Abstract
An international collection of Staphylococcus aureus of clonal complex (CC) 398 from diverse hosts spanning all continents and a 30 year-period is studied based on whole-genome sequencing (WGS) data. The collection consists of publicly available genomic data from 2994 strains and 134 recently sequenced Swiss methicillin-resistant S. aureus (MRSA) CC398 strains. A time-calibrated phylogeny reveals the presence of distinct phylogroups present in Asia, North and South America and Europe. European MRSA diverged from methicillin-susceptible S. aureus (MSSA) at the beginning of the 1950s. Two major European phylogroups (EP4 and EP5), which diverged approximately 1974, are the main drivers of MRSA CC398 spread in Europe. Within EP5, an emergent MRSA lineage spreading among the European horse population (EP5-Leq) diverged approximately 1996 from the pig lineage (EP5-Lpg), and also contains human-related strains. EP5-Leq is characterized by staphylococcal cassette chromosome mec (SCCmec) IVa and spa type t011 (CC398-IVa-t011), and EP5-Lpg by CC398-SCCmecVc-t011. The lineage-specific antibiotic resistance and virulence gene patterns are mostly mediated by the acquisition of mobile genetic elements like SCCmec, S. aureus Genomic Islands (SaGIs), prophages and transposons. Different combinations of virulence factors are present on S. aureus pathogenicity islands (SaPIs), and novel antimicrobial resistance gene containing elements are associated with certain lineages expanding in Europe. This WGS-based analysis reveals the actual evolutionary trajectory and epidemiological trend of the international MRSA CC398 population considering host, temporal, geographical and molecular factors. It provides a baseline for global WGS-based One-Health studies of adaptive evolution of MRSA CC398 as well as for local outbreak investigations.
Similar content being viewed by others
Introduction
Staphylococcus (S.) aureus is a Gram-positive coccoid bacterium commonly colonizing the skin and mucous membranes of humans and several animal species1,2. It is also recognized as one of the most important opportunistic bacterial pathogens in humans and causes a wide variety of infections, such as food-borne intoxication, skin and soft tissue infections, and life-threatening bacteremia3,4. Animals and humans are frequently asymptomatic carriers of S. aureus but can also develop a wide range of different infections and may act as a turntable for human transmission1,5. The pathogenicity of S. aureus is due to the production of a variety of virulence factors, such as hemolysins, leukocidins, exfoliative toxins, proteases, super-antigenic enterotoxins and toxic-shock syndrome toxin-1 (TSST-1), some of which are commonly carried by mobile genetic elements (MGEs), such as prophages or genomic islands (SaGIs), acquired via horizontal gene transfer. The immune evasion cluster (IEC) is a set of human-specific immune-modulating proteins, such as staphylokinase (SAK), enterotoxin (SEA), chemotaxis-inhibitory protein (CHP), and staphylococcal complement inhibitor (SCN), which are commonly encoded by hlb-converting lysogenic phages of the Siphoviridae family and mediate adaptation to the human host3,6. SaGIs can be integrated at different locations within the genome of S. aureus and might carry accessory genes, especially those encoding antibiotic resistance (S. aureus Resistance Islands [SaRIs]) or virulence proteins (S. aureus Pathogenicity Islands [SaPIs]). SaPI-encoded variants of the von Willebrand factor-binding protein (vWBP) play an important role in host adaptation1,7,8. Furthermore, S. aureus can acquire antibiotic-resistance genes, posing therapeutic challenges in healthcare institutions and veterinary settings1,4. Of major concern is methicillin-resistant S. aureus (MRSA), which is resistant to all beta-lactam antibiotics except for 5th-generation cephalosporins and is frequently resistant to other clinically important antimicrobial agents9,10,11. MRSA is characterized by the acquisition of the methicillin resistance genes mecA or mecC, which are encoded by Staphylococcal Cassette Chromosome mec (SSCmec) elements12,13,14. In animals, MRSA was initially acquired from humans; however, in the 2000s, a new lineage of the MRSA clonal complex (CC) (398) emerged through adaptation in livestock and is now known as livestock-associated MRSA (LA-MRSA)1,15. LA-MRSA animal adaptation was characterized by the occurrence of two recombination events, namely, the acquisition of the tetracycline resistance gene tet(M) and the loss of the prophage-mediated IEC16. However, the reacquisition of IEC determinants mediated by prophages or SaPIs is frequent, allowing for human readaptation17. In recent decades, the prevalence of the MRSA sequence type (ST)398 in livestock has increased steadily worldwide18,19,20,21,22. Pigs and, to a lesser extent, cattle are usually asymptomatic carriers of LA-MRSA ST398 and are considered the main reservoirs15. Horses, on the other hand, are particularly predisposed to develop postsurgical MRSA ST398 infections23. Animal owners and people working with animals are at risk of being colonized, creating a new biological environment of human-adapted clones that are currently also associated with clinical infections5,24,25,26,27,28,29,30.
In this study, the reconstruction of a time-calibrated phylogeny in combination with geographical and epidemiological data from more than 3000 strains was used to determine when and where successful lineages may have emerged. Additionally, an in silico screening for the presence of virulence and antimicrobial resistance genes was used to determine whether specific patterns were specifically associated with certain lineages and hosts. Representative strains of the major European lineages from the Swiss collection were hybrid assembled, and an in-depth analysis of the chromosome-associated mobilome, i.e., prophages and genomic islands, was performed to identify mobile genetic elements causing the gene carriage patterns observed. This approach, based on a deeper analysis of the epidemiological spread and recombination events involved in the acquisition of resistance and virulence genes, provides new insights into the temporal emergence and evolution of different lineages of MRSA CC398 at a global scale and will serve as an updated baseline for future analysis.
Results
CC398 population structure: time-calibrated phylogeny, molecular characterization, host, and geographic distribution
Based on our analysis, the molecular clock of the CC398 lineage was estimated to be 3.34 × 10−7 substitutions per site per year [95% higher posterior density (HPD) interval: 3.20–3.48 × 10−7] which is consistent with but lower than that estimated in previous studies for S. aureus31,32,33 and specifically CC39834. This phylogenetic analysis revealed the presence of six clades, seven major phylogroups, and several lineages and sublineages (Fig. 1).
The 3128 isolates were compared pairwise to the MRSA ST398 type strain reference genome (GenBank accession number NC_017333.1) and a cgSNP alignment was constructed after calling SNPs and removing recombination events, this alignment was used in combination with the isolation date to perform a MCMC simulation and reconstruct a Bayesian inferred phylogeny. Colored branches correspond to phylogroups; namely gray SP, dark pink AP, light pink AAP, orange EP1, violet EP2, turquoise EP3, red EP4, and blue EP5. carr. carriage, clin. clinical, c1-6: corresponds to major clades 1 to 6. Lpg: pig-associated lineage. Leq: equine-associated lineage. L1-5: major European lineages based on previous studies. LEa-g: minor lineages. External rings (from outer to inner one) represent metadata from the samples, namely MRSA, continent, country of isolation, host species, and spa type, respectively.
Clade one (c1) (n = 64) was formed mostly by Asian MSSA (only five MRSA) and was the most distant from the current MRSA clades. The divergence time from the most recent common ancestor of the remaining clades was estimated to be approximately 1949 [95% C.I.: 1942–1956]. The strains were isolated from different host species and sample materials, the majority carried AMR genes against several antibiotic classes and IECs were carried by less than 25% of the strains in this clade (most common chp-scn) (Fig. 2 and Supplementary Data 1).
In silico screening of AMR, virulence (Seemann T, Abricate, https://github.com/tseemann/abricate), spa87, and SCCmec typing71. SCCmec mec-containing staphylococcal chromosomal cassete, hIEC human-associated immune evasion cluster, eqIEC equine-associated immune evasion cluster, SaPI S. aureus pathogenicity islands, c1-6 clades 1 to 6, SP susceptible phylogroup, AP Asian phylogroup, AAP American-Asian phylogroup, EP1-5 European phylogroups 1 to 5, Lpg pig-associated lineage, Leq equine-associated lineage, L1-5 major European lineages based on previous studies, LEa-g minor lineages.
Clade two (c2) (n = 591) was also formed by MSSA distant from the MRSA clades, from which they diverged around the beginning of the 1950s [95% C.I.: 1947–1960]. It is composed of a major phylogroup of MSSA commonly isolated from human carriers in Europe and North America (called the susceptible phylogroup—SP), which are predominantly spa type t571 (t1451 predominant in Mexico), of which the majority (92%) carry the MLSB resistance gene erm(T) and an IEC (chp-scn combinations) (95%) which are traits indicative of the human-adapted clade.
Clade three (c3) (n = 355) evolved initially in Asia, where it later split approximately 1957 [95% C.I.: 1952–1964], giving rise to two distinct Asian subclades. The first one comprises a phylogroup of MRSA isolated mostly from human clinical samples in China (n = 148, Asian phylogroup—AP), which were predominantly spa type t034 (n = 107) and carried a chp-sak-scn-encoding IEC. The second subclade was instead formed by MSSA isolated from human carriers and animals (cattle and pigs); this subclade showed no spa type specificity and carried an IEC (chp-scn). Crucially, the first European MRSA phylogroup (n = 77, EP1), which was mostly isolated from human clinical samples in Denmark (also three from the Netherlands, two from Hungary, and four from New Zealand), evolved from the MSSA subclade and, specifically, from a susceptible strain isolated in the UK approximately 1972 [95% C.I.: 1965-1980]. Notably, all the strains from EP1 harbored an IEC (chp-sak-scn), the PVL, and lacked the tet(M)-carrying transposon Tn916, which are characteristic features of the so-called CC398 human-adapted clade17. The EP1 strains also carried the resistance genes mecA and blaZ, ant(9)-Ia, erm(A), and tet(K), which confer resistance to beta-lactams, aminoglycosides, macrolides-lincosamides-streptogramin B (MLSB), and tetracyclines, respectively. A lineage from EP1 containing 39 strains from Denmark that emerged in approximately 2010 also acquired the trimethoprim resistance gene dfrG (Fig. 2).
Between 1957 and 1956 approximately, the jump from humans to livestock which is characterized by the acquisition of the tet(M) gene and the loss of the φSa3 hlb-converting prophage encoding the IEC took place, and subsequently all the LA-MRSA CC398 clades and phylogroups identified in the study emerged35. Clade five (c5) (n = 159) emerged around 1965 [95% C.I.: 1961–1969] and consisted of the denominated European phylogroup 2 (EP2), with mostly MRSA strains (predominantly of spa type t108) that have acquired an SCCmec of type V. These strains were detected among human carriers in the Netherlands; however, they did not seem to have undergone human adaptation since none of them harbored an IEC.
Around 1974 [95% CI: 1971–1978] the remaining clade (c6), which is the main driver of CC398 spread in Europe, diverged forming three branches which included the remaining European MRSA phylogroups, EP3, EP4, and EP5 (Figs. 1 and 2).
The first branch of clade 6 further split forming two phylogroups, the AAP (American-Asian phylogroup) and EP3. AAP (n = 61) consisted of MSSA and MRSA (predominantly spa type t571) isolated mostly from pigs in Asia and North America which did not contain an IEC or a PVL but rather a plethora of resistance genes (Supplementary Data 1). EP3 (n = 104) contained strains belonging predominantly to either spa type t034 or t011 that were isolated mostly from pigs, cattle, chickens, and humans in several European countries (n = 97), such as Switzerland, Germany, Denmark, the Netherlands, Hungary, and Italy but also in China (n = 7). These strains harbored a specific pattern of resistance genes consisting of blaPC1, ant(9)-Ia, erm(A), vga(E), tet(M), tet(K) and dfrG, and only six strains harbored IECs with four different gene combinations (one with only chp, one with sak-scn-selp, one with chp-sak-scn, and three with sak-scn), suggesting individual and isolated reacquisition events of IEC-carrying prophages.
The second branch of c6 corresponded to the European phylogroup 4 (EP4; n = 652) that further diverged into four lineages and several sublineages, including the so-called Danish (also isolated in other countries) pig-associated lineages L1, L3, and L4 and the Italian lineage of spa type t899 (L5) (names of lineages with geographical association were previously defined as such)36. The four lineages of EP4 presented diverse profiles of AMR and virulence. L1 strains predominantly displayed a conserved pattern of genes, including aadD, ant(6)-Ia, spw, lnu(B), lsa(E), tet(M), tet(K), and dfrG, as well as mutations in the fluoroquinolone resistance-determining region of gyrA and grlA. Interestingly, a sublineage of L1 evolving in Germany additionally harbored erm(C), dfrK, and a tarP-carrying Sa1int prophage (Fig. 2). The L3 strains exhibited a conserved resistance pattern with blaPC1, ant(9)-Ia, lnu(B), lsa(E), tet(M), tet(K) and dfrG. Lineage L4 strains carried several different AMR genes and no specific conserved pattern. L5 was composed mostly of spa type t899 strains, which predominantly carried the blaPC, vga(A), and tet(M) resistance genes. It evolved into two sublineages that spread among different host species through Europe: one sublineage characterized by strains carrying a SCCmec type IV and fluoroquinolone resistance-associated mutations in grlA and by the presence of the IEC (chp-sak-scn) in 84% (n = 54) of them (Fig. 2). The second sublineage evolved locally in Germany, the Netherlands, and Italy; contained strains without IECs; and harbored either SCCmec type IV or SCCmec type V (Fig. 2).
From the third branch of c6 emerged around 1984 the larger European phylogroup (EP5, n = 1000) which included two major lineages, each composed of several sublineages. EP5 was characterized by the predominance of spa type t011 strains and, in several sublineages, the presence of SaPIPig2 (n = 549), which carries the scn and vwb genes. The two EP5 lineages diverged approximately 1993 [95% C.I.: 1991–1995]. The first one contained strains isolated from pigs (n = 334) but also from human carriers (n = 250) and was defined as a pig-associated lineage (Lpg, n = 634). Most of the strains (>90%) acquired a SCCmec of type V and contained blaZ, tet(M), and tet(K). The majority of sublineages belonging to Lpg contained strains that were identified in several European countries, such as Denmark, Finland, Switzerland, Germany, the Netherlands, Italy, Hungary, Belgium, the UK, France, Austria, Spain, Slovenia, Greece, Poland, and Luxembourg. However, one sublineage (LEd) of Lpg related to a Swiss sublineage was formed by strains from non-European territories, including strains from New Zealand (n = 15) and Australia (n = 11); half of the strains of this sublineage carried the erm(C) gene. One of the sublineages of Lpg contained strains with additional resistance genes (ant(9)-Ia, lnu(B), lsa(E), and erm(B)) and consisted of the already defined Danish L2 (n = 66, mostly associated with pigs). Our analysis revealed that L2 contained not only strains from Denmark (n = 30, from pigs, humans and cattle) but also from other European countries, such as Italy (n = 13, pig-associated), Hungary (n = 9, environment and humans), Belgium (n = 4, pigs), Finland (n = 2, pigs), Germany (n = 2, pigs), Slovenia (n = 1, human), and five strains from pigs in China.
The second major lineage of EP5 was named Leq (equine lineage, n = 301) since horses represented the principal animal species carrying strains (n = 75) of this lineage. A considerable number of human strains mostly originated from people working with horses as well as from human clinical samples (carriage, n = 137; clinical, n = 30). Leq strains were identified in the Netherlands (n = 130, 95% of which were isolated from humans); Switzerland (n = 59, with 29 from humans and 24 from horses); Germany (n = 39, where it was also reported in cattle); Belgium (n = 29, 10 from pigs, six from cattle, five from humans and chickens); France (n = 14 from humans, horses and dogs); Denmark (n = 13 from horses and human carriers); the UK (n = 11 from horses, humans and cattle); Spain (n = 1 from horse); Italy (n = 1 from pig); and the USA (n = 2 from horse and donkey). Leq strains were characterized by a SCCmec type IV and a conserved pattern of resistance genes (blaZ, aac(6′)-Ie-aph(2ʺ)-Ia, tet(M), and dfrK), as well as by the presence of SaPIpig2 in the majority (n = 266) of them and a Sa1int-tarP carrying prophage in 171 strains (see Comparative Genomics). Notably, the emergence of Leq sublineages with strains from different countries harboring mutations in the fluoroquinolone resistance-determining region of the gyrA and grlA genes (n = 66) and strains in Denmark that lost SaPIpig2 (11 out of 13) and harbored a chp-sak-scn IEC, as well as strains from horses and cattle in Germany that also carried the prophage φSaeq1-associated elements scn-eq, lukP and lukQ, shows the ongoing evolution and adaptation of this lineage to new niches (Fig. 2).
A closer analysis of the Swiss MRSA strains within the international phylogeny revealed that all but two (n = 133) of the strains belonged to the LA-MRSA CC398 clade 6, with the majority clustering within phylogroups EP3 and EP5. Together with other international strains, they formed sublineages LEc, LEd, LEe, LEf, and LEg within EP5 and LEa and LEb within EP3. LEe, LEf, and LEg belong to the denominated equine lineage (Leq) of EP5. LEe and LEf split approximately 1998 [95% C.I.: 1997-2000]. LEe (n = 77) contained strains isolated from humans, cattle, and pigs from several countries, such as Switzerland (n = 46, 22 from horses, 21 from humans), the Netherlands (n = 16, 14 from humans, and two from rodents), the UK (n = 6, horses and human clinical samples), France (n = 3, dog and horses), Belgium (n = 3, horse, pig, and human clinical sample), the USA (n = 2, horse and donkey) and Germany (n = 1, cattle). LEf (n = 56) included Swiss (n = 12, seven from humans, two from horses, one from a pig, one from a dog, and one from the environment) and German (n = 21, 20 isolated from horses and one from turkey) strains. LEg (n = 29) was formed by strains from Denmark (n = 13, 10 from horses and 3 from human carriers), Germany (n = 13, 5 from horses, 7 from cattle, and 1 from a human carrier), and single isolates from Spain (horse), the UK (human clinical) and Switzerland (human carrier). LEc and LEd clustered within the pig lineage (Lpg) of EP5 and contained strains from other animals (e.g., cattle, horses, chickens, rodents) and humans. The LEc split approximately 2002 [95% C.I.: 2001-2004] into three branches, the first consisting of strains from France (n = 2) (human carriage and pig), the second consisting of only strains from Switzerland (n = 12) (9 from cattle and 3 from pigs) and the third consisting of strains from Italy (n = 15, pigs) and France (n = 1, cattle). The sublineage LEd was formed by two branches that diverged approximately 2007 [95% C.I.: 2005–2010]. The first branch of LEd consisted of strains isolated from pigs (n = 7) and humans (n = 4) in Switzerland, and the second branch consisted of strains from human carriers in New Zealand (n = 15) and Australia (n = 11) (Fig. 2).
The two lineages of EP3 (LEa and LEb) have different epidemiological backgrounds. LEa contained only strains isolated in Switzerland from pigs (n = 10), cattle (n = 6), humans (n = 7) and pork meat (n = 2), and LEb contained strains from diverse animals, with half of them isolated from chickens (n = 8) in Switzerland, Germany, and the Netherlands. Less predominant Swiss strains (which may indicate isolated introduction events) belonged to sublineages L1 (n = 4, 3 from cattle and 1 from pig), L3 (n = 3, 2 from chicken and 1 from human), L5-t899 (n = 1, human carriage) and LEg (n = 1, human carriage); this lineage had one branch isolated in Denmark and one predominantly in Germany. Two clinical MRSA strains from humans in Switzerland clustered within the distant phylogroups EP2 and AP, representing isolated events.
Pangenome-based population structure
The reconstructed phylogeny and its respective estimated parameters (i.e., molecular clock and population model) already showed that the CC398 S. aureus population structure, especially that of MRSA, which is widely distributed throughout Europe, is very homogeneous, with an estimated low mean mutation rate of 3.3422 × 10−7 (95% C.I.: 3.0783 × 10−7–3.6188 × 10−7) substitutions per site per year and few nucleotide differences between phylogroups and lineages. Pangenome analysis of the entire collection revealed a total of 6317 genes (Fig. 3a), and the core genome (≥95%) consisted of 2246 loci, representing 385 loci more than the cgMLST scheme publicly available for S. aureus (1861 loci). The accessory genome consisted of 3669 shell genes (present in two or more strains) and 402 cloud genes that were present in only one strain. The pangenome estimated size was 6552 (Chao lower bound) and open with an estimated alpha coefficient of 0.70 (Fig. 3b), and the genomic fluidity between pairs of genomes was estimated to be approximately 0.0679 (S.D.: 0.022, number of simulations 10,000), which showed not only a high degree of overlap between genomes regarding gene cluster carriage but also the openness to assimilating novel determinants that might play a role in niche adaptation.
a Bar pie showing the structure of the pangenome, core genes are those present in >95% of strains, shell genes are present in at least two strains, and cloud genes are singletons. b Rarefaction curves showing the number of unique clusters in the pangenome (violet curve) and the core genome (green curve) as a function of the number of genomes observed. c, d PCA of the pangenome matrix. Circles represent genomes in the two first principal components (directions) of the space, colors represent phylogroups in c, and major lineages in d. Percentages indicate how much of the total variation is observed within each principal component.
A principal component analysis (PCA) based on the accessory genome showed agreement with the clustering resulting from the SNP-based analysis. Furthermore, PC1 revealed a clear separation between the Asian phylogroups (SP and AP) (Fig. 3c) and the majority of the European phylogroups (EP2, EP3, EP4, and EP5), and PC2 showed lineage separation between both major lineages of EP5, namely, Lpg and Leq (Fig. 3d).
Recombination events driving adaptation
Despite the clonal structure of the population, both the population description and the pangenome analysis revealed different patterns of gene content, which allowed us to further stratify the population into lineages and sublineages and confirmed that the acquisition/loss of traits is mostly mediated by MGEs. The chromosomally integrated MGEs carrying AMR and virulence genes were identified via structural analysis based on multiple sequence alignments of our de novo Swiss hybrid assemblies and the completely assembled genomes retrieved from the collection (n = 28). In addition to the already known chromosomal positions where the main drivers of population evolution and adaptation are located, namely, the tet(M)-carrying transposon Tn916 (a defining feature of the livestock-associated MRSA), the mecA-carrying SCCmec and the human IEC-carrying φSa3 prophages17, our analysis identified several other MGEs (i.e., prophages and genomic islands) and two additional recombination hotspots within the CC398 genome.
Prophages
Identification of prophages of the (formerly known as) Siphoviridae family, which are the most common in S. aureus, was performed by in silico screening for phage integrase genes of the most common types, namely, Sa1int (representative prophage φ55), Sa2int (φ47), Sa3int (φ42E), Sa4int (φSauS-IPLA35), Sa5int (φ29), Sa6int (φ77), Sa7int (φ53), Sa8int (φSauS-IPLA88), Sa9int (vB_SauS_Mh1) and Sa12int (MR11), followed by structural comparisons of some of the groups identified. The most abundant prophages were those containing integrase types 3 (n = 1099), 2 (n = 1050), 6 (n = 1001), 9 (n = 693) and 1 (n = 690). Less prevalent were those that possessed integrase types 7 (n = 64), 5 (n = 49), 4 (n = 28), 12 (n = 15), and 8 (n = 6). In this collection of S. aureus CC398 strains, the majority (93,3%, n = 2919) carried at least one prophage, 1130 (36,1%) carried two prophages, 279 (8,9%) harbored three prophages, 19 (<1%) possessed four prophages, and 7 (<1%) carried five different prophages. Notably, 209 (6,7%) of the isolates had no prophages integrated into their chromosome (Fig. 4b). The Sa1int prophages were predominantly found (i.e., in more than 50% of the strains) in EP3, L3 (EP4) and Leq (EP5); Sa2int were mostly identified in EP1, EP2 and Lpg (EP5); Sa6int, in c1 and EP3; most of the lineages of EP4 (except L5) and Leq (EP5); and Sa3int prophages were the most common type among the human-clade strains (Fig. 4a).
a Total carriage of different integrase families-associated prophages and a subtotal of those harboring the virulence factor tarP. b Multi-carriage of prophages. c, d, e, and f Detailed, lineage subdivided carriage of prophages of major phylogroups and the rest of the population studied, namely EP4, EP5, SP, and the remaining strains, respectively. g Hybrid tarP-Sa1int prophage harboring sea and sak virulence factors.
The prophages varied in terms of structure and gene content, and the insertion sites in the CC398 chromosome were quite conserved among the prophage integrase types. Sa1int prophages were mostly integrated at the 3’ end of the iron-sulfur (Fe-S) assembly protein-coding gene sufB; Sa2int prophages, at the 3’ end of the srrB gene; Sa3int, in the bacterial sphingomyelinase gene hlb, disrupting its expression (thus called hlb-converting bacteriophages); Sa4int, at the 3’ end of the ygdQ gene; Sa5int, near the 5’ end of the queE gene; Sa6int, 141 bp upstream of the 3’ end of the lipase-encoding gene geh; and Sa7int, at the intergenic region between the 5’ end of the isdB gene and the 3’ end of the rpmF gene. Sa8int and Sa9int were commonly found downstream of the 3’ end of a tRNA-Ser coding gene, and Sa9int prophages were also found integrated at the 3’ terminus of the tmRNA-coding gene ssrA, which is a common insertion for phages in Gram-negative bacteria. Sa12int was integrated at the 3’ end of the rclA gene.
In addition to Sa3int, whose spread among S. aureus CC398 strains has been widely documented and associated with virulence and host adaptation6, less common virulence-carrying prophages (i.e., non-Sa3int) were identified and further analyzed to determine the variability in structure and gene content. Sa1int phages were the fourth most abundant, and almost 70% of them (n = 482/690) carried the immune evasion factor gene tarP (Fig. 4a), which encodes a protein that glycosylates wall teichoic acid polymers, preventing them from becoming immunogenic and preventing host defense mechanisms from acting. The majority of tarP-Sa1int prophages (n = 335/482) were carried by EP5 strains, which suggests that after the acquisition, most of the lineages remained stable. However, in some sublineages, such as LEd, tarP-Sa1int was lost, and in LEg, a 110-bp deletion event occurred in the gene encoding the phage transcriptional activator RinA, which likely suppressed the formation of infectious phages37. The TarP determinant was additionally found in 193 Sa9int, 23 Sa3int, 3 Sa2int-like, 3 Sa5int, and 1 Sa12int prophage (Fig. 4a). A closer comparison of their structures revealed that tarP was the only common genetic feature among them, with the remaining structures differing for each integrase type.
In addition to the tarP-carrying and Sa3int strains harboring IEC determinants, most of the prophages identified had very diverse mosaic genomes, and the majority did not carry any virulence determinants. A few exceptions were a Sa7int variant carried by three related strains of EP4 isolated from pigs in Denmark that harbored the enterotoxin A and the staphylokinase genes sea and sak; a Cre recombinase associated with a pig strain in Taiwan that harbored the sak and hlb (phospholipase C) genes; and a tarP-containing Sa1int prophage in a pig strain from Denmark (EP4-L1) that also harbored the sea and sak genes. This prophage exhibited a hybrid structure with the module containing Sa1int and tarP and a fragment of approximately 25 kbp containing sea and sak, which shared 98.38% identity with the Sa3int prophage of a strain from Russia that harbored the sea-sak-scn virulence factors (Fig. 4g). To our knowledge, this is the first report of a Sa1int prophage carrying these important virulence factors.
Genomic islands
The in silico screening of this CC398 population revealed that 1335 (42.7%) strains carried at least one SaGI and that 50 (1.6%) of them harbored two. Several different SaGIs, including some not previously described, were identified. These SaGIs included integrases of one of four families (Supp. Fig. 1) that target four known CC398 chromosomal sites. The most common integrases belonged to the tyrosine recombinase family and targeted a site downstream of the guaA gene at minute 9 (9’, nomenclature in minutes taken from ref. 38) within the larger genomic island νSaα, which carries genes of a staphylococcal superantigen-like (ssl) cluster (Fig. 5). The most common SaPIs, namely, SaPIpig1 and SaPIpig2/SaPIbov5, were integrated at this site. SaPIpig1 was present in 435 (13.9%) strains, and SaPIpig2/SaPIbov5 was present in 706 (22.5%) strains. Hybrid SaPIpig1/2-like elements containing the integration/regulation module of SaPIpig1 and the virulence module of SaPIpig2 (SaPIpig1-2) or vice versa (SaPIpig2-1) were identified in 10 and four strains, respectively, revealing the plasticity and modularity of these pathogenicity islands. Notably, the SaPIpig1 isolated from a pig in Germany was inserted at a different chromosomal location truncating the zinc metalloproteinase aureolysin-encoding gene aur, indicating that the tyrosine recombinase integrase family responsible for mobilization of SaPIpig-like islands can also integrate at different sites in the chromosome. An additional 13,923-bp novel resistance island (SaRI), which shares the backbone of SaPIpig1 and harbors the tellurite resistance-associated gene tehA (Fig. 5), was found in 47 strains (Reference strain APHA03, GenBank acc. no. ERR1437617) from different countries and hosts. The second most common integrase family of SaGIs targets the intergenic region between the ktrB and groL genes (44’). SaPIs integrated at this site (n = 105) carried a variety of toxin-encoding genes, such as the enterotoxin genes sea, seb, sec, or the toxic-shock syndrome gene tsst, many of which have not been previously described. The third family of integrases targets a chromosomal site at the 3’ end of the metQ gene (19’). Thirteen SaGIs were identified at this position, 6 of which were SaPIs that carried the enterotoxin genes sek and seq (Fig. 5). The last family of integrases was located after the 3’ end of the rpsR gene, which encodes the small ribosomal subunit protein bS18 (8’). Among the eleven elements inserted at this site, seven were SaPIs harboring virulence genes such as tsst or scn. Notably, the 3’ end of the tyrosine recombinases mobilizing these SaGIs consists of a variable region of 1 to 4 kbps (additional variable end region—AVR), which may contain additional antimicrobial resistance genes. Among them, a novel SaRI was identified that carried the trimethoprim resistance gene dfrG flanked by IS1 (Fig. 5).
Pairwise analysis of CC398 SaGIs, including a novel dfrG-containing SaRI and several not previously described enterotoxins harboring SaPIs. Image constructed using clinker v.0.0.28. CDS coding sequence, IS insertion sequence, int integrase, scn staphylococcal complement inhibitor, SaPI S. aureus Pathogenicity Island, SaRI S. aureus Resistance Island, SaPRI S. aureus Pathogenicity and Resistance Island, AVR additional variable end region.
Recombination hotspots
Two chromosomal regions of CC398 were identified as hotspots for the insertion of variable transposable elements (TEs) (Fig. 6). The first hotspot was located at the 3’ end of the radC gene, where at least six different transposable elements were identified within the livestock-associated clade (Fig. 6a). All these TEs had a core of three transposition-associated genes (tnp-like) and a variable accessory gene content. The transposon Tn558, which carries the florfenicol-chloramphenicol exporter gene fexA, was found predominantly in strains of AAP, and in some lineages, it also carried the 23 S rRNA methyltransferase gene cfr (which provides resistance to chloramphenicol, florfenicol and linezolid) mobilized by IS256 (Fig. 6a). The transposon Tn554, which harbors the MLSB resistance gene erm(A) and the spectinomycin 9-adenyltransferase gene ant(9)-Ia, was predominantly present in strains from EP1. The Tn554 variant Tn6133, which additionally contains the lincosamide, streptogramin A, and pleuromutilin (LSAP) vga(E) genes, is characteristic of EP3 strains. The transposon Tn5406, which carries the LSAP resistance gene vga(A), was found predominantly in strains from the L5 lineage of EP4 (Fig. 6a).
Pairwise comparison of representative strains from identified phylogroups and lineages to identify recombinant fragments involved in the resulting differences observed in AMR and virulence profiles. Transposons inserted at the hotspot downstream of radC are shown in a and those downstream of a lactonase gene in b. Image constructed using clinker v.0.0.28. CDS coding sequence, AMR antimicrobial resistance, HMR heavy metal resistance, IS insertion sequence, Tn transposon, att attachment site, c1–6 clades 1 to 6, EP1-5 European phylogroups 1 to 5, Leq equine lineage, LE European lineages, L3 Danish lineage 3.
The second hotspot was located at the 5’ end of the lactonase gene, where TEs of different lengths and with different gene contents were inserted to truncate the yolD gene (Fig. 6b). For instance, strains from EP2, EP4, and EP5 carried Tn554-like transposons consisting of a backbone formed by the gene cluster blaI-blaR-blaZ and transposase genes. The majority of strains from EP5-Leq (n = 271/301) carried the same transposon backbone but with an additional module carrying the aminoglycoside resistance gene aac(6’)-Ie—aph(2”)-Ia flanked by IS256 inserted upstream of blaZ, and 96% (n = 97/101) of the EP3 strains had, instead of the resistance genes, an 8898 bp fragment containing an arsenic resistance gene cluster (arsR, arsB, arsC). Importantly, these arsenic-resistant strains of EP3 also carried blaZ, as part of a novel SCCmec type Vc variant. This novel variant acquired an additional recombinant fragment (9836 bp) that carries the cluster blaI-blaR-blaZ and the trimethoprim resistance gene dfrG (Fig. 7).
Pairwise nucleotide comparison of the SCCmec element from a representative strain from c6-EP3 as well as the reference SCCmec V extracted from the GenBank. Graphical representation created with clinker v.0.0.28. CDS coding sequence, AMR antimicrobial resistance, HMR heavy metal resistance.
Discussion
We performed an in-depth molecular characterization of a worldwide collection of S. aureus CC398 strains. A timed phylogeny based on a global molecular clock was reconstructed using the genomes of five continental regions-spanning collection of MSSA and MRSA CC398 strains. The available metadata, such as the isolation date and location, permitted the consideration of geographic and temporal dimensions to enrich the phylogenetic description. This time-calibrated whole-genome-based phylogeny showed the presence of five phylogroups, EP1 to EP5, which spread across Europe, diverged decades ago from other MSSA lineages prevalent in Asia and North America, and further evolved over time. These phylogroups were defined based on the identification of a specific epidemiologically relevant fingerprint and included several related lineages and sublineages, some of which were identified and described in other European countries, such as the Danish and Italian lineages39, as well as novel lineages (Lpg and Leq). For instance, Leq consisted of MRSA CC398-t011-IVa, which was predominantly isolated from horses and humans, and Lpg by MRSA CC398-t011-Vc, which was isolated from pigs, cattle, and humans. Each of the lineages also further evolved independently into sublineages according to their origin and type of host. Such geographic, host‐related, and time-framed analyses allowed for the determination of relatedness among strains of diverse hosts and geographical origins. Under the assumption of a coalescent population evolving according to a molecular clock which was consistent but lower than reported before for S. aureus, the clustering of strains from different geographical origins indicates international spread and the introduction of successful foreign CC398 clones into local populations. For instance, several strains isolated from pigs in Switzerland clustered within the EP4 lineages L1 and L3, which are mainly formed by MRSA CC398-t034-Vc isolated from pigs in Denmark and Hungary. The close relatedness of strains from Denmark, Hungary, and Switzerland indicates that a recent common ancestor was likely spread through animal trading. Indeed, a recent study from Hungary demonstrated the genetic relatedness between the MRSA CC398 strains from pigs from Denmark and those from farms in Hungary that imported pigs from Denmark40. Although Switzerland has an indigenous pig production system, a few hundred animals are imported each year for breeding purposes, which could also have contributed to the introduction of the L1 and L3 lineages into Swiss pig production. Otherwise, the majority of MRSA CC398 strains (>90%) isolated from pigs and cattle in Switzerland belonged to specific distant new lineages of clade 6 (Lpg in EP5 and LEa in EP3) indicating several introductions of MRSA CC398 in these animal populations. The presence of these novel lineages may also suggest that MRSA CC398 of Lpg and LEa evolved independently from MSSA in Switzerland. However, the absence of MSSA CC398 genomic data from pigs and cattle did not permit to verify this eventuality.
MRSA CC398 isolated from horses in Germany, Switzerland, Denmark, and the Netherlands belonged to the new specific lineage Leq within EP5. Crucially, MRSA of CC398-t011-IVa, which is the defining feature of Leq, has been detected and associated with sporadic outbreaks in equine settings all around Europe, including countries such as Austria41, Hungary42, Spain43, the UK44, France45, and the Netherlands46,47, which highlights a broader dissemination potential. Obtaining WGS data from previously identified but not yet sequenced as well as newly detected MRSA CC398-t011-IVa would allow us to map them into this phylogeny, refine the population description, and identify specific evolution of the Leq lineage. Thus, identifying global and local adaptation. Such an adaptation was observed particularly in the sublineage LEg in Germany, which acquired the eqIEC and has not only been isolated from horses but also from cattle. Furthermore, further studies are necessary to identify specific genetic traits that may contribute to the successful colonization and infection of equines by Leq strains and to understand their potential to be successful among other host species including humans. Indeed, genetic relatedness between MRSA CC398 strains from animals and those from humans was observed for each lineage and genetic evidence supported the idea of recent transmissions, a phenomenon mainly associated with occupational risk exposure, as already documented in several studies39,48,49,50. Notably, most human MRSA strains (carriage and clinical) in Switzerland as well as a considerable number in the Netherlands were CC398-t011-IVa and were closely related to strains from the horse-associated lineage (Leq). Considering that people working with horses are at risk of also being colonized and the fact that the exchange of clinic personnel from different countries through, for instance, residency and exchange programs between different universities and clinics are very common during academic formation, horse clinic personnel may contribute to the introduction of MRSA Leq into novel equine clinics and further dissemination to the community51,52. Livestock workers and veterinarians are also known to be at higher risk of being colonized and developing LA-MRSA infections15, however, further metadata are necessary to determine whether humans who developed infections with horse-related MRSA strains were at occupational risk or had contact with people associated with horses or other animals. Importantly, longitudinal studies are required to better understand if the cases of MRSA CC398 human carriage are due to transient or long-term host adaptation-mediated colonization. Although the reconstructed time-framed phylogeny indicated that spillover events occurred often and recently due to the close relatedness among strains, additional metadata would confirm this phenomenon. The creation of specific MRSA CC398 international databases aiming to share WGS, as well as metadata in a harmonized way, would facilitate a quicker and more effective analysis of the dynamics of S. aureus at a local and global level.
The in silico screening for AMR, virulence, and MGE further revealed the genetic features characterizing the strains of each lineage. The horse-associated MRSA strains of lineage EP5-Leq (CC398-t011-IVa) exhibited similar antimicrobial and virulence patterns. Most EP5 strains contained either SaPIpig1 or SaPIpig2/SaPIbov5, which are two variants of SaPIbov4. These SaPIs harbor the genes scn and vwb which code for the Staphylococcus complement inhibitor and the von Willebrand factor-binding proteins, respectively8,53. While scn may contribute to improved colonization and adaptation in humans, the vwb proteins of SaPIpig1, SaPIpig2, and SaPIbov4 have been shown to coagulate the plasma of several animals, including pigs, sheep, and bovines, but not equine plasma8,53. The equine-related scn-eq variant mobilized by the φSaeq1 prophage, which also carries the equine-specific leukocidin-encoding gene lukPQ, was only carried by MRSA from an Leq German sublineage (LEg) circulating in horses and cattle that also harbored a scn-encoding SaPI. The absence of additional equine-adapted genes suggests that additional unidentified factors may be responsible for adaptation and pathogenicity in both horses and humans. Nevertheless, the presence of scn may explain the ability of MRSA CC398 of lineage Leq to easily colonize and be maintained in the personnel working with horses with the risk of causing opportunistic infections in humans. On the other hand, SaPIpig2 and SaPIbov4 were not present in MRSA CC398 from pigs and cattle in Switzerland, although they were originally identified in porcine and bovine MRSA in other countries8,53,54,55, indicating the complex dynamics at play when following the evolution and adaptation of MRSA CC398 lineages and further highlighting local evolution a lineage.
In addition to factors that may have contributed to host adaptation, selection may also have been driven by long-term usage of antibiotics and the presence of antimicrobial resistance genes specific to the different lineages of MRSA CC398. For instance, the aac(6’)-Ie-aph(2”)-Ia tandem gene was exclusively detected among MRSA strains of the horse lineage Leq and confers resistance to aminoglycosides such as gentamicin, which is commonly used in combination with β-lactams in equine medicine56,57. Antimicrobial resistance profiles are also very well conserved among the strains of the different lineages, further emphasizing clonal dissemination. Notably, both the β-lactam and aminoglycoside resistance genes were located on the same mobile genetic element in some MRSA strains of Leq. Similarly, the multidrug resistance transposon Tn6133, which contains the lincosamide-pleuromutilin resistance gene vga(E), the MLSB resistance gene erm(A), and the spectinomycin nucleotidyltransferase gene ant(9)-Ia58, was only present in livestock MRSA from EP3. These antibiotics have been commonly used to treat infectious diseases in livestock, and their use may have contributed to the selection and maintenance of ST398-t034/t011-Vc harboring Tn6133 in food-producing animals. Notably, MRSA ST398 of EP3 also harbors a novel SCCmecVc element, which is characterized by the integration of the β-lactamase gene blaZ and the trimethoprim-resistant dihydrofolate reductase gene dfrG, further providing a broader selective advantage for maintenance of MRSA in the animal population when undergoing antimicrobial treatment. Importantly, a recent study has shown that MGEs carrying resistance and virulence genes such as transposons and phages may contribute to bacterial survival in blood and host adaptation59, indicating that AMR use might not be the only significant factor influencing the selection of MRSA CC398 and that MGE may have an impact in their ability to adapt to different host species.
This study provided a broad overview of the structure, phylodynamic trajectory, and chromosome-associated mobilome of the global population of CC398 S. aureus with a focus on the emergence and evolution of MRSA CC398. Considering the worldwide historical situation, lineages derived from both European phylogroups were identified in Switzerland. The global expansion of CC398 MRSA among livestock and horses during recent decades has been driven mainly by the expansion of the animal-adapted clade number 6 and specifically two major successful phylogroups (EP4 and EP5) that further diverged forming new lineages with specific AMR and adaptive traits. These results also confirmed the spread and adaptation of MRSA CC398 among livestock and the occurrence of spillover events between hosts and, crucially, humans, colonizing professionals at risk but also causing human infections. Nevertheless, a current limitation of this study is the restricted availability of MRSA CC398 WGS data from certain regions such as Africa, North America, or certain European and Asian countries. Data from European countries such as Denmark, the Netherlands, Germany, or Italy are abundant due to the fact that LA-MRSA CC398 has emerged and spread remarkably in their territories and that these countries have used a WGS-based approach to study the phenomenon. MRSA CC398 WGS data from other countries may so far not be available due to either the absence of MRSA CC398 in the animal and human population, lack of WGS-based MRSA surveillance programs, or absence of WGS-based analyses. Therefore, the reconstructed phylogeny obtained in this study is based on current available WGS data, resulting in some countries and regions being either overrepresented or underrepresented, thus highlighting the need for international collaborations to better understand the global expansion by refining this analysis and closely monitoring the evolution and spread of MRSA CC398. Importantly, this study reconfirmed previous findings and added novel data, providing an updated baseline for future epidemiological studies, which will permit us to determine at the local level whether the introduction of a strain into a clinical or farm setting is recent and what is the most likely origin of the strain. Additionally, such a WGS-based baseline could be used to further monitor the global expansion of some lineages in regions where MRSA CC398 may emerge or its presence has so far not been investigated. Therefore, continuous, and collaborative WGS-based long-term monitoring and surveillance of MRSA from a global One Health perspective are necessary to follow the evolutionary dynamics of MRSA CC398 and to identify successful and emerging lineages, as well as the evolutionary processes that gave rise to them, which may further spread and cause infections in both animals and humans.
Methods
Ethics statement
The samples from food-producing animals were taken at the slaughterhouse and the animal clinical samples were taken by veterinarians for diagnostic purposes therefore not requiring ethical approval or a permit for animal experimentation according to the current Swiss legislation (Federal Animal Protection Law, 455 (https://www.fedlex.admin.ch/eli/cc/2008/414/en).
Origin of bacterial strains and whole-genome sequences
A total of 134 CC398 MRSA strains representing all isolated from animals, meat, and people working with animals in Switzerland between 2007 and 2020 were obtained from the Center for Zoonoses, Animal Bacterial Diseases and Antimicrobial Resistance (ZOBA) at the University of Bern and from previous research projects. The animal MRSA strains originated from samples taken at slaughterhouses from the nasal cavities of healthy cattle (n = 21) and pigs (n = 23); from chicken meat (n = 10) and pork meat taken at retail (n = 3); from infection sites of dogs (n = 1) and cats (n = 1)60; from infection sites as well as nasal cavities of horses (n = 24) (ref. 51 and Institute collection) and from the environment (n = 2). MRSA from healthy people working with animals originated from the nasal cavities of veterinary healthcare providers (n = 23) and farmers (n = 6)27,48,51. The clinical human strains (n = 20) isolated from patients in the German-speaking part of Switzerland were already sequenced, and the WGS reads were obtained from the SPSP. Genomic data from previous international MRSA CC398 studies as well as all novel available (to the best of our knowledge) WGS reads or assembled genomes of international S. aureus CC398 strains were downloaded from the Sequence Read Archive (SRA) or the National Center for Biotechnology Information (NCBI) genome database. After the identification and removal of duplicates, the collection consisted of sequences from 2994 S. aureus CC398 strains (2046 MRSA and 948 MSSA) from diverse countries spanning five continental regions. The aim was to obtain a collection of MRSA CC398 (over MSSA) covering all parts of the world with available data. However, in certain regions (e.g., America or Asia) CC398 is not the predominant MRSA lineage and most of the available data belonged to MSSA CC398. As already shown by previous studies, MRSA CC398 evolved from MSSA so to understand the evolutionary dynamics and the geographical spread of CC398 we included MSSA data from underrepresented MRSA regions to increase the scope of the analysis. The available metadata, including geographical origin, host, genetic characteristics, BioProjects, or GenBank/SRA accession numbers of the S. aureus CC398 strains used in this study, are listed in Supplementary Data 1.
Whole-genome sequencing and assembly
The genome sequences of the MRSA CC398 Swiss strains were obtained from Nextera libraries prepared with the Illumina® DNA Prep (M) Tagmentation Kit (Illumina). Libraries were sequenced on an Illumina system to obtain 2 × 150 bp paired-end short reads at the Next Generation Sequencing Platform, Institute of Genetics, Vetsuisse Faculty, University of Bern. The remaining WGS reads were downloaded from the SRA (see above). Two isolates from each of the three prevalent Swiss lineages identified were additionally sequenced using Nanopore MinION long-read technology to obtain a complete assembled circular genome for use as a reference scaffold for comparative analysis. WGS reads that were not previously assembled were processed and analyzed using Trimommatic61 for the initial steps of removing sequencing adaptors and quality-based filtering of the raw reads with a window size of 20 bp and a quality threshold of 30. Subsequent assembly was performed using the normal mode of the Unicycler v0.4.9 assembler62 with default parameters. The six-lineage representative Swiss strains that were sequenced using both short- and long-read technologies were assembled using the hybrid assembly mode of Unicycler v0.4.9 with default parameters.
Time-calibrated phylogeny
A total of 3128 genome sequences, including those of the 136 S. aureus CC398 strains from Switzerland and the 2992 international strains, were subjected to comparative genomic analysis, and a time-calibrated phylogeny was constructed. All the de novo assemblies were aligned against the MRSA ST398 type strain reference genome (GenBank accession number NC_017333.1), and Snippy version 4.4.5 (https://github.com/tseemann/snippy) was used to call the core genome single nucleotide polymorphisms (cg-SNPs). Recombination events were detected and filtered out from the alignments using Gubbins version 2.4.163. Before inferring the time-calibrated phylogeny, the existence of a temporal signal in the data was assessed and confirmed by using TempEst v1.5.364 and Clockor265 which supported the utilization of a global clock (Supp. Figs. 2 and 3). Subsequently, based on the recombination-free cgSNPs (19,899 sites) obtained previously, a timed phylogenetic tree was constructed using BEAST266. For model selection, we used nested sampling (NS)67 to choose between the strict and optimized relaxed molecular clock models. The XML files were created with BEAUti2 using the isolation date, a GTR site, a strict clock (based on the NS results), and two different coalescent population models, constant and exponential (both with a log-normal prior). To account for constant sites, we used beast2_constsites (https://github.com/andersgs/beast2_constsites) to add this information and obtain the final XML file. The MCMC was set to run for 300 million steps logging every 10,000 with a 10% burn-in. Each MCMC configuration was run twice to confirm convergence, and completeness was assessed by checking the Effective Sample Size of each parameter of interest to be above 200 with Tracer v1.7.268. The results from both runs for each configuration were combined, and based on Bayes factor analysis, the model with a strict clock and a coalescent exponential population was chosen as the best model for fitting the data. A maximum clade credibility tree was constructed using TreeAnnotator v2.7.1 and edited with iTOL69.
spa and SCCmec typing
The assembled contigs were analyzed with the commercial software suite Ridom SeqSphere+ version 8.5.170 for spa typing. SCCmec typing was performed using the the staphophia-sccmec standalone version 1.0.0 from the Staphophia pipeline71 for all samples and the SCCmecFinder 1.2 at the Center for Genomic Epidemiology, Technical University of Denmark (https://cge.food.dtu.dk/services/SCCmecFinder/)72 for the Swiss MRSA and those that were not successfully typed by staphophia-sccmec.
Antimicrobial resistance and virulence gene profiling
Antimicrobial resistance (AMR) and virulence determinants were screened in silico using the abricate tool against Resfinder73 and the Virulence Factor Database (VFDB)74, which assess presence based on nucleotide similarity. AMR findings were reconfirmed using Resistance Gene Identifier (RGI) software based on the CARD database75. For some resistance and virulence factor genes that are not recognized by the abovementioned software (e.g., SaPI-associated virulence factors), manual screening was performed using BLAST.
Population structure
To establish a reference framework that would expand the current knowledge about CC398 population structure in Europe and consider the important role that recombination events have played in the evolution and adaptation of S. aureus populations, we defined a criterion that considered the genetic similarity and divergence time among strains as well as epidemiologically relevant information such as molecular characteristics (spa and SCCmec type), antimicrobial resistance (AMR) and virulence profile, geographical location, isolation host and available information from previous studies (i.e., lineage definitions). As a result, we established three nested layers to describe the population structure by means of clades, phylogroups, and lineages. Clades refer to the products of major divergence events that took place more than 45 years ago; phylogroups, are those that had specific genomic and epidemiological characteristics; and lineages, which denote smaller clades that diverged more recently (less than 45 years ago).
Pangenome reconstruction
All CC398 genomes were annotated using Prokka with default parameters76, and the graph-based clustering tool Panaroo77 was subsequently utilized to reconstruct the pangenome using the “strict” mode to remove putative contamination and erroneous annotations. After running the pipeline, we filtered out the sequences that were of unusual length or were flagged as pseudogenes from the resulting presence/absence matrix using the provided algorithm (panaroo-filter-pa), and the results with associated metadata were analyzed in R using the Pagoo framework to execute a principal component analysis (PCA), define the structure of the pangenome78, and use the micropan package79 to estimate pangenome size and openness according to Heap’s law model and compute genomic fluidity.
Identification of lysogenic prophages and genomic islands
To detect lysogenic prophages and SaGI, a manual approach termed “nested screening” was used and proceeded as follows. Known S. aureus phage and SaPI integrases available in the Nucleotide database of the NCBI were extracted including phage integrases Sa1int (representative prophage φ55; NC_007060, YP_240491), Sa2int (φ47; NC_007054, YP_240030), Sa3int (φ42E; NC_007052, YP_239890), Sa4int (φSauS-IPLA35; NC_011612, YP_002332364), Sa5int (φ29; NC_007061, YP_240566), Sa6int (φ77; NC_005356, NP_958623), Sa7int (φ53; NC_007049, YP_239679), Sa8int (φSauS-IPLA88; NC_011614, YP_002332477), Sa9int (vB_SauS_Mh1; OM439673, UKM36468.1) and Sa12int (MR11; NC_010147, YP_001604091.1) as well as SaPI integrases from SaPIpig1 (MW589252, QXJ78686.1), SaPIbov (AF217235, AAG29618.1), SaPIbov4 (HM211303, ADN95150.1), SaPIbov5 (HM228919, ADN53643.1), Ishikawa11 (AB716350, BAM66903.1), Hirosaki4 (AB716352, BAM66938.1), SaPIj11 (AB704541, BAM66859.1), IVM10 (AB716349, BAM66883.1), SaPIhhms2 (AB704540, BAM66842.1), and an initial in silico screening using BLAST (blastn/tblastn)80 was performed for their presence using a threshold for positive identification of >90% coverage and >90% identity, as well as manual inspection of known insertion sites along the chromosome, as described previously by Novick et al.81,82 (Step 1). Fragments containing an integrase with its flanking regions were extracted from positive strains in the collection and aligned with the same fragment (i.e., flanking regions) from a SaGI-negative strain to identify the whole structure of the elements and the exact site in the chromosome where they were inserted (Step 2). A region spanning 2-3 “representative” genes (add_frag) was extracted from each element, and a second screening was performed. Strains that had a hit for an integrase/flanking region but not for the add_frag region were manually inspected, allowing for the identification of novel structures (Step 3). Step 3 was iteratively performed for all the elements initially identified, and novel “representative” regions were identified for further screening.
Structural analysis
Complete or semicomplete assemblies (i.e., those that possessed contigs with complete MGEs and flanking chromosomal regions) were subjected to pairwise and multiple comparisons via the progressive algorithm of Mauve83 to describe in detail the structure of the identified prophages and SaGIs. For the final visualization of complete elements, we used Clinker v.0.0.2884.
Integrase gene relatedness
A multiple alignment of the nucleotide sequences of 22 SaGI integrases was performed using Clustal Omega 1.2.285, and a phylogenetic tree based on the alignment was reconstructed using RAxML 8.2.1186 with the rapid hill-climbing algorithm and a GTR GAMMA I substitution model.
Statistics and reproducibility
No statistical method was used to predetermine the sample size. No data were excluded from the analyses.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
The genomic data of all the strains used in this study are available in the NCBI GenBank and NCBI Sequence Read Archive databases under accession numbers listed in Supplementary Data 1. Other relevant data supporting the findings presented are available within the paper and its Supplementary material.
References
Haag, A. F., Fitzgerald, J. R. & Penades, J. R. Staphylococcus aureus in animals. Microbiol. Spectr. 7, 0060 (2019).
Mehraj, J. et al. Epidemiology of Staphylococcus aureus nasal carriage patterns in the community. Curr. Top. Microbiol. Immunol. 398, 55–87 (2016).
Cheung, G. Y. C., Bae, J. S. & Otto, M. Pathogenicity and virulence of Staphylococcus aureus. Virulence 12, 547–569 (2021).
Dayan, G. H. et al. Staphylococcus aureus: the current state of disease, pathophysiology and strategies for prevention. Expert Rev. Vaccines 15, 1373–1392 (2016).
Laumay, F., Benchetrit, H., Corvaglia, A. R., van der Mee-Marquet, N. & Francois, P. The Staphylococcus aureus CC398 lineage: an evolution driven by the acquisition of prophages and other mobile genetic elements. Genes (Basel) 12, 1752 (2021).
Rohmer, C. & Wolz, C. The role of hlb-converting bacteriophages in Staphylococcus aureus host adaption. Micro. Physiol. 31, 109–122 (2021).
Guinane, C. M., Penades, J. R. & Fitzgerald, J. R. The role of horizontal gene transfer in Staphylococcus aureus host adaptation. Virulence 2, 241–243 (2011).
Viana, D. et al. Adaptation of Staphylococcus aureus to ruminant and equine hosts involves SaPI-carried variants of von Willebrand factor-binding protein. Mol. Microbiol. 77, 1583–1594 (2010).
Lee, A. S. et al. Methicillin-resistant Staphylococcus aureus. Nat. Rev. Dis. Prim. 4, 18033 (2018).
Mlynarczyk-Bonikowska, B., Kowalewski, C., Krolak-Ulinska, A. & Marusza, W. Molecular mechanisms of drug resistance in Staphylococcus aureus. Int. J. Mol. Sci. 23, 8088 (2022).
Tsouklidis, N., Kumar, R., Heindl, S. E., Soni, R. & Khan, S. Understanding the fight against resistance: hospital-acquired methicillin-resistant Staphylococcus aureus vs. community-acquired methicillin-resistant Staphylococcus aureus. Cureus 12, e8867 (2020).
Lakhundi, S. & Zhang, K. Methicillin-resistant Staphylococcus aureus: molecular characterization, evolution, and epidemiology. Clin. Microbiol. Rev. 31, e00020–18 (2018).
Schwendener, S. & Perreten, V. The blaand mecfamilies of beta-lactam resistance genes in the genera Macrococcus, Mammaliicoccus and Staphylococcus: an in-depth analysis with emphasis on Macrococcus. J. Antimicrob. Chemother. 77, 1796–1827 (2022).
Liu, J. et al. Staphylococcal chromosomal cassettes mec (SCCmec): A mobile genetic element in methicillin-resistant Staphylococcus aureus. Micro. Pathog. 101, 56–67 (2016).
Chen, C. & Wu, F. Livestock-associated methicillin-resistant Staphylococcus aureus (LA-MRSA) colonisation and infection among livestock workers and veterinarians: a systematic review and meta-analysis. Occup. Environ. Med. 78, 530–540 (2021).
Price, L. B. et al. Staphylococcus aureus CC398: host adaptation and emergence of methicillin resistance in livestock. mBio 3, e00305–e00311 (2012).
Matuszewska, M. et al. Stable antibiotic resistance and rapid human adaptation in livestock-associated MRSA. Elife 11, e74819 (2022).
Lindsay, J. A. Genomic variation and evolution of Staphylococcus aureus. Int. J. Med. Microbiol. 300, 98–103 (2010).
Sharma, M. et al. Livestock-associated methicillin resistant Staphylococcus aureus (LA-MRSA) clonal complex (CC) 398 isolated from UK animals belong to European Lineages. Front. Microbiol. 7, 1741 (2016).
Sahibzada, S. et al. Transmission of highly virulent community-associated MRSA ST93 and livestock-associated MRSA ST398 between humans and pigs in Australia. Sci. Rep. 7, 5273 (2017).
Cuny, C. et al. Emergence of methicillin-resistant Staphylococcus aureus (MRSA) in different animal species. Int. J. Med. Microbiol. 300, 109–117 (2010).
Cuny, C. et al. Nasal colonization of humans with methicillin-resistant Staphylococcus aureus (MRSA) CC398 with and without exposure to pigs. PLoS One 4, e6800 (2009).
Witte, W., Strommenger, B., Stanek, C. & Cuny, C. Methicillin-resistant Staphylococcus aureus ST398 in humans and animals, Central Europe. Emerg. Infect. Dis. 13, 255–258 (2007).
Graveland, H., Duim, B., van Duijkeren, E., Heederik, D. & Wagenaar, J. A. Livestock-associated methicillin-resistant Staphylococcus aureus in animals and humans. Int. J. Med. Microbiol. 301, 630–634 (2011).
Graveland, H., Wagenaar, J. A., Bergs, K., Heesterbeek, H. & Heederik, D. Persistence of livestock associated MRSA CC398 in humans is dependent on intensity of animal contact. PLoS ONE 6, e16830 (2011).
Garcia-Graells, C. et al. Livestock veterinarians at high risk of acquiring methicillin-resistant Staphylococcus aureus ST398. Epidemiol. Infect. 140, 383–389 (2012).
Wettstein Rosenkranz, K. et al. Nasal carriage of methicillin-resistant Staphylococcus aureus (MRSA) among Swiss veterinary health care providers: detection of livestock- and healthcare-associated clones. Schweiz. Arch. Tierheilkd. 156, 317–325 (2014).
Wulf, M. W. et al. Prevalence of methicillin-resistant Staphylococcus aureus among veterinarians: an international study. Clin. Microbiol. Infect. 14, 29–34 (2008).
Davis, M. F. et al. Genome sequencing reveals strain dynamics of methicillin-resistant Staphylococcus aureus in the same household in the context of clinical disease in a person and a dog. Vet. Microbiol. 180, 304–307 (2015).
Smith, T. C. & Wardyn, S. E. Human infections with Staphylococcus aureus CC398. Curr. Environ. Health Rep. 2, 41–51 (2015).
McAdam, P. R. et al. Molecular tracing of the emergence, adaptation, and transmission of hospital-associated methicillin-resistant Staphylococcus aureus. Proc. Natl Acad. Sci. USA 109, 9107–9112 (2012).
Uhlemann, A. C. et al. Molecular tracing of the emergence, diversification, and transmission of S. aureus sequence type 8 in a New York community. Proc. Natl Acad. Sci. USA 111, 6738–6743 (2014).
Young, B. C. et al. Evolutionary dynamics of Staphylococcus aureus during progression from carriage to disease. Proc. Natl Acad. Sci. USA 109, 4550–4555 (2012).
Ward, M. J. et al. Time-scaled evolutionary analysis of the transmission and antibiotic resistance dynamics of Staphylococcus aureus clonal complex 398. Appl. Environ. Microbiol. 80, 7275–7282 (2014).
Stegger, M. et al. Rapid differentiation between livestock-associated and livestock-independent Staphylococcus aureus CC398 clades. PLoS One 8, e79645 (2013).
Sieber, R. N. et al. Genome investigations show host adaptation and transmission of LA-MRSA CC398 from pigs into Danish healthcare institutions. Sci. Rep. 9, 18655 (2019).
Ferrer, M. D. et al. RinA controls phage-mediated packaging and transfer of virulence genes in Gram-positive bacteria. Nucleic Acids Res. 39, 5866–5878 (2011).
Subedi, A., Ubeda, C., Adhikari, R. P., Penades, J. R. & Novick, R. P. Sequence analysis reveals genetic exchanges and intraspecific spread of SaPI2, a pathogenicity island involved in menstrual toxic shock. Microbiol. (Read.) 153, 3235–3245 (2007).
Sieber, R. N. et al. Drivers and dynamics of methicillin-resistant livestock-associated Staphylococcus aureus CC398 in pigs and humans in Denmark. mBio 9, e02142–18 (2018).
Albert, E. et al. High prevalence of livestock-associated methicillin-resistant in Hungarian pig farms and genomic evidence for the spillover of the pathogen to humans. Transbound Emerg. Dis. 2023, 5540019 (2023).
Cuny, C., Strommenger, B., Witte, W. & Stanek, C. Clusters of infections in horses with MRSA ST1, ST254, and ST398 in a veterinary hospital. Micro. Drug Resist. 14, 307–310 (2008).
Albert, E. et al. Outbreaks of a methicillin-resistant Staphylococcus aureus clone ST398-t011 in a Hungarian Equine Clinic: emergence of rifampicin and chloramphenicol resistance after treatment with these antibiotics. Micro. Drug Resist. 25, 1219–1226 (2019).
Gomez-Sanz, E. et al. First detection of methicillin-resistant Staphylococcus aureus ST398 and Staphylococcus pseudintermedius ST68 from hospitalized equines in Spain. Zoonoses Public Health 61, 192–201 (2014).
Bortolami, A. et al. Environmental surveillance identifies multiple introductions of MRSA CC398 in an Equine Veterinary Hospital in the UK, 2011-2016. Sci. Rep. 7, 5499 (2017).
Guerin, F. et al. Nationwide molecular epidemiology of methicillin-resistant Staphylococcus aureus responsible for horse infections in France. BMC Microbiol. 17, 104 (2017).
Van den Eede, A. et al. High occurrence of methicillin-resistant Staphylococcus aureus ST398 in equine nasal samples. Vet. Microbiol. 133, 138–144 (2009).
van Duijkeren, E. et al. Methicillin-resistant Staphylococcus aureus in horses and horse personnel: an investigation of several outbreaks. Vet. Microbiol. 141, 96–102 (2010).
Kittl, S., Brodard, I., Heim, D., Andina-Pfister, P. & Overesch, G. Methicillin-resistant strains in Swiss pigs and their relation to isolates from farmers and veterinarians. Appl. Environ. Microbiol. 86, e01865–19 (2020).
Parisi, A. et al. High occurrence of methicillin-resistant in horses at slaughterhouses compared with those for recreational activities: a professional and food safety concern? Foodborne Pathog. Dis. 14, 735–741 (2017).
Sun, J. et al. Longitudinal study of Staphylococcus aureus colonization and infection in a cohort of swine veterinarians in the United States. BMC Infect. Dis. 17, 690 (2017).
Sieber, S. et al. Evolution of multidrug-resistant Staphylococcus aureus infections in horses and colonized personnel in an equine clinic between 2005 and 2010. Micro. Drug Resist. 17, 471–478 (2011).
Khairullah, A. R. et al. A review of horses as a source of spreading livestock-associated methicillin-resistant to human health. Vet. World 15, 1906–1915 (2022).
Ji, X. et al. Comparative analysis of genomic characteristics, fitness and virulence of MRSA ST398 and ST9 isolated from China and Germany. Emerg. Microbes Infect. 10, 1481–1494 (2021).
Shoaib, M. et al. MRSA compendium of epidemiology, transmission, pathophysiology, treatment, and prevention within one health framework. Front. Microbiol. 13, 1067284 (2022).
Lienen, T., Schnitt, A., Hammerl, J. A., Maurischat, S. & Tenhagen, B. A. Genomic distinctions of LA-MRSA ST398 on dairy farms from different German federal states with a low risk of severe human infections. Front. Microbiol. 11, 575321 (2020).
Schnellmann, C. et al. Presence of new mecA and mph(C) variants conferring antibiotic resistance in Staphylococcus spp. isolated from the skin of horses before and after clinic admission. J. Clin. Microbiol. 44, 4444–4454 (2006).
Weese, J. S. et al. Methicillin-resistant in horses at a veterinary teaching hospital: frequency, characterization, and association with clinical disease. J. Vet. Intern. Med. 20, 182–186 (2006).
Schwendener, S. & Perreten, V. New transposon Tn6133 in methicillin-resistant Staphylococcus aureus ST398 contains vga(E), a novel streptogramin A, pleuromutilin, and lincosamide resistance gene. Antimicrob. Agents Chemother. 55, 4900–4904 (2011).
Ba, X. et al. High-throughput mutagenesis reveals a role for antimicrobial resistance- and virulence-associated mobile genetic elements in Staphylococcus aureus host adaptation. Microbiol. Spectr. 11, e0421322 (2023).
Wipf, J. R. & Perreten, V. Methicillin-resistant Staphylococcus aureus isolated from dogs and cats in Switzerland. Schweiz Arch. Tierheilkd. 158, 443–450 (2016).
Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30, 2114–2120 (2014).
Wick, R. R., Judd, L. M., Gorrie, C. L. & Holt, K. E. Unicycler: resolving bacterial genome assemblies from short and long sequencing reads. PLoS Comput. Biol. 13, e1005595 (2017).
Croucher, N. J. et al. Rapid phylogenetic analysis of large samples of recombinant bacterial whole genome sequences using Gubbins. Nucleic Acids Res. 43, e15 (2015).
Rambaut, A., Lam, T. T., Max Carvalho, L. & Pybus, O. G. Exploring the temporal structure of heterochronous sequences using TempEst (formerly Path-O-Gen). Virus Evol. 2, vew007 (2016).
Featherstone, L. A., Rambaut, A., Duchene, S. & Wirth, W. Clockor2: inferring global and local strict molecular clocks using root-to-tip regression. Syst. Biol. syae003, https://doi.org/10.1093/sysbio/syae003 (2024).
Bouckaert, R. et al. BEAST 2.5: an advanced software platform for Bayesian evolutionary analysis. PLoS Comput. Biol. 15, e1006650 (2019).
Russel, P. M., Brewer, B. J., Klaere, S. & Bouckaert, R. R. Model selection and parameter inference in phylogenetics using nested sampling. Syst. Biol. 68, 219–233 (2019).
Rambaut, A., Drummond, A. J., Xie, D., Baele, G. & Suchard, M. A. Posterior summarization in Bayesian phylogenetics using tracer 1.7. Syst. Biol. 67, 901–904 (2018).
Letunic, I. & Bork, P. Interactive Tree Of Life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res. 49, W293–W296 (2021).
Junemann, S. et al. Updating benchtop sequencing performance comparison. Nat. Biotechnol. 31, 294–296 (2013).
Petit, R. A. 3rd & Read, T. D. Staphylococcus aureus viewed from the perspective of 40,000+ genomes. PeerJ 6, e5261 (2018).
Kaya, H. et al. SCCmecFinder, a web-based tool for typing of Staphylococcal Cassette chromosome mec in Staphylococcus aureus using whole-genome sequence data. mSphere 3, e00612–e00617 (2018).
Zankari, E. et al. Identification of acquired antimicrobial resistance genes. J. Antimicrob. Chemother. 67, 2640–2644 (2012).
Chen, L., Zheng, D., Liu, B., Yang, J. & Jin, Q. VFDB 2016: hierarchical and refined dataset for big data analysis–10 years on. Nucleic Acids Res. 44, D694–D697 (2016).
Alcock, B. P. et al. CARD 2023: expanded curation, support for machine learning, and resistome prediction at the Comprehensive Antibiotic Resistance Database. Nucleic Acids Res. 51, D690–D699 (2023).
Seemann, T. Prokka: rapid prokaryotic genome annotation. Bioinformatics 30, 2068–2069 (2014).
Tonkin-Hill, G. et al. Producing polished prokaryotic pangenomes with the Panaroo pipeline. Genome Biol. 21, 180 (2020).
Ferres, I. & Iraola, G. An object-oriented framework for evolutionary pangenome analysis. Cell Rep. Methods 1, 100085 (2021).
Snipen, L. & Liland, K. H. micropan: an R-package for microbial pan-genomics. BMC Bioinforma. 16, 79 (2015).
Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic local alignment search tool. J. Mol. Biol. 215, 403–410 (1990).
Novick, R. P., Christie, G. E. & Penades, J. R. The phage-related chromosomal islands of Gram-positive bacteria. Nat. Rev. Microbiol. 8, 541–551 (2010).
Novick R. P. Pathogenicity islands and their role in staphylococcal biology. Microbiol. Spectr. 7, 0062 (2019).
Darling, A. E., Mau, B. & Perna, N. T. progressiveMauve: multiple genome alignment with gene gain, loss and rearrangement. PLoS One 5, e11147 (2010).
Gilchrist, C. L. M. & Chooi, Y. H. clinker & clustermap.js: automatic generation of gene cluster comparison figures. Bioinformatics 37, 2473–2475 (2021).
Sievers, F. et al. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol. Syst. Biol. 7, 539 (2011).
Stamatakis, A. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 30, 1312–1313 (2014).
Bartels, M. D. et al. Comparing whole-genome sequencing with Sanger sequencing for spa typing of methicillin-resistant Staphylococcus aureus. J. Clin. Microbiol. 52, 4305–4308 (2014).
Acknowledgements
We are grateful to Alexandra Collaud (Institute of Veterinary Bacteriology, University of Bern, Switzerland) and Helena Seth-Smith (Institute of Medical Microbiology, University of Zurich, Switzerland) for their technical support. We also thank Vinzenz Gerber and his team for providing horse clinic samples. This study was financed by the Swiss National Science Foundation (SNSF) grant no. 177504 “Development of a Swiss surveillance database for molecular epidemiology of multidrug-resistant pathogens” within the National Research Program NRP72 ‘Antimicrobial Resistance’ to A.E. and V.P. and by internal funds (REF-660-50) of the Institute of Veterinary Bacteriology, University of Bern, Switzerland to V.P.
Author information
Authors and Affiliations
Contributions
The project was conceived by V.P. and J.E.F. The data analysis was performed by J.E.F. under the supervision of V.P. G.O. provided strains and data for LA-MRSA from food-producing animals. A.E. supervised and provided the WGS data of the clinical human strains. The manuscript was drafted and edited by J.E.F. and V.P. and revised by all the authors.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Rory Cave, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Fernandez, J.E., Egli, A., Overesch, G. et al. Time-calibrated phylogenetic and chromosomal mobilome analyses of Staphylococcus aureus CC398 reveal geographical and host-related evolution. Nat Commun 15, 5526 (2024). https://doi.org/10.1038/s41467-024-49644-9
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-024-49644-9
- Springer Nature Limited