A CRISPR–Cas12a and Quantum‐Dot Lateral‐Flow Assay for Rapid Species‐Level Detection of Trichophyton rubrum

ABSTRACT

Background

Accurate species‐level diagnosis of dermatophytes, particularly Trichophyton rubrum, is important for guiding targeted antifungal therapy and improving the management of recurrent or atypical superficial mycoses. However, routine microscopy has only moderate sensitivity and cannot resolve fungal species; fungal culture is time‐consuming, and real‐time quantitative PCR (qPCR) requires costly instrumentation.

Objectives

To develop and clinically evaluate CRI‐RUB, a CRISPR–Cas12a–based assay integrating recombinase‐aided amplification (RAA) and a quantum‐dot (QD) fluorescent lateral‐flow strip for rapid species‐level detection of T. rubrum .

Patients/Methods

Analytical sensitivity and specificity were assessed using plasmid standards and a panel of common cutaneous fungi. A total of 140 clinical specimens were tested by both CRI‐RUB and TaqMan qPCR. Discrepant or grey‐zone qPCR results (Ct 33–34), together with a subset of concordant cases, were further arbitrated by internal transcribed spacer (ITS) sequencing to establish a composite reference standard.

Results

CRI‐RUB achieved a visual limit of detection of 2.3 × 10¹ copies/μL and showed no cross‐reactivity with other common cutaneous fungi. Among the 140 clinical specimens, CRI‐RUB showed a sensitivity of 96.1% (73/76), specificity of 100.0% (64/64), positive predictive value of 100.0% (73/73), negative predictive value of 95.5% (64/67) and Cohen's κ of 0.96 compared with the composite reference standard, indicating excellent concordance. The complete workflow required approximately 85 min.

Conclusions

CRI‐RUB provides a rapid, sensitive, and specific approach for species‐level detection of T. rubrum . By combining CRISPR–Cas12a detection with QD‐based lateral‐flow readout, this assay shows potential for near‐patient or point‐of‐care dermatophyte diagnosis.

1. Introduction

Dermatophyte infections are among the most common superficial mycoses worldwide, affecting an estimated 20%–25% of the population [1]. Among the etiologic agents, Trichophyton rubrum is the leading pathogen and causes a spectrum of manifestations, including tinea pedis, onychomycosis, tinea corporis and tinea cruris [2].

Accurate species‐level identification of dermatophytes, particularly Trichophyton rubrum, is therefore critical for guiding targeted antifungal therapy, reducing inappropriate use of broad‐spectrum agents and improving outcomes in patients with recurrent or atypical superficial mycoses. Clinical studies have shown that the optimal choice of systemic antifungal agent can depend on the infecting species; for example, terbinafine is more effective for Trichophyton infections, whereas griseofulvin is superior for Microsporum infections in tinea capitis [3, 4].

In addition, the recent emergence of terbinafine‐resistant Trichophyton species, including Trichophyton indotineae and resistant T. rubrum isolates, has highlighted antifungal‐resistant dermatophytosis as an emerging challenge that often presents with chronic, extensive, or recurrent disease [5, 6]. Despite the therapeutic importance of species‐level diagnosis, current diagnostic methods for dermatophyte infections have notable limitations: direct microscopy is operator‐dependent with suboptimal sensitivity; fungal culture, although highly specific, is time‐consuming, typically requiring 2–4 weeks to yield results [7]; and real‐time quantitative PCR (qPCR), while offering good sensitivity and specificity, relies on costly instrumentation and specialized laboratory infrastructure, limiting its use in resource‐limited settings or point‐of‐care testing (POCT) scenarios [8, 9, 10]. Therefore, there is an urgent need for a rapid, sensitive, highly specific and easy to use assay that can be deployed in both clinical laboratories and field or point‐of‐care settings to provide timely species‐level results.

In recent years, CRISPR–Cas systems particularly CRISPR–Cas12a—have become powerful tools for nucleic acid detection owing to their high specificity, programmability and target‐activated collateral (trans) cleavage activity [11, 12, 13]. When combined with isothermal amplification methods such as recombinase‐aided amplification (RAA), loop‐mediated isothermal amplification (LAMP) or rolling‐circle amplification (RCA), CRISPR–Cas12a generates detectable signals rapidly without a thermocycler, making it well suited to point‐of‐care testing (POCT) settings [14, 15, 16, 17, 18]. Meanwhile, quantum‐dot fluorescent lateral‐flow readouts provide higher sensitivity, clearer visualization and the potential for quantitative analysis compared with conventional gold‐nanoparticle colorimetric strips [19, 20, 21, 22]. Previous studies have demonstrated the feasibility of CRISPR‐based assays for a range of pathogens, including viruses, bacteria, and several fungal agents [23, 24, 25, 26, 27], yet applications to dermatophyte detection remain limited. Existing work on dermatophyte diagnostics has largely focused on broad‐range targets, has lacked highly specific species‐level platforms, or has relied on large laboratory instruments, thereby limiting clinical translation [28, 29]. In particular, there is still no clinically validated CRISPR‐based POCT assay that enables rapid species‐level identification of T. rubrum directly from dermatology specimens.

In this study, we developed and optimized a novel point‐of‐care testing (POCT) platform that integrates CRISPR–Cas12a, a quantum dot–based fluorescent lateral‐flow strip, and recombinase‐aided isothermal amplification (RAA), enabling rapid and accurate detection of T. rubrum and allowing systematic evaluation of its sensitivity, specificity and practical diagnostic utility in clinical specimens.

2. Materials and Methods

2.1. Reagents

Except for the TaqMan qPCR primers and probe, which were adopted from previously published studies, all other primers were designed de novo using a combination of the Oligo 7 and Primer3Plus online tools. Target gene sequences were retrieved from the NCBI Gene database and compared using NCBI BLAST. All nucleotide sequences used in this study are listed in Table S2. Primers, probes, and plasmids were synthesized by Sangon Biotech (Shanghai, China), and crRNAs and Cas12a protein were provided by Guangzhou EDITGene Technology Co. Ltd. (Guangzhou, China). The recombinase‐aided amplification (RAA) kit was provided by Jiangsu Qitian Gene Biotechnology Co. Ltd. (Jiangsu, China). The lysis buffer, extraction buffer, and other buffers were prepared by Shanghai Xin Yu Biotech Co. Ltd. (Shanghai, China); quantum dots, the fluorescence reader, and lateral‐flow strip assembly were supplied by Shanghai Kundao Biotechnology Co. Ltd. (Shanghai, China). Streptavidin was obtained from Sigma‐Aldrich (St. Louis, MO, USA). Goat anti‐mouse IgG and anti‐FAM antibodies were purchased from Cell Signalling Technology (Danvers, MA, USA). GoTaq Master Mixes were obtained from Promega (Madison, WI, USA). The PCR thermocycler was supplied by Thermo Fisher Scientific (Shanghai) Co. Ltd, the Roche LightCycler 480 instrument was purchased from Roche Diagnostics (Basel, Switzerland), and the UV lamp was manufactured by Philips (Amsterdam, the Netherlands).

2.2. Sample Collection and Preparation

This study was approved by the Shanghai Dermatology Clinical Trial Ethics Committee (approval number 202204). A total of 140 clinical specimens from patients with suspected superficial dermatophyte infection were collected. Fungal cell‐wall disruption during DNA extraction followed the optimized protocol described by Tian et al. [30]. Briefly, 0.1 g acid‐washed quartz sand and 300 μL buffer were added to each specimen and shaken at 70 Hz for 60 s, five times in total. Then 100 μL lysis buffer preheated to 35°C was added, the mixture was vortexed for 5 s, and briefly centrifuged. The tube was secured with a safety clamp and heated in a 600 W microwave oven for a total of 105 s in seven 15‐s intervals, with thorough mixing after each heating and brief venting to release pressure. After microwaving, 400 μL extraction buffer preheated to 65°C was added and mixed for 5 s, followed by addition of 600 μL Tris‐saturated phenol–chloroform for extraction. The suspension was vortexed for 5 s and centrifuged at 12,000 rpm for 5 min, and the aqueous phase was transferred to a sterile 1.5‐mL microcentrifuge tube. An equal volume of isopropanol was then added to precipitate the DNA; the pellet was collected by centrifugation and washed once with 70% ethanol. The residual ethanol was removed, and the pellet was air‐dried at room temperature until fully dry. Finally, the DNA was resuspended in 30 μL TE buffer and stored at −20°C until use.

2.3. Design and Amplification of Target DNA Fragments

Because the internal transcribed spacer (ITS) region exhibits high interspecific variability among fungi, it is commonly used for species identification [31]. Accordingly, ITS sequences for Trichophyton rubrum and other clinically common fungi were retrieved from the GenBank database. Sequences were compared in DNAMAN to identify hypervariable ITS regions that distinguish T. rubrum from other fungal species. Isothermal amplification primers were then designed in‐house targeting these regions and validated with NCBI Primer‐BLAST [32], demonstrating high specificity; these regions were consequently selected as targets for typing. All primers, probes, and crRNAs used in this study are summarized in Table S2.

For the recombinase‐aided amplification (RAA) assay, each 50 μL reaction contained 25 μL reaction buffer V, 2 μL forward primer (10 μM), 2 μL reverse primer (10 μM), 8 μL extracted DNA, 8 μL nuclease‐free water, and 5 μL magnesium acetate. Reactions were incubated at 37°C for 40 min. Amplicons were analysed by agarose gel electrophoresis, gel‐excised for DNA sequencing, and the resulting sequences were compared using NCBI BLAST.

2.4. TaqMan qPCR

TaqMan qPCR was performed in a 20 μL reaction containing 10 μL AceQ Probe Mix V2, 0.4 μL each of forward and reverse primers (10 μM), 0.2 μL TaqMan probe (10 μM), 1 μL sample eluate, and 8 μL nuclease‐free water (ddH₂O). Thermocycling comprised 40 cycles with fluorescence acquisition at 60°C on a Roche LightCycler 480 instrument. Results were interpreted as follows: Ct < 33 with a smooth sigmoidal amplification curve was considered positive; Ct 33–34 was considered indeterminate and retested; Ct > 34 was considered negative.

2.5. Fabrication of Lateral‐Flow Chromatographic Strips

The strip consisted of a PVC backing card, a nitrocellulose membrane, and a glass‐fibre sample pad. Sample pads (20 mm width) were pretreated with buffer (0.05 M Tris‐HCl, 2% [w/v] PEG‐8000, 1.5% [v/v] Tween‐20; pH 8.0) until fully saturated (~5 min) and then dried at 50°C for 24 h. Streptavidin (2.5 mg/mL) and goat anti‐mouse IgG (0.8 mg/mL) were striped onto the nitrocellulose to form the test (T) and control (C) lines, respectively, with a 5 mm spacing. Components were laminated onto the PVC backing and the membrane laminate was slit into 5‐mm‐wide strips. Finished strips were conditioned at 37°C, stored desiccated, and assembled into cassettes immediately before use.

2.6. PCR

PCR was performed in a 20 μL reaction containing 10 μL of 2× PCR premix, 0.5 μL each of forward and reverse primers (10 μM), 2 μL of template (eluted DNA), and nuclease‐free water (ddH₂O) to a final volume of 20 μL. Thermocycling conditions were: initial denaturation at 95°C for 5 min; 40 cycles of 95°C for 30 s (denaturation), 58°C for 30 s (annealing), and 72°C for 1 min (extension); followed by a final extension at 72°C for 5 min and a hold at 4°C. Amplicons were analysed by agarose gel electrophoresis, gel‐purified for DNA sequencing, and compared using NCBI BLAST.

2.7. Preparation of Fluorescent Quantum Dots

Based on our previous protocols [33], 100 μL of QDs (10 mg/mL) was diluted with 200 μL phosphate buffer (PB; 0.02 M, pH 6) and activated with 6 μL of 1‐ethyl‐3‐(3‐dimethylaminopropyl)carbodiimide hydrochloride (EDC·HCl; 20 mg/mL). After brief bath sonication and incubation at 37°C for 30 min, the activated QDs were collected by centrifugation (15,000 rpm, 10 min, 4°C) and reconstituted in 100 μL PB (0.02 M, pH 6). The suspension was then mixed with 100 μL mouse anti‐FAM antibody (1 mg/mL) and 100 μL PB (0.02 M, pH 6). Following a gentle spin (800 rpm, 5 min, 4°C), the anti‐FAM–QDs were resuspended in 300 μL of 0.5% casein and incubated overnight. The fluorescent QD–antibody conjugates were stored at 4°C.

2.8. CRISPR–Cas12a Assay

Each 30 μL CRISPR–Cas12a reaction contained 3 μL of 10×reaction buffer, 1 μL Cas12a (1 μM), 1.2 μL ssDNA reporter (10 nM), 1 μL crRNA (10 μM), 1 μL RAA amplicon, and nuclease‐free water (ddH₂O) to a final volume of 30 μL. Reactions were incubated at 37°C for 30 min, after which 30 μL of the quantum dot–antibody conjugate (QDs–Ab) was added to the same tube.

2.9. Statistical Analysis and Visualization

Naked‐eye readouts were acquired under a UV lamp (320–400 nm; 14.2 mJ/cm²). Statistical analyses and visualization were performed in GraphPad Prism 10.1.2, with data reported as mean ± standard deviation (mean ± SD). Between‐group differences were assessed using one‐way ANOVA for continuous variables and Fisher's exact test for categorical variables in SPSS 27, with two‐sided p < 0.05 considered statistically significant. Positive and negative predictive values (PPV, NPV) and receiver‐operating characteristic (ROC) curves with 95% confidence intervals were generated in Prism. Figures were prepared in BioRender and Adobe Illustrator 2023.

3. Results

3.1. Workflow of the CRI‐RUB Assay

We developed the CRISPR‐Cas12a‐rubrum (CRI‐RUB) assay using an anti‐FAM–quantum‐dot conjugate (QDs–Ab) as the tracer. The reporter is an ssDNA labelled with FAM and biotin at opposite ends (Figure 1). The lateral‐flow strip was assembled as described in Section 2.5: streptavidin was dispensed on the control (C) line to capture biotinylated fragments, and goat anti‐mouse IgG on the test (T) line to capture the mouse anti‐FAM–QDs–Ab, with a 5 mm spacing between lines. After specimen collection (e.g., nail debris), fungal cells were lysed by microwave treatment, DNA was extracted, and recombinase‐aided amplification (RAA) generated target amplicons. Amplicons and tracer were then added to the CRISPR‐Cas12a reaction. In the presence of target DNA, Cas12a activation induces cis and trans cleavage of the reporter, preventing tethering of the QD tracer to the C line; the free anti‐FAM–QDs–Ab migrates and is captured at the T line, yielding a detectable fluorescent signal, while biotinylated fragments are retained at the C line. In the absence of target, the intact FAM–biotin reporter bridges the anti‐FAM–QDs–Ab to the C line via the biotin–streptavidin interaction, producing a bright C line and a negative T line. The presence of T. rubrum DNA is thus determined from the relative fluorescence at the T and C lines.

FIGURE 1.

Workflow and readout principle of the CRISPR–Cas12a–coupled lateral flow assay (LFA) for rapid detection of T. rubrum . Nail and skin‐scale specimens were collected from suspected lesions, followed by microwave lysis and DNA extraction. The DNA eluate was amplified by recombinase‐aided amplification (RAA). The resulting amplicons were introduced into the CRI‐RUB reaction, in which the ssDNA reporter is cleaved in positive samples, whereas no cleavage occurs in negative samples. The quantum‐dot tracer was then added, and the mixture was applied to a lateral‐flow strip. Results can be read by eye under UV illumination, and fluorescence intensity can be quantified with a fluorescence reader.

3.2. Validation of the CRI‐RUB Assay

To enable subsequent evaluation of the performance and species‐level specificity of the CRI‐RUB assay for Trichophyton rubrum infection, we first aligned the ITS region of T. rubrum with ITS sequences from other clinically common cutaneous fungi and designed T. rubrum –specific isothermal amplification primers on this basis. The fungal species included in this comparison and their corresponding ITS‐region NCBI accession numbers are listed in Table S1.

As described in Section 2.2, we compared a microwave lysis–based extraction method with a commercial DNA extraction kit using clinical specimens. As shown in Figure 2c, the universal primers ITS1 and ITS4 yielded clear amplification bands from DNA extracted by the microwave method, whereas no distinct specific bands were obtained from DNA extracted with the commercial kit.

FIGURE 2.

Target design and primer validation of the CRI‐RUB assay. (A) Representative clinical lesions from patients with suspected dermatophyte infection and the corresponding fluorescence microscopy images, showing bright blue fungal hyphae in the stratum corneum. (B) Schematic illustration of crRNA binding sites within the T. rubrum ITS target region. The upper sequences show the double‐stranded DNA targets containing the PAM motifs (green), with key bases in the original ITS sequence underlined, and the lower sequences depict the complementary crRNA‐1 and crRNA‐2 guide regions (purple). (C) Comparison of PCR amplification from DNA extracted by different methods. M, DNA marker; lanes 1–3, DNA extracted from clinical specimens using the microwave lysis method and amplified with ITS1/ITS4, yielding clear specific bands; lanes 4–6, DNA extracted using a commercial kit, showing markedly weaker or absent amplification bands. (D) Agarose gel verification of RUB‐RAA primer pairs. M, DNA marker; lanes 1–3, amplification products obtained with the RUB‐RAA‐1 primer pair; lanes 4–6, amplification products obtained with the RUB‐RAA‐2 primer pair, all generating specific bands of the expected size.

Subsequently, the designed RUB‐RAA1 and RUB‐RAA2 isothermal amplification primers were used to perform RAA on microwave‐extracted clinical DNA (Figure 2d), successfully generating T. rubrum –specific amplicons. All gel bands were subjected to DNA sequencing, and the resulting sequences were confirmed by NCBI BLAST analysis. On the basis of these validated RAA amplicon sequences, we designed the crRNAs required for the CRISPR module of the assay, as illustrated in Figure 2b.

3.3. Optimization of the CRI‐RUB Assay Conditions

To further improve the performance of the CRI‐RUB detection platform and shorten the overall turnaround time, we first systematically optimized the working concentrations of the fluorescent quantum dots and the reporter probe. Different concentrations of the quantum dot–antibody conjugate (QDs–Ab) were added to the lateral‐flow assay, and the fluorescence intensity at the T line was compared with the background signal. Under the premise of maintaining a normal C‐line signal, we selected the QDs–Ab concentration that produced a clearly distinguishable T line with minimal background as the optimal working concentration. As shown in Figure 3a, a QDs–Ab concentration of 0.3 mg/mL yielded a sharp T line with the lowest background; therefore, 0.3 mg/mL was adopted as the optimal concentration of QDs–Ab. On this basis, gradient concentrations of the ssDNA reporter were tested under the optimized QDs–Ab condition to evaluate their effects on the C‐line signal and background. As shown in Figure 3c, a final reporter concentration of 10 nM was sufficient to generate a stable and clear C line without excessive background or signal saturation, and was thus selected as the optimal working concentration of the reporter probe.

FIGURE 3.

Optimization of CRI‐RUB assay conditions and component‐omission controls. (A) Titration of the working concentration of the quantum dot–antibody conjugate (QDs–Ab). Numbers above the strips indicate QDs–Ab concentrations (mg/mL). T and C lines are marked to identify the concentration that yields a clear T line with minimal background. (B) Optimization of Cas12a reaction time. Numbers above the strips indicate the Cas12a incubation time (min). The line graph on the right shows the fitted fluorescence ratio (T/C) at different reaction times, used to determine the optimal cleavage time. (C) Optimization of the ssDNA reporter probe concentration. Numbers above the strips indicate the final reporter concentration (nM); NC denotes the negative control. The goal was to select a probe concentration that produces a stable, clear C line without excessive background. (D) Optimization of the readout time after sample application to the lateral‐flow strip. Numbers above the strips indicate the time (min) elapsed after loading the reaction mixture. The line graph on the right depicts the T/C fluorescence ratio as a function of time, used to define an appropriate readout window. (E) Component‐omission control experiments. The table on the left summarizes the presence (+) or absence (−) of Cas12a, crRNA, reporter probe, and DNA target in reactions 1–5. The corresponding strips on the right show that omission of Cas12a, crRNA or the DNA target abolishes or markedly reduces the T‐line signal, whereas both the complete reaction (all components present) and the reaction lacking the reporter probe exhibit a characteristic T/C band pattern (in the latter case, the QD conjugate is no longer retained at the C line and is predominantly captured at the T line).

We next evaluated the Cas12a reaction time and the time window for lateral‐flow readout. The reaction mixture was applied to the strip after different incubation times, and T/C line signals were recorded at various time points. As shown in Figure 3b,d, the T/C fluorescence ratio reached a plateau after approximately 25 min of Cas12a incubation, and reading the strip 10–15 min after sample application provided reliable discrimination between positive and negative samples while minimizing the total assay time.

In addition, to verify the requirement of each key component and to exclude nonspecific band formation, we performed component‐omission control experiments in which Cas12a, crRNA, RAA amplicons, or the ssDNA reporter were individually omitted while the remaining components were retained. As shown in Figure 3e, omission of Cas12a, crRNA, or the RAA product abolished the specific T‐line signal, whereas the complete reaction system produced the characteristic T/C band pattern. Notably, in the absence of the reporter probe, the QDs–Ab could no longer be immobilized at the C line via the reporter–biotin–streptavidin axis and instead migrated with the flow and was captured by antibodies at the T line, resulting in a visible T‐line signal accompanied by a markedly weakened or absent C line. These findings indicate that, in the intact assay system, positive signals arise from target‐triggered CRISPR–Cas12a–mediated cleavage of the reporter and its redistribution between the T and C lines, rather than from nonspecific binding of the quantum dots or the lateral‐flow strip itself, thereby further supporting the rationality and specificity of the CRI‐RUB platform design.

3.4. Analytical Sensitivity and Specificity of CRI‐RUB

A plasmid containing the Trichophyton rubrum ITS sequence was constructed as a reference standard, and the copy number concentration was calculated using Formula (1). The assay achieved a visual limit of detection of 2.3 × 10¹ copies/μL for T. rubrum (Figure 4a), with corresponding fluorescence intensities quantified as shown. For analytical specificity, plasmids bearing ITS regions from other clinically common cutaneous fungi were tested; all non‐ T. rubrum targets were negative, while T. rubrum was positive (Figure 4b). In parallel, TaqMan qPCR exhibited a robust standard curve with good linearity across the tested plasmid dilutions (Figure 4c), and its limit of detection was on the same order of magnitude as that of the CRI‐RUB assay (10¹ copies/μL).

$$\text{Copy number}\left(\text{copies}/\mathrm{\mu L}\right)=\left(\mathrm{C}\times 6.022\times {10}^{23}\right)/\left(\mathrm{L}\times 660\times {10}^9\right)$$ (1)

In Formula (1), C is the plasmid DNA concentration (ng/μL), L is the plasmid length (bp), and 660 g/mol is the average molecular weight of one base pair of double‐stranded DNA.

FIGURE 4.

Analytical sensitivity and specificity of the CRI‐RUB assay. (A) Analytical sensitivity of the assay. Left: Naked‐eye appearance of lateral‐flow strips after testing plasmid standards with serial copy numbers (10⁷–10⁰ copies/μL) using CRI‐RUB; NC, negative control. Right: Bar graph showing the corresponding quantified T/C fluorescence ratios. (B) Analytical specificity of the assay. Left: Lateral‐flow strips obtained after testing DNA from clinically common superficial fungi (T. rubrum, T. mentagrophytes, M. canis, E. floccosum, T. tonsurans, T. violaceum, and T. soudanense) and a negative control (NC); species names are indicated below the strips. Right: Bar graph showing the quantified T/C fluorescence ratios for each species, demonstrating a marked positive signal only for T. rubrum. (C) TaqMan qPCR standard curve. Ct values were plotted against log₁₀ (copies/μL) of the plasmid standards to generate the calibration curve. The slope, coefficient of determination (R2), and amplification efficiency are shown and were used to evaluate the analytical performance and sensitivity of the qPCR assay.

3.5. Clinical Performance on Patient Specimens

Representative clinical lesions and fluorescence microscopy findings from patients with suspected dermatophyte infection are shown in Figure 2a. We next evaluated the clinical performance of CRI‐RUB using 140 patient specimens, with the overall results summarized in Figure 5a and TaqMan qPCR initially serving as the reference method. Each specimen was tested in triplicate; a sample was considered qPCR‐positive if all three replicates had Ct values < 33 or if the mean Ct was < 33. The Ct distribution of individual specimens is presented in Figure 5b. Specimens with discrepant results between CRI‐RUB and qPCR, borderline qPCR results (Ct 33–34), or concordant negative results by both assays underwent additional workup consisting of repeated qPCR testing and PCR amplification of cultured isolates followed by ITS sequencing. Borderline qPCR results were retested, and if repeat Ct values remained within the 33–34 range with a typical amplification curve, the specimen was adjudicated as positive in the composite reference standard. On the basis of this composite standard, 76 specimens were classified as T. rubrum –positive and 64 as negative.

FIGURE 5.

Diagnostic performance of CRI‐RUB and TaqMan qPCR in clinical specimens. (A) Distribution of CRI‐RUB T/C fluorescence ratios for clinical specimens classified as positive or negative according to the composite reference standard. Each dot represents one specimen; the dashed line indicates the decision threshold. (B) Distribution of TaqMan qPCR Ct values for specimens classified as positive or negative by the composite reference standard. Each dot represents one specimen, illustrating the separation between the two groups. (C) Receiver operating characteristic (ROC) curves for TaqMan qPCR and CRI‐RUB based on clinical testing results. The red curve (AUC1 = 0.9859) corresponds to TaqMan qPCR, and the blue curve (AUC2 = 0.9803) corresponds to CRI‐RUB, indicating high diagnostic accuracy for both assays.

Against this composite reference, CRI‐RUB correctly identified 73 of 76 positive specimens and all 64 negative specimens (73 true positives, 3 false negatives, 0 false positives, and 64 true negatives), as summarized in Table 1. This corresponded to a sensitivity of 96.1% (73/76; 95% CI 89.0%–98.6%) and a specificity of 100% (64/64; 95% CI 94.3–100.0%), with a positive predictive value (PPV) of 100% (73/73) and a negative predictive value (NPV) of 95.5% (64/67). The Youden's index was 0.96, and the Cohen's κ coefficient was 0.96, indicating excellent agreement with the composite reference.

TABLE 1.

Comparison of CRI‐RUB and the composite reference standard.

	Composite standard		Sensitivity% (95%)	Specificity% (95%)	PPV%	NPV%	Kappa
	Positive	Negative
CRI‐RUB
Positive	73	0	96.05	100	100	95.52	0.9569
Negative	3	64

For TaqMan qPCR, 75 true positives, 1 false positive, 1 false negative, and 63 true negatives were observed, as shown in Table 2, yielding a sensitivity of 98.7% (75/76; 95% CI 92.9%–99.8%) and a specificity of 98.4% (63/64; 95% CI 91.7%–99.7%), with a PPV of 98.7% (75/76) and an NPV of 98.4% (63/64). As shown in Figure 5c, ROC analysis for CRI‐RUB yielded an AUC of 0.9803, with an optimal decision threshold of 0.266 determined by the Youden index. In addition, compared with TaqMan qPCR, CRI‐RUB reduced the overall turnaround time by approximately 60 min, while maintaining high diagnostic accuracy and perfect specificity without false‐positive results.

TABLE 2.

Comparison of TaqMan qPCR and the composite reference standard.

	Composite standard		Sensitivity% (95%)	Specificity% (95%)	PPV%	NPV%	Kappa
	Positive	Negative
TaqMan qPCR
Positive	75	1	98.68	98.44	98.68	98.44	0.9712
Negative	1	63

4. Conclusion

In this study, we validated CRI‐RUB, a CRISPR–Cas12a–coupled quantum‐dot lateral‐flow assay for species‐level detection of T. rubrum using 140 clinical specimens. Compared with the composite reference standard consisting of TaqMan qPCR and ITS sequencing, CRI‐RUB showed excellent agreement (sensitivity 96.1% [73/76], specificity 100.0% [64/64], κ = 0.96). Relative to TaqMan qPCR (sensitivity 98.68%, specificity 98.44%, κ = 0.97), the main advantages of our assay are its shorter turnaround time and lack of dependence on expensive real‐time PCR instruments, requiring only a simple heating device and a portable fluorescence reader. In routine outpatient practice, KOH microscopy typically has a sensitivity of only around 60% and cannot reliably identify the infecting species, whereas qPCR and ITS sequencing provide species‐level resolution but are limited by equipment requirements, assay cost, and turnaround time [34, 35]. Culture‐based identification usually requires 1–4 weeks, further underscoring the potential value of a rapid, culture‐independent diagnostic method that could be adapted for use in outpatient, primary‐care, and resource‐limited or decentralized laboratory settings. The comparative characteristics of CRI‐RUB and conventional diagnostic methods are summarized in Table 3.

TABLE 3.

Comparison of CRI‐RUB with conventional diagnostic methods.

Item/Method	Direct microscopy	Fungal culture	TaqMan qPCR	CRI‐RUB
Species‐level identification	No	YES	YES	YES, T. rubrum
Typical clinical sensitivity	Moderate (≈60%)	Variable; dependent on fungal viability and sampling quality	High	High
Total turnaround time	15‐30 min	1–4 weeks	3‐4 h	85 min
Main equipment	Light/fluorescence microscope	Incubator, culture media	Real‐time PCR thermocycler	Simple heater, pipettes, QD LFA strips
Suitability for POCT	Limited	Poor	Limited	High

Note: Comparative information for direct microscopy, fungal culture, and TaqMan qPCR was summarized from previous reports [35, 36, 37, 38]. Data for CRI‐RUB were generated in the present study.

These diagnostic limitations have important implications for antifungal management. In daily practice, clinicians confronted with tinea‐like lesions often initiate topical or systemic antifungal therapy based on clinical appearance combined with microscopy, usually without species‐level testing [26, 27]. Prescription database studies indicate that a substantial proportion of antifungal use is not supported by laboratory evidence, raising concerns about unnecessary drug exposure and selection pressure for resistance [39, 40]. At the same time, multiple clinical trials and treatment guidelines demonstrate that the optimal systemic therapy may depend on the infecting species: infections caused predominantly by Trichophyton species, most often T. rubrum , respond better to oral terbinafine, whereas Microsporum infections are more effectively treated with griseofulvin [4]. T. rubrum is also the leading cause of dermatophyte onychomycosis worldwide, and several guidelines recommend oral terbinafine as the first‐line systemic agent for these patients [41, 42]. Against the backdrop of widespread terbinafine use, terbinafine‐resistant dermatophytes—most notably T. indotineae and resistant T. rubrum isolates—have been increasingly reported in recent years, highlighting the need for more judicious antifungal prescribing and more robust species‐level laboratory diagnostics [6, 43]. In this context, a rapid platform such as CRI‐RUB, which enables near–point‐of‐care, culture‐independent, species‐level detection of T. rubrum , may contribute to more pathogen‐informed and individualized antifungal therapy, rather than relying solely on species‐unknown, microscopy‐based empiricism.

In recent years, several CRISPR/Cas‐based assays for fungal pathogens, including dermatophytes, have been described, demonstrating the feasibility of combining RPA/RAA with Cas12‐based detection for clinical samples [44, 45]. However, most previously reported methods have focused on veterinary dermatophytosis in companion animals, keratomycoses diagnosed from corneal scrapings, or fluorescence readouts requiring well‐equipped laboratory settings. By contrast, CRI‐RUB was specifically developed for the detection of T. rubrum in human superficial dermatophytosis and was systematically evaluated in 140 skin and nail specimens using a composite reference standard incorporating both qPCR and ITS sequencing. By coupling CRISPR–Cas12a with a quantum dot–based lateral‐flow strip, CRI‐RUB provides a visually interpretable, portable readout that is better aligned with routine dermatology workflows and the needs of resource‐limited environments than CRISPR assays relying solely on laboratory fluorescence detection. The entire workflow can be completed in approximately 85 min using only a simple heating device and a small fluorescence reader, suggesting that, after further simplification, it could be feasible for use in dermatology outpatient clinics, community hospitals, and peripheral laboratories.

Guided by the intended clinical application of this study, a quantum dot–based lateral‐flow readout was selected. CRISPR‐based nucleic acid detection can be coupled with several signal readout strategies, including tube‐based fluorescence detection, electrochemical readouts, and colorimetric systems [46, 47]. Tube‐based fluorescence detection is sensitive and useful for laboratory validation, but it generally requires fluorescence instrumentation and is less intuitive for endpoint interpretation in routine dermatology workflows. Electrochemical readouts are promising for miniaturized sensing, but they usually require electrode preparation, interface modification, and dedicated electrochemical instrumentation, which may complicate assay standardization and clinical deployment [47]. Conventional colorimetric systems, particularly gold‐nanoparticle lateral‐flow strips, are simple and instrument‐free but generally provide lower signal intensity, weaker signal‐to‐background contrast, and limited semi‐quantitative capability. In contrast, the QD‐LFA format used in CRI‐RUB combines the advantages of strip‐based visual interpretation with enhanced fluorescence signal intensity and the potential for portable semi‐quantitative readout [48], making it suitable for rapid species‐level detection of T. rubrum in clinical dermatology specimens.

This study also has several limitations. First, CRI‐RUB was evaluated in a single‐center cohort and targets only T. rubrum ; its performance for other dermatophyte species and for strains circulating in different geographic regions remains to be established. Second, the current workflow still involves multiple manual steps and separate reaction stages, which may introduce operator variability and limit throughput. In addition, we did not directly assess the impact of incorporating CRI‐RUB into routine care on antifungal prescribing patterns or clinical outcomes. Although the ITS region was selected as the primary target in this study because it remains the most widely used fungal barcode region and is consistent with the ITS sequencing‐based reference framework, future expansion of the assay panel could involve broader sequence comparisons during crRNA design rather than being restricted to the ITS region. TEF1‐α or other secondary loci may be incorporated to expand species coverage and improve discrimination among closely related dermatophyte species [49, 50].

Future work will focus on integrating the reactions into more streamlined one‐tube or lyophilized formats, incorporating simplified DNA extraction procedures, and optimizing reaction conditions—for example, by exploring Cas13a‐based detection schemes to circumvent PAM constraints and broaden target design, and by adding dithiothreitol (DTT) or other chemical enhancers to further increase Cas trans‐cleavage activity and potentially improve the limit of detection [51, 52, 53]. In parallel, resistance‐associated targets, such as mutations in the squalene epoxidase (SQLE) gene, could also be included to enable simultaneous detection of dermatophyte species and clinically relevant antifungal resistance markers. Further evaluation of the performance and feasibility of CRI‐RUB in other centers will also be important. Interventional studies assessing how the implementation of CRI‐RUB influences antifungal selection, treatment duration, and recurrence will also be helpful to more clearly define its role in antifungal stewardship. Despite these limitations, our findings indicate that CRI‐RUB has promising clinical utility and translational potential as a practical species‐level diagnostic tool that can complement microscopy, culture, and qPCR, and could contribute to more judicious and rational use of antifungal agents in outpatient, primary‐care, and resource‐limited settings.

Author Contributions

Peng Xu: data curation. Jingwen Tan: data curation. Lianjuan Yang: data curation, funding acquisition. Xiaoping Hu: writing – original draft, visualization, validation, methodology, investigation, data curation, formal analysis. Xuepeng Sun: data curation. Qixia Wang: supervision. Fei Tan: project administration, funding acquisition. Xiaoyun Chai: data curation. Mengyuan Shen: data curation.

Funding

This work was supported by the Program of the Science and Technology Commission of Shanghai Municipality ‘Science and Technology Innovation Action Plan’ Medical Innovation Research Special Project (No. 22Y11905700); the Shanghai Shen Kang Clinical Technology Promotion Program (No. SHDC12026111); the Center for Basic Research and Innovation of Medicine and Pharmacy (MOE) Grant (No. YXX2024‐KF01‐04); and the Xi'an Municipal Health Commission General Research Project (No. 2026ms03).

Ethics Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Shanghai Dermatology Clinical Trial Ethics Committee (approval number 202204).

Consent

Informed consent was obtained from all subjects involved in the study.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Fungal species and GenBank accession numbers.

Table S2: Oligonucleotide sequences used in this study.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.