Determination of Characteristics of Erythromycin Resistant Streptococcus pneumoniae with Preferred PCV Usage in Iran

Amongst 100 Streptococcus pneumoniae isolated from clinical cases and nasopharynx of healthy individuals, 60 erythromycin resistant strains were isolated and characterized using MLST, PFGE, transposon analysis and Quellung reaction. Most of the S. pneumoniae erythromycin resistant (80%) were found to be attributable to the ermB-edncoded ribosome methylase activity which differs from the dominant mechanism of macrolide resistance seen in North America. The most predominant transposons were; Tn1545/6003 (27%), Tn6002 (22%), Tn2009 (20%), Tn2010 (17%). Number of the clinical isolates carrying Tn2010 was more significant than the normal flora. The serotypes found were; 14 (33%), 3 (22%), 23F (15%), 19F (15%), 19A (7%), 6A (3%), 9V (3%) and 6B (2%). The most prevalent serotypes among the clinical (n = 28) and normal flora (n = 32) isolates were serotypes 14 (46%) and 3 (31%), respectively. The most prevalent vaccine serotypes amongst the clinical isolates and the healthy individuals were pneumococcal conjugate vaccines (PCV) 13 and PCV10, respectively. PFGE revealed 34 pulsotypes with 9 common and 25 single types. Significant number of the normal isolates belonged to CT5 and CT6. On the other hand, significant number of clinical isolates belonged to CT8 as compared to the normal flora isolates. MLST showed 2 dominant sequence types. ST3130 (23%) and ST180 (22%) were the most predominant sequence types in the clinical and normal isolates, respectively. There was no significant difference in other sequence types between clinical and normal flora isolates. Three polyclonal complexes including Sweden15A -25, Spain23F-1 and Spain9V-3 constituted 58% of the isolates. Our results suggest that the genetic diversity and transposon distribution were high among S. pneumoniae, particularly in the isolates containing erm(B) and double antibiotic resistant genes (erm/mef). The results presented here could influence the change in the current vaccination practices in Iran which currently calls for vaccination with PCV7 or PCV10.

Introducation Streptococcus pneumoniae is a common bacterial pathogen responsible for a numerous community acquired (CA) infections especially among children and elderly [1]. The pneumococcal illness varies from self-limiting mucosal infections to life-threatening invasive pneumococcal diseases [2] and eventual appearance of antibacterial resistance strains.
Macrolides are the most frequently prescribed antibiotic among CA pneumococcal infections [3]. Macrolide-resistance (MR) in pneumococci is mediated by two major mechanisms including target site modification by methylases, encoded by ermB (MLS B phenotype) and efflux pump mechanism mediated by mef (M phenotype) genes with various transposons [4][5][6].
A recent trend of the rise of multiple drug resistant (MDR, defined as resistance to at least 3 antimicrobial classes) clinical pneumococcal strains expressing both ermB and mef carried on Tn2010 [similar to Tn2009 with an additional ermB] has been reported [5]. These isolates are clonally disseminated and often carry resistance to macrolide, lincosamide and streptogramin B (MLS B ) [7].
The aim of this study was to compare serotyping, antibiotic resistance, clonal diversity and the presence of genetic elements in the erythromycin resistant S. pneumoniae (ERSP) isolates obtained from the clinical cases and healthy unvaccinated PCV individuals in Tehran, Iran.

Sample collection and identification
A total of 100 clinical and normal flora pneumococcal isolates (50 each) were collected from hospitals and private clinical laboratories in Tehran, Iran in a two-year study (2011)(2012)(2013). The collected patients' samples were from confirmed cases of meningitis, pneumonia, or bacteremia. None of the patients or healthily subjects received any forms of PCV. The normal flora isolates were collected from healthy nasopharynx individuals who did not have any antibiotic treatments for at least 6 months preceding to the sampling and had no serious nasopharynx infections. All procedures were done in accordance with the Declaration of Helsinki (1975) amended in 2013 [8]. The study was approved by the Ethics Committee of Iran University of Medical Sciences. For privacy reason, the identity of patients and clinical laboratories form whom/where the isolates had been collected remained anonymous throughout the study. A written consent, however, was obtained from the control volunteers. Standard microbiological techniques were performed for species identification, including hemolysis, Gram staining, bile solubility and susceptibility to optochin (1μg) disc [9], and identification of isolates was confirmed by lytA and ply genes using species-specific primers for polymerase chain reaction (PCR) [10]. Serotyping was performed using the Quellung reaction with antisera (Statens Serum Institut Copenhagen, Denmark).

Macrolide resistant phenotype
MR phenotype was determined by the double disk diffusion method with erythromycin and clindamycin discs (D test) to differentiate constitutive and inducible resistance phenotype [7].
Erythromycin and tetracycline resistance genes PCR amplification of MR genes including ermB, mefA/E (mef), tetM and transposon related genes were done using previously published papers [4,12].

Statistical analysis
Differences in the proportions among categorical variables were assessed using Fisher's exact test or the χ 2 test as appropriate. Where needed statistical analysis was performed between the patient and normal flora samples.

Bacterial isolates
In all, 60 (60%) ERSP were isolated with 28 (47%) and 32 (53%) isolates belonging to the clinical and normal flora groups, respectively. The number of ERSP isolates (clinical and normal flora) obtained was significantly higher in the individuals with <15 (76.5%) years old than in adults) !15 years, P<0.05). The rate of ERSP isolates was higher among the invasive (75%) than non-invasive (43%) clinical isolates.

ERSP phenotypes
On the basis of the erythromycin-clindamycin double disk test and MIC, 48 ERSP isolates were assigned to MLS B phenotype (MIC >256 μg/ml) (80%) and 12 isolates were assigned to M phenotype (MIC of 1.5 to 16 μg/ml) (20%) (Fig 1). No significant difference in the distribution of MLS B phenotype amongst the normal (54%) and clinical isolates (46%) was found. On   the other hand, M phenotype was significantly present in the normal (67%) than the clinical isolates (33%).

Erythromycin resistant genes distribution
No significant difference in the overall rate of erythromycin resistant genes was found between the clinical and normal flora isolates. The ermB, ermB/mef and mef were carried in 50%, 36% and 18% of the clinical isolates, respectively. The predominant double resistant genes, ermB/ mef, in the clinical isolates were found in serotype 14 (60%) followed by 19F. Amongst the ERSP nasopharynx normal flora, 47%, 28% and 25% were ermB, ermB/mef and mef, respectively. In the normal flora isolates, the major serotypes carrying double resistant genes (ermB/ mef) were 19A and 3 (22% each) (Fig 1).

Transposons distribution
Amongst the total 60 ERSP isolates, Tn1545/Tn6003 (27%), Tn6002 (22%) and Tn2009 (20%) were the most dispersed transposons (Table 2). Tn2010 was the most common transposon in 29% of the clinical isolates followed by Tn1545/6003 (25%). Significant differences between the clinical and normal flora isolates carrying Tn2009 and Tn2010 were observed. The least common type of transposons was Tn1116 and Tn3872+MEGA (3% each) in both clinical and normal flora isolates (Table 2). No significant correlation was found between the type of transposon and the antibiotic resistance pattern (Table 3) in the clinical nor normal flora isolates. Resistance to kanamycin was found only in the isolates carrying Tn1545/Tn6003. PFGE PFGE analysis of total 60 ERSP isolates showed the presence of 34 pulsotypes with nine common types (CT) constituting 35 isolates and 25 single isolates (Fig 1). With the exception of serotype 3, no correlation was found between the serotype and PFGE type in the clinical or normal flora isolates. All serotype 3 isolates (clinical and nonclinical isolates) were found to be belonged to a single pulsotype (CT5).
The results showed that all of the CC180 were serotype 3 MLS B phenotype (Fig 1) which were mostly isolated from the normal flora (77%). No new alleles were found in our isolates.

Discussion
In the present study, the overall rate of ERSP was 60% which did not significantly differ between the clinical and nasopharynx normal flora isolates. The presence of high number of MDR ERSP isolates (98%) was an indicative of indiscriminate use of antibiotics in general, and macrolide in particular, in clinical practices with eventual dissemination of the resistant strains in Iran. Similar to the results obtained in Asia and some European countries [3,[14][15][16], we also found that the most common mechanism (48%) for macrolide resistance in our pneumococcal isolates was ermB-encoded ribosome methylase activity. On the contrary, efflux mediated pump gene (mef), which has been reported to be the dominant macrolide resistant in North American countries [17][18] and Germany [19], was found only in 20% of our isolates. Double resistant genes were present in 32% of our isolates which was significantly higher than some  Serotype 3130  180  81  63  166 557 2678 878  6354 337 199  271  329  926  1152  2253  3772  5004  7577  Total   14  6  0  0  4  2  2  3  0  1  0  0  0  0  0  0  1  1  0  0  20   19A  0  0  0  0  1  0  0  0  0  0  1  1  0  0  0  0  0  1  other investigators [20]. The prevalence of dual resistance genes in the pneumococcal isolates amongst the macrolide resistant isolates has been reported around 16% globally [17,21] and up to 63% in China in recent years [22]. Some degree of genetic linkage was observed in this study. Our pneumococcal isolates carrying mef were found to be associated with 7 different serotypes (out of 8 serotypes identified), suggesting mef to be less discriminative than ermB which was found in 5 different pneumococcal serotypes (mostly in serotypes 3 and 14). Furthermore, it was observed that all isolates carrying Tn2010 (clinical or normal flora) also harbored double resistant genes (ermB/mef). On the other hand, no association between the serotypes and transposons or the source of isolation was found. Moreover, we found that many of our isolates could carry wide range of transposons. For examples, serotype 14 in both clinical and normal flora isolates could carry all major transposons including Tn6002, Tn1545/6003, Tn1116, Tn2009 and Tn2010. In contrary to the suggestion by Chancey and colleagues [23] that the genetic linkages of Tn2010 with Tn2009 and Tn6002 may exist, we did not found any coexistence of transposons in our isolates. Table 2 shows significant difference (p<0.01) in the rate of Tn2010 in the clinical as compared to normal flora isolates. On the other hand, significant number of the normal flora carried Tn2009 than the clinical isolates.
Comparative analysis between the report from Asian for Surveillance of Resistant Pathogens (ANSORP) on S. pneumoniae isolated from 11 Asian countries [24], which is the most comprehensive study done so far in the Asia, with the present investigation resulted in several interesting observations. The frequency of the invasive pneumococcal isolates with 19F serotype in our and ANSORP studies was similar (13.3% vs 13.7%). The differences between ANSORP and our studies were; i) their frequency for serotype 19A was 8.2%, whereas no 19A serotype was found in our clinical isolates, ii) they found serotype 14 in 7.3% of the cases, we found in 46% of our clinical isolates (67% for the invasive isolates), iii) serotype 6A (4.2%) was the most prominent non-PCV7 serotype in the ANSORP report, whilst we observed no 6A serotype in our clinical isolates, and iv) 52% of ANSORP isolates showed PCV7 serotypes, we found significantly higher frequency (89%). These observations collectively suggest that several common S. pneumococcus serotypes have disseminated throughout the Far East, South East Asia, Middle East countries. On the other hand, some serotypes are highly localized in parts of Asian countries. Similar to the other reports from Middle East countries [3,25] the frequencies of our serotypes included in PCV7 and PCV10 were 89% for clinical, 50% for normal flora and 68% for total sisolates. Serotype 3 was the only non-PCV7 serotype found in our clinical study. The frequency of the serotypes included in PCV13 was 100% for both clinical and normal flora. For invasive isolates, the rates of serotypes covered by PCV7or PCV10 and PCV13 were 93% and 100%, respectively. At the present time pneumococcal conjugated vaccine is not routinely used in Iran and it is only given to a limited number of people who are at high medical risks. Our results suggested that PVC13 would be the preferred vaccine over PVC7 or PVC10 and can be implemented in the Iranian national vaccine program.
By using Pneumococcal Molecular Epidemiology Network classification,7 international resistant clones, including 19 different STs were identified among our clinical isolates. The most frequent group (23 isolates, 38%) was related to the Sweden 15A -25 ST63 international clone with 57%, 39% and 4% harboring erm(B) alone, mef alone and ermB/mef, respectively. The second most frequent group was Spain 23F -1 ST81 clone (6 isolates, 10%). Additionally, MLST analysis (at least six common MLST alleles out of 7) revealed nineteen STs with ST3130, ST180, ST81 and ST63 were the frequent STs covering 62% of the total cases. In a study by Calatayud and colleagues in Spain [26], 10 ST clones were reported during 1999-2007 period with Spain 23F -1 ST81, Sweden 15A -25 ST63 and Spain 9V -3 ST156 constituting 30% of their invasive isolates, we found the 2 international ST clones (Sweden 15A -25 ST63, and Spain 9V -3 ST156) constituting 73% of our invasive isolates. On the other hand, ST81, ST283 and ST236 have been reported as the most frequent STs in some Asian countries [27,28]. Overall, our results suggest that our pneumococcal clonal types were similar to the clones in Europe as compared to the reports in South Africa, South Korea and United States where high prevalence of Taiwan19F-14 clonal complex 271 (CC271) has been reported [20].

Conclusions
The present study showed that the rate of resistance of S. pneumoniae strains to erythromycin was significantly high and the related isolates were clonally disseminated with majority having MLS B and ermB phenotypic and genotypic characteristics, respectively. Additionally, we showed that within the ERSP isolates, serotypes 14, 3, 19F and 23F constituted majority (85%) of the total ERSP isolates which were belonged to ST3130 (CC63), ST180, ST3130 and ST81 clones, respectively. The data presented here suggest that the use of PVC13 could have a significant protection coverage and should be the recommended vaccine for the Iranian population living in Tehran.