CRISPR/Cas13a-assisted amplification-free miRNA biosensor via dark-field imaging and magnetic gold nanoparticles

Abstract

MicroRNAs (miRNAs) are short (about 18-24 nucleotides) non-coding RNAs and have emerged as potential biomarkers for various diseases, including cancers. Due to their short lengths, the specificity often becomes an issue in conventional amplification-based methods. Next-generation sequencing techniques could be an alternative, but the long analysis time and expensive costs make them less suitable for routine clinical diagnosis. Therefore, it is essential to develop a rapid, selective, and accurate miRNA detection assay using a simple, affordable system. In this work, we report a CRISPR/Cas13a-based miRNA biosensing using point-of-care dark-field (DF) imaging. We utilized magnetic-gold nanoparticle (MGNPs) complexes as signal probes, which consist of 200 nm-sized magnetic beads and 60 nm-sized gold nanoparticles (AuNPs) linked by DNA hybridization. Once the CRISPR/Cas13a system recognized the target miRNAs (miR-21-5p), the activated Cas13a cleaved the bridge linker containing RNA sequences, releasing 60 nm-AuNPs detected and quantified by a portable DF imaging system. The combination of CRISPR/Cas13a, MGNPs, and DF imaging demonstrated amplification-free detection of miR-21-5p within 30 min at a detection limit of 500 attomoles (25 pM) and with single-base specificity. The CRISPR/Cas13a-assisted MGNP-DF assay achieved rapid, selective, and accurate detection of miRNAs with simple equipment, thus providing a potential application for cancer diagnosis.

Fig. 1. CRISPR/Cas13a-assisted magnetic nanoparticle (MGNP)-dark-field (DF) assay. A. Schematic diagram for the CRISPR/Cas13a-assisted MGNP-DF assay. In the presence of target miR-21-5p, the hybridization of the crRNA to miR-21-5p activates Cas13a proteins. Activated Cas13a proteins then cleave the bridge linkers between magnetic beads and gold nanoparticles (AuNPs). Subsequent magnetic separation of the magnetic beads releases AuNPs. These released AuNPs are captured onto a glass slide with black PDMS wells. The number of released AuNPs is quantified using a custom-built, portable DF imaging system. The entire assay can be completed within 30 min. B. Bacterial expression and purification of LwaCas13a. Purified LwaCas13a was shown by SDS-PAGE, followed by silver staining. C. Synthesis of miR-21-5p crRNA by in vitro transcription was confirmed by denaturing PAGE and SYBR gold staining. D. Scanning electron micrograph (SEM) of MGNP. Scale bar, 100 nm. E. Photograph of the portable DF imaging device. The overall size is 140 mm (L) × 85 mm (W) × 250 mm (H). F. 10× objective lens and slide glass holders coupled with a white LED. G. Representative zoomed-in DF images of AuNPs detached from magnetic beads by mock (left) or miR-21-5p treatment (right). Scale bar, 100 μm.

Fig. 2. Optimization of the CRISPR/Cas13a-assisted MGNP-DF assay. A. DF images of released AuNPs with transparent and black PDMS (left) chambers. Comparison of plot profiles between PDMS and black PDMS chambers (right). Scale bar, 100 μm. B. Comparison of the size of magnetic beads. Magnetic beads with a 200 nm diameter showed a higher signal of AuNPs than magnetic beads with a 1 μm diameter. *P < 0.05 compared with 200 nm and 1 μm, as assessed by unpaired t-test. Bar graphs are shown as mean ± SD. C. Comparison of the size of AuNPs. AuNPs with a 60 nm diameter showed a higher signal than AuNPs with an 80 nm diameter by miR-21-5p-mediated Cas13a activation. ns, not significant; ****P < 0.0001 compared with the mock sample, as assessed by two-way ANOVA with Bonferroni's multiple comparisons tests. D. The number of AuNPs released from 3 different designs of MGNPs (1; linker only, 2; bridge without poly-T, 3; bridge with poly-T) depending on the amount of miR-21-5p. The reaction volume was 20 μl. Error bars are shown as mean ± SD. E. The number of AuNPs detached from magnetic beads depending on the Cas13a reaction time. Error bars are shown as mean ± SD.

Fig. 3. Evaluation of sensitivity and specificity of the CRISPR/Cas13a-assisted MGNP-DF assay. A and B. Measurement of released AuNPs by seven different concentrations of miR-21-5p. A. Representative zoomed-in images for the 4 different target miRNA amounts. Scale bar, 100 μm. B. Quantitative analysis for the numbers of released AuNPs. The dotted line indicates the number of mean + 3 × standard deviation (s.d.) of the mock sample. Error bars are shown as mean ± s.d. from three independent experiments. C. Detection specificity was evaluated using wildtype (WT), single-mismatch (SM), and double-mismatches (DM) of miR-21-5p, miR-421, and miR-9-5p. Bar graphs are shown as mean ± SD from the three experiments. ****P < 0.0001 compared miR-21-5p (WT) sample, as assessed by two-way ANOVA with Bonferroni's multiple comparisons test.

Fig. 4. miR-21-5p detection from breast cancer cell lines. A. Representative zoomed-in DF images of released AuNPs by miR-21-5p-activated Cas13a. Scale bar, 100 μm. B. The number of released AuNPs in five different breast cancer cell lines. Bar graphs are shown as mean ± SD from the three independent experiments. C. Relative abundances of miR-21-5p, as analyzed by RT-qPCR. U6 snRNA levels were used as an internal control. Bar graphs are shown as mean ± SD from the three independent experiments. D. Dot plot for correlation coefficient. The number of released AuNPs from the CRISPR/Cas13a-assisted MGNP-DF assay and the relative abundance from RT-qPCR are indicated in the y- and x-axis, respectively. The black dashed line indicates the best linear fit. The two blue dashed lines indicate the 95% confidence interval.