Increasing Resistance and Changes in Distribution of Serotypes of Streptococcus agalactiae in Poland

Streptococcus agalactiae is responsible for serious infections in newborn babies, pregnant women, and other patients. The aim of this study was to evaluate antimicrobial susceptibility, serotype distribution, and virulence determinants of the S. agalactiae isolates derived from clinical specimens considering the global increase of both antibiotic resistance and virulence. A total of 165 isolates were identified and serotyped by PCR techniques. Antimicrobial susceptibility was assessed by disk diffusion method, gradient diffusion method and VITEK® System. Virulence associated genes were investigated by PCR; ability to form biofilm was assessed using a microtiter plate assay. The highest observed MIC value for penicillin G was 0.12 µg/mL, seen in 8.5% of isolates. Resistance to erythromycin and clindamycin were found in 30.38% and 24.8% of the strains, respectively. The serotype III (32.73%), V (25.45%), and Ia (18.18%) were found as the most frequently represented. Previously unidentified strains in Poland, belonging to serotypes VI (three strains) and VII (one strain) were recognized. The presence of genes encoding various virulence factors as well as diverse ability to form biofilm were found. In conclusion, macrolide-resistance and decreased susceptibility to penicillin G were revealed signifying the increasing resistance among group B streptococci. Moreover, the presence of genes encoding various virulence factors and the ability to form biofilm were confirmed indicating their role in the pathomechanisms of the evaluated GBS infections.


Introduction
Despite the screening and introduction of prenatal prophylaxis, S. agalactiae (Group B streptococci, GBS) still remains an important etiological factor of infections in newborns. Infections caused by S. agalactiae become also a growing problem among patients who are older than 60 years, especially among those who suffer from coexisting other diseases and whose immunity system is weakened [1][2][3].
Ten serological types, marked as Ia, Ib, and II-IX, can be differentiated among GBS strains on the basis of structure differences of capsular polysaccharides and protein antigen C. Various serotypes, identified with different frequency, depending on a geographical region, display a substantial antigen diversity. Serotypes Ia, II, III, and V were responsible for the majority of infections in newborns and pregnant women [4-6].

Detection of GBS Isolates with Decreased Sensitivity to Penicillin and Macrolides-Lincosamides-Streptogramins B Resistance
According to EUCAST criteria, all of the tested 165 GBS strains were sensitive to penicillin ( Table 1). The MICs ranged from 0.032 µg/mL to 0.125 µg/mL. The EUCAST susceptible breakpoint for GBS is 0.25 µg/mL. The results obtained with disc diffusion method revealed a high-level resistance to tetracycline in 154 GBS strains. All strains were sensitive to levofloxacin and tigecycline. Antimicrobial susceptibility results for all tested strains are shown in Supplementary Table S1.
Among the tested strains, 50 and 41 isolates were resistant to erythromycin and clindamycin, respectively. Based on the mechanisms of resistance, various resistance phenotypes to macrolides, lincosamides, and streptogramin B were identified. The M phenotype was found in nine strains, the constitutive MLS B mechanism was identified in 30 strains, whereas 11 of S. agalactiae strains showed the inducible MLS B mechanism ( Figure 1). * MIC value 0.125 obtained using the E-test was confirmed by the VITEK ® 2 (0.12 value was obtained for these isolates). The sensitivity limit of the MIC according to EUCAST is 0.25 µg/mL. Detailed results for each material tested are presented in Supplementary Table S1.
Pathogens 2020, 9, x FOR PEER REVIEW 3 of 14 * MIC value 0.125 obtained using the E-test was confirmed by the VITEK ® 2 (0.12 value was obtained for these isolates). The sensitivity limit of the MIC according to EUCAST is 0.25 µg/mL. Detailed results for each material tested are presented in Supplementary Table S1.

The Rib Gene was the Most often Identified in the Tested Strains
Among the tested strains, isolates with the rib gene (n = 72) predominated, followed by the alp2/alp3 (n = 48), epsilon (n = 37) and bca (n = 8) genes. The presence of the alp4 gene was not confirmed ( Table 2).

Majority of S. agalactiae Strains Identified as Moderate-Biofilms Producers
All of the S. agalactiae strains showed the ability to form biofilms. More than half of the tested isolates (n = 86) demonstrated moderate strength of biofilm formation; 72 and seven were found as weak and strong biofilm-producers, respectively (Table 2), (Supplementary Table S3).

Correlations of the Recognized Features
Statistically significant differences were observed by analyzing occurrence of genes encoding different types of fimbriae among strains representing individual serotypes. Genes encoding Pl-1 fimbriae together with Pl-2a were the most frequently found in strains of serotypes II, and V, while genes encoding fimbriae Pl-2a (occurring singly) were most frequently observed in strains of serotype Ia (p = 1.34e-10). The genes encoding fimbriae Pl-1 together with Pl-2b were the most often recognized in strains representing serotypes III (p < 2.2e-16). Moreover, occurrence of genes encoding tested proteins from Alp family among strains representing individual serotypes was statistically significant (epsilon (p < 2.2e-16), rib (p = 8.802e-07), alp 2/3 (p = 1.851e-09), and bca (p = 1.573e-07)). When analyzing the participation of individual serotypes among erythromycin resistant strains, statistically significant differences were also observed (p =3.483e-05). Almost half of these strains belonged to serotype V.
Assessing the occurrence of strong, moderate and weak biofilm producers among the strains belonging to the most common serotypes, no statistically significant differences were found. Moreover, comparing the presence of genes encoding particular types of fimbriae and genes encoding surface proteins among strains differing in their ability to form biofilms, no statistically significant differences were obtained. Correlation analyses' results are presented in Figure 2.
Assessing the occurrence of strong, moderate and weak biofilm producers among the strains belonging to the most common serotypes, no statistically significant differences were found. Moreover, comparing the presence of genes encoding particular types of fimbriae and genes encoding surface proteins among strains differing in their ability to form biofilms, no statistically significant differences were obtained. Correlation analyses' results are presented in Figure 2. The distribution of the strong (indicated in naval blue), moderate (red), and weak (yellow) biofilm producers in the most frequently observed serotypes (shown in the circles, placed with the Ia close to the center, through Ib, II, III, and V the most distally, respectively) was evaluated. No statistically significant difference in the incidence of the serotypes was found (p > 0.05). (B) The occurrence of genes encoding various types of fimbriae (marked with naval blue, red, yellow and green) among the most frequently isolated serotypes, Ia, Ib, II, III and V, VI (presented in circles) was analyzed. (C) The occurrence of genes coding the tested proteins from the Alp family (indicated by colors) was analyzed in the strains The distribution of the strong (indicated in naval blue), moderate (red), and weak (yellow) biofilm producers in the most frequently observed serotypes (shown in the circles, placed with the Ia close to the center, through Ib, II, III, and V the most distally, respectively) was evaluated. No statistically significant difference in the incidence of the serotypes was found (p > 0.05). (B) The occurrence of genes encoding various types of fimbriae (marked with naval blue, red, yellow and green) among the most frequently isolated serotypes, Ia, Ib, II, III and V, VI (presented in circles) was analyzed. (C) The occurrence of genes coding the tested proteins from the Alp family (indicated by colors) was analyzed in the strains representing individual serotypes (shown in the circles). Statistically significant differences were found in the occurrence of epsilon (p < 2.2e-16)* (naval blue), rib (p = 8.802e-07)* (red), alp 2/3 (p = 1.851e-09)*(yellow) and bca (p = 1.573e-07)* (green) genes among individual serotypes. (D) Analyzing the participation of individual serotypes (color-coded) among erythromycin-resistant strains (presented in a circle), it was observed that almost half of these strains (n = 28) belonged to serotype V (brown), followed by serotypes Ia (naval blue), III (green), and II (yellow). The observed differences in prevalence were statistically significant (p = 3.483e-05)*. (E) The presence of genes encoding different types of fimbriae (indicated by colors) was compared among strains differing in the ability to form biofilms (shown in the circles). The differences found were not statistically significant. (F) The presence of genes encoding surface proteins (color coded) in the strains differing in the ability to form biofilm structure (shown in the circles) was compared. In all analyzed groups dominated isolates with the rib gene (red). The observed differences were not statistically significant. * Statistically significant.

Discussion
Here the assessment of the antibiotics' susceptibility, distribution of serotypes, ability to produce biofilm, and the presence of genes encoding fimbriae and surface proteins from the Alp family and fibrinogen binding proteins as well as correlations between the identified features were investigated in S. agalactiae strains.
The susceptibility testing of GBS strains showed that all tested strains were sensitive to penicillin as recorded in accordance with the EUCAST guidelines; however, reduced susceptibility to penicillin with the MIC values of 0.12 µg/mL was detected in more than eight percent of the tested strains (n = 14). That indicates the necessity of monitoring the susceptibility of GBS to penicillin, as it was previously indicated [7][8][9][10][11][12][13].
Beta-lactams are the first-line antibiotics for the prevention and treatment of GBS infections, but in the case of reduced susceptibility to penicillin and beta-lactam, allergy macrolides are recommended as the second-line agents. During the last decades resistance to macrolides and lincosamides emerged. The percentages of strains resistant to erythromycin and clindamycin varied between European countries and were reported as 43.7% and 32.2% in Italy [17], 30.0% and 28.0% in Switzerland [29], 23.1% and 21.3% in Serbia [30], and 14.5% and 14.0% in Sweden [31]. Moreover, outside Europe, a worrying trend towards increasing resistance to these antibiotics group have been found, especially in China (78.6%-92.5% of strains resistant to erythromycin and 64.3%-87.5% resistant to clindamycin) [5,18,19]. In contrast, a relatively low percentage of the resistant strains was observed in South American countries such as Brazil (4.1% strains resistant to erythromycin and 3.0% resistant to clindamycin) [32] and Chile (9.5% and 13.7%, respectively) [33].
In previous studies in Poland, 16-24% of erythromycin resistant and 10-20% of clindamycin resistant strains were reported [34,35]. Our findings revealed that the resistance to erythromycin was 30.3%, whereas the resistance to clindamycin was 24.8%. Indicating the increase in the percentage of strains resistant to the evaluated groups of antibiotics in Poland. The results of our study also confirm that testing of susceptibility to these antibiotics should be performed on regular basis.
In this study, three mechanisms of macrolide resistance were identified in the tested strains. Most of them were diagnosed with constitutive MLS B and ten times less with inducible MLS B . The constitutive MLS B mechanism was also the most commonly found mechanism of macrolide resistance by other authors in Poland [34,35], Serbia [30], Germany [36], and France [37]. Other groups of antibiotics were also assessed in this study. All tested strains were sensitive to levofloxacin and tigecycline, but a high percentage of strains resistant to tetracycline (93.3%) was observed. The obtained results are consistent with the available literature data [17,20,34,38]. The occurrence of individual serotypes (Ia, Ib, and II-IX) is different, depending on the geographical region of the world and ethnicity assessed; it also changes for a given region over the time [39][40][41][42]. Among 353 strains isolated from pregnant women from the Małopolska province in Poland, the serotypes III (35%) dominated among the tested strains, followed by Ia (20%), V (17%), II (15%), Ib (8%), and IV (5%) [43]. The authors did not find strains belonging to the VI, VII, and VIII serotypes in the study group. Moreover, Wolny and Gołda-Matuszak carried out an assessment of the occurrence of serotypes Ia, Ib, II, III, IV, and V among 380 strains isolated from genital tracts from women of reproductive age from Krakow. The serotypes III (50%), Ia (18%), and V (14%) were predominant; the authors also noted the presence of serotypes: IV (4%), Ib (3%), and II (3%) [44]. The above presented data is consistent with studies in the Czech Republic and Italy, where serotypes III (33.2%), Ia (22%), and V (13.9%) [45] and III (32.9%), V (26.1%), and Ia (20.5%) [46] dominated, respectively.
Here we found the distribution of GBS serotypes consistent with other results from our country and from other European populations. Among evaluated S. agalactiae strains serotypes III, V, and Ia predominated. However, it should be noted that for the first time, strains belonging to the serotypes VI (three strains) and VII (one strain) were recognized in Poland. Understanding the prevalence of individual GBS serotypes for the population is important for epidemiological reasons, as well as for the design of a multivalent vaccine against GBS. In S. agalactiae three types of fimbriae (Pl-1, Pl-2a, and Pl-2b) have been identified, and each strain produces simultaneously one or two types of fimbriae [25][26][27][28]. Martins et al. evaluated the presence of various types of fimbriae among 898 GBS strains derived from colonized pregnant women, from non-invasive infections of adults and from invasive infections of neonates and adults from Spain [27]. Strains having fimbriae Pl-1 occurring together with Pl-2a predominated in the study group (49%), followed by Pl-2a (30%), Pl-1 together with Pl-2b (21%), and Pl-2b (0.6%). The strains belonging to the serotype Ia most often had fimbriae Pl-2a (92.5%), whereas fimbriae type Pl-1 together with Pl-2a were most often identified among strains belonging to serotypes Ib, II, IV, and V (94.4%, 92.8%, 64.7%, and 82.6%, respectively) [27]. Minor differences in relation to the above-mentioned studies were obtained by Lu et al. [28]. They assessed the incidence of different types of fimbriae in 160 GBS strains isolated from colonization from pregnant women in China. In the tested strains, they most often identified fimbriae type Pl-2a (43.8%) and fimbriae Pl-1 together with Pl-2a (41.9%). In strains of serotype Ia dominated fimbriae Pl-2a, from strains belonging to serotype Ib, III, and V, the authors identified the most fimbriae Pl-1 in combination with PI-2a [28].
Our study showed that the detection of various types of fimbriae was similar to that reported by other researchers, and the isolates with genes encoding both Pl-1 and Pl-2a fimbriae (55% of all tested strains) predominated. Based on our results, it can be concluded that in the strains belonging to serotype Ia, the Pl-2a fimbriae type are frequently present. In the strains belonging to the serotypes Ib, II, V, and VI, Pl-1 simultaneously with Pl-2a fimbriae were the most commonly found. The observed differences were found to be statistically significant and consistent with the previously reported results.
Moreover, in this study, among the tested strains, those in which rib, alp2, or alp3 and epsilon genes were found dominated. Analyzing the presence of genes encoding proteins from the Alp family among the studied S. agalactiae strains, it was observed that there was a statistically significant relationship between the presence of specific genes and individual GBS serotypes. Similar observations have been made by other authors [22,47,49]. Creti et al. observed the occurrence of the rib gene most often in strains with serotype III; the bca gene among the isolates belonging to serotypes Ia, Ib, and II; and the gene alp3 in serotypes V and VIII [22]. Moreover, it has been noticed that the most common rib gene among serotype III strains is epsilon in serotype Ia, bca in strains representing serotypes Ib and II, and alp3 in strains belonging to serotype V [47,49].
Biofilm formation by GBS strains seems to support colonization of female genital tracts, providing protection against difficult environmental factors and host defense mechanisms. In vitro, it depends on various environmental factors, including the type of a medium used, the conditions of pH, temperature, and osmolarity [50][51][52][53]. Methodological differences and various interpretation criteria applied in different studies make the data comparisons difficult [26,51,[54][55][56]. Ho and co-authors evaluated the ability to form biofilm in 80 GBS strains derived from colonization from pregnant women, classifying them into groups of very strong (51%), strong (28%), and moderate biofilm producers (18%) and strains which do not produce biofilm (4%). However, these results are not comparable with the results obtained by other authors due to differences in the adopted criteria of analysis of the results [51]. In their studies Parker et al., qualifying the tested GBS strains into two groups, strong and weak biofilm producers, noted that 43.0% of the 242 strains isolated from humans produced biofilm in a strong and 57.0% in a weak degree. It was also observed that a greater percentage (45.9%) of strains from the colonization of pregnant women in comparison to those from invasive infections of newborns (32.5%) was strong biofilm producers [56]. Similar observations suggesting a greater ability to produce biofilm among strains from colonization (76.5%) compared to strains derived from infections (34.6%) are presented in the work of Kaur et al. [55].
Moreover, the results obtained in this study confirmed the ability of GBS strains for biofilm formation. It was found that more than half of the tested strains produced moderate biofilms and about 4% strong ones. Analyzing the ability to form biofilm by strains representing different serotypes, it was observed that more than half of the strains representing serotypes Ib, II, III, V, and VI formed moderate biofilms; they accounted for 75.0%, 52.6%, 51.8%, 57.1%, and 66.7% respectively. Of the seven isolates showing a strong ability to form biofilm, three belonged to serotype III, two serotype V, and one for serotypes Ia and one for II. However, the results were not statistically significant. Jiang et al., studying 87 GBS strains, observed that only 13.8% of them were able to form biofilms [57]. They did not observe the relationship between the ability of strains to create biofilms, serotype, and drug resistance [57].
Concerning the correlation analyses performed in this study, it was observed that genes encoding Pl-2a fimbriae occurring together with genes encoding fimbriae Pl-1 (85.7%) or individually (14.3%) were present in all strains producing biofilm to a strong degree. In addition, 57.0% of moderate biofilm strains possessed Pl-1 fimbriae together with Pl-2a, and 23.3% of Pl-2a fimbriae were found alone. However, the observed differences were not statistically significant. It should be highlighted that the ability to form biofilm depends on the expression of these genes. Moreover, Rga, a RofA-Like Regulator not evaluated in this study, is important in fimbriae gene expression process [58,59].
Summarizing, in this study, increase in macrolide-resistance compared to the previously published reports was found. Moreover, reduced susceptibility to penicillin G was observed. These findings indicate the increasing resistance among group B streptococci. The previously unidentified serotypes in Poland were recognized, and the occurrence of genes encoding various virulence factors was confirmed in the evaluated strains. Moreover, the study results indicate that S. agalactiae strains have the ability to form biofilm in vitro, differing in the degree of its formation.

Strains Collection
A total of 165 S. agalactiae clinical strains derived from Polish pregnant women, newborns, and adults affected by non-perinatal infections were collected, as presented in details in Table 3.

GBS Identification
Initial identification of the strains was conducted on the basis of macroscopic assessment of bacterial colonies on Columbia Agar with 5% sheep blood. Subsequently, the GBS were identified using a latex agglutination test (OXOID, UK) in accordance with the manufacturer's instructions and the PCR techniques described previously by Kong et al. [60].

Antimicrobial Susceptibility Testing
Assessment of antibiotics susceptibility of S. agalactiae was carried out using disc diffusion method (penicillin (1 IU), erythromycin (15 µg), clindamycin (2 µg), tigecycline (15 µg), tetracycline (30 µg), levofloxacin (5 µg), and antimicrobial gradient diffusion method (penicillin G) according to the guidelines of the EUCAST and the National Reference Center for Antimicrobial Susceptibility Testing (KORDL) [61,62]. The MICs of penicillin G were confirmed using the VITEK ® 2 System with AST-ST03 cards (bioMérieux, Durham, NC, USA) in accordance with the manufacturer's instructions. Interpretation of the results of susceptibility testing was carried out in accordance with the EUCAST and KORDL recommendations [61,62].
Resistance to macrolides, lincosamides, and streptogramins B was assessed using the double disc diffusion method in accordance with the EUCAST and the KORDL guidelines [61,62].

S. agalactiae Strains DNA Isolation
Bacterial DNA was extracted using the thermal method, subjecting bacteria cell wall to the process of lysis by warming up (60 ± 2 • C) and cooling down (−20 ± 2 • C) during six cycles lasting 30 s each. The obtained DNA concentrations were measured using the NanoDrop spectrometer (Thermo Scientific, Wilmington, DE, USA) and stored in −20 ± 2 • C for further analyses.

Serotyping and Detection of Virulence Genes of S. agalactiae Strains
Supplementary Table S4 shows PCR techniques, primers, and reaction conditions used for serotyping of S. agalactiae strains and detection of the virulence genes encoding surface proteins from the Alp family (epsilon, rib, alp2/3, bca, alp4) and different types of fimbriae. PCR products were purified using an ExoSAP-IT ® for PCR product Clean-Up (Affymetrix, Santa Clara, CA, USA), and the specificity was confirmed by Sanger sequencing (performed at Genomed, Warsaw, Poland).

Biofilm Assay
The experiments were performed using polystyrene microtiter plate assay, based on the techniques described by Borges et al. [63]. Briefly, three wells of 96-well flat-bottomed plastic plates were filled with 200 µL of each tested bacterial suspension in Todd-Hewitt broth. The plates were covered and incubated aerobically for 24 h at 37 • C. Then, the content of each well was aspirated, and each well was washed three times with 250 µL of sterile physiological saline. The plates were shaken in order to remove all non-adherent bacteria. The biofilms were stained for 15 min with 200 µL of 0.1% crystal violet. Then, each well was washed three times with 250 µL of sterile physiological saline once again and air-dried. After that, 200 µL of 99% methanol per well was added. The positive and negative controls were Staphylococcus epidermidis ATCC 35,984 with proven biofilm capacity, and the Todd-Hewitt broth without bacterial suspension, respectively. The optical density (OD) of the solubilized dye was measured at 590 nm using plate reader-Infinite M200 (Tecan, Grödig, Austria). Interpretation of biofilm production was performed according to the criteria described by [64]. The mean optical density (OD) of the negative control +3 standard deviations of negative control was considered the cut-off (ODc = 0.13). The tested strains were classified on the basis of the following criteria:

Correlations and Data Analyses
Statistical analysis was performed using R statistical package R. The χ2 homogeneity test and Fisher's exact two-sided test were applied to check the relationship between categorized variables. In the situation when the assumptions of the χ2 test were not met, Fisher's exact test was used to check whether the distribution of cases between the categories of one variable is independent of their distribution between the categories of the other variable. The differences were considered statistically significant at p value < 0.05.
In correlation analyses of the obtained results, the following aspects were assessed: the participation of strong, moderate, and weak biofilm producers; the presence of genes encoding different types of fimbriae; and occurrence of genes coding for tested proteins from the Alp family among strains belonging to the most common serotypes. Moreover, the incidence of individual serotypes among erythromycin resistant strains was analyzed. Furthermore, the presence of genes encoding individual types of fimbriae and the presence of genes encoding surface proteins were compared among strains differing in their ability to form a biofilm.

Conclusions
The observed changes in susceptibility are alarming and indicate the need for tracking the susceptibility of the S. agalactiae strains. Although the GBS serotype distribution is consistent with results relating to other European populations, the serotypes previously not reported in Poland have been revealed. Identification of serotypes occurring in the population is important for epidemiological reasons, as well as for the design of a multivalent GBS vaccine. The presence of genes encoding various virulence factors involved in colonization and the ability to form biofilm were confirmed indicating their role in the pathomechanisms of the evaluated GBS infections.
Supplementary Materials: The following are available online at http://www.mdpi.com/2076-0817/9/7/526/s1. Table S1: Antimicrobial susceptibility testing results obtained by disc diffusion method and antimicrobial gradient diffusion method (E-test). Table S2: Results of serotyping and detection of virulence factors performed in the assessed strains. Table S3: Percentage of Streptococcus agalactiae strains with different biofilm forming capacity. Table S4: Primer sets used for identification, serotyping, and detection of virulence genes in S. agalactiae.