Genomic analysis of almost 8,000 Salmonella genomes reveals drivers and landscape of antimicrobial resistance in China

ABSTRACT Foodborne Salmonella infection remains a major public health concern worldwide. With rising antimicrobial resistance, genomic surveillance is key to tracking outbreaks and monitoring transmission, but there is no comprehensive national surveillance scheme for Salmonella involving humans, food, animals, and the environment in China. Moreover, the association between antimicrobial resistance and climate, social, and economic factors has rarely been investigated. Here, we use 1,962 Salmonella isolates collected from 22 Chinese provinces and add 6,035 publicly available genomes to build a Chinese local Salmonella genome database version 2 (CLSGDB v2) representing 30 Chinese provinces, covering 1905–2022. Using the CLSGDB v2, we mapped the landscape and spatiotemporal dynamics of antimicrobial resistance markers, virulome, and mobilome in Salmonella. We identified 317 mcr positive and 745 azithromycin resistance genes positive Salmonella isolates out of 7,997 isolates. We further uncovered the geographic distribution veil of mcr-1, fosA7, fosA3, mph(A), and bla CTX-M-55 genes in China, all of them resistant to the critically important antimicrobials including colistin, fosfomycin, azithromycin, and the third-generation cephalosporins. Interestingly, economic, climatic, and social factors can drive the rise of antimicrobial resistance was observed. Finally, we release the CLSGDB v2 as an open-access database and thus can assist surveillance studies tracking 164 Salmonella enterica serovars and 295 sequence types across the globe. The CLSGDB v2 is freely available at https://nmdc.cn/clsgdbv2. IMPORTANCE We established the largest Salmonella genome database from China and presented the landscape and spatiotemporal dynamics of antimicrobial resistance genes. We also found that economic, climatic, and social factors can drive the rise of antimicrobial resistance. The Chinese local Salmonella genome database version 2 was released as an open-access database (https://nmdc.cn/clsgdbv2) and thus can assist surveillance studies across the globe. This database will help inform interventions for AMR, food safety, and public health.


Figure S1 .
Figure S1.Summary of the 3721 Salmonella enterica strains of human origin in the CLSGDB v2.

Figure S2 .
Figure S2.Summary of the 1811 Salmonella enterica strains of chicken origin in the CLSGDB v2.

Figure S3 .
Figure S3.Summary of the 1123 Salmonella enterica strains of pig origin in the CLSGDB v2.

Figure S4 .
Figure S4.Summary of the 373 Salmonella enterica strains of aquatic animal origin in the CLSGDB v2.

Figure S5 .
Figure S5.Summary of the 230 Salmonella enterica strains of duck origin in the CLSGDB v2.

Figure S6 .
Figure S6.Summary of the 367 Salmonella enterica strains of environment origin in the CLSGDB v2.

Figure S7 .
Figure S7.Shared and unique serovars among aquatic animals, chickens, ducks, pigs, humans, and the environment in the CLSGDB v2.

Figure S8 .
Figure S8.ST prevalence is grouped by sampling periods (A), hosts (B), and regions (C) in the CLSGDB v2.

Figure S9 .
Figure S9.Shared and unique STs among aquatic animals, chickens, ducks, pigs, humans, and the environment in the CLSGDB v2.

Figure S11 .
Figure S11.Temporal changes in AMR among different isolation sources in the CLSGDB v2.

Figure S12 .
Figure S12.The distribution of the number of ARGs per isolate among humans, the different animal host groups, and the environment in the CLSGDB v2.

Figure S13 .
Figure S13.Differences in MDR rates between (A) geographic regions and (B) STs in the CLSGDB.

Figure S14 .
Figure S14.ARG prevalence grouped sampling period among different isolation sources in the CLSGDB v2.

Figure S15 .
Figure S15.Shared and unique ARGs among aquatic animals, chickens, ducks, pigs, humans, and the environment in the CLSGDB v2.

Figure S16 .
Figure S16.Temporal changes in virulence genes among different isolation sources in the CLSGDB v2.

Figure
Figure S17.VG prevalence was grouped by sampling periods, hosts, serovars, and regions in the CLSGDB v2.

Figure S18 .
Figure S18.The distribution of the number of virulence genes per isolate among the top

Figure S19 .
Figure S19.SPI prevalence was grouped by sampling periods, hosts, regions, and serovars in the CLSGDB v2.

Figure S20 .
Figure S20.Temporal changes in plasmid replicons among different isolation sources in the CLSGDB v2.

Figure S21 .
Figure S21.The distribution of the number of plasmid replicon genes per isolate among the top 10 serovars in different time groups.

Figure S22 .
Figure S22.The distribution of the number of plasmid replicons per isolate among different hosts.

Figure S23 .
Figure S23.Plasmid prevalence grouped sampling period among different isolation sources in the CLSGDB v2.

Figure S24 .
Figure S24.Shared and unique plasmid replicons among aquatic animals, chicken, duck, pig, humans, and environment in the CLSGDB v2.

Figure S27 .
Figure S27.The correlation of annual grain output with ARGs, VGs, and MGEs.

Figure S28 .
Figure S28.The correlation of mean counts of ARGs, plasmid replicons, MGEs, and VGs detected from genomes in the CLSGDB v2.

Figure S1 .
Figure S1.Summary of the 3721 Salmonella enterica strains of human origin in the CLSGDB v2.A, Origins of the 3721 Salmonella enterica isolates from 25 Chinese provinces.B, All isolates were grouped by serogroup.C, Top 10 serovars of human origin among the CLSGDB v2.A total of 117 serovars were detected.A total of 188 STs were detected.

Figure S2 .
Figure S2.Summary of the 1811 Salmonella enterica strains of chicken origin in the CLSGDB v2.A, Origins of the 1811 Salmonella enterica isolates from 20 Chinese provinces.B, All isolates were grouped by serogroup.C, Top 10 serovars of chicken origin among the CLSGDB v2.A total of 58 serovars were detected.A total of 79 STs were detected.

Figure S3 .
Figure S3.Summary of the 1123 Salmonella enterica strains of pig origin in the CLSGDB v2.A, Origins of the 1123 Salmonella enterica isolates from 21 Chinese provinces.B, All isolates were grouped by serogroup.C, Top 10 serovars of pig origin among the CLSGDB v2.A total of 54 serovars were detected.A total of 75 STs were detected.

Figure S4 .
Figure S4.Summary of the 373 Salmonella enterica strains of aquatic animal origin in the CLSGDB v2.A, Origins of the 373 Salmonella enterica isolates from 8 Chinese provinces.B, All isolates were grouped by serogroup.C, Top 10 serovars of aquatic animal origin among the CLSGDB v2.A total of 81 serovars were detected.A total of 107 STs were detected.

Figure S5 .
Figure S5.Summary of the 230 Salmonella enterica strains of duck origin in the CLSGDB v2.A, Origins of the 230 Salmonella enterica isolates from 13 Chinese provinces.B, All isolates were grouped by serogroup.C, Top 10 serovars of duck origin among the CLSGDB v2.A total of 20 serovars were detected.A total of 25 STs were detected.

Figure S6 .
Figure S6.Summary of the 367 Salmonella enterica strains of environment origin in the CLSGDB v2.A, Origins of the 367 Salmonella enterica isolates from 13 Chinese provinces.B, All isolates were grouped by serogroup.C, Top 10 serovars of environment origin among the CLSGDB v2.A total of 77 serovars were detected.A total of 98 STs were detected.

Figure S7 .
Figure S7.Shared and unique serovars among aquatic animals, chickens, ducks, pigs, humans, and the environment in the CLSGDB v2.

Figure S8 .
Figure S8.ST prevalence was grouped by sampling periods (A), hosts (B), and regions (C) in the CLSGDB v2.A total of 295 STs were identified in the CLSGDB v2, and STs that were detected at > 1% prevalence in Salmonella genomes in the CLSGDB v2 are shown.

Figure S9 .
Figure S9.Shared and unique STs among aquatic animals, chickens, ducks, pigs, humans, and the environment in the CLSGDB v2.

Figure S10 .
Figure S10.Temporal changes in AMR phenotypes among different isolation sources in the CLSGDB v2.

Figure S11 .
Figure S11.Temporal changes in AMR among different isolation sources in the CLSGDB v2.Detail of dynamic trends in antimicrobial resistance in the CLSGDB v2.

Figure S12 .
Figure S12.The distribution of the number of ARGs per isolate among humans, the different animal host groups, and the environment in the CLSGDB v2.

Figure S13 .
Figure S13.Differences in MDR rates between (A) geographic regions and (B) STs in the CLSGDB v2.

Figure S14 .
Figure S14.ARG prevalence grouped sampling period among different isolation sources in the CLSGDB v2.

Figure S15 .
Figure S15.Shared and unique ARGs among aquatic animals, chickens, ducks, pigs, humans, and the environment in the CLSGDB v2.

Figure S17 .
Figure S17.VG prevalence was grouped by sampling periods, hosts, serovars, and regions in the CLSGDB v2.VGs that were detected from 1% to 86% in Salmonella genomes in the CLSGDB v2 are shown.

Figure S18 .
Figure S18.The distribution of the number of virulence genes per isolate among the top 10 serovars in different time groups.The mean VGs count for each group is noted above.The '*' on the top represents Pvalues.*: P < 0.05; **: P < 0.01; ***: P < 0.001.

Figure S19 .
Figure S19.SPI prevalence was grouped by sampling periods, hosts, regions, and serovars in the CLSGDB v2.SPI2 and SPI3 which were detected 100% in the top 10 serovars of Salmonella genomes in the CLSGDB v2 are not shown.

Figure S21 .
Figure S21.The distribution of the number of plasmid replicon genes per isolate among the top 10 serovars in different time groups.The mean plasmid replicons count for each group is noted above.The '*' on the top represents P-values.*: P < 0.05; **: P < 0.01; ***: P < 0.001.

Figure S22 .
Figure S22.The distribution of the number of plasmid replicons per isolate among different hosts.

Figure S23 .
Figure S23.Plasmid prevalence grouped sampling period among different isolation sources in the CLSGDB v2.

Figure S24 .
Figure S24.Shared and unique plasmid replicons among aquatic animals, chickens, ducks, pigs, humans, and the environment in the CLSGDB v2.

Figure S25 .
Figure S25.MGE prevalence was grouped by sampling periods (A), hosts (B), serovars (C), and regions (D) in the CLSGDB v2.A total of 50304 plasmid replicons classified into 1002 types were identified in the CLSGDB v2, replicons that were detected at >1% prevalence (except for MITEEc1 and cn_5129_ISVsa3) in Salmonella genomes in the CLSGDB v2 are shown.

Figure S26 .
Figure S26.Shared and unique MGEs among aquatic animals, chickens, ducks, pigs, humans, and the environment in the CLSGDB v2.

Figure S28 .
Figure S28.The correlation of mean counts of ARGs, plasmid replicons, MGEs, and VGs detected from genomes in the CLSGDB v2.A, The correlation between plasmid replicons and ARG counts.B, The correlation between MGEs and ARG counts.C, The correlation between MGEs and plasmid replicons counts.D, The correlation between ARGs and VGs counts.E, The correlation between plasmid replicons and VGs counts.F, The correlation between MGEs and VGs counts.