DNA-assisted CRISPR-Cas12a enhanced fluorescent assay for protein detection in complicated matrices

Abstract

Proteins are important biological macromolecules that perform a wide variety of functions in the cell and human body, and can serve as important biomarkers for early diagnosis and prognosis of human diseases as well as monitoring the effectiveness of disease treatment. Hence, sensitive and accurate detection of proteins in human biospecimens is imperative. However, at present, there is no ideal method available for the detection of proteins in clinical samples, many of which are present at ultra-low (less than 1 pM) concentrations and in complicated matrices. Herein, we report an ultra-sensitive and selective DNA-assisted CRISPR-Cas12a enhanced fluorescent assay (DACEA) for protein detection with detection limits reaching as low as attomolar concentrations. The high assay sensitivity was accomplished through the combined DNA barcode amplification (by using dual-functionalized AuNPs) and CRISPR analysis, while the high selectivity and high resistance to the matrix effects of our method were accomplished via the formation of protein-antibody sandwich structure and the specific recognition of Cas12a (under the guidance of crRNA) toward the designed target ssDNA. Given its ability to accurately and sensitively detect trace amounts of proteins in complicated matrices, the DACEA protein assay platform pioneered in this work has a potential application in routine protein biomarker testing.

Graphical Abstract

1. Introduction

Proteins are biological macromolecules that consist of long chains of amino acid residues and are often described as the building blocks of life because they are found everywhere in the body, such as cells, bones, muscles, tendons, ligaments, blood, hormones, and enzymes. They perform a wide variety of functions within organisms, ranging from responding to stimuli to catalyzing metabolic reactions, and from acting as messengers to control or regulate specific physiological processes to providing structure to cells and organisms¹. Accordingly, a wide variety of physiological and pathological processes will occur if there is a change in protein structure and/or concentration. Hence, proteins are important biomarkers for the early diagnosis of diseases, monitoring treatment progress & outcomes, and assessing prognosis. For example, the current tumor marker tests used in cancer diagnosis and prognosis involve the measurement of the concentrations of several proteins in the human blood, including carcinoembryonic antigen, cancer antigen 125, alpha-fetoprotein, cancer antigen 19–9, and prostate-specific antigen^2–6. Their normal levels are typically less than 2.5 ng/mL, 35 U/mL, 40 ng/mL, 37 U/mL, and 4.0 ng/mL, respectively. Other important protein tests include c-reactive protein (CRP), procalcitonin (PCT), ptau181, and troponin, just to name a few, with their normal concentrations usually less than 3 μg/mL, 0.1 ng/mL, 2.4 pg/mL and 0.04 ng/mL, respectively. High levels of CRP, PCT, ptau181, and troponin in the blood may indicate that we may have a serious health condition that causes inflammation, or maybe a sign of a serious bacterial infection (or sepsis), Alzheimer’s disease, and heart muscle damage, respectively^7–10.

Thus far, a wide range of methods for protein detection is available, including western blotting, enzyme-linked immunosorbent assay (ELISA), mass spectrometry (MS), and various nano-biosensing techniques^11–17. However, at present, there is no ideal method available for the detection of proteins in clinical samples, many of which are present at ultra-low (less than 1 pM) concentrations and in complicated matrices. For example, western blotting is widely used in biological research for the identification of specific proteins from a complex mixture. However, western blotting is time-consuming and suffers from low sensitivity and lack of reproducibility. Mass spectrometry is another well-established technique for protein biomarker discovery and proteomics research due to its high accuracy and specificity, but MS involves complicated and time-consuming sample preparations such as protein extraction/purification and digestion. Other disadvantages of MS include low sensitivity and the use of very expensive instrumentation. Compared with western blotting and MS, ELISA is more sensitive. In fact, it is the most popular technique and gold standard for protein detection, due in large part to its high resistance to matrix effects, which is one of the major concerns in real-world sample analysis. The disadvantage of ELISA lies in that it involves time-consuming and labor-intensive procedures. Moreover, its sensitivity is still not good enough to detect a wide variety of protein biomarkers in human biospecimens. To address these issues, numerous colorimetric, fluorescence, electrochemical, and nanopore-based protein biosensing methods have been developed these years^18–25. Although the majority of them have reported significantly lower limits of detection (LODs) than ELISA, their practical application in clinical sample analysis is severely limited by their low resistance to matrix effect interference. Taken together, given the important roles proteins play in early disease diagnosis and prognosis as well as monitoring the effectiveness of disease treatment, the development of a new protein detection method, which is not only ultra-sensitive but also highly resistant to complicated matrices, is imperative.

Herein, we report an innovative DNA-assisted CRISPR-Cas12a enhanced assay (DACEA) for protein detection. The CRISPR-Cas system is an adaptive immune system in bacteria and archaea that recognizes and degrades foreign nucleic acid through the guidance of crRNA, a specific RNA sequence^26,27. One popular CRISPR-Cas12a biosensing system consists of three key components: Cas12a enzyme, crRNA, and a dye-labeled ssDNA reporter substrate²⁸. Among them, crRNA recognizes the target DNA region of interest via base pairing. Our DACEA platform can accurately quantify attomolar concentrations of proteins in complicated matrices, offering potential use in routine protein biomarker testing. In our design, the magnetic beads-based antibody-protein-antibody sandwich architecture enables fast and convenient magnetic separation of the target protein from other matrix components and provides high protein detection specificity, while the dual-functionalized gold nanoparticles (AuNPs) are used for conversion and amplification of the analyte signal. Moreover, the Cas12a-crRNA system offers further improvement in the assay sensitivity and selectivity. In the latter case, only the designed DNA sequence activates the CRISPR sensor, while trace amounts of nucleic acid contaminants, if present, would not. In the former case, once the Cas12a is activated (by formation of Cas12a-CrRNA-target DNA complex), it will cleave the dye-labeled ssDNA reporter substrate, thus enhancing the sensitivity of DNA analysis.

2. Experimental section

2.1. Materials and Reagents

CRP protein, as well as paired monoclonal anti-CRP and anti-IL-6 antibodies (capture and detection), were obtained from Medix Biochemica (St. Louis, MO, USA). CrRNA, LbaCas12a, and all the other DNA samples were purchased from Integrated DNA Technologies (Coraville, IA, USA). Magnetic beads were acquired from Invitrogen (Carlsbad, CA, USA), while the gold colloidal solution of 30-nm diameter was procured from Nanopartz (Loveland, CO, USA). All the other chemicals were obtained from Sigma-Aldrich (St. Louis, MO, USA).

2.2. Preparation of functionalized magnetic beads (MBs)

Functionalized magnetic beads (MBs) were prepared by conjugating CRP/IL-6 monoclonal capture antibody to carboxylated MBs via the well-known EDC/NHS coupling reaction, with the detailed procedure described previously²⁹. Briefly, 200 μL of MBs were washed with MES buffer (200 μL, 25 mM, pH 5.0) three times, followed by being resuspended in 200 μL of the same MES solution. Next, 100 μL of EDC (50 mg/mL) and 100 μL of NHS (50 mg/mL) were introduced into the suspension. This mixture was then incubated at room temperature with agitation for 30 minutes. After removing the supernatant, the magnetic beads were washed three times with the MES buffer (100 μL). Subsequently, the capture antibody (100 μL, 1 mg/mL) for CRP or IL-6 was added, and the mixture was incubated with rotational agitation at room temperature for 30 minutes, followed by three times washes with PBS buffer (100 μL, pH 7.4). To block non-specific binding, the beads were further treated with BSA (50 μL, 0.05%) for 10 minutes with vortexing. Finally, the antibody-functionalized beads were resuspended in 500 μL of PBS and stored at 4 °C.

2.3. Preparation of dual functionalized AuNPs

A three-step conjugation procedure was used to prepare dual-functionalized AuNPs. First, the detection antibody (Ab₂) was immobilized onto AuNPs through spontaneous adsorption. This was achieved by incubating the detection antibody (2.5 μL, 100 μg/mL), colloidal AuNPs (100 μL), and NaOH (5 μL, 2 M) at room temperature for 30 minutes. It should be noted that the spontaneous adsorption of antibodies onto citrate capped AuNPs is initially driven by the long-range electrostatic interaction³⁰, and when they come into close proximity to each other, it is reinforced by the formation of stable S–Au covalent bonds between AuNPs and the cysteine residues within the antibodies³¹. Second, 20-mer disulfide-terminated ssDNA P1 (sequence: 5’-HS-S-(CH₂)₆-TTTTTTTTTTTTTTTTTTTT-3’) was reduced using TCEP (short for tris(2-carboxyethyl)phosphine) at room temperature for 1 hour, resulting in thiol-terminated ssDNA. This thiol-terminated ssDNA was then combined with the antibody-functionalized AuNPs (Ab₂-AuNPs), and the mixture was incubated at ambient temperature for 1 hour. Subsequently, the “aging-salting” process was performed to enhance AuNP uptake. The mixture was gradually brought to a NaCl concentration of 0.3 M by adding 2 M NaCl and incubated overnight, followed by separating and removing unreacted materials via centrifugation for 10 minutes at 10,000 × g. Third, the resulting product (Ab₂-AuNP-P₁) was further modified by adding 10 μL of 1 mM complementary ssDNA P2 (sequence: 5’-AAAAAAAAAAAAAAAAAAAA-3’), resulting in Ab₂-AuNP-P₁-P₂. The final product was stored at 4°C after being centrifuged three times and resuspended in HPLC-grade water.

2.4. Characterization of Ab₂-coated AuNPs

Transmission electron microscopy (TEM) was recorded according to the procedure given by Khashaya et al.³² using JEOL JEM-1400 instrument. The experimental procedure for negative staining of Ab₂-conjugated AuNPs has been described by Munusamy and co-workers.²⁹ Briefly, the Ab₂-AuNPs sample was covered by application of uranyl acetate solution (~5 – 10 μL, 0.5%) onto the TEM grid, followed by incubation for 30 to 60 seconds. The grid was then air-dried at room temperature. Dynamic light scattering (DLS) was recorded according to the procedure given by Conrad et al.³³ using Zetasizer Nano ZS from Malvern Panalytical. Infrared spectra were acquired on an FTIR Perkin Elmer Spectrum Two: UATR Two spectrometer using 1 cm⁻¹ resolution. AuNPs and Ab₂-AuNP (10 μL each) were placed directly onto the ATR crystal and allowed to dry until a thin film remained.

2.5. Procedure for the DACEA-based protein assay

After being washed three times with 500 μL of assay buffer (pH 7.2), which contained 0.1 M NaCl, 10 mM PBS, 0.1% BSA, and 0.025% Tween 20, 50 μL of the capture antibody-functionalized MBs (Ab₁-MBs) were resuspended in 1 mL of assay buffer, and then a CRP/IL-6 standard solution at 0.1 −10 pg/mL was added. The mixture was incubated at room temperature for 1 hour on a rotary mixer. After incubation, the magnetic field was applied to remove the supernatant, and the MBs-Ab₁-protein complex was washed three times with assay buffer before being resuspended in 500 μL of assay buffer. Subsequently, dual-functionalized gold nanoparticles (Ab₂-AuNP-P₁-P₂) were added, and the mixture was incubated at room temperature for 1 hour. The produced sandwich complex (MB-Ab₁-protein-Ab₂-AuNP-P₁-P₂) was then separated from unreacted materials and washed with assay buffer for three times. Next, 100 μL of DI water was added, and the mixture was heated to 70°C for 15 minutes to release the single-stranded DNA P₂ from the sandwich complex (note that under such a condition, the P1 strand would not dissociate from the AuNP surface significantly according to Li et al.’s report³⁴). The supernatant containing P₂ was collected and used for CRISPR analysis. Briefly, we began the trans-cleavage activity experiment of Cas12a nuclease by incubating the LbCas12a with the corresponding crRNA in 1× NEB buffer 3.1 at final concentrations of 1 μM and 3 μM, respectively, for 30 minutes. Then, the collected supernatant was added to the Cas12a/crRNA complex solution and allowed to react for 30 minutes at room temperature. After that, the dye-labeled ssDNA reporter substrate was introduced to the mixture and incubated for another 3 hours, followed by fluorescence measurement with λ_ex/em = 492/528 nm at room temperature.

3. Results and discussion

3.1. Principle of the DACEA-based fluorescent assay for proteins

As shown in Scheme 1, five major steps are involved in our pioneered DACEA protein. (1) The target proteins in the clinical samples (take blood for example) are captured by capture antibody-conjugated magnetic beads (MBs) after incubation of the sample with MBs, while the spent blood, which contains other components, is discarded; 2) Gold nanoparticles (AuNPs) dual-functionalized with detection antibody and dsDNA are mixed with MBs. Accordingly, AuNPs, target proteins, and MBs will form sandwich structures, thus introducing dsDNA; 3) The sandwich product is heated at an appropriate temperature such as 70 – 90 °C to de-hybridize the duplex DNA on AuNPs. The target ssDNA will be released into the solution, while the remaining sandwich complexes will be removed after magnetic separation; and 4) the released target ssDNA will be recognized by Cas12a-crRNA, thus 5) activating Cas12a cleavage of a fluorescently labeled reporter ssDNA substrate, emitting fluorescence.

Scheme 1.

Schematic representation of the principle of the ultra-sensitive DACEA sensing platform for the detection of protein biomarkers in complicated matrices.

3.2. Detection of CRP

To demonstrate this concept, our initial experiment was carried out to detect C-reactive protein (CRP). CRP is a hepatic acute-phase protein whose levels rise significantly in response to systemic inflammation. Quantifying CRP levels in blood serves as a valuable diagnostic and prognostic marker for various inflammatory conditions, such as infections and autoimmune disorders. Elevated CRP concentrations can indicate the presence and intensity of inflammatory processes, thereby assisting clinicians in diagnosing and managing underlying pathological states^35,36. To separate CRP from matrix components, functionalized magnetic beads (MBs) were prepared by conjugating CRP monoclonal capture antibody (Ab₁) to carboxylated MBs via the well-known EDC/NHS coupling reaction^37,38. Note that the carboxylic acid groups on MBs allow covalent amide bond formation to proteins via primary amino- or sulphydryl groups. To confirm whether CRP capture antibody-conjugated MBs were successfully synthesized the UV-Vis absorbance of the CRP antibody solution was measured before and after MBs conjugation. Based on the difference in the absorbance at 280 nm (Fig. S1), we found that ~ 68% of CRP antibodies (1 mg/mL) were successfully immobilized to MBs of 2.8-μm diameter.

To introduce DNA for CRISPR-Cas12a analysis, dual-functionalized AuNPs conjugated with CRP detection antibody and dsDNA were then synthesized. In principle, the number of DNA molecules, which could be potentially conjugated to AuNPs, increases with an increase in the AuNPs surface area, thus leading to a larger DNA-to-antibody ratio and an increase in the analyte signal amplification. As a proof-of-concept, we used a three-step conjugation procedure to immobilize CRP detection antibody (Ab₂) and dsDNA to spherical citrate-capped AuNPs of 30 nm diameter. Briefly, Ab₂-functionalized AuNPs was first prepared in a basic solution via spontaneous adsorption. Although we could not determine the exact number of antibodies conjugated on the surface of AuNPs due to the very low concentration of Ab₂ immobilized, successful coupling of Ab₂ to AuNPs was supported by the UV–Vis experiment and FT-IR spectroscopy. Briefly, the surface plasmonic resonance absorption peak of the AuNP solution before and after modification showed a redshift of 4 nm (from 521 nm to 525 nm). Moreover, a reduced absorption intensity was also observed (Fig. S2). The FT-IR spectrum (Fig. 1A) of the unmodified AuNPs shows a distinct peak at 836 cm⁻¹, corresponding to the metal-oxygen stretching vibration, which confirms the coordination of citrate ions to the gold nanoparticle surface via RCOO⁻ → Au interaction³⁹. Additional peaks at 1563 cm⁻¹ and 1109 cm⁻¹, attributed to the R-CO₂ and C-O stretching vibrations, respectively, further corroborate the presence of citrate capping on the AuNPs. Shifts observed at 1599 cm⁻¹ and 1100 cm⁻¹, relative to the control, are indicative of ligand’s coordination to the nanoparticle surface⁴⁰. In the FT-IR spectrum of the Ab₂-functionalized AuNPs (Ab₂-AuNP), the presence of an Amide I band near 1696 cm⁻¹, which corresponds to the C=O stretching vibration typical of amide groups, signifies the successful conjugation of the antibody to the AuNPs⁴¹. Additionally, a peak at 1242 cm⁻¹ is observed, characteristic of C-N-C stretching vibrations within the antibody structure. DLS measurements were further carried out to determine the mean hydrodynamic diameter and to monitor antibody adsorption onto the AuNPs. The bare AuNPs exhibited a hydrodynamic diameter of approximately 68 nm, as shown in Fig. S3. Following conjugation with Ab₂, the hydrodynamic size increased to approximately 79 nm, indicating successful antibody attachment to the AuNP surface. TEM analysis was additionally employed to elucidate the size and morphology of the gold nanoparticles. The TEM images revealed that the AuNPs were well-dispersed, with an average diameter of 29.8 ± 2.1 nm (Fig. 1B). A histogram of the particle size distribution generated from the TEM images (Fig. 1C) confirmed this observation. Upon conjugation with Ab₂, the TEM data (Fig. 1D) showed a material layer surrounding the AuNPs, corresponding to the attached antibody, with an average particle size of 31.9 ± 3.1 nm (Fig. 1E). This result was further substantiated by negative staining of the Ab₂-AuNP samples, confirming the effective coupling of the antibody to the AuNPs (Fig. 1F). Then, disulfide-terminated DNA P1 was reduced (by adding TCEP) to yield thiol-terminated DNA as described in our published work⁴². The CRP detection antibody-functionalized AuNPs were then mixed with the thiol-terminated DNA to produce Ab₂-AuNP-P₁. After that, a complementary ssDNA strand P₂, a 20-mer poly(dA) (for short A₂₀), was added to yield the Ab₂-AuNP-P₁-P₂ complex. By measuring the variation in the DNA solution’s absorbance at 260 nm before and after AuNPs’ modification (Fig. S4) and determining the ratio of the concentration of the DNA immobilized on AuNPs to that of AuNPs²⁹, we found that an average of ~1,100 DNA molecules were conjugated per AuNP under our experimental condition. To confirm the successful conjugation of DNA P₁ to AuNPs, FTIR spectra were additionally collected for unmodified AuNPs, free P₁, and the Ab₂-AuNP-P₁ complex (Fig. S5). We found that the FTIR spectrum of P₁ showed a characteristic band at 1704 cm⁻¹, which corresponds to the amine groups in the nitrogenous bases⁴³. Additionally, bands at 1059 and 1026 cm⁻¹ are typically associated with ribose vibrations, while the band at 1223 cm⁻¹ is attributed to the antisymmetric stretching vibration of phosphate groups (PO₂⁻). In the spectrum of the Ab₂-AuNP-P₁ complex, the bands at 1083, 1059, and 1028 cm⁻¹, likely resulting from the combination of the DNA bands at 1059 and 1026 cm⁻¹ and a band near 1100 cm⁻¹ from Ab₂, exhibit broadening. This broadening indicates the interactions between DNA and AuNPs. Further evidence of such interactions is provided by the shift and intensity change in the Amide I band near 1696 cm⁻¹ and the band corresponding to nitrogenous DNA bases (1704 cm⁻¹). After complexation, these bands shifted to 1712 cm⁻¹ and 1687 cm⁻¹, respectively, suggesting successful binding of DNA to AuNPs. It should be noted that the functionalized AuNPs maintained the red color of the free AuNPs, indicating no aggregation occurred during the modification process. Further TEM analysis showed that, after attaching P1 to antibody-modified AuNPs, their average particle size increased by 2.8 nm (from 31.9 ± 3.1 to 34.7 ± 2.3 nm, Fig. S6).

Figure 1.

(A) FT-IR spectrum of AuNPs and Ab₂-AuNPs; (B) Representative TEM image and (C) Size distribution of the citrate-capped AuNPs; (D) Representative TEM image and (E) Size distribution of Ab₂-AuNPs; and (F) Representative TEM image of Ab₂-AuNPs after negative staining with 0.5% uranyl acetate solution.

3.3. Optimization of the experimental conditions for CRISPR analysis

To develop a highly sensitive CRISPR-Cas12a sensing platform for DNA analysis, some important detection parameters such as the effects of the molar ratio of Cas12a to crRNA (sequence: rUrArArUrUrUrCrUrArCrUrArArGrUrGrUrArGrArUrUrUrUrUrUrUrUrUrUrUrUrUrUrUrUrUrUrUrUrU) and ssDNA reporter (sequence: FAM-TTATT-IABkFQ) and the incubation time on detection of A₂₀ were then investigated. Our experiments showed that the fluorescence intensity increased significantly with an increase in the concentrations of crRNA in the enzymatic reaction buffer until the molar ratio of crRNA to Cas12a reached 3:1 (Fig. 2A). In another separate study, where the concentrations of Cas12a and crRNA were kept constant at 1 μM and 3 μM, respectively, but varied the reporter substrate concentration, we found that the fluorescence intensities of both the control (without A₂₀) and the A₂₀ standard solution increased with increasing concentration of the reporter. The maximum signal-to-background (S/B) ratio of 3.76 was achieved at a Cas12a-to-reporter ratio of 1:10 (Fig. 2B). Moreover, the incubation time study showed that the fluorescence signal of the DNA sample increased with increasing reaction duration until 3 hours, after which the fluorescence intensity of the solution did not change significantly (Fig. 2C).

Figure 2.

Optimization of the Cas12a-crRNA analysis conditions. (A) Plot of fluorescence intensity as a function of Cas12a-to-crRNA molar ratio. The concentrations of A20 and the reporter substrate used were 1 μM and 5 μM, respectively. (B) Effect of Cas12a nuclease-to-reporter molar ratio on fluorescence intensity. The concentration of A20 used was 1 μM. (C) Plot of fluorescence intensity versus reaction time. The concentrations of Cas12a, crRNA, and reporter substrate used were 1 μM, 3 μM, and 10 μM, respectively, while that of A20 was 0.5 μM.

3.4. Sensor sensitivity and selectivity

Using the 1:3:10 molar ratio of Cas12a to crRNA and ssDNA reporter and 3 hr incubation, a series of experiments with use of the DACEA sensing platform to examine various concentrations of CRP proteins (ranging from 0.1 pg/mL to 2 pg/mL) was carried out. Briefly, as outlined in Scheme 1, CRP proteins and as-prepared Ab₁-MBs and Ab₂-AuNPs-P₁-P₂ were first incubated at room temperature to form sandwich structures. After heat denaturation to release the target ssDNA P₂, Cas12a-crRNA, and ssDNA reporter substrate were added, followed by fluorescence measurement. Our experimental results (Fig. 3A) showed that if the solution did not contain CRP, we only observed a small background-level fluorescence signal. In contrast, if the solution contained a trace amount of CRP, it produced a significantly larger fluorescence intensity. Moreover, as the analyte concentration increased, the fluorescence signal of the CRP sample also increased. The dose-response curve is shown in Fig. 3B. We found that the limit of detection (LOD) of the CRP sensor was as low as 16 fg/mL (which is equivalent to ~140 aM). Such a LOD, which is determined by dividing the difference between the intercept of the least-squares regression equation and the sum of the mean of the blank signal and three times the standard deviation of blank by the slope of the calibration curve⁴⁴, is about 9 orders of magnitude lower than the CRP concentrations (10 – 100 μg/mL) in the serum of a healthy population. A similar LOD was obtained by dividing three times the standard deviation of the blank by the slope of the regression equation.⁴⁵

Figure 3.

DACEA-based detection of CRP. (A) Fluorescence spectra, showing the fluorescence intensity changed with the concentration of CRP in the solution, (B) dose-response curve, and (C) selectivity study. The concentrations of the proteins in Fig. 3c were 2 pg/mL each.

To assess the selectivity of this CRP sensor, five proteins were examined, including bovine serum albumin (BSA), human serum albumin (HSA), procalcitonin (PCT), interleukin-6 (IL-6), and interleukin-1β (IL-1β). Among them, HSA and BSA are the most abundant blood proteins in humans and bovines, respectively, while PCT, IL-6, and IL-1β are important inflammatory markers^46–51. As shown in Fig. 3C, with the exception of the target CRP, all the other protein samples only produced background-level fluorescence signals, suggesting that our sensor is highly selective for CRP.

3.5. Effect of number of DNA on AuNP surface on sensor sensitivity

It should be noted that the performance of the DACEA sensing platform is highly dependent on the dual-functionalized AuNPs. In principle, a larger DNA/antibody molar ratio on the surface of AuNPs, an enhanced sensor sensitivity will be achieved. To evaluate the impact of the number of DNA strands on the AuNPs on protein detection, a series dual-functionalized AuNPs with varying number of DNAs but constant number of antibodies was prepared and utilized for analysis of 2 pg of CRP. We found that the AuNPs functionalized with only 110 DNA strands exhibited a marginal increase in the fluorescence intensity compared to the control (Fig. 4). When the number of DNAs increased from 110 to 1100, a four-fold increase in the fluorescence signal was obtained, thus supporting the hypothesis that the sensitivity of our DACEA detection method is dependent on the number of DNA strands conjugated to the AuNPs. Therefore, optimizing the number of DNA strands attached to AuNPs is crucial for maximizing the sensitivity of protein detection. The ability to fine-tune the number of DNA strands on AuNPs provides a valuable parameter for enhancing assay performance, thereby offering a robust strategy for sensitive biomolecular detection.

Figure 4.

Effect of the number of DNAs attached on the surface of AuNPs on the sensitivity of CRP detection. The concentration of analyte CRP tested was 2 pg/mL each. In the prepared series of CRP detection antibody/DNA dual-functionalized AuNPs, the same amount of CRP antibody (2.5 μL, 100 μg/mL) was used.

3.6. IL-6 detection

To demonstrate that our developed DACEA sensing platform can be used as a generic tool for ultra-sensitive detection of proteins, IL-6 was then analyzed. IL-6, a 21-kDa glycoprotein composed of 184 amino acids, is an important multifunctional cytokine. Owing to its potent ability to induce the acute phase response and its wide variety of immune and hematopoietic activities^52,53, IL-6 plays a critical role in host defense. Elevated IL-6 levels have been reported to be associated witha variety of diseases, including psoriasis, rheumatoid arthritis, multiple myeloma, Castleman’s disease, and post-menopausal osteoporosis^54,55. In healthy populations, the blood IL-6 levels are about 5 pg/mL. However, in abnormal states (e.g., in response to inflammation), IL-6 levels in human fluids can rise to between 100 pg/mL and 1 ng/mL. By only changing the paired antibodies coupled to MBs and AuNPs, we found that our constructed DACEA fluorescent sensor can detect IL-6 at concentrations as low as ~13 fg/mL (equivalent to ~650 aM). Furthermore, the sensor was highly selective to IL-6; other proteins did not interfere with its detection (Fig. 5C).

Figure 5.

IL-6 detection. (A) Fluorescence spectra, showing the effect of IL-6 concentration on fluorescence intensity; (B) plot of fluorescence intensity as a function of IL-6 concentration; and (C) selectivity study. In the sandwich formation, IL-6 capture antibody functionalized MBs, and IL-6 detection antibody/DNA dual-functionalized AuNPs were used. In the former, ~56% of IL-6 antibodies (1 mg/mL) were immobilized to the MBs. In the latter, an average of ~1,100 DNA molecules were conjugated per AuNP. The concentrations of the proteins in Fig. 5c were 4 pg/mL each.

3.7. Simulated serum sample analysis

Three mock human serum samples were used to examine the feasibility of our developed DACEA sensor in real-world applications. Briefly, IL-6 standard solutions were spiked in human serum (from human male AB plasma, USA origin, Sigma-Aldrich, St. Louis, MO) with final concentrations obtained ranging from 0.5 to 2.0 pg/mL. These samples were analyzed by our DACEA sensor without any treatment, with the experimental results summarized in Table 1. From the table, we could see that the recoveries (ranging from 90 to 101%) of IL-6 from serum samples determined by use of our developed DACEA assay were satisfactory even when the spiked IL-6 concentration is very low (note that IL-6 was not detected in the unspiked serum control). The results suggest that the matrix components in the serum would not significantly affect IL-6 detection, and our developed IL-6 sensing platform has a potential application in real-world sample analysis.

Table 1.

Simulated human serum sample analysis by use of the DACEA sensor. Each experimental value represents the mean of three replicate analyses ± one standard deviation.

Sample Number	Theoretic Value (pg/mL)	Experimental Value (pg/mL)	Recovery (%)
1	0.5	0.49 ± 0.02	98 ± 2
2	1.0	0.90 ± 0.04	90 ± 4
3	2.0	2.02 ± 0.11	101 ± 11

4. Conclusion

In summary, a pioneering DACEA fluorescent assay was developed for the highly sensitive and selective detection of proteins in clinical samples with LODs reaching as low as ~13 fg/mL. Our developed protein detection strategy takes advantage of magnetic beads for fast magnetic separation of the target protein from other matrix components, sandwich binding of two antibodies for high specificity, DNA barcode amplification for high detection sensitivity, and Cas12a-crRNA system for further improvement of the sensor performance. Several things are worth mentioning here. First, the DACEA fluorescent assay constructed in this work can be easily expanded as a large-scale multiplexing tool for high-throughput early detection and diagnosis of human diseases such as cancer, heart and neurodegenerative diseases. For example, by using a 96-well polystyrene plate coated with 96 capture antibodies and 96 AuNPs dual-functionalized with detection antibodies and dsDNAs, simultaneous measurement of 96 proteins can be readily accomplished. Second, the purpose of this study is proof of concept, and the DACEA sensor reported in this work was not optimized. Many experimental conditions and detection parameters such as the diameter of the MBs, the loading density of the protein capture antibody on MBs, the size, shape and surface function of AuNPs, the length and sequence of the barcode DNA, the sequence of crRNA and so on can be investigated to further improve the sensor sensitivity. Third, in this work, protein detection was accomplished by converting the protein-antibody interaction event into barcode ssDNA, followed by fluorescent determination of its concentration after incubation with Cas12a-crRNA and the fluorescently labeled reporter substrate. An improved protein detection LOD and sensitivity can be visualized if the collected ssDNA is further amplified by using chemical reactions such as hybridization chain reaction^56,57 before CRISPR analysis. Alternatively, other readout formats might be used. Fourth, aptamers might be used instead of antibodies to form protein sandwich structures. Given its ability to sensitively and accurately determine trace amounts of proteins in complicated matrices, the DACEA protein sensing platform developed in this work offers the potential as a tool for routine protein biomarker testing.