Introduction

Listeria monocytogenes is a Gram-positive bacterial foodborne pathogen [1–4] that belongs to the genus Listeria along with L. grayi, L. innocua, L. welshimeri, L. seeligeri, L. ivanovii and contains thirteen serotypes (1/2a, 1/2b, 1/2c, 3a, 3b, 3c, 4a, 4ab, 4b, 4c, 4d, 4e, and 7) [5, 6]. L. monocytogenes causes listeriosis in humans, and L. ivanovii mainly infects ruminants [7–9]. The prevalent serotypes of L. monocytogenes are 1/2a, 1/2b, and 4b, which account for >95% of Listeria spp. isolated from humans [10]. L. monocytogenes, which can grow at low temperatures (4 °C), has been detected in dairy products, raw meat, and seafood [11, 12]. The standard detection method of L. monocytogenes is culture-based and requires chromogenic and biochemical confirmation, which are time-consuming, labor-intensive, and complex [5, 11]. More rapid and effective detection methods are required in the food industry and medical field.

Immunoassays and biosensors based on antibody-antigen reactions represent rapid and powerful tools for the analyses of pathogens and related toxins [12–16]. Compared with PCR, immunoassays are cost-effective and do not rely on sophisticated instrument or trained technicians [17–20]. The enzyme linked immunosorbent assay (ELISA) for L. monocytogenes detection is time-consuming (3–4 h) and limited to the laboratory [21, 22]. In contrast, the lateral flow immunochromatographic (ICG) strip assay is fast (10 min), simple, and portable [23, 24]. The ICG strip developed by Shim and co-authors was based on monoclonal antibodies (mAbs) that had stronger affinity towards L. monocytogenes than towards other Listeria spp., but cross-reacted with L. innocua and L. ivanovii at higher concentrations [25]. Kim and co-authors developed an ICG strip based on commercial mAbs and improved sensitivity using immuno-magnetic separation (IMS) [26]; however, the cross-reactivity of this strip was not clear because only L. monocytogenes was evaluated. Several biosensors have greatly improved the sensitivity of L. monocytogenes detection [3, 26–32]. The majority of mAbs are either specific to a certain serotype or cross-react with non-virulent Listeria spp. [33–35], which limit the analyses of unknown samples. MAbs against all L. monocytogenes serotypes have been prepared; however, it is challenging to obtain paired mAbs, which are essential for immunoassays and biosensors. Therefore, the mAbs for detection of L. monocytogenes are still limited.

We prepared mAb 1C1, which specifically recognized all L. monocytogenes serotypes, using bovine serum albumin (BSA) conjugated with peptide PepD as the immunogen. MAb 1C1 was paired with mAb 10E7, which was prepared against P60 protein present in all Listeria spp. including L. monocytogenes. Based on the signals obtained with gold-labeled nanoparticles (AuNPs), an ICG strip assay was successfully established for the rapid and accurate detection of L. monocytogenes without cross reactivity with other bacteria including Listeria spp.

Material and methods

Reagents and instruments

Peptides were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China, www.sangon.com). 4-(N-Maleimidomethyl) cyclohexanecarboxylic acid N-hydroxysuccinimide ester (SMCC), BSA, goat anti-mouse IgG antibody, horse radish peroxidase (HRP), Freund’s adjuvant, chloroauric acid (HAuCl4), and sodium citrate were obtained from Sigma-Aldrich Co., LLC (St. Louis, MO, USA, www.sigmaaldrich.com). All the other chemicals (analytical grade) were acquired from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China, www.sinoreagent.com). The High-binding 96-well microplate was obtained from GuoSheng Bio-Engineering Co., Ltd. (Wu Xi, China, www.gsbio.cn), and the strip reader was obtained from Huaan Magnech Bio-Tech Company Corp (Beijing, China, www.magnech.com). The UV-Vis spectrophotometer (Evolution 60S) was supplied by Thermo Fisher Scientific (Miami, USA, www.thermofisher.com). A transmission electron microscope (TEM; JEM-2100, JEOL, Japan, www.jeol.co.jp/en/) was used to visualize and characterize the AuNPs.

Bacteria and growth conditions

The bacterial strains are listed in Table 1. Campilobacteri jejuni was cultured at 37 °C for 2–3 d in a micro-aerobic environment (4% O2, 10% CO2 and 86% N2) in a three-gas incubator (Binder CB210, Tuttlinger, Germany, www.binder-world.com). L. monocytogenes and other bacterial strains were cultured overnight at 37 °C in Brain-Heart Infusion broth (Oxoid, Basingstoke, UK, www.oxoid.com/UK/blue/).

Table 1 The bacterial strains in this work
Full size table

Preparation of mAbs specific to L. monocytogenes and Listeria spp.

A peptide of the L. monocytogenes P60 protein (PepD; QQQTAPKAPTE) was synthesized with cysteine on the N-terminal end [36] and conjugated to BSA with SMCC. Briefly, 20 mg BSA was dissolved in 0.1 M phosphate buffered saline (PBS, 100 mM phosphate buffer containing 150 mM NaCl, pH 7.2). SMCC (6 mg) was dissolved in 100 μL N,N-dimethylformamide, added dropwise to the BSA solution, and allowed to react at room temperature for 1 h under constant stirring. Excess SMCC was removed by performing ultrafiltration twice (Milipore-Amicon, cut off 3000, www.merckmillipore.com) at 7000×g for 25 min and the SMCC activated BSA on the hyperfiltration membrane was dissolved in 0.1 M PBS. Subsequently, 12.88 mg and 6.44 mg of the BSA solution were respectively mixed with 500 μL peptide (10 mg⋅mL−1, in 0.1 M PBS) and allowed to react for 12 h. The conjugates with different reaction ratios were dialyzed prior to protein electrophoresis (Bio-Rad, Shanghai, China, www.bio-rad.com) and both of them were respectively used as immunogens in BALB/c mice [37]. Following cell fusion, positive cell lines were selected against culture medium of different Listeria spp. (diluted 4 × for indirect ELISA) and confirmed with recombinant P60 protein of L. monocytogenes (0.3 μg⋅mL−1 for coating). MAb 10E7 against the whole P60 protein was produced in our laboratory. The purified mAb was conjugated to HRP with oxidation method by sodium periodate [38].

AuNPs and development of the ICG strip

The synthesis of 15 nm AuNPs was performed using a citrate reduction method [39]. HAuCl4 solution (0.0 1%, 100 mL) was boiled in a conical flask for 10 min under constant stirring. Trisodium citrate (10%, 200 μL) was quickly added, and the solution was boiled for another 15 min. Once the solution developed a red wine color, it was allowed to cool at room temperature and stored at 4 °C. The diameter of AuNPs was determined by TEM.

MAb 1C1 against L. monocytogenes was conjugated to AuNPs [23]. Briefly, the pH of the AuNP solution (20 mL) was adjusted to 7.5 with 110 μL of 0.1 M K2CO3. Subsequently, 67 μL of mAb (3 mg⋅mL−1) was added and allowed to react at room temperature for 2 h under gentle stirring. BSA (10% w/v, 1 mL) was added and incubated for 2 h under gentle stirring. The solution was washed twice with washing buffer (0.02 M PBS, 1% BSA, and 2% sucrose) and centrifuged at 6000×g for 30 min at 4 °C. AuNP-labeled mAbs were suspended in 200 μL of 0.02 M PBS containing 0.02% NaN3 and stored at 4 °C.

The ICG strip was developed by attaching an absorption pad, a nitrocellulose membrane (NC membrane), and a sample pad (JieYi Biotechnology Co., Ltd., Shanghai, China, www.joey-bio.cn) to a polyvinylchloride sheet (Fig. 1). MAb 10E7 (4 mg⋅mL−1) and goat anti-mouse IgG antibody (0.5 mg⋅mL−1) were respectively sprayed onto the NC membrane as test line (T line) and control line (C line) using a BioJet Quanti3000 dispenser (Kinbio Tech Co., Ltd., Shanghai, China, www.kinbio.com). The distance between the T line and C line was 10 mm. The membrane was air-dried in an oven (37 °C, 2 h) and cut into individual strips (4 mm wide) using a CM4000 Guillotine Cutting Module (Kinbio Tech Co., Ltd., Shanghai, China, www.kinbio.com). The strips were stored with a desiccant at room temperature.

Fig. 1
figure 1

Scheme of the gold nanoparticle-based paper assay for identification of Listeria monocytogenes from listeria spp.

Full size image

Evaluation of the performance of the ICG strip

To evaluate the sensitivity and cross-reactivity of the ICG strip, culture solution of eight L. monocytogenes serotypes, four Listeria spp., and six other common bacteria were centrifuged (5000×g, 10 min) and the supernatants were tested directly or diluted with 10 mM PBS (3×, 9×, or 27×) before detection. In addition, recombinant P60 protein of L. monocytogenes was diluted to 5, 10, 25, 50, 100, and 250 ng⋅mL−1 with PBS and analyzed with the ICG strip. PBS was used as a control.

For each single test, 7 μL of AuNP-labeled mAb 1C1 was mixed with 43 μL of suspension buffer (10 mM PBS, 2% BSA, 0.1% Tween, and 0.2% sucrose) and 100 μL of sample. Following incubation at 37 °C for 5 min, the solution was loaded onto the strip and allowed to react for 10 min. The results were visualized by bare eyes. Two red bands on both the T line and C line were indicative of a L. monocytogenes-positive sample, while one red band on the C line was indicative of a L. monocytogenes-negative sample.

Analysis of spiked milk samples using the ICG strip

To test the effectiveness of the ICG strip, real samples were analyzed. Pure milk was purchased from a local market and confirmed to be free of L. monocytogenes and Listeria spp. by a culture-based method [40]. Different strains of L. monocytogenes (ATCC 19111, ATCC 19115, ATCC 19118, and CMCC 54003) were cultured and added to pure milk to simulate samples contaminated with L. monocytogenes. One to nine colony-forming unit⋅mL−1 (CFU⋅mL−1) were added to the milk samples by serial dilution. Subsequently, 25 mL of the spiked samples were mixed with 225 mL of L. monocytogenes enrichment broth (Oxoid) and cultured at 37 °C. Samples (1 mL) of each culture solution were collected after 8 and 12 h and centrifuged (5000×g, 10 min). The supernatant of each sample was analyzed by the ICG strip.

Results and discussion

Principle of the ICG strip

The detection of L. monocytogenes was based on an antibody-antigen reaction on the NC membrane (Fig. 1). First, the sample was added to a tube with a pipette and reacted with the Au NPs labeled mAb of L. monocytogenes in the tube. The sample pad of the strip was then loaded into the sample solution. With the capillary force, AuNPs labeled mAb and the proteins in the sample flowed toward the absorption pad. The results can be quickly visualized as the color of the T and C lines change. MAb 10E7 against the P60 protein of all Listeria spp. (including L. monocytogenes) was coated on the T line, and mAb 1C1 against the P60 protein of L. monocytogenes was labelled with AuNPs and used as a detector probe. In L. monocytogenes-positive samples, P60 proteins first react with the detector probe and are captured by mAb 10E7 on the T line. Red bands on both T and C lines are indicative of positive samples. When Listeria spp. are present, P60 proteins cannot react with the detector probe. Even though P60 proteins are captured on the T line, a red band only appears on the C line (excess AuNP-labeled mAb captured by the goat anti-mouse antibody), which is indicative that the sample is L. monocytogenes-negative. Furthermore, negative results are obtained in samples containing bacterial strains devoid of P60 protein.

MAbs specific to L. monocytogenes pep D and Listeria spp. P60

L. monocytogenes-specific mAb was prepared against the peptide Pep D of L. monocytogenes P60 protein. Pep D was conjugated to BSA and used as an immunogen. The NH2 group of BSA first reacted with the N-hydroxysuccinimide ester group of SMCC. Following the removal of excess SMCC, maleimidomethyl group-activated BSA was conjugated with the cysteine SH group on the N-terminal end of Pep D. Conjugation of the peptide to the BSA plays a key role in preparing an effective immunogen, therefore we optimized two reaction ratios between BSA and the peptide(MW: 1300) with 1: 20 and 1: 40. The SDS-PAGE image in Fig. 2a shows that the molecular weight of SMCC-activated BSA increased compared with that of BSA, and increased further following conjugation with Pep D, which confirmed the successful conjugation of Pep D with BSA. Immunization of mice with the two immunogens revealed the titer of antibodies against L. monocytogenes was higher with the conjugates having higher reaction ratio (1:40). After cell fusion, mAb 1C1 was obtained. MAb 1C1 had high binding affinity towards both the recombinant P60 protein and the supernatant of L. monocytogenes. While specific to P60 protein secreted by L. monocytogenes, mAb 1C1 did not react with P60 protein of other Listeria spp. or with surface-associated P60 protein on L. monocytogenes. Pep D is a specific peptide of all L. monocytogenes serotypes, and is not accessible when P60 proteins are embedded in the cell membrane [36, 41].

Fig. 2
figure 2

a Characterization of the prepared immunogen. 1, BSA; 2, SMCC activated BSA with reaction ratio at 50:1; 3, PepD and BSA conjugate with reaction ratio at 20:1; 4, PepD and BSA conjugate with reaction ratio at 40:1. b Detection of P60 protein from L. monocytogenes with gold nanoparticle-based paper assay. From left to right, the concentrations of P60 protein were 250, 100, 50, 25, 10, 5, 0 ng⋅mL−1 in 10 mM PBS

Full size image

MAb 10E7, prepared with the whole P60 protein, was cross-reactive with both secreted and surface-associated P60 protein of all Listeria spp. Horseradish peroxidase-conjugated mAb 1C1 paired with mAb 10E7 had a detection limit (signal to ratio ≥ 2.1) of 1.2 ng⋅mL−1 P60 protein (Fig. S1) based on sandwich ELISA results. This ELISA was specific to the supernatant of L. monocytogenes culture solution but not to that of other Listeria spp. (Figure S2). The supernatant still can be detected after 1000 times dilution.

Identification of L. monocytogenes from Listeria spp. and other bacteria

The TEM image in Fig. S3a shows that AuNPs had a diameter of 15 nm ± 2.8 nm. The UV data in Fig. S3b reveals that the maximum absorption of AuNP-labeled mAb 1C1 and AuNPs was 524 and 521 nm, respectively. The red shift in the absorption band revealed the successful conjugation of mAb with AuNPs. Recombinant P60 protein was analyzed with the strip assay. Figure 2b shows that the visual detection limit of P60 protein was 25 ng⋅mL−1 in PBS.

The presence of other Listeria spp. and bacterial strains may interfere with the detection and identification of L. monocytogenes. Supernatant of eight L. monocytogenes strains including ATCC 19111 (serotype1/2a), CMCC 54007, ATCC 19115 (4b), CMCC 54003 (1/2a), CMCC 54004, CMCC 54002 (1/2c), ATCC 19118 (4e), and a wild strain isolated from frozen beef were all detected by the lateral flow assay (Fig. 3a). As shown in Fig. 3b, the gray values decreased along with the decrease of the equivalent bacteria numbers. Based on the gray value, the practical detection limit (P/N ≥ 2.1) was 27X of the supernatant, which is equivalent to 3.7 × 106 CFU⋅mL−1 of L. monocytogenes because the original concentration was 1 × 108 CFU⋅mL−1 after culturing overnight (12 h). Therefore, the assay had broad cross-reactivity among different L. monocytogenes. The detection limit was relatively high because only secreted P60 proteins can be detected (P60 proteins that were surface-associated were not accessible to mAb 1C1). The detection sensitivity was also limited to the affinity of mAb 1C1 due to the short incubation time of the ICG strip assay.

Fig. 3
figure 3

a Gold nanoparticle-based paper assay for detection of eight strains of L. monocytogenes. The supernatants of culture solution were tested directly or after diluted at 3 times, 9 times, 27 times with PBS, control check was the 10 mM PBS. b The plot of grey value on the test line versus the equivalent CFU values, Error bar was the standard deviation and the gray value was the average of six replicates of measuring the same bacterial sample

Full size image

In spite of the presence of P60 protein in the supernatant, L. ivanovii, L. innocua, L. welshimeri, and L. grayi did not interfere with the detection of L. monocytogenes (Fig. 4a). When other Listeria spp. presented with L. monocytogenes together in the food and were tested with the strip after enrichment, the Au NPs labeled 1C1 selectively reacted with the P60 protein from the L. monocytogenes other than the listeria strains. When they moved to the test line, both the P60 protein from the L. monocytogenes and from listeria strains were equally captured by the mAb 10E7 but only P60 protein from the L. monocytogenes generated color on the test line. The equal competition slightly decreased the color on the test line, but the result was still positive. Additionally, there was no cross-reactivity with other bacterial strains such as Escherichia coli, E. coli O157:H7, Cronobacter sakazakii, Salmonella enteritidis, Staphylococcus aureus, or C. jejuni (Fig. 4b). The results confirmed the specificity of the method (Fig. 4c). MAbs 1C1 and 10E7 were prepared against the conserved peptide (Pep D) of L. monocytogenes P60 protein and whole P60 protein of Listeria spp., respectively. The combination of the two mAbs contributed to a very specific detection method. Of the 13 different serotypes in L. monocytogenes, 1/2a, 1/2b, 1/2c, and 4b account for 95% of human listeriosis cases [10]. Therefore, the ICG strip represents a portable and effective tool for the identification of L. monocytogenes.

Fig. 4
figure 4

Cross-reactivity of the gold nanoparticle-based paper assay: a with four Listeria strains, 1, L. monocytogenes; 2, Uncultured medium as control check; 3, L. grayi; 4, L. ivanovii; 5, L. welshimeri; 6, L. innocua. b with other bacteria, 1, L. monocytogenes; 2, E.coli; 3, E. coli O157:H7; 4, Cronobacter sakazakii; 5, Salmonella enteritidis; 6, Staphylococcus aureus; 7, Campylobacter jejuni. Strains were cultured overnight and tested without dilution after centrifugation. c The overall performance of the gold nanoparticle-based paper assay. Error bar was the standard deviation and the gray value was the average of six replicates of measuring the same bacterial sample

Full size image

Analysis of milk samples spiked with L. monocytogenes

Pure milk samples spiked with low levels (1 to 9 CFU⋅mL−1) of four L. monocytogenes strains were analyzed by the ICG strip following a short enrichment period (8 and 12 h). Figure 5 shows that L. monocytogenes strains ATCC 19111, ATCC 19115, ATCC 19118, and CMCC 54003 were detected following a 12-h enrichment period. On the other hand, the 8-h enrichment period was too short to detect P60 proteins in the culture solution. The evaluation of the tested strains with popular serotypes (1/2a and 4b) demonstrated that our method was effective for the analysis of low concentrations of L. monocytogenes in foods. The results obtained by the ICG strip were consistent with those obtained by the culture-based method but more efficient and simple (Table S1). Nowadays, PCR assays are available and reliable, but need skilled training and sophisticated instrument. In contrast, immunoassays like lateral flow assay are very simple and portable but sometimes not accurate due to poor quality of the mAbs. Therefore, it is necessary to develop qualified mAbs against the pathogens and make the immunoassays also very dependable. Furthermore, once obtained, these antibodies can be produced in a large scale with homogenous quality. The Blazkova and coauthors developed an effective nucleic acid lateral flow immunoassay for listeria spp. and L. monocytogenes, which was also portable and simple [42]. However, the nucleic acid needs further enrichment by the PCR before lateral flow assay. This lowered the simplicity of the method and increased the cost. In our work, the sample was analyzed without PCR after enrichment by culture broth, which was more timesaving and use-friendly.

Fig. 5
figure 5

Analyzing low level of L. monocytogenes in Milk sample with gold nanoparticle-based paper assay after 8 h and 12 h enrichment: 1 PBS as control check, 2 L. monocytogenes ATCC 19111, 3 L. monocytogenes ATCC 19115, 4 L. monocytogenes ATCC 19118, 5 L. monocytogenes CMCC 54003

Full size image

L. monocytogenes mAbs against different surface antigens such as autolysin IspC [35], internalin A [21, 34], and P60 protein [22, 36, 41] have been prepared. A sandwich ELISA against L. monocytogenes based on P60 mAb and polyclonal Ab (pAb) was developed by Yu; however, it’s based on pAb and sensitivity was not reported. Recently, a phage display antibody prepared by Tu [10] recognized the popular serotypes 1/2a, 1/2b, and 4b of L. monocytogenes, and the detection limit of the sandwich ELISA was 104 CFU⋅mL−1. However, the ICG strip test reported by Shim [25] cross reacted with Listeria spp. (107 CFU⋅mL−1) and was not strictly specific to L. monocytogenes. The ICG strip assay developed by Kim [26] used IMS for pre-concentration. The sensitivity decreased to 6.6 × 103 CFU⋅mL−1 following IMS. However, the cross-reactivity against different serotypes of L. monocytogenes was unknown. As reviewed in Table 2, the sensitivity of our method was a little lower but still comparable with the classical lateral flow assays. However, the existing immunoassays are all insufficient to detect the low contaminant level (1–9 CFU⋅mL−1) in the food sample and enrichment was inevitable. The advantage of our method was the specific and homogenous identification of different serotypes of L. monocytogenes, which provided a resolution to the current cross-reactivity issue with other existing ICG strip assay.

Table 2 An overview on recently reported immunoassay for determination of L. monocytogenes
Full size table

The Pep D pAb prepared by Bubert [36] recognized L. monocytogenes, and the P60 protein mAb prepared by Yu [22] recognized L. monocytogenes and Listeria spp. These results were confirmed in our study. It is noteworthy that the two types of mAbs can be paired in the sandwich ELISA and be effective in the lateral flow assay, probably because the epitopes of the two types of mAbs are different and both are part of the P60 protein. Our results overcome the main limitations of the current ICG strip immunoassay including the lack of mAb specific to L. monocytogenes and the lack of a mAb pair.

Conclusion

We prepared mAb 1C1 that recognized specific peptide Pep D of L. monocytogenes and mAb 10E7 that recognized P60 protein of Listeria spp. The mAb pair was used in the development of an AuNP-based ICG strip assay. Based on the results, the ICG strip assay detected the eight tested L. monocytogenes strains of different popular serotypes (1/2a, 1/2b, 4b, 4e), without any cross-reactivity with other Listeria spp. (L. ivanovii, L. innocua, L. welshimeri, and L. grayi) or common Gram-positive and Gram-negative bacteria. Milk samples spiked with L. monocytogenes were detected by the ICG strip assay in 13 h. Therefore, the ICG strip represents a portable and effective tool for the identification of L. monocytogenes in foods.