Nanotechnology

CPA-Cas12a-based lateral flow strip for portable assay of Methicillin-resistant Staphylococcus aureus in clinical sample | Journal of Nanobiotechnology


Principle of CPA-Cas 12a lateral flow assay strategy

CPA and CPA-Cas 12a sensing strategies were proposed for detection of nuc and mecA genes in MRSA respectively. According to the sequence of the target gene, we designed two PAM sites and four pairs of primer according to the principles of double-crossing CPA and trans-cleavage of CRISPR/Cas 12a. The details of PAM sites and primers are shown in Fig. 2 and Table S1 respectively.

Fig. 2
figure 2

Schematic diagram of primer design for CPA and CPA-Cas 12a assay. (A) Primers of nuc and PAM, (B) Primers of mecA and PAM.

The outline of double signal amplification CPA-Cas 12a assay for MRSA in the secretions of infected patients is described in Scheme 1. As shown in Scheme 1 A, 500 µL lysate solution containing 20 mg mL− 1 lysozyme and proteinase K was added to enrolled patients’ secretions for 30–60 min reaction at room temperature. Next, 500 µL mixture (50 mg mL− 1 silicified magnetic beads, 8 M GuHCl) was introduced to as-prepared lysate solution. The reaction was incubated at room temperature for 20 min and stirred slightly at every 5 min. Subsequently, magnetic beads were separated and cleaned twice with cleaning solution (25 mM NaHCO3, 1% casein, 0.4 mM EDTA, 0.2% α-cyclodextrin, 0.2% Triton-100, 0.4 M urea, 0.1% sodium azide, pH 9.6). Finally, 200 µL eluent (35% absolute ethanol) was used to eluate the nucleic acid absorbed on the surface of magnetic beads, which was the analyte in the following study.

The process of CPA is shown in Scheme 1B. Efficient isothermal amplification of target sequence was able to be achieved by using two pairs of cross-primers. The produced four kinds of dsDNA involving target sequence could be recognized by CRISPR/Cas12 system under the guidance of CrRNA. The repeat sequence and programmable target-specific sequence designed in CrRNA promote dsDNA to form an R loop. In presence of Cas 12a, formed Cas12a-CrRNA duplex can specifically recognize and cis-cut target dsDNA (Scheme 1 C). After the reaction, nonspecific ssDNA can be trans-cut by duplex. In this study, a large amount of dsDNA generated by CPA was recognized by Cas 12a protein and CrRNA, resulting in dissociation of reporter DNA. Therefore, the established lateral flow strip biosensor could convert the disintegration of reporter DNA into visual detection (Scheme 1D).

Scheme 1
scheme 1

Schematic illustration of CPA-Cas 12a mediated lateral flow assay for MRSA. (A) Sample preparation from patient. (B) Design of CPA system. (C) CRISPR/Cas12a activity. (D) Visualization principle of lateral flow strips

Verification of CPA integrated CRISPR/Cas system

As shown in Fig. S1A, we designed two pairs of primers labeled with FITC and digoxigenin to evaluate the feasibility of amplification reaction of double-crossing CPA system. After reaction, the products were examined by agarose electrophoresis as shown in Fig. S1B. A large amount of dsDNA products with different chemical base could be seen in lane 5 and 7 at the expense of primers depletion, which demonstrated the successful development of this CPA isothermal amplification method.

On the other hand, a total of 100 nM CrRNA was preincubated with 75 nM CRISPR/Cas 12a protein in 1 × NE Buffer a 37 ℃ for 10 min. After the successful formation of CRISPR/Cas 12/CrRNA complex, the CPA-Cas 12a trans-cleavage activity was performed in a 100 µL mixture containing 25 µL of 2 × NE Buffer, 1.25 µL DNA reporter molecule (FITC-ATTAGCACTTGTAAGCACACCTTCA, 10 µM), 1 µL aforementioned CPA amplification products, 9 µL CRISPR Cas 12a/CrRNA protein mixture, 13.75 µL deionized water containing RNA enzyme inhibitor (1 U µL− 1) reaction system. The reaction was carried out at 37℃ for 30 min and the product was characterized by agarose electrophoresis. As shown in Fig. S2, lane 1, lane 2, lane 3, lane 4, lane 5, lane 6 and lane 7 were the bands of marker, CrRNA, CrRNA with Cas12a, CPA product, DNA reporter, CPA with CRISPR/Cas 12a product and Cas12a-CrRNA duplex without CPA products respectively. By comparing the bands in lanes 5, 6 and 7, DNA reporter was able to be trans-cleavage by Cas12a-CrRNA duplex to form fragment in the presence of CPA product containing target dsDNA, resulting in disappearance of bands marked as red arrow. In addition, in order to demonstrate that the CPA products could initiate ssDNA trans-cleavage activity of Cas12a-CrRNA duplex, a fluorophore and quenched labeled ssDNA reporter denoted as Beacon2 with 5 bases (Table S1) of approximately 1.7 nm in distance, was designed and added in the above-mentioned 100 µL mixture without DNA reporter molecule. As shown in Fig. 3C, the fluorescence was recovered when only CPA products and Cas12a-CrRNA duplex simultaneously presented to start the trans-cleavage activity, which further demonstrates the excellent specificity of proposed CPA-Cas 12a strategy.

Optimization of the assay conditions

To achieve the best signal output performance, various factors including concentrations of CRISPR/Cas 12/CrRNA complex and beacon in solution, reaction time, and the concentration of beacon on strip were optimized through repeated assays.

We firstly optimized the concentration of CRISPR/Cas 12/CrRNA complex according to our previous work [30]. As shown in Fig. 3A, the magnitude of fluorescence increased from 5 nM, and started to stabilize at 50 nM. After 50 nM the fluorescence showed no obvious change. In contrast, the negative control (without bacterial DNA extraction solution) signal did not change significantly with the increase in the concentrations of CrRNA and Cas 12a, further demonstrating the high specificity of this assay. Therefore, a concentration of 50 nM was chosen for CRISPR/Cas 12/CrRNA complex. To ensure high sensitivity and efficiency, we varied the concentration of beacon from 50 to 400 nM. Figure 3B shows the concentration of beacon dependence of fluorescence response in the presence and absence of target respectively. According to the signal/background, which was to quantify the sensitivity of assays, beacon with 200 nM afforded the highest efficiency. Thereby, 200 nM beacon which could provide the signal-to-background ratio (S/B) of 11 was selected in our experiment. Furthermore, the time-dependent change of fluorescence signal was recorded after target was introduced. Figure 3 C showed that after the addition of target, the fluorescence kept increasing gradually in early time and its fluorescence intensity reached the maximum when the incubation time was close to 1800 s, while the negative control without bacterial DNA extraction solution hardly affected the fluorescence intensity in the duration of 3500 s. In addition, the concentration of beacon on the strip exhibited the same optimal concentration of 200 nM in homogeneous phase as that selected in solution phase as shown in Fig. 3D. It is worth mentioning that when the beacon concentration increased to be 10 µM, the signal output showed a false-negative, which may be related to the hook effect. Therefore, this CPA via Cas 12a mediated fluorescent signal and its coupled lateral flow assay has the same disadvantage of high-concentration inhibition. In order to circumvent this drawback, we proposed this CPA-CrRNA based reverse signal output mode to switch the concentration of DNA reporter molecules to visual signal, where the target dsDNA and the signal collected from test line had inverse proportional relationship.

In short, the following experimental conditions were found to give best results: (a) 50 nM of CRISPR/Cas 12/CrRNA complex; (b) 200 nM of beacon in both homogeneous phase and lateral flow strip; (c) reaction time of 30 min.

Fig. 3
figure 3

Optimizations of concentrations of (A) CrRNA and Cas 12a, (B) beacon, (C) reaction time of double-crossing CPA-Cas 12a assay and (D) beacon on lateral flow strip

Analytical performance of CPA-Cas 12a lateral flow assay

Under optimal experimental conditions, the analytical performance of this lateral flow strip assay was evaluated by detecting different concentrations of S. aureus, which was prepared by diluting 0.5 MCF (~ 0.5 × 108 CFU mL− 1) of S. aureus bacterium suspension with different volume of PBS. As shown in Fig. 4A, germ culture was used to prepare every diluted bacterium suspension and their bacterium colony was counted by blood plate. The concentration of 500 µL bacterial suspension was adjusted to 1, 10, 102, 103, 104 and 105 CFU mL− 1 to extract DNA and initiate CPA-Cas 12a reaction. The lateral flow strip detection was executed as described in the last section. All experiments were repeated in three times. Figure 4B and C exhibit the performance comparison of CPA based single-amplification assay and double amplification CPA-Cas 12a based assay respectively. S. aureus spiked in PBS could be successfully identified on both kinds of lateral flow strips by naked eyes within 3 h. The intensity of test and control lines of lateral flow strips representing as densitograms was analyzed by Image J after homogenizing as show in Fig. 4D and E respectively.

The signal intensity of CPA based assay increased continuously with the increase of S. aureus concentration (Fig. 5A). As shown in Fig. 5B, a good linear correlation between the value of signal and the S. aureus concentration between 1 and 103 CFU mL− 1 was obtained with regression equation of ΔI = 12.01×lgC + 1.69 (ΔI represents the intensity of test line, and C represents the bacterial concentration) with a correlation coefficient (R2) of 0.97, and the detection limit was calculated to be 2.26 CFU mL− 1, evaluated by the definition of the rule of 3σ/k (σ represents the standard deviation and k represents the intercept of the fitting curve) [31]. But given the potential limitations of visual inspection and actual detection capability, we estimated the LOD of our CPA based biosensor by naked eye was 5 CFU mL− 1.

Fig. 4
figure 4

Verification of CPA-Cas12a mediated lateral flow assay (nuc). (A)S. aureus absolute colony culture counts of S. aureus at different concentrations. (B) Results of CPA mediated lateral flow strips for detecting bacterial suspensions with different concentrations. (C) Results of CPA-Cas 12a mediated lateral flow strips for detection of bacterial suspensions with different concentrations. (D) Densitograms of line in CPA mediated lateral flow strips; (E) Densitograms of line in CPA-Cas 12a mediated lateral flow strips

Since the gene of S. aureus could be amplified with CPA, the resultant CPA amplicon could activate the ssDNA trans-cleavage activity of Cas12a-CrRNA duplex and produced the beacon from breakage of bridge connection. Thus, the CPA product originated from S. aureus could be visually analyzed by observing the color intensity in the test line coated with AuNPs-streptavidin conjugates as well as anti-FITC by hapten-anti-hapten interaction. To quantify the magnitude of color intensity in the test line on the lateral flow strips, densitograms of its corresponding pixel were employed to calculate its optical density as shown in Fig. 4E. The optical density of test line exhibited a nonlinear relationship with the quantity of bacterial, which no signal was detectable in the presence of S. aureus at the concentration larger than 1 CFU mL− 1. This is attributed to this ultra-sensitive double-amplification CPA-Cas 12a sensing strategy. Furthermore, this ultra-sensitive “0” or “1” binary output mode could significantly increase true positive rate, reduce false negative rate, and show an excellent advantage for semi-quantitative analysis by visual inspection (LOD = 5 CFU mL− 1). The analytical performance of this method was compared with other detection strategies and the corresponding results were listed in Table S2. This proposed assay obviously offers very high sensitivity due to the CPA-Cas 12a assisted double-amplification strategy.

Fig. 5
figure 5

Sensitivity evaluation of the assays. (A) Color intensity of T line in CPA mediated lateral flow strip for different concentrations of bacterial suspension. (B) Linear curve fitting of color intensity with bacterial concentrations ranging from 1 to 103 CFU mL− 1

Good specificity, which directly affects the clinical application prospect, is one of the indispensable characteristics to evaluate biosensor. Several other gram-negative, gram-positive, and candida bacteria were involved in the selectivity study of this proposed method. As illustrated in Table 1, benefiting from the specificity design of the two pairs of primers, this assay had good selectivity for the detection of S. aureus.

Overall, the above results proved that our lateral flow strip had high sensitivity and good selectivity for pathogen detection.

Table 1 Specificity evaluation of CPA and CPA-Cas 12a sensing strategy

S. aureus analysis in clinical sample

To evaluate the clinical application potential of this CPA and Cas 12a mediated lateral flow assay, the No. 4 quality control (QC) sample of the first Chongqing external quality control (EQC) collected in 2020 and No. 3 QC sample of the second national EQC collected in 2018 were then assessed by the proposed lateral flow assay. The results (Table 1) indicated that all samples were distinguished as MRSA, which matched with results published online. Subsequently, 202 pieces of secretions collected from 166 infected patients were tested by the proposed lateral flow assay again and culture identification of these secretions was carried out simultaneously for comparison. The characteristics of these patients and their secretions are shown in Table S3. (1) The median age of 166 patients was 48 years old (range, 9 ~ 81 years), and 59.6% of them were males; (2) 39.2% of patients came from orthopedics and trauma centers and 39.2% of them came from the center of otorhinolaryngology respectively; (3) According to the diagnosis, 80.1% of the secretions derived from open wounds, and the rest of them originated from closed wounds; (4) According to culture identification, the positive rate of clinical samples was 13.4%, of which 12 samples contained gram-positive bacteria, 11 were S. aureus (2 were MRSA), and 1 sample contained streptococcus.

The result of 202 samples analyzed by CPA and CAP-Cas 12a mediated lateral flow assay are illustrated in Table 2. Compared with the traditional culture identification method, the CPA and CPA-Cas 12 A mediated lateral flow assays had a higher checkout rates of 6.4% (13 samples were identified by RT-PCR showed in Fig. S3), and the reasons were summarized as follows: (1) CPA and CPA-Cas 12a assay adopted molecular diagnostic have high sensitivity by extracting characteristic genes from S. aureus and amplifying corresponding DNA; (2) Traditional culture identification adopted smear methods that can only “feel” surface of a sample, so the inoculation is incomplete and there is a certain degree of false-negatives; (3) Compared with smear inoculation method, we adopt homogeneous reaction system, so the bacteria on secretion swabs was released completely; (4) CPA and CPA-Cas 12a based lateral flow biosensors employing isothermal amplification strategy could significantly amplify the signal of bacterial target gene and then enhance the sensitivity of bacteria detection.

According to the propose of original design, CPA-Cas 12a mediated lateral flow assay employing double amplification strategy should have higher sensitivity and lower detection limit than CPA mediated lateral flow assay that only amplifies signal one time theoretically. However, in our study, the checkout rate of CPA-Cas 12a based lateral flow assay for clinical samples was not significantly improved compared to that of CPA method, which is not an expected result. The possible reason may be attributed to this it that the infective doses of pathogenic bacteria in real samples is usually ranging from 102 to 106 CFU mL− 1 [27, 28]. This much higher quantity of bacteria than the detection limit of CPA-Cas 12a based method does not fully exert the advantage of high sensitivity of this method, which leaves us the impression here that the detection performance of both CPA-Cas 12a and CPA based lateral flow biosensors was comparable. Although the advantage of CPA-Cas 12a based lateral flow biosensor was not completely showed here, its ability for clinical testing had been demonstrated successfully.

Table 2 Test results of actual samples

Performance of CPA assay accommodated in microfluidic devices

A microfluidic device had been designed to integrate the CPA based sensing strategy to simplify the operation process of multiple genes detection of S. aureus in clinical samples and realize this process automatically. The schematic illustration of the microfluidic device is shown in Fig. 6A, and its corresponding digital image is shown in Fig. 6B. The structure of microfluidic device is composed of three layers, which are polycarbonate (PC) membrane layer, channel layer, and response layer respectively. Figure 6C1 and 6C2 are microscopic images of the magnetic bead enrichment cell. Figure 6C3 is the microscopic images of the magnetic bead enrichment cell filled with magnetic beads. The magnetic beads randomly distributed well in the enrichment cell as shown in Fig. 6C. The adsorption of FITC-labeled DNA on magnetic beads is shown in Fig. 6C5 and the corresponding green fluorescence signal derived from FITC could be obviously observed in Fig. 6C6. It proved that homodisperse magnetic beads had successfully adsorbed with DNA extracted from bacteria.

We then explored the voltage dependent temperature of microheater (R = 10 Ω) as shown in Fig. 6E. The increasing in the voltage correspondingly increased the current linearly, and then the temperature of microheater. Figure 6D shows a captured infrared image of the microdevice mounted on the system where the microheater driven by a DC power controller was used to heat the CPA chamber. When the voltage of the microheater was set at 2.1 V, the temperature of CPA reaction cell E was steadily maintained at about 63.0 °C at the central top area while the heat diffused to surrounding area and the temperature there was slightly lower than its counterpart. Then we injected clinical positive samples into this microfluidic device as above-mentioned for analysis. These samples were determined as 3 pieces of MRSA and 10 pieces of S. aureus, which agreed with previous detection results.

Fig. 6
figure 6

Schematic diagram and characterization of microfluidic device. (A-B) Structure diagram and digital image of microfluidic device. (C) Micrograph of magnetic bead enrichment cell and characterization of DNA isolation. (D) Microheater characterized by IR camera. (E) Current and temperature of the microheater powered at different voltages

Guiding clinical medication

S. aureus is able to access the underlying tissues or bloodstream to cause infection when the cutaneous and mucosal barriers are disrupted [16]. The antibiotics is usually employed to cure disease related to S. aureus. However, the widespread use of spectrum antibiotics might lead to an increase in drug resistance, and their use should be under monitor for drug resistance during treatment. Therefore, early microbiological diagnosis of S. aureus infections and detection of drug resistance are essential for the efficient treatment of disease.

To verify the usefulness of our proposed assay for the guideline of clinical antibiotics medication, the detection results of secretions collected from patients were given back to their clinicians as well as some evidence for rational use of antibiotics. The clinical data and antibiotic usage of 13 S. aureus infected patients (P1-P13) were collected and shown in Table S4. 10 (76.92%) of infected patients were empirically administered antibiotics upon receipt of analysis results. Most selected antibiotics were penicillin or second-generation cephalosporin antibiotics combined with mupirocin or fusidic acid. This treatment regimen was used to treat infections by most gram-negative bacteria and gram-positive bacteria. In particular, the topical antibiotics of mupirocin and fusidic acid were especially effective for skin or wound infections caused by gram-positive bacteria.

Due to poor physical condition and severe infection of P1 and P9 patients, clinicians chose cefuroxime or piperacillin-sulbactam combined with vancomycin for anti-infection therapy at the early stage of infection, which covered most gram-negative and positive bacteria, and eventually achieved good efficacy. This demonstrated that the evidence of S. aureus infection we provided within 3 h was very important for further confirming the clinician’s choice of antibiotics.

In addition, there was a case of inconsistency between analysis results and clinical symptoms. The P11 patient had no obvious sign of MRSA infection, while MRSA was detected from its nasopharyngeal secretion. It might be attributed to colonization bacteria or sample contamination. S. aureus is found in the human commensal microbiota of the nasal mucosa in 20–40% of the general population[32].

Moreover, debridement and regular wound care are essential for recovery of infected patient. Each patient had received effective debridement, cleaning or changing dressing of infected site, which could effectively inhibit the growth of pathogenic bacteria. Although the P4 patient did not use antibiotics, the regular clean of necrotic tissue and wound in the nasal cavity is useful for treatment. The plantar infected patient (P8) was given antifungal treatment by doctor, but our result indicated that he was infected by S. aureus. Although there is a contradiction, the P8 patient still recovered well due to carefully daily debridement without changing his treatment regimen. Unfortunately, homogenization analysis of antibiotics therapeutic effect could not be implemented because different patients have diverse statuses. In a word, these results imply that the proposed assay for MRSA and S. aureus has a promising feature for antibiotic selection for physicians within 3 h.