findings indicated that all isolates were typed using the SC typing method. These isolates were distinguished into 4 types of SC. The predominant SC type was II (34.9%, 30/86), followed by III (32.6%, 28/86), V (20.9%, 18/86), and I (11.6%, 10/86). All the MSSA isolates belonged to SC type III. All MS phenotype MRSA isolates belonged to SC type I. Among iMLS_B phenotype MSRA isolates, the most predominant SC types were II and V representing 25.6% (22/86) and 8.1% (7/86) of isolates. cMLS_B phenotypes were distributed in SC types III, II, and V accounting for 32.6%, 9.3%, and 12.8%, respectively.

Resistance encoding genes analysis showed that the most prevalent gene was mecA (75.6%, 65/86), followed by tet(M) (50%, 43/86), erm(C) (40.7%, 35/86), ant (4΄)-Ia (29.1%, 25/86), aac (6΄)-Ie/aph (2˝) (20.9%, 18/86), msr(A) (20.9%, 18/86), aph (3΄)-IIIa (14%, 12/86), erm(B) (14%, 12/86), msr(B) (9.3%, 8/86), erm(A) (8.1%, 7/86), mupA (7%, 6/86), fusB (3.5%, 3/86), and fusC (1.2%, 1/86). Our findings showed that no PCR products were found for the resistance genes vanA, vanB, mupB, fusA, and mecC. The distribution of SCCmec types in the 65 MRSA isolates showed that SCCmec type III was the most prevalent type found in 33 isolates (50.8%), followed by type II in 20 isolates (30.8%), and type IV in 12 isolates (18.4%). SCCmec types I and V were not found in our isolates. The resistance profile and distribution of coa and SCCmec types in MRSA and MSSA isolates with inducible and constitutive phenotype are presented in Table 3.

Table 3 Distribution of coa, SCCmec Types and Resistance Profiles of Nosocomial MRSA and MSSA Isolates with Inducible and Constitutive Phenotype

Discussion

Little is known about the emergence, distribution and molecular types of constitutive and inducible clindamycin resistance S. aureus strains in Iran. Accordingly, our study focused on the identification of molecular characteristics and understanding of cMLS_B, iMLS_B S. aureus isolates epidemiology. Our research highlighted several new findings in relation to MRSA and MSSA isolates with inducible and constitutive resistance phenotype including a relatively high prevalence of inducible and constitutive resistance and distinct molecular types with genetic diversity. According to the evidence, the prevalence rate of iMLS_B phenotype among S. aureus isolates was markedly varied across the geographical region and among health-care settings. The current finding showed a prevalence rate of 14.2% for iMLS_B which is higher than the reported rate in Nepal (11.48%),³ Iran (8.6%),⁵ Egypt (7.7%),¹⁵ Turkey (7.8%)⁴ and lower than those reported in India (37.5%).¹⁶ However, the results of previous studies conducted in Iran noted significant variation in iMLS_B prevalence rates in different areas^10,17 ranging from 6% to 32.3%. This work presented a relatively high prevalence of iMLS_B phenotype, which highlighted the need to prescribe of macrolides in a logical manner, in order to change in resistance pattern. However, the true evaluation of iMLS_B S. aureus isolates prevalence depends on the accurate diagnosis, geographical variation, characteristic of health-care setting, and population under study.

In the current research, the prevalence of cMLS_B among S. aureus was found to be 23%, which was similar to reported rate, by Delialioglu et al (24.3%),⁴ and Eksi (20.4%).¹⁸ However, a lower and higher percentage of cMLS_B were also reported in previous studies performed by Khashei et al (82.9%),⁵ Adhikari et al (29.25%),³ Mansouri et al (28.4%),¹⁹ Sedaghat et al (32.1%),¹⁷ and Sasirekha et al (13.1%).⁶ These variations in the prevalence of cMLS_B among S. aureus in different parts of the world could be attributed to the difference in consumption of macrolides in community and hospital settings, study design, population and geographical distribution, and the spread of specific molecular types. As presented in Table 2, the cMLS_B phenotype was higher in MRSA (30.2%) as compared to MSSA (24.4%) and iMLS_B and MS phenotypes were found only in MRSA strains, which is consistent with previous reports from Nepal,³ Iran,¹⁹ Egypt,¹⁵ and Turkey.⁴

In our study, low resistance rate to mupirocin (9.3%) was noted which is in agreement with other studies conducted in Iran (6%),⁷ India (5%),⁸ and Jordan (2.6%).²⁰ Furthermore, the findings of the present study demonstrate that 7% of tested isolates were found to have resistance to mupirocin at a high level which are quite similar to the results of a study conducted by Liu and colleagues in China. They reported a prevalence rate of 6.6% for isolates with high-level mupirocin-resistant.²¹ However, there has been a higher prevalence of HLMUPR strains in Iran (25%)²² and Egypt (61.5%).²³ Different results of these studies may be due to study design (patient characteristics and specimen types), specific type dissemination among patients and unrestricted policies in taking mupirocin. This study showed that all HLMUPR isolates were mupA-positive (7%). González-Domínguez et al (27.2%)²⁴ and Abbasi-Montazeri et al (34%)⁷ reported a higher percentage of mupA. Shahsavan et al reported that mupA was responsible for the resistance to mupirocin only in MRSA strains with cMLS_B phenotype. Conversely, in the present research, this gene was present in MRSA strains with iMLS_B and MS resistance phenotype.

Tigecycline is a reliable treatment option against many infections caused by MDR isolates especially MRSA. So far, there have been few reports published on the emergence of S. aureus strains with reduced susceptibility and resistance to tigecycline. We found the low numbers of tigecycline resistant isolates (2.3%). However, different resistance rates to tigecycline among S. aureus isolates are reported in Libya (3.6%),²⁵ Turkey (2%),²⁶ and Iran (6.6%).⁹

Recently published data from Asian countries indicated a low prevalence of resistance to fusidic acid (<10%).^27,28 We found a low prevalence (4.7%) of resistance to fusidic acid among our isolates carrying fusB (3 isolates), and fusC (1 isolate) genes. This observation is consistent with data from a recent multicenter study in Iran that showed a low prevalence of MRSA fusidic acid-resistant (3%) among 726 studied S. aureus isolates.²⁸ Various resistance rates to fusidic acid have been described in many countries such as Greece (62.4%), Ireland (19.9%), Australia (7.0%), Canada (7.0%), and the United States (0.3%).²⁹ Notably, this study showed that fusidic acid resistance was only seen among MSSA isolates which was in contrast to a report from China that indicated the prevalence of fusidic acid resistance among MRSA isolates was higher significantly than that among MSSA isolates.³⁰ Yu and colleagues also reported a 10.5% incidence of fusB genes while fusC and fusA genes were not detected in any of the isolates examined.³⁰ This indicates that fusB is the predominant determinant responsible for resistance to fusidic acid among MSSA isolates with cMSL_B resistance phenotype in Iran.

ant (4΄)-Ia as the most prevalent aminoglycoside resistance gene was present in 25 strains (29.1%) which was higher than those reported in Turkey (24%)³¹ and lower than that reported rate in India (9%).³² However, much higher rates have also been reported by Nezhad et al (94.7%)¹² and Ida et al (84.5%).³³ In our research, the prevalence of aac (6΄)-Ie/aph (2˝) as the second commonly detected aminoglycoside resistance gene was found to be 20.9% which is lower than that reported by Ardic et al (60.5%).³¹ Akpaka et al reported that in MRSA strains, 20% harbored ant (4΄)-Ia gene.³⁴ Many reports from different parts of Asia such as Turkey (8%),³¹ and India (9%)³⁴ have shown a low prevalence of the ant (4ʹ)-Ia gene. Likewise, in the current work, 14% of isolates were found to carry aph (3΄)-IIIa. This variation in aminoglycoside resistance determinant frequency displayed that factors such as study design, specimen types, different policies in aminoglycosides consumption, and horizontal gene transfer among the strains may be involved.

Based on the literature, inducible and constitutive resistance in S. aureus strains is mediated by both erm and msr genes. Our analysis recorded erm(C) (40.7%) and erm(A) (8.1%) as the highest and lowest erythromycin resistance gene. Although there is a discrepancy in the prevalence rate of erm(C) gene in different parts of the world, a similar finding from Iran has also been reported.¹⁷ These findings are in concordance with those described by Fasihi et al, which reported prevalence of erm(C) and erm(A) genes to be 20.5% and 11% for MRSA strains.³⁵ In accordance with our results, Schmitz and colleagues reported that the erm(A) gene was more common in MRSA isolates compared to MSSA isolates (88% vs 38%) and occurred mainly in cMLS_B expression strains.³⁶ A high percentage of msr(A) (20.9%) was obtained in the current research, similar to a study conducted by Sedaghat et al (43.6%),¹⁷ Nezhad et al (47.3%).¹² In accordance with several investigators who reported the prevalence of erm(B) gene at a low level, we detected erm(B) gene in 5.8% and 8.1% inducible and constitutive resistant strains, respectively. A study from Texas also showed a high frequency of erm(B) gene in inducible resistant strains (46.3%).³⁷ This discrepancy could be because of the distribution of some specific clonal lineages in communities and hospitals and might be more closely related to the usage of particular macrolides and ketolides in our health-care settings.

Our data related to SCCmec types are in concordance with many studies that demonstrated SCCmec types I, II, III are related to hospital-acquired S. aureus infections while IV and V are prominent types in community-acquired S. aureus infections.^12,21,24 We confirmed SCCmec type IV in MRSA isolates with cMLS_B (5.8%) and iMLS_B (8.1%) phenotype. This finding supports a shift in these isolates from our community to hospital.

According to the coa typing results, predominant SC type was II (34.9%), followed by III (32.6%), V (20.9%), and I (11.6%). This was in comparison to the previous report by Hirose et al in Japan which indicated coa type II, VII and I accounted for 91.9%, 3.9% and 1.7% of isolates.¹⁴ We detected 2 SC types (II–V) among iMLS_B S. aureus strains suggesting the clonal distribution of tested isolates in this region of Iran. In research involving 157 S. aureus strains from clinical specimens, nine different patterns of coa gene have been detected.³⁸ In another study from Thailand on 129 MRSA isolates from 17 hospitals, Janwithayanuchit et al determined four different genotypes from coagulase gene typing of tested strains.³⁹ They showed that the most prevalent coa type was III (82.2%) followed by IV (14%), II (2.3%), and I (1.5%). The analysis was carried out by Younis Omar et al on 75 MRSA isolated from different ICUs grouped into three different types based on the polymorphism of coa gene products by PCR.⁴⁰ The similarity between coa genes of examined strains highlighted that it may be as a predictor for specific inducible resistant S. aureus strains.

Conclusion

Given the presence of various types of MLS_B resistance in our survey, special attention should be given to diagnosing these resistance types in order to judicious use of clindamycin. However, some resistance patterns were related to certain SCCmec and coa types. These strains have low genetic variability with a predominance of coa type II and SCCmec type III. Further researches should be performed in other regions of Iran to keep track of the emerging coa types.