Serotype distribution and antibiogram of Streptococcus parauberis isolated from fish in South Korea

Abstract

Streptococcus parauberis is the dominant etiological agent of streptococcosis, the most devastating bacterial disease in the olive flounder farming industry in South Korea. In this study, the distribution of serotypes, antimicrobial susceptibility, and presence of antimicrobial resistance genes (ARGs) in S. parauberis isolates obtained between 1999 and 2021 was thoroughly investigated to gain insight into the dynamics of their presence and the relationship between serotypes and antimicrobial resistance. Disk diffusion testing of 103 isolates against 10 antimicrobial agents was performed, and epidemiological cut-off values generated through normalized resistance interpretation analysis were used to classify wild-type (WT) and non-wild-type (NWT) populations. Principal component analysis and hierarchical clustering were implemented to achieve an understanding on the relationship between serotypes and antimicrobial resistance patterns. PCR-based serotyping showed that serotype Ia (67.1%) was the most prevalent in South Korea, followed by serotypes Ib/Ic (25.2%) and II (7.7%). The highest proportion of isolates was assigned to NWT against amoxicillin (80.6%), followed by oxytetracycline (77.7%) and erythromycin (48.5%). The time-scale data showed that recently obtained serotypes Ib/Ic and II isolates tended to be categorized as NWT populations resistant to more antibiotics, possibly due to microbial adaptation to antibiotic pressure. ARGs responsible for resistance to oxytetracycline and erythromycin were found only in NWT populations in serotype Ia [tet(S) and erm(B), respectively], and serotype II [tet(M) and mef(J)-msr(I), respectively]. We also found that the mef-msr gene pair in S. parauberis serotype II might be involved in low-level resistance to erythromycin. IMPORTANCE This study presents serotype distribution and antimicrobial susceptibility data along with the antimicrobial resistance genes (ARGs) of Streptococcus parauberis, which is an important bacterial fish pathogen worldwide. In particular, almost all oxytetracycline and erythromycin non-wild-type (NWT) populations harbored tet(S) or tet(M), and erm(B) or mef(J)-msr(I), respectively. Interestingly, these ARGs were distributed in a highly serotype-dependent manner, resulting in a clear correlation between the antibiogram and serotype distribution. Moreover, recent isolates belonging to serotypes Ib/Ic and II tended to be more frequently categorized as NWT against antimicrobials, including amoxicillin and cefalexin compared to old isolates, while a dramatic decrease in erythromycin and clindamycin NWT frequencies was observed in recent serotype Ia isolates, which lacked erm(B). These variations might be attributed to shifts in the antibiotics employed in South Korean aquaculture over time. The overall findings would provide important background knowledge for understanding the epidemiology of S. parauberis infection in aquaculture.

Keywords: Streptococcus parauberis; antimicrobial resistance gene; antimicrobial susceptibility; serotype.

Fig 1

Principal component analysis (PCA; A) and hierarchical clustering analysis (B) based on the antibiogram of 103 Streptococcus parauberis isolates against 10 antibiotics. (A) The PCA plot captures the variance in the data set composed of normalized inhibitory zone diameters obtained from the disk diffusion tests of 103 S. parauberis strains. The X and Y-axes show principal components 1 and 2, which explain 67.8% and 16.9% of the total variance, respectively. The colors of the circles represent the serotype. Open and closed circles indicate the isolated year, categorized as 1999–2007 and 2013–2021, respectively. (B) Heatmaps show the wild-type (beige) and non-wild-type (red) distributions of S. parauberis isolates according to serotype. The black and pink ellipses located on the right side of the heatmap indicate the year of isolation, categorized as 1999–2007 and 2013–2021, respectively. Abbreviations for the antibiotic agents are shown in Table 2.

Fig 2

Year-wide distribution of non-wild-type (NWT) populations (%) of three serotypes of Streptococcus parauberis isolates against 10 antibiotics. The color of the dots indicates the NWT proportion, while the size indicates the number of NWT populations in the respective isolation year. Abbreviations for the antibiotic agents are shown in Table 2.

Fig 3

Distribution of antibiotic resistance genes (ARGs) in 31 Streptococcus parauberis genome sequences. (A) The heatmap presents the presence and absence of ARGs in each isolate, as shown in the red and blue boxes, respectively. (B) Gene organization of four ARG-carrying mobile element structures in S. parauberis serotypes Ia and II. The direction of the arrow indicates the direction of transcription. ARGs and mobile element-related genes are shown in red and yellow arrows.

Fig 4

Distribution of inhibitory zone diameters of 103 Streptococcus parauberis isolates to oxytetracycline (A) and erythromycin (B) shown with the proportion of isolates harboring antimicrobial-resistance genes. Inhibitory zone diameters of 103 S. parauberis isolates are shown in filled circles [tet(S)⁺ tet(M)⁻ and erm(B)⁺ mef(J)-msr(I)⁻ in panels A and B, respectively], filled triangles [tet(S) ⁻tet(M)⁺ and erm(B)⁻ mef(J)-msr(I)⁺ in panels A and B, respectively], and empty squares [tet(S)⁻ tet(M)⁻ and erm(B)⁻ mef(J)-msr(I)⁻ in panels A and B, respectively]. The horizontal red lines indicate the epidemiological cut-off values (ECVs). The proportion of respective genotypes in every serotype is shown below the scatter plots.