Rapid and sensitive detection of Yersinia pestis by lateral-flow assay in simulated clinical samples

Construction of recombinant p18RMAF1

Two DNA fragments containing the Y. pestis caf operon [35, 36], were amplified by PCR (Perkin Elmer DNA Thermal cycler 480; 30 amplification cycles: 95 °C for 1-min, 54 °C for 1-min, and 72 °C for 2-min) with Y. pestis ATCC 19428 template DNA and oligonucleotide primer pairs including: F1OU(F) (5′-GTTCCGGAATTCTTCCGAACATAAATCGGTTCAGTGGCC-3′) and F1OU(R) (5′-GGCGTA TTCCTCGCTAGCAATGTTTAACG-3′), or F1OD(F) (5′-ATCGTTAAACATTGCTAGCGAGGA ATACGCC-3′) and F1OD(R) (5′-GTTCCGCTGCAGTGAACCTATTATATTGCTTCGCGC-3′). Amplified DNA fragments were first ligated together using the unique Nhe I restriction site in the caf operon, and then cloned into a high-copy-number vector, pUC18 that obtained from Promega (Medison, WI, USA). This was followed by Eco RI/Pst I digestion (with T4 ligase, Invitrogen) to generate the plasmid p18RMAF1. To confirm the coding sequences of caf operon genes, DNA sequencing was performed, and the deduced amino acid sequences of the caf operon genes of Y. pestis (ATCC 19428) were shown to be identical to previously reported sequences of Y. pestis CO92 (Accession No. AL590842, data not shown). The recombinant p18RMAF1 plasmid was transformed into E. coli strain DH5 to generate the E. coli (p18RMAF1) strain.

Expression and purification of recombinant F1 antigen

All cultures, tests, and treatments involving Y. pestis were performed in a bio-safety level 3 (BSL-3) laboratory. For expression of the F1 antigen, a seed culture was prepared as follows (culture medium (Luria-Bertani broth, LB) was purchased from BD (Becton Dickinson, USA)): LB broth (10 ml) containing ampicillin (100 μg/ml) was inoculated with a loop of E. coli (p18RMAF1) from a plate stock, and grown for 16~ 18-h at both 28 and 37 °C with shaking. LB broth (2-L) containing ampicillin was subsequently inoculated with the seed culture and grown at 28/37 °C for 48-h with shaking at 200 rpm. The cultures were harvested by centrifugation at 12,000×g for 5-min at 4 °C. Following centrifugation, ammonium sulfate (final concentration 30% w/v) was added to the bacteria-free supernatant to generate the F1 precipitant (4 °C, overnight). The F1-containing precipitate (if there contains any) was harvested by further centrifugation (12,000×g), dissolved and dialyzed (overnight at 4 °C) against 1 mM PBS (potassium buffer saline, pH 7.2), followed by filtration through a 0.22 μm membrane. Finally, the crude F1 antigen was further purified through a low-molecular-weight cutoff filter (10 kDa, Amicon Ultra, Millipore Co. Bedford, MA), a size-exclusion chromatography Superose-6 column (SEC, separation based on molecular size, GE Healthcare), and an anion-exchange Q-Sepharose FF column (AEC, separation by molecular charge, GE Healthcare) [37]. The purity of the F1 protein was evaluated using sodium-dodecyl-sulfate-polyacrylamide gel electrophoresis (SDS-PAGE, 12%) and western blotting [38], using a mouse anti-Y. pestis F1 antibody (YPF19, Abcam Inc). Finally, the F1-protein was quantified by Coomassie Plus – The Better Bradford Assay Kit (Thermo Scientific, USA).

Generation of anti-F1 monoclonal antibodies

Monoclonal antibodies against the F1-antigen were generated as previously described [39]. Briefly, F1-antigen (100μg/ml, 200 μl) was first mixed with an equal volume of complete (first inoculation) or incomplete (subsequent inoculation) Freund’s adjuvant (Difco). The antigen-adjuvant mixture was then subcutaneously injected into BALB/c mice. The eight-week-old BALB/c mice (male, number of mice: 5) for ascites production were purchased from Biolasco, Taipei, Taiwan. After 4 weeks, the mice were boosted with F1-antigen (0.1 mg/ml, 100 μl) by intravenous inoculation. Mice were sacrificed (inhalation with isoflurane and then CO₂ suffocation) 3 days after the intravenous boost, and spleens were removed and homogenized. The spleen cells were subsequently fused with mouse myeloma NS-1 cells to produce hybridoma cells.

For hybridoma selection, culture medium collected from hybridoma clones was centrifuged at 500×g for 10-min, and the supernatants were assayed by the ELISA screening. First, flat-bottom 96-well plates (ICN Biomedicals) were coated with 50 μl of purified F1 protein (2 μg/ml) in 0.05 M NaHCO₃ (pH 9.6), at 4 °C overnight. The plates were decanted, washed (dd H₂O) and blocked with 200 μl of blocking buffer (5% BSA in PBST (PBS containing 0.1% Tween-20)) at 37 °C for 1-h. Subsequently, the wells were washed four times with 250 μl of PBST, and were incubated with 50 μl of culture supernatant for 1-h at 37 °C. The plates were washed again (4×) with 250 μl of PBST, and incubated with 50 μl of goat-anti-mouse IgG (H + L)-HRP at a 1:2000 dilution in PBST for 30-min at 37 °C, followed by decanting and final washing. Substrate, 50 μl sodium perborate (Sigma P-4922) in phosphate-citrate buffer containing O-phenylenediamine (Sigma P-6787) and 0.03% H₂O₂, was added. After 10~ 20 min incubation in the dark (room temperature), the reaction was stopped by adding 100 μl 4 N H₂SO₄ and the absorbance was determined by ELISA reader at 490 nm.

After selection, the hybridoma cells were injected intraperitoneally into five BALB/c mice with incomplete adjuvant for ascites production. Various anti-F1 monoclonal antibodies were purified from mice ascites through an IgG-specific immunosorbent, thiophilic gel (Pierce, Rockford, USA). (5).

Animal operations

The animals (for ascites production) were placed as a whole (i.e. housed in a 362 × 235 × 195 mm cage with sufficient food and water) and given 5 days to accommodate the environment in the housing facility. The environmental conditions were as followed: temperature: 21 °C. ± 2 °C., humidity 55% ± 10%, illuminance 350 lx, ratio of light to dark cycle was 1:1, and light was turned on and off at 0700 and 1900 h. All animal experiments were conducted in compliance with the regulations of IPM Institutional Animal Care and Use Committee (IPM-IACUC) of the National Institute of Allergy and Infectious Diseases, National Institutes of Health, and with licenses of IPM-IACUC ref. AN102–15 (2013) and AN-104-15 (2015). All sections of this report adhere to the ARRIVE Guidelines for reporting animal research.

A completed ARRIVE guidelines checklist is included as an addition file.

When ascites has been taken twice, or when the tumor was too big to affect physiological conditions (e.g., activity, breathing, etc.), the experiment is terminated and euthanized by anesthesia (with isoflurane) to prevent animals from to suffering. The animals are then packaged in plastic bags, autoclaved and subsequently incinerated.

Conjugation of monoclonal anti-F1 antibody to colloidal gold particles

Monoclonal anti-F1 antibodies were conjugated to colloidal gold particles (acquired from Aurion (Wageningen, Netherlands)) as described previously [29, 31, 39]. Briefly, 25 nm colloidal gold particles were applied to the conjugation process. The anti-F1 IgG (1 mg/ml, 0.1 ml) was gently added to 0.9 ml of colloidal gold solution (1% w/v, pH 8.5) and incubated with shaking for 30-min at room temperature. The colloidal gold particles were next precipitated by centrifugation for 30-min at 4 °C (1550×g; 8178 swing-out rotor, Labofuge 400R, Heraus Instrument, USA), and suspended in 1 ml working buffer (20 mM Tris/HCI buffer (pH 8.2) containing 1% w/v BSA). The optical density of the suspension was adjusted to 5.0 at 520 nm. The prepared anti-F1 IgG-coated colloidal gold probes were stored at 4 °C until use (30 μl/cm on the conjugation pad).

Preparation of immunochromatographic test strips

The strip elements, including high-flow nitrocellulose membranes (AE-98), glass fiber conjugation pads (AccuFlow™G), sample application pads (#12-S), and reagent adsorption pads (470 Zuschnitte), were all purchased from Schleicher & Schuell GmbH (Dassel, Germany) and had been described previously [28, 29, 31]. Briefly, 1 μl of goat anti-mouse IgG (whole molecule, purchased from Sigma (St. Louis, MO, USA)) and mouse anti-F1 monoclonal antibodies (1 mg/ml each) were sprayed onto a nitrocellulose membrane independently using a BioDot dispensing apparatus (BioDot XYZ 3000 1414) to create a control region (C) and a test region (T). The membrane was incubated in 1% w/v polyvinyl alcohol for 30-min at room temperature to block remaining active sites, followed by a quick wash of the strip with ddH₂O and subsequent drying. The membrane was then adhered to an adhesive paper plate (2.44 × 11.81 in., Adhesive Research Inc., Taiwan), with an additional reagent adsorbent pad, a colloidal gold conjugate pad (containing F1-IgG probes), and a sample application pad. The paper plate was then cut into 5-mm-wide strips (CM4000 cutter, BioDot) and mounted on a plastic cassette. At this point, the device was ready for use.

Bronchoalveolar lavage fluid collection

The extraction of mouse BAL fluid was based on the method described by Baughman [40], with a minor modification. Briefly, two 12-week-old ICR mice (male, Biolasco, Taiwan) were anaesthetized with 0.03 ml Zoletil-50 (Zolazepam + Tiletamine: 2.5 × dilution with PBS, pH 7.4) through intramuscular injection, followed by heart exsanguination (blood was aliquot in two tubes containing anticoagulant EDTA). The mice were then anatomized, cut the neck skin and pushed back the neck muscles to expose the trachea. An IV catheter (24 GA, 0.75 in, Angiocath Plus™, BD REF 591836, Korea) was then inserted, fixed with the black nylon threads, washed twice with normal saline (1 and 0.5 ml each), and the lavage fluids (1.2 ml final volume) were collected.

The living environment for ICR mice was identical to that of the BALB/c mice.

Sensitivity and specificity of the F1 test strips

Sensitivity and specificity assays for the test strip have been reported elsewhere [29, 31, 41].

Briefly, appropriate amounts (100 μl) of various concentrations of F1 proteins (2~ 20 ng/ml), or Y. pestis (10³~ 10⁵ CFU/ml as measured by plate count) were applied individually to the strips. To facilitate sample migration, a half volume (50 μl) of tracing buffer (potassium carbonate buffer, PCB) was applied when samples were almost drained. In addition, various samples containing Yersinia strains (e.g. Y. pestis yreka strain IPM00722; Y. mollaretii (ATCC43969); Y. frederiksenii (ATCC 29912); Y. pseudotuberculosis (ATCC29910); Y. enterocolitica (ATCC 27729) and Y. intermedia (ATCC29909) were also assayed by F1 strips for specificity evaluation. The Enterobacteriaceae bacterial strains were kindly supplied by Dr. Shih-Shiung Huang, from the Institute of Preventive Medicine, National Defense Medical Center, Taiwan. To verify the reproducibility of the strip analysis in this section, the sensitivity and specificity tests were performed in triplicate from the same (or different) batch of strips (intra- and inter-assay).

Evaluation of the F1 test strip with simulated and field-captured rat samples

To assess the feasibility of F1-strips in various biological conditions, several different matrices were employed. The samples (100 μl of F1 or Y. pestis) were mixed with the appropriate volume of human sera (purchased from Sigma: S1-M EMD Millipore, Sigma-Aldrich, 1:5 dilutions in PCB), mouse blood (0.5 ml from one BALB/c mouse, 1:10 dilutions in normal saline), or mouse BAL (1:10 in normal saline). Mouse blood was pre-treated with distilled water for 10 min to allow the blood cells to burst and hence reduce interference and false-positive test results, and subsequently diluted with normal saline. Control samples were prepared by direct application of PCB/normal saline to the matrices. The F1-strips were also evaluated using sera from the rodents captured in Penghu Island (48 rodents, 13 with flea). The rodents were anesthetized before exsanguination, and same as before, the sera were diluted with five volumes of PCB followed by applying onto the homemade Mab/Pab F1-strips. In order to validate the efficiency of the F1-strips, samples were also analyzed by ELISA. The F1 immunized and pre-immunized NC-C2 rats were used as the control group.

Construction, expression, and purification of recombinant F1 protein

Figure 1 shows the constructed plasmid. Three structural subunits, caf1, caf1M, caf1A, and the regulatory protein (caf1R) were cloned into the pUC18 vector to generate the 8-Kb recombinant plasmid p18RMAF1. The caf gene cluster could thus express the F1 structural protein. This plasmid was then transformed into E. coli for F1 protein production.

The F1 protein was subsequently expressed and purified as described in the materials and methods section. To test the presence of F1, we cultivated the Y. pestis at different temperatures (37 °C and 28 °C, same incubation time). A previous study mentioned that “F1 capsule production is strongly influenced by temperature--little or no capsule is detected at < 35°C in vitro or in the flea vector” [42]. The cultivation results show that F1 can only be detected in the bacterial cells cultured at 37 °C, but not at 28 °C (similar bacteria concentration, judged by McFarland detection kit), which was consistent with the previous report (data not shown).

After Superose (for size exclusion, one peak) and Sepharose (for ion exchange, two peaks) purification (Fig. 2a), the expressed F1 recombinant protein was assayed by SDS-PAGE followed by western blotting. If the order of these two columns was changed, no difference in purity was observed (data not shown). The results of SDS-PAGE (Fig. 2b) and western blot (Fig. 2c) showed that P1 and P2 had the same pattern and both displayed the predicted molecular weight of F1-antigen (15.5 kDa) [11]. P1 and P2 were then pooled and used as an antigen for mice immunization.

Production of F1-MAbs

Pairing of MAbs on strip device

Sensitivity and specificity of F1 test strips

For the sensitivity assay of F1 test strips, samples with different amounts of Y. pestis and F1 protein were applied to previously prepared F1-strips. Figure 3 presents the assay results. All the analyses were accomplished in less than 10-min, and the detection limit for the bacteria and F1 protein was 10³ CFU/ml and 4 ng/ml, respectively. While in the absence of Y. pestis, there was no visible band in the test area and was therefore considered as negative. For specificity analysis, various Yersinia strains (10⁵ CFU/ml of each) were applied to the F1-strips. Figure 4 shows that only Y. pestis presented a positive result, whereas all other Yersinia species resulted in negative. As for inter- and intra-assays, at least three assays were included. This means that the intra-assays were to test the strips generated from the same batch at different times (e.g. once a week, at least 3 weeks). The number of strips used per batch was not less than 30. As for inter-assays, strips produced from different batches (at least 3 different batches, under the same generation conditions) were assayed simultaneously. Likewise, at least 30 strips were used for each assay. All tests results were confirmed by the six members of the experimental group. In addition, in this section, no sensitivity or specificity variation occurred in the inter- and intra-assays.

The sensitivity and specificity assay results demonstrated that this format has a remarkable increased detection sensitivity compared to that of Pab-based methods [39]. That is, in the previous study, a LFA based on a polyclone-antibody format for detection of Y. pestis F1 antigen and bacteria obtained the detection limit of 50 ng/ml and 10⁵ CFU/ml, respectively. Whereas in this study (MAb-based), the detection sensitivity was improved to 4~ 5 ng/ml for F1 antigen and 10³ CFU/ml for bacteria. Although silver enhancement may improve the detection sensitivity; it needs at least 1 h to obtain results.

If there is sufficient time, for sure it can be applied in field situations thus increase the detection limit; or it needs to be proceeded on a laboratory bench in other ways.

Evaluation of simulated sample and wild-captured rat serum by F1 test strips

F1 protein or Y. pestis spiked into human serum, mouse blood, and mouse bronchoalveolar lavage fluid was assayed by the F1-strips. Results (Fig. 5 and Table 3) revealed that in human serum, both F1 protein and Y. pestis maintained the same detection sensitivity as that in PCB (Fig. 5a); however, some sensitivity regression was observed when assayed with mouse blood and BAL. As shown in Fig. 5b, when using F1-strips to analyze F1 or Y. pestis in mouse blood, the detection limits dropped to 100 ng and10⁵ CFU/ml, respectively. In BAL (Fig. 5c), the level of sensitivity for Y. pestis decreased to 10⁴ CFU/ml, and the sensitivity to F1 protein also been affected (data not shown).

Specimen types	PCB				Human serum				Mouse blood				Bronchoalveolar lavage
	Y. pestis (CFU mL− 1)		F1 protein (ng mL− 1)		Y. pestis (CFU mL− 1)		F1 protein (ng mL− 1)		Y. pestis (CFU mL− 1)		F1 protein (ng mL− 1)		Y. pestis (CFU mL− 1)		F1 protein (ng mL− 1)
	106	+	10	+	106	+	10	+	2 × 107	+	1000	+	2 × 107	+	1000	+
	105	+	5	+	105	+	5	+	2 × 106	+	100	+	2 × 106	+	100	+
	104	+	4	+	104	+	4	+	2 × 105	+	10	+	2 × 105	+	10	+
	103	+	3	±	103	+	3	±	2 × 104	±	5	±	2 × 104	+	5	±
	102	±			102	±							2 × 103	±

In addition, the Mab/Pab F1-strips were applied to field captured rodent specimens (the rodents were sacrificed with the method same as that of ICR mice). The results revealed that when these two types of strips were used for plague detection, same results were obtained. Although fleas were detected in some of these rodents (13/48), all rodents were F1 negative (data not shown. On the other hand, half of the samples were re-assayed by ELISA (blind test, data not shown), and the results were identical to the strip assay. These results implied that Y. pestis does not exist in selected areas; hence, the plague is a rare occurrence on Penghu Island.