A dominant clonal lineage of Streptococcus uberis in cattle in Germany

Abstract

Bovine mastitis causes enormous economic losses in the dairy industry with Streptococcus uberis as one of the most common bacterial pathogens causing clinical and subclinical variations. In most cases mastitis can be cured by intramammary administration of antimicrobial agents. However, the severity of the clinical manifestations can vary greatly from mild to severe symtoms. In this study, a comparative genomic analysis of 24 S. uberis isolates from three dairy farms in Germany, affected by different courses of infection was conducted. While there were sporadic mild infections in farm A and B, a large number of infections were observed within a very short period of time in farm C. The comparison of virulence genes, antimicrobial resistance genes and prophage regions revealed no features that might be responsible for this severe course. However, almost all isolates from farm C showed the same, novel MLST profile (ST1373), thus a clonal outbreak cannot be excluded, whereby the actual reason for the particular virulence remains unknown. This study demonstrates the importance of extensive metagenomic studies, including the host genomes and the environment, to gain further evidence on the pathogenicity of S. uberis.

Keywords: Bovine mastitis; Comparative genomics; Multilocus sequence typing; Prophage regions; Resistance; Streptococcus uberis; Virulence.

Fig. 1

There is a high statistically significant difference (p < 0.01) between groups A and B and between groups B and C. Groups A and C differed significantly (p < 0.05). There is a high statistically signifant difference between all genomes currently found in the NBCI database and the total genome size of group A, B and C combined. Boxplot and calculation was created with R-Studio

Fig. 2

Core genome phylogenetic tree. Core genes of these genomes were computed in EDGAR 3.0 based on muscle alignment. An approximately-maximum-likelihood phylogenetic tree was calculated using the FastTree software. The core genome analysis was based on 1567 genes per genome in 24 strains plus the reference strains 0140J and EF20. The core had 489,494 amino acid residues/bp per genome, 12,726,844 in total. Bar, 0.001 nucleotide substitutions per site. The values at the branches are Shimodaira-Hasegawa support values. Isolates are highlighted according to group affiliation (Group A: red; Group B: orange; Group C: blue)

Fig. 3

Pan versus core development plot to gain insight into the stability of the pan- and core-genome of the isolates. Results were obtained using EDGAR 3.0. Starting with the first contig, consecutive numbers for the core and pan genome size were calculated and plotted. The plot shows an increase of the pan-genome (black line) and a decrease of the core-genome (grey line) as more genomes are added

Fig. 4

Overview of the average nucleotide match between genomes. The isolates were grouped according to their match. Within each box, the identity was given in percent. Darker heat map colours indicate higher relatedness. Results were obtained using EDGAR 3.0 based on a BLASTN comparison of the genome sequences

Fig. 5

All virulence genes that were not found in all isolates with a gene coverage of 100 % are listed. The gene-coverage in percent is indicated in the boxes. The heatmap was created with R-Studio from the results generated with ABRicate

Fig. 6

All prophage sequences found are listed. The score (< 70: incomplete; 70–90: questionable; > 90: intact) for each sequence is indicated in the boxes. The heatmap was created with R-Studio from the results generated with PHASTER