What this is
- This research develops a new visual detection method for Yersinia pestis, the plague-causing bacterium.
- The method integrates technology with for sensitive and cost-effective detection.
- It aims to improve rapid diagnostics in resource-limited settings, offering a viable alternative to traditional methods.
Essence
- The study presents a novel visual detection platform for Yersinia pestis that combines and G4 DNAzyme technology, achieving 100% sensitivity and specificity in detecting the pathogen in simulated clinical samples.
Key takeaways
- The RCCD detection platform achieved a limit of detection (LOD) of 1 copy/reaction, indicating high sensitivity for detecting Yersinia pestis. This performance is critical for timely diagnosis in clinical settings.
- The method demonstrated 100% sensitivity and specificity when tested with DNA-spiked blood samples and uninfected controls, showcasing its reliability for plague detection.
- Visual differentiation between positive and negative samples was achieved; positive results appeared colorless while negative samples turned dark green, simplifying interpretation without complex equipment.
Caveats
- The study utilized simulated clinical samples due to the unavailability of actual patient samples, which may limit the generalizability of the findings.
- The long reaction time for the Cas12a-G4 colorimetric reaction may hinder rapid testing, necessitating further optimization for practical applications.
Definitions
- CRISPR/Cas12a: A gene-editing technology that can be used for precise DNA detection through its ability to cleave specific nucleic acid sequences.
- G-quadruplex DNAzyme: A DNA molecule that can catalyze reactions and is used in biosensing applications due to its stability and ability to produce color changes.
AI simplified
INTRODUCTION
Plague is a deadly infectious disease caused by Yersinia pestis, known for its rapid transmission and high fatality rate (1). Throughout history, three pandemics of plague have resulted in over 100 million deaths. According to statistics from the World Health Organization’s official website and reports from various countries, a total of 58,752 cases of plague were reported globally between 1987 and 2020, including 4,888 deaths (2–4). Since the 21st century, more than 95% of plague cases have been concentrated in Africa, with Congo, Madagascar, and Peru emerging as the three most heavily impacted countries (5, 6). In August 2017, a severe pulmonary plague epidemic broke out in Madagascar. During the period from 1 August to 26 November 2017, 2,417 cases were recorded, including confirmed, possible, and suspected cases, along with 209 deaths (7). The prevalence of plague remains a challenge to public health. Therefore, it is crucial to develop a rapid and sensitive diagnostic method for plague.
Traditional methods for detecting Y. pestis involve isolating and culturing the bacteria from clinical samples, conducting phage lysis assays, and utilizing serological detection methods that rely on identifying antibody-mediated F1 antigens, such as enzyme-linked immunosorbent assay (ELISA) (8), passive hemagglutination assay (PHA) (9), and fluorescent antibody (FA) assay (10). However, these conventional bacteriological detection techniques are prone to misjudgment with other bacteria in the Enterobacteriaceae family (11). Furthermore, they are time-consuming and require costly equipment. Consequently, there is an imperative need to develop a rapid, uncomplicated, and efficient diagnostic method for Y. pestis detection in resource-limited environments.
Nucleic acid detection methods offer a promising alternative for detecting and distinguishing Y. pestis, allowing for the differentiation of Yersinia species closely related to Y. pestis. Nucleic acid detection methods such as real-time fluorescent PCR (RT-PCR), recombinase polymerase amplification (RPA), recombinase-aided amplification (RAA), and loop-mediated isothermal amplification (LAMP) have played crucial roles in the diagnosis of infectious diseases. However, the practical application of pathogen detection puts forward a higher demand for detection simplicity and detection sensitivity. The clustered regularly interspaced short palindromic repeats (CRISPR) detection technology is a newly developed technique in recent years. During the COVID-19 pandemic, CRISPR detection technology underwent rapid development, with multiple COVID-19 virus detection reagents based on this technology being approved (12). The basic principle relies on the trans-cleavage characteristics of the Cas proteins in the CRISPR system. It involves designing crRNA that is complementary to the target nucleic acid sequence to be detected. The Cas protein binds with the crRNA to form a complex, which recognizes and binds to the target nucleic acid sequence through complementary pairing, thereby activating the trans-cleavage activity of the Cas protein in the system (13). This, in turn, cleaves the reporter molecules in the system that carry fluorescent reporters and quenchers, releasing detection signals for molecular diagnosis (14). The combination of RAA and CRISPR/Cas12a has been successfully used for the detection of various pathogens, such as hepatitis B virus (HBV) (15), severe acute respiratory syndrome coronavirus 2 variables (SARS-CoV-2 variables) (16, 17), African swine fever virus (ASFV) (18), Vibrio vulnificus (19), Vibrio parahaemolyticus (20), and Listeria monocytogenes (21).
DNAzyme is a class of DNA molecules with catalytic functions. Activated Cas12a can trans-cleave single-stranded DNA and inactivate the catalytic functions of DNAzyme. Therefore, it is suitable for combining Cas12a and DNAzyme to build a biosensor. A commonly used DNAzyme, G-quadruplex (G4), is an intricate structure formed by stacking single-stranded DNA rich in guanine and stabilized by monovalent cations like K+ and Na+ (22, 23). The G4 DNAzyme demonstrates exceptional thermodynamic resilience, maintaining structural integrity and catalytic activity across a wide thermal gradient, which ensures reliable chromogenic performance under diverse environmental conditions. Furthermore, the colorimetric reaction between G4 and hemin is more cost-effective than traditional signal reporting systems, such as fluorescence and lateral flow (LF) test strips, while generating a robust colorimetric signal that is easily visible to the naked eye. G4 is known for its simple synthesis, cost-effectiveness, and robust thermal stability in the design and development of highly sensitive molecular diagnostic methodologies. It has been used for signal transduction and target recognition and further developed into G4-based biosensors with easy operation and visualization capabilities (24–26). The interaction of G4 with specific fluorescent ligands can significantly amplify its fluorescence signal and be used as a fluorescent tracer in constructing highly sensitive fluorescent biosensors (27–30). The formation of DNAzymes by combining G4 with hemin has the characteristics of inducing color changes in substrates, catalyzing luminol-H2O2 chemiluminescence reactions, and quenching the fluorescence signals of nanoparticles. Based on these characteristics, G4 DNAzyme has been used to develop biosensing platforms for various detection methods, such as electrochemistry (31), chemiluminescence (32), fluorescence detection (33), and colorimetry (34). Therefore, it is supposed that combining Cas12a with G4 in a liquid detection system will greatly reduce the detection cost and improve the detection throughput of the CRISPR-based visual detection.
The aim of this study was to develop a sensitive, highly specific, rapid, and cost-effective Y. pestis detection method based on CRISPR/Cas12a and G4 DNAzyme technologies. Our results suggest that the RAA-CRISPR/Cas12a-DNAzyme (RCCD) detection platform holds promise for successful implementation in clinical environments.
MATERIALS AND METHODS
Materials and reagents
RAA nucleic acid amplification kit (basic version) and RAA nucleic acid amplification kit (fluorescent method) were purchased from Jiangsu Qitian Gene Biotechnology Co., Ltd. (Wuxi, China). LbCas12a nuclease was purchased from GenScript (Nanjing, China). NEBuffer r2.1 was purchased from New England Biolabs (Beijing, China). Hemin was purchased from Solarbio (Beijing, China). EL-ABTS Chromogenic Reagent kit (containing 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid), ABTS) and 3,3',5,5'-tetramethylbenzidine (TMB) Chromogen Solution (for ELISA) were purchased from Sangon (Shanghai, China). Dithiothreitol (DTT) was purchased from Beyotime (Shanghai, China). H2O2 was purchased from HANNO (Huainan, China). Plasmids, crRNAs, primers, reporters, and G4 ssDNA were chemically synthesized by Sangon company, and the DNA and RNA sequences used in the work are listed in Table 1. All other chemical reagents were purchased from Sangon company. Genomic DNAs of Y. pestis (Microtus 201) and other bacterial species (Table 2) were extracted from the corresponding strains using a QIAwave DNA Blood & Tissue Kit (QIAGEN, Hilden, Germany). The concentrations of the extracted DNAs were determined according to the absorbance at a wavelength of 260 nm measured by the microspectrophotometer (KAIAO, K5600). All genomic DNAs were diluted with elution buffer to 10 ng/µL. Fifteen blood samples were collected from blood donors in a hospital for this study. The use of these samples was reviewed and approved by the Ethics Committee of the Huadong Research Institute for Medicine and Biotechniques (approval no. 2022005).
| Names | Sequence (5′−3′) |
|---|---|
| CH57_3927-F2 | GCATAAGCTCTTGTAGTAACTCTGGCGATAT |
| CH57_3927-F3 | AACTCTGGCGATATTTTGTCGGGGTAGTAAT |
| CH57_3927-F6 | CTCTTGTAGTAACTCTGGCGATATTTTGTCGG |
| CH57_3927-R2 | CTCAACATGACGAGCAATTAGTGGTATGTGGC |
| CH57_3927-R3 | TAGTGGTATGTGGCGTTCCTATCGGTGAA |
| CH57_3927-R4 | TATTGAAACCTGCTTTCCTTATGCCTCTG |
| CH57_3927-R5 | ATGTGGCGTTCCTATCGGTGAAGGGAGTAT |
| CH57_3927-R6 | TCCTTATGCCTCTGCATTTTCTCGACCTTG |
| CH57_3927-R7 | TGAAAAATCAGGCATAACTCGGGTCAATAT |
| CH57_3927-R8 | GATCACATTGTTACTCAACATGACGAGCAA |
| CH57_3927-R9 | TTACTTACAGTACCCATTTTAAAGATCACA |
| CH57_3927-P1 | /i6FAMdT//idSp//iBHQ1dT/TAATTCGGCTGGAGTCAAAAGGAGGGGTTACTTCAAAAGTCCC |
| CH57_3927-crRNA1 | UAAUUUCUACUAAGUGUAGAUUCGGGGUAGUAAUAUUCCAGUGGUU |
| CH57_3927-crRNA2 | UAAUUUCUACUAAGUGUAGAUAUUCGGCUGGAGUCAAAAGGAGG |
| CH57_3927-crRNA3 | UAAUUUCUACUAAGUGUAGAUCUUCAAAAGUCCCUUUUAGUGUA |
| CH57_3927-crRNA4 | UAAUUUCUACUAAGUGUAGAUGUGUACUGACUUCAUUGGGGUUG |
| CH57_3927-crRNA5 | UAAUUUCUACUAAGUGUAGAUAAUAUUGACCCGAGUUAUGCCU |
| CH57_3927-crRNA6 | UAAUUUCUACUAAGUGUAGAUAACCAAUAUUCUGAAGCAAUUA |
| CH57_3927-crRNA7 | UAAUUUCUACUAAGUGUAGAUAAAUGGGUACUGUAAGUAAAGGG |
| CH57_3927-crRNA8 | UAAUUUCUACUAAGUGUAGAUCUAUAUCGCAAUGUGCCCCGAC |
| CH57_3927-crRNA9 | UAAUUUCUACUAAGUGUAGAUGAUAAUCUCGUCGGCAGUUUC |
| CH57_3927-crRNA10 | UAAUUUCUACUAAGUGUAGAUCCUGAAGUAGAGGAACCAGUGAU |
| CH57_3927-crRNA11 | UAAUUUCUACUAAGUGUAGAUAACUCAUGAAUAAUGCGUGGAU |
| CH57_3927-crRNA12 | UAAUUUCUACUAAGUGUAGAUCUAAUCGCUUUAACCGACGUA |
| FAM-12T-BHQ1 reporter | 5′6-FAM --BHQ1-3′TTTTTTTTTTTT |
| G-rich ssDNA-1 | CTGGGAGGGAGGGAGGGA |
| G-rich ssDNA-2 | TTAGGGTTAGGGTTAGGGTTAGGGTTA |
| G-rich ssDNA-3 | TTTGGGAAGGGCGGGTAGGGT |
| G-rich ssDNA-4 | TGGGTAGGGCGGGTTGGGAAA |
| G-rich ssDNA-5 | GGTTGGTGTGG |
| FAM-G-rich ssDNA-4-BHQ1 | 5′6-FAM --BHQ1-3′TGGGTAGGGCGGGTTGGGAAA |
| Species | Strains | Sources |
|---|---|---|
| Yersinia rohdei | 43380 | ATCC |
| Yersinia mollaretii | 43969 | ATCC |
| Yersinia kristensenii | 33638 | ATCC |
| Yersinia frederiksenii | 33641 | ATCC |
| Yersinia ruckeri | 29473 | ATCC |
| Yersinia intermedia | 29909 | ATCC |
| Yersinia pseudotuberculosis | 28933 | ATCC |
| Yersinia enterocolitica | 9610 | ATCC |
| Yersinia pestis | 201Microtus | Our laboratory |
| Escherichia coli | O157:H7 | Our laboratory |
| Staphylococcus aureus | 25923 | ATCC |
| Pseudomonas aeruginosa | 10145 | ATCC |
| Bacillus subtilis | 6051 | ATCC |
| Bacillus thuringiensis | 10792 | ATCC |
| Vibrio parahaemolyticus | 17802 | ATCC |
Specific gene sequences screening
To screen specific DNA fragments of Y. pestis, the genome sequences of all species within the genus Yersinia were collected and aligned in Mauve version 20150226 (The Darling lab at the University of Technology Sydney) (35). All the specific sequences of over 300 bp in Y. pestis were selected and further aligned with all the public genome sequences of Y. pestis strains using BLAST software online (http://blast.ncbi.nlm.nih.gov/↗) to ensure the sequence is conserved within the species. To test the out-of-genus specificity of the selected sequences, they were aligned with all the gene sequences from non-Yersinia species in GenBank. The sequence showing the highest specificity and conservation was selected as the target gene, chemically synthesized by Sangon Biotech Company, and linked to pUC57 plasmid to construct a positive template for further detection method development.
Cas12a-based detection method development
A series of crRNA sequences were designed according to the restricted PAM sequences for Cas12a and chemically synthesized by Sangon. Each crRNA featured a 21 bp universal sequence (5′-UAAUUUCUACUAAGUGUAGAU-3′) (16) and a complementary sequence to the target gene. The initial Cas12a-based detection system was formulated with 20 nM of LbCas12a, 100 nM of crRNA, 50 nM of FAM-TTTTTTTTTTTT-BHQ1 reporter, and the positive plasmids (1010 copies). All the components were mixed in 20 µL of 1 × NEB buffer 2.1 and incubated at 37°C. Real-time fluorescence signals were measured by a F1620 fluorescent reader (Qitian Gene Biotechnology Co., Ltd.) at 20-s intervals for an hour. A negative control using plasmid pUC57 as the template was conducted in each test. The dynamic change of fluorescence value with reaction time was plotted, and the slope was calculated. A higher slope indicated a higher detection efficiency. The optimal crRNA was selected with the highest slope ratio between the positive group and the negative group.
Optimization of Cas12a-G4 colorimetric reaction development
Five experimental conditions were optimized, including the concentrations of G4, Cas12a, and hemin, the types of color-developing substrates, and the sequence of G-rich ssDNA. First, the concentration ratio of Cas12a to G4 was optimized by testing various Cas12a concentrations (223 nM and 111 nM) and G4 concentrations (0.25 nM, 0.5 nM, and 1 nM). The colorimetric reaction involved testing different concentrations of Cas12a and G4, 100 nM crRNA, and the positive plasmids (1010 copies). All the components were mixed in 20 µL of 1× NEB buffer 2.1 and incubated at 37°C for 2 hours. After incubation, 1 µL of hemin (50 µM) and 2 µL of KCl (500 mM) were added to activate the peroxidase activity. Then, 50 µL of the EL-ABTS chromogenic reagent was added, and the mixture was incubated at room temperature for another 10 min. The color change of the solution was observed, and the absorbance was measured by a Spark microplate reader (TECAN, Männedorf, Switzerland). The optimal concentration of Cas12a and G4 was determined based on visual color changes and absorbance differences between the positive plasmid and plasmid-free control groups.
Subsequently, the hemin concentration and the type of chromogenic agent were optimized following the above experimental procedure. Specifically, experiments were conducted with varying hemin concentrations (2.5 µM, 5 µM, and 10 µM) and different chromogenic reagents (ABTS and TMB). The optimal experimental conditions were determined based on visual color observation and absorbance differences. Finally, under the optimal conditions established above, G-rich ssDNAs with various DNA sequences were tested to identify the G4 sequence with the best color rendering effect.
RAA assay development
To develop an RAA assay that can be combined with the Cas12a-G4 reaction, a series of RAA primers were designed on either side of the optimal crRNA site in the target sequence according to the manufacturer’s instruction. An RAA probe was also designed with FAM and BHQ1 labeled. An RAA nucleic acid amplification kit (fluorescent method) was used to evaluate the gene amplification efficiency of each primer pair. In the RAA reaction system, 4.2 µL of each primer pair (10 µM), 0.6 µL of RAA probe (10 µM), 1 µL of template, and 25 µL of buffer were mixed and complemented with ddH2O to 47.5 µL. The reaction mixture was then transferred into the tube containing the lyophilized RAA enzyme mix. Subsequently, 2.5 µL of magnesium acetate was added to initiate the reaction, and the tubes were placed into the B6100 Oscillation mixer (Qitian Gene Biotechnology Co., Ltd.) for pre-amplification for 4 min. The tubes were moved to the F1620 fluorescent reader for fluorescence measurements at 20-s intervals over 20 min at 37°C. The optimal primer pair was selected with the highest reaction efficiency.
RCCD detection assay development and preliminary evaluation
The optimized RAA assay was combined with the Cas12a-G4 reaction to construct a visual detection method. Briefly, the target fragment was amplified using the optimized RAA assay. Here, the RAA nucleic acid amplification kit (basic version) was used without the probe added. Then, various volumes of the amplification product were added to the optimized Cas12a-G4 colorimetric reaction system for colorimetric detection. Genomic DNAs of various Yersinia species were used as templates in the constructed RCCD detection assay for specificity evaluation. In addition, to evaluate the limit of detection (LOD) of the assay, serially diluted genomic DNA (with concentrations of 103, 102, 10, and 1 copies/µL) of Y. pestis in elution buffer was used as templates, and the minimum concentration that could be detected was the LOD. Elution buffer served as the template of negative control during both experiments. In the experiments above, each reaction was performed with at least two replicates.
RCCD detection assay evaluation using simulated clinical samples
The detection performance of the established assay was evaluated using simulated clinical samples due to the unavailability of clinical samples from Y. pestis-infected patients. Y. pestis genomic DNA was added to 200 µL of uninfected blood samples, achieving final concentrations of 30 copies/µL. DNA was subsequently extracted from the simulated clinical samples using the QIAamp DNA Blood & Tissue Kit, with a final elution volume of 200 µL. The RCCD detection assay was then performed on the extracted DNA. The original blood samples served as the negative controls, with each test conducted in duplicate.
Data processing and statistical analysis
The UV-visible absorption curve was plotted using Origin 2018 software (OriginLab, Northampton, MA, USA). The OD values among various groups were compared using one-way ANOVA or two-way ANOVA using GraphPad Prism 8.3.0 software (GraphPad Software, Boston, MA, USA), and the difference was considered significant with a P value of <0.05.
RESULTS
Detection principle
The principle of the RCCD detection platform is shown in Fig. 1. The target gene fragment is initially amplified through RAA amplification. The amplified target sequence binds to crRNA and activates the Cas12a nucleases for trans-cleave G4. Consequently, the cleaved G4 is unable to bind with hemin to exert peroxidase activity, thus impeding the catalysis of the ABTS2– colorimetric reaction. The color of the solution will remain transparent. Conversely, in the absence of the target gene, G4 will not be cleaved and can bind with hemin to form G4/hemin DNA enzyme, catalyzing the color reaction of ABTS2– and resulting in a color change from transparent to blue-green. The colorimetric signal could not only be detected by an ELISA reader but also observed by the naked eye.
Schematic illustration of the visual detection system ofcombining RAA, CRISPR/Cas12a, and G-quadruplex DNAzyme. Y. pestis
Specific gene sequence screening
A novel specific gene named CH57_3927 (nucleotides from 4338643 to 4339629 bp in Y. pestis A1122 references CP009840.1↗) of Y. pestis was successfully identified as the target. As shown in Fig. S1, this sequence was shared by all 81 strains of Y. pestis but not by any other Yersinia species or public sequences, indicating that it is an ideal target sequence for developing nucleic acid detection methods.
Optimal crRNA screening for Cas12a detection
Twelve crRNAs were designed and synthesized according to the sequence of CH57_3927 gene. The detection efficiency of each crRNA was individually tested. As shown in Fig. 2, a significant fluorescence increase was only observed in the groups using the positive plasmid as a template. Among the 12 crRNAs, crRNA2 showed the highest detection efficiency, with a slope ratio of 10.85 between the positive group and the negative control, representing superior sensitivity among the tested crRNAs (Table S1).
crRNA selection in CRISPR experiments. Each crRNA was individually tested with the CH57_3927 DNA. The negative control was treated with DEPC water instead of CH57_3927 DNA, with the same other components. All experimental data are represented as mean ± standard deviation (SD) of two technical replicates.
Optimization of Cas12a-G4 DNAzyme visualization system
The feasibility of integrating G4 DNAzyme with the Cas12a detection method was evaluated. First, the peroxidase activity of the G4/hemin DNAzyme and its ability to produce a color change for visual detection were tested. As shown in Fig. 3A, in the presence of ABTS2– and H2O2 simultaneously, the peroxidase activity of G4 and hemin can catalyze the oxidation of ABTS2–, producing the color change from transparent to dark green. The absorption peaks ranged from 400 nm to 420 nm. It is crucial to ensure compatibility between the Cas12a detection system and the G4 DNAzyme colorimetric system. Various components in the Cas12a detection system were separately added to the G4 DNAzyme colorimetric system to evaluate their ability to inhibit color change. As shown in Fig. 3B, Cas12a significantly inhibited the color development process, which was consistent with previous studies (34, 36). This interference could be attributed to the presence of DTT in Cas12a, a reducing agent employed to protect Cas12a’s activity, which can directly neutralize ABTS-free radicals (37). It is imperative to determine the optimal Cas12a nuclease concentration that does not affect the color development process. The concentrations of the Cas12a were optimized. As shown in Fig. 3C and D, the optimal concentration of Cas12a was 111 nM and the optimal concentration of G4 was 0.25 µM, the positive and negative groups showing the most distinct color contrast. After the optimization of Cas12a, the feasibility of the integrated Cas12a-G4 detection system was further evaluated by removing a single component. As shown in Fig. 3E, G4 was degraded into fragments in the presence of CH57_3927, Cas12a, and crRNA. Consequently, the fragmented G4 failed to engage with hemin to form the G4/hemin DNA enzyme with peroxidase activity, and a very light color change, as well as a low absorption value at 405 nm, was observed.
TMB and ABTS were evaluated for their effects as substrates in visual detection. As shown in Fig. 3F and G, the color difference between positive and negative groups was more obvious using ABTS as the substrate. In addition, various concentrations of hemin in the detection system were evaluated. The positive and negative groups can be clearly distinguished by color at a hemin concentration of 2.5 µM (Fig. 3F and G).
In addition, to achieve the best detection performance, five kinds of G-rich ssDNA sequences that have been reported to be able to form G4 structures were evaluated, and G-rich ssDNA-4 performed the best as shown in Fig. 3H. We further evaluated the efficiency of Cas12a in cleaving G-rich ssDNA-4. FAM and BHQ1-labelled G-rich ssDNA-4 was used as a fluorescent reporter in the Cas12a detection system. As shown in Fig. 3I, the positive group showed a significant fluorescence increase, indicating that G-rich ssDNA-4 could be cleaved efficiently by the activated Cas12a and was suitable for use in this visual detection system.
Subsequently, the LOD of the method was evaluated using a 10-fold series dilution of the positive plasmid (ranging in concentrations from 1010 to 104 copies/μL) as templates. The results showed that when the concentration of DNA template per reaction exceeded 109 copies, an obvious color difference between the experimental and control groups was observed, indicating that the LOD of this method was 109 copies/reaction (Fig. 3J).
Optimization of the Cas12a-G4 visual detection system. () UV-visible absorption curve of G4 catalyzed ABTS-HOcolor reaction. () Impact of 250 nM crRNA, 111 nM Cas12a, and 2 nM CH57_3927 (corresponding to the 20 µL CRISPR reaction system) on color development. The color reaction was conducted in a 100 µL solution containing 5 pmol G4. The control group lacked any CRISPR system components. () Various concentrations of Cas12a (223 nM in C and 111 nM in D) and G4 (0.25 µM, 0.5 µM, and 1 µM) on color development. () UV-visible absorbance values of G4 under various reaction conditions. () The influence of different concentrations of hemin (2.5 µM, 5 µM, and 10 µM) and types of color reagents (TMB in F and ABTS in G) on color development. () Impact of G4 with different sequences on color development. () Fluorescence intensity generated by the CRISPR/Cas12a system cleaving FAM-G-rich ssDNA-4-BHQ1. The negative control was treated with DEPC water instead of CH57_3927 DNA, with the same other components. Data are represented as the mean ± SD of two biological replicates. () Detection of a series of gradient dilutions of CH57_3927 using the CRISPR/Cas12a-G4 system. All experimental data in B-H are represented as mean ± standard deviation (SD) of two technical replicates. Differences among groups in B were analyzed by one-way ANOVA with Dunnett’s multiple comparisons test. Differences among groups in C, D, F, J, and H were analyzed by two-way ANOVA with Bonferroni’s multiple comparisons test. ns, not significant; *,< 0.05; **,< 0.01; ***,< 0.001; ****,< 0.0001. A B C and D E F and G H I J 2− 2 2 P P P P
RAA assay establishment
RAA assay was established for pre-amplification of the target sequences to improve the sensitivity of the Cas12a-G4 detection. We devised an RAA probe and a series of RAA primers both around the crRNA2 binding site. Through fluorescent RAA assays, we evaluated the amplification efficiency of various primer pairs. As a result, the primer pair F2/R6 showed the strongest fluorescence signal and the shortest positive judgment time, demonstrating its superior performance in target sequence amplification (Fig. 4A and B). The RAA amplification assay was established using this primer pair.
Optimization of the RAA and RCCD visual detection system. () Real-time fluorescence curves of 14 RAA primer combinations in RAA nucleic acid amplification experiments. () Reversal of color inhibition caused by DTT with HO. The colorimetric reaction was performed in a 100 µL solution comprising 50 nM G4, 500 nM hemin, and 2.5 mM DTT. () Color reaction of hemin with varied concentrations. Conduct gradient dilutions of DNA for RAA amplification and subsequently employ the amplification products for color reaction with diverse hemin concentrations. () Effect of different volumes of RAA amplification products (1 µL, 0.5 µL, and 0.25 µL) on color development. All experimental data in A through G are represented as mean ± standard deviation (SD) of two technical replicates. Data in E through G were analyzed by one-way ANOVA with Dunnett’s multiple comparisons test. **,< 0.01; ***,< 0.001; ****,< 0.0001. A and B C D E through G 2 2 P P P
Optimization of RCCD visualization system
To combine the RAA assay and the Cas12a-G4 visual detection assay, their compatibility was evaluated. The target sequences, pre-amplified by RAA, were added to the Cas12a-G4 visual detection system, and color change was observed. However, after the reaction, the color of both the positive and negative control groups did not change (), indicating that the reaction systems were incompatible. Fig. S2
We evaluated the ability of various components in the RAA system to inhibit the G4/hemin-based color development and found that DTT was the key inhibitor (Fig. 4C). We added various amounts of hydrogen peroxide (H2O2) to the DTT-inhibited system and found that the inhibitory effect could be reversed (Fig. 4C). The investigations revealed that the optimum color development enhancement was achieved when incorporating 35 mM of H2O2 (Fig. 4C).
Furthermore, we discovered that the quantity of hemin implemented also played a pivotal role in color development. High concentrations of hemin in the system could catalyze ABTS2– color development independently of G4. We optimized the hemin concentration again. As shown in Fig. 4D, in the absence of G4s, 75 µM and 18.75 µM of hemin could catalyze ABTS2– coloration, producing notable green products. Conversely, 7.5 µM of hemin rendered the solution almost colorless, leading us to select this concentration.
Lastly, we optimized the amount of RAA products added to the follow-up Cas12a-G4 DNAzyme colorimetric system (Fig. 4E through 4G). The results showed that introducing 0.25 µL of the RAA amplification products yielded the optimal color contrast between positive and negative samples.
Limit of detection and specificity analysis
To evaluate the LOD of this method for detecting Y. pestis DNA, we used serially diluted Y. pestis genomic DNA as templates, with the copy number ranging from 1 to 103 copies/µL. As a result, only the negative control group showed a dark green color, indicating the LOD of the established assay as 1 copy/reaction (Fig. 5A).
The specificity of this method was also evaluated, and the established method could distinguish the genomic DNA of Y. pestis and the other Yersinia species, with only the Y. pestis group being colorless and the other groups being dark green (Fig. 5B). Additionally, we conducted an expanded specificity evaluation by testing six common environmental and clinical bacterial strains, including Escherichia coli, Bacillus subtilis, Staphylococcus aureus, Vibrio parahaemolyticus, Bacillus thuringiensis, and Pseudomonas aeruginosa. The results demonstrated that this method specifically detected Y. pestis without cross-reacting with other bacterial species (Fig. S3).
Limit of detection and specificity ofdetection by the RCCD visualization system. () LOD of the RCCD visualization system in detectinggenomic DNA. () Specificity of the RCCD visualization system in detectinggenomic DNA. Mix bacteria is a mixture of the genomes from eight bacteria, excluding. All experimental data in A and B are represented as mean ± standard deviation (SD) of two technical replicates. Differences among groups in A were analyzed using one-way ANOVA with Dunnett’s multiple comparisons test. ****,< 0.0001. Y. pestis Y. pestis Y. pestis Y. pestis P A B
Sensitivity and specificity evaluation using simulated clinical samples
The sensitivity and specificity were further evaluated using 15 Y. pestis genomic DNA-spiked blood samples and 15 uninfected blood samples. With a concentration of 30 copies of Y. pestis genomic DNA per µL, all the simulated samples were detected to be positive, showing a sensitivity of 100% (15/15, Fig. 6). Meanwhile, all the uninfected blood samples were detected to be negative, showing a specificity of 100% (15/15, Fig. 6).
Detection ofin simulated clinical samples by RCCD detection assay. Tested samples,DNA-spiked blood samples; NTCs, uninfected blood samples as no-template controls. Each test was performed in duplicate. Y. pestis Y. pestis
DISCUSSION
Y. pestis, a highly pathogenic microorganism, is classified by the WHO as one of the potential bioterrorism agents. Although Y. pestis has been effectively controlled worldwide, cases continue to be reported in regions such as Congo, Madagascar, and Peru in recent years. These countries, with their limited economic and healthcare resources, require sensitive, rapid-response, and cost-effective detection methods to strengthen plague surveillance and prevention efforts, ensuring that large-scale epidemics do not occur again.
CRISPR detection is a novel technology, and its biggest advantage is the portability. It can be combined with various upstream amplification methods and downstream signal recognition systems. The typically used visual CRISPR detection method based on fluorescence-labeled ssDNA reporters and LF test strips greatly increases the detection cost (38). Also, it is not suitable for the detection of batch samples. In light of this, the integration of G4/hemin DNAzyme emerges as a promising and cost-effective alternative. Also, the experimental results showed good compatibility between the two systems. However, the sensitivity of CRISPR detection is low, and it is necessary to combine nucleic acid amplification technology to improve sensitivity (39). Here, we observed that a distinct signal was only discernible when the copy number of the target gene exceeded 109 copies within the reaction system. A step of pre-amplification was necessary. However, our initial evaluation showed that the RAA system inhibited the G4/hemin DNAzyme reaction and was not suitable for direct use. Further analysis proved that the reducing agent DTT in the RAA system was the key inhibitor. Therefore, in the follow-up test, we added the oxidizing agent H2O2 to conquer the problem successfully. Finally, the introduction of RAA significantly improved the sensitivity, with LOD as low as 1 copy/reaction. In our study, we integrated CRISPR-Cas12a, RAA amplification, and G4 DNA enzyme to establish an advanced ELISA-like visual detection method. Compared with traditional fluorescence detection and lateral flow detection, this method has the advantages of not relying on fluorescence detection equipment, low cost, and batch detection.
The Cas12a-G4 visual detection system exhibits an inverse concentration-dependent response within the range of 108 to 1010 template copies per reaction, wherein increasing template copies result in progressively attenuated chromogenic signals, indicating the quantitative or semi-quantitative analysis potential of the system (Fig. 3J). However, this concentration-dependent trend disappears upon integration of the RAA reaction (Fig. 4E through G), likely due to signal saturation caused by the high amplification efficiency of RAA, which impairs the quantitative capability of the system. To achieve quantitative or semi-quantitative detection, a two-stage optimization strategy is required: first, defining the linear dynamic range of the colorimetric system; second, adjusting the RAA amplification parameters (e.g., primer concentration, reaction duration) to ensure the amplified product remains within this quantifiable range. Substantial experimental optimization is necessary to implement this process.
In nucleic acid detection, the target sequence is one of the key factors that determine detection specificity. In the detection of Y. pestis, screening specific gene sequences for distinguishing Y. pestis and other pathogenic bacteria within the same genus, such as Y. pseudotuberculosis and Y. enterocolitica, is a challenge due to the high genomic sequence similarity. At least 97% of sequence homology between Y. pestis and Y. pseudotuberculosis is found in 2,976 genes (40). Specific genes like pla, pst, and caf1 were used as specific identifiers for Y. pestis in previous studies (41–43). However, the pla gene has been detected in other bacteria like Citrobacter koseri (41) and Escherichia coli (42), potentially leading to false-negative results when used as a detection target. Additionally, caf1 and pst are situated on plasmids pMT1 and pPCP1 (43), which are not present in all Y. pestis strains. Therefore, more dependable and specific molecular targets are needed to help researchers accurately identify Y. pestis and distinguish it from closely related species.
In our study, we compared and analyzed the homology of genome sequences of all Yersinia species using Mauve software. We divided the sequence of each reference genome into millions of fragments for comparative analysis between different strains. Through an in-depth comparison within Y. pestis strains, we screened a specific gene CH57_3927 that is consistently present in all Y. pestis strain genomes and only exists in Y. pestis. The established detection method based on this sequence did not recognize other Yersinia species except Y. pestis, proving its specificity and a reliable molecular marker, which can effectively distinguish Y. pestis from closely related species. As far as we know, this is the first time CH57_3927 has been used for Y. pestis detection, which provides a valuable reference for detecting Y. pestis.
One distinctive feature of this detection system is that the positive test result appears colorless, while the negative one shows a dark green color. During the experiment, failure to add the correct reagents or degradation of reagents may result in false positive outcomes. Under such circumstances, using internal quality control may be a very good solution. However, in a single colorimetric system, internal quality control cannot be compatible with the samples in a single reaction system. Therefore, we recommend incorporating positive and negative controls into the experimental system to address this issue. In addition to improving the objectivity of result interpretation, colorimetric readouts can be systematically analyzed using calibrated reference cards or smartphone-based image analysis applications. This approach effectively eliminates ambient light interference and minimizes inter-operator interpretation variances through standardized digital/analog quantification protocols.
In theory, the RCCD platform can be easily adapted for the detection of any target pathogen by simply redesigning primers and crRNAs, without altering the core mechanism. All other reaction steps and components remain unchanged, making the platform highly versatile. To validate the consistency and stability of the RCCD platform, multiple experimental replicates were performed with reagents from distinct production batches (e.g., G4, H₂O₂, DTT;), demonstrating robust reproducibility across independent tests. Table S2
There are some limitations in the present study. One limitation is the long reaction time requirement for the Cas12a-G4 colorimetric reaction. This problem likely arises from the relatively dense structure of G4, which impedes cleavage efficiency and extends the reaction time. To address this, we are exploring the use of split G-quadruplex and G-triplex structures as alternatives to G4. Another limitation of this study is the lack of clinical sample validation. Due to the unavailability of clinical samples from Y. pestis-infected patients, we used simulated clinical samples. However, further investigation is needed to determine whether this approach can be effectively applied to the detection of other complex sample matrices to evaluate its specificity and sensitivity. Such validation is critical for assessing the potential of this method as a reliable diagnostic tool in clinical settings.
CONCLUSIONS
In the present study, we established an advanced ELISA-like visual detection method that integrates CRISPR-Cas12a, RAA amplification, and G4 DNAzyme for cost-effective and highly sensitive detection of Y. pestis. Except for the visualization detection as well as good specificity and sensitivity, the established method has a lower cost and is suitable for batch sample detection like ELISA. Moreover, it does not require complex instruments, which is convenient for rapid and on-site screening of plague outbreaks, enabling effective support for plague detection, prevention, and control at primary-level medical and health care institutions.
ACKNOWLEDGMENTS
This work was financially supported by the Medical Science and Technology Project (JK2023gk002).
Contributor Information
Yuexi Li, Email: liyxi2007@126.com.
Yong Qi, Email: qslark@126.com.
Erin McElvania, Endeavor Health, Evanston, Illinois, USA.
SUPPLEMENTAL MATERIAL
The following material is available online at https://doi.org/10.1128/jcm.00274-25↗.
ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.