Metagenomic Sequencing Revealed the Potential Pathogenic Threats of Banknotes.

Banknotes have long been suspected to be biologically "dirty" due to their frequent human contact, which may transmit human microbial pathogens. Still, it is an unsettled issue whether the microbes on banknotes pose a real threat to human health. In several previous studies, metagenomic sequencing was used to reveal the diversities of microbes on banknotes but live microorganism culture and functional verification were lacking. In this study, we collected banknotes of RMB in China as well as dollar bills in the United States and analyzed the microbial biodiversity and drug resistance genes carried by the identified microbes by metagenomic sequencing and in vitro culture methods. We identified eight major genera of drug-resistant bacteria through screening of 30 antibiotics, and the blood agar plate culture uncovered six pathogenic fungal species. Numerous phage and six dangerous viral sequences were also found. These results should substantiate our concern about the potential risk of banknotes to human health.

Introduction

Banknotes are paper currency. The earliest banknotes, called “Jiaozi”, originated from the Northern Song Dynasty in China. Today, there are more than 200 types of banknotes circulating around the world. Paper banknotes are prone to contamination due to frequent human contact. Banknotes are not only the carrier but also the potential transmitter of many diseases.[1] There have been news reports that bank employees suffered from Condyloma Acuminatum without washing their hands after counting money.[2,3] Of particular concern are contagious microbial contaminants that pose serious health hazard. Some bad habits, such as touching food after using banknotes and not washing hands after counting money, have increased the chances of being infected with diseases.

Paper-based banknotes are excellent substrates for microbe attachment and absorption of various nutrients for microbial growth. It is likely that microbes on banknotes may carry antibiotic resistance genes and thus lead to the spread of antibiotic resistance to humans and other environments. Back in 1972, a JAMA published research paper[4] reported that coins may carry human pathogens. Since then, several studies have been published on microbial contamination of banknotes. Gabriel et al.[5] used the traditional microbial culture method to study the prevalence and drug resistance of Staphylococci on 155 European banknotes. Gedik et al.[6] used traditional microbial culture techniques to study the distribution and survival of methicillin-resistant Staphylococcus aureus, vancomycin-resistant Enterococcus, extended spectrum beta-lactamase (+) Escherichia coli, and other microbes with significant pathogenicity on currencies of different countries. Jalali et al.,[7] using a metagenomics approach in a study on Indian banknotes, identified 78 species of pathogens and 78 antibiotic resistance genes. However, the metagenomic DNA in this study was extracted using the traditional CTAB method, which only provides a biased representation of the microbial species. Heshiki et al.[8] found that the Hong Kong banknotes covered nearly 50% of the total genera in all the environmental samples. They also found that the most abundant bacteria genus was Acinetobacter and the chloramphenicol resistance gene was the most abundant antibiotic resistance gene. This study aims to provide a preliminary survey on the diversity of pathogenic microorganisms and their resistance genes on two widely circulated but little studied banknotes, the US dollar and the Chinese RMB, using a metagenomics approach and wet bench experimental validation.

Results

Species Found on the Banknotes of this Study

We randomly collected apparently heavily used 12 one-dollar bills in the United States and 12 one-yuan RMB bills in mainland China as two representative banknote groups. The classic STE method and a Mobio DNA extraction kit were used to obtain the metagenomic DNA, resulting in four metagenomic DNA samples. The yield of DNA obtained by different extraction methods varied significantly; the DNA yields of SteD, SteR, KitD, and KitR were 14, 39, 18, and 52 μg, respectively. In the sample names of SteD, KitD, SteR, and KitR, the “STE” represents the STE extraction method and the “Kit” represents the Mobio kit extraction method; the “D” and “R” represent dollar and RMB, respectively. NGS sequencing was used to obtain sequencing reads from the metagenomic DNA samples. All NGS sequencing raw data was uploaded to the NCBI-SRA database under the accession number of SRP128023. For the four samples, we performed species annotation at the phylum and genus level.

We analyzed the diversity of microbes in the four samples from the “kingdom” to “species” level. We found many bacteria with significant abundance in the RMB samples but not in US dollar samples. At the genus level, the microbes with the highest presentation in the KitR samples are Pseudomonas, but the microbes with the highest abundance in the KitD, SteR, and SteD samples are Acinetobacter. Species annotation analysis at the family level clearly showed the microbial diversity difference between the RMB and US dollar samples (Figure a). Clustering analysis of species at the genus level also showed the obvious difference between the dollar bills and RMB bills (Figure b). We also found that the presentation of species is dependent on the DNA extraction method (Figure b). However, regardless of the sample preparation methods, the diversity of microbes on RMB bills is significantly greater at all taxonomic levels than that on the US dollar bills (Table ). A more detailed list of species annotation can be found in Tables S1–S7.

Figure 1

Species annotation at the family and genus level of the four samples. (a) Statistical chart of species annotation results at the family level. The vertical axis represents the relative proportion of species annotated to a certain type; the horizontal axis represents the name of the sample; and the species category corresponding to each color block is shown in the legend on the right. (b) Clustering analysis of microbial species at the genus level. According to the species annotation and abundance information of all samples at the genus level, we selected the first 35 genera and their abundance information in each sample to draw a heatmap, and clustering was carried out from two levels of classification information and differences between samples, with sample information in the horizontal direction and species annotation information in the vertical direction. The cluster tree on the left side of the graph was a species clustering tree; the cluster tree above was a sample clustering tree.

Table 1

Comparison of Microbial Species of Each Level in Taxonomy per Sample

taxonomy level	Ste dollar	Ste RMB	Kit dollar	Kit RMB
kingdom	3	4	3	4
phylum	13	28	12	28
class	23	41	21	41
order	53	91	49	91
family	86	184	84	184
genus	170	452	180	449
species	288	885	314	890

We compared the effects of different extraction methods on microbial species and abundance. The STE DNA extraction method is a relatively mild procedure, which uses SDS to lyse cells at high temperature (55–65 °C), denature proteins, and release DNA. The strong alkali and vortices used in the Mobio PowerWater DNA Isolation Kit to extract DNA are quite violent. For some microorganisms with relatively thick cell walls, such as recalcitrant soil bacteria, it may not be efficient to extract their DNA. The kit method, in contrast, can lyse microorganisms with thicker cell walls, but the strong shear force may break the genomes of the easy-to-lyse microorganisms into small fragments.[9]

We found that on the RMB yuan bills, at the family level, about 65% of microbial species belonged to the family Enterobacteriaceae, Pseudomonadaceae, and Moraxellaceae. However, the abundance of Moraxellaceae and Pseudomonadaceae is different for the KitR and SteR samples. On the dollar bills, at the family level, about 35% of microbial species belonged to the family Moraxellaceae (Figure a). This may indicate that Enterobacteriaceae, Pseudomonadaceae, and Moraxellaceae are more likely to survive and/or reproduce on banknotes.

Presence of Phages on Banknotes

It is conceivable that wherever there are bacteria, there will be phages. We generated nonredundant de novo spliced nucleic acid sequences and searched them against the phage database containing 1534 phage genomic sequences.[10] We screened nucleic acid sequences with identity greater than 85, length greater than 200, gap less than 3, and E value less than 0.00001. Indeed, we found many kinds of phage sequences on banknotes. In the dataset of SteR and KitR, we found all 1534 kinds of phages. Interestingly, in the dataset of SteD and KitD, we detected only 23 kinds of phages. We counted the most abundant phage sequences in the four samples (Figure ) and found that the most abundant phage on RMB samples was Escherichia phage and the most abundant phage on the US dollar samples was Enterobacteria phage, all of which belonged to Enterobacteria phage. Enterobacteria are one of the most common Gram-negative bacteria, which can explain why Enterobacteria phage was found the most abundant phage on banknotes. On RMB samples, the most abundant phage found by two extraction methods is Escherichia phage EC1-UPM. However, on the US dollar samples, the most abundant phage found by the STE method is Enterobacteria phage ime09 and the most abundant phage found by the kit method is Enterobacteria phage Sf6. This suggests that the extraction method has little effect on the analysis results of phage sequences. A list of phages detected in the SteD and KitD sample could be found in Table S8. The huge difference in the number of phages between the dollar and RMB samples was unexpected and intriguing and could not be accounted for by the diversity of detected bacteria species.

Figure 2

Most abundant phages of the four samples. The vertical axis represents the number of the phages; the horizontal axis represents the name of the sample; and the name of the phage corresponding to each color block is shown in the legend on the right. In the sample names of SteD, KitD, SteR, and KitR, the “STE” represents the STE extraction method and the “Kit” denotes the Mobio kit extraction method; the “D” and “R” represent dollar and RMB, respectively.

Pathogenic Prokaryotic Microorganisms on Banknotes

Prokaryotic microbial infection is very common in clinics, but most infection can be effectively treated by proper use of antibiotics. The real threat to humans is microbial drug resistance. In this study, we attempted to identify known antibiotic resistance genes from the NGS data and verify the results using the standard blood agar culture method to grow bacteria to test drug resistance on plates containing antibiotics.

We found 37 drug resistance mechanisms (Table S9) by mapping our NGS data to the ARDB[11] and CARD database[12,13] (version: 2015-February-10). These resistance mechanisms are divided into the following categories: (1) efflux pump conferring antibiotic resistance, (2) gene conferring antibiotic resistance via molecular bypass, (3) antibiotic resistance gene clusters, cassettes, or operons, (4) antibiotic inactivation enzymes, (5) antibiotic target modifying enzymes, (6) antibiotic target protection proteins, and (7) antibiotic target replacement proteins. By mapping our NGS data to the ARDB and CARD databases (Figure ), we found seven types of drug resistance mechanisms. The efflux pump conferring antibiotic resistance and the antibiotic target modifying enzymes accounted for the highest proportion, accounting for 53% and 30%, respectively (Figure a). When compared with the CARD database, the efflux pump conferring antibiotic resistance and the antibiotic resistance gene clusters, cassettes, or operons accounted for the highest proportion, 43% and 45%, respectively (Figure b). The ARDB database has not been updated since 2009, so the mapping result is not as complete as the CARD database. Our results indicate that the most common antibiotic resistance mechanism is the efflux of antibiotics (aminoglycoside and macrolide antibiotics, etc.) or generation of antibiotic resistance gene variants (aminoglycoside resistance gene, chloramphenicol resistance gene, etc.).

Figure 3

Statistical chart of seven drug resistance mechanisms obtained by mapping our NGS data to the ARDB and CARD databases. (a) Seven drug resistance mechanisms obtained by mapping our NGS data to the ARDB databases. (b) Seven drug resistance mechanisms obtained by mapping our NGS data to the CARD databases. The vertical axis represents the number of resistance mechanisms; the horizontal axis represents the categories of the drug resistance mechanisms; and the type of drug resistance mechanism corresponding to each color block is shown in the legend on the right.

Based on the mechanisms of drug resistance, we selected 30 antibiotic drugs and designed the antibiotic resistance tests to isolate the corresponding drug-resistant microbes on blood agar plates. Taking tobramycin as an example, we designed four control groups and two experimental groups to screen drug-resistant bacteria on blood agar plates.

Blood agar medium is the most commonly used medium for laboratory testing to isolate and culture pathogenic microorganisms from clinical specimens. Due to the special growth factors provided by defibrinated sheep blood, most pathogenic microorganisms can grow on blood agar plates. Microorganisms that do not grow on blood agar plates are likely unharmful to human health.

For microbes grown on blood agar plates with drugs, we chose single colonies of 660 microbes with different morphologies and colors and applied the Gram stain for classification. After classifying and removing redundancy, we took 350 single colonies to extract genomic DNA and performed PCR to amplify the 16s rDNA full-length sequences. Sanger sequencing and NCBI online Blast were applied for accurate taxonomic identification of each clone. We identified nine major categories of drug-resistant bacteria: Enterobacter sp., Enterococcus sp., Enhydrobacter sp., Klebsiella sp., Leclercia sp., Acinetobacter sp., Pantoea sp., Pseudomonas sp., and Moraxella sp. (Table S10). The screening process is shown in Figure .

Figure 4

Screening process of drug-resistant bacteria on banknotes.

We compared the metagenomic data with those of the cultivation study at the genus level and found that the two results were basically the same. However, the cultured Leclercia sp. and Enhydracter sp. were not found in the metagenomic data. The possible reason is that the content of Leclercia sp. and Enhydracter sp. is relatively small. This indicates that the metagenomic data, though comprehensive, is still far from panoramic. Some factors such as the thickness of the cell wall, the number of bacteria, and the content of endogenous DNA enzymes of the bacteria will affect the reliability of the metagenomic results.

We also found 63 species anaerobic microbes belonging to four genus, Alkaliphilus, Alistipes, Bacteroides, and Bifidobacterium, on RMB and dollar samples through NGS data (Table S11). Among them, Bacteroides is the normal flora of the human intestine, oral cavity, upper respiratory tract, and the reproductive tract. However, it may cause endogenous infection only when the immune function of the body is compromised or the flora is imbalanced. Since banknotes are generally exposed to air, it is likely that most of these anaerobic microbes might not be viable on banknotes and thus we did not attempt to study the status of anaerobic microbes.

Pathogenic Viruses on Banknotes

We downloaded all nucleotide sequences of the 490 human infectious viruses from the NCBI data base (https://www.ncbi.nlm.nih.gov/genomes/GenomesGroup.cgi?taxid=10239&host=human) and assembled a local Blast database to detect virus sequences in the banknote microbiome dataset. The banknote sequence reads were processed to remove low quality bases and adapter sequences to generate “clean reads”, which were then Blast searched against the 490 virus sequences. We found six viruses, Alphapapillomavirus 7 (human papillomavirus 18, HPV18), Human alphaherpesvirus 1 (herpes simplex virus type 1, HSV-1), Macaca mulatta polyomavirus 1 (SV40), Molluscum contagiosum virus subtype 1 (MCVI), Orf virus, and Shamonda orthobunyavirus (SHAV), in the banknote dataset (Figure ). These are all disease-causing or disease-associated virus in humans. The HPV is a spherical DNA virus that causes proliferation of human cutaneous and mucosal squamous epithelia.[14] HPV infection is associated with cancers such as cervical cancer, head and neck cancers, and carcinoma of the oropharynx and tonsils.[15] HSV is a virus that causes herpes, which is categorized into two types: HSV-1 and HSV-2. HSV is widely distributed around the world. HSV-1 can cause oral herpes and genital herpes and its infections are life-long. SV40 is a small DNA tumor virus belonging to the Polyomaviridae family, which has transforming and tumorigenic properties.[16] Whether SV40 can cause human cancer has been controversial.[17] MCVI is a DNA virus, a member of the Poxviridae subfamily. Molluscum contagiosum is a human disease caused by MCVI, which has been the principal poxvirus cause of human disease since the eradication of smallpox.[18]Orf virus is a DNA virus belonging to the Parapoxvirus genus,[19] commonly known as sore mouth disease in sheep and cattle.[20] It can be transmitted from animals to humans and cause orf, such as contagious pustular dermatitis, infectious pustular dermatitis, or Ecthyma contagiosa.[20] However, human to human transmission is very rare.[19] SHAV is an RNA virus that belongs to the genus Orthobunyavirus of the family Bunyaviridae.(21) SHAV is an arthropod-borne virus (arboviruses), mainly transmitted by mosquitoes, and is teratogenic in ruminants.[22]

Figure 5

Six viruses found on banknotes and their association with human diseases.

Interestingly, among the six viruses found, one is an RNA virus. This could indicate that the DNA produced in the life cycle might also be captured and detected by the high-throughput sequencing of this study. The virus DNA detected could either be the contaminating intermediate of the RNA virus’ life history or that the virus could still reproduce and produce DNA on banknotes. These findings suggest that other more common RNA viruses such as HIV and hepatitis C virus should also be present on banknotes in a more stable DNA form.

At present, it is not known whether these viruses are infectious to humans. Banknotes as a medium for virus disease transmission are worth further study.

Pathogenic Eukaryotic Microorganisms on Banknotes

Fungi are an important category of human and animal pathogens. Previous studies[5,6,23] were generally concerned with bacteria on banknotes. We analyzed the diversity of microbes in the four samples at the family level. We found that the number of fungi on RMB samples was slightly higher than that on the US dollar samples, but there was no significant difference in the diversity of fungi. Species annotation analysis at the family level clearly showed the eukaryotic diversity difference between the RMB and US dollar samples (Figure ). At the family level, the fungi with the highest presentation in the KitR and SteR samples are Saccharomycetaceae, the fungi with the highest abundance in the KitD samples are Sordariaceae, but the fungi with the highest abundance in the SteD samples are unclassified. This shows that the DNA extraction method has a significant influence on the representation of fungi.

Figure 6

Statistical chart of eukaryotic annotation results at the family level.

The vertical axis represents the relative proportion of species annotated to a certain type; the horizontal axis represents the name of the sample; and the species category corresponding to each color block is shown in the legend on the right. In the sample names of SteD, KitD, SteR, and KitR, the “STE” represents the STE extraction method and the “Kit” denotes the Mobio kit extraction method; the “D” and “R” represent dollar and RMB, respectively.

To survey the prevalence of viable fungi on banknotes, we used the classic PDA medium plate to enrich the fungi. We extracted genomic DNA from single colonies and amplified the 18s rDNA full-length region by PCR. We then used Sanger sequencing and NCBI online Blast analysis to identify fungal species. We found the following fungi: Pichia guilliermondii, Aureobasidium pullulans, Rhodotorula mucilaginosa, Meyerozyma guilliermondii, Debaryomyces hansenii, Candida parapsilosis (C. parapsilosis) (Table S12). We found that the viable fungi on plates were mainly yeast and C. parapsilosis. Compared to those of the found bacteria, the number of fungus species and their abundance were much lower, which was consistent with the data obtained from metagenomic sequencing data. We noticed that C. parapsilosis is a pathogen that can cause recurrent skin diseases in humans. These findings suggest that banknotes could be an important medium to transmit fungal disease due to frequent contact with human hands.

Discussion

We are taught to avoid microbial diseases by minimizing personal contact and sharing of personal items. However, banknotes seem to be the only exception. Banknotes are a daily used item that transferred among many people during their lifetime without sanitation. Ironically, rarely do we realize the fact that banknotes are a good carrier of microbes and banknote circulation could be an important medium of spreading contagious microbial diseases. All countries in the world have inspection and quarantine for the import of goods and personnel to prevent species invasion and pathogen entry. However, there is no country where the inspection and quarantine departments would check banknotes carried by the passengers. This study showed that on banknotes, there are abundant viable and drug-resistant human pathogenic microbes that may pose potential threats to humans.

In this study, we found six viruses on banknotes, among which HPV18 and HSV-1 are known harmful agents. HPV has many subtypes, and HPV18 is considered high risk. HPV18 infection is closely related to cervical cancer in women. According to the World Health Organization (WHO), about 70% of cervical cancers are caused by HPV16 and HPV18. Cervical cancer is the most common gynecological malignancy, with an estimated 570,000 cases[24] in 2018 (accounting for 84% of new cases worldwide). HSV infection is extremely common in the population with more latent and recurrent infections. Approximately 45–90% of the population carries HSV-1, which is more frequent in developing countries.[25] According to WHO estimates, 3.7 billion people under the age of 50 (67%) are infected with the HSV-1 virus worldwide. Most oral and genital herpes infections are asymptomatic but can still be transmitted to others. The discovery of HPV18 and HSV1 on banknotes indicates that the virus may spread through the banknotes, endangering people’s lives and health.

WHO has published a list of “PRIORITY PATHOGENS”, listing 12 families of bacteria that are most threatening to human health. In this study, we found Acinetobacter baumannii (A. baumannii) and Shigella sp. on the banknotes of both RMB and US dollar. A. baumannii ranks first in the WHO’s list at the “Priority 1: CRITICAL” risk level. A. baumannii is a Gram-negative opportunistic nosocomial pathogen that causes 2–10% of all Gram-negative infections in hospitals.[26]A. baumannii is one of the six most important worldwide multidrug-resistant microorganisms in hospitals.[27] The found A. baumannii on banknotes was resistant to multiple antibiotics, including norfloxacin, ampicillin, fusidic acid, daptomycin, tetracycline, mupirocin, and vancomycin. A. baumannii can adhere to abiotic surfaces such as glass, plastics, etc. and form biofilms.[28]A. baumannii has great vitality and can survive even long-term exposure to dry conditions and nutrient deficiencies. A. baumannii infection can cause skin and wound infection, pneumonia, soft tissue infections, endocarditis, urinary tract infection, meningitis, and septicemia[29] and even lead to death and thus has raised considerable concerns as there are no antibiotics for the infection of this microbe. Shigella is a group of Gram-negative short bacilli, which is mainly prevalent in developing countries. It is the most common pathogen of human bacillary dysentery, which invades the human colon and can rapidly acquire antibiotic resistance.[30] The presence of Shigella on banknotes is a potential risk factor for epidemic bacterial dysentery.

C. parapsilosis, an opportunistic human pathogen, is a yeast-like fungus of the genus Candida. It is an important cause of hospital-acquired infections that are commonly found on the skin of healthy people. C. parapsilosis is the main species causing infection of neonatal and adult people with low immunity in hospitals.[31] The mortality rate of Candida infection in hospital-acquired infections is approximately 40%, of which approximately 15% is caused by C. parapsilosis.[32] Even more frightening is that the species showed significant resistance to fungal drugs.[33] The presence of C. parapsilosis on banknotes means that the probability of C. parapsilosis infection has increased.

A dollar bill is made of a mix of 75% cotton and 25% linen,[34] while an RMB bill is made of short-staple cotton. Cotton-based banknotes provide fiber surfaces, which provide ample opportunities for bacteria to adhere. Linen has a rougher surface than cotton, and microbes are expected to adhere more easily on to the dollar bill. However, our results show that the number of microorganisms on a US dollar bill is much smaller than that on an RMB bill, even without considering the fact that the size of the one dollar bill is about 20% larger than that of the one yuan RMB bill. This indicates that the types and quantities of microorganisms carried by banknotes are related not only to their materials but also to other factors, such as transmission frequency and environmental exposure, which are reflected by the economic situation of a country, the social and municipal health infrastructure, and the average life of the banknotes.[35] The type and number of microorganisms on the banknote are the result of the combined effects of the above factors.

It is true that pathogens need to reach a certain amount and have certain virulence to infect humans and animals. The area of banknotes is small, so the number of pathogens carried by them may not be numerous and some of these microorganisms may be dead. Therefore, there was a view that microorganisms on banknotes are harmless. However, in currency circulation, the pathogens on banknotes may be exposed and transferred to an environment that is rich in nutrients and suitable for their settlement and rapid growth and reproduction, resulting in massive proliferation of pathogens. This is common in densely populated developing countries in which sanitation conditions are poor and management is not standardized.

Conclusions

Though banknotes may not lead to human infection directly, its circulation may cause rapid spread of pathogens. In hospitals, pharmacies, and food and beverage outlets, there are no disinfection measures for banknotes. Cashiers are often also chefs or shop assistants. Direct or indirect contact of banknotes with food or drugs is inevitable. Food may be stored for a long time before eating, giving the opportunity for pathogens to reproduce. Drugs are used for patients, who may have a weak immune system during their illness and are vulnerable to invasion by external pathogens. For example, some injectable drugs or medical devices that need to puncture the epidermis may introduce pathogens into the body.

Of particular concern is that banknotes may be a loophole that can be used by terrorists as a stealthy vehicle to deliver highly life-threatening biological agents such as Bacillus anthracis. It is anticipated that an eventual solution to these concerns is to use electronic currency and phase out banknotes completely.

Methods

Sample Collection and DNA Preparation

We collected RMB samples in China and US dollar samples in the United States, one in the eastern hemisphere and the other in the western hemisphere. The dollar samples and the RMB samples are treated separately to avoid cross contamination. In this study, we collected 12 one yuan bills of RMB in the city of Fuzhou, China, and 12 one dollar bills around the Dallas area in the United States. The dollar and RMB bills were of apparently similar age but relatively worn out. The surface of each bill was washed with sterile water, and the liquid was filtered through a 0.22 μm filter to collect the microbes. Extraction of metagenome was performed for high throughput sequencing. In order to obtain the most complete information on the metagenomic DNA, we used two genomic DNA extraction methods (Scheme S1), the classic STE buffer (sodium chloride, Tris–HCl, and EDTA) and Mobio kit to isolate bacterial genomic DNA from banknotes. The STE is suitable for bacteria, especially Gram-negative strains. The kit from Mobio is advantageous for some tough-to-lyse microbes. However, the harsh cell grinding and disrupting procedure in this method may damage the genomic DNA of some fragile microbes. In this study, four DNA samples of the metagenome were studied, which were labeled as follows: SteD: metagenomic DNA from dollar samples extracted using the STE method; KitD: metagenomic DNA from dollar samples using the Mobio Kit; SteR; metagenomic DNA from RMB samples using the STE method; and KitR: metagenomic DNA from RMB samples using the Mobio Kit. The extracted DNA samples were sequenced and analyzed separately.

Resistance Gene Annotation

The sequencing methods can be found in Scheme S1. The bioinformatics analysis method for NGS data of this study is described in Scheme S2. We Blast searched the Unigenes to ARDB database[11] (http://ardb.cbcb.umd.edu/) and CARD database[12,13,36] (https://card.mcmaster.ca/) with the default parameter setting blastp, -e 1e-5,[37−40] then filtered the aligned result, and chose the identity value bigger than the lowest identity value from the aligned result of each sequence to obtain reliable results of resistance gene annotation. Based on the aligned result, the relative abundance of different resistance genes was calculated. Finally, based on the abundance of resistance genes, the abundance bar charts, the abundance cluster heatmap, and the resistance genes’ number difference between groups were displayed. In the same way, the resistance genes’ abundance distribution in each sample, the species attribution analysis of resistance genes, and the resistance mechanism of resistance genes analysis were also conducted.

Screening of Antibiotic-Resistant Microorganisms

Blood agar medium is generally used for the isolation and culture of clinical specimens. First, the samples were subjected to serial dilution with dilution factors of 2, 4, 8, and 16. We used blood agar medium as well as 30 antibiotics to culture pathogenic microorganisms in the diluted sample, using JM109 E. coli as a control. The list of antibiotics is shown in Table . In this experiment, we picked a total of 660 single colonies and expanded culture in blood medium. Subsequently, Gram staining was performed, and single colonies were classified according to colony morphology, color, and Gram properties.

Table 2

Thirty Antibiotics and Their Working Concentration

antibiotic	working concentration (μg/mL or U/mL)
ampicillin	50
streptomycin	100
kanamycin sulfate	50
tobramycin	10
gentamycin sulfate	20
neomycin sulfate	50
penicillin g sodium salt	100 U/mL
netilmicin	10
amikacin	30
sulbactam	25
fusidic acid	5
colistin sulfate	10
daptomycin	2
tetracycline hydrochloride	15
rifampin	100
novobiocin	20
mupirocin	8
nalidixic acid	30
ciprofloxacin	1
norfloxacin	2
polymyxin B	300 U/mL
trimethoprim	10
spectinomycin	100
sisomicin	10
paromomycin sulfate	100
fosfomycin	30
enoxacin	3
erythromycin	20
ribostamycin	25
vancomycin	10

Genome Amplification and Sanger Sequencing Verification

According to the microscopic examination results, the resistant bacteria were classified into bacteria and fungi. Bacterial genome was amplified with 16S rRNA universal primers 27F, 1492R, and 2 × Taq Plus MasterMix (CWBIO, China). Each reaction was carried out for 10 min at 94 °C; 30 cycles of 30 s at 94 °C, 30 s at 53 °C, and 90 s at 72 °C; and 10 min at 72 °C. Fungal genomes were amplified with 18S universal primers 18S+, 18S- or NS1, NS4 or NS1, NS8, and 2 × Taq Plus MasterMix (CWBIO, China). Each reaction was carried out for 10 min at 94 °C; 30 cycles of 30 s at 94 °C, 30 s at 53 °C, and 2 min at 72 °C; and 10 min at 72 °C. PCR products were detected using a 1% agarose gel and subjected to Sanger sequencing for accurate taxonomic identification. All primers used in this article are shown in Table S13.