Tóm tắt Luận án Antimicrobial resistance characteristics and related genes of multidrug-Resistant salmonella

elevant drug resistance. 1.3. Contamination and antibiotic resistance of Salmonella in food 1.3.1. In the world There have been many different studies on Salmonella contamination rate in food which have been published. In general, the research results show that Salmonella strains are distributed differently depending on the geographical region and the food source. 6 The common serotypes vary by geographic region and their rates of antibiotic resistance in Salmonella are increasing each year. 1.3.2. In Vietnam Recent reports show that Vietnam food is contaminated Salmonella at different rates, including multiple antibiotic-resistant species. The prevalence of Salmonella in food and the rate of multi- antibiotic resistance in Salmonella isolates are increasing every year. Therefore, it is necessary to isolate and determine the antibiotic resistance characteristics of Salmonella in each food type in a particular geographical region at different time periods. 1.4. Methods are commonly used in gene expression research Currently, there are four methods are used: Reverse transcription PCR (RT-PCR), Real-time PCR (qPCR), Microarray, and next generation sequencing (NGS) to study gene expression. Among the above techniques, the next generation sequencing has the most advantage, this technique can overcome the disadvantages of the remaining three techniques and is the only one capable of detecting new genes. Next generation sequencing NGS is a direct measurement of nucleic acid sequences present in the sample. There is the linear relationship between the number of sequences and the concentration of nucleic acid sequences present in the sample. Moreover, NGS is not dependent on the information of nucleic acid sequences and highly homologous genes that can be expressed in the sample. Thus new genes can be detected. Among the next generation sequencing technologies, the Illumina technology produces the most accurate data, the procedure is simple, and is widely applied in many different research fields. Currently, 7 more than 90% of the sequence data in the world is generated by Illumina technology. Chapter 2. MATERIAL AND METHODS 2.1. MATERIAL 2.1.1. Samples A total of 90 meat samples, including 30 porks (from TL-1 to TL-30), 30 chicken meats (from TG-1 to TG-30), and 30 beef samples (from TL-1 to TL-30). symbols from TB-1 to TB-30), randomly collected at 10 markets in Hanoi. 2.1.2. Culture media, chemicals, antibiotics, and kits Culture media, antibiotics, and kits are purchased from reputable companies globally. 2.1.3. Research equipment The equipments are used at prestigious lab in Vietnam, such as Vietnam Academy of Science and Technology, National Institute of Burn, Vietnam Medical Military University. 2.2. METHODS 2.2.1. Sampling Samples were collected according to TCVN 4833-2002, from 7-8 am in the winter season, from October to December 2016. 2.2.2. Identification of Salmonella Salmonella was detected according to ISO 6579: 2002. 2.2.3. Antibiotics susceptibility Salmonella’s antibiotics susceptibility was testing using Kirby-Bauer diffuse method. 2.2.4. Nex generation sequencing We choosed some of multidrug resistant Salmonella to transcriptome sequence using Illumina's technology. 8 2.2.5. Bioinformatics methods The sequence quality is checked by FastQC software. The adapter sequences and noise sequences were removed by Trimmomatic 0.32 software. The sequence at Q20 quality score was de novo assembled using Geneious R11 software. The de novo sequence was annotated by different databases such as RAST, PATRIC 3.5.2, BASys, and Geneious R11 software. Identification of antibiotic resistance genes using: ResFinder, ARG-ANNOT, CARD, and PATRIC 3.5.2. Identify the gene mutation resistance to quinolone, by ResFinder tool. 2.2.6. Confirm antimicrobial resistant gene mutations using Sanger sequencing To confirm the predicted results of antibiotic resistance gene mutations obtained from the next generation sequencing method, the new gene mutations related to quinolone resistance will be confirmed using Sanger sequencing from cDNA of the samples. 2.2.7. Statistical analyzed Using SPSS 16.0 software to calculate the χ2, Fisher exact test, p values. Chapter 3. RESULTS AND DISCUSSION 3.1. Isolate and identify the Salmonella serotype 3.1.1. Identify Salmonella results After non-selective enrichment, selective enrichment and biochemical confirmation, we have obtained Salmonella spp. from the research samples. The results are presented in Table 3.1 9 Table 3.1. Results of identify Salmonella in the research samples Sample Total (number) Positive Negative number rate (%) number rate (%) Chicken 30 11 36,7 19 63,3 Pork 30 9 30,0 21 70,0 Beef 30 5 16,7 25 83,3 Total 90 25 27,8 65 72,2 χ2 = 3,102; df = 2; p = 0,212 The list of Salmonella positive isolated sources is presented in Table 3.2 Table 3.2. List of samples positive for Salmonella The prevalence of Salmonella in this study was 27.8%, this result is in line to that of Do Ngoc Thuy et. al. (30%). However, this rate is lower than other studies such as Ta et. al. (48.7%), Nguyen et. al. (69.7%), Boomar et. al. (80%). This lower result may be due to collecting samples in the morning, when meat was fresh, limiting bacterial infection. Moreover, the time of sample collection is in the winter season, the low temperature and humidity, combined with the dry air, these factors pose inhibited to the growth of bacteria. Among Salmonella positive samples, chicken samples were the most prevalence (36.7%), followed by pork samples (30%), and beef samples were the lowest rate (16.7%). This result is in line with the research results of Do Ngoc Thuy et. al., Zhao et. al., Miranda et. al. However, this result is different from other studies such as Phan et. al. (pork, beef, chicken) and Nguyen et. al. (pork, chicken, beef). The Salmonella positive rate in chicken is often higher than pork, which Sample sources Sample ID 1 Chicken (11 samples) TG-1, 2, 3, 4, 5, 6, 14, 25, 28, 29, 30 2 Pork (9 samples ) TL-1, 2, 3, 4, 5, 15, 26, 29, 30 3 Beef (5 samples ) TB-1, 2, 3, 15, 29 10 can be due to chickens are plucked, slaughter and bleeding directly on the cement floor, there is no separation between these areas, so bacteria from feces, environment easily contaminate into meat. From the above research results it can be seen that the prevalence of Salmonella in retail meats is very different, depending on the geographic area of sample collection. 3.1.2. Salmonella serotype results Identify serotype from the Salmonella spp. we obtained, we obtained 9 different serovars. The most common was S. Typhimurium (11/25 strains). Following by S. Derby (4/25 strains), S. Warragul, S. Indiana, S. Rissen (2/25 strains), S. London, S. Meleagridis, S. Give, S. Assine (1/25 strains). We found that in different studies the common serovar was also different. For example, Moe et. al. (S. Albany), Patchanee et. al. (S. Rissen). From the above research results, it is shown that the common serovars in the studies are different, depending on the geographical area and time of sample collection. 3.2. Antibiotic resistance characteristics of Salmonella isolated 3.2.1. The level of antibiotic resistance of Salmonella isolated Base on the antibiotic susceptibility results, we obtained 52% (13/25) strains resistant to at least one antibiotic, of which the multiple resistance rate was 36%. The rate of resistance to Streptomycin (STR) and tetracycline (TE) was highest (44%). This is understandable because these two antibiotics are widely used in Vietnam, both in treatment and animal husbandry. All Salmonella are susceptible to ceftazidime (CAZ), thus this is a good antibiotic that can be used to treat Salmonella infection. It is needed to monitor the use of CAZ. Antibiotic-resistant Salmonella is a source of transmission of resistance genes to other organisms, more dangerous is to humans 11 through the consumption of food. The rate of antibiotic resistance in this study (52%) is lower than the results of some studies in Vietnam (62.2%) and Japan (89.9%). The rate of multi-resistance (36%) is lower than some other research results such as Nguyen et. al. (41.1%) and Katoh et. al. (90.2%). This difference may be due to the overuse of antibiotics in animal husbandry and treatment between nations, which increases selective pressures, resulting in emergence different rates of antimicrobial resistance Salmonella. The number of samples in this study is small (25 samples). Therefore, further research on resistance rates and multiple antibiotic resistance should be conducted with a larger number of samples. 3.2.2. The number of each Salmonella resistant to antibiotics according to the isolation source Determining the rate of antibiotic resistance of Salmonella according to isolated sources helps to evaluate in detail their antibiotic resistance characteristics by food source. The results showed that Salmonella isolates from pork were most resistant to antibiotics, with 66.7% (6/9 strains), of which 44.4% (4/9 strains) were multi-antibiotic resistant. Following by the isolated from chicken with the rate of 36.4% (4/11 strains) of which 27.3% (3/11 strains) are resistant. There is only one S. Typhimurium from beef that is antibiotic resistant and is multi-resistant. S. Typhimurium accounts for a large proportion of all three isolates (11 strains), but only 3 strains of antibiotic resistance are also multi-resistant. 3.2.3. Number of Salmonella resistant to each antibiotic according to the isolation source The determination of Salmonella antibiotic resistance according to isolated sources helps to evaluate in detail their antibiotic 12 resistance characteristics by food source. Accordingly, all Salmonella isolated from chicken were antibiotic resistant (except CAZ). No strains were isolated from pork and beef resistant to CIP. Salmonella isolated from three meat sources that are resistant to AM, STR, C, TE and SXT, among which isolates from pork have the highest rate of antibiotic resistance. All Salmonella isolates from beef are sensitive to CAZ, GN and CIP. Salmonella was isolated from pork resistant to AM, STR, and TE with the largest percentage. Salmonella from chicken is the most C resistant, especially the CIP resistant strains were found only in chicken. Beef is the least contaminated Salmonella and these strains are most susceptible to antibiotics. 3.2.4. Antibiotic resistance phenotypic Base upon antibiotic susceptibility results, we have identified the antibiotic resistance pattern of Salmonella studied. Two common phenotypes of antibiotic resistance are TE, STR, AM (2/9), and C, TE, SXT, STR, AM (3/9). TE, STR, AM phenotypes are only found in Salmonella isolated from pork. The two common antibiotic resistant phenotypes in this study differ from the antibiotic resistance phenotypes published in the Miranda et. al., Kim et. al. studies. From the above results it can be said that the common patterns of antibiotic resistance are different between the studies. Identifying the antibiotic resistance phenotype in the research samples is important. This phenotypic result combined with genomic analysis results will show antibiotic resistance genes, mutations related to antibiotic resistance phenotypes. From the phenotypic results, we have obtained 9 multi-resistant Salmonella, of which S. Typhimurium is the most common. Pork is the most isolated source of 13 multi-resistant Salmonella (5 serovars), followed by chicken (3 serovars), beef (1 serovar). Antibiotic resistance rates, multidrug resistance rates, and antibiotic resistance patterns are different among published studies. This difference, according to some researchers, may be due to the overuse of antibiotics in treatment and animal husbandry, increasing the selection pressure on bacteria leading to the emergence of different strains of multi-antibiotic resistant Salmonella by region. geography. 3.3. Results of analyzing gene categories in some multidrug resistance Salmonella Transcriptome sequencing is rarely used to identify antibiotic resistance genes. Instead, researchers often use the entire whole genome sequencing. Transcriptome sequencing allows us to study the function of genes better than DNA sequencing. In this study, we want to focus on the functional genome of bacteria. This is important because the unnecessary genes will not be expressed and tend to be lost or degraded. Moreover, the whole genome sequencing technique could not distinguish the genes that were inactivated in the genome and other related functions of that gene. In addition, in our opinion, the expression gene will be related to the phenotype, particular resistant or sensitive phenotypes, higher than the non-expression gene. We have sequenced transcriptome of multidrug resistant Salmonella using method which was published by Marcelino et. al. (the author also sequenced the transcriptome of bacteria in bird gut in Australia to identify of antibiotic resistance genes). It is very important that antibiotic resistance genes are also expressed in antibiotic sensitive bacteria under normal culture conditions. Therefore, in this study we did not use antibiotic sensitive strains for comparison. 14 Instead, we use the same phenotype that is sensitive and resistant to each antibiotic in the strains to compare. From that, we predict which genes, or mutations, might be related to antibiotic resistance in this bacterium. In order to analyze the gene groups in multidrug-resistant Salmonella, several resistant serovars should be selected to sequence transcriptome. However, in the framework of this thesis, we selected only 6 strains according to the criteria of high infection rate, resistance to as many antibiotics as possible and by isolated sources, including S. Derby, S. Give, S. Indiana, S. Typhimurium S384, S. Typhimurium S360, and S. Typhimurium S181. After extracted RNA from 6 research samples, we conducted quality control by concentration measurement at OD260/280, and electrophoresis. Results showed that these RNA samples were of good quality. This sample was then synthesized cDNA and tested for integrity. The results showed that the RIN (RNA Integrity Number) was above 8.0, qualified for sequencing. The results are as follows: 3.3.1. Number of read From the raw sequence results, we conducted trimming sequence by Trimmomatic software: this yield total of 160,043,486 read, total number of read at Q20 score is 146,080,642. The number of read is highest in Sal 4 and at least in Sal 6. The read sequences at Q20 will be used for de novo assembly and used for further analysis. 3.3.2. De novo assembly Results of de novo assembly showed that the transcriptome size ranged from 4.69 Mb (Sal 4) to 5.1 Mb (Sal 11), GC values around 52%, N50 values are high and L50 values are low. This result is similar to the sequencing result of S. Derby 07CR553 published by 15 Kérouanton et. al. From that, it can be concluded that the sequence of 6 research samples is of good quality, eligible for subsequent analysis. 3.3.3. Genes annotation To avoid the missing annotation genes, gene analysis tools have been used as many as possible. The number of genes detected in these tools is different. The findings of these tools' genes differ because of their different methods of analysis, and there is currently no standard method for annotating genes accepted among researchers around the world. The number of coding sequences found by different databases is shown in Table 3.3. Table 3.3. Coding sequences in 6 research samples Gene analysis tool Number of coding sequences Sal 4 Sal 6 Sal 7 Sal 8 Sal 11 Sal 12 RAST 4.807 4.917 5.154 5.097 5.357 5.000 PATRIC 4.807 4.917 5.154 5.049 5.357 5.000 BASys 4.973 5.100 5.336 5.253 5.566 5.209 Genious R11 4.400 4.513 5.022 5.207 5.304 5.309 In 2017, Baek et. al. announced that many genes that encode proteins less than 100 amino acids undetectable when annotating the bacterial genome. Therefore, further studies of the above genes are needed in the research samples. 3.3.4. Gene categories analysis in multidrug resistant Salmonella In addition to housekeeping genes...We have identified important gene categories expressed in multi-resistant Salmonella: 3.3.4.1. Antibiotic resistance genes Results of antibiotic resistance genes and antibiotic resistance phenotype are presented in Table 3.4. Accordingly, a total of 107 16 antibiotic resistance genes (list of genes not shown in this summary). Including 22 β-lactam resistance genes, 46 aminoglycoside resistance genes, 8 quinolone resistance genes, 7 phenicol resistance genes, 6 cycline resistance genes, 3 sulfonamide resistance genes, 3 trimethoprim resistance genes. Furthermore, we have found 12 antibiotic resistance genes which resistant to macrolides, rifamycin, fosfomycin, lincosamide, polymyxin, and peptides. The number and diversity of antibiotic resistance genes in this study are similar to the results of the study by Saskia et. al. (2018). A total of 42 phenotypes were identified from antibiotic susceptibility results. There are 29 antibiotic resistant phenotypes have expression of antibiotic resistance genes. There are 12 antibiotic- sensitive phenotypes have expression of antibiotic resistance gene. The only sensitive phenotype is that there is no expression of the antibiotic resistance gene, Sal 6 is susceptible to SXT and there is no expression of SXT resistance gene. Thus, the genotype and phenotype accordant is (29 + 1) / 42 = 71.4%. And the genotype and antibiotic resistance phenotype not accordant is 12/42 = 28.6%. The correlation between genotype and phenotype is similar to the study results of Owen et al., 2017 (72.7%). Our research results are lower than those of some other authors like McDermott et. al, 2016 (99%), Zankari et. al., 2013 (99.74%). The genotypes and phenotypes not accordant in this study can be explained by the inadequate expression of antibiotic resistance genes, due to the multiple antibiotic resistance mechanisms involved in resistance to one antibiotic, and the other mechanisms of antibiotic resistance have not been found. 17 Table 3.4. Summarize results of antibiotic resistance genes and antibiotic resistance phenotype in the research samples KS Mẫu nghiên cứu (Kiểu hình kháng kháng sinh/gen kháng kháng sinh) Sal 4 Sal 6 Sal 7 Sal 8 Sal 11 Sal 12 AM (R) blaOXA-1 blaTEM family, PBPE** (R) blaTEM family (R) blaTEM family PBPE** (R) blaTEM family (R) blaTEM family PBPE** (R) blaTEM family PBPE** GN (R) aac family* aph family* ant family*, kdpE (S) aac(6')-Iy kdpE (S) aac (6')-Iy, aac6-Iy, aadA8, aadA17; kdpE (S) aac (6')-Iaa, aac3-IIa, aadA17, aadA8b aph3-IIa, kdpE (R) aac family* aph family* kdpE (S) aac(6')-Iaa, aac6- Iaa, aph(6)-Id, aph(3'')-Ib strA, strB, kdpE STR (R) aac family* aph family* ant family*, kdpE (R) aac(6')-Iy kdpE (R) aac (6')-Iy, aac6-Iy aadA8, aadA17 kdpE (R) aac (6')-Iaa, aac3-IIa aadA17, aadA8b aph3-IIa, kdpE (R) aac family* aph family* kdpE (R) aac(6')-Iaa, aac6-Iaa aph(6)-Id, aph(3'')-Ib strA, strB, kdpE CIP (R) aac(6')Ib-cr, gyrA, gyrB, parC (S) gyrA, gyrB, parC (S) qnr-S1, qnr-S3, gyrA, gyrB, parC, parE (S) qnrS1, qnr-S3, qnr-S5, gyrB, parC (S) gyrA, gyrB, parC (S) gyrA, gyrB, parC C (R) floR, cmlA1, catB3, catB4, catB8 (S) floR, cmlA1 (R) floR, cmlA1, cmlA5, cat2 (R) floR, cmlA1, cmlA5, cat2 (R) floR, cmlA1 (S) cmlA1 TE (R) tet(A), tet(R) (R) tet(A), et(M), tet(S) (R) tet(A), tet(M), tet(R), tet(S) (R) tet(A), tet(B), tet(C), tet(M), tet(R), tet(S) (R) tet(A), tet(B), tet(C), tet(R) (R) tet(A) SXT (R) sul1, sul2, dfrA12 (S) (R) sul2, sul3, dfrA12 (R) sul2, sul3, dfrA12 (R) sul2, dfrA14, dfrA5 (S) sul2 Abbreviation: *Aac (Acetylation) family; Aph (Phosphorylation) family; Ant (Adenylylation) family; ** Penicillin Binding Protein E. coli. KS (kháng sinh), AM (ampicillin), GN (gentamycin), STR (streptomycin), CIP (ciprofloxacin), C (chloramphenicol), TE (tetracyclin), SXT (sulfamethoxazol/trimetoprim). 18 3.3.4.2. Quinolone resistance gene mutations Quinolone is commonly used in the treatment of Salmonella infections in humans. For food-borne Salmonella, quinolone resistance is the most concerned and has been mentioned in the list of the most important antibiotics in the field of medicine in 2016 by WHO. One of the quinolone resistance mechanisms in Salmonella is caused by mutations of the gyrA, gyrB and parC genes. Thus, we found mutations of these genes, the results are presented in Table 3.5. Table 3.5. Results of the quinolone resistance gene mutations Samples SR Gene mutations gyrA parC parE Sal 4 R S83F;D87G T57S; S80R; T255S; A628S Sal 6 S T57S; T255S; N395S Sal 7 S S83Y T57S; T255S; A352V S592N Sal 8 S T255S; N395S; S469A; A620T Sal 11 S T255S; N395S; S469A; A620T Sal 12 S T255S; N395S; S469A; A620T Abbreviation: A (Alanine), N (Asparagine), R (Arginine), S (Serine), T (Threonine), V (Valine), SR (susceptibility results), R (resistant), S (sensitive). The results of Table 3.5 show that the list of mutations was identified: S83F, S83Y, D87G, S80R, T57S, T255S, N395S, S469A, A620T, A628S, S592N. To date, there are no reports of mutations A628S, T255S, N395S, S469A, and A620T have been published. However, the mutations T255S, N395S, S469A, A620T were identified in CIP-sensitive strains Sal 6, Sal 7, Sal 8, Sal 11, and Sal 19 12 proving that these mutations have no role in CIP resistance in research samples. The parC mutation (A628S) may have a role in CIP resistance because it appears only in Sal 4, the only CIP resistant strain. This result is not only new but also has high scientific significance, paving the way for further research on antibiotic resistance in Salmonella. 3.3.4.3. The gene categories involved in the efflux pumps We identified the expression of 41 efflux pumps-related genes, including 37 genes in S. Typhimurium S181, 25 genes in S. Typhimurium S384, 27 genes in S. Typhimurium S360, 27 genes in S. Give, 23 genes in S. Derby, 26 genes in S. Indiana. Efflux pumps were detected in research samples belonging to 3 families: MFS (MefB, EmrAB, tetA, tetB, MdtD), SMR (QacE), and RND (AcrAB-TolC, AcrAD-TolC, AcrEF-TolC, MdtABC-TolC, MexPQ-OpmE). Understanding of efflux pumps is essential for the development of interventions to limit antibiotic resistance in Salmonella. Some strains of Salmonella are susceptible to antibiotics but still have expression of efflux pumps, like AcrAB-TolC, AcrEF-TolC, and MdfA in C-sensitive strains. AcrD in GN-sensitive strains. MexPQ-OpmE in CIP-sensitive and C-sensitive strains. Therefore these channels do not play a role in resistant to these antibiotics. Some antibiotic resistant strains have the expression of efflux pumps, like AcrD in AM, STR-resistant strains, MexPQ-OpmE in TE- resistance strains, EmrAB in TE, STR and AM-resistant strains. Therefore, we predicted that these channels may be related to resistance to the respective antibiotics. Further studies on the drug dispensing canal system in Salmonella need to be conducted in the future. 20 New finding the role of the above efflux pumps on antibiotic resistance is the novelty of the thesis. 3.3.4.4. Analysis results of some other gene categories In addition to the antibiotic resistance gene category and the genes related to efflux pumps, we identified many other gene groups related to antibiotic resistance and bacterial toxins. The number of genes in each of these functional categories is shown in Table 3.6. Table 3.6. Expression of related gene groups. Gene category Number of genes Sal 4 Sal 6 Sal 7 Sal 8 Sal 11 Sal 12 Toxin genes 118 124 125 130 131 133 flagellar 38 38 38 38 38 38 Cell wall and LPS 52 39 48 52 51 51 Response to Selenium 2 2 2 2 2 2 Response to Tellurite 3 3 3 3 3 3 Response to Formaldehyde 2 2 2 2 2 2 Mobile elements, phage and prophage 12 24 22 27 26 26 ABC transporter 21 21 21 21 21 21 Abbreviation: Sal 4 (S. Indiana), Sal 6 (S. De