Silence as a way of niche adaptation: mecC-MRSA with variations in the accessory gene regulator (agr) functionality express kaleidoscopic phenotypes

Abstract

Functionality of the accessory gene regulator (agr) quorum sensing system is an important factor promoting either acute or chronic infections by the notorious opportunistic human and veterinary pathogen Staphylococcus aureus. Spontaneous alterations of the agr system are known to frequently occur in human healthcare-associated S. aureus lineages. However, data on agr integrity and function are sparse regarding other major clonal lineages. Here we report on the agr system functionality and activity level in mecC-carrying methicillin resistant S. aureus (MRSA) of various animal origins (n = 33) obtained in Europe as well as in closely related human isolates (n = 12). Whole genome analysis assigned all isolates to four clonal complexes (CC) with distinct agr types (CC599 agr I, CC49 agr II, CC130 agr III and CC1943 agr IV). Agr functionality was assessed by a combination of phenotypic assays and proteome analysis. In each CC, isolates with varying agr activity levels were detected, including the presence of completely non-functional variants. Genomic comparison of the agr I-IV encoding regions associated these phenotypic differences with variations in the agrA and agrC genes. The genomic changes were detected independently in divergent lineages, suggesting that agr variation might foster viability and adaptation of emerging MRSA lineages to distinct ecological niches.

Figure 1

The core genome phylogeny based on the Maximum Common Genome. The core genome phylogeny based on the Maximum Common Genome comprising 2,094 orthologous genes present in all isolates show four distinct clusters, whereby the genetic diversity within the clusters is rather low. Furthermore, the isolates metadata show no significant association with the core genome clusters. The 2,003 accessory genes show a distribution pattern that is highly correlated with the core genome clusters (right side), suggesting a lineage-specific gene content. Genes for aureolysin (aur), leucotoxins D and E (lukD, lukE), gamma-haemolysin component A–C (hlgA, hlgB, hlgC) and proteases SpIA or SpIB are present in all isolates. All isolates belonging to ST-1943 as well as some CC130 and CC599 were positive for different variants of the Staphylococcal pathogenicity island (SaPI) harbouring a toxic shock toxin encoding gene (tst), which were variants of tst-bov. Moreover, 48.5% of the mecC-positive isolates harboured staphylococcal enterotoxins (SE). The protease SpIE can just be found in 23/33 isolates and is not associated with any sequence type. The epidermal cell differentiation inhibitor B (edinB) cannot be determined in the isolates of ST-1943, ST-2361, ST-49 and ST-599.

Figure 2

Sequence alignments of agr regions in mecC-MRSA. Sequence alignments for (a) CC599 (agr I), (b) CC49 (agr II), (c) CC130 (agr III) and (d) CC1943 (agr IV) isolates. First row shaded in yellow indicates the wild type (wt) exemplarily shown for all isolates sharing 100% coverage and 100% nucleotide and amino acid sequence identity. For each of the non-wt isolates on the display, changes within the upper gray line indicates a nucleotide sequence alteration while changes in the second line indicates amino acid sequence alteration. For all details and the reference sequences used for each CC see Supplemental Table 4. (a) First row, wt agr I in CC599 shared by IMT32509, IMT32513, IMT36943, IMT 36,947, IMT36948, IMT38116, IMT39825, RKI5962, RKI5963 and RKI5964. Second row, IMT31818 has a triplet nucleotid deletion resulting in ΔN177 in AgrA; third row, IMT39824 shows a non synonymous substitution (from A to G in position 289) leading to F196S in AgrA. (b) First row, wt agr II in CC49 represented by RKI5972 while the second row shows IMT36945 with an insertion creating a premature stop codon in AgrA. (c) First row, wt agr III in CC130 shared by IMT34480, IMT34488, IMT34491, IMT36946, IMT36950, IMT38115, IMT38119, IMT39816, IMT39819, IMT40506, IMT41554, RKI5965, RKI5967 and RKI5968. Rows 2, 4, 9 and 10 show amino acid variations in AgrC for IMT34479 (N5S), RKI5966 (Q16H), IMT32929 (E216K) and IMT32510 (G284D) generated by corresponding non-synonymous substitutions. RKI5967 (row 3), IMT38119 (row 5) and IMT38115 (row 6), harbour a nucleotide substitution (c to a) at position -2 upstream the RNAIII sequence start, respectively. Rows 7, 8 and 11 harbour variants of AgrA for IMT34489 (G68D), IMT34478 (S215P) caused by non-synonymous substitutions and IMT31819 shows an insertion at position 711 bp in agrA causing an alternate stop codon. (d) First row, WT agr IV in CC1943 shared by RKI5973, IMT34485, IMT40507, IMT40504, IMT38113. Second to fourth row (upper gray line) show a synonymous substitution (a to t at position 442) in agrA for RKI5971, RKI5970 and RKI5969 while the fifth row indicates a further variation of agrA (C199R) generated by a non-synonymous substitution.

Figure 3

Haemolytic activities of S. aureus isolates on sheep blood agar (SBA) plates. (a) Scheme for assessment of haemolytic activity based on Geisinger et al.. The isolates were tested by cross-streaking perpendicularly to S. aureus RN4220 on sheep blood agar (SBA) plates. The turbid zone induced by β-haemolysin production of RN4220 enhanced lysis by δ-haemolysin and PSMs (clear zone at the intersection) and inhibited α-haemolysin (V-shaped zone at the intersection). (b) Haemolytic activity of mecC-positive S. aureus belonging to different agr types on SBA plates. The lack of a corresponding phenotype in the isolate collection is indicated by a grey rectangle. One exemplarily image was used to illustrate the differences, respectively.

Figure 4

Illustration showing protein abundance values, relevant genomic variation and phenotype results of mecC-MRSA. (a) δ-haemolysin (Hld) protein abundance and the synergistic production of different haemolysins (SPDH test). (b) δ-haemolysin (Hld) protein abundance and the isolates’ capability for colony spreading.

Figure 5

Colony spreading assay results for mecC-S. aureus with different agr functionalities on soft agar plates. A TSA soft agar plate (0.24%) was inoculated with 2 µl overnight culture of S. aureus. Examples shown here include isolate RKI5966 associated with a weak agr functionality, which was not able to spread on semisolid agar plates, while isolate IMT38119 (strong agr functionality) showed spreading. As controls we have employed the standard laboratory strains RN4220 (weak agr functionality) and USA300 (strong agr functionality).

Figure 6

Agr activity and biofilm formation differences among mecC-MRSA belonging to clonal complex 130. (a) Macroscopic camera image of an S. aureus isolate (mecC-MRSA) harbouring the agr III wild type (wt) variant (IMT38119) grown in a 24-well plate and (b) the confocal laser scanning picture showing the biofilm profile at the indicated spot (red square). (c) Isolate harbouring a agr III variant (non-wt agrC variant) lacking agr activity (RKI5966) and (d) its corresponding biofilm. (e) Isolate lacking agr activity (IMT31819) and (f) its corresponding biofilm.