Magnetic Bead-Quantum Dot (MB-Qdot) Clustered Regularly Interspaced Short Palindromic Repeat Assay for Simple Viral DNA Detection.

We have developed a novel detection system that couples clustered regularly interspaced short palindromic repeat-Cas recognition of target sequences, Cas-mediated nucleic acid probe cleavage, and quantum dots as highly sensitive reporter molecules for simple detection of viral nucleic acid targets. After target recognition and Cas-mediated cleavage of biotinylated ssDNA probe molecules, the probe molecules are bound to magnetic beads. A complementary ssDNA oligonucleotide quantum dot conjugate is then added, which only hybridizes to uncleaved probes on the magnetic beads. After separating hybridized quantum dots, the collected supernatant is illuminated by a portable ultraviolet flashlight, and it provides a simple "Yes-or-No" nucleic acid detection answer. By using a DNA target matching part of the African swine fever virus, detection limits of ∼0.5 and ∼1.25 nM are achieved in buffer and porcine plasma, respectively. The positive samples are readily confirmed by visual inspection, completely avoiding the need for complicated devices and instruments. This work establishes the feasibility of a simple assay for nucleic acid screening in both hospitals and point-of-care settings.

Introduction

Intensified infectious disease outbreaks occurred more frequently in the past few decades due to the increased human population and global travel and included viruses as SARS,[1,2] Ebola,[3,4] Zika,[5,6] MERS,[7,8] and the novel coronavirus (SARS-CoV-2).[9,10] The ongoing COVID-19 crisis has evolved as a global pandemic since it was first reported in December 2019. Due to the shortages of testing kits, vaccines, and treatments, the number of cases still rapidly increases in many countries. Since the development of vaccines and treatments for novel viruses can take as long as 18 months, the most efficient immediate approach to slow down and contain epidemics is to provide rapid and widespread testing followed by the isolation of positive cases and their close contacts. The current standard tests for viral infection use highly sensitive and specific real-time polymerase chain reaction (RT-PCR).[11,12] Although these tests are highly sensitive accurate, they require a laboratory setting for use, bulky instruments, sophisticated operations, and they are therefore not field-deployable. These issues present a significant bottleneck to rapid and widespread testing, hindering our ability to respond to emerging pandemics. Recently, rapid point-of-care (POC) systems using isothermal amplification were demonstrated for viral pathogen detection;[13] however, they have very limited throughput and are not field-deployable due to the complexity of the instrumentation. Given the limitations of existing viral pathogen testing methods, there is a significant need for simpler testing methodologies that do not require complex instrumentation and that can be performed easily in the field.

An ideal POC device that can combat modern era pandemic crises should be simple, inexpensive, and accurate and must be easily operated and understood by people without special training. A recently discovered technology called clustered regularly interspaced short palindromic repeats (CRISPR) in conjunction with CRISPR-associated proteins (Cas) showed great promise as an alternative to RT-PCR detection.[14,15] This method is significantly simpler than RT-PCR since it does not require complicated instrumentation to perform. Certain Cas nucleases such as Cas12a and Cas13a can be activated to indiscriminately cleave single-stranded DNA or RNA substrates once they identify and bind with the target DNA or RNA.[16,17] The activated Cas enzymes cleave a nucleic acid probe, releasing a fluorophore from a quencher molecule, thereby generating signals. CRISPR-Cas systems have also been linked with isothermal amplification methods to achieve highly sensitive and specific detection of pathogens.[18,19] However, most of the CRISPR-Cas detection systems utilize reporter probes with organic dyes and quenchers that require external instruments and possess a high fluorescence background, which limits overall sensitivity.[20]

In this study, we develop a novel probe system for CRISPR-Cas nucleic acid assays that use quantum dots (Qdots) as a reporter. Compared to conventional organic fluorescence dyes, Qdots have unique electronic and optical properties such as stable signal intensity, a narrow emission spectrum, a continuous broad excitation spectrum, and high quantum yields, making them an ideal fluorescence indicator for in vitro diagnostics.[21,22] After target recognition and Cas-mediated cleavage of biotinylated ssDNA probe molecules, the probe molecules are bound to streptavidin-coated magnetic beads. A complementary ssDNA oligonucleotide quantum dot conjugate is then added, which only hybridizes to uncleaved probes on the magnetic beads. After separation of hybridized from unhybridized quantum dot conjugates by magnetic sequestration, the signal is measured fluorometrically to provide a signal to indicate the cleavage activity and therefore the presence of target nucleic acid. Our use of magnetic bead sequestration effectively reduces the background signal, enabling use of highly sensitive quantum dots as reporters. As a proof of principle for viral pathogen detection, a double-stranded DNA segment matching a portion of the African swine fever virus (ASFV) genome was selected as the detection target.

A CRISPR-Cas12a system, considered as a revolutionary tool for biosensing applications, can identify double-stranded DNA targets and indiscriminately degrade single-stranded DNA.[23,24,14] Unlike Cas9 proteins, Cas12a proteins are guided by a single-stranded RNA sequence without any trans-activating CRISPR RNAs (tracrRNAs), thereby making them simpler to design and operate.[25,26] In this work, a Cas12a enzyme is chosen mainly because it can be activated by dsDNA, corresponding to the structure of ASFV. Generally, the Cas12a enzyme relies on a short DNA sequence, also called as a protospacer adjacent motif (PAM), adjacent to the complementary crRNA-dsDNA regions for target recognition. Since the Cas12a enzyme requires a T nucleotide-rich PAM sequence adjacent to the complementary crRNA-dsDNA regions for nucleic acid identification,[23,24,14] we chose the sequence following TTTC in the ASFV genome as the matching regions for Cas12 identification (Figure S2). Previous studies showed that CRISPR-Cas12a can distinguish single- or double-base mismatches between the crRNA and double-stranded DNA target.[27,28] In this work, we chose the matched target and mismatched sequences both from the B646L gene of ASFV (accession number NC_044959). Without target amplification, we achieved a detection limit of 0.5 nM by using a small and inexpensive UV flashlight as an excitation source. No fluorescence signal was detected in mismatched samples, demonstrating the high selectivity of our strategy. Even though a vision difference can lead to inevitable ambiguity for the visual detection method, our sensitivity is determined based on the photographs that were taken by the digital cameras with much lower light sensitivity than human eyes.[29,30] To verify the reliability of visual detection, the fluorescence intensity of every sample was also measured by a laser-based fluorometer and was consistent with naked-eye observation. Our new protocol adds a bead-washing protocol, thus significantly reducing the high fluorescence background originated from dye-quencher substrates in bodily fluids.[31] A comparable sensitivity in porcine plasma and buffer is achieved without additional sample preparation steps. This simple and colorimetric protocol paves the way for next-generation POC and field-deployable viral pathogen detection applications.

Results

Figure a shows the workflow of the Qdots-CRISPR-based colorimetric nucleic acid detection. Cas12a first binds with crRNA, forming a Cas12a-crRNA complex. Then, oligonucleotide DNA linker probes with a biotin label are added into the complex as substrates. With the presence of a double-stranded target DNA (part of the ASFV genome), the Cas12a-crRNA complex is activated and indiscriminately cleaves the single-stranded linker probes in the assay. After the reaction, streptavidin-coated magnetic beads are added, which bind both cleaved and uncleaved linker probes. A complementary oligonucleotide DNA reporter probe, which is conjugated to quantum dots, is then added. The reporter probe only binds to uncleaved linker probes on the magnetic beads. An external magnet is then applied to isolate the magnetic beads, leaving the unbound reporter probes in the eluate for fluorescence analysis. Since the cleavage of the linker probe is related to the amount of reporter probes remaining in the eluate, the fluorescence signal indicates the presence of the starting target in the sample.

Figure 1

(a) Scheme for the Qdots-CRISPR-based colorimetric nucleic acid detection strategy. (b) Schematic of the simple detection by using Qdots. (c) TEM images: (i) streptavidin-coated Qdots, (ii) streptavidin-coated magnetic beads, and (iii) MB-Qdot conjugation. The MB-Qdot conjugation includes magnetic beads, linker probe A, reporter probe B, and Qdots. Scale bars are 50 nm.

Figure b further shows the workflow of our detection scheme: cleaved and uncleaved linker probes are immobilized on the magnetic beads followed by the hybridization with Qdot reporter probes. Magnetic isolation separates unhybridized reporter probes, leaving probes in the eluate that indicates the presence of the starting target nucleic acid. Then, fluorescence measurement can be performed with the illumination of an inexpensive and portable flashlight.

Transmission electron microscopy (TEM) images of the streptavidin-coated quantum dots (mean diameter of 17.5 nm), streptavidin-coated magnetic beads (mean diameter of 1 μm), and hybridization of the Qdot reporter probe onto linker probes bound to the magnetic beads are presented in Figure c. Numerous Qdots are immobilized on the magnetic bead by the complementary hybridization of reporter probes and linker probes.

A custom-designed fluorometer was used to confirm the successful formation of conjugates. As shown in Figure a, streptavidin-coated Qdots (156 fmol) show a fluorescence intensity of ∼24 counts (Qdot, green). MB-Qdot conjugation composed of magnetic beads (100 μg), linker probes A (25 pmol), reporter probes B (12.5 pmol), and Qdots (156 fmol) was mixed in a solution of 60 μL and pulled down by a magnet. Since the input Qdots were saturated by the other three components, no fluorescence signal was detected in the supernatant (conjugation, red). To confirm that the MB-Qdot conjugation was made by the hybridization of two DNA strands, we mixed magnetic beads, reporter probes, and Qdots without the input of linker probes. After magnetic isolation, a high fluorescence intensity (∼14 counts) was detected, indicating that the MB-Qdot conjugate is mainly achieved by DNA complementary hybridization rather than non-specific absorption (control, blue). In the future, streptavidin-coated magnetic beads can be treated with blocking reagents such as bovine serum albumin[32] to reduce the non-specific binding. To further confirm the MB-Qdot conjugation, the beads were washed 5 times followed by DNase I (EN0521, ThermoFisher Scientific) treatment to degrade the hybridized DNA strands. As shown in Figure b, a distinct fluorescence peak was detected in the conjugation sample (red), while no signal was observed in the control sample (blue). The integrated signals of the fluorescence spectra (550–700 nm) were plotted (inset of Figure b), where the conjugation sample shows ∼2000 counts, which is approximately 13 times higher than that of the control sample. In addition, for the conjugation sample, the bright pink fluoresence signal is easily observed by using a flashlight.

Figure 2

Validation of the MB-Qdot conjugation formation. (a) Emission curves of collected supernatants. The conjugation sample consists of magnetic beads, DNA probes, and Qdots. The control sample consists of magnetic beads, Qdots, and reporter probes. The Qdot sample is the streptavidin-coated Qdots (2.6 nM). (b) Fluorescence signals of released Qdots by DNase treatment versus streptavidin-coated magnetic beads. The inset shows the integrated fluorescence signals and photographs of the suspensions under UV illumination. Error bars are standard deviation of the mean.

We then optimized the MB-Qdot conjugation assay to decrease the fluorescence background, which is essential for the CRISPR-Cas12a-based detection strategy. Linker probes (100 pmol) with 3′ FAM modification (/5BioTinTEG/TTATTCTTATTGTGTGAACTGCTCCTTCTTGACTCCACC/36-FAM/) were introduced to bind with the magnetic beads at weights of 0, 25, 100, 400, and 600 μg at room temperature for 15 min. After incubation, the magnetic beads were isolated, and the fluorescence intensity of the unbounded ssDNA was measured by using a commercial spectrofluorometer (JASCO FP-8500, USA). As shown in Figure a, the integrated fluorescence intensity (500–700 nm) dramatically drops with an increase in the magnetic bead input from 0 to 400 μg. Adding more magnetic beads does not decrease the fluorescence intensity, indicating that the saturated ratio of magnetic beads to linker probe A is ∼4 μg:1 pmol. Next, we optimized the ratio of reporter probes B to Qdots, which is the key to eliminate the fluorescence background. One pmol of Qdots was selected, and the load of reporter probe B was varied from 0 to 80 pmol. As shown in Figure b, a ratio of 80:1 between reporter probe B and Qdots shows the lowest background and is used for nucleic acid detection. To determine the optimized hybridization condition of linker probe A and reporter probe B, Qdots with a load of 0.5 pmol were premixed with 40 pmol of reporter probe B (ratio of 1:80) and then allowed to hybridize with magnetic bead-linker probe A conjugation (ratio of 4 μg:1 pmol). The retrieved Qdots were then measured by a custom-designed fluorometer. The fluorescence intensity of the supernatant drops significantly by increasing the load of linker probe A from 20 to 50 pmol (Figure c). However, the intensity only shows a slight change when the load is further increased to 80 and 100 pmol. Thus, we concluded that the saturated ratio of linker probe A to reporter probe B is ∼2:1.

Figure 3

Optimization of the MB-Qdot conjugation. (a) Integrated fluorescence signal of unbound linker probe A versus the input load of magnetic beads (0–600 μg). (b) Integrated fluorescence signal of unbound Qdots versus the input load of reporter probe B (0–80 pmol). (c) Integrated fluorescence signals of the collected Qdot supernatant versus the input load of linker probe A (0–100 pmol). Insets show the photographs of Qdots under UV illumination. Error bars represent standard deviation of the mean.

With the optimized MB-Qdot conjugation assay in hand, we then applied this protocol for CRISPR detection. We added target DNA into a tube containing 50 nM Cas12a, 62.5 nM crRNA, and 1.25 μM linker probe A with a total volume of 20 μL. The detection was performed in the dark, and the flashlight was placed ∼40 cm over the Qdot supernatants. Photographs of the supernatants were taken by a digital camera with 1/25 s exposure time and an 800 iso value and used as the standard to determine the detection sensitivity. As presented in Figure a, under flashlight illumination, a distinct pink color is observed with a target concentration higher than 0.5 nM. No distinct fluorescence difference with a concentration higher than 0.75 nM was detected. This is caused by the high turnover rate of the CRISPR-Cas12a complex, degrading almost all the linker probes and leaving the same amount of Qdots-reporter probe B conjugates in the supernatant. On the other hand, the collected supernatants do not show any fluorescence signals without a target input or with a mismatched target sequence. These results were then validated by using the custom-designed fluorometer and are presented in Figure b. The intensities of matched target DNA samples with concentrations of 0.5, 0.75, 1, 2.5, 10, and 50 nM are significantly higher than those of the mismatched or blank samples. The experiment was repeated three times on different days, and each sample was measured three times to ensure the accuracy of the results. These results were further confirmed by TEM imaging. As shown in Figure c, for blank and mismatched samples, Qdots with an average diameter of 17.5 nm were linked on the magnetic bead via linker probe A and reporter probe B hybridization. On the other hand, for samples containing target DNA, Qdots are not observed in the TEM image, indicating that the linker probes were denatured and left the Qdots suspended in the solution for detection.

Figure 4

Qdots-CRISPR-based nucleic acid detections. (a) Fluorescence images of samples containing (i) matched target DNA (positive), (ii) mismatched target sequence (negative), and (iii) blank (no target input). The samples were illuminated by a UV flashlight (wavelength of 395 nm). (b) Integrated fluorescence signal (624–1022 nm) of (i) matched target DNA (black triangle), (ii) mismatched target sequence (green triangle), and (iii) blank sample (0 nM). The red dashed line shows the logistic fit of the integrated fluorescence intensity versus DNA concentration. (c) TEM images of the magnetic beads mixed with (i) matched target DNA, (ii) mismatched target sequence, and (iii) blank (no target input).

To verify that our MB-Qdot assay can directly work with bodily fluids, we spiked target DNA into porcine plasma (P2891-10 mL, Sigma-Aldrich, MO, USA) and applied our MB-Qdot assay. As shown in Figure , a clear pink color is observed with a target concentration higher than 1.25 nM, whereas the blank sample shows no color. To validate the visual readout, the supernatants were measured by a custom-designed fluorometer. The fluorescence intensities of target concentrations at 1.25, 1.5, 2, and 10 nM are all higher than those of the samples without a target.

Figure 5

Integrated fluorescence spectra (624–1022 nm) of Qdots-CRISPR-based nucleic acid detection in porcine plasma, measured by a custom-designed fluorometer. The samples are composed of matched target DNA spiked into porcine plasma. The red dashed line represents the logistic fit of integrated fluorescence signals versus DNA concentration. (insets) Fluorescence images of each sample with flashlight illumination (excitation of 395 nm).

Discussion

The work presented here leverages a CRISPR-Cas12a assay and uses Qdots as an indicator. The product of the quantum yield and absorption coefficient in Qdots is ∼3 fold higher than that of organic dyes, demonstrating Qdots as an ideal reporter for simple visual detection.[33,21,34] CRISPR-Cas12a cleavage is a robust target recognition method that can maintain high cleavage efficiency even after 16 h incubation.[31] Preassembled crRNA and Cas12a can also protect crRNA from degradation.[35] Target-specific CRISPR cleavage was performed at 37 °C, while other experiments were performed at room temperature. Unlike methods such as PCR,[36,37] electrochemical sensing,[38,39] and surface-enhanced Raman spectroscopy,[40,41] our assay does not require time-consuming sample processing, bulky sensing instruments, or complicated expert operation and data interpretation. These unique advantages enable us to demonstrate a simple diagnosis assay, ideal for on-field applications. Our strategy can provide a simple “Yes-or-No” answer without trained personnel, ideal for massive screening during outbreaks. It can be used in many locations, such as stadiums, airports, and customs without relying on bulky and complicated instruments.

Using our Qdot-CRISPR-based DNA detection strategy, we report a very low detection background compared to conventional CRISPR-Cas systems that use fluorophore-quencher substrates as reporters. The cleaved ssDNA probes linked to the Qdots are detected by isolating the magnetic beads, thus avoiding the high background issues from the uncleaved fluorophore-quencher probes in the assay. This is essential to enhance the overall detection sensitivity of CRISPR-Cas systems and is especially useful for simple detection in resource-limited environments.[42,43] More importantly, the high background originated from fluorophore-quencher substrates in unpurified bodily fluids is greatly reduced, which is the key to reduce complicated sample preparation for “real-world” diagnostics.

We previously developed an integrated POC system for ASFV detection by using a laser-based fluorometer and conventional dye-quencher substrates.[31] A high fluorescence background originated from dye-quencher probes in unpurified porcine plasma was observed, and it severely affected the detection sensitivity. In this work, we developed a magnetic bead-Qdot conjugation protocol. Qdots provide much higher fluorescence signals than organic dyes and can be excited by a portable flashlight. The utilization of magnetic beads reduces the background from the substrates in plasma, achieving comparable sensitivity in plasma and buffer solutions.

Unlike PCR that requires ultra-purified buffer solutions,[44,45] one of the unique advantages of CRISPR-based detection is that the collateral cleavage process is not affected much by the presence of nucleases in bodily fluids. For example, numerous studies have shown the sensitive detection of nucleic acid samples in urine,[46] saliva,[17] and plasma[47] samples without stringent purification. However, to meet the requirements of “real-world” POC diagnostics, a nucleic acid concentration step may still be needed as the lysis process could result in the dilution and loss of target samples, especially when the viral load is low. In addition, CRISPR-based detection can only work with known genomic sequences, which may take time to respond during emerging outbreaks.

Currently, our approach can achieve a detection limit of ∼0.5 nM without any target amplification. To improve the detection sensitivity, our assay can be easily integrated with well-established nucleic acid amplification methods such as polymerase chain reaction (PCR),[48] recombinase polymerase amplification (RPA),[16,49] and loop-mediated isothermal amplification (LAMP).[50,51] The amplification reagents can be mixed with CRISPR-Cas12a proteins and crRNA as a “one-pot” assay.

Furthermore, our assay is poised to be developed for high-throughput multiplexing detection purposes. The current method can achieve high-throughput detection as multiple samples can be excited at the same time. In the future, robotic pipetting[52,53] and automated microfluidics[54,55] can be integrated with our assay to further increase the throughput. Because CRISPR-Cas12a assay is an indiscriminate DNase activity,[14,31] linker probe A and reporter probe B can be used as universal reporters for the detection of various DNA targets by only changing the crRNA sequence to match the target sequence. In addition, Cas13a proteins demonstrate a similar indiscriminate RNase activity to Cas12a.[56,16] Our assay could work with RNA-based viruses by including a well-established reverse transcription process or changing the DNA probes to RNA probes.[57,58] Thus, our assay presented here should be able to detect both DNA and RNA targets for multiplexing detection.

Conclusions

To avoid bulky and complicated sensing instruments, in this work, we developed a simple visual detection system coupled with quantum dots as an ultra-brightness indicator and CRISPR-Cas12a assay for isothermal viral DNA target sensing. Illuminated by a handheld flashlight, the positive samples are visually distinguishable from negative samples. Multiple samples can be simultaneously excited, making it ideal for rapid and large-scale POC screening. Furthermore, by introducing a simple magnetic bead purification process, a low fluorescence background in plasma samples was detected, achieving a comparable detection sensitivity within buffer. Our work lays down the foundation for future molecular diagnostics without using expensive, tedious, and time-consuming sample purification and detection instruments.

Materials and Methods

Reagents

Both the matched target and mismatched sequence were chosen from the B646L gene in the African swine fever virus strain (accession number NC_044959), corresponding to the previous outbreaks.[59] CRISPR RNA (crRNA) contains two regions, a short direct repeat and a spacer that has a complementary sequence to the ASFV target. Single-stranded linker DNA probes and reporter probes were both labeled with biotin tags. Binding buffer (1×) is composed of 50 mM NaCl, 5 mM MgCl2, 10 mM Tris-HCl, pH 7.5, and 0.1 mg/mL BSA. Synthetic RNA oligonucleotides were purchased from IDT, Inc., and the pellets were suspended in TE buffer (10 nM Tris, 0.1 nM EDTA, pH 8.0). Nuclease-free water and RNase-free microtubes were used to prevent RNA degradation.

Instruments

A 395 nm flashlight (brand, Vansky; part number, VS-FL03-V; spectral bandwidth of 30 nm) with 5 W output power was used to illuminate the Qdots. A commercial spectrofluorometer (JASCO FP-8500, USA) and a laser-based fluorometer were both used to measure the fluorescence signals. The design of the sensitive fluorometer was presented in our previous works.[31,56] Briefly, liquid samples were loaded in a disposable cartridge and then placed on top of the fluorometer. The fluorometer comprises a continuous laser (488 nm, 3 mW power) to excite the samples, an off-axis parabolic (OAP) mirror to collect fluorescence signals, and a 488 nm notch filter to filter excitation light. An optical fiber (M93L01, Thorlabs) was used to transmit the fluorescence signals, and a USB spectrometer (USB 2000+, Ocean Optics) was used to record the spectra.

CRISPR-Cas12a-Based Linker Probe A Cleavage

The total volume of CRISPR-Cas12a-based cleavage assay was 20 μL including LbCas12a proteins (New England BioLabs, Inc.), crRNA, linker probe A, and binding buffer. Briefly, 50 nM LbCas12a was assembled with 62.5 nM crRNA at room temperature for 10 min. Afterward, 1.25 μM linker probe A, 1× binding buffer, 14.5 μL of nuclease-free water (IDT, Inc.), and target DNA (concentration ranging from 0.1 to 50 nM) were mixed with the LbCas12a-crRNA complex to initiate the reaction. The mixture was incubated in a water bath at 37 °C for 120 min to ensure optimized cleavage activity. All the nucleic acids used in the assay were purchased from IDT, Inc., and the sequences are given in Table .

Table 1

Sequences of DNA and crRNA Used in This Study

linker probe A	/5BioTinTEG/TTATTCTTATTGTGTGAACTGCTCCTTCTTGACTCCACC
reporter probe B	/5BioTinTEG/CTGTCGTGGTTCTAGGGAGTCAAGAAGGAGCAGTTCACATTT
matched target DNA (positive control)	NTS: 5′-...TTTCATCGGTAAGAATAGGTTTGC...-3′
	TS: 5′-...GCAAACCTATTCTTACCGATGAAA...-3′
mismatched target sequence (negative control)	NTS: 5′-...TTTGACCACCCCAGTCATATCCGT...-3′
	TS: 5′-...ACG GAT ATG ACT GGG GTG GTC AAA... -3′
crRNA	uaauuucuacuaaguguagauAUCGGUAAGAAUAGGUUUGC

Linker Probe A Isolation via Magnetic Beads

First, 10 μL of magnetic beads (Dynabeads MyOne Stretpavidin C1, ThermoFisher Scientific) was washed with PBS buffer (Gibco TM, PH 7.4) three times and resuspended in 10 μL of PBS buffer. Next, the washed magnetic beads were transferred to a new 0.2 mL PCR tube, and the supernatant was discarded by isolating the magnetic beads with a magnetic rack (NEBNext magnetic separation rack, New England BioLabs, Inc.). The collateral cleaved products by CRISPR were then added to the magnetic beads at room temperature for 15 min. After incubation, the magnetic bead-linker probe A conjugation was washed and resuspended in 20 μL PBS buffer for later use.

Reporter Probe B Conjugation with Quantum Dots (Qdots)

Qdots (Qdot 605 ITK streptavidin conjugate kit, ThermoFisher Scientific) with a concentration of 15.6 nM were mixed with 1.25 μM reporter probe B (IDT, Inc.) to reach a 10 μL reaction mixture. To avoid photo-bleaching issues, the mixture was incubated in a dark environment at room temperature for 15 min.

Hybridization of Linker Probe A and Reporter Probe B

The supernatant of the magnetic bead-linker probe A conjugation was first evacuated followed by mixing with the reporter probe B-Qdot conjugation. The mixture was placed on a tube rotator (BS-RTMNI-1, STELLAR SCIENTIFIC) and allowed to react at room temperature under mild shaking for 90 min. After that, the supernatant was collected and transferred to a new PCR tube.

Colorimetric Detection and Fluorescence Quantification

For naked-eye detection, the samples were placed 40 cm away from a portable flashlight (wavelength of 395 nm) in a dark room. Sensitivity is determined based on the photographs taken by a digital camera with 1/25 s exposure time and an 800 iso value in the dark. To validate the naked-eye detection, every sample was also quantified with a custom-designed and calibrated fluorometer. As reported before,[31] the sample was added in a disposable cartridge and excited by a continuous wave semiconductor laser (wavelength of 488 nm). The collected fluorescence signal was analyzed by a portable spectrometer (USB 2000+, Ocean Optics).

TEM Characterization

Transmission electron microscopy (TEM, JEOL 2010) with a point resolution of 0.23 nm was used to characterize the samples. The supernatant containing magnetic beads was deposited onto a carbon-coated copper grid and left to evaporate. The TEM images were then captured through an AMT XR81 bottom mount camera with a 60 kV electron microscope.