Antimicrobial Resistance Patterns and Serological Diversity of Shigella Species from Patient Isolates at University Teaching Hospital in Lusaka, Zambia

Abstract

Background: Shigella species are the leading cause of human shigellosis. In Zambia, more than 30% of children experiencing diarrhea are infected with Shigella species. The increasing resistance of Shigella species to the recommended therapy is of great concern. Therefore, this study investigated the antibiotic resistance profiles and phenotypic and genotypic characteristics of Shigella isolates at the largest referral hospital in Zambia. Methodology: Of the forty-eight archived presumptive Shigella isolates, thirty-two were serologically confirmed and subjected to antimicrobial susceptibility testing using the Kirby Bauer disk diffusion method. Thereafter, polymerase chain reaction was performed to detect the bla genes. Results: Most isolates were Shigella flexneri (16/32, 50%) and Shigella sonnei (14/32, 44%), while Shigella boydii and Shigella dysenteriae were rare. High resistance rates were noted for sulfamethoxazole/trimethoprim (78%) and tetracycline (75%), while 15.6% of the isolates showed resistance to ciprofloxacin and/or azithromycin. The bla_TEM gene encoding beta-lactamase was detected in 7/32 (22%) of isolates. Conclusions: In this study, a significant number of multidrug-resistant isolates were identified. Additionally, Shigella species resistant to the World Health Organization-recommended drugs call for strengthened laboratory diagnosis and close monitoring of these pathogens to guide the clinical management of shigellosis.

1. Introduction

Shigella species are small Gram-negative rod-shaped bacteria belonging to the Enterobacteriaceae family. The genus Shigella consists of four species—Shigella dysenteriae, Shigella flexneri, Shigella boydii, and Shigella sonnei—all of which are pathogenic to humans and cause shigellosis, a severe form of bacillary dysentery characterized by diarrhea, fever, and abdominal pain [1,2]. The four Shigella species are classified into serotypes and subserotypes based on the O-antigen of their cell envelope lipopolysaccharides [3]. S. dysenteriae comprises 15 serotypes, with Shigella dysenteriae type 1 being the most pathogenic, as well as S. flexneri (19 serotypes and subserotypes), S. boydii (19 serotypes), and S. sonnei (one serotype) [3,4,5,6]. Globally, Shigella species cause approximately 80 to 165 million cases of infection and 600,000 deaths annually [7]. In Zambia, more than 30% of children experiencing diarrhea are infected with Shigella species [8,9]. Generally, shigellosis is self-limiting, but antibiotic use shortens the duration of symptoms and the shedding of pathogens, consequently lowering the likelihood of transmission [2]. In essence, shigellosis prevails in developing nations and spreads through the consumption of contaminated food, inadequate sanitation, or direct person-to-person contact [10]. Shigella infections affect everyone, but high-risk groups include children, the elderly, and immune-compromised persons. Shigella species have a notable resistance to stomach acid, and an infectious dose can be as low as 10 bacteria [11]. Once ingested, Shigella invades the epithelial cells and secretes virulence factors that induce inflammatory colitis; this decreases colonic absorption or net colonic secretion, leading to the accumulation of fluids in the colon, causing watery diarrhea [12,13]. Consequently, damage to the epithelial cells lead to vascular damage to the intestines and hemorrhage characterized by blood and fecal leukocytes in stool [10]. Treatment of Shigella infections with antibiotics plays a vital role in reducing the prevalence and death rates of the disease [14].

Several risk factors can increase antimicrobial resistance (AMR) in developing countries, including the prescription of substandard imported antimicrobial drugs, antimicrobial misuse and overuse, and self-prescription [15,16,17,18]. Drug-resistant Shigella species, on the other hand, have become a global concern. Since 2013, Shigella infections have been progressively resistant [7]. Unfortunately, Shigella is challenging to control because it easily spreads between people, including through sexual activity [19]. According to the World Health Organization (WHO) treatment guidelines for shigellosis, ciprofloxacin is recommended as a first-line option, and either pivmecillinam, ceftriaxone, or azithromycin (in adults only) as a second-line antibiotic with zinc supplementation [20]. Nevertheless, these recommended antibiotics, with a wide range of effectiveness in diverse clinical scenarios, face a challenge due to the escalating drug resistance observed in Shigella species, particularly in developing nations [21].

In Zambia, access to clean water and sanitation infrastructure is limited; the majority of the population has inadequate sanitary facilities, with only 64% using basic drinking water services and 33% using basic sanitation services, while 10% practice open defecation [22]. Additionally, only 24% of the population can access basic hygiene services such as hand washing facilities with soap and water [22]. Therefore, contaminated water sources, inadequate sewage disposal, and limited access to healthcare facilities increase the frequency and spread of Shigella. Public health efforts in Zambia involve initiatives to improve water and sanitation infrastructure, promote hygiene practices, and enhance healthcare services to diagnose and treat Shigella infections promptly [23,24]. Surveillance and monitoring systems have been put in place to track the incidence of Shigella and implement appropriate interventions [25]. While other studies have reported AMR in Shigella species isolated from children in Zambia [9,26,27], reports on Shigella species in adults are rare, and there is inadequate information on the circulating serotypes and their resistance patterns in the country. This study, therefore, investigates the antibiotic resistance profiles and phenotypic and genotypic characteristics of Shigella isolates at the University Teaching Hospital, the largest referral hospital in Zambia.

2. Materials and Methods

2.1. Isolation of Shigella Species, Biochemical Tests, and Detection of Strain Serotypes

Suspected Shigella isolates previously collected in 2021 from 48 patients in 10 health facilities from three districts (Table S1) were identified using standard microbiological methods [28]. Briefly, the isolates were plated on xylose lysine deoxycholate (XLD) agar (Himedia, Mumbai, India) and MacConkey agar (MAC) (Himedia, Mumbai, India) and aerobically incubated for 24 h at 37 °C; the obtained colonies (pink and colorless colonies from XLD agar plates and MAC plates, respectively) were identified using the standard biochemical tests, including triple sugar iron (TSI), lysine iron agar (LIA) urease, and indole and motility. Serological identification was performed by slide agglutination as described by the manufacturer (Mast Group Ltd., Bootle, UK). The information on the archived isolates such as age, sex, and residence was collected from the laboratory registers.

2.2. Antimicrobial Susceptibility Testing (AST): The Kirby–Bauer Disk Diffusion Method

The bacterial isolates identified as Shigella species underwent antimicrobial susceptibility testing using the Kirby–Bauer disk diffusion method, as detailed by [29]. The following antimicrobial agents were included: ampicillin (AMP 10 µg), ciprofloxacin (CIP 10 µg), tetracycline (TET 30 µg), amoxicillin-clavulanic acid (AMC 20 µg), sulfamethoxazole/trimethoprim (SXT 25 µg), cefepime (FEP 30 µg), azithromycin (AZM 15 µg), ceftriaxone (CRO 30 µg), and imipenem (IPM 10 µg) (Mast Group Ltd., Bootle, UK). A bacterial suspension equivalent to 0.5 McFarland standard [30], was prepared and inoculated on Mueller–Hinton agar plates. According to the Kirby–Bauer disk diffusion method described by [29], inoculation involves dipping a sterile swab in a 0.5 McFarland-standardized bacterial suspension and pressing against the tube to remove excess inoculum, but it does not accurately measure the inoculum on the swab, thus we could not determine the amount plated. The plates were allowed to dry for 10 min, and then antimicrobial disks were placed on the medium. The culture plates were incubated at 37 °C for 24 h. The results were interpreted using guidelines stipulated by the Clinical and Laboratory Standards Institute (CLSI) [31]. For quality control, a whole-genome sequenced laboratory strain of Escherichia coli (ATCC 25922), recommended by CLSI for antimicrobial susceptibility testing, was utilized.

2.3. DNA Extraction, Polymerase Chain Reaction, and Gel Electrophoresis

Genomic DNA was extracted from overnight cultures prepared in Luria–Bertani (LB) agar (Himedia, Mumbai, India) using a HiPurA Stool DNA Purification Kit (Himedia, Mumbai, India). Conventional Polymerase Chain Reaction (PCR) was performed using Takara PCR Thermal Cycler Dice Touch (Takara, Japan). Primers (TEM-F; 5′-CCCCGAAGAACTTTTC-3′ and TEM-R; 5′-AGAAGTGGCCTGCAACTTT-3′; CTX-MA1; 5′-CSATGTGCAGYACCAGTAA-3′, CTX-MA2; 5′-CCGCRATATGRTTGGTGGTG-3′) [32,33,34] were used to screen for bla genes using KOD One Master Mix (TOYOBO, Osaka, Japan). The PCR conditions were as follows: an initial denaturation of 95 °C for 2 min, 25 cycles, 98 °C for 10 s, 59.5 °C for 5 s, 68 °C for 1 s, and a final hold at 4 °C to infinity. The amplicons were visualized under UV light after 30 min of electrophoresis at 100 volts using 1.5% agarose gel. For the PCR positive control, a previously characterized strain of Escherichia coli using whole genome sequencing with Strain-ID, Zam_UTH_03 was employed [35].

2.4. Analysis of AMR Profiles

Microsoft Excel version 21 was used to record the obtained data from the Kirby–Bauer disk diffusion method. The data were then exported into R software version 4.4.0, where the resistant profiles for all antibiotics, frequency tables, and graphs were generated. The R package dplyr was used for data manipulation before visualizing with ComplexHeatmap.

3. Results

3.1. Demographic Features and Shigella Isolate Serogroup Distribution

Of the 48 presumptive Shigella isolates, 32 (66.7%) tested positive on species-specific serology. The 32 seropositive isolates exhibited a diverse age distribution; 16 (16/32, 50%) from children (1–12 years), seven (7/32, 22%) from adults (18–59 years old), and two (2/32, 6%) from older adults (60 years and above), while seven (7/32, 22%) had missing age data (Figure 1). The serogroup-based species classification of Shigella isolates is depicted in

Table 1. Shigella isolate serogroup distribution (n = 32).

Name of Species	Most Prevalent Age Category	Number (n)	Percentage (%)
S. flexneri	Children (6/16; 37.5%) *	16	50
S. sonnei	Children (10/14; 71.4%)	14	44
S. dysenteriae	Unknown (1; 100%)	1	3
S. boydii	Older adult (1; 100%)	1	3
Total		32	100

* In brackets are the proportions and percentages for the most prevalent age category.

. Of these, the most predominant serogroup was Group B, S. flexneri (16/32, 50%), followed by Group D, S. sonnei 14 (14/32, 44%), and both Group A, S. dysenteriae, and Group C, S. boydii (1/32, 3%).

3.2. Antibiotic Resistance Rates to Commonly Used Antibiotics and WHO Priority Shigella Isolates

Generally, the resistance rates varied across different antibiotic classes, with the highest being sulfamethoxazole/trimethoprim (SXT) (25/32, 78%) and tetracycline (TET) (24/32, 75%). Much lower resistance rates were observed against ampicillin (12/32, 38%), ciprofloxacin (CIP) (5/32, 16%), amoxicillin-clavulanic acid (AMC) (3/32, 9%), and azithromycin (AZM) (2/32, 6%) (Figure 2). Notably, resistance to third-generation cephalosporins (3GCs) was low, with only one (1/32, 3%) isolate resistant to ceftriaxone. The ceftriaxone-resistant isolate also exhibited intermediate resistance to the drug of last resort, imipenem. Among the 16 S. flexneri isolates, 13 (13/16, 81%) were resistant to SXT, 11 (11/16, 69%) to ampicillin (AMP), and 11 (11/16, 69%) to TET. Resistance to first- and second-line treatment (CIP and AZM, respectively) was detected only in S. flexneri isolates. Contrarily, resistance among S. sonnei species was observed only against SXT (11/14 78.5%) and TET (13/14, 93%), with occasional intermediate resistance to AMC, cefepime (FEP), AMP, AZM, and CIP. Two isolates (MN_12 and MN_13) from the UTH Adult Hospital shared the same AMR pattern (Figure 3).

3.3. Proportion of Multidrug Resistance Shigella Species

Multidrug resistance (MDR) was defined as non-susceptibility to at least one agent in three or more antimicrobial classes [36]. The Shigella MDR phenotypes detected in this study are shown in

Table 2. MDR phenotype patterns and facility of referral for S. flexneri isolates.

MDR Phenotype	Strain ID	Facility of Origin	No. of Isolates	Percentage (%)
SXT/TET/AMP/CIP	MN_12	UTH Adult	2	25
	MN_13	UTH Adult
SXT/AMP/TET	MN_14	Matero Hospital	2	25
	MN_28	UTH Paeds
AMC/TET/SXT	MN_22	UTH Paeds	1	12.5
CIP/TET/SXT	MN-10	Chilanga Hospital	1	12.5
CIP/AZM/AMC/CRO/TET	MN_11	UTH Adult	1	12.5
AMC/AZM/CIP/TET	MN_30	State House Clinic	1	12.5

. There were six combinations of multidrug-resistant phenotypes involving eight (8/32, 25%) isolates, all being S. flexneri. The most predominant MDR phenotypes were SXT/TET/AMP/CIP and SXT/AMP/TET (2/8, 25%, each). Phenotypes AMC/TET/SXT, AMC/AZM/CIP/TET, CIP/AZM/AMC/CRO/TET, and CIP/TET/SXT all had one (1/8, 13%, each).

3.4. β-Lactamase-Encoding Genes Were Detected Among Shigella Species

Genomic DNA from the 32 clinical isolates of Shigella was screened for bla_CTX-M and bla_TEM genes by PCR using previously described primers [32,33,34]. The expected bands (i.e., 480 bp) for the bla_TEM gene were obtained in seven (7/32, 22%) isolates (Figure S1). The bla_TEM gene was present in seven out of twelve (7/12, 58.3%) AMP-resistant isolates, and only in one out of three (1/3, 33.3%) AMC-resistant isolates (Figure 3). However, none of the strains tested positive for the bla_CTX-M gene.

4. Discussion

This study revealed the sero-diversity and resistance profiles of clinical Shigella isolates from the highest referral hospital in Zambia. The clinical isolates exhibited a diverse age distribution, the majority (16/32, 50%) being children (1–12 years), followed by adults (7/32, 22%) (18–59). According to a recent CDC report, Shigella is most predominant in children globally [2]. This study, along with others conducted in different locations [37,38] also identified S. sonnei as the most predominant species among children. In this study, the most predominant serogroup was S. flexneri (16/32, 50%), and the least predominant were S. dysenteriae and S. boydii, at 3% (1/32) each. This is similar to what is reported elsewhere, in Iran [39], China [3,40], and India [41]. The second most predominant species in this study was S. sonnei, which has been reported to be the most prevalent species in the developed world [42]. The lower proportion of S. sonnei observed in this study and in many developing countries could be related to cross-immunity associated with frequent exposure to Plesiomonas shigelliodes [43]. Nevertheless, there has been a recent shift towards increasing rates of S. sonnei infections being reported in low- or middle-income countries [14], which is attributed to increased international travel [42,44].

Generally, the resistance rates varied across different antibiotic classes, the highest being SXT (25/32, 78%) and TET (24/32, 75%). The observed high resistance level could be related to antimicrobial overuse. For instance, SXT is often used as prophylaxis for Pneumocystis jirovecii pneumonia in HIV-exposed infants [45] and those with very low CD4 counts. Additionally, this drug class has been on the market for a long time [46], giving microorganisms time to develop resistance [47]. In addition, the low cost of this drug and poor restrictions on its access have led to overuse and misuse by the general public [48].

Although TET is not among the recommended treatments for Shigella infection, our results show a high resistance rate, often coinciding with SXT resistance. This observation could be caused by the co-selection of genes encoding resistance to the two drug classes. For instance, the tet and dfr genes have been reported to co-exist on the same plasmid of chromosomal insertion [35], suggesting that selection pressure created by one drug could be selected for both. Notably, resistance to 3GCs was low, with only one (1/32, 3%) isolate resistant to ceftriaxone. The ceftriaxone-resistant isolate also exhibited intermediate resistance to the drug of last resort, imipenem, suggesting the need for close monitoring.

This study reports MDR Shigella with a prevalence of 25% (8/32). Despite the lower prevalence of MDR reported in this study compared to other studies reported elsewhere [49,50], there is a need for monitoring the observed phenomenon. Curiously, all the MDR phenotypes were observed only in S. flexneri isolates, and majority of the isolates with similar MDR profiles were from the same health facility. Furthermore, these facilities are located in densely populated urban areas with inadequate sanitation and hygiene services [51,52]. This may suggest the clonal spread of multidrug-resistant S. flexneri as a nosocomial or community-acquired infection. This could also indicate the spread of MDR among S. flexneri via the horizontal gene transfer of AMR-gene-carrying plasmids. Therefore, further investigations, such as molecular epidemiology using whole-genome sequencing and detailed demographic history, are needed to confirm this observation.

WHO treatment guidelines for shigellosis recommend CIP as a first-line option, together with either pivmecillinam, CRO, or AZM (in adults only) as a second-line antibiotic with zinc supplementation [16]. Similarly, the Zambia Standard Treatment Guidelines recommend including ciprofloxacin in treating shigellosis in both adults and children [53], while AZM is reserved for second-line treatment in adults. High resistance to CIP and AZM was detected in this study and was only observed in S. flexneri isolates. In addition, intermediate resistance was observed in both S. flexneri and S. sonnei. The observed resistance to CIP and AZM is of great concern, and it must be closely monitored to prevent the situation from escalating and causing challenges in managing shigellosis. Our results suggest that the current treatment regimens may be ineffective, thus there is a need to strengthen Shigella surveillance and revise shigellosis management guidelines based on updated antibiograms.

The bla_TEM gene was present in seven (7/32, 22%) Shigella isolates, suggesting β-lactamase-mediated resistance. β-lactamase inhibitors such as clavulanic acid inhibit β-lactamases by irreversibly binding to the active site, rendering the enzyme ineffective [54]. Expectedly, most (6/7, 85.7%) of the bla_TEM-positive isolates were susceptible or only exhibited intermediate resistance to AMC. The observation that the bla_TEM gene was not found in all the penicillin-resistant isolates highlights that other mechanisms are responsible for β-lactam resistance. Therefore, future studies should explore more robust approaches like whole-genome sequencing.

5. Conclusions

We characterized 32 Shigella isolates collected in 2021 from patients at the UTH. Speciation of the isolates revealed a serological diversity of the four Shigella species—S. flexneri, S. sonnei, S. boydii, and S. dysenteriae. We found that S. flexneri was the most predominant isolated species, with half exhibiting MDR with a prevalence of 25% (8/32). Consequently, the resistance of Shigella species to CIP and AZM is of public health concern, as these are the recommended treatment options for shigellosis. The bla_TEM gene was detected in 58.3% (7/12) AMP-resistant strains, but none possessed the bla_CTX-M gene. Hence, there is a need for systematic AMR surveillance for Shigella to guide clinical management in Zambia. Additionally, we recommend improved sanitation services as key to infection prevention and control, particularly in endemic areas with a high prevalence of shigellosis.

6. Limitation of This Study

The small sample size affected the generalization of the findings and prevented us from identifying significant differences or relationships. Furthermore, the absence of comprehensive epidemiological data hindered the contextualization of the results, making it challenging to account for confounding variables such as socioeconomic factors and underlying clinical conditions. Consequently, future studies should involve larger sample sizes and robust epidemiological data to make meaningful statistical inferences.