A study analyzing 60 samples of vegetables obtained from local markets in China, including cabbage, cucumber, cauliflower, leek, and other commonly consumed vegetables, found that a 33 percent of the samples contained detectable levels of antibiotics (Food Analytical Methods 2018;11:2857–2864). The vegetables are likely to have absorbed the antibiotics from soil contaminated by antibiotics.
Antibiotics are still routinely added to animal feed to prevent or treat microbial infections, as well as promote animal growth in livestock production. Most (50 to 90 percent) antibiotics and their primary metabolites are rapidly excreted and ultimately end up in sewage and manure. Some of this is then spread on agricultural fields as fertilizer for growing crops. Vegetables elsewhere, including corn, potatoes, and lettuce, have also been found to contain antibiotic residues. Worryingly, there are currently no regulations to check and monitor for antibiotics in food products. Moreover, antibiotics have been detected in groundwater leading to concerns over their entry into food chain. Antibiotic residue levels should be monitored in fertilizer, the soil, and vegetables for risk assessment and control (Environ Pollut. 2006;143:565–571, Scientific American. January 2006).
Despite efforts to curtail the use of antibiotics in the era of antibiotic-resistant microorganisms, antibiotics are still widely used to treat human and animal diseases. Antibiotic resistance poses a global threat to public health; antibiotic resistance is responsible for 25,000 annual deaths in the European Union and 23,000 annual deaths in the U.S. There are numerous causes of antibiotic resistance, including over-prescribing, patients not taking antibiotics as prescribed, poor infection control in hospitals, poor hygiene and sanitation practices, lack of rapid laboratory tests, and unnecessary antibiotic use in agriculture.
The analysis to detect the antibiotics in the vegetables used a novel highly sensitive method devised to detect 49 target antibiotics, which fall into different classes, including sulfonamides, quinolones, macrolides, beta-lactams, and tetracyclines. Of these 49 antibiotics, five were most commonly detected across 20 samples: oxytetracycline, doxycycline, sulfamethoxazole, enrofloxacin, and chlortetracycline.
The highest concentration was of oxytetracycline in cabbage, found to be 126 μg/kg and roughly 1% of the usual daily dose (1000 mg) for an adult. While this does not sound like much, it could become substantial if exposure is chronic. Oxytetracycline is a broad-spectrum antibiotic and is associated with gastrointestinal and skin-sensitivity side effects. It is contraindicated in pregnancy because it can cross the placenta and may have toxic effects on fetal tissues (Natl Health Stat Report. 2018;122:1–16). Although lower compared with the oxytetracycline, doxycycline, sulfamethoxazole, enrofloxacin, and chlortetracycline were also detected, at concentrations ranging between 2.0 and 12.8 μg/kg in the vegetables (Food Analytical Methods 2018;11:2857–2864).
Method for Detecting Antibiotics in Vegetables
The method used to detect and identify this wide range of antibiotics in vegetable samples is a relatively new one, involving the quick, easy, cheap, effective, rugged, and safe (so-called QuEChERS) procedure to prepare the sample for liquid chromatography and mass spectroscopic analysis using SCIEX ExionLC and QTRAP 4500 systems (Food Analytical Methods 2018;11:2857–2864). The QuEChERS technique is a simple, rapid, and cost-efficient method of extracting and preparing the sample for liquid chromatography tandem mass spectrometry (LC-MS/MS) (Anal Chem. 2012;84(13):5677–5684). It requires less time and solvent than other methods to detect antibiotics, including solid-phase extraction (SPE) after ultrasonic, vortex, or vibration extraction. For the LC-MS/MS analysis of multiple antibiotic residues in different vegetable samples, the extraction timing and buffer system, dispersive solid-phase extraction (d-SPE) clean-up, and other parameters, such as those controlling for matrix effects, were also optimized (see Figure 1).
Along with the improved extraction procedure, the research team also optimized the LC-MS/MS technique. It is common practice to use LC to separate out the analytes in the sample, and then transfer them into a triple quadrupole-based mass spectrometer (triple-quad) to further separate and scan the discrete analytes using a multiple reaction monitoring (MRM). However, using the triple-quad approach to detect and identify multiclass antibiotics can result in type I errors (false positives) due to interferences that have MRM transition signatures that coincide with those of the antibiotics. Type II errors (false negatives) may also occur, should the antibiotic analyte be present at a very low concentration, thus producing a weak response in the second transition (Food Analytical Methods 2018;11:2857–2864; Anal Chem. 2012;84(13):5677–5684). Therefore, the team used a quadrupole linear ion trap mass spectrometer, which combines the rapid, multiple scanning functionality of a triple-quad with the sensitivity of a linear ion trap mass spectrometer (Food Analytical Methods 2018;11:2857–2864; Anal Chem. 2007;79(24):9372–9384). With such an advanced hybrid system, the SCIEX QTRAP 4500, coupled with the SCIEX ExionLC ultra-high performance LC system, the team were able to develop and validate their method to simply and reliably detect and identify multiple antibiotic residues from different classes (Food Analytical Methods 2018;11:2857–2864).
The method was validated by analyzing 17 sulfonamides, 16 quinolones, 6 macrolides, 5 beta-lactams, and 5 tetracyclines, with 7 isotope-labelled internal standards for all the antibiotic classes tested. The QuEChERS-based LC-MS/MS method was confirmed to be highly accurate and precise with recoveries of 70–100 percent and reproducibility of less than 20 percent for relative standard deviation (RSD) for most of the sulfonamide, macrolide, beta-lactam, and tetracycline antibiotics. Although they are still considered acceptable at higher than the SANTE/11813/2017 guideline standard of 30 percent, the recoveries of the quinolones were lower than those of the other antibiotic classes in different vegetables. However, this was not unexpected as similar findings have been reported with both SPE and QuEChERS methods (Food Analytical Methods 2018;11:2857–2864). The reproducibility and thus, precision was especially good for the analyses of the macrolide and beta-lactam antibiotic residues, with RSDs that were lower than the other antibiotic classes, particularly at low concentrations of 5 μg/kg. The limit of quantification (LOQ) was 2 μg/kg for most (~74 percent) of the antibiotics tested, and 5 μg/kg for the remaining (~26 percent) residues. The method is accurate for a wide range of concentrations, with the linearity range being 1–200 μg/L. The coefficient of determination (r2) was the requisite value higher than 0.995 for each residue; which guarantees the accurate quantification of each of the 49 antibiotics through the application of this method (Food Analytical Methods 2018;11:2857–2864).
To confirm the accuracy of the qualitative results, the MS/MS spectra of the putative antibiotic residues in the positive samples were compared with the spectra of known target analytes housed in a reference library. This helped disqualify type I errors and confirm true positives. This final step, was facilitated by the simultaneous acquisition of the MRM scan data alongside the full scan MS/MS spectra in enhanced product ion (EPI) mode using information-dependent acquisition (IDA), which was uniquely possible with the use of the SCIEX QTRAP instruments. This final confirmatory step helps validate the utility and reliability of this method (Food Analytical Methods 2018;11:2857–2864).
Fulfilling a Need
According to research, antibiotic resistance may cause 10 million deaths annually by 2050 (PLOS Medicine. 2016;13(11):e1002184). The startling figures show that greater efforts need to be made to eliminate the injudicious application of antibiotics. Moreover, further research and understanding of the presence of antibiotics in the environment is required since antibiotics can leach from the soil into aquifers or groundwater due to run-off. All organisms—human, animal, or vegetable—are therefore susceptible to being exposed unnecessarily and unknowingly to antibiotics. As such, they can unwittingly contribute to the development of antibiotic-resistant bacteria and other microbes (Scientific American. January 2006).
Not only is there a need for better standards and regulation, there is also a need for tools such as the method described here to allow scientists, regulators, farmers, retailers and even consumers to identify antibiotics in their food. A united effort needs to be made to protect our environment as well as human and animal health, while maintaining food safety. This could include the exploration of other ways to combat bacterial infection, using innovative new technologies such as clustered regularly interspaced short palindromic repeats (CRISPR) and the development of precision medicines (Nature Medicine. 2019;25:730–733). The development of our methodology, using QuEChERS and LC-MS/MS, is just one tool in the arsenal in the fight against antibiotic resistance.
Professor Liu is executive deputy director of the Agro-environmental Quality Supervision and Testing Center at the Agro-Environmental Protection Institute (AEPI), Ministry of Agriculture and Rural Affairs, in China.