HOMEBRED: A unified CRISPR platform for field-ready shadowing of infectious agents and oncogenic mutations

Summary

Emerging infectious diseases and late-stage cancers demand rapid, sensitive, and deployable diagnostics. Here, we present HOMEBRED (highly sensitive & specific omnipresent multipurpose endonuclease based reliable detection), a CRISPR-based platform capable of detecting human and animal pathogens (viral/bacterial, RNA/DNA) alongside oncogenic transcripts. HOMEBRED advances veterinary medicine by integrating CaSilico, an automated tool that designs crRNAs by identifying conserved, mutant-resistant regions across pathogenic genomes. By combining CaSilico with extraction-free processing, the platform achieved field-ready FMDV (foot-and-mouth disease virus) detection with a visual readout using only a handheld blue light. Furthermore, HOMEBRED’s dual-crRNA architecture differentiated Brucella vaccine strains from field isolates, overcoming long-standing DIVA (differentiating infected from vaccinated animals) limitations to prevent unnecessary culling of high-breeding-value livestock. In oncology, HOMEBRED distinguished BCR-ABL1 transcripts, establishing a foundation for early diagnosis and personalized therapy. Matching gold-standard PCR sensitivity with reduced complexity and cost, HOMEBRED offers a versatile, portable solution for precision diagnostics in point-of-care settings.

Graphical abstract

Highlights

• •HOMEBRED: a field-ready CRISPR diagnostic platform with automated crRNA design
• •Detects crude samples, offering a rapid, extraction-free alternative to RT-qPCR
• •Dual-crRNA architecture enables DIVA testing to prevent livestock culling
• •Differentiates cancer oncogenic transcripts for personalized oncology

Microbiology; Biotechnology; Cancer

Introduction

Microbial infections and cancer are among the most pressing global health threats, each responsible for substantial morbidity, mortality, and economic burden. In 2022 alone, cancer accounted for nearly 10 million deaths worldwide,¹ while infectious disease outbreaks continue to challenge healthcare systems with their unpredictability and severity. Despite their distinct etiologies, both disease types share a critical obstacle: the difficulty of early and accurate diagnosis. Delayed or imprecise detection contributes to high mortality rates and limits the effectiveness of treatment strategies.^2,3 Therefore, the development of sensitive, specific, and rapid diagnostic means remains a cornerstone for improving clinical outcomes and strengthening public health preparedness.

Nucleic acid detection remains the gold standard for diagnosing infectious diseases and identifying cancer biomarkers. However, conventional PCR-based methods, while sensitive and specific, are limited by their dependence on specialized equipment, trained personnel, and centralized laboratories, barriers that hinder timely diagnosis, especially in resource-limited settings. Notably, the existence of only one FDA-approved at-home nucleic acid-based test (Lucira COVID-19 & Flu Home Test) underscores the ongoing difficulties in creating deployable diagnostic tools that are simple to set up, equipment free, and capable of delivering results in a user-friendly format with minimal risk of error.⁴ Consequently, there is an urgent need to develop novel diagnostic approaches that overcome these limitations while maintaining the sensitivity and specificity of current gold-standard methods.

Recent advances in molecular diagnostics, particularly the emergence of CRISPR-Cas systems, have opened new avenues for accessible, sensitive, and rapid disease detection. When combined with isothermal amplification techniques such as loop-mediated isothermal amplification (LAMP) or recombinase polymerase amplification (RPA), CRISPR-based diagnostics offer high specificity, speed, and potential for field deployment. Numerous CRISPR-Cas systems have been developed, employing various Cas proteins, reporter molecules, pre-amplification strategies, and detection techniques.⁵ Pioneering platforms like specific high sensitivity enzymatic reporter unlocking (SHERLOCK) and DNA endonuclease-targeted CRISPR trans reporter (DETECTR) exemplify capabilities of this approach, but many current implementations still require complex workflows, including sample processing and lab infrastructure, limiting their usability in decentralized settings.⁶

In response to these challenges and aligned with the ASSURED criteria (affordable, sensitive, specific, user friendly, rapid, equipment free, and deliverable) set by the World Health Organization (WHO), we developed HOMEBRED (highly sensitive & specific omnipresent multipurpose endonuclease based reliable detection), a portable, instrument-free diagnostic platform optimized from SHERLOCK. While conventional CRISPR setups face severe bottlenecks due to fragile crRNA target windows vulnerable to mutations, HOMEBRED effectively overcomes these limitations. It integrates an automated computational pipeline via CaSilico tool, which applies stringent design criteria to identify highly conserved, mutant-resistant regions across diverse pathogenic genomes.⁷ By neutralizing mutational escape, this optimized crRNA design architecture allows HOMEBRED to seamlessly pair with flexible amplification formats, enabling both two-step and single-step reactions with streamlined visual readouts (Figure 1).

Figure 1.

Schematic overview of the HOMEBRED platform: Advancing health equity in low-resource settings

HOMEBRED leverages CRISPR-Cas13a technology while eliminating the need for advanced laboratory equipment. To ensure diagnostic precision and eliminate the risk of mutational escape, target-specific crRNAs are engineered using CaSilico, an automated R package. By executing high-throughput screening across rapidly evolving pathogens, CaSilico facilitates a high-velocity, automated design pipeline that grants the platform deployment speed against outbreaks, allowing diagnostic adaptation ahead of shifting evolutionary pressures. Furthermore, HOMEBRED introduces a high-resolution dual-crRNA architecture engineered to overcome long-standing limitations in differentiating infected from vaccinated animals (DIVA). The platform enables flexible testing with or without nucleic acid extraction across both single-step or two-step formats, accommodating a range of disease-specific sample types. For example, vesicular fluid, epithelial tissue, and saliva for foot-and-mouth disease (FMD); blood, aborted fetuses, and vaginal secretions for brucellosis; nasopharyngeal/oropharyngeal swabs and saliva for COVID-19; and whole blood for BCR-ABL1 oncogenic transcripts. As illustrated, nucleic acid preparation is followed by isothermal target amplification via RPA or RT-RPA, and downstream detection using a reaction mixture containing T7 RNA polymerase, Cas13a, crRNA, and an RNA reporter. The single-step format integrates amplification and detection into a one-pot reaction. The entire process is instrument-free, with collateral trans-cleavage signals visually interpreted using handheld blue lights or lateral flow strips. RT-RPA, reverse transcriptase-recombinase polymerase amplification; C, control line; T, test line.

The resulting platform is highly adaptable to emerging sequence variations and applicable to a wide range of viral and bacterial pathogens in both humans and animals, across diverse sample types. It also enables the detection of cancer biomarkers and can distinguish between different cancer types, making it a promising tool for personalized medicine. We demonstrate the versatility of HOMEBRED across a diverse range of targets, including foot-and-mouth disease virus (FMDV), a highly contagious viral disease affecting livestock⁸; Brucella, a bacterial pathogen with significant zoonotic potential⁹; SARS-CoV-2, the viral causative agent of COVID-19; and BCR-ABL1 fusion transcripts, a hallmark genetic alteration in chronic myeloid leukemia (CML) and other hematologic malignancies.¹⁰ Notably, the platform’s dual-crRNA architecture differentiates Brucella vaccine strains from field isolates, overcoming long-standing DIVA (differentiating infected from vaccinated animals) limitations to prevent the unnecessary culling of high-breeding-value livestock. The system achieves sensitivity on par with PCR and sequencing, even from unprocessed samples, and enables subtype differentiation using easy result interpretation. These capabilities position HOMEBRED as a promising tool for rapid, accessible diagnostics and precision medicine applications across both clinical and field-based or resource-limited settings.

Results

Development and scope of the HOMEBRED diagnostic platform

Motivated by the necessity for a robust, versatile, and field-deployable diagnostic platform, we developed HOMEBRED, an SHERLOCK-based platform designed to detect viral and bacterial pathogens, distinguish vaccine-derived from wild-type strains, and identify cancer biomarkers, including the discrimination of oncogenic transcript variants (Figure 2A). To enhance field applicability, HOMEBRED was optimized to enable direct sample analysis without nucleic acid extraction and to allow result visualization without the need for specialized equipment or infrastructure. To illustrate HOMEBRED’s potential, it was applied across a diverse panel of targets, including three infectious agents: FMDV, Brucella, and SARS-CoV-2 as well as the BCR-ABL1 cancer biomarker, characteristic of CML.

Figure 2.

Scope and optimization of the HOMEBRED diagnostic platform

(A) HOMEBRED’s applicability for detecting various agents in diverse settings.

(B) Illustration of RPA primer and crRNA design for detecting multiple BCR-ABL1 fusion transcripts. The reverse primer (Ra2R) is common to all transcript variants, while the primer pair Re13a2F/Ra2R enables amplification of both e13a2 and e14a2 transcripts.

(C) Schematic of a typical crRNA hairpin structure, featuring a stable secondary fold with base-paired stems and a loop region.

(D) Predicted secondary structures of CR3D1 and CR3D2: the left images show the MFE structures, while the right images display the centroid secondary structures.

Design and optimization of crRNAs and RPA primers

FMDV, a virus known for its high genetic variability, presents significant challenges for diagnosis and disease control. Conservation analysis has identified the 3D gene as a plausible candidate for crRNA design, as it contains extensive conserved regions.¹¹ Using the CaSilico R package, we analyzed 707 3D gene sequences extracted from various FMDV serotypes, applying a 98% conservation threshold and obtaining 41 potential target sites (Table S1). From this pool, two crRNAs were selected based on stringent design criteria, including high GC content, high U-rich content with target site accessibility within the RPA amplicon length, no off-target effects, and favorable secondary structure predictions, permitting no more than a single mismatch (Table S2).

Brucellosis, another focus of this project, is caused by various Brucella species, with four important species relevant to human and animal health including B. melitensis, B. abortus, B. suis, and B. canis. Among commonly used amplification targets, such as omp2, omp28, omp31, 16S rRNA, IS711, and bcsp31, the bcsp31 gene stands out due to its high conservation and diagnostic reliability across Brucella species.¹² On the other hand, the inability to distinguish between infected and vaccinated animals poses a major challenge for disease diagnosis, complicating effective disease management, and potentially resulting in unnecessary culling or the undetected spread of infections. Therefore, the development of diagnostics to differentiate between infected and vaccinated animals is essential.¹³ Whole genome sequence analysis of the RB51 vaccine strain compared to other Brucella species revealed that a deletion mutation in RB51 narJ gene may serve as a key marker for distinguishing the vaccine strain from the others.¹⁴ Accordingly, we designed two crRNAs: one to recognize all four main species and another to differentiate them from the vaccine strain (Table S3).

To demonstrate this capability of the HOMEBRED platform to detect oncogenic markers, we targeted the BCR-ABL1 fusion transcripts, which are characteristic of CML. These transcripts result from different breakpoints within the BCR and ABL1 genes, producing major isoforms including p210 (e13a2/b2a2 and e14a2/b3a2) and p190 (e1a2).¹⁵ To specifically identify each of these transcripts, we designed three distinct crRNAs. Additionally, a crRNA targeting the ABL1 gene was designed to function as a control, ensuring the accuracy and reliability of the results (Table S3).

RPA primer design followed strict criteria, including primer and amplicon lengths, GC content for primer and target region, melting temperature, and avoidance of repetitive sequences, palindromes or heterodimers. Since RPA primers can tolerate minor mismatches, we designed FMDV primers to meet a 91% conservation threshold in each position of the 3D gene, allowing for up to two mismatches between the primer and target in regions other than the 3′ end. In contrast, the Brucella primers were designed without mismatches for maximal specificity. To identify BCR-ABL1 isoforms, we developed separate RPA primer pairs for each isoform, with a shared reverse primer targeting exon 2 of ABL1. Since the forward primer for e13a2 is located on exon 13 of the BCR gene, it is capable of amplifying not only e13a2 but also e14a2, producing a longer amplicon. Therefore, e14a2 can be amplified by two different primer pairs: one specific to e14a2 and the other designed for e13a2. Additionally, an RPA primer pair spanning the exon-exon junction of the ABL1 gene was designed (Figure 2B and Table S4).

Two-step HOMEBRED assay

SARS-CoV-2 detection

The development of diagnostic tests requires high reliability and must ensure consistent results across varied testing conditions. To assess this within the context of CRISPR-based diagnostics, we detected SARS-CoV-2 under our experimental conditions using previously reported crRNA and RPA primers.¹⁶ Since these sequences were designed early in the pandemic, we examined their conservation using publicly available Iranian genomes via the global initiative on sharing all influenza data (GISAID) database as of September 2024. After conducting multiple sequence alignments (MSA) of 1,374 complete genomes, the crRNA sequence was found to be fully conserved in 99.42% of the genomes, while the RPA primers were predicted to amplify 98.91% of them. To validate assay performance, we tested eight clinical samples collected from symptomatic Iranian SARS-CoV-2 patients (Table S5). HOMEBRED demonstrated complete qualitative agreement with reference RT-qPCR results using both fluorescent and lateral flow readouts, correctly identifying all five RT-qPCR-positive samples and all three RT-qPCR-negative samples (Figure S1).

FMDV detection and structural crRNA optimization

FMDV detection was pursued using two crRNAs, CR3D1 and CR3D2, targeting the 3D gene. The crRNAs must fold into a stable secondary structure, typically including a hairpin loop with base pairing, as shown in Figure 2C, which is critical for recognition by the Cas protein. Any atypical interaction between hairpin bases or between the direct repeat sequence and the spacer sequence compromises crRNA functionality.⁷ Accordingly, we analyzed the structure of all crRNAs by generating the minimum free energy (MFE) structure using RNAfold software, ensuring a natural hairpin loop structure and the absence of interactions between the spacer and direct repeat sequences. Although CR3D1 possessed all the key features required for a functional crRNA, it failed to detect the virus. Further examination of its thermodynamic properties revealed that the centroid secondary structure of CR3D1 did not meet the specified criteria. Similar instances of non-functional crRNAs have been observed in previous studies. For example, Arizti-Sanz et al. tested four crRNAs for SARS-CoV-2 detection, one of which (S gene-crRNA3) failed to produce a significant fluorescent signal. This highlights the importance of in silico analysis such as MFE and centroid secondary structures of the designed crRNAs.¹⁷ As a result, we designed CR3D2 by considering both the MFE and centroid secondary structures (Figure 2D).

We tested 11 clinical samples from FMDV serotypes, as well as O and A seedstocks (Table S5). This method accurately identified all positive and negative samples, achieving complete concordance with RT-qPCR results. Furthermore, the fluorescent and lateral flow readouts showed consistent agreement. To determine the limit of detection (LOD), which is essential for effective screening of infected animals, we utilized in vitro-transcribed RNA as input for the RPA reaction at concentrations ranging from 10⁶ to 10 cp/μL, prepared in a 1:10 dilution series. Dilutions from 10⁵ to 10 cp/μL were tested using paper-based lateral flow assay, while those from 10⁶ to 10 cp/μL were evaluated using fluorescent assay. Synthetic targets were detectable at the lowest dilution in both methods (Figures 3A–3D, S2, and S3).

Figure 3.

Two-step HOMEBRED assay for FMDV detection

(A) Colorimetric detection of extracted RNA after 1.5 h.

(B) Fluorescent detection of extracted RNA after 1.5 h, with images captured using a smartphone camera. Time-course data at 0.5, 1, 2, and 3 h are available in Figure S2.

(C) Colorimetric detection of synthetic RNA using two-step HOMEBRED after 3 h.

(D) Fluorescent detection of synthetic RNA using two-step HOMEBRED after 3 h.

For (A) and (C), heat maps represent signal intensity, where a value of 1 corresponds to the maximum intensity of the test band and serves as the reference for normalizing all other bands; C, control line; T, test line. For (B) and (D), each sample was tested in duplicate with upper tube strips representing the first replicate, and lower strips representing the second; Rep, replicate. For (C) and (D), input concentration is given in cp/μL; NTC, non-template control.

Brucella detection and differentiation from vaccine strains

To evaluate diagnostic performance for Brucella, we tested 14 samples, comprising bacterial cultures, along with negative and positive clinical samples exhibiting different titers, as determined by qPCR assay (Table S5). HOMEBRED demonstrated complete qualitative concordance with reference PCR and qPCR assays, correctly identifying all positive and negative samples across both lateral flow and fluorescent readouts. These results confirm high qualitative accuracy across diverse sample concentrations.

To differentiate RB51 vaccine strain from other strains, we designed a new crRNA, and the test was repeated on all samples using both strip-based lateral flow and fluorescent assay. As expected, both assays produced signals for positive samples, including bacterial cultures of Melitensis, Abortus, and Suis, while no signals were observed for the vaccine strain or negative samples (Figures 4A, 4B, S4, and S5). We utilized synthesized standard DNA as input for the RPA reaction at concentrations ranging from 10⁶ to 10 cp/μL, prepared through a 1:10 dilution series to establish the LOD. For the strip-based lateral flow assay, dilutions ranging from 10⁵ to 10 cp/μL were tested, while for the fluorescent assay, an additional lower dilution of 10⁶ cp/μL was also included. At the lowest dilution level, successful detection of synthetic targets was achieved using both methods (Figures 4C and S6). These results sufficiently demonstrate the broad applicability of HOMEBRED for rapid and highly sensitive detection of low-concentration pathogens.

Figure 4.

Two-step HOMEBRED assay for Brucella and BCR-ABL1 detection

(A) Strip-based lateral flow detection of Brucella DNA (upper image) and differentiation of RB51 vaccine strain from wild strains (lower image) after 1.5 h of incubation.

(B) Fluorescent detection of Brucella DNA (upper 8-tube strip) and differentiation of RB51 strain (lower 8-tube strip) after 1.5 h. Time-course data at 0.5, 1, 2, and 3 h are provided in Figure S5.

(C) Detection of synthetic Brucella DNA using both lateral flow and fluorescent assays after 3 h.

(D) Detection of BCR-ABL1 fusion transcripts from KCL-22 and K-562 cell lines using colorimetric and fluorescent signals within 1.5 h. Fluorescent signal data at 0.5, 1, 2, and 3 h are shown in Figure S7. Samples 1–3 correspond to KCL-22, and samples 4–7 to K-562; lowercase letters indicate specific RPA primer and crRNA combinations: (a) Re13a2F/Ra2R/CRe13a2, (b) Re13a2F/Ra2R/CRe14a2, (c) Re14a2F/Ra2R/CRe14a2, and (e) RABL1F/RABL1R/CRABL1; NTC, non-template control.

(E) Detection of fusion transcripts from clinical samples using colorimetric signals within 1.5 h; primer/crRNA combinations are labeled as: (a) Re13a2F/Ra2R/CRe13a2, (b) Re13a2F/Ra2R/CRe14a2, (d) Re1a2F/Ra2R/CRe1a2, and (e) RABL1F/RABL1R/CRABL1.

(F) Detection of synthetic e14a2 transcripts using colorimetric and fluorescent assays after 3 h.

For (A, C, D, and F), test band intensity was normalized by assigning a value of 1 to the highest observed signal, with all other bands calibrated accordingly; C, control line; T, test line. For (C and F), input concentrations are expressed in cp/μL; NTC, non-template control. In (C, D, and F), duplicate testing was performed, with upper tube strips representing the first replicate and lower strips representing the second; Rep, replicate.

Detection of BCR-ABL1 fusion transcripts in leukemia

We next assessed HOMEBRED for detecting complex cancers with isoform diversity, focusing on BCR-ABL1 fusion transcripts which serve as key biomarkers in CML. Given the expression of e13a2 and e14a2 fusion transcripts in KCL-22 and K-562 cell lines, respectively, these lines serve as valuable models for diagnostic applications. The platform was evaluated using RNA extracted from these two cell lines. To enhance accuracy, we conducted nested PCR following a previously validated protocol¹⁸ on the extracted RNA, and the resulting amplified transcripts were subjected to Sanger sequencing (Data S1).

We successfully detected both isoforms using the common RPA primer set (Re13a2F/Ra2R) and specific crRNA, sensing with distinct signal readouts and no reactivity observed with non-specific crRNA for either transcript. Additionally, detection of the e14a2 isoform was achieved using the specific RPA primer set (Re14a2F/Ra2R). Finally, diagnosis of the ABL1 gene in both cell lines using the control crRNA demonstrated a comprehensive diagnostic assessment (Figure 2B, 4D, and S7). We further validated the platform using 14 clinical samples, previously analyzed by RT-qPCR. The type of transcript for these samples was determined through RT-PCR and Sanger sequencing, as previously described (Figure S8, Table S5, and Data S1). Using the transcript-specific test, we accurately identified all isoforms e13a2, e14a2, and e1a2 among the clinical samples. Surprisingly, three samples (BA1, BA3, and BA11) that tested negative by RT-qPCR were diagnosed as positive by HOMEBRED. Notably, the assay also identified two patients (BA2 and BA11) co-expressing both e13a2 and e14a2 using the shared RPA primer set. Furthermore, the control crRNA detected all samples, including negative ones. The results demonstrated that the proposed platform offers superior sensitivity compared to RT-qPCR and competitive specificity relative to gold-standard methods such as Sanger sequencing, while maintaining field-deployable capabilities (Figure 4E).

To assess LOD, we tested a gradient of in vitro-transcribed fusion transcripts e14a2, using the Re13a2F/Ra2R RPA primer set, and e1a2 at different concentrations with 10-fold dilutions. The e14a2 assay demonstrated sensitivity down to 10² cp/μL with a colorimetric signal and 10 cp/μL with a fluorescent signal. The LOD for e1a2 was 10³ cp/μL for the lateral flow readout and 10² cp/μL for the fluorescence-based readout (Figures 4F, S9, and S10).

Single-step HOMEBRED assay

The single-step assay combined amplification and Cas13-based detection within a single tube, minimizing contamination risks and offering a significantly less time- and labor-intensive alternative to the two-step assay.

To assess performance, we conducted a head-to-head comparison of single-step and two-step results using FMDV clinical samples. Both fluorescent and lateral flow readouts were employed. The single-step assay achieved 77% agreement with the two-step assay, with three positive samples (F4, F6, and F7) yielding false-negative results (Figures 5A, 5B, and S11). Evaluation of the assay’s LOD using synthetic RNA targets showed reduced sensitivity for both fluorescence and lateral flow formats (LOD 10³ cp/μL; Figure S12).

Figure 5.

Diagnosis by single-step and extraction-free HOMEBRED assay

(A) Colorimetric detection of extracted FMDV RNA using single-step HOMEBRED assay after 1.5 h.

(B) Fluorescent detection of extracted FMDV RNA after 1.5 h. Time-course data at 0.5, 1, 2, and 3 h are provided in Figure S11.

(C) Strip-based lateral flow detection of Brucella DNA (upper image) and differentiation of RB51 vaccine strain from wild strains (lower image) after 1.5 h.

(D) Fluorescent detection of Brucella DNA (upper 8-tube strip) and differentiation of RB51 strain (lower 8-tube strip) after 1.5 h. Time-course data at 0.5, 1, 2, and 3 h are available in Figure S13.

(E) Performance comparison of HOMEBRED for detecting FMDV and Brucella across both two-step and single-step formats; LOD, limit of detection.

(F) Detection of BCR-ABL1 fusion transcripts from KCL-22 and K-562 cell lines using colorimetric and fluorescent signals within 1.5 h. Fluorescent signal data at 0.5, 1, 2, and 3 h are detailed in Figure S15. Samples 1–2 correspond to KCL-22; samples 3–5 to K-562; lowercase letters indicate specific RPA primer and crRNA combinations: (a) Re13a2F/Ra2R/CRe13a2, (b) Re13a2F/Ra2R/CRe14a2, (c) Re14a2F/Ra2R/CRe14a2, and (e) RABL1F/RABL1R/CRABL1; NTC, non-template control.

(G) Two-step detection of non-extracted FMDV samples via colorimetric and fluorescent readouts after 1.5 h. Time-course fluorescent data at 0.5, 1, 2, and 3 h are provided in Figure S16.

For (A, C, F, and G), heat maps represent test band intensity, normalized by assigning a value of 1 to the highest signal and adjusting all other intensities accordingly; C, control line; T, test line. For (B, F, and G), samples were tested in duplicate, with upper tube strips representing the first replicate and lower strips the second; Rep, replicate.

We next evaluated single-step performance for Brucella detection. Both the strip-based lateral flow and fluorescent assay accurately detected all positive and negative samples, achieving 100% concordance with the two-step assay. This finding was consistent during the differentiation of the vaccine strain from other strains, with neither colorimetric nor fluorescent signals observed for the vaccine strain, as expected (Figures 5C, 5D, and S13). Additionally, the sensitivity of the Brucella single-step assay was assessed using synthetic DNA, demonstrating a decreased detection limit of 10² cp/μL for both readout types (Figure S14).

Comparison of two-step and single-step assay results for FMDV and Brucella, including the LOD experiment, revealed that the single-step assay exhibited lower sensitivity and posed a risk of false negatives (Figure 5E). Due to the crucial need for precise identification of cancer transcripts in patients, the single-step method was not pursued for clinical samples. Instead, we tested it on BCR-ABL1-positive KCL-22 and K-562 cell lines. Utilizing colorimetric and fluorescent readouts, the common RPA primer set enabled the successful detection of e13a2 and e14a2 isoforms. Meanwhile, the e14a2 isoform was distinctly identified through its specific RPA primer set (Figures 5F and S15).

Across all two-step and single-step assays, we performed parallel testing on identically processed aliquots (single freeze-thaw cycle), ensuring consistency and reproducibility.

Optimization of HOMEBRED for non-extracted animal-derived matrices

SHERLOCK-based diagnostics are compatible with HUDSON (heating unextracted diagnostic samples to obliterate nucleases), which integrates heat inactivation with chemical reduction to denature nucleases and release nucleic acids from infectious particles.¹⁹ While HUDSON’s compatibility with various sample types has been established, further testing and validation are needed for animal-derived matrices, particularly foot-and-mouth disease (FMD) collection matrices.

In this study, we paired the HOMEBRED assay with the HUDSON technique to streamline the processing of FMDV seedstocks and epithelial samples. Accordingly, FMDV seedstock A and seven clinical samples underwent treatment using the HUDSON method, followed by testing via the two-step assay with two distinct readouts: colorimetric and fluorescent. Additionally, FMDV seedstock O, at an initial concentration of 3.23 × 10⁷ plaque-forming units per milliliter (PFU/mL), was serially diluted at 10-fold intervals, heat-inactivated using the HUDSON method, and tested following the same procedure.

We successfully identified all positive and negative samples across both outputs. Furthermore, utilizing this faster two-step protocol, we detected as low as 3.23 × 10⁴ and 3.23 × 10³ PFU/mL of unextracted seedstock with colorimetric and fluorescent outputs, respectively. These findings demonstrate improved results compared to previous studies. For instance, Casati et al. observed reduced sensitivity and a higher LOD when testing unextracted samples,²⁰ or Arizti-Sanz et al. detected HUDSON-treated SARS-CoV-2 viral seedstocks down to 1.31 × 10⁵ PFU/mL¹⁶ (Figures 5G, S16 and S17).

Evaluating the robustness of HOMEBRED

To assess the robustness of HOMEBRED in the face of genetic variability, we benchmarked its performance against sequencing data for both FMDV and BCR-ABL1 targets. The FMDV crRNAs were designed using CaSilico, which identifies conserved regions in highly variable genomes. To assess this capability, FMDV samples underwent Sanger sequencing to determine how effectively the tool could capture a conserved region for crRNA design, an essential factor in mitigating virus mutation. Sequencing data revealed that the designed crRNA remained fully conserved across different strains, except for one sample (F6) that exhibited a single base substitution (Figure S18). Despite this mismatch, HOMEBRED accurately detected the target, demonstrating tolerance to limited sequence variation.

For BCR-ABL1 fusion transcripts, we used nested PCR¹⁸ and Sanger sequencing to validate transcript identity and inspect for potential protospacer mutations that may impair CRISPR-based detection. Two clinical samples (BA8 and BA12) harbored point mutations within the protospacer area of the breakpoint region. Nevertheless, HOMEBRED accurately identified both samples, with no loss in signal (Figure S18). These findings indicate that HOMEBRED maintains high diagnostic accuracy even in the presence of minor mutations within the target region, highlighting its robustness for real-world applications. This reliability was further supported by overall concordance analysis across technical replicates, yielding 100% qualitative agreement (111/111 duplicate reactions) for positive and negative call determinations across all runs.

Discussion

Global public health is perpetually challenged by the dual threats of emerging infectious diseases and cancers. Rapid pathogen spread and high mortality from late-stage cancer diagnoses highlight the urgent need for diagnostic tools that are fast, sensitive, and adaptable across diverse settings. CRISPR-based diagnostics offer a transformative solution, delivering precise, rapid, and cost-effective detection across virology, bacteriology, and oncology. In this study, we introduced HOMEBRED, a versatile SHERLOCK-based diagnostic platform engineered to detect a wide range of targets, including human and animal viruses, bacterial pathogens, and cancer biomarkers. The platform exhibited high sensitivity in identifying highly infectious agents such as SARS-CoV-2 and FMDV as well as Brucella, a stealth pathogen known for its asymptomatic presentation in both humans and animals, enabling silent transmission. Although research has focused on the standardization, reproducibility, and optimization of CRISPR-based assays to ensure consistency, studies specifically evaluating lab-to-lab variation in CRISPR diagnostic tests remain limited. To validate accuracy and reproducibility at the interlaboratory level, a fundamental published protocol for detecting COVID-19¹⁶ was successfully tested using our facilities, confirming that this diagnostic approach performs consistently under varied conditions. Furthermore, the platform accurately detected and distinguished various fusion transcripts of the BCR-ABL1 oncogene, underscoring its dual utility in both infectious disease surveillance and precision oncology.

Animal farms, where laboratory infrastructure is limited and continuous disease monitoring is essential, represent an ideal setting for CRISPR-based tests as rapid, deployable diagnostic solutions. While CRISPR has been previously explored for detecting pathogens like FMDV and Brucella, HOMEBRED introduces critical advancements. Unlike the Cas12a-based system and reverse transcriptase enzyme-assisted recombinant isothermal amplification (RT-ERA) used by Xiong et al. for FMDV detection,²¹ our platform leverages Cas13a with reverse transcriptase-recombinase polymerase amplification (RT-RPA). Cas13a, being protospacer flanking site (PFS)-free, does not require a protospacer adjacent motif (PAM) for target recognition, offering greater flexibility in crRNA design. In contrast, Cas12a typically depends on aT-rich PAM sequence for cleavage, limiting its targetable space, particularly when isolating specific, low-variability junctions like BCR-ABL1 fusion transcripts. Furthermore, Cas13a’s robust trans-cleavage activity suggests enhanced enzymatic velocity even at suboptimal temperatures, positioning it as a highly viable, instrument-free competitor to less sensitive antigen assays in point-of-care (POC) settings.²² On the other hand, given the competitive nature of ERA primers, RPA offers enhanced structural adaptability for parallelized assays, facilitating the simultaneous detection of multiple targets.²³ However, we deliberately avoid single-reaction multiplexing due to inherent biochemical disadvantages, such as target-primer kinetic competition within the isothermal amplification phase and cross-talk or signal dilution during simultaneous collateral nuclease activation. Instead, HOMEBRED capitalizes on spatial multiplexing across isolated reaction chambers to screen for multiple targets or transcript variants concurrently, preserving maximum analytical sensitivity without biochemical interference.

Practically, the HUDSON-HOMEBRED workflow became compatible with animal sample processing, completely eliminating extraction and enabling direct target detection. This streamlined approach allows testing with minimal equipment, reducing both operational cost and turnaround time. The assay demonstrated sensitivity comparable to the two-step method using purified samples, reinforcing its potential as a superior alternative to the less sensitive single-step assay. Furthermore, the simplified testing process, coupled with an easy visual readout using only a handheld blue light, enhances feasibility for on-site applications, such as farms where timely and cost-effective diagnosis is essential. Crucially, HOMEBRED circumvents major operational vulnerabilities that restrict the deployment of other contemporary CRISPR diagnostics. It is well-established in the literature that bypassing traditional purification typically compromises sensitivity due to target dilution and severe enzymatic inhibition by crude matrix components like hemin or polysaccharides. While prominent extraction-free frameworks frequently suffer from elevated LOD in unextracted backgrounds, as seen in the accurate detection of evolving SARS-CoV-2 through SHERLOCK optimization (ADESSO) procedure by Casati et al.²⁰ or viral assays by Arizti-Sanz et al.¹⁶ our platform maintains high analytical sensitivity without raw sample processing. By lowering detection threshold down to 3.23 × 10³ PFU/mL in crude backgrounds, HOMEBRED advances existing literature standards by up to two orders of magnitude. This performance advantage is rooted in our strategic use of epithelial tissue homogenates for direct FMDV detection. While solid tissue processing may seem counterintuitive for point-of-care testing, epithelial components are significantly more compatible with HUDSON-HOMEBRED chemistry than the destructive inhibitors found in sample types like blood or feces.

Despite strict implementation of control measures, brucellosis remains endemic worldwide. Conservative estimates suggest that over 300 million of the 1.4 billion cattle are infected with Brucella.²⁴ Humans can easily acquire brucellosis through direct contact with animals or contaminated animal products. Despite the complexity of treatment in humans, a vaccine has yet to be developed. So, early detection in animals remains the cornerstone of brucellosis management. While some studies have explored CRISPR-based detection for Brucella infection, a critical gap in these efforts is the lack of consideration for the challenge of DIVA.^25,26,27,28 Diagnostic failures in this regard often lead to unnecessary culling, which is particularly problematic for livestock with high breeding value. To overcome this issue, our platform is equipped with two crRNAs: one for detecting all Brucella strains, including the RB51 vaccine strain, and another for distinguishing the vaccine strain from field isolates. In contrast to previous studies, this study has designed RPA primers and crRNAs based on a comprehensive analysis of sequences extracted from NCBI, ensuring the inspection of all potential genomic variations. Additionally, this study provides a comparative analysis of two-step versus single-step detection methods. Beyond these innovations, our detection system integrates a fluorescent-emitting tube method, utilizing only a low-cost portable blue light alongside colorimetric paper-based strip readout. This user-friendly approach enhances field deployability while minimizing the risk of sample contamination, as it eliminates the need to open tubes containing amplified products. Looking ahead, future strategies should prioritize the development of lyophilized reagents with preserved performance to enhance further the practicality of CRISPR diagnostics for use in animal farms.

HOMEBRED provided a versatile diagnostic approach applicable to both pathogen and cancer mutation detection. This study pursued two key objectives related to BCR-ABL1: facilitating early detection and advancing targeted therapy for cancer patients. Despite significant advancements, many cancers are still diagnosed at advanced stages. Early identification of BCR-ABL1 enables timely intervention with tyrosine kinase inhibitor (TKI) therapy, significantly improving survival rates and preventing mutations that drive TKI resistance. The specific BCR-ABL1 transcript type influences molecular and cytogenetic responses to various TKI regimens, directly impacting survival outcomes.^15,29 Innovations in early detection, such as HOMEBRED, play a crucial role in addressing health inequities, particularly in resource-limited settings. Furthermore, our platform supports personalized medicine by enabling targeted cancer therapies through precise differentiation of oncogenic transcript types.

Limitations of the study

Bioinformatic analyses utilized during the computational design of the Brucella crRNAs targeted hyper-conserved loci, enabling universal detection across all major Brucella species while maintaining the resolution to differentiate the RB51 vaccine strain from wild-type counterparts. Due to the regional source limitations, clinical isolates of Brucella canis could not be physically evaluated in this study. To mitigate this, MSA confirmed 100% nucleotide identity within the bcsp31 and narJ target windows between B. canis and our validated strains. Furthermore, we successfully validated the assay’s analytical performance using synthesized DNA fragments containing the identical, conserved bcsp31 target domain as an operational surrogate for B. canis verification. Additionally, while the BCR-ABL1 diagnostic panel was successfully validated using the K562 and KCL-22 cell lines to detect the e14a2 and e13a2 oncogenic isoforms, respectively, the lack of a commercial or accessible cell line expressing the minor e1a2 fusion transcript precluded further expansion of our in vitro diagnostic benchmarking panel.

Resource availability

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Mohammad Reza Bakhtiarizadeh (mrbakhtiari@ut.ac.ir).

Materials availability

All unique/stable reagents generated in this study are available from the lead contact without restriction.

Data and code availability

• •The datasets generated during the current study have been deposited at NCBI repository (accession numbers SAMN56814468 to SAMN56814485 and SAMN56815408 to SAMN56815439 in BioProject dataset) and are publicly available as of the date of publication. Also, they all have been presented in Data S1.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this article is available from the lead contact upon request.

Acknowledgments

We would like to thank I. Ashrafi for kindly providing some FMDV and brucellosis samples used in this study. We thank H.R. Ajideh and M. Arabzadeh for their contributions to the SARS-CoV-2 and BCR-ABL1 RT-qPCR assays. Additionally, we express our sincere gratitude to A. Razavirad for valuable guidance and insights across several experimental procedures. This study was supported by the research fund of the Tehran University of Medical Sciences with grant no. 1404-9-228-90819.

Author contributions

Conceptualization, A.A. and M.S.; data curation, A.A.; formal analysis, A.A., H.F., M.S., M.H.M., and M.R.B.; funding acquisition, A.A., M.S., and R.S.; investigation, A.A.; methodology, A.A., H.F., M.S., R.S., A.S., and M.R.B.; project administration, A.A. and M.S.; resources, A.A., M.S. and R.S.; software, A.A.; supervision, R.S. and M.R.B.; validation, A.A.; visualization, A.A.; writing – original draft, A.A.; writing – review and editing, A.A., H.F., M.S., R.S., M.H.M., A.S., and M.R.B.; all authors reviewed and approved the final manuscript.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Bacterial and virus strains
FMDV-serotype O	Dr. Ashrafi laboratory	N/A
FMDV-serotype A	Dr. Ashrafi laboratory	N/A
B. melitensis	Dr. Ashrafi laboratory	N/A
B. abortus	Dr. Ashrafi laboratory	N/A
B. suis	Dr. Ashrafi laboratory	N/A
RB51 vaccine strain	Dr. Ashrafi laboratory	N/A
Biological samples
FMDV samples	This manuscript	Table S6
Brucella samples	This manuscript	Table S6
SARS-CoV-2 samples	This manuscript	Table S6
BCR-ABL1 samples	This manuscript	Table S6
Chemicals, peptides, and recombinant proteins
HiScribe T7 High Yield RNA Synthesis Kit	NEB	Cat# E2040S
RNase-free DNase I	Thermo Fisher Scientific	Cat# EN0521
Monarch RNA Cleanup Kit	NEB	Cat# T2030S
Dulbecco’s Modified Eagle Medium (DMEM)	Thermo Fisher Scientific	Cat# 10564011
Fetal Bovine Serum (FBS)	Thermo Fisher Scientific	Cat# A3160402
Penicillin-streptomycin	Thermo Fisher Scientific	Cat# 15-140-122
Trypsin, TPCK treated	Thermo Fisher Scientific	Cat# 20233
RPMI 1640	Thermo Fisher Scientific	Cat# 11875093
QIAamp Viral RNA Mini Kit	Qiagen	Cat# 52904
Total DNA Extraction Kit	Viragene	Cat# 0201
Whole Blood RNA Mini Kit	Favorgen	Cat# FAWBR050
TRIzol Reagent	Thermo Fisher Scientific	Cat# 15596026
RevertAid reverse transcriptase	Thermo Fisher Scientific	Cat# EP0441
Taq DNA Polymerase Master Mix Red	Ampliqon	Cat# A180301
Gel extraction kit	Favorgen	Cat# FAGCK 001-2
REALQUALITY RQ-BCR-ABL p210 One-Step Kit	AB Analitica	Cat# RQ-105-6M
REALQUALITY RQ-BCR-ABL p190 One-Step Kit	AB Analitica	Cat# RQ-115-6M
TwistAmp Basic Kit	TwistDx	Cat# TABAS03KIT
murine RNase inhibitor	NEB	Cat# M0314S
LwaCas13a protein	GenScript	Cat# Z03486
rNTP	NEB	Cat# N0466S
NextGen T7 RNA polymerase	Lucigen	Cat# 30223-1
RNase H	NEB	Cat# M0297S
HybriDetect lateral flow strips	Milenia Biotech	Cat# MGHD 1
TCEP-HCl	Thermo Fisher Scientific	Cat# PI20490
EDTA	Sigma	Cat# ED2SS
RealQ Plus Master Mix Green	Ampliqon	Cat# A324402
dNTP mix	Bio Basic	Cat# DD0056
Taq DNA Polymerase	Thermo Fisher Scientific	Cat# EP0401
Deposited data
Sanger sequencing of FMDV	NCBI database	BioProject ID: PRJNA1446202
Sanger sequencing of BCR-ABL1 gene	NCBI database	BioProject ID: PRJNA1446216
Experimental models: Cell lines
BHK-21	ATCC	Cat# CCL-10
KCL-22	ATCC	Cat# CRL-3349
K-562	ATCC	Cat# CCL-243
Oligonucleotides
crRNA spacer sequences	This manuscript	Table S3
Primer sequences	This manuscript	Table S4
RPA primers	This manuscript	Table S4
T7 RNA promoter: GAAATTAATACGACTCACTATAGGG	This manuscript	Table S5
Biotinylated FAM Cleavage Reporter: /56-FAM/rUrUrUrUrUrUrUrUrUrUrUrUrUrU/3Bio/	This manuscript	Table S5
Software and algorithms
CaSilico	Asadbeigi et al.7	https://doi.org/10.3389/fbioe.2022.957131
MAFFT	GraphPad	https://www.graphpad.com/
ImageJ	ImageJ Software	https://imagej.net/
SnapGene	SnapGene Software	https://www.snapgene.com/
Inkscape	Inkscape Software	https://inkscape.org/

Experimental model and study participant details

Cell and bacterial culture

To prepare and titer FMDV stocks from serotypes O and A, Baby Hamster Kidney Cells (BHK-21; ATCC) were used as previously described.¹¹ Briefly, BHK-21 cells were cultured to full confluence in 25 cm² culture flasks (SPL Life Sciences). For virus propagation, cells were infected at a multiplicity of infection (MOI) of 0.2 plaque-forming unit (PFU) per cell for 1.5 h in serum-free medium. Following infection, complete Dulbecco’s Modified Eagle Medium (DMEM; Gibco) supplemented with 2% heat-inactivated fetal bovine serum (FBS; Gibco) was added. Cultures were maintained at 37°C in a humidified atmosphere containing 5% CO₂. Viral supernatants were harvested upon the appearance of severe cytopathic effects (CPE), centrifuged at 1000 g for 10 min to eliminate cellular debris, aliquoted, and stored at -80°C for future use. Infectivity titers were determined using the median tissue culture infectious dose (TCID₅₀) method.

Human leukemia cell lines KCL-22 and K-562 (ATCC) were cultured in RPMI-1640 medium (Gibco) supplemented with GlutaMAX, 15 mM HEPES, 10% FBS (Gibco), and penicillin-streptomycin (Thermo Fisher Scientific). Cultures were maintained at 37°C in a humidified 5% CO₂ incubator.

The cell lines used in this study were kindly provided by our collaborator’s laboratory. While independent authentication was not performed in our laboratory after transfer, the cell lines were originally obtained by the donor lab from the cell repository. Cell lines were maintained at low passage numbers, and cell morphology was consistently monitored throughout the study to verify identity. All cell lines were cultured and routinely tested for mycoplasma contamination using a PCR-based assay. All cell lines tested negative prior to experimentation.

Lyophilized stocks of B. melitensis, B. abortus, B. suis, and the RB51 vaccine strain were rehydrated to establish active cultures. A small aliquot of each rehydrated sample was inoculated into broth tubes and incubated with gentle shaking at 37°C in a 5% CO₂ atmosphere. Upon visible growth, a portion of the bacterial suspension was plated onto Brucella agar and incubated under the same conditions for colony formation.

Method details

crRNA and RPA primer design

All CRISPR-RNAs (crRNAs) used in this study were designed for type VI-A of class II CRISPR-Cas system. To achieve this, complete genomic sequences of FMDV serotypes O, A, C, Asia 1, SAT 1, SAT 2, and SAT 3 were retrieved from the NCBI database. Genomes were annotated using the reference genome with accession number NC039210. Considering 3D gene as the most conserved gene among others,¹¹ all 3D sequences were analyzed using CaSilico, an R package developed for crRNA design. This tool enables simultaneous targeting of multiple sequences by identifying conserved regions.

For Brucella, crRNA design targeted the bcsp31 gene, widely used in PCR assays due to its high specificity and sensitivity. This crRNA was able to detect all strains of B. melitensis, B. abortus, B. suis, and B. canis. Additionally, a crRNA targeting the narJ gene was designed to distinguish the RB51 vaccine strain from non-vaccine strains. Gene sequences for bcsp31 and narJ were extracted using BLAST from Brucella genomes with average size of ∼3.3 Mb. Succinctly, bcsp31 and narJ gene sequences from reference genome for each species (RefSeq numbers GCF_000007125.1, GCF_000369945.1, GCF_000007505.1, and GCF_000018525.1 from NCBI database) were blasted against their respective RefSeq Genome Database. Hits with 100% coverage were retained for downstream analysis. Multiple sequence alignments (MSA) for each gene were performed using MAFFT (v7.475), ensuring inclusion of vaccine strain sequences.³⁰ crRNAs were then designed within fully conserved regions. Given the high similarity between narJ sequences of RB51 and wild-type strains, crRNA design focused on a known deletion mutation unique to the RB51 strain, enabling specific detection.

Four crRNAs were designed for BCR-ABL1 diagnostic panel. Each crRNA specifically targeted the recombinant region between the BCR and ABL1 genes, covering the e13a2, e14a2, and e1a2 fusion transcripts. An additional crRNA was designed to target ABL1 gene, serving as an internal control. Given the presence of reported mutations within the BCR-ABL1 kinase domain, all crRNAs were carefully designed to exclude these mutation-prone regions.^31,32

For recombinase polymerase amplification (RPA) primer design, MSA was conducted for the 3D gene, alongside bcsp31 and narJ, to account for genomic variations in FMDV and Brucella. Based on the MSA results, RPA primers were designed within conserved regions. Additional primers were developed to amplify BCR-ABL1 fusion transcripts and the internal control. All RPA primers ranged from 28 to 35 bp in length, with amplicon sizes between 90 and 140 bp. Primer GC content was optimized between 35% and 65%, while the amplicon region maintained a GC content of 40% to 60%. Melting temperatures were controlled within a range of 55°C to 70°C, and thermodynamic properties were carefully considered. To enable T7 transcription, a T7 RNA polymerase promoter sequence (Table S6) was appended to the 5′ end of each forward RPA primer.

To ensure specificity and minimize cross-reactivity, the designed crRNAs and RPA primers were evaluated for potential off-target effects. For crRNAs, the BLASTn tool was employed using a word size of 7 and an E-value cutoff of 30,000. Primer specificity was assessed using Primer-BLAST. These analyses were conducted for both the FMDV and Brucella diagnostic targets, as well as for the BCR-ABL1 fusion transcripts. Off-target searches were performed against a curated list of organisms relevant to each diagnostic objective (Table S2). For crRNA analysis, the RefSeq RNA database (refseq_rna) was used, while the nucleotide collection (nr/nt) database was applied for primer evaluation. The crRNA design also prioritized high U-rich content and target site accessibility within the RPA amplicon region to enhance Cas13a efficiency.

Additionally, SARS-CoV-2-specific crRNAs and RPA primers were designed following previously established protocol.¹⁶

crRNA synthesis and purification

DNA oligonucleotides containing an appended T7 promoter sequence were ordered from GenScript Biotech as templates for crRNA production. Each oligo was annealed with a T7 promoter primer (Table S6) and subjected to in vitro transcription (IVT) using the HiScribe T7 High Yield RNA Synthesis Kit (NEB). Post-transcription, residual DNA was removed by treatment with RNase-free DNase I (Thermo Scientific), and crRNAs were purified using Monarch RNA Cleanup Kit (NEB) in accordance with the manufacturer’s instructions. The length and integrity of crRNAs were verified by electrophoresis on a 10% denaturing polyacrylamide/urea gel. crRNA concentrations were quantified spectrophotometrically using a NanoDrop instrument (Thermo Fisher Scientific). To avoid repeated freeze-thaw cycles, crRNAs were aliquoted at working concentrations and stored at -80°C.

Specimen collection, nucleic acid extraction, and ethical approval

Bovine epithelial samples from fresh vesicular lesions caused by FMDV were collected from dairy farms and homogenized and passed through a 0.22 μm filter to remove cellular debris. The clarified filtrate was aliquoted and stored at −80°C for downstream applications. Separately, nasopharyngeal swabs were collected from symptomatic individuals suspected of SARS-CoV-2 infection and preserved in viral transport medium (VTM). Nucleic acids were extracted from FMDV seedstocks, processed bovine epithelial samples, and SARS-CoV-2–containing VTM using the QIAamp Viral RNA Mini Kit (Qiagen). RNA was extracted from 140 μl of input material, and eluted in 60 μL of nuclease-free water.

Blood samples from animals suspected of Brucella infection were collected from animal farms and stored at 4°C until DNA extraction. Whole-genomic DNA was extracted either from a single bacterial colony of cultured Brucella strains or from 200 μL of blood using the Total DNA Extraction Kit (Viragene), following the manufacturer’s instructions. DNA was eluted in 50 μL of nuclease-free water and quantified using a NanoDrop (Thermo Fisher Scientific).

Clinical samples for BCR-ABL1 testing were obtained from patients referred for molecular diagnostics and stored at 4°C for immediate RNA extraction. RNA was isolated from 400 μL of blood using the Whole Blood RNA Mini Kit (Favorgen), yielding 25 μL of eluate. Additionally, KCL-22 and K-562 cell lines were harvested by centrifugation at 125 g for 5 min, and total RNA was extracted using TRIzol Reagent (Thermo Fisher Scientific). All RNA samples were quantified using Nanodrop (Thermo Fisher Scientific).

To avoid repeated freeze-thaw cycles, all extracted nucleic acids were aliquoted. DNA samples were stored at −20°C, and RNA samples were stored at −80°C until further use.

All animal sampling was conducted according to the University of Tehran Research Center policies and under ethical guidance from the university’s Animal Care and Use Committee. Ethical approval for all procedures involving human subjects was granted by the Ethics Committee of Tehran University of Medical Sciences. Specifically, the clinical diagnostic protocols for SARS-CoV-2 experiments were approved under reference number IR.TUMS.VCR.REC.1399.315, whereas the oncology protocols for BCR-ABL1 experiments were approved under reference number IR.TUMS.IKHC.REC.1404.429.

Synthetic standard preparation

Extracted RNA from the FMDV O seedstock was reverse-transcribed into cDNA using RevertAid reverse transcriptase (Thermo Fisher Scientific). PCR amplification was performed using a forward primer containing a T7 promoter sequence (Table S6) and Taq DNA Polymerase Master Mix Red (Ampliqon). Thermocycling conditions were as follows: initial denaturation at 95°C for 5 min, 40 cycles of 95°C for 30 s, 57°C for 35 s, and 72°C for 24 s, followed by a final extension at 72°C for 5 min. PCR products were resolved on a 0.9% agarose gel, and the target fragment was purified using a gel extraction kit (Favorgen). The purified dsDNA served as the template for IVT, conducted overnight at 37°C using the HiScribe T7 High Yield RNA Synthesis Kit (NEB), following the manufacturer’s protocol. Template DNA was removed with DNase I (Thermo Scientific Scientific), and the resulting RNA was purified using the Monarch RNA Cleanup Kit (NEB). The standard RNA was quantified using a Nanodrop and stored at −80°C.

PCR-based synthesis of Brucella reference was performed using DNA extracted from the RB51 strain and Taq DNA Polymerase Master Mix Red (Ampliqon). The thermocycling conditions included: initial denaturation at 95°C for 5 min, 40 cycles of 95°C for 30 s, 57°C for 30 s, and 72°C for 30 s, followed by a final extension at 72°C for 5 min. PCR products were resolved on a 0.9% agarose gel, and the target fragment was purified using a gel extraction kit (Favorgen). DNA concentration was measured using a Nanodrop and stored at −20°C.

Positive controls from the REALQUALITY RQ-BCR-ABL p210 & p190 One-Step Kit (AB Analitica), containing DNA fragments corresponding to e14a2 and e1a2 transcripts, were used for standard generation. These controls were verified by Sanger sequencing (Data S1). Standard RNAs were synthesized following the same protocol used for FMDV, with modified thermocycling conditions: for e14a2, the protocol included an initial denaturation at 95°C for 5 min, followed by 40 cycles of 95°C for 30 s, 57°C for 40 s, and 72°C for 27 s, with a final extension at 72°C for 5 min; for e1a2, the conditions were 95°C for 5 min, followed by 40 cycles of 95°C for 60 s, 57°C for 60 s, and 72°C for 60 s, with a final extension at 72°C for 5 min.

Copy numbers of synthetic references were calculated based on fragment length and concentration. All targets were subsequently diluted tenfold for use in sensitivity evaluation experiments.

Two-step detection assay

The two-step detection assay consisted of: (1) isothermal amplification via RPA and (2) LwaCas13a-mediated detection with a biotinylated or quenched FAM cleavage reporter. RPA reactions were carried out using the TwistAmp Basic Kit (TwistDx). Each lyophilized pellet was resuspended with 29.5 μL rehydration buffer, 2.4 μL of each 10 μM RPA forward and reverse primer, 5.6 μL nuclease-free water, 1 μL RevertAid reverse transcriptase (Thermo Fisher Scientific), and 1.5 μL murine RNase inhibitor (NEB). For DNA input experiments, reverse transcriptase and RNase inhibitor were omitted. Finally, 3 μL MgOAc was added. The RPA mixture was then divided into four 9 μL aliquots, each supplemented with 1 μL of extracted sample (∼200 ng for BCR-ABL1) or Hudson-treated material. Incubation was performed on a thermocycler for 25 min at 41°C for RNA samples and 37°C for DNA samples, except for BCR-ABL1, which required a 45-minute incubation.

For the detection step, 1 μL of the RPA product was combined with 19 μL of detection master mix, comprised of: 1x cleavage buffer (10x CB: 400 mM Tris pH 7.5, 10 mM DTT), 1 mM each rNTP (NEB), 1.23 U/μL murine RNase inhibitor (NEB), 1 μM biotinylated FAM reporter (IDT; Table S6) or 125 nM RNase Alert v2 (Thermo Fisher Scientific), 47.3 nM LwaCas13a protein (GenScript) diluted in storage buffer (SB, 50 mM Tris pH 7.5, 600 mM NaCl, 5% glycerol, 2 mM DTT), 1.58 U/μL NextGen T7 RNA polymerase (Lucigen), and 22.5 nM crRNA. Magnesium chloride was added last to a final concentration of 9.46 mM.

Single-step detection assay

In the single-step detection assay, amplification and detection were integrated into a unified reaction. The master mix for five reactions was prepared with the following components: 1x reaction buffer (5x RB, 0.1 M HEPES pH 8.0, 300 mM KCl, 25% PEG-8000), 2.11 mM of each rNTP (NEB), 1.05 U/μL murine RNase inhibitor (NEB), 1 mM biotinylated FAM reporter (IDT) or 125 nM RNase Alert v2 (Thermo Fisher Scientific), 47.3 nM LwaCas13a protein (GenScript) diluted in storage buffer (SB, 50 mM Tris pH 7.5, 600 mM NaCl, 5% glycerol, 2 mM DTT), 4.23 U/μL RevertAid reverse transcriptase (Thermo Fisher Scientific), 0.11 U/μL RNase H (NEB), 1.05 μL NextGen T7 RNA polymerase (Lucigen), 120.45 nM of each RPA forward and reverse primer, and 22.5 nM crRNA. The master mix was used to resuspend one lyophilized pellet from the TwistAmp Basic Kit. MgOAc was added last to a final concentration of 14 mM to complete the reaction mixture. Extracted samples (∼200 ng for BCR-ABL1) or Hudson-treated samples were added to the final master mix at a ratio 1:19.

For both the two-step and single-step formats, detection was performed using either lateral flow or fluorescence-based readouts. For colorimetric detection, reactions were incubated at 37°C for 1.5 h, diluted 1:4 with HybriDetect Assay Buffer, and applied to HybriDetect lateral flow strips (Milenia Biotech). Images were captured 5 min after incubating. For fluorescent detection, reactions were incubated at 37°C, and images were acquired at 30, 60, 90, 120, and 180 min using a LED blue light transilluminator (Padideh Nogen Pars).

Rapid nuclease inactivation and sample lysis

The HUDSON method was employed to rapidly inactivate nucleases and lyse FMDV particles while preserving nucleic acids for subsequent CRISPR-based detection. Non-extracted FMDV samples and seedstocks were treated with 90 mM TCEP (Thermo Fisher Scientific), 0.9 mM EDTA (Thermo Fisher Scientific), and 0.7 U/μL murine RNase inhibitor (NEB). The mixture was incubated at 50°C for 15 min, followed by an additional heat step at 95°C for 5 min. After cooling, the treated samples were used in both two-step and single-step detection assays.

Parallel experiments for side-by-side comparison

Extracted nucleic acids were analyzed for the presence of infectious agents or oncogenic transcripts using nucleic acid amplification techniques.

For FMDV detection, extracted RNA was reverse-transcribed into cDNA using RevertAid Reverse Transcriptase (Thermo Fisher Scientific). Quantitative PCR (qPCR) was performed in two technical replicates using RealQ Plus Master Mix Green (Ampliqon; Table S4). Thermocycling conditions included an initial denaturation at 95°C for 15 min, followed by 40 cycles of 95°C for 15 s and 69°C for 30 s. Viral load was quantified using a standard curve as previously described.¹¹

Brucella DNA samples were analyzed using both conventional PCR and qPCR. PCR was conducted with Taq DNA Polymerase Master Mix Red (Ampliqon) under the following conditions: 95°C for 5 min, followed by 40 cycles at 95°C for 30 s, 60°C for 30 s, and 72°C for 30 s, concluding with a final extension step at 72°C for 10 min. qPCR was performed in duplicate using RealQ Plus Master Mix Green (Ampliqon), with an initial denaturation at 95°C for 15 min, followed by 40 cycles of 95°C for 15 s and 53°C for 60 s. To generate a standard curve for bacterial load quantification, 10-fold serial dilutions of PCR-amplified RB51 target fragments were prepared, ranging from 10¹¹ to 10 copies/μL. Three independent dilution series were created, and each dilution was amplified in duplicate (Table S4).

For BCR-ABL1 transcript analysis, 1.4 μg of extracted RNA was reverse-transcribed using RevertAid Reverse Transcriptase. Nested PCR was performed following the validated protocol by van Dongen et al.¹⁸ Briefly, the first-round reaction with volume of 50 μl included 3 μl of cDNA, 400 nM of each primer, 200 μM dNTP mix, 0.03 U/μl Taq DNA Polymerase (Thermo Fisher Scientific), and 5 μl of 10X Taq Buffer with 2.5 mM MgCl₂. Thermocycling conditions were: initial denaturation at 95°C for 30 s, 35 cycles of 94°C for 30 s, 65°C for 60 s, and 72°C for 60 s. The second round used 1 μL of the first-round product and internal primers under identical conditions. GAPDH as a control gene, amplified under the same conditions as the first round of nested PCR but with specific primers listed in Table S4. Thermocycling conditions included initial denaturation at 95°C for 5 min, 35 cycles of 95°C for 30 s, 65°C for 10 s, and 72°C for 6 s, followed by a final extension at 72°C for 5 min (Table S4). RT-qPCR for BCR-ABL1 was also performed as part of routine clinical diagnostics using the REALQUALITY RQ-BCR-ABL p210 & p190 One-Step Kit (AB Analitica), following the manufacturer’s instructions.

For SARS-CoV-2 detection, RT-qPCR was conducted using the GA SARS-CoV-2 OneStep RT-PCR Kit (Geneova), targeting the N gene with primer and probe sequences not publically available. Thermocycling conditions included reverse transcription at 53°C for 15 min, heat activation at 95°C for 3 min, followed by 45 cycles of 94°C for 5 s and 60°C for 25 s.

PCR-amplified products targeting FMDV and BCR-ABL1 were further validated by Sanger sequencing (Table S4 and Data S1).

Quantification and statistical analysis

Lateral flow results were interpreted visually and band intensity was quantified by measuring the pixel intensity of the test line using ImageJ image processing software. To evaluate assay precision and intra-assay variability, overall qualitative concordance was calculated using the formula: Overall Concordance (%) = (Number of Concordant Pairs/Total Number of Pairs [k]) × 100, where k represents the total number of sample pairs evaluated in duplicate and a Concordant Pair is defined as both replicates yielding identical calls (Pos/Pos or Neg/Neg). Selected multi-panel figures were re-assembled using the Inkscape vector graphics environment to ensure high-resolution formatting. For RT-qPCR analysis, standard curves were used to convert cycle threshold (Ct) values from FMDV and Brucella samples into RNA or DNA copy numbers. Copy number for each biological replicate was calculated as the average of duplicate Ct values. Sanger sequencing data were analyzed using SnapGene tool (version 8.0.2).