English

• Pathogen monitoring and control is a critical step in ensuring food safety and the health of the general public. Over the years, there have been many threats to human health from bacterial infections such as Escherichia coli (E. coli), leading to a surge of ingenious strategies for detecting microorganisms in connection with food safety, medical diagnostics, water quality, and bioterrorism [1–4]. Commonly, the conventional method for detecting pathogenic bacteria is primarily based on bacteria culture, which is sensitive and inexpensive but bears the limitation of time-consuming and labor-intensive procedures [5]. Immunological assay such as enzyme linked immunosorbent assay (ELISA) is a rapid method but suffers from relatively low sensitivity [6–9]. Nucleic acid-based assays such as polymerase chain reaction (PCR) have shown advantages of high sensitivity and high specificity for pathogenic bacteria detection but suffering expensive equipment, complicated procedures, and the need for skillful technicians [10,11]. Hence, it is urgent to develop a novel, simple, inexpensive and time saving approach to determine bacterial with high selectivity and sensitivity.

  Fluorescence resonance energy transfer (FRET) is a distance-dependent energy transfer phenomenon (from a donor in excited energy state to a nearby acceptor) which has been exploring to design various sensors [12–14]. In recent years, nanomaterials derived energy donors and acceptors have been employed to construct FRET optical sensors for bacterial detection [15–23]. Wang et al. used antibody functionalized two-color quantum dots (QDs) as donor and acceptor to detect Salmonella enteritidis in eggshell [24]. Heli et al. developed a FRET immunosensor based on antibody functionalized QDs and gold nanoparticles (AuNPs) for sensitive identifying of E. coli [25]. However, the inherent defects of antibody, including low screening efficiency and low stability, make the antibody-based strategies limited for application [19,20,24–26].

  As a group of proteins, lectins can strongly bind to specific carbohydrate moieties on the surface of bacteria [27,28] and are particularly interesting candidates for using as molecular recognition elements because of their ease of production and intrinsic stability [29,30]. Concanvalin A (Con A), a type of legume lectin, could specifically bind with terminal α-D-mannosyl and α-D-glucosyl groups on the surface of bacteria and has been employing as recognition molecule to develop sensors for detecting pathogen [29–34]. However, the sensitivity of current Con A based methods is not very satisfied, and moreover the methods involve complex operation and time-consuming. Aptamers are single-stranded DNA or RNA oligonucleotides that can recognize and bind to specific targets, including whole cells, proteins, peptides and small molecules [35], and have been attracting great attention in the fields of biosensor due to their unique characteristics such as ease of synthesis and modification, good reproducibility [36,37].

  In this study, a new FRET sensor based on aptamer-modified quantum dots (Aptamer-QDs) as energy donor and Con A-modified gold nanoparticles (Con A-AuNPs) as energy receptor was constructed to detect bacteria sensitively. As a great threat to human health, E. coli needed to be strictly monitored and employed as the testing model in this study. The Con A and aptamer molecules exhibit broad-spectrum and high-specific binding on the surface of E. coli, respectively, providing an opportunity to construct the FRET sensor on the surface of bacteria. As shown in Scheme 1, in the presence of target of E. coli, the aptamer and Con A bound to E. coli simultaneously, making the energy donor (Aptamer-QDs) and acceptor (Con A-AuNPs) dramatically closed to each other and subsequently the FRET "on", thus achieving the detection of E. coli. Moreover, based on the proposed FRET sensor, E. coli in milk and orange juice could be detected sensitively and rapidly in one-step. The mode of the FRET sensor is also suitable for the detection of other bacteria, which is of great significance for the development of bacterial detection methods.

  Scheme 1

  

  Scheme 1.  Illustration of dual-recognition FRET sensor for detecting E. coli based on Aptamer-QDs and Con A-AuNPs.

  DownLoad: Full-Size Img PowerPoint

  Con A-AuNPs were prepared through a one pot process via reduction of HAuCl₄ by NaBH₄ in the presence of Con A which served as stabilizer. UV-vis spectra revealed the maximum absorption peak of Con A-AuNPs at 518.5 nm (Fig. 1A). The diameter of spherical Con A-AuNPs was around 8.7 nm as indicated by TEM images and statistical analysis (Fig. 1B). The zeta potential of Con A-AuNPs was about −43.4 mV ± 0.6 mV (Fig. 1C), providing enough surface charge to stabilize the nanoprobes. FTIR spectra showed that Con A-AuNPs had the similar infrared characteristic spectrum as Con A. Specifically, the stretching vibration of the groups of N-H, O-H and C-O showed typical absorption peaks at 3467, 1639, 1399 cm⁻¹ (Fig. 1D), which means Con A was involved in the particles of Con A-AuNPs. Then the binding affinity of the generated Con A-AuNPs toward bacteria was further examined by incubating Con A-AuNPs with E. coli (B), S. typhimurium (C), P. aeruginosa (D), V. parahemolyticus (E), S. enteritidis (F), respectively for 0.5 h, following subjected to centrifugation at 1100 g for 10 min. As shown in Fig. S1 (Supporting information), red precipitates were observed at the bottom of the tube except the tube containing the Con A-AuNPs alone after centrifugation (top row) while white precipitates were observed in the tube containing bacteria alone after centrifugation (bottom row), suggesting the binding of Con A-AuNPs with bacteria. The complex of Aptamer-QDs could be easily obtained according the published procedures [38], and the aptamer sequence information shown in Table S1 in Supporting Information. As shown in Fig. 1E, Aptamer-QDs in agarose gel shifted fast compared with streptavidin-QDs due to the couple of negative charged of DNA aptamer onto streptavidin-QDs. The absorbance spectrum of Con A-AuNPs showed spectral overlap with the emission spectrum of Aptamer-QDs (Fig. 1F), suggesting that the Aptamer-QDs and Con A-AuNPs could serve as energy donor-acceptor pair in FRET sensor.

  Figure 1

  

  Figure 1.  (A) The absorption spectrum (the inset is a digital photo of Con A-AuNPs); (B) TEM photo, (C) zeta potential, and (D) FTIR of Con A-AuNPs; (E) Digital image of agarose gel electrophoresis; (F) The typical emission spectrum of Aptamer-QDs and the absorbance spectrum of Con A-AuNPs.

  DownLoad: Full-Size Img PowerPoint

  The feasibility of Aptamer-QDs and Con A-AuNPs based FRET strategy for E. coli detection was first investigated by measuring the fluorescence response after they incubated with the target E. coli. As shown in Fig. 2A, the fluorescence of the QDs was reduced in the presence of the target bacteria E. coli with fluorescence quenching efficiency about 26.50% (the fluorescence quenching efficiency of η was calculated by the equation η = (F₀ − F)/F₀ × 100%, where F₀ and F were the fluorescence intensity of the FRET sensor system in the absence and presence of E. coli or other bacteria, respectively). However, when there were only non-target bacteria such as S. aureus or S. typhimurium, or when the aptamer was replaced with a random DNA sequence (RanSeq) which could not recognize E. coli, only a slight fluorescence intensity decrease was observed. The η for all the bacteria samples were summarized with a bar graph for more clear understanding (the inset in Fig. 2A). Moreover, the fluorescence emission of Aptamer-QDs alone, Aptamer-QDs with E. coli, Aptamer-QDs with S. aureus, and Aptamer-QDs with S. typhimurium were measured, respectively. As shown in Fig. S2 (Supporting information), after incubating with E. coli, S. aureus, and S. typhimurium, the fluorescence intensity of Aptamer-QDs did not change compared with that of Aptamer-QDs alone, indicating that the fluorescence intensity quenching of Aptamer-QDs was not caused by bacteria but the FRET occurred between the Aptamer-QDs and Con A-AuNPs once they bound onto the surface of E. coli. Meanwhile, the fluorescence quenching phenomenon for the target E. coli was visually observed under a fluorescence microscope. As shown in Fig. 2B, bright green fluorescence was observed around E. coli (a, b) after E. coli were treated only with Aptamer-QDs while no obvious fluorescence around E. coli (c, d) was observed when treated only with RanSeq-QDs, indicating the specific recognition other than non-specific binding of Aptamer-QD to E. coli. Interestingly, the green fluorescence around E. coli (e, f) was dramatically subdued when E. coli treated with Aptamer-QDs/Con A-AuNPs together, suggesting the occurrence of FRET between Aptamer-QDs and Con A-AuNPs.

  Figure 2

  

  Figure 2.  (A) Fluorescence spectra of different nanoprobes before and after incubation with bacteria at excitation wavelength of 350 nm (emission wavelength range from 450 nm to 650 nm), The inset is a bar graph of the η for different samples. (B) Fluorescence microscope images of E. coli after treated with Aptamer-QDs (a, b), RanSeq-QDs (c, d), Aptamer-QDs/Con A-AuNPs (e, f).

  DownLoad: Full-Size Img PowerPoint

  To obtain the best assay performance, the dosages of Aptamer-QDs and Con A-AuNPs and the incubation time were optimized through the orthogonal experiment method employing the L9(3³) orthogonal layout, in which fluorescence quenching efficiency was used as the evaluating index. Each of three major factors (dosages of Aptamer-QDs and Con A-AuNPs and the incubation time) was studied at three levels (Table S2 in Supporting information). The orthogonal experiment program, results, and analysis were shown in detail in Tables S3 and S4 (Supporting information). As shown in Table S4, the value of η reached a maximum when the dosages of Aptamer-QDs and Con A-AuNPs were 4 nmol/L and 30 nmol/L, respectively, and the optimal reaction time was 0.5 h.

  After optimizing the assay conditions, the linear range and detection limit of the proposed strategy for E. coli detection were investigated. Specifically, a series of E. coli solutions with different concentrations (0, 10², 10³, 10⁴, 10⁵, 10⁶, 10⁷, 5 × 10⁷, 10⁸ and 2 × 10⁸ cfu/mL, respectively) were incubated with Aptamer-QDs and Con A-AuNPs simultaneously for 0.5 h, followed by the determination of the fluorescence intensity of each sample. As shown in Fig. 3, the η of the FRET sensor was linearly dependent upon the logarithm of the concentration of E. coli from 10² cfu/mL to 2 × 10⁸ cfu/mL, with limit of detection (LOD) 45 cfu/mL (η = 0.05498log₁₀N − 0.09095, R = 0.9920, N stands for the quantity of E. coli in cfu/mL); LOD was determined by the equation LOD = 3 S/K, where S was the standard deviation of the blank samples (n = 10) and K was the slope of the calibration curve).

  Figure 3

  

  Figure 3.  The calibration curve of the η on the logarithm of E. coli concentration (log₁₀N, cfu/mL).

  DownLoad: Full-Size Img PowerPoint

  In nature, E. coli may be usually coexist with other bacteria, including both Gram-positive (G⁺) and Gram-negative (G⁻) bacteria; therefore, the selectivity of the FRET sensor for E. coli should be verified. The selectivity of the FRET sensor for E. coli was tested by comparing the η of FRET sensor for detection of the E. coli and other interfering bacteria. Each of the four interfering bacteria (S. typhimurium (G⁻), P. aeruginosa (G⁻), L. monocytogenes (G⁺) and S. aureus (G⁺)), a mixture 1 of the four interfering bacteria together, and another mixture 2 (containing E. coli and four interfering bacteria) were assayed with the proposed FRET sensor. As shown in Fig. 4, the fluorescence intensity changes for the groups of interfering bacteria (individual interfering bacterium or the mixture 1) were almost negligible. The fluorescence intensity change of mixture 2 was almost the same as that of the E. coli. The above results indicate that the proposed FRET sensor has good selectivity for the detection of E. coli, which could be attributed to the specific recognition of aptamer to E. coli.

  Figure 4

  

  Figure 4.  The η of FRET sensor for detecting E. coli, interfering bacteria, and the mixtures of them, respectively.

  DownLoad: Full-Size Img PowerPoint

  In order to demonstrate the applicability of the proposed FERT strategy for E. coli detection in real samples, two test samples (milk and orange juice) spiked with different concentrations of E. coli were analyzed. The method was performed after the samples were properly diluted without other pretreatment, and the interference of the real samples (the background signals produced by the pure real sample without spiking any E. coli) to the FRET-based sensor was also explored. Ten-fold dilution of milk and orange juice produced comparable background signals with that of binding buffer (blank) (Fig. 5A). The 10-fold dilution real samples spiked with E. coli were subjected to analysis with the constructed FRET sensor. The recoveries varied from 83.1% to 112.5%, with relative standard derivations (RSDs) in the range of 0.05%–0.50% (Table 1), indicating that the proposed FRET strategy could be applied for detection of E. coli in real samples. The LODs of the FRET sensor for E. coli in milk and orange juice sample were 300 cfu/mL and 200 cfu/mL, respectively (Fig. 5B), suggesting that the sensor has a good application for the detection of E. coli in real samples. Meanwhile, the reported methods for E. coli assay plus with the proposed method in this work were summarized and made a simple comparison (Table S5 in Supporting information). Compared with these methods reported, the as-proposed strategy in this study is more sensitivity and time-saving for detection of E. coli.

  Figure 5

  

  Figure 5.  (A) Fluorescence response (ΔF) of FRET sensor in different sample with different-fold dilution (ΔF = F₀ - F_a, where F₀ and F_a are the fluorescence intensity of the FRET sensor in binding buffer and different sample with different-fold dilution, respectively). (B) The η of FRET sensor for E. coli detection in different sample with different concentrations of E. coli.

  DownLoad: Full-Size Img PowerPoint

  Table 1

  

  Table 1.  Recovery efficiency of the proposed strategy for E. coli detection in real samples.

  DownLoad: CSV

  In this study, a rapid and facile of FRET strategy was developed for the detection of E. coli. With this strategy, pathogen of E. coli could be specifically detected within 0.5 h over a range from 10² cfu/mL to 2 × 10⁸ cfu/mL with a detection limit of 45 cfu/mL in one step. E. coli spiked in authentic samples was also quantified with good recoveries. Since Con A is a broad-spectrum recognition molecule for most bacteria, this work demonstrates a universal platform for pathogenic bacteria assay by simply changing the aptamer to detect different target bacteria.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (Nos. 22174116, 21974110), and Chongqing Science Funds for Distinguished Young Scientists (No. cstc2021jcyj-jqx0024), the Innovation Research Group at higher Education Institutions in Chongqing, Chongqing Education Committee (No. CXQT21006).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.108102.

  
  
• 1. [1]

     J. Zhou, L.J. Yin, Y.N. Dong, et al., Anal. Chim. Acta 1127 (2020) 225-233. doi: 10.1016/j.aca.2020.06.041

  2. [2]

     M.J. Hajipour, A.A. Saei, E.D. Walker, et al., Adv. Sci. 8 (2021) 2100556. doi: 10.1002/advs.202100556

  3. [3]

     L. Varadi, J.L. Luo, D.E. Hibbs, et al., Chem. Soc. Rev. 46 (2017) 4818-4832. doi: 10.1039/C6CS00693K

  4. [4]

     A.L. Furst, M.B. Francis, Chem. Rev. 119 (2019) 700-726. doi: 10.1021/acs.chemrev.8b00381

  5. [5]

     Y. Wang, J.K. Salazar, Compr. Rev. Food Sci. Food Saf. 15 (2016) 183-205. doi: 10.1111/1541-4337.12175

  6. [6]

     B. Pang, C. Zhao, L. Li, et al., Anal. Biochem. 542 (2018) 58-62. doi: 10.1016/j.ab.2017.11.010

  7. [7]

     S. Umesha, H.M. Manukumar, Crit. Rev. Food Sci. Nutr. 58 (2018) 84-104. doi: 10.1080/10408398.2015.1126701

  8. [8]

     T.X. Wei, D. Du, M.J. Zhu, et al., ACS Appl. Mater. Interfaces 8 (2016) 6329-6335. doi: 10.1021/acsami.5b11834

  9. [9]

     I.H. Cho, J. Irudayaraj, Int. J. Food Microbiol. 164 (2013) 70-75. doi: 10.1016/j.ijfoodmicro.2013.02.025

  10. [10]

      A. Garrido, M.J. Chapela, M. Ferreira, et al., Food Control 24 (2012) 128-135. doi: 10.1016/j.foodcont.2011.09.015

  11. [11]

      J.D. Barak, K. Sananikone, M.J. Delwiche, Lett. Appl. Microbiol. 41 (2005) 112-118. doi: 10.1111/j.1472-765X.2005.01746.x

  12. [12]

      L.L. Wu, C.S. Huang, B.P. Emery, et al., Chem. Soc. Rev. 49 (2020) 5110-5139. doi: 10.1039/C9CS00318E

  13. [13]

      C.H. Zhang, P.J. Ni, B. Wang, Chin. Chem. Lett. 33 (2022) 757-761. doi: 10.1016/j.cclet.2021.08.017

  14. [14]

      Y. Xian, M.X. Zhou, S.C. Han, Chin. Chem. Lett. 31 (2020) 468-472. doi: 10.1016/j.cclet.2019.08.012

  15. [15]

      J.Y. Shi, C.Y. Chan, Y.T. Pang, et al., Biosens. Bioelectron. 67 (2015) 595-600. doi: 10.1016/j.bios.2014.09.059

  16. [16]

      B.R. Jin, S.R. Wang, M. Lin, et al., Biosens. Bioelectron. 90 (2017) 525-533. doi: 10.1016/j.bios.2016.10.029

  17. [17]

      N. Duan, S.J. Wu, S.L. Dai, et al., Microchim. Acta 182 (2015) 917-923. doi: 10.1007/s00604-014-1406-3

  18. [18]

      M.Q. Yu, H. Wang, F. Fu, et al., Anal. Chem. 89 (2017) 4085-4090. doi: 10.1021/acs.analchem.6b04958

  19. [19]

      Z. Dehghani, J. Mohammadnejad, M. Hosseini, B. bakhshi, A.H. Rezayan, Food Chem. 309 (2020) 125690. doi: 10.1016/j.foodchem.2019.125690

  20. [20]

      E.J. Kim, E.B. Kim, S.W. Lee, et al., Biosens. Bioelectron. 87 (2017) 150-156. doi: 10.1016/j.bios.2016.08.034

  21. [21]

      F. Fu, Y.Q. Zhang, L.Y. Li, et al., Anal. Chem. 92 (2020) 11462-11468. doi: 10.1021/acs.analchem.0c02695

  22. [22]

      X.Q. Tao, Z.Y. Liao, Y.Q. Zhang, Chin. Chem. Lett. 32 (2021) 791-795. doi: 10.1016/j.cclet.2020.07.020

  23. [23]

      L.S. Liang, M. Chen, Y.L. Tong, W.G. Tan, Z.G. Chen, Anal. Chim. Acta 1186 (2021) 339090. doi: 10.1016/j.aca.2021.339090

  24. [24]

      B.B. Wang, Q. Wang, Y.G. Jin, et al., J. Photochem. Photobiol. A 299 (2015) 131-137. doi: 10.1016/j.jphotochem.2014.10.020

  25. [25]

      B. Heli, A. Ajji, Sci. Rep. 10 (2020) 14367. doi: 10.1038/s41598-020-71285-3

  26. [26]

      E. Heyduk, T. Heyduk, Anal. Biochem. 396 (2010) 298-303. doi: 10.1016/j.ab.2009.09.039

  27. [27]

      J. Hirabayashi, J. Biochem. 144 (2008) 139-147. doi: 10.1093/jb/mvn043

  28. [28]

      J.Q. Gao, D.J. Liu, Z.X. Wang, Anal. Chem. 82 (2010) 9240-9247. doi: 10.1021/ac1022309

  29. [29]

      M. Gamella, S. Campuzano, C. Parrado, et al., Talanta 78 (2009) 1303-1309. doi: 10.1016/j.talanta.2009.01.059

  30. [30]

      Z.M. Li, Y.C. Fu, W.H. Fang, Y.B. Li, Sensors 15 (2015) 19212-19224. doi: 10.3390/s150819212

  31. [31]

      H. Kearns, R. Goodacre, L.E. Jamieson, et al., Anal. Chem. 89 (2017) 12666-12673. doi: 10.1021/acs.analchem.7b02653

  32. [32]

      X. Liu, Z. Lei, F.Y. Liu, et al., Biosens. Bioelectron. 58 (2014) 92-100. doi: 10.1016/j.bios.2014.02.056

  33. [33]

      L.B. Zheng, Y. Wan, P. Qi, et al., Talanta 167 (2017) 600-606. doi: 10.1016/j.talanta.2017.03.007

  34. [34]

      H.Y. Yang, H.F. Zhou, H.Y. Ha, et al., Sens. Actuators B: Chem. 229 (2016) 297-304. doi: 10.1016/j.snb.2015.08.034

  35. [35]

      E.J. Cho, J.W. Lee, A.D. Ellington, Annu. Rev. Anal. Chem. 2 (2009) 241-264. doi: 10.1146/annurev.anchem.1.031207.112851

  36. [36]

      A.B. Iliuk, L.H. Hu, W.A. Tao, Anal. Chem. 83 (2011) 4440-4452. doi: 10.1021/ac201057w

  37. [37]

      Y. Zhang, X.Z. Hu, Q.J. Wang, Y. Zhang, Chin. Chem. Lett. 33 (2022) 2817-2831. doi: 10.1016/j.cclet.2021.11.033

  38. [38]

      H.P. Huang, J.J. Zhu, Biosens. Bioelectron. 25 (2009) 927-930. doi: 10.1016/j.bios.2009.08.008

• 

  Scheme 1  Illustration of dual-recognition FRET sensor for detecting E. coli based on Aptamer-QDs and Con A-AuNPs.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  (A) The absorption spectrum (the inset is a digital photo of Con A-AuNPs); (B) TEM photo, (C) zeta potential, and (D) FTIR of Con A-AuNPs; (E) Digital image of agarose gel electrophoresis; (F) The typical emission spectrum of Aptamer-QDs and the absorbance spectrum of Con A-AuNPs.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (A) Fluorescence spectra of different nanoprobes before and after incubation with bacteria at excitation wavelength of 350 nm (emission wavelength range from 450 nm to 650 nm), The inset is a bar graph of the η for different samples. (B) Fluorescence microscope images of E. coli after treated with Aptamer-QDs (a, b), RanSeq-QDs (c, d), Aptamer-QDs/Con A-AuNPs (e, f).

  下载: 全尺寸图片 幻灯片
  

  Figure 3  The calibration curve of the η on the logarithm of E. coli concentration (log₁₀N, cfu/mL).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  The η of FRET sensor for detecting E. coli, interfering bacteria, and the mixtures of them, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (A) Fluorescence response (ΔF) of FRET sensor in different sample with different-fold dilution (ΔF = F₀ - F_a, where F₀ and F_a are the fluorescence intensity of the FRET sensor in binding buffer and different sample with different-fold dilution, respectively). (B) The η of FRET sensor for E. coli detection in different sample with different concentrations of E. coli.

  下载: 全尺寸图片 幻灯片

  Table 1.  Recovery efficiency of the proposed strategy for E. coli detection in real samples.

  下载: 导出CSV
•