Temporal and geographical distribution of Salmonella Typhi isolatesAll blaTEM harboring genomic sequences were downloaded, re-confirmed serovar Typhi using SISTR and classified as S. Typhi harboring blaTEM isolates (n = 6079; 31.4%). While isolates that did not harbor blaTEM were classified as non-blaTEM harboring S. Typhi (n = 13,275; 68.6%) (Fig. 1 & Supplementary Table 1). Most blaTEM variant were blaTEM-1B (n = 6037), but we also found blaTEM-135 (n = 8), blaTEM-116 (n = 1), blaTEM-215 (n = 1) and blaTEM-others (n = 32). Among all blaTEM-positive isolates, there were 851 isolates showing co-existence with ESBL genes. One of the most common ESBL genes was blaCTX-M-15β-lactamase gene (n = 834/851) (Supplementary Table 2).Fig. 1Acquisition and distribution of Salmonella Typhi isolates (harboring vs non-harboring blaTEM). Acquisition of S. Typhi harboring blaTEM isolates from publicly available online database (A). The distribution of S. Typhi isolates (harboring vs non-harboring blaTEM) shows as the world map (B), year of isolation (C) and country of origin (D). Legend, green is S. Typhi harboring blaTEM isolates (n = 6079) and grey is non-harboring blaTEM isolates (n = 13,275).Our dataset revealed information of S. Typhi isolates, spanning from 1916 to 2023 with the first blaTEM-positive isolates observed from Chile in 1983 (Supplementary Table 3). Although the number of blaTEM harboring isolates has been increasing based on sequencing input, the proportion of blaTEM-positive isolates remained relatively stable over time. The blaTEM-positive isolates were predominantly from Asia (Southern, South-eastern) and Africa (Eastern, Western and Southern) and had a more geographically restricted distribution than blaTEM-negative isolates (Fig. 1).Association of bla
TEM to genotypesThe distribution of blaTEM-positive S. Typhi isolates was also compared to genotypes. The blaTEM-negative isolates exhibited greater genotypic diversity of genotype than the blaTEM-positive isolates and the later isolates were mainly found in genotype 4.3.1.1, 4.3.1.1.EA1, 4.3.1.1.P1, 3.1.1, 4.3.1.3.Bdq and and 2.5.1, mainly from Asia and Africa (Fig. 2 and Supplementary Fig. 1 & Supplementary Table 4).Fig. 2Distribution of Salmonella Typhi harboring blaTEM isolates. The distribution of the selected isolates (n = 6079) is divided by year (A), country (B) and genotype (C).Associations between phylogeny, genotypes, plasmid and resistance profiles of bla
TEM positive isolatesThe 6079 S. Typhi harboring blaTEM isolates were used for constructing maximum-likelihood phylogenetic tree. SNP-analyses showed large genomic diversity among the blaTEM-positive isolates, with the phylogenetic tree in general clustering with the defined genotypes (Supplementary Fig. 2). However, some discrepancies were also observed. Thus, especially genotype 4.3.1.1 clustered in groups into different parts of the phylogeny, genotypes 2.3.1, 2.3.4, 2.5.1, 3, 3.1.1 and 3.2.1 closely together, as did genotypes 4.3.1, 4.3.1.2, 4.3.1.2.EA2, 4.3.2.2.EA3, 4.3.1.3 and 4.3.1.3.Bdq. There was also some clustering of isolates from the same countries, but in several cases, clones were observed across several countries and even continents. The antibiotic resistance gene (ARG) profile also showed association to phylogeny. Almost all blaTEM-positive isolates also harbored catA1, dfrA7, sul1 and sul2. However, exceptions were observed (Supplementary Fig. 2).A limited number of isolates harbored blaCTX-M-15, but even though this was identified in two different resistance profiles and across the different phylogenetic clusters, no clear association to plasmids was observed. One exception was a very clear association between resistance profile 10 (negative for catA1, tet(B) and sul2) and plasmid profile 6 (IncFIB(K) among isolates from Bangladesh (Supplementary Fig. 2). All ARG profiles were classified in to 16 profiles (Supplementary Fig. 3 and Supplementary Table 5). All plasmid profiles were classified in to 13 plasmid profiles (Supplementary Table 6).Analysis of bla
TEM flanking region patternsIt was only possible to obtain sufficient flanking regions (2000 bp on each side) for 740 isolates (12.2%) (Supplementary Fig. 4 and Supplementary Table 7). Those isolates were not equally distributed across the phylogenetic tree, but were mainly from genotype 3.1.1 in Nigeria and Ghana, 4.3.1.3.Bdq in Bangladesh, 4.3.1.1.P1 in Pakistan and 4.3.1.1.EA1 in Zambia and Tanzania (Fig. 3 and Supplementary Figs. 5 and 6). Flanks of 5000 bp were available from 577 isolates (Supplementary Fig. 7 and Supplementary Table 8).Fig. 3Association of flanking region patterns with analyzed genomic data. The flanking region patterns (1) are linked to genotype (2), antibiotic resistance gene (ARG) profile (3) and plasmid replicon profile (4) (A). The S. Typhi harboring blaTEM isolates with flankophile outputs (n = 740) are used for constructing a SNP Tree using CT18 as reference strain. The tree has 36,350 SNP and 76.5% coverage. The distribution of flanking region pattern of S. Typhi harboring blaTEM isolates shows as the world map (B). The dynamics of flanking region patterns is shown between 2004 and 2023 (C). The symbols of seven common flanking region patterns, pattern 01 (five point star), pattern 02 (triangle), pattern 03 (square), pattern 04 (circle), pattern 05 (oval), pattern 06 (hexagon) and pattern 07 (multipoint star), depict linkage of the flanking region pattern with other metadata.The flankophile outputs were then categorized into 13 different patterns (01–13). Among them, there were 7 dominant patterns, namely pattern 01 (n = 240), pattern 02 (n = 167), pattern 03 (n = 48), pattern 04 (n = 22), pattern 05 (n = 128), pattern 06 (n = 19) and pattern 07 (n = 80) (Supplementary Fig. 4 and Supplementary Table 7).The seven dominant flanking patterns (01–07) were closely linked to genotype, ARG profile and plasmid profile, but less to country of origin and isolation year (Fig. 3 and Supplementary Figs. 5 and 6).Flanking region pattern (Pattern) 01 clustered with genotype 3.1.1-mainly isolated from Nigeria and Ghana, plasmid profile 5 and ARG profile 5. However, Pattern 03 clustered phylogenetically here and with a different plasmid profile namely IncY (profile 1) and ARG profile 14. Pattern 03 was only linked with genotype 3.1.1 and was mainly isolated from Nigeria. However, in 3 isolates the pattern was linked to USA (n = 2) and UK (n = 1) (Fig. 3 and Supplementary Figs. 5 and 6).Pattern 02 was associated with 4.3.1.3.Bdq, mainly originated from Bangladesh and IncFIB(K) (plasmid profile 6) and ARG profile 10. Pattern 04 was observed among 22 isolates and distributed quite broadly across the phylogeny, but associated to plasmid profile 7 (IncN) (Fig. 3 and Supplementary Figs. 5 and 6).Pattern 05 clustered with genotype 4.3.1.1.P1, was mainly isolated from Pakistan and associated with ARG profile 4 and plasmid pattern 1 (IncY). Pattern 06 was also related to genotype 4.3.1.1.P1, ARG profile 4 and plasmid profile 1 and was observed from various countries, Pakistan (n = 5), UK (n = 5), USA (n = 3). Pattern 07 clustered with 4.3.1.1.EA1, was isolated from Africa (Zambia and Tanzania), mainly associated to ARG profile 15 and plasmid profile 6 (IncFIB(K) (Fig. 3 and Supplementary Figs. 5 and 6).The major clustering of SNP-phylogeny and flanking region pattern is also confirmed and illustrated using tanglegram (Fig. 4).Fig. 4Clustering of flanking region patterns of S. Typhi harboring blaTEM isolates. The clustering is linked between a flankophile tree constructed using Flankophile and a SNP tree constructed using CSIPhylogeny. The clustering represents association between flanking region pattern (1) and genotype (2), ARG profile (3), plasmid replicon profile (4) and country (5). The symbols of seven common flanking region patterns, pattern 01 (five point star), pattern 02 (triangle), pattern 03 (square), pattern 04 (circle), pattern 05 (oval), pattern 06 (hexagon) and pattern 07 (multipoint star), depict linkage of the flanking region pattern with other metadata. We further analyzed the genetic contexts surrounding blaTEM, specifically focusing on insertion sequences, within a 5000 bp flanking region using the Flankophile tool. Pattern 01 was included upstream Tn2 (n = 240). Pattern 02 was contained upstream Tn2 and downstream ISVsa3 and IS5075 (n = 160). Pattern 03 was contained upstream ISAs25 and ISEc63 (n = 47). Pattern 04 was included upstream Tn2 (n = 20). Pattern 05 was comprised upstream ISEcp1 and downstream IS5075 (n = 11). Pattern 07 was upstream ISEc63 and downstream IS5075 (n = 75). Overall, Tn2 emerged as the most common upstream element, while IS5075 was identified as the dominant downstream element (Supplementary Fig. 7 and Supplementary Table 8).All flanking region patterns were illustrated individually for clustering with metadata profile (genotype, ARG profile, plasmid profile and country) (Supplementary Figs. 8–20).
The dynamics of blaTEM resistance genes in Salmonella Typhi
