Virulence-Related Gene Distribution Among Shigella Isolates in Anhui, China: The Association with Antimicrobial Resistance

Introduction

Shigellosis is one of the most important causes of diarrhea worldwide.¹ The annual number of deaths due to shigellosis has been estimated at 1.1 million, and shigellosis-associated mortality is particularly prevalent among children in developing countries.² Shigellosis can be caused by four Shigella species: S. dysenteriae, S. flexneri, S. boydii, and S. sonnei. Shigella sonnei is the predominant serogroup in developed countries, while S. flexneri is the main serogroup found in developing countries.³ However, the frequency of S. sonnei has recently increased in several areas that have undergone rapid socioeconomic improvements.^4,5

Antimicrobial agents have been increasingly effective in alleviating the dysenteric syndrome associated with shigellosis.¹ However, the effectiveness of antimicrobial agents has become limited due to the emergence of multidrug resistance (MDR).³ Antimicrobial resistance among Shigella species, including resistance to fluoroquinolones and third-generation cephalosporins, is prevalent in China.^6–8 Moreover, antimicrobial resistance differs among regions.⁹ Consequently, it is important to understand the nature of the local antimicrobial resistance when choosing an appropriate antibiotic for the treatment of shigellosis.

The type III secretion system (T3SS) of Shigella has been reported to be associated with the pathogenesis of shigellosis, and is regulated by multiple virulence-related genes (VRGs). The differential distribution of VRGs among Shigella isolates can lead to different clinical manifestations.^10,11 Invasion plasmid antigen (ipa) genes and invasion associated locus (ial) are important mediators of Shigella cell invasion in the intestinal epithelium.¹² Shigella also produces distinct enterotoxins such as Shigella enterotoxin 1 (ShET-1), encoded by the set gene, and Shigella enterotoxin 2 (ShET-2), encoded by the sen gene.¹⁰ VirF and VirB (InvE) are transcriptional activators of plasmid-bourne genes that control the expression of invasion-related genes.¹³ The icsA gene is the major determinant of Shigella actin-based motility, and has been shown to interact with icsB through the prevention of autophagic recognition of icsA.¹⁴ Finally, the genes encoding serine protease autotransporters of Enterobacteriaceae (SPATE) are phylogenetically classified into two main classes: class 1 SPATE genes, comprising the Shigella IgA-like protease homolog (sigA) and secreted autotransporter toxin (sat); and class 2 SPATE genes, comprising Shigella extracellular protein A (sepA) and protease involved in intestinal colonization (pic).¹⁵ Overall, Shigella isolates harboring these VRGs can induce inflammation and extensive mucosal damage during intestinal infections, especially when the isolates encode more than one VRG. Furthermore, the detection of additional VRGs in Shigella spp. together with the concomitant increase in resistance against more antimicrobials suggests that there is a link between virulence and antimicrobial resistance of shigellosis.¹¹ Hence, understanding the distribution of VRGs in Shigella isolates is useful for designing new antimicrobial therapies based on gene targets.

To the best of our knowledge, this is the first report of the VRG profiles of Shigella isolates in Anhui, China. The aim of this study was to investigate the antimicrobial resistance profiles and VRG distribution among Shigella isolates in Anhui, China, and identify the correlation between the VRGs and antimicrobial resistance to help guide the treatment of shigellosis.

Materials and Methods

Bacterial Isolates and Serotyping

A total of 525 non-duplicate Shigella isolates (449 S. flexneri, 68 S. sonnei, 3 S. boydii, and 5 S. dysenteriae) were collected from the stool samples of patients in Anhui Province, China between September 2011 and September 2015. Individual isolates were identified through standard microbiological and biochemical methods. All Shigella isolates were confirmed via the VITEK 2 Compact system (bioMérieux, Marcy l’ Étoile, France) and serotyped using commercially available antisera (Denka Seiken Co., Ltd, Tokyo, Japan). Escherichia coli ATCC 25922 was stored at the Anhui Center for the Surveillance of Bacterial Resistance (Hefei, Anhui, China). All clinical samples were part of the routine hospital laboratory procedure.

The study was conducted in accordance with the guidelines of the Declaration of Helsinki, the principles of Good Clinical Practice, and Chinese regulatory requirements, and was approved by the local Ethics Committees of the First Affiliated Hospital of Anhui Medical University (Hefei, China). All patients gave written informed consent, and parental/legal guardian informed consent was obtained for patients under the age of 18.

Antimicrobial Susceptibility Testing

Bacteria were cultured to the log-phase (approximately 1.5 × 10⁸ CFU/mL). The minimum inhibitory concentrations (MICs) of the following 17 antimicrobial agents were determined by the agar dilution method using Mueller-Hinton agar (Oxoid Ltd, Basingstoke, UK) containing a series of twofold diluted antibiotics: ampicillin (AMP), nalidixic acid (NAL), tetracycline (TCY), chloramphenicol (CHL), trimethoprim/sulfamethoxazole (SXT), cefotaxime (CTX), ceftriaxone (CRO), ceftazidime (CAZ), cefepime (FEP), cefoxitin (FOX), ciprofloxacin (CIP), levofloxacin (LEV), gentamicin (GEN), amikacin (AMK), piperacillin–tazobactam (TZP), fosfomycin (FOS), and imipenem (IMP) (all from Sigma–Aldrich, Shanghai, China). The plates containing the bacterial colonies and the antibiotics were incubated at 37 °C for 18–20 h. The results were interpreted according to CLSI breakpoints published in 2019. Escherichia coli ATCC 25922 cells were used as the quality control strain. The antimicrobial susceptibility of the Shigella isolates was accepted only if the MICs of the quality control strains that were tested in parallel were within the acceptable ranges set by the CLSI guidelines (2019). MDR was defined as resistance to three or more antimicrobial agents.

DNA Extraction and VRG Detection

Bacterial DNA was extracted using the boiling method and plasmid DNA was extracted using the Qiagen Plasmid Purification Kit (QIAGEN, Hilden, Germany). The 16 VRGs (ipaH, ipaA-D, ial, virB, virF, set, sen, icsA, icsB, sigA, sat, pic, and sepA) of the Shigella isolates were identified by polymerase chain reaction (PCR). The primer sequences are shown in Table 1. The cycling conditions included polymerase activation, then an initial denaturation at 95 °C for 5min, followed by 30 cycles of denaturation at 95 °C for 45s, annealing for 45 s (annealing temperatures are shown in Table 1), and extension at 72 °C for 1 min, with a final extension at 72 °C for 10 min. All the PCR products were purified using the QIAquick PCR Purification Kit (QIAGEN) and directly sequenced. Sequences were compared with those of the GenBank nucleotide database to identify the VRGs.

Table 1 Sequences of the Primers Used for PCR Amplification

Statistical Analysis

Statistical analysis focused on the relationships between Shigella serotype and antimicrobial resistance patterns, and the relationships between VRG profiles and antimicrobial resistance. Data were analyzed using SPSS version 16.0 (SPSS Inc., Chicago, IL, USA). Univariate analysis was performed by the chi-square test or Fisher’s exact test when appropriate. P-values were based on two-tailed test results, and P-values <0.05 were considered statistically significant.

Results

Distribution of Shigella Isolates

A total of 525 non-duplicate Shigella isolates were collected between September 2011 and September 2015. Shigella flexneri was the most abundant (n=449, 85.5%), followed by S. sonnei (n=68, 13.0%). Shigella dysenteriae (n=5) and S. boydii (n=3) were relatively uncommon. Figure 1 shows the yearly distribution of the Shigella isolates. Shigella flexneri was the most frequently isolated species each year. The proportion of S. sonnei isolates increased yearly. Two S. dysenteriae were isolated in 2013 and one in each of 2011, 2012, and 2014. One S. boydii was identified in each of 2012, 2014, and 2015.

Figure 1 The yearly distribution of the proportions of the Shigella isolates.

Age and Gender

Shigella isolates were collected from specimens from all age groups that presented bacillary dysentery. The median age was 1 year (range: 4 months to 97 years). Infants and children between 1 and 4 years of age accounted for the highest proportion of Shigella isolates (n=331, 63.1%), followed by 5–19-year-old patients (n=72, 13.7%), patients >60 years of age (n=64, 12.2%), and patients between 20 and 59 years of age (n=58, 11.0%) (Figure 2A). Shigella species showed the greatest male predominance. The proportion of S. flexneri was higher in males than in females (57% vs 43%; P<0.0001). There was no significant difference in the proportion of S. sonnei between male and female patients (47.1% vs 53.9%; P=0.49) (Figure 2B).

Figure 2 Age and gender distribution of Shigella isolates in China from 2011 to 2015. (A) The distribution of the proportions of Shigella isolates by age group. (B) Comparison of the proportion of S. flexneri and S. sonnei by gender group.

Antimicrobial Resistance

A comparison of the resistance rates among the Shigella isolates is shown in Table 2. Shigella flexneri was most commonly resistant to AMP (97.3%), followed by NAL (96.9%), TCY (93.8%), and CHL (86.6%). Shigella sonnei was most commonly resistant to AMP (100%), followed by NAL (98.5%), TCY (94.1%), SXT (92.6%), and GEN (92.6%). Shigella flexneri was most sensitive to TZP (99.3%), followed by IPM (98.2%) and AMK (98%). Shigella sonnei was most sensitive to TZP (100%) and AMK (100%), followed by IPM (98.5%) and FEP (98.5%).

Table 2 MIC50 and MIC90 of 17 Antimicrobial Agents Against S. Flexneri and S. Sonnei

Among the 449 S. flexneri isolates, 417 (92.9%) displayed MDR, and the most common resistance pattern was AMP–NAL–TCY (90.2%). Among the 68 S. sonnei isolates, 66 (94.1%) exhibited MDR, and the most common resistance pattern was AMP–NAL–TCY (92.6%).

Resistance to CHL, CIP, and LVX was more common among S. flexneri isolates than among S. sonnei isolates (86.6% vs 8.8%, P<0.0001; 49.7% vs.19.1%, P<0.0001; 30.5% vs 10.3%, P=0.001, respectively). However, resistance to STX and GEN was more common among S. sonnei isolates than among S. flexneri isolates (76.2% vs 92.6%, P<0.0001; 28.7% vs 92.6%, P<0.0001, respectively).

Prevalence of VRG

Invasion-Associated Genes

The detection of invasion-associated genes among the 449 S. flexneri isolates showed that ipaH had the highest frequency (100%), followed by ipaB (99.8%), ipaC (98.7%), ipaD (98.7%), ipaA (98.4%), and ial (96.0%) (Table 3). The detection of invasion-associated genes among the 68 S. sonnei isolates showed that ipaH had the highest frequency (100%), followed by ipaB (98.5%), ipaC (97.1%), ipaD (95.6%), ipaA (95.6%), and ial (75.0%) (Table 3).

Table 3 Distribution of Virulence-Related Genes Among Shigella Isolates in Anhui, China

Regulatory Genes

Among the 449 S. flexneri isolates, 439 (97.8%) were positive for the virB gene, while 431 (96.0%) were positive for the virF gene (Table 3). Moreover, a total of 413 isolates (92.0%) were positive for both genes. Among the 68 S. sonnei isolates, 66 (97.1%) were positive for the virB gene, while 51 (75.0%) were positive for the virF gene (Table 3). Fifty (73.5%) isolates were positive for both genes.

Among the 449 S. flexneri isolates, 448 (99.8%) were positive for the icsA gene, while 439 (97.8%) were positive for the icsB gene (Table 3). A total of 437 (97.3%) isolates were positive for both genes. Among the 68 S. sonnei isolates, 67 (98.5%) were positive for the icsA gene and 66 (97.1%) for the icsB gene (Table 3). Moreover, a total of 63 (91.3%) isolates were positive for both genes.

SPATEs

Among the 449 S. flexneri isolates tested, 421 (93.8%) contained class I (sigA, sat) and/or class II (sepA, pic) SPATE genes (Table 3). Moreover, a total of 415 (92.4%) had two class I SPATE genes, and 394 (87.8%) had both class II SPATE genes. The most common SPATE gene among S. flexneri isolates was sat (93.8%), followed by sigA (92.4%), pic (88.2%), and sepA (87.8%). Among the 68 S. sonnei isolates, 64 (94.1%) harbored only one class I SPATE gene (sigA) (Table 3). The sat, pic, and sepA genes were not detected among the S. sonnei isolates in this study.

Enterotoxin Genes

The set gene was present in 420 of the 449 (93.5%) S. flexneri isolates, while the sen gene was present in 413 (92.0%) of the S. flexneri isolates and 60 of the 68 (88.2%) S. sonnei isolates. The set gene was not detected among the S. sonnei isolates in this study.

VRGs and Antimicrobial Resistance

Resistance to CHL and CIP was more common among VRG-positive (ial, ipaA, ipaC, ipaD, virB, virF, set, icsB, sat, pic, and sepA) S. flexneri isolates than among VRG-negative S. flexneri isolates (P<0.05) (Figures 3 and 4). Furthermore, resistance to CRO and CAZ was more common among sat- and set-positive S. flexneri isolates than among sat- and set-negative S. flexneri isolates (P<0.05) (Figure 3E, F). However, CEN resistance was more frequent among VRG-negative (ial, virF, set, sat, pic, and sepA) S. flexneri isolates than among VRG-positive S. flexneri isolates (P<0.05) (Figure 3). With fewer VRGs-negative strains among S. flexneri (ipaH, ipaB, and icsA) and among S. sonnei, no correlation tests were done between these VRGs and antimicrobial resistance. The resistance rates were similar between VRG-positive S. flexneri and VRG-positive S. sonnei isolates, except for the resistance to CRO and CAZ, which was more commonly found among VRG-positive S. flexneri than among VRG-positive S. sonnei (P<0.05) (Supplementary Figure 1 and 2).

Figure 3 Comparison of the antimicrobial resistance rates between virulence-related gene (VRG)-positive (ial, virF, sepA, pic, sat, and set) and VRG-negative S. flexneri. (A) Comparison of the antimicrobial resistance rates between ial-positive and ial-negative S. flexneri. (B) Comparison of the antimicrobial resistance rates between virF-positive and virF-negative S. flexneri. (C) Comparison of the antimicrobial resistance rates between sepA-positive and sepA-negative S. flexneri. (D) Comparison of the antimicrobial resistance rates between pic-positive and pic-negative S. flexneri. (E) Comparison of the antimicrobial resistance rates between sat-positive and sat-negative S. flexneri. (F) Comparison of the antimicrobial resistance rates between set-positive and set-negative S. flexneri. *, P<0.05; **, P<0.001.

Figure 4 Comparison of the antimicrobial resistance rates between virulence-related gene (VRG)-positive (ipaA, ipaC, ipaD, virB, icsB, sigA, and sen) and VRG-negative S. flexneri. (A) Comparison of the antimicrobial resistance rates between ipaA-positive and ipaA-negative S. flexneri. (B) Comparison of the antimicrobial resistance rates between ipaC/ipaD-positive and ipaC/ipaD-negative S. flexneri. (C) Comparison of the antimicrobial resistance rates between virB-positive and virB-negative S. flexneri. (D) Comparison of the antimicrobial resistance rates between icsB-positive and icsB-negative S. flexneri. (E) Comparison of the antimicrobial resistance rates between sigA-positive and sigA-negative S. flexneri. (F) Comparison of the antimicrobial resistance rates between for sen-positive and sen-negative S. flexneri. *, P<0.05; **, P<0.001.

Discussion

Shigellosis is an invasive bacterial infection of the human colon that manifests a spectrum of clinical presentations, ranging from short-lasting watery diarrhea to inflammatory bowel disease.³ In this study, we explored the incidence rate of shigellosis during 2011–2015 in Anhui, China, and showed that there was an obvious decline. Shigella flexneri was the most commonly isolated species, similar to that previously reported for several regions in China.^16–18 However, we found that the detection rate of S. sonnei, which is the predominant species in developed countries, showed a clear yearly increase. This was likely due to improvements in socioeconomic levels and environmental conditions.^18,19

In this study, shigellosis was detected in all age groups; however, children under 5 years of age had higher incidence rates, which was consistent with previous reports.^11,17,20 Children under 5 years of age are likely to be more susceptible to shigellosis because of low immune function and the lack of previous exposure.^21,22 In this study, we observed that the incidence of S. flexneri was markedly higher among males than among females, similar to that reported in previous studies.^16,17 One explanation may be that males have a higher risk of exposure to Shigella-contaminated environments, such as men having sex with men.²³ However, there was no significant difference in the incidence of S. sonnei between males and females, which may have been due to the small number of isolated samples.

Effective antimicrobial therapy for shigellosis can significantly relieve symptoms and interrupt bacterial shedding.³ The emergence and dissemination of antimicrobial resistance among Shigella strains in recent years have complicated the therapeutic management of severe shigellosis.^6,7,24 Our data showed that Shigella isolates in China had a high resistance to AMP, NAL, TCY, and SXT, suggesting that these drugs can no longer be recommended as empirical therapy for shigellosis. In this study, the Shigella isolates were observed to be susceptible to TZP, IMP, AMK, and FOX, indicating that these antibiotics may be the preferred treatment options in Anhui, China.

In our study, the most frequently detected antimicrobial resistance pattern was AMP–NAL–TCY (90.2% in S. flexneri, 92.6% in S. sonnei), which indicated that MDR among Shigella isolates was widespread in Anhui, China. According to the WHO guidelines for the control of shigellosis, fluoroquinolones and third-generation cephalosporins are effective treatments for this disease.² However, several studies have reported that the rate of resistance to CIP and CTX in Shigella isolates has gradually increased in recent years,^11,25 which was also observed in our study. Of note, our data showed that the susceptibility to CIP and LVX was higher among S. sonnei isolates than among S. flexneri isolates (80.9% vs 50.3%, P<0.0001; 89.7% vs 69.5%, P=0.001, respectively), suggesting that resistance to fluoroquinolones in Anhui, China, was more pronounced for S. flexneri than for S. flexneri. Moreover, the susceptibility to GEN was higher among S. flexneri isolates than among S. sonnei isolates (71.3% vs 7.4%, P<0.0001), suggesting that GEN may be effective as a second-line treatment for S. flexneri infection in Anhui, China.

All Shigella species possess a large virulence plasmid and a single circular chromosome. The T3SS-associated ipa genes are both necessary for the invasion of epithelial cells and the development of shigellosis.²⁶ The ipaH gene, which encodes invasion plasmid antigen H, is commonly used for the molecular identification of Shigella spp. using PCR assays. In our study, all the isolates harbored ipaH, which was consistent with previous reports.²⁷ Most of the Shigella isolates in this study harbored ipaA, ipaB, ipaC, and ipaD, genes that regulate the secretion and translocation of several effector proteins and play a key role in intracellular actin polymerization and depolymerization.²⁸ The ial gene was reported to be responsible for Shigella epithelial cell penetration and cell-to-cell dissemination.²⁹ In our study, the ial gene was present in 96.0% of the S. flexneri isolates and 75.0% of the S. sonnei isolates. Notably, the prevalence of the ial gene in S. sonnei was significantly lower than that in S. flexneri, suggesting that S. sonnei may be less aggressive.

The expression of Shigella virulence genes is regulated by a heat-stable, nucleoid-structuring protein. When the conditions are favorable for invasion, a transcriptional cascade is initiated through the activation of the virF gene, which subsequently turns on the transcription of the regulatory virB gene.³⁰ In this study, 92.0% of the S. flexneri isolates and 73.5% of the S. sonnei isolates were found to harbor both virF and virB, indicating that there might be other pathways for regulating gene expression. Five of the 449 S. flexneri isolates tested possessed virF, but not virB, suggesting that virF may regulate the expression of virulence genes through additional pathways. Interestingly, virB, but not virF, was detected in 3.0% of the S. flexneri isolates and in 21.1% of the S. sonnei isolates, which may have been due to the loss of the virF gene.

IcsB was previously shown to be required for S. flexneri evasion of autophagy at late stages infection (4–6 h) through inhibition of the binding of the autophagy protein Atg5 to the Shigella surface protein IcsA.³¹ In this study, 97.3% of S. flexneri and 91.3% of S. sonnei harbored both icsA and icsB, indicating that there might be other pathways involved in the prevention of the autophagic recognition of icsA.

The SPATE genes comprise an extended family of secreted autotransporters in Gram-negative bacteria. They exert various functions, including as proteases or mucinases, that can directly or indirectly be toxic to intestine cells.³² In this study, most of the S. flexneri isolates harbored two or more SPATE genes, which is consistent with the results of previous studies.^33,34 We also found that the sat gene was the most commonly identified SPATE gene among the S. flexneri isolates, but was absent among the S. sonnei isolates. This indicated that sat-encoded toxin might be a major contributor to the virulence of S. flexneri isolates. Similar to sat, the class II SPATE genes (pic and sepA) were also absent among the S. sonnei isolates in the present study, which indicated that pic and sepA toxins played minor roles in the pathogenicity of S. sonnei.

Two novel enterotoxins (ShET-1 and ShET-2) have recently been described in Shigella isolates. ShET-1 (encoded by set) and ShET-2 (encoded by sen) can alter electrolyte and water transport in the small intestine, which is closely related to the symptoms of dehydration in shigellosis.^10,35 In our study, no significant difference was found in the distribution of the sen gene between S. flexneri and S. sonnei. Interestingly, the set gene was not detected among the S. sonnei isolates in this study, suggesting that S. sonnei may be less aggressive than S. flexneri.

This is the first study to demonstrate the correlation between VRGs and antimicrobial resistance among S. flexneri in China. With fewer VRG-negative strains among S. flexneri (ipaH, ipaB, and icsA) and among S. sonnei, no correlation tests were done between these VRGs and antimicrobial resistance. In this study, we observed that resistance to CHL and CIP was more common among VRG-positive (except for sigA and sen) S. flexneri than among VRG-negative S. flexneri; this indicated that, except for sigA and sen, all the S. flexneri VRGs were positively correlated with resistance to CHL and CIP. Furthermore, resistance to CRO and CAZ was more commonly detected among sat- and set-positive S. flexneri isolates than among sat- and set-negative S. flexneri, indicating that the sat and set genes were also positively correlated with resistance to third-generation cephalosporins. Interesting, CEN resistance was more frequent among VRG-negative (ial, virF, set, sat, pic, and sepA) S. flexneri than among VRG-positive S. flexneri, implying that the ial, virF, set, sat, pic, and sepA VRGs in S. flexneri were negatively correlated with CEN resistance.

Conclusion

Our study demonstrated that S. flexneri remains the predominant species in Anhui, China and that the most frequent MDR pattern was AMP–NAL–TCY. This study also highlighted that S. flexneri strains harboring several VRGs were associated with antimicrobial resistance. In summary, we identified the correlation between the VRGs and antimicrobial resistance, thereby improving our understanding of Shigella virulence. Further investigation of VRG expression and the genetic mechanisms underlying the antimicrobial resistance of Shigella isolates is required to help guide empirical antimicrobial therapy for the treatment of shigellosis.