Description and Design
Abstract
Global food safety problems emerge one after another, and people's requirements for quality of life are
constantly improving. At this time, it is very important to put forward a method that can quickly and
directly detect pathogenic bacteria. We hope that this project can strengthen people's attention to food
safety and reduce the tragedies caused by food safety. We used phage receptor binding protein
(RBP)-nanomagnetic bead complexes (RBP-MBs) combined with paper-based sensors to isolate and
colorimetrically detect four types of foodborne pathogenic bacteria simultaneously. Compared with the
existing detection technology, it is quicker, simpler and more efficient in identifying multiple
foodborne pathogens at one time.
1 Microbial pathogens
According to the statistics of the World Health Organization in 2020, 600 million people (almost one in
every 10 people) in the world suffer from diseases due to eating contaminated food every year, and
420,000 people die, resulting in a loss of 33 million healthy life years (WHO, 2022). pathogenic
Escherichia coli (E. coli, EC), Staphylococcus aureus (S. aureus, SA), Vibrio parahaemolyticus (V.
parahaemolyticus, VP) and Salmonella enteritidis (S. enteritidis, SE) are four common foodborne
bacterial pathogens.
1.1 Pathogenic E. coli
E. coli are mostly harmless bacteria that live in the intestines of people and animals and contribute to
intestinal health. However, eating or drinking food or water contaminated with certain types of E.
coli can cause mild to severe gastrointestinal illness. Some types of pathogenic (illness-causing) E.
coli, such as Shiga toxin-producing E. coli (STEC), can be life-threatening. Different types of E.
coli tend to contaminate different types of foods and water. Previous U.S. outbreaks of pathogenic E.
coli have included leafy greens, sprouts, raw milk and cheeses, and raw beef and poultry.
Some wildlife, livestock, and humans are occasional carriers of EC and can contaminate meats and food
crops. Ruminant animals such as cattle, goats, sheep, deer or elk, as well as other animals such as pigs
or birds are known carriers of EC, such as STEC, and are often the pathway as to how STEC is introduced
into the environment.
Figure1. Recent outbreaks of E. Coli
(https://www.fda.gov/news-events/public-health-focus/e-coli-and-foodborne-illness)
1.2 S. aureus
S. aureus is the third most prevalent microbial pathogen for food-borne illnesses and
accounts for about 25% of microbial food poisoning events (Wei, 2011). Enterotoxin produced by SA in
food can cause Staphylococcal food poisoning (SFP). Symptoms of SFP include an acute onset of nausea,
severe vomiting, abdominal pain, and diarrhea (Tong, 2015). SA has a certain tolerance to high
temperature (up to 80℃) and high salt concentrations (up to 15% concentration of NaCl).
Figure 2 SEM image of Staphylococcus aureus taken from a vancomycin intermediate resistant
culture
(CDC/ Matthew J. Arduino, DRPHPhoto Credit:Janice Haney Carr, Public domain, via Wikimedia Commons)
1.3 V. parahaemolyticus
V. parahaemolyticus is a type of marine bacteria, mainly found in fish, shrimp, crab, shellfish, seaweed
and other seafood. Raw seafood is a high-risk food for VP infection. The clinical manifestations of food
poisoning caused by VP are acute onset symptoms, mainly including abdominal pain, vomiting, diarrhea,
and even shock and death (Ghenem, 2017). During 1990–2019, VP was responsible for more than 40 global
outbreaks (Pazhani, 2021). According to the statistics released by the China National Center for Food
Safety Risk Assessment (CFSA), in recent years, VP has become the leading microbial pathogen causing
foodborne diseases in China (Wu, 2014). In 2020, a survey in China showed that the prevalence of VP in
fish and shrimp reached 14.9% in summer and 7.3% in winter (Li, 2020).
Figure 3 SEM image of Vibrio parahaemolyticus (Sathiyamoorthi, 2021)
1.4 S. enteritidis
S. enteritidis is the most prevalent strain of Salmonella and the predominant cause of foodborne
Salmonella diseases (Shah, 2017). In the United States, SE accounts for
about 32% of Salmonella outbreaks (Coveny, 2022).
Salmonella infections are common to humans and animals, and are mainly caused by eating contaminated
food. According to statistics, among all bacterial food poisoning in the world, Salmonella poisoning
often ranks first (CFSA, 2020). Salmonella is also one of the most common bacterial foodborne pathogens
in the inland areas of China (Lin, 2016). According to the Centers for Disease Control and Prevention
(CDC), 9 multistate outbreaks of Salmonella food poisoning were reported in the United States in the
year of 2021, causing more than 1,200 people sick and 250 Americans hospitalized (News desk, 2022).
Typical symptoms of Salmonella infections include fever, nausea, vomiting, diarrhea and abdominal colic
(Lin, 2016). Salmonella can survive for 15 minutes at 60 ℃, 2-3 weeks in water and 3-4 months in
refrigerator (Hebei Institute of Food Inspection, 2016)
Figure 4 Colorized SEM of SE. Blue is growth medium. Picture is colored in false colors to
illustrate
difference. (Photo by Jean Guard, U.S. Department of Agriculture, Public domain, via Wikimedia Commons)
2 Current detection approaches for microbial pathogens
Pathogen testing is important to ensure the safety of food being taken and monitor the hygiene quality of
food processing and shelf-life stability.
The golden standard method for food-borne pathogen detection is the conventional microbiological test
based on bacteria culturing. It is reliable and accurate, but it takes heavy workload and needs 2-3 days
to get results and another 7-10 days for confirmation (Adzitey, 2013).
Figure 5 Representative culture media for Salmonella Typhi strains on: (a) XLD agar; (b) SS
agar
(Salman, 2021)
Immunodiagnostics such as Enzyme-Linked Immune Sorbent Assay (ELISA) uses antibodies to specifically
identify pathogens. It has high specificity and sensitivity, and can be used in large scale. Though it
is relatively fast compared with the conventional culturing tests, it is still not in “real-time”. In
addition, low sensitivity and affinity of the antibody to the pathogens can occur and contaminants may
interfere with the results (Umesha, 2018).
Polymerase chain reaction (PCR) technology targets nucleic acid like DNA of the pathogens. PCR is
sensitive, specific, accurate, and can sense low level DNA in the sample over the above two conventional
methods. PCR takes hours to over a day in processing. The reaction specificity is affected by conditions
such as magnesium concentration and cross contamination. Multiplex PCR (mPCR) and broad-range PCR assays
are developed based on PCR to simultaneously detect multiple pathogens. Since they are multitasking,
they can save time and labor work compared with PCR when testing for several pathogens (Umesha, 2018).
Metabolomic approaches, such as liquid chromatography mass spectrometry (LCMS), gas chromatography mass
spectrometry (GC-MS), matrix assisted laser desorption/ionization tandem time of flight mass
spectrometry (MALDI-TOF-MS) and nuclear magnetic resonance (NMR), are used to detect the metabolites of
pathogens. These methods are currently used more in pathogenic fungi than bacteria (Oyedeji, 2021). The
cost for equipment and maintenance are high for these methods and the processing is time-consuming.
Due to the complexity of the testing methods above, they are incapable to give results in real time,
require laboratory equipment and professional staff together with laboratory management and quality
control measures. Testing costs are not cheap.
ATP bioluminescence technology is a fast (within minutes), portable and convenient tool commonly used to
analyze the overall quantity of microorganisms in food via ATP detection in cells. But it has the
disadvantages of low sensitivity and inability to detect specific strains. In addition, the results are
affected a lot by factors like temperature change (Sun, 2022).
3 Aim of our project
Our project aims to develop a sensitive and rapid detection tool for four microbial pathogens, S. aureus
(SA), V. parahaemolyticus (VP), pathogenic E. coli (EC), and S. enteritidis (SE) in food. Compared with
other available approaches, our tool is sensitive, rapid, cheap, portable and easy to handle, with no
requirements for devices, laboratory setup, and professional personnel. Therefore, our tool can be
widely used in various conditions such as household settings and is especially friendly in remote areas
with poor economic condition and laboratory resources.
4 Overall design of our tool
Our detection tool consists two major components.
- Four receptor binding protein (RBP)-nanomagnetic bead complexes (RBP-MBs): The RBP side of the
complexes can bind specifically to SA, VP, EC, and SE, respectively and the nanomagnetic bead side
can be attracted by magnets. The complexes can isolate the four strains of interest from samples
rapidly and easily while maintaining the viability of the bacteria.
- Paper-based sensors: The sensors contain bacterial cell lysis reagent and four different
chromogenic substrates. It shows different color signals when in contact with the pathogens
respectively.
Figure 6 Overall design of our detection tool for SA, VP, EC and SE strains in food (images
from
https://pixabay.com)
To develop our tool, we used three steps: Construction of plasmid vectors for RBP production, formation
of RBP-nanomagnetic bead complexes and development of paper-based sensors.
5 Components of our tool
5.1 RBPs
RBPs are special tail proteins of bacteriophages. The phage-specific recognition of its host bacteria is
realized by RBPs. Different RBPs can specifically bind to different bacteria strains. RBPs have been
used to develop various diagnostic tools for its efficiency and sensitivity in identifying bacterial
pathogens.
In our project, four RBPs, gp13, gp15, CBD(C-terminal cell-binding domain), and TSP (tailspike protein),
are initially selected to detect EHEC O157:H7(a typical strain of pathogenic E. coli), VP, SA, and SE,
respectively, due to their high specificity in binding with the corresponding bacteria strains. After
expert advice, our project abandoned the EHEC O157:H7 strain for strain testing for safety concerns, and
replaced the relatively safe E.coli strain K12 MG1655 before entering the laboratory for a formal
experiment.
gp13 is a putative RBP within the genome of Phage EP335. Researches have demonstrated that GFP-gp13 bind
strongly and evenly to the cell surfaces of E. coli O157 strains prone to EP335 infection (Witte, 2021).
Because we changed the EC strain into K12 MG1655, we replaced gp13 with a more broad-spectrum RBP
protein TFP (tail fiber protein) that can bind to multiple pathogenic E.coli strains.
Figure 7 Fluorescence and phase contrast images of GFP-gp13 cell binding to different E. coli
O157
strains such as TW01286, 396, 999/1 and 777/1. Scale bars represent 5 mm (Witte, 2021).
TFP is a receptor binding protein (RBP) from the genome of Escherichia coli (EC) phage T7. Researches
have demonstrated that GFP-TFP can bind to the cell surfaces of several E. coli strains such as the
pathogenic E.coli O157 and K12 MG1655.
gp15 is an RBP found in the genome of the vB_VpaP_GHSM17 phage. vB_VpaP_GHSM17 is a recently isolated
phage that infects VP (Liang, 2022). gp15 is a protein of the phage that can specifically identify VP.
CBD is the C-terminal part of the endolysins from gram-positive phages. Studies have shown that CBDs of
the bacteriophages adhere to the bacterial cell wall and grant endolysins access into the bacteria
(Stone, 2019). CBDs can bind specifically with target bacteria with great affinity, and can be easily
manufactured in an E. coli expression system. Therefore, CBDs are widely investigated in biosensor
development for bacteria detection. The CBD used in this project is from the genome of the phage P108
which infects SA. It has showed broad-spectrum detecting capability toward SA strains (Wang, 2020).
Figure 8 Characterization and purification of endolysin LysP108 derived from phage P108
of SA strain
XN108. (A) Schematic illustration of structure of phage endolysin LysP108. (B) 3D structure of endolysin
protein LysP108. (Lu, 2021)
Tailspike protein (TSP, also named as gp9) is an RBP from the genome of bacteriophage Salmonella P22. TSP
is used by the P22 phage to attach to the lipopolysaccharides of its target bacteria (Seul,2014). TSPs
play an important role in the initial stage of phage P22’s infection of Salmonella enterica serovar
Typhimurium, which help in the formation of a channel to translocate viral genome into the cytoplasm
(Wang, 2019).
Figure 9 Phage P22 binds obliquely to the cell surface of Salmonella enterica serovar
Typhimurium. PG:
peptidoglycan cell wall; OM: outer membrane of the bacteria; gp9: TSP (Wang, 2019).
We used the pET28a plasmid as the expression vector for the RBPs. And we constructed plasmid vectors of
pET28a-TFP, pET28a-gp15, pET28a-CBD, and pET28a-TSP via seamless cloning to produce TFP, gp15, CBD and
TSP proteins, respectively. To verify the binding ability of these proteins with the target bacteria,
plasmid vectors of pET28a-GFP-TFP, pET28a-GFP-gp15, pET28a-GFP-CBD and pET28a-GFP-TSP are constructed to
express green fluorescent protein (GFP) fusion proteins of TFP, gp15, CBD and TSP. Under the confocal
laser scanning microscope, green fluorescence on the cell surface of the target bacteria after
incubation with GFP fusion proteins indicates specific binding of the proteins to viable bacteria.
Figure 10 Plasmid vectors of pET28a-TFP and pET28a-GFP-TFP
In our project, we chose RBP(TFP, gp15, CBD, TSP) from several phages to accurately capture the specific
target bacteria (EC,VP,SA,SE).Through the construction of GFP-RBPs (GFP-TFP, GFP-gp15, GFP-CBD, and
GFP-TSP), the four RBPs were proven to be effective in our experiment. RBPs were constructed as part
number BBa_K4430000~BBa_K4430003,and GFP-RBPs were constructed as part number BBa_K4430005~BBa_K4430008.
Click here for more details
5.2 RBP-nanomagnetic bead complexes
Carboxyl magnetic beads (MBs) contain carboxyl functional groups on their surface and are
superparamagnetic nanomagnetic beads. These beads can be covalently conjugated with primary amine from
proteins, nucleic acids and other molecules via stable amide bond. The conjugated nanomagnetic bead
complex has magnetic property and can be separated by magnets.
In our project, the RBPs, TFP, gp15, CBD and TSP, produced by the plasmid vectors are conjugated to the
nanomagnetic beads and four types of RBP-nanomagnetic bead complexes (RBP-MBs) are formed. The RBP side
of an RBP-MB binds specifically to the target bacteria and the MB side can be attracted by a magnet.
Therefore, in a solution with the target bacteria, the bacteria can be screened and captured by the
RBP-MBs and can be isolated via magnetic separation. Throughout the process, the purified bacteria
remain viable, and thus can be used for further testing requirements of various kinds such as tests for
drug resistance and genetic information.
Figure 11 Illustration of magnetic isolation of analyte using nanomagnetic beads
5.3 Paper-based sensors
Paper-based sensors (e. g. pH values and pregnancy test paper) are currently the most promising sensors
in the chemical/bioanalysis-related fields, as analysts prefer to perform field and real-time /visual
detection without the help of analytical instruments. Sensors built with paper as a substrate are
low-cost and flexible with short response time. Moreover, they are biodegradable and suitable for mass
deployment in resource-limited areas and can be easily used by unskilled operators. Paper is also a
great medium for immobilization and trapping and, in some cases, for binding with biomolecules. Its
porous structure with large connected pores composed of cellulose fibers allows transporting liquid by
means of capillary forces that result in short response time. The porous structure of paper also allows
any nano- and microparticles to remain immobilized in the paper structure. Paper can be functionalized
with certain materials such as nitrocellulose paper used for immobilizing nucleic acids for selective
sensing. Paper-based potentiometric sensors are reported for detecting many ions and proteins.
Paper-based pathogen and virus sensors are also easy to incinerate. They can be used as one-time-use
front end in sensor systems that can be peeled off and replaced (Dolai, 2020).
Figure 12 Office paper platform for bioelectrochromic detection of electrochemically active
bacteria
using tungsten trioxide nanoprobes (Marques, 2015)
Our team designed paper-based sensors that can show different visual signals for a solution sample
containing EC, VP, SA, and SE. To develop the sensor, we prepared a reaction system with the optimal
reaction concentrations of a lysing agent and one of the four chromogenic substrates, chlorophenol
red-β-D-galactopyranoside (CPRG), X-β-glu, 4-nitrophenyl β-D-glucopyanoside (pNPG), and magenta
caprylate on a paper-based carrier. When drops of resuspension buffer of the target bacteria purified by
RBP-MBs are added to the sensor, bacteria are lysed and enzymes within the cells are released into the
system: β-galactosidase (β-gal) released from EC, β-glucosidase (β-glu) from VP, α-glucosidase (α-glu)
from SA, and esterase from SE, respectively. The products of the catalytic reactions of the chromogenic
substrates by these enzymes give different visual signals that can be detected by naked eye.
5.4 Tools used for color detection of the paper-based sensor
Snipaste, a software application from Windows system for screen capturing, and the screen capture tool
from Wechat are both appropriate tools for the analysis of visual signals of the paper-based sensors.
They can list out RGB (red, green, blue) color values, which distinguish subtle color differences that
naked eyes can’t detect. These tools make our paper-based sensors more sensitive in detecting low
concentrations of target bacteria. In order to precisely analyze the color signals of our paper-based
sensors, our modeling group developed a software called Image Colorimetric Detection (ICD) with the help
of the research team from South China University of Technology. Bacteria concentration can be calculated
by the data given by ICD.
6 Application of our tool
Canned foods and meat products are the two most common food groups that cause foodborne illnesses. Main
reasons of microbial contamination of canned foods include lax quality control of production process
(such as poor sealing, failed sterilization, and improper cleaning process of equipment), low acidity of
canned fruit, prolonged storage of opened food, large temperature fluctuations, and microbial
contamination of raw and auxiliary materials (China Canned Food Industry Association, 2022). Meat is
rich in nutrients, such as sugar, protein and water, which are needed for microbial growth (Nan,2004).
Canned meat foods are especially vulnerable for bacterial contamination. Our product can be used to
detect EC, VP, SA and SE in canned meat foods.
Figure 13 A meat can (Ll1324, CC0, via Wikimedia Commons)
To apply our tool for detection of pathogens in canned foods, a food sample is added into a tube with
buffer. RBP-MBs are added. After 30-minute incubation at 37℃, magnets are applied outside the bottom of
the tube for 30s to isolate the target bacteria. The supernatant is removed and the purified viable
target bacteria are resuspended. A drop of the resuspension solution is added on to a paper-based sensor
and a visual signal can be detected by naked eye. The purified bacteria can also be used for further
tests of various kinds if necessary.
Compared with other methods, our product has several advantages. First of all, it is very accurate to
detect the target bacteria strains because of the high selectivity of bacteriophage RBPs. It is also a
rapid method with the whole process in less than 1 hour (Li, 2010). Moreover, it is very cheap and
portable since the tool only uses tubes, magnets, reagents and a paper-based sensor. In addition, the
procedure is easy and convenient to perform, so no professional staff is required and the training for
its usage is simple. This tool can be used in all regions including remote areas with poor economic
condition and laboratory setup and by ordinary non-professional personnel. Therefore, our product can be
applied widely and cheaply for potentially contaminated food and helps to reduce the incidence of
foodborne diseases in all regions.
Reference
- WHO. Food safety. (2022). https://www.who.int/news-room/fact-sheets/detail/food-safety
- Wei L, Wu HJ, Li BM, et al. Pollution of four food borne pathogens and research progress of new
detection techniques. Food Science. 2011; 32 (19): 302-6.
- Tong SY, Davis JS, Eichenberger E, et al. Staphylococcus aureus infection: epidemiology,
pathophysiology, clinical manifestations and management. Clint Microbiological Research Revised
Edition. 2015; 28(3):603-61.
- CDC/ Matthew J. Arduino, DRPHPhoto Credit:Janice Haney Carr, Public domain, via Wikimedia Commons.
https://commons.wikimedia.org/wiki/File:Staphylococcus_aureus_VISA_2.jpg
- Ghenem L, Elhadi N, Alzahrani F, et al. Vibrio parahotlyticus: a review of distribution,
pathogenesis, virulence determinants, and epidemiology. Saudi J Medical Science. 2017; 5(2):93-103.
- Pazhani GP, Chowdhury G, Ramamurthy T. adaptation of Vibrio parahaemolyticus to stress during
environmental survival, host colonization, and infection. Anterior Microorganisms. 2021; 12: 737299.
- Wu YN, Wen J, Ma Y, et al. Epidemiology of foodborne disease outbreaks caused by Vibrio
parahaemolyticus, China, 2003–2008. Food Control; 2014; 46:197-202.
- Li Y, Xie T, Pang R, et al. Prevalence, antibiotic sensitivity and genetic characteristics of
foodborne Vibrio parahaemolyticus in China. Anterior Microorganisms. 2020; 11:1670.
- Sathiyamoorthi E, Faleye OS, Lee JH, et al. Antibacterial and Antibiofilm Activities of
Chloroindoles Against Vibrio parahaemolyticus.Frontiers in Microbiology. 2021;12: 714371.
- Shah DH, Elder JR, Chiok KL, et al. Genetic basis of salmonella enteritidis pathogenesis in
chickens. A chapter in the book of Producing Safe Eggs. 2017
- Coveny T. The five most common Salmonella strains. 2022.
- China National Center for Food Safety Risk Assessment (CFSA). [World Food Safety Day] Foodborne
pathogens that cannot be ignored: Salmonella. 2020.
https://www.cfsa.net.cn/Article/News.aspx?id=A9955B2F6251D76CDCAAF69390A1984B2DD4C40A9FEB65AA
- Lin ZY, Pediatric medical team of Taiwan Chang Gung Memorial Hospital, Zhou Y. Encyclopedia of
Chinese parenting. China: Beijing United Publishing Company. 2016: 365.
- News Desk. Nine multistate salmonella outbreaks in 2021 sickened more than 1,200. January, 2022.
- Hebei Institute of Food Inspection, edited by the Productivity Promotion Center of Food Industry.
Development and quality safety of puffed food. China: China Economic Publishing House. 2016: 86.
- Photo by Jean Guard, ARS. U.S. Department of Agriculture, via Wikimedia Commons.
https://commons.wikimedia.org/wiki/File:Salmonella_enteritidis_in_color.jpg
- Adzitey F, Huda N, Ali GRR. Molecular techniques for detecting and typing of bacteria, advantages
and application to foodborne pathogens isolated from ducks. 3 Biotech. 2013; 3(2):97–107.
- Salman HA, Abdulmohsen AM, Falih MN, Romi ZM (2021) Detection of multidrug-resistant Salmonella
enterica subsp. enterica serovar Typhi isolated from Iraqi subjects, Veterinary World, 14(7):
1922-1928.
- Umesha S, Manukumar HM. Advanced molecular diagnostic techniques for detection of food-borne
pathogens: Current applications and future challenges. Critical Reviews in Food Science and
Nutrition. 2018; 58:1, 84-104.
- Oyedeji AB, Green E, Adebiyi JA, et al. Metabolomic approaches for the determination of
metabolites from pathogenic microorganisms: A review. Food Research International. 2021; 140:
110042.
- Sun Z, Guo J, Wan WB, et al. A system of rapidly detecting Escherichia coli in food based on a
nanoprobe and improved ATP bioluminescence technology. Nanomaterials. 2022; 12(14): 2417.
- Witte S, Zinsli LV, Gonzalez-Serrano R, et al. Structural and functional characterization of the
receptor binding proteins of Escherichia coli O157 phages EP75 and EP335. Comput Struct Biotechnol
J. 2021;19:3416-26.
- Liang X, Wang Y, Hong B, et al. Isolation and characterization of a lytic vibrio
parahaemolyticus phage vB_VpaP_GHSM17 from sewage samples. Viruses. 2022; 14(8): 1601.
- Stone E, Campbell K, Grant I, et al. Understanding and exploiting phage-host interactions.
Viruses. 2019; 11(6):567.
- Wang Y, He Y, Bhattacharyya S, et al. Recombinant bacteriophage cell-binding domain proteins for
broad-spectrum recognition of methicillin-resistant staphylococcus aureus strains. Anal Chem. 2020;
92:3340−5.
- Seul A, M¨uller JJ, Andres D, et al. Bacteriophage P22 tailspike: structure of the complete
protein and function of the interdomain linker. Acta Cryst. (2014). D70, 1336–1345
- Wang C, Tu J, Liu J, et al. Structural dynamics of bacteriophage P22 infection initiation
revealed by cryo-electron tomography. Nature Microbiology, (2019),
https://doi.org/10.1038/s41564-019-0403-z.
- Lu Y, Wang Y, Wang J, et al. Phage Endolysin LysP108 Showed Promising Antibacterial Potential
Against Methicillin-resistant Staphylococcus aureus. Front Cell Infect Microbiol. 2021;11.
- Dolai S, Tabib‐Azar M. Whole virus detection using aptamers and paper-based sensor potentiometry.
Med Devices Sensors. 2020; 3(6): e10112.
- Marques A, Santos L, Costa M, et al. Office paper platform for bioelectrochromic detection of
electrochemically active bacteria using tungsten trioxide nanoprobes. Scientific reports. 2015; 5.
9910. 10.1038/srep09910.
- China Canned Food Industry Association. Risk factors analysis of canned food production safety.
2022. http://www.topcanchina.com/hyzx/13675
- Nan Q, Dai RT. Microbial safety of meat products in China. Meat Res. 2004; 4:33-5.
- Li M, Wang JX, Lin H. Advances in bacteriophage detection of foodborne pathogens. Food Science.
2010; 31: 23 439.
- figure 1 E. coli and Foodborne Illness Information for the Public, FDA Actions, and
Recommendations https://www.fda.gov/news-events/public-health-focus/e-coli-and-foodborne-illness)