Label-Free Isothermal Amplification Assay for Specific and Highly Sensitive Colorimetric miRNA Detection.

We describe a new method for the detection of miRNA in biological samples. This technology is based on the isothermal nicking enzyme amplification reaction and subsequent hybridization of the amplification product with gold nanoparticles and magnetic microparticles (barcode system) to achieve naked-eye colorimetric detection. This platform was used to detect a specific miRNA (miRNA-10b) associated with breast cancer, and attomolar sensitivity was demonstrated. The assay was validated in cell culture lysates from breast cancer cells and in serum from a mouse model of breast cancer.

Introduction

Micro RNAs (miRNAs) are small noncoding single-stranded RNA molecules, typically 21–25 nucleotides in length, that regulate gene expression.[1] The first miRNA was discovered in 1993 in Caenorhabditis elegans by Lee et al.[2] Since this discovery, miRNAs have been shown to be naturally abundant and evolutionarily conserved in both plants and animals.[3a−3c] In fact, bioinformatic studies predict that the human genome may contain up to 1000 miRNAs, of which 706 have already been identified. Accordingly, it is estimated that miRNA sequences, which are spread throughout the genome, account for 2–5% of human genes.[4] The function of miRNAs is to downregulate gene expression by binding to the 3′ untranslated regions (UTRs) of target mRNAs.[5a,5b] miRNA expression profiles exhibit unique temporal and spatial patterns that are specific for developmental stages and tissue types.[6a,6b] The expression of miRNAs is determined by both intrinsic cellular factors and diverse environmental variables.[1,7a,7b] Notably, it is estimated that miRNAs regulate 10–30% of all protein-coding genes.[8] In humans, miRNAs typically exert their effect by binding to imperfect complementary sites within the 3′UTRs of their target protein-coding mRNAs, thereby hindering translation.[9,10] Therefore, miRNAs can reduce the protein levels of their target genes without affecting the mRNA levels. On the contrary, in plants, miRNAs bind to protein-coding mRNA sequences that are exactly complementary to the miRNA, consequently inducing the RNA-mediated interference pathway, leading to cleavage of mRNA by Argonaute in the RNA-induced silencing complex.[11a,11b]

Several studies have demonstrated that miRNAs are important regulators of a variety of fundamental biological processes, such as embryonic development,[12] cell proliferation,[13] cell death,[13,14] fat metabolism,[15] hematopoiesis,[16a,16b] stress resistance,[17a,17b] neuronal development,[18] and tumorigenesis.[13,19] Furthermore, accumulating evidence demonstrates a crucial role of miRNAs in cancer and other diseases.[20a−20c] In contrast to the tightly regulated patterns of miRNA expression during development and in normal tissues, miRNAs are often misregulated under pathological conditions. For example, miRNAs that are overexpressed in cancer usually function as oncogenes, whereas miRNAs with tumor-suppressing activity are frequently downregulated.[21a,21b] Some studies provide functional evidence that overexpression of a specific miRNA, miRNA-10b, can contribute to the development of breast cancer metastasis.[22]

Traditional methods for detecting miRNAs are northern blotting, reverse-transcription polymerase chain reaction (RT-PCR), and microarrays.[23] However, all of these methods display some limitations. For instance, northern blotting is a long and complex procedure that requires radiolabeling and usually has a low detection limit, whereas RT-PCR and microarrays have good sensitivity but require costly instrumentation. Consequently, there is an urgent need to develop efficient and low-cost miRNA detection methods. The rapid development of nanotechnology has resulted in new tools for DNA[24] and miRNA detection, including nanoparticle-derived probes,[25a−25c] electrochemical methods,[26a−26c] and DNAzyme-based reporters.[27a−27e] Such newly developed methods combine an isothermal amplification step with a signal output component to achieve a high detection efficiency.[28a−28c]

Nicking enzyme amplification reaction (NEAR) is a commonly used method for isothermal amplification of miRNA.[27a,29] This technique exploits DNA polymerase and a nicking enzyme to rapidly generate and release a desired sequence, obtaining an amplification efficiency comparable to that of PCR.[30] Notably, NEAR is especially suitable for the development of portable devices for point-of-care (POC) testing, as it can proceed at a constant temperature and does not require specialized instrumentation. In this study, NEAR was combined with a gold nanoparticle (AuNP)/magnetic microparticle (MMP) system to obtain highly sensitive colorimetric miRNA detection in biological samples.

The miRNA detection process used in this study is illustrated in Figure . In the first step of the process, a DNA probe hybridizes with the target miRNA, and brief RNase digestion is carried out to reduce unspecific interference. Subsequently, polymerization is initiated by the addition of a DNA polymerase. This latter enzyme recognizes the 3′-OH end of miRNA and extends it using the adjacent sequence as a template. Polymerization introduces a nicking enzyme recognition site, which is readily cleaved by an endonuclease. Cleavage of the single strand generates a new 3′-OH group for subsequent initiation of strand extension. A short single-stranded DNA sequence is released during polymerization. This process of strand extension, cleavage, and displacement is repeated for many cycles, thereby producing a large quantity of short DNA strands (linker DNA). In the second step of the process, the linker DNA is hybridized with complementary oligonucleotides immobilized on AuNPs and MMPs. Magnetic washing is used to separate the unattached AuNPs from the MMPs. AuNPs attached to the surface of MMPs serve as a colorimetric marker for miRNA. Indeed, AuNPs are superior for colorimetric assays, as they possess extinction coefficients that are several orders of magnitude higher than those of traditional chromophores.[31]

Figure 1

miRNA detection process. (A) Schematic illustration of the steps required for colorimetric detection of miRNA. In step A, the probe binds to the target miRNA and DNA polymerase extends the strand, displacing the linker. Polymerization introduces a nicking enzyme recognition site. Subsequently, enzymatic cleavage generates another linker and 3′-OH end that can be used for further polymerization. Strand displacement amplification produces single-stranded DNA products (linker) that trigger step B. In step B, the linker is used to attach AuNPs to MMPs. Half of the linker sequence is complementary to the oligonucleotides immobilized on AuNPs, whereas the other half is complementary to the oligonucleotides immobilized on MMPs. A red-colored solution after magnetic retrieval of the MMPs indicates binding of AuNPs and the presence of target miRNA in the sample. (B) Polyacrylamide gel electrophoresis (PAGE) of strand displacement amplification components. First lane, target miRNA; second lane, probe; third lane, negative control (reaction performed without target); fourth lane, linker (reaction performed with 100 nM of target miRNA).

Therefore, the proposed assay is based on isothermal amplification and a barcode system composed of MMPs conjugated with AuNPs by hybridization of a universal linker sequence, namely, a linker with a fixed sequence, independent of the target miRNA. This allows the development of multiplexed assays, as the probe sequence is the only assay component that is target-specific.

Experimental Section

AuNP Functionalization with Thiolated DNA Oligonucleotides

AuNPs conjugated with DNA oligonucleotides were prepared as previously described.[32] Briefly, thiolated DNA oligonucleotides were digested with Tris(2-carboxyethyl)phosphine (10 mM) for 3 h at room temperature. The digested oligonucleotides were then incubated with AuNPs (2000:1 molar ratio) overnight at room temperature in an electric shaker (400 rpm). A solution of NaCl (0.3 M) in phosphate buffer (10 mM; pH 7.4) with sodium dodecyl sulfate (SDS; 0.01%) was added stepwise to the AuNP–DNA mixture over the course of 8 h. After additional overnight incubation at room temperature, the DNA-conjugated AuNPs were centrifuged (17 300g, 15 min) and washed with NaCl (0.3 M) in phosphate buffer (10 mM) with SDS (0.01%) to remove excess unbound DNA. The probes were stored at 4 °C until use. To determine the density of DNA probes on the nanoparticle surface, DNA-conjugated AuNPs (10 μL) were digested overnight with dithiolthreitol (1 mM) in phosphate buffer (90 mM, pH 8) at 40 °C. The mixture was centrifuged (17 300g, 15 min), and the amount of oligonucleotides in the supernatant was estimated using the Quant-iT OliGreen ssDNA kit (Invitrogen). The concentration of AuNPs was measured by UV–vis spectroscopy and inductively coupled plasma atomic emission spectroscopy (ICP-AES). On average, 540 copies of DNA oligonucleotides were linked to each nanoparticle. The size and zeta potential of pre- and post-functionalized AuNPs were measured using a Zetasizer Nano ZS90 (Malvern).

ICP-AES

AuNPs (10 μL) were added to aqua regia solution (500 μL), and the mixture was incubated overnight at room temperature. After digestion, the total volume was brought to 5 mL and the samples were analyzed using an Agilent Technologies 700 series ICP-AES instrument. A wide calibration curve (0–3 mg/mL) for Au3+ ions was obtained, with three absorbance peaks (λAu = 242.794, 267.594, and 208.207 nm; λAr = 737.212 nm).

Agarose Gel Electrophoresis

Gel electrophoresis was carried out on agarose gels (0.7%), using sodium boric acid buffer (pH 8.5), for 80 min at 70 V. AuNP samples (10 μL; 1.17 × 1012 particles/mL) containing glycerol (25%) were loaded into each well. A band with clearly reduced electrophoretic mobility was observed for the DNA–AuNP probes (Figure S1) compared with the AuNPs without immobilized oligonucleotides, confirming successful oligonucleotide conjugation (Figure S2A).

Preparation of MMPs Conjugated with DNA Oligonucleotides

The biotinylated DNA oligonucleotide stock was conjugated to streptavidin-coated MMPs following the manufacturer’s instructions. Briefly, the MMPs (10 mg/mL) were washed three times in RNase-free hybridization buffer (HB; 100 mM Tris–HCl pH 7.4 and 0.1% Tween 20) to remove the preservative storage buffer. Next, they were suspended in RNase-free buffer and incubated at a final concentration of 5 mg/mL with a 2 mM solution of oligonucleotides for 30 min at room temperature.

Finally, the obtained MMP–DNA oligonucleotide complex was washed three times with RNase-free buffer to remove the unbound DNA strands. To confirm successful conjugation of the oligonucleotides on the MMP surface, the MMP–DNA oligonucleotide complex was further characterized with a Zetasizer Nano ZS90 (Malvern) (Figure S2B). The binding capacity of the MMPs was estimated through DNA quantification using a Quant-iT OliGreen ssDNA kit (Invitrogen). Briefly, the immobilized DNA molecules were released from the MMPs following the manufacturer’s instructions (2 min at 90 °C in 10 mM EDTA, pH 8.2, with 95% formamide) and quantified through fluorescence measurements after MMP removal.

Transmission Electron Microscopy (TEM)

The sample solution (10 μL) was deposited onto a copper grid for 10 min, after which the excess solution was removed with a filter paper. Imaging was performed using a JEOL 1200 microscope (TEM, JEOL; Peabody) operated at an acceleration voltage of 200 kV (Figures S1A,B and S4).

Probe Preparation

The miRNA probe was prepared by mixing sequences 1 and 2 at a final ratio of 1:0.95 (stock concentration 1 μM) in a total volume of 100 μL. The mixture was heated at 95 °C for 10 min and slowly cooled to 4 °C (0.1 °C/s).

RNase Protection Assay

An RNase protection assay was performed by incubating the probe (10 μL) in a stock solution with 100 nM miRNA in 1× NEB buffer (10 mM Tris–HCl, pH 7.9; 100 mM KCl; 10 mM MgCl2; and 0.1 mg/mL bovine serum albumin) at 37 °C (15 min). RNase was inactivated through heating at 70 °C for 15 min.

NEAR

All of the amplification conditions, such as the nicking enzyme and DNA polymerase concentrations, were optimized, starting from a previously published protocol.[27a] The final miRNA probe mixture (50 μL, 200 nM probe) was incubated at 37 °C (30 min) with Klenow exopolymerase (0.25 units/μL; Bsm DNA polymerase), nicking enzyme (0.125 units/μL; Nb.Bpu 10I), and dNTP mix (200 μM) in 1× NEB buffer (10 mM Tris–HCl, pH 7.9; 100 mM KCl; 10 mM MgCl2; and 0.1 mg/mL bovine serum albumin).

Native PAGE

Polyacrylamide gels were prepared, containing a 12% acrylamide/bisacrylamide mix (19:1; Bio-Rad) in 1× Tris/borate/EDTA (TBE) (Sigma-Aldrich). The amplified samples (5 μL) were loaded onto the gels with a 1× DNA gel loading dye (Thermo Fisher Scientific), and the gels were run in 1× TBE at 90 V for 3 h. The gels were then stained with GelRed (Biotium) for 15 min, and images were acquired with a gel doc system (Bio-Rad). To confirm the size of the DNA bands, a synthetic single-strand DNA oligonucleotide marker was used.

Hybridization and Colorimetric Detection of miRNA

The isothermal amplification product (10 μL) was incubated with oligonucleotide–MMPs (4 μL, 10 min, room temperature). Subsequently, oligonucleotide–AuNPs (10 μL, 1.7 × 1012 particles/mL) were added to the solution (30 min, room temperature). The mixture was then washed two times with HB and once with 50 mM Tris–HCl, pH 7.4, and 0.1% Tween 20 buffer. Quantification was performed using a UV–vis spectrophotometer. The sequences of all probes and miRNAs used in this study are reported in Table .

Table 1

Oligonucleotide Sequences Used in This Worka

In sequence 1, the probe sequence is underlined. Single-base mutations are shown in red.

Cell Culture

MDA-MB-231 and 4T1 cells were obtained from American Type Culture Collection (ATCC; Manassas, VA). The cells were maintained at 37 °C with 5% CO2 and cultured in Dulbecco’s modified Eagle’s medium (DMEM) with high glucose (Corning Cellgro), supplemented with fetal bovine serum (10%, Atlas Biological, Fort Collins, CO) and penicillin/streptomycin solution (1%; Sigma-Aldrich). MCF10A cells were purchased from ATCC (Manassas, VA) and grown in DMEM medium supplemented with 5% of horse serum (Atlas Biological, Fort Collins, CO), 1% of penicllin/streptomycin solution, 20 ng/mL epidermal growth factor (Sigma-Aldrich), 10 μg/mL insulin (Sigma-Aldrich), and 250 μL hydrocortisone (Sigma-Aldrich).

Extraction of small RNAs (sRNAs) from the cell culture lysates was performed with the miRVANA isolation kit (Ambiogen). The concentration and purity of the extracted RNA were measured using a UV spectrophotometer. The assay was performed with 1 or 10 μL of extracted sRNA from 4T1 cell culture lysates and with 10 μL of extracted sRNA from MDA-MB-231 cell culture lysates. The sRNA was incubated with the miRNA probe and 1.25 U/μL RNase to avoid interferences from unspecific miRNA present in the sample. For qRT-PCR, 1 or 10 μL of extracted sRNA from 4T1 cell culture lysates and 10 μL of extracted sRNA from MDA-MB-231 cell culture lysates were analyzed using the Advanced TaqMan miRNA Assay (Applied Biosystems) according to the manufacturer’s instructions. PCR was performed in duplicates for 40 cycles using an ABI Prism 7900HT Sequence Detection System (Applied Biosystems). Subsequently, the quantity of miRNA-10b was determined by relative quantification using miRNA-16 as a reference.

Mouse Breast Cancer Model

Animal studies were performed following a protocol approved by the Animal Care and Use Committee at the Houston Methodist Research Institute and in adherence to the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Female athymic nude mice (6–8 weeks of age) were acquired from Charles River. MDA-MB-231 cells were suspended in 50% phosphate-buffered saline/matrigel solution and injected into the mammary fat pad of mice (3 million cells/mouse). Blood was collected by cardiac puncture 2, 3, or 4 weeks after tumor cell inoculation. The blood was kept at room temperature for 30 min, after which it was centrifuged (1500g, 10 min), and the serum was collected and stored at −80 °C until use. sRNAs were extracted from 350 μL of serum and processed as described in the Cell Culture section. For Advanced TaqMan miRNA assays, 1 μL of RNA was reverse-transcribed using the TaqMan Advanced miRNA cDNA Synthesis Kit (Thermo Fisher Scientific) following the company’s guidelines.

Limit of Detection (LoD)

The LoD was estimated in accordance with the standard International Union of Pure and Applied Chemistry definition (doi:10.1351/goldbook. L03540) as 10 times the standard deviation of the blank signal (five independent measurements).

Results and Discussion

In this study, the AuNP/MMP colorimetric method was used to detect miRNA-10b. A probe was designed, containing the following three components: (1) a sequence complementary to miRNA-10b, (2) a recognition site for a nicking enzyme, and (3) a linker sequence. PAGE was initially carried out to confirm that the designed NEAR reaction can successfully be completed within 30 min at 37 °C. Indeed, a band corresponding to the size of the linker DNA can be detected in the gel (Figure B).

The AuNPs used in the assay were synthesized as previously described.[33a−33c] The size and morphology of the AuNPs pre and post functionalization were assessed by TEM (Figure S1A,B), dynamic light scattering (Figure S1C), and UV–vis spectroscopy (Figure S1D). The nanoparticles had a uniform size distribution, with an average diameter of ∼40 nm. Successful conjugation of thiolated oligonucleotides was confirmed by an agarose gel retardation assay (Figure S2A).

After NEAR optimization (Figure S3), the detection limit of the colorimetric assay was determined using various concentrations of miR-10b (100 aM–100 fM) (Figure ). Notably, quantification of miRNA using conventional methods is usually challenging because of its small size (17–25 nucleotides) and low concentration (usually

Figure 2

Colorimetric detection of synthetic miRNA-10b at various concentrations (100 aM–100 fM). Calibration curve of UV–vis absorbance (λ = 540 nm) at different miRNA-10b concentrations. The solid line indicates a logarithmic correlation, where absorbance A = 0.0218 ln(miRNA concentration) + 0.1114 (R2 0.969). Data are presented as mean ± SD of triplicates.

Moreover, the sequence specificity of the proposed system was evaluated using miR-10b with artificially introduced mutations. It is important to be able to distinguish miRNAs that differ in a few nucleotides, as single-nucleotide polymorphisms (SNPs) are frequently occurring variations that may lead to unique traits and disease phenotypes.[34a−34c] In particular, as miRNAs can regulate hundreds of genes, SNPs can have drastic consequences on cell functions and physiological processes.[35a,35b] The existence of miRNA SNPs poses a great challenge for high-specificity detection. The results in Figure clearly suggest that this assay is capable of discriminating specific targets from mismatched strands. The signal of the complementary target was approximately 3.4- and 4.4-fold higher than that of the single-base mismatched strand and two-base mismatched strand, respectively. It is worth noting that sequence specificity was seen despite the use of unnaturally high concentrations of mutated miRNA. In addition, to confirm the specificity of our assay in biological samples, we detected synthetic miRNA-10b (100 fM) after mixing with total sRNA extracted from MCF10A cells, which lack miRNA-10b. As shown in Figure S5, the interference associated with nontarget miRNA is very low, demonstrating the high specificity of our method also under complex experimental conditions, in which similar interfering sequences are present. The observed specificity suggests that this assay could potentially be used for the analysis of miRNA SNPs.

Figure 3

Detection specificity of miRNA-10b, miRNA-10b with one mutation, and miRNA-10b with two mutations. Absorbance of miRNA-10b at a concentration of 100 nM. The blank value was subtracted from the absorbance results. Data are presented as mean ± SD of triplicates.

Furthermore, the assay was applied successfully for the detection of miRNA in biological samples, indicating its practicality for bioanalysis. Specifically, cell culture lysates from breast cancer cells and serum collected from a mouse model of breast cancer were analyzed. MDA-MB-231 human breast cancer cells and 4T1 mouse breast cancer cells were used in this study, as they have previously been reported to express high levels of miRNA-10b.[35]Figure shows that miRNA-10b can be detected in both 4T1 and MDA-MB-231 cell lines. The performance of the colorimetric assay was compared to that of RT-PCR analysis, a gold standard technique for miRNA detection, finding good agreement (Table ).

Figure 4

Colorimetric detection of miRNA-10b in cell culture lysates and serum samples. miRNA-10b detection in cell culture lysates from 4T1 and MDA-MB-231 (MDA) breast cancer cells. The volume (μL) refers to the amount of extracted RNA. The blank value was subtracted from the absorbance results. Data are presented as mean ± SD of triplicates.

Table 2

Relative Quantification of miR-10 by RT-PCR, Using miR-16 as an Endogenous Controla

	CT mean miR-l0b	CT SD	CT mean miR-16	CT SD	Log 2 fold change (mR-10b/mR-16)
MDA-MB-231	21.84	0.78	24.88	1.45	7.48
4T1	18.49	0.98	24.4	0.81	10.35
control	27.93	1.66	23.49	0.85	1
week 2	14.17	1.1	22.03	0.69	12.3
week 3	14.49	1.14	21.73	1.22	11.68
week 4	13.52	0.91	21.94	1.51	12.86

Data are presented as mean ± SD of triplicates.

Next, serum samples were collected from mice bearing orthotopic MDA-MB-231 breast cancer tumors. Specifically, blood was collected 2, 3, or 4 weeks after tumor inoculation. Detection of miRNAs in blood samples can be used for cancer diagnosis, as tumor-derived miRNAs frequently reach the circulation.[36] The results demonstrated that the signal from control serum (tumor-free mouse) was similar to that from the background buffer, whereas the levels of miRNA-10b were substantially increased in tumor-bearing mice (Figure ). In particular, miRNA-10b could be detected in the blood as early as 2 weeks after cancer cell inoculation, and the levels were relatively constant in mice assayed 3 and 4 weeks after inoculation. The levels of miRNA-16 were also measured in the samples to ensure that the presence of tumors did not cause a general increase in the serum levels of miRNAs (Table ).[39] The results obtained from the colorimetric assay reflected the relative quantification achieved with RT-qPCR (Table ). Although miRNA-10b did not allow the differentiation of tumor stages, as its serum level does not directly correlate with tumor size,[37a,37b] the colorimetric detection of its upregulation is indeed of diagnostic interest, as it enables early cancer detection, even before the tumor becomes palpable or visible. In essence, the proposed assay can be used for rapid quantitative isothermal detection of miRNA over a wide range of concentrations that are clinically relevant.[38]

Figure 5

miRNA-10b detection in serum from mice bearing orthotopic MDA-MB-231 breast cancer tumors. The x axis represents weeks after tumor inoculation. A non-tumor-bearing mouse was used as a control. The blank value was subtracted from the absorbance results. Data are presented as mean ± SD of triplicates.

This NEAR-based colorimetric assay exhibits a higher clinical sensitivity and specificity than those of previously proposed isothermal amplification-based strategies.[27a,40a,40b] Our system provides three distinct advantages. First, the readout of the assay is based on a clear color change, which can be observed by the naked eye, avoiding the use of expensive facilities and skilled labor. Second, the assay results can be obtained by simply mixing the assay components together, so no complicated operations are needed. Third, we successfully tested the colorimetric assay with biological samples (cellular extracts and mouse serum), achieving a highly sensitive, specific, and semiquantitative detection of miRNA-10b. Assays with low-cost and simple operation and readout are equally critical for managing human health, especially in resource-poor areas. Our colorimetric assay showed interesting potential in POC applications.

Conclusions

In summary, this study describes the development of a colorimetric isothermal NEAR assay for rapid and ultrasensitive detection of miRNA. The results reveal that the assay is capable of detecting aM levels of miRNA in biological samples without the use of conventional methods. The performance of this AuNP/MMP-based assay was assessed using various concentrations of miRNA-10b, and sequence sensitivity was evaluated using single-base-mismatched sequences. Moreover, the ability of the colorimetric assay to quantify miRNA levels in cell culture lysates from breast cancer cells and in serum samples from tumor-bearing mice was similar to that of RT-qPCR. Accordingly, this platform is a promising analytical tool for the detection and quantification of miRNAs in biological samples without the use of expensive and complex equipment.