Antibiotics Resistance Pattern of Food-Borne Bacteria Isolated from Ice Cream in Bangladesh: AMultidisciplinary Study

Ice cream is one of the most popular food items consumed during the summer season in Bangladesh by all ages but mostly schoolgoing students. Due to the ingredients and handling process of ice cream, it acts as a good shelter for pathogenic and nonpathogenic microorganisms. ,erefore, we aimed to investigate the microbial count, prevalence, isolate and characterize multidrug-resistant bacteria in ice cream samples collected from nearby shops of schools in Tangail district, Bangladesh. Ice-cream consumer and nonconsumer students were selected by providing questionnaires. Total viable count (TVC) and total coliform count (TCC) were determined by pour plate methods, where conventional methods were performed for bacterial identification. ,e Kirby-Bauer disk diffusion method was used to determine the antimicrobial susceptibility of bacterial isolates. Kado and Liu method, with some modifications, was used to extract plasmid from the isolated bacteria and visualized through gel electrophoresis. ,e demographic characteristics showed that the degree of symptoms regarding microorganisms mediated disorders and rate of antibiotics intake in ice cream consumers were significantly higher than the nonconsumers. ,e range of TVC and TCC in the ice cream samples was found 0–9.9×10 CFU/ml and 0–900CFU/ml, respectively. Interestingly, 93.75% of the total ice cream samples also showed fungal positive. A total of 12 different bacterial species were identified, including Proteus spp, E. coli, V.cholera, Pseudomonas spp, Shigella spp, Klebsiella spp, Aeromonas spp, V. Parahemolyticus. Salmonella paratyphi, Citrobacterspp, Plesiomonasspp, and Staphylococcus aureus. ,e antimicrobial susceptibility assay showed the multiple resistance frequency of these isolates to different antimicrobial drugs. All individual isolates were screened for plasmid DNA, and we found that seven strains harbored a single or more than two plasmids sized approximately between 1.9 and 140MDa, indicating a possible connection between resistance phenotype pattern and genotype.


Introduction
Food-borne disease is an emerging infectious disease caused by microbial pathogens, i.e., fungus and bacteria, leading to threats to global public health in the whole world [1]. Centre for Disease Control reported that around 48 million people get sick from a food-borne illness every year, 128,000 are hospitalized, and 3,000 have died from more than 250 related food-borne diseases globally [2]. Frequencies of foodborne diseases are very common in many developing countries like Bangladesh, and it has become a significant public health-related issue. According to a recent estimate, about 30 million people in Bangladesh are affected by foodborne diseases each year due to adulterated food from vendors, malnutrition, inadequate public health, and sanitation facilities, and other conditions leading to morbidity as well as causes mortality, especially in children [3,4].
Ice cream is one of the most popular dairy products consumed by all age groups, particularly school-going children worldwide [5]. But most of these ice creams may be contaminated with microbes during either production or postproduction [6,7]. Additionally, the components of ice cream like water, fat, sugar, nonfat milk solids, emulsifier and stabilizer, flavoring agents, coloring materials, and thickeners may be a potent source for contamination of ice cream [8]. Additionally, bacteria can cause spoilage of milk and milk products because of extracellular or intracellular thermoresistant enzymes (proteases, lipases, and phospholipases) produced by them [9]. It has also been reported that people are at risk from pathogenic bacteria found in milk, which account for over 90% of all dairy-related disorders [10].
Bacteria having plasmid DNA play an essential role in spreading antibiotic-resistant genes to others [23]. Plasmids may contain resistance genes for single or multiple antimicrobial agents, and these mobile genetic elements can be transferred to other nonresistant bacterial cells via horizontal gene transfer mechanisms, including conjugation, transformation, and transduction [24]. In Bangladesh, contamination of ice cream by pathogenic bacteria is becoming an emerging food safety problem.
Bangladesh is still an underdeveloping country, where about 70% of people are educated, so the awareness of food safety and security is vulnerable. As a tropical country, the summer season is very long with hot weather in Bangladesh. Moreover, a huge number of unauthorized ice cream factories are developed here, particularly in the summer season. As a sweetened and cheap food item, ice cream is very popular among school-going students during the summer season. However, contamination of ice cream by pathogenic bacteria and their antibiotic-resistance pattern were not clearly documented in the previous research studies, but little information is reported regarding bacterial load and their biochemical identification [7,25]. erefore, in this study, we demonstrated the overall prevalence of bacterial contamination and their characterization with antibiotic susceptibility patterns and the genotypic characteristics of antibiotic-resistant bacterial strains in ice cream samples. We also showed the demographic characteristics of the ice cream consumer and nonconsumer primary and high school students. is is the first report that will assess the microbiological quality and safety of ice cream and will make sense to the regulatory authorities with effective approaches to adopting proper supervision during ice cream preparation for better children and public health management.

Materials and Methods
2.1. Questionnaire Preparation, Data Collection, and the Ethical Statement. An empirical study was done to compare isolated microorganisms and observational data. A standardized questionnaire was generated based on some literature search about food-borne diseases caused by several food-borne microorganisms [12]. e variables included in the questionnaire were age, sex, degree of symptoms, antibiotics intake, and body mass index (BMI) of school-going students with their age range 8-14 years, where the total number of student were 232 on. Ethical approval was received from the Department of Biochemistry and Molecular Biology, Mawlana Bhashani Science and Technology University, Santosh, Tangail, Bangladesh (MBSTU/BMB/TEST/ 2014/06(2).

Collection, Preparation, and Processing of Sample.
Sixteen (16) ice cream samples from sixteen (16) different companies, including Bangladesh standards and testing institution (BSTI) approved and non-BSTI approved, were collected during six months period (april, 2019 to september, 2019) by using an icebox from popular ice cream shops as well as steed-vendors ice creams at around nearby schools in the Tangail district, Bangladesh. e respective samples were transported to the microbiology laboratory in aseptic conditions for further investigation [5,8].

Mycological (Fungus) Evaluations.
e analysis of ice cream samples was done by direct plating to detect micromycetes undiluted 100 μl ice cream samples were inoculated into Potato Dextrose Agar media (HIMEDIA, India) and incubated at optimum growth conditions [26,27]. Morphology of all fungal isolates was visualized and noted down based on phenotypic characterization, including colony, color, texture, and pigmentation [27].

Enumeration of Total Viable Count (TVC) and Total
Coliform Count (TCC). TVC was determined by traditional methods (serial dilution, spread plate technique, colony count) as described by Shabnum Shaheen et al. [28], and TCC was measured by using MacConkey's agar according to K. M. (2012) [29] and International Organization for Standardization (ISO) rules [30].

Isolation and Biochemical
Identification. Bacterial colonies were isolated by the streak plate method on various selective agar media (EMB, SS, MAC, MSA, PALCAM Listeria identification agar base, and Listeria selective supplement; all from Himedia, India). Growth and morphological characteristics were recorded according to the Biology Libre Texts technique [31] and the biochemical assay results (Sugar fermentation KIA, Motility, indole, urea, and citrate utilization tests) for the identification of bacterial isolates [32].

Extraction, Separation, and Visualization of Bacterial
Plasmid DNA. Isolated bacteria were cultured in Luria broth media and incubated overnight at 37°C in a rotatory incubator. Plasmid DNA was isolated with some modifications of the alkaline lysis method of Kado and Liu through repetitive centrifugation [37,38]. e reagents used for plasmid DNA isolation were Solution I (40 mM Tris-HCl, 2 mM EDTA, pH 7.4 0), Solution II (50 mM Tris-HCl, 3% SDS, pH 12.9, and Solution III (Phenol: chloroform: Isoamyl alcohol (25 : 24 :1) (manual of icddr,b). Plasmid DNA was electrophoresed in 7% agarose gel; prepared in 1 x TBE (Tris-borate EDTA buffer) and mixed with 0.5 µg/mL of ethidium bromide [23,39]. Plasmid samples were mixed with 1x dye were electrophoresed for 2.5-3 h at 80 V in 1 x TBE buffer [23]. e molecular mass of the plasmid DNA bands was measured by comparing the mobility of known molecular mass plasmids of E. coli V517.

Statistical Analysis.
e parametric data were expressed as the mean ± standard error of the mean (s.e.m.), and the differences between consumer and nonconsumer and comparison of antimicrobial resistances were determined by t-test analysis of variance. Bacteria counts data were transformed to logarithm 10 of colony-forming units per milliliter of the sample (log 10 cfu/ml), and the results were presented as mean ± standard error mean (sem) by t-test. P-value of less than 0.05 was considered to indicate a statistically significant difference. Statistical analysis was performed using Microsoft Excel 2013 and statistical package for the social sciences (spss) version 20. Table 1 shows the general characteristics of the ice cream consumer (n � 122) and nonconsumer (n � 110) students in Tangail district, Bangladesh. e participants were selected under the exclusion and inclusion criteria with the consent of their guardians regarding normal hygiene issues during the summer season. Ice cream nonconsumer students were considered a reference group to measure the degree of symptoms related to bacterial infections compared to the ice cream consumer students group. e average BMI level of the nonconsumer and consumer students was 19.76 ± 2.53 and 19.43 ± 2.71, respectively, where the ages were 12.45 ± 1.85 and 12.65 ± 1.87 years. No significant differences were found in average BMI and ages between both groups. e degrees of symptoms of bacterial infections in ice cream consumer students (5.53 ± 1.28) were significantly higher than that in ice cream nonconsumer students (2.27 ± 1.01). On the other hand, the percentage of antibiotics intake in ice cream consumer students was significantly greater than in the nonconsumer group.

Determination of Total Viable and Total Coliform
Count of the Ice Cream Samples. In Bangladesh, BSTIapproved permissible total viable count for ice cream is about less than 4.23 log10 CFU/ml. According to our study, the microbiological load of collected samples was recorded after overnight incubation, and the overall range of TVC was between 0.00 (Polar)-10 (Doikulfi) log10 CFU/ml; about 0-3 fold higher than the safety limits (Figure 1(a)). e TVC of BSTI-approved samples ranged between 0.00 (Polar)-9.53 (Promi Ice Lolly) log10 CFU/ml as against 4.06 (Grameen pipe) to 10 (Doikulfi) log10 CFU/ml in the non-BSTI-approved samples. TVC value of only three ice cream samples was within the BSTI recommended range CFU/ml, where 2 samples (Cocola Mango, Polar) were included in BSTI approved and 1 (Grameen pipe) comprised in non-BSTI approved. e rest of the ice cream samples Lovello Chocobar, Kwality Magnum, Igloo, Kwality Birds, Choco Papa (BSTI) and Ripen Mango, Juice Ripen Litchi Juice, Kironmala, Munni Chocobar, Malai ice cream, Chocobar (Non-BSTI), showed higher microbial load than the normal limit indicated the contamination of ice cream by the pathogenic microorganism. e coliform counts (TCC) of the total of 7 ice cream samples were crossed the BSTI-approved permissible limits (≤1.17 log10 CFU/ml) of the TCC (Figure 1(b)).

Prevalence of Bacterial Isolates from Ice Cream Samples.
A variety of biochemical assays were conducted to make a comprehensive view, and we identified a total of 26 bacterial strains from which 12 different identical species are listed in Journal of Food Quality Table 2. e data in Table 2 showed that Proteus spp. and E. coli were the most frequently (4, 15.38%) identified species. e next magnitude was Staphylococcus aureus (3, 11.54%). e frequency and percentage of Vibrio cholerae, Pseudomonas spp., Shigella spp., Klebsiella spp,. Aeromonas spp., and V. parahemolyticus bacterial strains were found 2 and 7.69%, respectively. Some other human pathogenic bacterial strains, including Salmonella paratyphi and Citrobacter spp. were also identified (1, 3.85%) from several ice cream samples.

Determination of the Presence of Fungus in the Ice Cream
Samples. According to our result, 93.75% of samples flourished with positive results with a high percentage of fungus grown, while the polar alike total viable count of bacteria (Figure 1(a)) did not expose any visible fungal growth. e fungal isolates found in the culture plate differed according to their colony, surface, size, height, shape, pigmentation, and variation of growth rate ( Figure 2).

Antibiotics Susceptibility Pattern of Bacterial Isolates from Ice Cream Samples.
e overall bacterial isolate susceptibility profiles have been shown in Table 3. A total of 17 antibiotics were used for our study, where cefuroxime, rifampicin, erythromycin, ampicillin, and methicillin had the highest resistance pattern, which accounted for 100% of resistance, and streptomycin showed the 2 nd highest resistance pattern (82.6%). A couple of antimicrobial drugs, including kanamycin, nalidixic acid, and gentamycin, have less resistance behavior with a value of 30.43%, and chloramphenicol, norfloxacin, clotrimazole, ciprofloxacin, tetracycline, and moxifloxacin have least resistance values with high sensitivity. Although the five antibiotics were 100% resistant against 23 different isolates, only one antibiotic, namely levofloxacin, was 100% sensitive against all the isolates. Moxifloxacin, clotrimazole, ciprofloxacin, gentamycin, chloramphenicol, tetracycline, and norfloxacin showed around 75% sensitivity profile against the entire tested strains isolated from the ice cream samples. Our experimental data showed that in the 100% sensitivity and resistance scale of the antimicrobial test, the resistant isolates were significantly higher than the sensitive isolates. However, some antibiotics also showed intermediate values regarding sensitivity and resistance scale against isolated bacterial strains.
e specific antimicrobial resistance rates of the bacterial species are shown in Table 4. e isolates were resistant to one or more antibiotics of different structural classes called multidrug resistance (MDR). Plesiomonas spp, a rarely isolated bacterial strain showed the highest resistance rate against 65% of antibiotics used in this study. e other five most common isolates, including Aeromonas spp., V. Parahemolyticus, Citrobacter spp., Pseudomonas spp., and V. cholera spp., exhibited a resistance pattern against more than 50% of antibiotics. However, the four isolated bacterial strains, including Proteus spp., S. Paratyphi, Shigella spp., and Klebsiella spp. displayed a resistance pattern between 35 and 47%.

Determination of Antibiotic-Resistance Patterns in Bac-
terial Isolates by Plasmid Analysis. Antibiogram profile of bacterial strains isolated from ice cream samples showed multiple antibiotic-resistance patterns (Tables 3 and 4). ese phenotypic resistance patterns were validated by the molecular level through the analysis of plasmid DNA extracted from isolated bacterial strains. In this study, a total of 7 (63.63%) from 11 bacterial strains displayed their plasmid DNA in agarose gel and were found MDR-resistant to more than three antibiotics. e plasmid-positive bacterial strains also displayed the seven different plasmid band patterns with varying sizes of molecular weight ranging from 1.9 to 140 MDa ( Figure 3). Bacterial strains, including Paraheamolyticus spp., Citrobacter spp., Proteus spp., E. coli, and Proteus spp. showed only one band size ranging from 1.9 to 140 MDa. e remaining strains had plasmids with a different number of bands. e following isolates Vibrio, Aeromonas spp., Proteus spp., and Alcalegenes spp. showed two bands ranging from above 3.1 MDa to 140 MDa. e plasmid of one isolate, Klebsiella spp., showed its bands ranging from 3.4 MDa to 4.8 MDa (Figure 3). e rest of 4 (36.37%) bacterial species did not show any plasmid DNA, but they were resistant to several antibiotics. Our data suggested that plasmid-positive bacterial strains may possess antibiotic-resistant genes and showed their resistance pattern against different antimicrobial drugs. Moreover, our result has also shown that some isolates did not contain any plasmid but had multidrug resistance phenotypically, which is summarized in Table 5.

Discussion
Microbial contaminations are responsible for serious human health concerns worldwide mediated by antibiotic-resistance patterns. Ice-cream is a milk-based product that serves as a potential shelter for microbial growth. Due to the proper nutrient elements as well as nutrient compositions, almost neutral pH (pH 6-7), and extended storage duration, it may easily be contaminated by pathogenic microorganisms. In this study, the data from the demographic characteristics of ice cream consumers and nonconsumers (    the control group, nonconsumers. ese data suggested that ice cream consumption has a greater role in food-borne diseases among school-going students in Bangladesh. e rate of antibiotics intake without any prescribed consultant and incomplete the recommended full courses of antibiotics is also a vital issue that influences the generation of antibiotic-resistant bacterial strains. It is generally recommended that patients complete their course of antibiotics even if they feel better, preventing reinfection with the same species and reducing the risk of becoming resistant to antibiotics [40]. Our experimental data regarding bacterial load showed high levels of TVC and TCC both in BSTI-approved and non-BSTI-approved ice cream samples, deviating from a permissible limit of BSTI standard (Figures 1(a) and 1(b)). A total of 93.75% ice cream samples were also possess fungal positive. However, such a high microbial load in the ice cream samples recorded in our study indicates the contamination of the samples by pathogenic microorganisms [28,[41][42][43]. e occurrences of indicator organisms in the ice cream constitute a severe threat to the ice cream consumers and are responsible for various food-borne diseases, including salmonellosis, hemorrhagic colitis, listeriosis, shigellosis perfringens poisoning, and campylobacteriosis.
In the overall prevalence study (Table 2), the most frequently isolated strains Proteus spp. and E. coli imply that they may have come from the contaminated water used in the ice cream industry. ese types of pathogens cause a series of diseases like diarrhea and gastrointestinal illness [44]. Milk is the major source of Staphylococcus aureus in the dairy industry that cause numerous food-borne diseases by producing toxins, especially when the dairy product is kept at room temperature. In our study, we also found that the tested ice cream samples were contaminated by Vibrio cholera, possibly due to improper handling, undercooking, and washing with unhygienic water [45]. Contamination of ice cream by this bacterium may cause diarrheal diseases, nausea, stomach cramping, and fever among ice cream consumers. Milk, different raw materials, fruit juice, and fruits pulps are frequently used in the ice cream industry, which are significant sources of Klebsiella spp. and Aeromonas spp. and contribute to developing pneumonia, septicemia, wound infections, and gastroenteritis like diarrhea, abdominal pain, headache, vomiting, or fever [46,47]. However, some other bacteria such as Salmonella paratyphi, Citrobacter spp., and Shigella spp. may contaminate ice cream from utensils, use of raw or undercooked components, and improper pasteurization process during ice cream production and may cause diarrhea, gastroenteritis, and vomiting among the consumers; school-going children [44]. Moreover, some of these identified bacterial species are also associated with neonatal sepsis and meningitis; few are the causes of sporadic pneumonia and sometimes cause milkborne infections [48,49]. e disc diffusion method showed a relatively higher antibiotic-resistance frequency among the identified bacterial isolates (Tables 3 and 4). Interestingly, we found that the most frequent isolated bacteria were highly resistant to cefuroxime, rifampicin, erythromycin, ampicillin, and methicillin, indicating that these commercial drugs did not possess antimicrobial activity and failed to halt the bacterial growth in human bodies. Proteus spp., Aeromonas spp., and Plesiomonas spp. were found to be resistant to cefuroxime, rifampicin, erythromycin, ampicillin, and methicillin but sensitive to levofloxacin, ciprofloxacin, norfloxacin, moxifloxacin, clotrimazole, and gentamycin. In contrast, the different antibiotic-resistant patterns reflect that the ice cream sample from the different retail shops might be contaminated with antimicrobial-resistant bacteria. e use of inadequate doses,   Journal of Food Quality inappropriate prescribing, extensive agricultural use, and inadequate regulation of antibiotics intake may have a vital role in achieving and generating antibiotic-resistant genes to the commonly infected bacteria. A significant correlation was also found between the phenotypes and genotypes of antibiotic resistance in bacteria that may indicate bacteria contain a pool of mobile genetic elements (Figure 3), and the transfer of antibiotic-resistance genes can easily occur between bacteria through conjugation or transformation. In this study, the isolated bacterial strains contain different sizes of plasmid DNA that may indicate the presence of antibiotic-resistance genes [23]. e phenotypic properties indicated that maximum bacteria were resistant to a couple of antibiotics (Table 5). Van TT (2007) conducted an experimental study and reported that high-molecular-weight plasmids are conjugative and contain many antibiotic-resistance genes [36]. Bacteria can also contain a number and size of plasmids for a long time and possesses antibiotic-resistant genes that can be transferred to the daughter cells in equal numbers [50]. Our electrophoresis data showed that the identified bacterial strains have large plasmids, so it may contain antibioticresistance genes and could have contributed to the spread of resistance genes. Previous studies have reported that plasmids with lower molecular weight are nonmotile or may lose during conjugation but have antibiotic-resistant genes [23,51]. However, in this study, we found that some of the isolated strains did not possess any plasmid band in gel electrophoresis but showed their antibiotic-resistant pattern phenotypically. erefore, we may conclude that some of these isolated bacterial strains indicated that multidrugresistant determinants were due to chromosomal DNA or other reasons instead of plasmid DNA [23], and it could be a potential risk for public health issues.   e data of this study will provide essential information regarding safety and quality issues of ice cream and ice cream like food items. ese may also help the regulatory authorities to take effective steps to adopt proper supervision during ice cream preparation and marketing for better children and public health management.

Conclusion
Most of the isolated bacterial strains from the ice cream samples are pathogenic. erefore, it can be concluded that microorganisms in ice cream have the potential role in causing severe illness, increased risk of complications, and hospital admission by consuming these types of food items. e correlations between the phenotypic and the genotypic susceptibility were also explored that the majority of the antibiotic resistances were due to acquiring plasmid-carrying antibiotic-resistance genes. is situation is alarming for Bangladesh, where health care facilities, surveillance for antibiotics medication, and facilities to detect MDR are underdeveloped. Our results highlighted the need to improve the hygiene level during pre-and postproduction, distribution, and retail storage practices to ensure the microbiological safety of ice cream and other food items.

Data Availability
e data used to support all the findings of this study are included within the article.

Conflicts of Interest
e authors declare that there are no conflicts of interest.

Authors' Contributions
AAM and MS contributed to the conceptualization and study designing; MS, MA, and MFH conducted all laboratory experiments, manuscript writing, and draft preparation; MS, MJI, and AI equally participated in data analysis, table and figure generation; SM and KI provided their logistics support; AAM participated in the writing, reviewing, editing, visualization, and supervision.

Supplementary Materials
Materials and method. Isolation and biochemical identification: Bacterial colonies were isolated by the streak plate method on Mannitol salt agar and PALCAM Listeria identification Agar base mixed with Listeria selective supplement (HIMEDIA, India) for the isolation of Grampositive bacteria Staphylococcus aureus and Listeria spp. Growth and morphological characteristics were recorded according to the Biology Libre Texts technique [30]. e isolated colonies were identified based on some biochemical assays including sugar fermentation, KIA, Motility, indole, urea, and citrate utilization tests [31]. Isolation and biochemical identification of Gram-positive bacteria Staphylococcus aureus and Listeria spp. from ice cream samples. Initially a total of five from sixteen ice cream samples (Kironmala, Doi Kulfi, Cocola Mango, Grameen Pipe, and Chocobar) were positive for bacteria culture and the rest of the samples did not show any colony (Supplemental Figure 1). After performing a biochemical assay, we identified that three (Kironmala, Doi Kulfi, and Chocobar) of the five colony-positive samples possess Gram-positive bacteria Staphylococcus aureus and the remaining two were unidentified (Supplemental Figure 2). On the other hand, we did not find any positive colony on the culture plate of PALCAM Listeria identification Agar base (mixed with Listeria selective supplement) media for the isolation and identification of Gram-positive bacteria Listeria (Supplemental Figure 3). (Supplementary Materials)