1. Introduction

Commercially-available assays for pathogen nucleic acid diagnostics heavily rely on biological amplification steps to increase the amount of target material via replicating ribonucleic acid (RNA) or deoxyribonucleic acid (DNA) of the pathogen genome [1,2]. These biological amplification steps increase the test complexity and processing time and may sometimes generate false positive results [3]. On the other hand, adaptation of conventional photonic and plasmonic biosensing platforms [4 –6] for early-stage detection poses challenges of returning false negative results if the amount of viral RNA in the sample is below the sensor detection limit. Standard photonic biosensors make use of surface functionalization by probes that can selectively capture target molecules, and the detection limit is defined by how many molecules must be captured before a response from the sensor can easily be distinguished from its inherent noise. Furthermore, when new pathogens emerge, it may take up to several months to develop new probes specific to the target [7].

As a potential solution for realizing reliable early-stage detection with conventional optical biosensors, we recently proposed via simulations a new cleavage-based sensing approach, high contrast cleavage detection (HCCD), that can provide a highly sensitive readout based on the removal of reporter molecules from a photonic biosensor surface [8]. The HCCD sensing modality is the reverse of standard binding/adsorption-based label-free optical biosensing methods [9], which depend on the small difference in refractive index between samples such as blood or saliva and biomolecules binding to the sensor surface. In HCCD, one specific molecular detection event triggers the non-specific cleavage of numerous reporters from the sensor surface in a manner similar to that employed in clustered regularly interspaced short palindromic repeats (CRISPR)-based diagnostics [10]. By using nucleic acid reporters that are labeled with high contrast nanoparticles, such as semiconductor nanocrystals, quantum dots, or metal nanoparticles, HCCD enables additional signal amplification beyond that achieved by unlabeled reporters due to their larger refractive index change and/or stronger perturbation of the optical mode when they are cleaved from the photonic sensor surface. Hence, HCCD offers all the advantages of miniaturization and resonant optical signal amplification of conventional photonic sensors, while additionally significantly increasing the sensitivity and signal-to-noise ratio. Furthermore, since sensor functionalization with the high contrast reporters can be done prior to the sample collection from the patient or the environment, the time to detection can also be shortened significantly.

In order to validate the working principles of HCCD and its possible applications in RNA and DNA detection, deoxyribonuclease (DNase) enzyme sensing experiments were performed using a porous silicon (PSi) interferometric sensor functionalized with either high-contrast ds-DNA-quantum dot (Qdot) or low-index ds-DNA-fluorescein (FAM) reporters. DNase serves as a model for activated CRISPR-Cas complexes (e.g., CRISPR-Cas12 and CRISPR-Cas13) that perform non-specific, collateral cleavage of nearby DNA or RNA when triggered by the presence of target pathogen DNA or RNA that is complementary to the guide RNA in the CRISPR-Cas complex. Hence, the present experiments prove that when collateral cleavage is initiated, the HCCD approach works on a traditional optical biosensing platform. PSi interferometer sensors are a well-studied, simple, and promising optical biosensing platform for low-cost diagnostics [11 –13]. This platform offers large surface-to-volume ratio, simple optical reflectance readout, and surface functionalization by using well-developed surface chemistries. Although DNA detection has been demonstrated using PSi sensors, challenges in achieving high sensitivity detection remain. Typically, experimental detection of µM concentrations of DNA are reported. While signal amplification through quantum dot attachment to target DNA [14,15] and electrokinetic focusing [16] to concentrate target DNA lead to improved sensitivity, these approaches are not compatible with simple, real-time diagnostics. In this work, we not only demonstrate that HCCD detection is a promising approach to improve the sensitivity of PSi sensors in a straightforward manner, but we also importantly establish the feasibility of the HCCD detection approach for future implementation on resonant silicon-based optical sensors, which have the potential to reach the sensitivity necessary for amplification-free PCR-quality assays.

2. Working principle of PSi reflective interferometer with the HCCD mechanism

Single layer PSi interferometers work on the principle of thin film interference. Light reflected from the top (air/PSi) and bottom (PSi/silicon substrate) interfaces of the PSi layer interfere, giving rise to characteristic Fabry-Perot reflectance fringes (Fig. 1(a)). The fringe positions and the periodicity of the interferometric pattern are determined by the effective optical thickness of the PSi film:

 

Fig. 1. (a) Interference spectrum collected from an oxidized PSi layer under normal illumination by a broadband white light source. (b) Schematic of the HCCD-based reflectance interferometry detection platform with a PSi layer as the optical transducer. The PSi layer is functionalized with quantum dot-labeled reporters using the appropriate surface chemistry and the biotin-streptavidin conjugation mechanism, as described in the methods section. The Qdot reporters are then cleaved by the exposure of the sensor to the DNase enzyme and washed away, which generates a measurable optical signal change in the form of the Fabry-Perot fringes spectral blueshift. Here, DNase serves as a model for activated CRISPR-Cas12, which causes collateral cleavage of DNA when triggered by the presence of target pathogen DNA or RNA that is complementary to the guide RNA in the CRISPR-Cas complex. The inset shows a fluorescence image of the Qdots.

Download Full Size | PDF

At the same time, the pore size is chosen to be larger than both the molecular target and the molecules used in the surface functionalization process, to allow sufficient space for the molecular species to infiltrate into the large surface area of the porous material. When molecular species enter the pores, the effective refractive index of the PSi layer increases, and the fringes redshift. Conversely, when species (or dielectric nanoparticles) are removed from the pores, the effective refractive index of PSi decreases and the fringes blueshift. In the proof-of-concept experiment carried out in this work to validate the HCCD detection approach, as illustrated in Fig. 1(b), a PSi interferometer is first functionalized with high-index biotin-dsDNA-Qdot reporters, leading to a redshift of the reflectance fringes. When an enzyme (DNase) is used to cleave the DNA, leading to removal of the Qdots, the reflectance fringes blueshift. The magnitude of the reflectance shifts are proportional to the quantity of material being added to or removed from the pores, and can provide a mechanism for a sensitive and quantitative optical readout. In the following sections, we describe the methodology in detail.

3. Experimental methods and procedure

3.1 Porous silicon fabrication and surface modification

PSi thin films used for the sensing experiments in this work were fabricated by electrochemical etching of boron doped p-type silicon wafers ($100$, 0.01–0.02 Ω cm, 500-550 µm thick) in a solution of 15% hydrofluoric acid (49% aqueous) in ethanol at a current density of 70 mA/cm² for 100 s. We note that prior to fabricating these PSi films, a sacrificial PSi layer was first etched and then removed in NaOH solution to ensure that there were large pores openings at the surface [17]. After the removal of the sacrificial layer and again after formation of the PSi film used for sensing, chips were washed with ethanol and dried with nitrogen. Surface modification followed a previously established procedure [15]. First, to provide surface passivation and prepare for silanization, freshly etched PSi chips were thermally oxidized at 800 °C for 30 min. Chips were subsequently rinsed with ethanol and dried with nitrogen. The thickness, average pore diameter, and porosity of the PSi layer were measured by scanning electron microscopy (SEM), and found to be 3.99 µm, 50 nm, and 73%, respectively. Electrochemical etching typically produces a Gaussian-like distribution of pore sizes with a standard deviation of approximately 10 nm [15]. The nominal 50 nm pore size used in this work was selected to balance the desire for a high internal surface area where the reporters can be attached with the mass transport challenges that arise when the pore size is too small compared to the relevant molecule sizes.

The oxidized samples were incubated in freshly prepared 4% 3-aminopropyltriethoxysilane (APTES) solution (1 mL deionized water, 920 µL methanol and 80 µL APTES) for 15 min. The samples were then soaked in methanol for 15 min. Excess APTES was removed by rinsing with deionized (DI) water and ethanol. The samples were dried with nitrogen, and then annealed at 150 °C for 10 min, rinsed with methanol and dried with nitrogen. Next, the samples were incubated in a 2.5% glutaraldehyde solution (50 µL of 25% glutaraldehyde and 950 µL 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer, pH = 7.4) for 30 min. Excess glutaraldehyde was removed by incubating each sample in 1 µL of sodium cyanoborohydride (5 M cyanborohydride + 1M NaOH) in 100 µL HEPES buffer solution for 30 min, which is also necessary to stabilize the Schiff base. Finally, streptavidin was attached to the surface by exposing the samples to a 10 µM streptavidin solution in phosphate-buffered saline (PBS) buffer (pH = 7.4) for 30 min, and the Schiff base was stabilized by repeating the reducing step with sodium cyanoborohydride. To minimize non-specific binding, the sample was incubated in 3 M ethanolamine in buffer with pH = 9.0 for 30 min. The sample was then thoroughly rinsed with DI water and dried with nitrogen.

3.2 Synthesis of quantum dot and FAM dye labeled reporters

To synthesize biotin-dsDNA-Qdot reporters, 0.2 µM Qdots (QdotTM 605 ITK Streptavidin Conjugate Kit, ThermoFisher Scientific) were first mixed with 5 µL of 100 µM biotinylated ssDNA (5’ biotin- CTGTCGTGGTTCTAGGGAGTCAAGAAGGAGCAGTTCACATTT 3’; IDT, Inc.) in PBS buffer (pH = 7.4) to reach a total volume of 36.2 µL. The mixture was incubated in the dark at room temperature for 15 min. Next, 10 µL of complementary biotinylated ssDNA (IDT, Inc.) in 1x TE buffer (10 mM Tris + 0.1 mM EDTA) with a concentration of 100 µM was added and reacted with the mixture under mild shaking with a tube rotator (BS-RTMI-1, STELLAR SCIENTIFIC) for 90 min. We note that the specific DNA sequence does not strongly influence the cleavage-based sensing mechanism, and the particular length of the DNA molecule was selected to enable a relatively large signal change upon cleavage while maintaining reasonable mass transport in the pores.

To synthesize FAM dye labeled reporters, two biotinylated complementary DNA strands at 100 µM concentration with FAM dye tags (IDT, Inc.) were hybridized in PBS buffer (pH = 7.4) via a 90-minute incubation under mild shaking with the tube rotator.

3.3 Reporter attachment to porous silicon

Biotinylated dsDNA-Qdot reporters were attached to the streptavidin-modified PSi surface by incubating the sample in 24 µL of 135 nM reporter solution overnight. The footprint of PSi exposed to the reporter solution was 0.25 cm². After incubation, the supernatant was collected, and the sample was rinsed thoroughly with DI water and dried with nitrogen. The same procedure was repeated on a separate PSi sample using 135 nM biotinylated dsDNA-FAM reporters. The 135 nM reporter solution concentration was selected to ensure a measurable reflectance change resulted from cleavage of the reporters from the PSi surface.

3.4 DNase cleavage of the reporters

Each PSi sample was incubated in 12.5 nM DNase I solution (2 µL DNAse solution (1U/µL), 2 µL buffer and 16 µL DI water) for 2 h at 310 K in a moist environment. The samples were rinsed thoroughly with DI water and dried with nitrogen.

3.5 Measurements

Reflectance measurements were carried out using a Newport Q series lamp and fiber-coupled Ocean Optics USB-4000 spectrometer. Attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectroscopy measurements were taken using a Hyperion 2000/Vertex 70v system with a silicon prism. Fluorescence measurements were performed on a Jobin Yvon Fluorolog-3 FL3-111 spectrofluorometer.

3.6 Simulations

We validated the experimental measurements with transfer matrix calculations [18]. The refractive index of the functionalized PSi matrix with and without reporters was modeled with the two-component and the three-component Bruggeman’s effective medium theory [19], respectively. The thickness of the organic layer that is used to capture the Qdot reporters (APTES-glutaraldehyde-streptavidin) was estimated to be 3.5 nm based on an assumed 50% surface coverage of streptavidin molecules in the pores and near monolayers of APTES and glutaraldehyde. This leaves a 54% volume fraction of air in the total sensor volume, assuming the initial PSi porosity of approximately 73%. According to Bruggeman’s effective medium theory, the effective refractive index of the PSi matrix with the organic layer (${n_m})\; $ can be calculated as follows:

 

Fig. 2. (a) FFT of reflectance spectrum of PSi after various surface modification steps. (b) Corresponding ATR-FTIR spectra of PSi measured after each surface modification step in (a); important peaks are labeled.

Download Full Size | PDF

The addition of biotin-dsDNA-Qdot reporters – which can only occupy the space within the air pores – was modeled via the three-component Bruggeman’s model:

Combining all these effective index approximations, the filling fraction of the reporters attached in the void space of PSi was calculated based on the comparison to experimental spectra taken after the Qdot attachment step. The simulated reporter filling fraction was adjusted such that the simulated spectrum after reporter attachment matched the experimentally measured spectrum. The total number of Qdots bound inside the PSi was then calculated and compared to the total number of Qdots estimated to be in the pores based on the change of the measured fluorescence intensity of the reporter solution before and after exposure to PSi, as discussed in section 4.

4. Results and discussion

To prepare the surface of the PSi films for attachment of the biotin-dsDNA-Qdot or biotin-dsDNA-FAM reporters, the films were first thermally oxidized and then APTES, glutaraldehyde, and streptavidin were sequentially attached to the oxidized PSi (OPSi) surface. Reflectance measurements and ATR-FTIR measurements were taken after each surface modification step, as shown in Figs. 2(a) and (b). Additionally, after the streptavidin surface modification step, the PSi was soaked in DI water for 24 h and then remeasured to ensure that the surface modification was stable before proceeding to reporter functionalization.

The reflectance spectra of the PSi thin films are characterized by Fabry-Perot fringes (see, for example, Fig. 1(a)). The spacing of these fringes is determined by the optical thickness of the PSi film. For large changes in the optical thickness of the PSi film that cause large fringe shifts (e.g., due to the attachment of a large number of molecules in the pores), it can be challenging to analyze the exact spectral change due to the possibility of the fringes shifting by more than the fringe spacing. Accordingly, for the initial surface modification steps, we analyzed the optical thickness changes by using the reflective interferometric Fourier transform spectroscopy technique [12]. In this approach, taking the FFT of the reflectance spectrum gives a peak whose position corresponds to twice the optical thickness of the PSi film (i.e., 2nL where n is the effective refractive index of the PSi and L is the physical thickness of the PSi film). As shown in Fig. 2(a), the 2nL peak position increases after each surface modification step, suggesting that APTES, glutaraldehyde, and streptavidin are attached to the OPSi. Negligible shift in the 2nL peak position was measured after the DI water soak following streptavidin attachment, suggesting that the PSi surface modification is stable.

ATR-FTIR measurements confirm that the measured optical thickness changes are indeed caused by the different molecular attachments. N-H bending peaks around 1580 cm⁻¹ and 1650 cm⁻¹, shown in Fig. 2(b), indicate the presence of amine in APTES. An increase in C-H stretching peaks at 2863 cm⁻¹ and 2966 cm⁻¹ and appearance of a carbonyl peak at 1715 cm⁻¹ confirm the presence of glutaraldehyde on the surface. Finally, the attachment of streptavidin is shown by O-H, carbonyl, and amide bands peaks at 3300 cm⁻¹, 1721 cm⁻¹, and 1470 cm⁻¹ and 1600 cm⁻¹, respectively.

After incubating the streptavidin-modified PSi in the 135 nM biotin-dsDNA-Qdots reporter solution, the supernatant was collected for fluorescence measurements and the reflectance of the PSi was measured after the film was rinsed and dried. As shown in Fig. 3(a), the Fabry-Perot fringes redshift by approximately 3.1 nm after exposure to the reporters, suggesting the biotin-dsDNA-Qdots were attached to the streptavidin-modified PSi. Note that due to the excess biotinylated DNA in the reporter solution, some of the streptavidin molecules on the PSi surface may capture biotinylated DNA instead of biotin-dsDNA-Qdots. Fluorescence measurements comparing an equal volume of the supernatant with 135 nM biotin-dsDNA-Qdots solution that was not exposed to PSi are shown in Fig. 3(b). A clear decrease in fluorescence intensity for the supernatant confirms that the biotin-dsDNA-Qdots were attached to the streptavidin-modified PSi. To ensure the reporters were not non-specifically trapped in the pores, a streptavidin-modified PSi film was exposed to 135 nM streptavidin-coated Qdots. As shown in Figs. 3(c) and (d), there was negligible change in reflectance after exposure to the streptavidin-coated Qdots and the fluorescence intensity of the supernatant of the streptavidin-coated Qdots exposed to the PSi was the same as that of an equal volume of 135 nM streptavidin-coated Qdots solution not exposed to PSi. Hence, we conclude that there is negligible non-specific reporter binding occurring in the PSi film.

 

Fig. 3. (a) Reflectance spectra before and after attachment of 135 nM biotin-dsDNA-Qdot reporters to streptavidin-modified PSi. (b) Fluorescence intensity of 135 nM biotin-dsDNA-Qdots reporter solution compared to equal volume of biotin-dsDNA-Qdots supernatant collected after exposure of the reporters to streptavidin-modified PSi. The lower fluorescence intensity of the supernatant is consistent with reporters being attached to the streptavidin-modified PSi during incubation. (c) Reflectance spectra before and after attachment of control solution of 135 nM streptavidin-coated Qdots to streptavidin-modified PSi. No change in reflectance suggests that there is no non-specific binding. (d) Fluorescence intensity of 135 nM streptavidin-Qdots solution compared to equal volume of streptavidin-Qdots supernatant collected after exposure to streptavidin-modified PSi. No change in fluorescence intensity is consistent with negligible non-specific binding of the Qdots in the PSi. (e) Reflectance spectra before and after attachment of 135 nM biotin-dsDNA-FAM reporters to streptavidin-modified PSi. (b) Fluorescence intensity of 135 nM biotin-dsDNA-FAM reporter solution compared to equal volume of biotin-dsDNA-FAM supernatant collected after exposure of the reporters to streptavidin-modified PSi. The lower fluorescence intensity of the supernatant is consistent with reporters being attached to the PSi during incubation.

Download Full Size | PDF

Figure 3(e) shows the reflectance measurements of a streptavidin-modified PSi film before and after attachment of 135 nM biotin-dsDNA-FAM reporter solution: a 1.7 nm redshift results from attachment of these reporters. Fluorescence measurements of the supernatant of the biotin-dsDNA-FAM reporter solution after incubation on the PSi film showed decreased intensity compared to an equal volume of solution of the biotin-dsDNA-FAM reporters that was not exposed to PSi (Fig. 3(f)). This decrease in fluorescence intensity is consistent with reporter attachment to the streptavidin-modified PSi. Comparing the Fabry-Perot fringe shifts in Figs. 3(a) and (e), the redshift for reporter attachment is nearly 2 times larger for the biotin-dsDNA-Qdots than the biotin-dsDNA-FAM, suggesting that a larger change in the optical thickness occurs when the biotin-dsDNA-Qdots are attached. This result indicates that the higher refractive index and larger size of the Qdots compared to FAM make them a superior high contrast label for the reporters. This conclusion is confirmed in the DNase cleavage detection experiment, as discussed below.

As a proof of principle demonstration of HCCD, DNase was used to cleave the dsDNA and release the reporters from the pores. First, the experiment was carried out with the PSi sample functionalized with biotin-dsDNA-Qdot reporters. Figure 4(a) shows a nearly 2.2 nm blueshift of the Fabry-Perot fringes that results after the Qdots and some portion of the DNA were removed from the PSi surface and rinsed out of the pores. The fringes do not shift fully back to their position prior to reporter attachment as some portion of the biotinylated DNA remains attached to the streptavidin-modified PSi surface. Moreover, the DNase enzyme likely did not cleave every DNA tether and thus some reporters may remain bound to the pore walls. The simulations discussed below and shown in Fig. 4(b) provide insights on the quantity of reporters attached and subsequently cleaved from the PSi sample. Importantly, in the control experiment shown in Fig. 4(c), in which a biotin-dsDNA-Qdot-functionalized PSi film was soaked in DI water for 2 h – a duration equivalent to the DNase exposure – there was only a minor change to the reflectance spectrum (∼ 0.4 nm blueshift), which may be due to the fact that DI water and not specifically DNase-free water was utilized for the soak. Thus, we believe that the reporter-functionalized PSi surface is stable and DNase leads to removal of material from the pores. When PSi functionalized with biotin-dsDNA-FAM reporters was exposed to DNase solution, a nearly 1.2 nm blueshift of the Fabry-Perot fringes was measured, as shown in Fig. 4(d). The magnitude of the blueshift was much less for the PSi with biotin-dsDNA-FAM reporters after exposure to DNase compared to that for the PSi with biotin-dsDNA-Qdot reporters, confirming that Qdots are the preferable label compared to FAM dye for HCCD sensing.

 

Fig. 4. (a) Reflectance spectra showing redshift of Fabry-Perot fringes after biotin-dsDNA-Qdot reporter attachment to streptavidin-modified PSi and blueshift of fringes after exposure to DNase and removal of DNA and Qdots from the PSi. (b) Simulated reflectance spectra showing redshift of Fabry-Perot fringes after 0.6 picomoles of biotin-dsDNA-Qdot reporters are attached to streptavidin-modified PSi and blueshift of fringes after removal of approximately 67% of DNA-Qdots from the PSi. (c) Reflectance spectra for control experiment demonstrating minimal change in fringe position when reporter-functionalized PSi is exposed to DI water instead of DNase. (d) Reflectance spectra showing redshift of Fabry-Perot fringes after biotin-dsDNA-FAM reporter attachment to streptavidin-modified PSi and blueshift of fringes after exposure to DNase and removal of DNA and FAM from the PSi.

Download Full Size | PDF

To quantitatively relate the observed spectral shifts in Fig. 4(a) to the amount of bound and removed reporters, we carried out analytical simulations to model the Fabry-Perot fringe shifts due to biotin-dsDNA-Qdots attachment and DNase cleavage, using a combination of the analytical transfer matrix method and the Bruggeman effective index approximation, as detailed in section 3.6. Figure 4(b) shows that the simulated reflectance spectra are in good agreement with the measured data in Fig. 4(a) when a biotin-dsDNA-Qdot reporter fill fraction of $v$ = 1.31% is assumed. By considering the footprint of the PSi sample (0.25 cm²) and thickness of the PSi film (3.99 µm), as well as the volume of the reporters, which includes the volume of the streptavidin-coated Qdots and biotinylated DNA, we then estimated the total number of Qdots bound in the pores to be 0.6 picomoles.

The estimated number of Qdots bound in the pores as determined by the reflectance measurements and simulations was then correlated with the fluorescence data in Fig. 3(b). Using a calibration curve produced with stock Qdot solution, we determined the decrease in fluorescence intensity of the Qdot reporter solution after exposure to the PSi sensor during functionalization corresponds to 0.85 picomoles of Qdots in the collected supernatant. Since the reporter solution initially contained 3.24 picomoles of Qdots, it can be inferred that the maximum number of Qdots attached in the PSi is 2.34 picomoles. The actual number of Qdots attached in the pores is likely less because the PSi was thoroughly rinsed with DI water after collection of the supernatant and any unbound Qdot reporters in the pores would have been rinsed out of the pores but not collected for the fluorescence measurements. Accordingly, we consider the simulations and fluorescence data to be in good qualitative agreement.

There are two key differences between HCCD and a traditional DNA sensing assay: the direction and magnitude of the spectral change for a detection event. In a traditional sensing assay, DNA hybridization between DNA probe and target molecules causes a spectral redshift. In contrast, for HCCD, using activated enzymes to cleave DNA molecules that tether the Qdots to the surface leads to removal of part of the DNA tether and the Qdot from the surface, causing a spectral blueshift. This spectral blueshift – the magnitude of which can be controlled by the choice of the reporter size and composition – is advantageous for handling biofouling, a major concern for sensors operating in complex media. Since a HCCD detection event leads to a spectral blueshift, non-specific attachment of proteins or other low-index biomolecule species that result in a biofouling-induced spectral redshift would not cause a false positive sensor response. As long as the sensor surface is stable, as was shown in this work, using a spectral blueshift to signify a detection event has a significant advantage for sensing in saliva, serum, and blood. Further, by taking advantage of the dual amplification achieved from single enzymes cleaving large numbers of reporters and the high refractive index of the nanoparticles or Qdots used to label the cleavage reporters, it is possible to overcome the sensitivity limitations of traditional biosensing systems.

Since the magnitude of the cleavage-induced spectral change can be increased by the judicious choice of the high-contrast reporters, HCCD leads to a larger signal-to-noise ratio than the standard sensing modality based on detecting low-index biological material added to the system in the process of DNA hybridization. In prior PSi sensor studies, micromolar level target DNA concentrations were required to cause a similar magnitude fringe shift as the DNase cleavage of the 135 nM dsDNA-Qdots reported in this work (Fig. 4(a)) [24,25]. The experiments reported here lay an important foundation for HCCD that could be performed with CRISPR-Cas12 or CRISPR-Cas13 selective cleavage [26,27] of nucleic acids using on-chip optical biosensors or other cleavage assays such as toehold-switch [28] and DNA-zyme [29], resulting in a highly specific nucleic acid test. The peak shift observed after 2 hours of DNase cleavage at a 12.5 nM concentration is 20 times the resolution of the spectrometer used for the reflectance measurements, showing that there is a path to highly sensitive and fast, real-time readout of assays that result in a cleavage event. Additional optimization may be achieved by decreasing the proportion of ds-DNA linkers to Qdots in order to increase the average refractive index change per occupied streptavidin binding site on the PSi surface.

5. Conclusion

This work demonstrates the practical potential of high contrast cleavage detection as a new sensing modality. PSi interferometers were functionalized with biotin-dsDNA-Qdot reporters that served as the high contrast species whose cleavage from the PSi led to a large change in the reflectance spectrum. The reporter cleavage was triggered by exposure to 12.5 nM DNase, which served as a model in the reported experiments for activated CRISPR-Cas12 complexes that can enable highly specific detection of pathogens. The magnitude of the measured reflectance blueshift after DNA cleavage was much larger than the magnitude of the reflectance redshift that would have resulted from hybridization of a nM concentration of DNA in traditional PSi sensing assays. Moreover, we showed that larger and higher refractive index inorganic quantum dots serve as more effective reporter labels than smaller and lower refractive index organic FAM dye. HCCD has the potential to enable new applications for optical biosensors by providing a sensing modality that is robust to noise from surface fouling or complex patient sample media, generates large signal amplification via the use of reporters with high-contrast nanoparticles and multiplicative enzymatic activity, and takes advantage of the high target specificity that can be provided by enzymatic assays.