# Fileset

[Iwanaga_ACSNano19(2025)38841.pdf](https://mdr.nims.go.jp/filesets/706638d5-9f20-47a5-a45b-7b7bcb33a978/download)

## Creator

[Masanobu Iwanaga](https://orcid.org/0000-0002-8930-6940)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Single-MicroRNA Detection on High-Selectivity Metasurface Fluorescence Biosensors](https://mdr.nims.go.jp/datasets/d6a8435c-4ba3-4e8a-bcf5-05ad18983232)

## Fulltext

Single-MicroRNA Detection on High-Selectivity Metasurface Fluorescence BiosensorsSingle-MicroRNA Detection on High-Selectivity Metasurface FluorescenceBiosensorsMasanobu Iwanaga*Cite This: ACS Nano 2025, 19, 38841−38848 Read OnlineACCESS Metrics & More Article Recommendations *sı Supporting InformationABSTRACT: Next-generation diagnostics is expected to use theabundant data on living bodies and provide sufficiently usefulhealthcare information. A significant portion of the data areconsidered to be collected from microRNAs (miRNAs), which playcrucial roles in various activities inside the body. Here, wedemonstrate single-miRNA detection using metasurface fluorescence(FL) biosensors, which are optimized all-dielectric nanostructuredsurfaces featuring excellent FL detection capability. Ultimate high-sensitivity discrimination of one miRNA from zero miRNA isachieved at the subattomolar level by employing optimized reversetranscription (RT) of miRNAs, polymerase chain reaction (PCR)suppressing false reactions, and highly efficient and target-selectiveFL detection of the miRNA amplicons on the metasurface biosensorsusing appropriately designed oligo DNA probes. This degree of precision has never been obtained using any other technique,such as digital PCR, which is currently one of the most efficient techniques. Furthermore, we demonstrate the specificdetection of a cancer-correlated miRNA that is deeply mixed with another miRNA. We also examine and discuss othermethods that possibly work for miRNA detection at femtomolar or lower concentrations, such as chromatography anddifferent amplification methods, including handy one-step RT-PCR.KEYWORDS: metasurface, biosensor, microRNA, single-molecule detection, selective biosensingMicroRNAs (miRNAs), comprising approximately 20bases, are currently crucial biosensing targets becauseof their involvement in diverse activities in livingbodies, indicating disease-related signatures, even in the earlystages. Substantial volumes of the information on miRNAs havebeen accumulated in databases, one of which is an open Website.1,2 Many miRNAs are most likely correlated with cancers,and a part of them is considered to serve as markers at the earlystage diagnostics3−7 and noninvasive examinations,8 which arebeing pursued extensively and have not yet been established asmedical examinations. For example, hsa-miR-15a-5p and hsa-miR-143-3p were suggested to be associated with severaldiseases such as hepatocellular carcinoma, colorectal cancer, andpancreatic cancer.1Typical procedures to obtain miRNA involve several steps, asillustrated in Figure 1A. From the sampling of a biopsy to thefinal collection of the miRNA, substantial effort is required.Various trials using magnetic beads, porous materials, and so onare commercially underway to improve these processes. Thesepretreatments for collecting miRNAs are out of the scope of thisstudy.Several methods were pursued to detect collected miRNAswith a high precision. RT of miRNAs to complementary DNAs(cDNAs) and PCR of the cDNA are currently the most commontechnique. In conventional quantitative PCR (qPCR), fluo-rescence (FL) probes are added to amplicons, and FL signals aredetected as the amplification progresses. In the early stage of RT-PCR trials, it was reported that ten types of miRNAs were tested,one of them showed a limit of detection (LOD) of 0.2femtomolar (fM), six showed an LOD of 2 fM, and three showedan LOD of 20 fM;9 thus, the typical LOD was approximately 2fM. Although trials to attain lower LODs in miRNA detectionare rare, RT-PCR for long RNA of hundreds of bases wasextensively tested during the pandemic due to COVID-19. TheReceived: September 15, 2025Revised: October 20, 2025Accepted: October 21, 2025Published: October 28, 2025Articlewww.acsnano.org© 2025 The Author. Published byAmerican Chemical Society38841https://doi.org/10.1021/acsnano.5c15853ACS Nano 2025, 19, 38841−38848This article is licensed under CC-BY 4.0Downloaded via NATL INST FOR MATLS SCIENCE (NIMS) on November 11, 2025 at 23:09:52 (UTC).See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.https://pubs.acs.org/action/doSearch?field1=Contrib&text1="Masanobu+Iwanaga"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/showCitFormats?doi=10.1021/acsnano.5c15853&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?goto=articleMetrics&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?goto=recommendations&?ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?goto=supporting-info&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=tgr1&ref=pdfhttps://pubs.acs.org/toc/ancac3/19/44?ref=pdfhttps://pubs.acs.org/toc/ancac3/19/44?ref=pdfhttps://pubs.acs.org/toc/ancac3/19/44?ref=pdfhttps://pubs.acs.org/toc/ancac3/19/44?ref=pdfwww.acsnano.org?ref=pdfhttps://pubs.acs.org?ref=pdfhttps://pubs.acs.org?ref=pdfhttps://doi.org/10.1021/acsnano.5c15853?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://www.acsnano.org?ref=pdfhttps://www.acsnano.org?ref=pdfhttps://acsopenscience.org/researchers/open-access/https://creativecommons.org/licenses/by/4.0/https://creativecommons.org/licenses/by/4.0/https://creativecommons.org/licenses/by/4.0/RT-qPCR for the virus RNA was reported to have an LOD of30−50 copies/test.10 An improved PCR technique is digitalPCR (dPCR), which uses fractionation plates and implementselaborate statistical analysis to obtain an improved LOD.11−13Further trials to improve the LOD are conducted using dropletsfor PCR, called droplet dPCR (ddPCR). The best performancewas claimed to be 5 copies/test.14Apart from the PCR techniques, another approach for miRNAdetection was based on loop-mediated isothermal amplification(LAMP);15 as the best performance, detection of 6 copies/testwas claimed. As is widely known, LAMP is more elaborate thanPCR,16,17 requiring four types of primers, whereas PCR usesonly two types. Although attempts have been made to attainhigher sensitivity for DNA using LAMP,18−20 single DNAdetection has not been succeeded so far. As a simpler and moreimproved procedure compared with the previous study,15 theRT-PCR for miRNAs to use two types of primers (primers 1 and2) was adopted, as shown in Figure 1B.To pursue extreme high-precision capability enabling single-miRNA detection, metasurface FL biosensors21,22 are acandidate because they are successfully detected in single cell-free DNA (cfDNA),23 which is a short fragment of the full-length gene. Figure 1D shows a photograph of a metasurface FLbiosensor chip, which is placed in a holder and comprises a self-absorbed pair of a metasurface substrate and a transparentmicrofluidic (MF) chip made of polydimethylsiloxane (PDMS);six areas (blue in the dotted-line box) aligned in the verticaldirection are metasurfaces. Manipulation of liquid flows in theMF channels and FL measurement on the metasurfacebiosensors are automated.23 Figure 1E illustrates a standardmolecular configuration to detect FL signals on the metasurfaceFL biosensors. Initially, binding molecules of cysteine-streptavidin (Cys-SA) are immobilized; subsequently, biotiny-lated amplicons are effectively captured via biotin−streptavidinbinding; finally, LED-light excitation induces the FL signals. Themetasurface consists of a periodic silicon-nanocolumn array ofFigure 1. Schematics of microRNA (miRNA) collection, amplification of miRNA, hybridization with probes, and metasurface fluorescence (FL)biosensors. (A) A typical way of sampling of miRNA. (B) Two-step reverse transcription (RT) polymerase chain reaction (PCR) to amplify thetarget miRNA(s) and yield amplicons. (C) Hybridization of biotin- and FL-probes with the amplicons. (D) Photo of a six-channel metasurfaceFL biosensor chip with lateral dimension of 45 mm × 45 mm, which is set in a holder. (E) Illustration of FL detection on the metasurfacebiosensor, comprising a periodic array of silicon nanocolumns. After immobilization of binding molecules, the target amplicons with biotinlabels are captured and exhibit enhanced FL emission.ACS Nano www.acsnano.org Articlehttps://doi.org/10.1021/acsnano.5c15853ACS Nano 2025, 19, 38841−3884838842https://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig1&ref=pdfwww.acsnano.org?ref=pdfhttps://doi.org/10.1021/acsnano.5c15853?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-as200 nm height on a silicon dioxide layer. The metasurfacebiosensors exhibit outstanding FL-intensity enhancement24among numerous trials for metasurface nanofabrication,25functioning as efficient FL biosensors.23,26−30 Further detailsof the FL detection, the metasurface structures, and thefabrication process are described in the Experimental Section.In this study, we aimed at demonstrating single-miRNAdetection employing metasurface FL biosensors. For conductingexplicit quantitative experiments, we used synthesized single-strand miRNAs as detection targets. As representative miRNAsamong numerous miRNAs known to date,1,2 hsa-miR-15a-5pand hsa-miR-143-3p are here set to be the targets. AlthoughmiRNAs exist in cells and biopsy samples in reality, theconcentrations are undetermined, which prevents us from theirquantitative evaluation in practice. After extensive explorationsof protocols in two-step RT-PCR and FL detection on themetasurface biosensors, we achieved single-miRNA detectionand substantiated robust detection under mixed miRNAconditions.RESULTS AND DISCUSSIONSingle miRNA Detection. Figure 2A shows a representativeset of miRNA-detection FL images of the metasurfacebiosensors, acquired by a noncooling CCD camera; the coloredareas correspond to the metasurfaces; clearly, areas outside themetasurfaces are dark, suggesting the FL background due tononspecific absorptions is very low. The target miRNA was hsa-miR-15a-5p, the sequence of which is listed in the ExperimentalSection. The target concentrations were in an attomolar (aM)range from 500 attomolar (aM) to 0 molar (M).In Figure 2B, we show the quantitative net FL intensities(orange bars), which were evaluated by subtracting the FLintensities shown in Figure 2A from the background intensitiesmeasured immediately after the binding-molecule immobiliza-tion. The miRNA concentration of 0.5 aM corresponds to 1miRNA/test in the experiment. The miRNAs of 0 M representnegative control. Evidently, the FL intensity at 1 miRNA/test isdistinct from those at 0 miRNA/test.In Figure 2C, we present experimental data (orange dots witherror bars) in an extremely low-concentration range from 5000to 0.5 aM, which corresponds to a range from 11,800 to 1.18copies/test. Practically, 1.18 copies/test is equivalent to a singlemiRNA test. In this measurement, we implemented thedetection of miRNAs at extremely low concentrations withconducting 45-cycle PCR after the RT reaction. The detailedexperimental conditions are provided in the ExperimentalSection. The measured data plotted on a log−log scale werefitted using a Hill curve (dashed black curve), defined in eq 1;the profile was described using the Hill equation because itquantifies acceptor−analyte coupling products under equili-brium conditions;31,32 indeed, the capture reaction of biotin-labeled amplicons on the metasurface biosensors in microfluidicchannels under low flow rate occurs under an equilibriumcondition:y y S yxx K( )nn n0 0D= ++ (1)where y is the FL intensity; y0 is the FL intensity at zeroconcentration, representing with negative control; x is theconcentration of miRNA; S is the saturation value; KD is thedissociation factor; and n is an index representing cooperative/anticooperative reaction. By fitting the experimental data inFigure 2C, a set of parameters was determined such that y0 =52.7, S = 12,689, KD = 11.7, and n = 0.778. We note that thevalue of y0 was experimentally determined using the averaged FLintensities at 0 M, which was a negative control; the value wasless than 0.1% in the detection range of the CCD camera. A shorthorizontal bar indicates the 3σ line (σ: standard deviation) fromthe negative control. Furthermore, the inset presents the dataand Hill curve on a semilinear scale at 0−100 aM, including thedata at 0 M. Obviously, the FL intensity of 1 copy/test is abovethe 3σ line, which statistically guarantees that the single-miRNAFigure 2. A representative result of miRNA detection on the metasurface FL biosensors. (A) FL images of target miRNA, hsa-miR-15a-5p,concentrations from 500 attomolar (aM) to 0 molar (M), displayed from left to right. (B) Net FL intensity evaluated from the FL images in (A).(C) Measured FL intensity versus target miRNA concentration in aM (bottom) or copies/test (top), plotted on a log−log scale. The fitted Hillcurve defined in eq 1 is shown with a dashed curve. The horizontal bar indicates the 3σ line from the zero-concentration data (σ: standarddeviation). Inset provides a magnified view at low concentrations of 0−100 aM on a semilinear scale, including the data point at 0 M.ACS Nano www.acsnano.org Articlehttps://doi.org/10.1021/acsnano.5c15853ACS Nano 2025, 19, 38841−3884838843https://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig2&ref=pdfwww.acsnano.org?ref=pdfhttps://doi.org/10.1021/acsnano.5c15853?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-assignal is discriminated from the zero-miRNA signal. The value ofn < 1 implies that the binding reaction between the immobilizedCys-SA and the miRNA amplicons with the biotin probes wasanticooperative, which was often observed in configurations touse low-concentration analytes.23,27,29,33 Owing to the largePCR cycles, the dissociation factor KD, which gives the centerconcentration of the S-shape Hill curve on the linear scale of they axis, was reduced to approximately 12 aM. We remark that theHill equation is equivalent to the 4-parameter equation,26 whichis frequently used to analyze concentration-dependent biosens-ing data.In Figure 3, comprehensive detection results of miRNA, hsa-miR-143-3p, are shown. Figure 3A shows the miRNA detectionin a wide range of concentrations from 10 picomolar (pM) to 50aM; the concentrations are represented on a semilog scale. Tocover this wide range more than six orders of concentrations,different PCR cycles were conducted; open black squares, closedgreen triangles, and closed red circles correspond to 35, 40, and45 PCR cycles, respectively. Dashed curves are fitted to Hillcurves (eq 1) for each measured set. The parameters KD were480.3, 59.5, and 6.7 fM for the PCR cycles of 35, 40, and 45,respectively, indicating that the detection range of concen-trations can be changed by varying the cycles. Although the 5-cycle increase, in principle, results in 32-fold amplification, theKD values mean less than 10-fold amplification. Thus, the PCRprocess deviated from the ideal amplification due to thecombination of the miRNA, primers, and reagent kit.We remark that the FL-intensity range in Figure 3A, which isat most 4 orders of magnitude, is determined primarily not bythe detection capability of metasurface biosensors but by the setof reagents and primers in the RT-PCR. This is understood asfollows: the metasurface biosensors comprise a square array of300 nm periodicity, thereby having 1.11 × 107 Si nanocolumns/mm2; each Si nanocolumn functions as an FL-enhancing opticalresonator;24 consequently, when FL-labeled analytes are ideallyimmobilized on all the nanocolumns, more than 107 sites emitFL; thus, the intrinsic detection range (or dynamic range) of themetasurface biosensors is, in principle, more than 7 orders ofmagnitude, which is considered to be limited in Figure 3A by theimmobilization efficiency and actual range of analyte concen-trations.Figure 3B,C shows the detection profile of miRNA of hsa-miR143-3p on the log−log and semilinear scales, respectively,measured after implementing 50 PCR cycles; orange dots witherror bars indicate the measured FL intensities, a dashed blackcurve fits the FL intensities using eq 1, and a horizontal bardenotes the 3σ line from the FL-intensity level at theconcentration of 0 M. The FL signal at 1 copy/test (i.e., 0.5Figure 3. (A) A series of miRNA, hsa-miR-143-3p, detections in a wide range from 10,000 fM to 50 aM on a semilog scale. Data shown with blacksquare, green triangles, and red circles were measured after going through 35, 40, and 45 PCR cycles, respectively. For clarity, the FL intensitiesat the highest concentration in the three series are set to be equal. All the dashed curves are Hill curves, defined in eq 1. (B,C) Data plotsregarding single-miRNA detection, shown on log−log and semilinear scales, respectively. Measured data in a range from 500 to 0.5 aM arepresented with orange dots with error bars, shown together with the fitted Hill curve (dashed black), which is defined in eq 1. The horizontal barindicates the 3σ line from the zero-concentration data.ACS Nano www.acsnano.org Articlehttps://doi.org/10.1021/acsnano.5c15853ACS Nano 2025, 19, 38841−3884838844https://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig3&ref=pdfwww.acsnano.org?ref=pdfhttps://doi.org/10.1021/acsnano.5c15853?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-asaM) is definitely discriminated from the FL signal at 0 copy/test,thus substantiating the single-miRNA detection. In fitting theexperimental data of Figure 3B using eq 1, a set of parameterswere determined such that y0 = 386.9, S = 12,200, KD = 2556.2,and n = 0.465. We note that the value of y0 was experimentallydetermined similarly to that in Figure 2C. The factor KDremained at a large value of 2556.2 aM even under 50 cyclesof PCR amplification, implying that the RT-PCR is less efficientfor hsa-miR-143-3p than for hsa-miR-15a-5p, despite thesimilarity of their primer designs. Generally, the efficiency ofRT-PCR depends on the target miRNAs and primers.Selective Detection of miRNAs. Figure 4 presents a seriesof specific detection results of hsa-miR-15a-5p miRNA as a 3Dbar graph. In this experiment, the target miRNA was mixed withanother miRNA, hsa-miR-143-3p. We started solutionscontaining the target and counter miRNAs and used primersonly for the target in the two-step RT-PCR. The targetconcentrations were varied from 50 fM to 50 aM, whereas theconcentrations of the mixed miRNA were set to 5 fM, 500 aM, or50 aM. The PCR cycles were set to 45 cycles. Clearly, the threeseries of target miRNAs were detected regardless of the mixed-miRNA concentrations. In particular, the target miRNAs weredetected even when the concentration of the mixed miRNA at 5fM was 100-fold higher than that of the target miRNA at 50 aM.These results declare that the present scheme for miRNAdetection, which combines two-step RT-PCR with the metasur-face FL biosensors, is a robust method.Discussion on Other Detection Techniques. One-StepRT-PCR. We performed one- and two-step RT-PCRs usingcommon reagents, such as polymerases, as described in theExperimental Section. As a result, we found that the one-stepRT-PCR was inferior to the two-step RT-PCR because the one-step process frequently yielded unstable reactions and some-times false-positive reactions. The background FL levelsometimes rose and could not be completely suppressed, evenafter adjustment of the one-step conditions. Therefore, weadopted two-step RT-PCR in this study.One-step real-time RT-PCR is a conventionally available,handy technique. We examined this, as described in the results inthe Supporting Information (Section S1). Summing up theresults, it turned out that the one-step real-time RT-PCR was anunreliable technique to detect the present miRNAs atconcentrations of 1 pM or less because of the poor or false-positive reactions irrespective of the miRNA concentrations(Figure S1).In terms of short-time and point-of-care PCR, infrared-light-heating metasurfaces were recently reported.34 The rapidheating at 16.6 K/s and cooling at 7.7 K/s may make PCRcycling more feasible in a compact setup with low-powerconsumption.LAMP. One of the lowest concentrations detected wasreported using RT-LAMP.15 We followed the primer designsand LAMP procedures to detect miRNAs in this study.Eventually, we could not reproduce the main claim of 6copies/test detection;15 only 1686 copies/test (or 2 fM) andhigher concentrations were detectable. Our typical RT-LAMPresults are shown in the Supporting Information (Figure S2)together with the experimental details. As shown in Figure 1B,we substantially simplified the elaborate RT-LAMP that requiresthe four types of primers, leading to single-miRNA detection inthis study.Chromatography for RT-PCR Products. After RT-PCR,chromatography is a possible technique to detect the amplicons.We tested it using an instrument that uses MF chips to obtainimproved chromatography results (Labchip GX Touch 24,Revvity, Waltham, MA, USA) and found that the signalscorresponding to the amplicons appeared, irrespective of thetarget miRNA concentrations. The experimental results areshown in the Supporting Information (Figures S3 and S4)together with detailed descriptions. These results most likelycome from the fact that the instrument cannot discriminate thegenuine amplicons from unintended amplified DNAs; indeed,the RT-PCR products tend to include a smear. Thechromatography is designed to analyze the multicomponentproducts based only on the mass. In contrast, the metasurface FLbiosensors incorporate biotin- and FL-probes for the RT-PCRproducts, as shown in Figure 1C, and ensure highly selectivedetection of the genuine amplicons, as shown in Figures 2−4.Other Reported Techniques. At the end of the discussion, werefer to two techniques for miRNA detection reported so far,which took different approaches from those described above.They detected miRNA without using nucleic-acid amplificationtechniques.One technique used gold triangular nanoparticles,35 whichhave local surface plasmon resonances (LSPRs) and show aresonant wavelength shift in accordance with the amount ofcaptured miRNAs. The resonant shift is the principle of LSPR-based detection. The LOD was claimed to be 1 fM. However,the miRNA concentrations were varied by six orders (or from 1nanomolar (nM) to 1 fM), while the change of resonance shiftwas at most 2%; therefore, the dynamic range of the detectionsignals was very narrow, and the detected signals at differentconcentrations overlapped largely to each other. Therefore, theresonance-shift technique is a nonquantitative method, which isa definite drawback. Another study using gold nanospheres36was stimulated by a similar motive to that of the previous study35and concluded that the LOD was 1 pM. These reports suggestthat the resonance-shift techniques are unlikely for to be efficientat low concentrations below 1 fM.Figure 4. 3D-bar presentation of specific detection of the targetmiRNA, hsa-miR-15a-5p, which was mixed with another miRNA,hsa-miR-143-3p. The probe set was used only for the target.Selective detection of the target miRNA was obtained, even whenthe target miRNA was 100-fold lower in the concentration than themixed miRNA.ACS Nano www.acsnano.org Articlehttps://doi.org/10.1021/acsnano.5c15853ACS Nano 2025, 19, 38841−3884838845https://pubs.acs.org/doi/suppl/10.1021/acsnano.5c15853/suppl_file/nn5c15853_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsnano.5c15853/suppl_file/nn5c15853_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsnano.5c15853/suppl_file/nn5c15853_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsnano.5c15853/suppl_file/nn5c15853_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?fig=fig4&ref=pdfwww.acsnano.org?ref=pdfhttps://doi.org/10.1021/acsnano.5c15853?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-asThe other technique was based on chemical luminescence(CL).37 Target miRNAs were first incubated together withbiotin-labeled counter DNA probes and anti-DNA/RNAantibodies that were immobilized in advance on magneticbeads; after washing the DNA probes that were not hybridizedwith the miRNA, streptavidin-labeled luciferase was added asthe CL source and incubated; finally, the luciferase unboundedto the DNA probes was rinsed; then, the CL was measured usinga commercial instrument.37 A range of miRNA concentrationsfrom 1 nM to 10 fM was evaluated, and the LOD wasdetermined to be 6.3 fM, suggesting that this technique issuitable for miRNAs at 10 fM and higher concentrations.CONCLUSIONSWe have demonstrated single-miRNA detection, which wasobtained by going through the two-step RT-PCR suppressingfalse-positive reactions and the highly selective, efficient FLdetection on the metasurface biosensors. In particular, the singlemiRNA was discriminated from zero, which is an ultimatesensitivity that has never been attained. Furthermore, robustdetection under mixed miRNA conditions was tested andsubstantiated. Several other techniques to detect miRNAs werealso discussed together with our experimental examinations.Thus, we reached an ultimate goal in biosensing technology forpromising target miRNAs, making the best use of themetasurface FL biosensors with both high sensitivity andselectivity.EXPERIMENTAL SECTIONmiRNA and Probes. Table 1 lists the sequences of the miRNAs,primers, and probes used in this study, which were synthesized andpurified through high-performance liquid chromatography (EurofinGenomics, Tokyo, Japan). The roles of the primers and probes areillustrated in Figure 1B,C. The primers for miRNAs generally needsequences of more than 60 bases because the miRNAs areapproximately 20 bases. In contrast, ordinary PCR for DNAs, whichare mostly 100 bases or longer, allows the primers to be short (∼20bases). Considering these points, we adapted a primer design from thestudy on LAMP for miRNA,15 where four types of primers wereprepared to conduct LAMP for one type of miRNA. The primers wereconceived to reduce false reactions by introducing the loop structure, asseen in Figure 1B. Our results following this LAMP technique aredescribed in the Discussion section.Primer 1 in Table 1 contributed to the RT reaction, designed tohybridize with 10 bases of the target miRNA and produce cDNA. In thedesign, accidental matching ratio to other miRNAs was suppresseddown to 9.77 × 10−7. Primer 2 worked in the PCR, hybridizing with 11bases of the elongated primer 1. Both primers were thus involved in theamplification procedure, as illustrated in Figure 1B.RT-PCR. The RT reaction in this study was conducted at 37 °C for10 min using M-MLV reverse transcriptase (28025013, Thermo FisherScientific, Waltham, MA, USA). The PCR reaction was carried outusing a KAPA2G Fast PCR kit (KK5500, Roche, Basel, Switzerland);the protocol was implemented using a thermal cycler, such as hot startat 95 °C for 3 min, thermal cycles of (95 °C for 10 s → 45/50 °C for 15/5 s → 72 °C for 15 s) ×N cycles, and final elongation at 72 °C for 1 min,in which the annealing temperature was set to 45 and 50 °C for hsa-miR-15a-5p and hsa-miR143-3p, respectively, and the cycle number Nwas set in a range from 35 to 50 in each experiment. In the two-step RT-PCR, the RT reaction was conducted first; next, primer 2 and PCRreagents were added to the RT-reaction solution; finally, PCR wasconducted, as noted above.For the two-step RT-PCR, reaction solutions were prepared asfollows. The RT-reaction solution of total 10 μL per test contained 2 μLof 20 pmol primer 1, 0.25 μL of 50-unit M-MLV reverse transcriptase, 2μL of buffer for M-MLV, 1 μL of each 10 nmol dNTP mixture, 4 μL oftarget miRNA diluted using nuclease-free distillated water (314-09291,Nippon Gene, Tokyo, Japan) with 1 unit/μL RNase inhibitor (0317L,New England Biolabs, Ipswich, MA, USA), and 0.75 μL of RNase-freewater with the RNase inhibitor for adjustment of the total amount. ThePCR reaction solutions contained the 10 μL RT reaction solution and15 μL of solution consisting of 2 μL of 20 pmol primer 2, 0.6 μL of 3-unit PCR-polymerase KAPA2G, 5 μL of buffer for the KAPA2G, 1 μL ofdNTP mixture associated with the KAPA2G, and 6.4 μL of adjustingnuclease-free water.After these reactions, the amplicons were hybridized with biotin- andFL-probes, as shown in Figure 1C. In both cases of hsa-miR-15a-5p andhsa-miR-143-3p, the hybridization condition was set to be 95 °C for 3min → 45 °C for 30 min. As shown in Figure 1C, the probes weredesigned to couple with the elongated sequences in the RT-PCR,suppressing an increase in unnecessary background FL.For the one-step RT-PCR, all the primers and reagents were mixedfirst, being adjusted to a total of 25 μL comprising the 20 μL mixtureand 5 μL target miRNA; then, they went through the thermal processwithout pausing. After this thermal cycling, the probe hybridization wasimplemented similarly to that in the two-step case.Metasurface FL Biosensors. The all-dielectric metasurfacebiosensors in this study were fabricated using silicon-on-insulator(SOI) wafers composed of the top SOI layer, middle buried-oxide SiO2layer of 375 nm thickness, and bottom Si wafer of 675 μm thickness.The top SOI layer was selectively fabricated through electron-beamlithography for on-top negative resist (NEB-22A, Sumitomo Chemical,Tokyo, Japan) and selective deep reactive ion etching (RIE) only forthe SOI layer. The metasurfaces were designed to have 300 nmperiodicity and 220 nm diameter of nanocolumns (Figure 1E), beingalmost faithfully realized, as reported previously.26,39 As was examinedpreviously,24 the metasurfaces have prominent, almost optimalcapability for FL-intensity enhancement at wavelengths of 560−600nm, where the FL-probe TAMRA (Table 1) emits FL. The large FL-enhancing capability reaching 1000-fold in comparison with a referenceflat Si substrate was attained with optimizing the total process fromphotoexcitation to FL emission.24 The analyses and consideration forthe whole photoexcited dynamics on metasurfaces were detailedpreviously.21,24,40,41Table 1. Sequences of miRNA, Primers, and Probes, WhichAre Displayed from 5′-End (Left) to 3′-End (Right)cspecies sequencehsa-miR-15a-5p UAGCAGCACAUAAUGGUUUGUGprimer 1a GCTGACGACTCCTTTTGTTGTCTGG-AAGTGTGACGCGATTTAGGACTCGT-CAGCTTTTTCACAAACCATTprimer 2b TAGCAGCACTGACTTTGTAATAGG-ACTGTCCGCCGCACTTTGTCAGTG-CTGCTATTTTTAGCAGCACATbiotin-probe 1 [Bio]AAATCTGGAAGTGTGACGCGATbiotin-probe 2 [Bio]AAATGTAATAGGACTGTCCGCCFL-probe 1 CAGCTTTTTCACAAACCAAAT[TAM]FL-probe 2 CTGCTATTTTTAGCAGCATTA[TAM]hsa-miR-143-3p UGAGAUGAAGCACUGUAGCUCprimer 1a GCTGACGACTCCTTTTGTTGTCTGG-AAGTGTGACGCGATTTAGGACTCGT-CAGCTTTTTGAGCTACAGTprimer 2b CACTGACTTTGTAATAGGACTGTCC-GCCGCACTTTGTCAGTGCTGCTATT-TTTTGAGATGAGATGAAGCbiotin-probe 1 [Bio]AAATCTGGAAGTGTGACGCGATbiotin-probe 2 [Bio]AAATGTAATAGGACTGTCCGCCFL-probe 1 GGACTCGTCAGCTTTTTGTTT[TAM]FL-probe 2 GTGCTGCTATTTTTTGATAT[TAM]aPrimer for RT reaction. bPrimer for PCR. cSymbols [Bio] and[TAM] denote biotin and FL-molecule TAMRA,38 respectively.ACS Nano www.acsnano.org Articlehttps://doi.org/10.1021/acsnano.5c15853ACS Nano 2025, 19, 38841−3884838846www.acsnano.org?ref=pdfhttps://doi.org/10.1021/acsnano.5c15853?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-asSlight nanofabrication deviation from the design hardly affected themeasured FL signals on the metasurface biosensors; indeed, we usedthe metasurfaces of diameters of 210−220 nm in this study; thediameters varied from substrate to substrate, and a typical variation on asubstrate was 2−3 nm, being suppressed fairly well. Thus, the FL datacoming from the same sample were quite uniform on a substrate. Theheight of Si nanocolumns was determined by the height of the SOI layer(200 ± 5 nm) because the deep RIE etched only the SOI layer withoutthe resist coat. Importantly in practice, the metasurface substrates werereused frequently after washing in piranha solution, which did notinduce any detectable damage on the Si nanocolumns forming themetasurfaces.The MF chips were made of transparent PDMS, which have sixchannels in accordance with the number of metasurface areas (Figure1D), which were typically 2.1 × 0.7 mm in lateral size. The height ofeach MF channel was set to 30 μm, and the total thickness of the MFchips was 2 mm. The MF chips were commercially produced accordingto our design. Using the MF chips, small volumes (e.g., 50 μL) of liquidsamples can be manipulated at low flow rates (e.g., 10 μL/min), whichincreases FL-measurement reproducibility. Furthermore, each MFchannel is isolated, thereby reducing contamination that could happenin handling the amplified products.The binding molecules, Cys-SA, were flowed at 20 μg/mL, whichwere diluted using phosphate-buffered saline of pH 7.4, andimmobilized onto the metasurfaces (Figure 1E). Owing to the highlyefficient biotin−streptavidin coupling, the miRNA amplicons werecaptured selectively and efficiently on the metasurfaces. Theimmobilization and capture performance of Cys-SA was explicitlyconfirmed in the reference experiment in the Supporting Information(Section S4).ASSOCIATED CONTENT*sı Supporting InformationThe Supporting Information is available free of charge athttps://pubs.acs.org/doi/10.1021/acsnano.5c15853.Experimental results and details of the one-step real-timeRT-PCR, LAMP, chromatography, and reference experi-ment on Cys-SA (PDF)AUTHOR INFORMATIONCorresponding AuthorMasanobu Iwanaga − Research Center for Electronic andOptical Materials, National Institute for Materials Science(NIMS)RINGGOLD, Tsukuba 305-0044, Japan; orcid.org/0000-0002-8930-6940; Email: iwanaga.masanobu@nims.go.jpComplete contact information is available at:https://pubs.acs.org/10.1021/acsnano.5c15853NotesThe author declares no competing financial interest.ACKNOWLEDGMENTSThe author thanks Takashi Hironaka for data acquisition. Thisstudy was partially supported by the NIMS Priority ResearchProject “Biomaterials” and by JSPS KAKENHI numberJP24K01389. Nanofabrication and characterization of themetasurfaces were conducted at the Advanced ResearchInfrastructure for Materials and Nanotechnology (ARIM) ofthe Ministry of Education, Culture, Sports, Science andT e c h n o l o g y ( M E X T ) , J a p a n , P r o p o s a l N u m b e rJPMXP1225NM5134. A part of illustrations in Figure 1A wasadapted from Mind the Graph under CC BY-SA 4.0 license.REFERENCES(1) miRBase miRNA database (miRBase), Available at https://www.mirbase.org (accessed September 4, 2025).(2) Kozomara, A.; Birgaoanu, M.; Griffiths-Jones, S. miRBase: frommicroRNA sequences to function. Nucleic Acids Res. 2019, 47, D155−D162.(3) Li, G.; Luo, J.; Xiao, Q.; Liang, C.; Ding, P. Predicting microRNA-disease associations using label propagation based on linearneighborhood similarity. J. Biomed. Inf. 2018, 82, 169−177.(4) Sohel, M. M. H. Circulating microRNAs as biomarkers in cancerdiagnosis. Life Sci. 2020, 248, 117473.(5) Takizawa, S.; Matsuzaki, J.; Ochiya, T. Circulating microRNAs:Challenges with their use as liquid biopsy biomarkers. CancerBiomarkers 2022, 35, 1−9.(6) Metcalf, G. A. D. MicroRNAs: circulating biomarkers for the earlydetection of imperceptible cancers via biosensor and machine-learningadvances. Oncogene 2024, 43, 2135−2142.(7) Naranbat, D.; Herdes, E.; Tapinos, N.; Tripathi, A. Review ofmicroRNA detection workflows from liquid biopsy for diseasediagnostics. Expert Rev. Mol. Med. 2025, 27, No. e11.(8) Sembler-Møller, M. L.; Belstrøm, D.; Locht, H.; Pedersen, A. M. L.Distinct microRNA expression profiles in saliva and salivary gland tissuedifferentiate patients with primary Sjögren’s syndrome from non-Sjögren’s sicca patients. J. Oral Pathol. Med. 2020, 49, 1044−1052.(9) Sharbati-Tehrani, S.; Kutz-Lohroff, B.; Bergbauer, R.; Scholven, J.;Einspanier, R. miR-Q: a novel quantitative RT-PCR approach for theexpression profiling of small RNA molecules such as miRNAs in acomplex sample. BMC Mol. Biol. 2008, 9, 34.(10) Bruce, E. A.; et al. Direct RT-qPCR detection of SARS-CoV-2RNA from patient nasopharyngeal swabs without an RNA extractionstep. PLoS Biol. 2020, 18, No. e3000896.(11) Sykes, P. J.; Neoh, S. H.; Brisco, M. J.; Hughes, E.; Condon, J.;Morley, A. A. Quantitation of targets for PCR by use of limitingdilution. BioTechniques 1992, 13, 444−449.(12) Kalinina, O.; Lebedeva, I.; Brown, J.; Silver, J. Nanoliter scalePCR with TaqMan detection. Nucleic Acids Res. 1997, 25, 1999−2004.(13) Vogelstein, B.; Kinzler, K. W. Digital PCR. Proc. Natl. Acad. Sci.U.S.A. 1999, 96, 9236−9241.(14) Yin, H.; Wu, Z.; Shi, N.; Qi, Y.; Jian, X.; Zhou, L.; Tong, Y.;Cheng, Z.; Zhao, J.; Mao, H. Ultrafast multiplexed detection of SARS-CoV-2 RNA using a rapid droplet digital PCR system. Biosens.Bioelectron. 2021, 188, 113282.(15) Gao, K.; Zhang, P.; Wang, H.; Wang, H.; Su, F.; Li, Z.Ultrasensitive homogeneous detection of microRNAs in a single cellwith specifically designed exponential amplification. Chem. Commun.2021, 57, 5570−5573.(16) Notomi, T.; Okayama, H.; Masubuchi, H.; Yonekawa, T.;Watanabe, K.; Amino, N.; Hase, T. Loop-mediated isothermalamplification of DNA. Nucleic Acids Res. 2000, 28, No. e63.(17) Tomita, N.; Mori, Y.; Kanda, H.; Notomi, T. Loop-mediatedisothermal amplification (LAMP) of gene sequences and simple visualdetection of products. Nat. Protoc. 2008, 3, 877−882.(18) Hardinge, P.; Murray, J. A. H. Reduced False Positives andImproved Reporting of Loop-Mediated Isothermal Amplification usingQuenched Fluorescent Primers. Sci. Rep. 2019, 9, 7400.(19) Hardinge, P.; Murray, J. A. H. Full Dynamic RangeQuantification using Loop-mediated Amplification (LAMP) byCombining Analysis of Amplification Timing and Variance betweenReplicates at Low Copy Number. Sci. Rep. 2020, 10, 916.(20) Hu, F.; Li, J.; Zhang, Z.; Li, M.; Zhao, S.; Li, Z.; Peng, N.Smartphone-Based Droplet Digital LAMP Device with Rapid NucleicAcid Isolation for Highly Sensitive Point-of-Care Detection. Anal.Chem. 2020, 92, 2258−2265.(21) Iwanaga, M.; Hu, Q.; Tang, Y. Metasurface biosensors: Statusand prospects. Appl. Phys. Rev. 2025, 12, 021305.(22) Iwanaga, M. Productive biosensing techniques empowered by all-dielectric metasurfaces. Front. Bioeng. Biotechnol. 2025, 12, 1484638.ACS Nano www.acsnano.org Articlehttps://doi.org/10.1021/acsnano.5c15853ACS Nano 2025, 19, 38841−3884838847https://pubs.acs.org/doi/suppl/10.1021/acsnano.5c15853/suppl_file/nn5c15853_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?goto=supporting-infohttps://pubs.acs.org/doi/suppl/10.1021/acsnano.5c15853/suppl_file/nn5c15853_si_001.pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Masanobu+Iwanaga"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://orcid.org/0000-0002-8930-6940https://orcid.org/0000-0002-8930-6940mailto:iwanaga.masanobu@nims.go.jpmailto:iwanaga.masanobu@nims.go.jphttps://pubs.acs.org/doi/10.1021/acsnano.5c15853?ref=pdfhttps://www.mirbase.orghttps://www.mirbase.orghttps://doi.org/10.1093/nar/gky1141https://doi.org/10.1093/nar/gky1141https://doi.org/10.1016/j.jbi.2018.05.005https://doi.org/10.1016/j.jbi.2018.05.005https://doi.org/10.1016/j.jbi.2018.05.005https://doi.org/10.1016/j.lfs.2020.117473https://doi.org/10.1016/j.lfs.2020.117473https://doi.org/10.3233/CBM-210223https://doi.org/10.3233/CBM-210223https://doi.org/10.1038/s41388-024-03076-3https://doi.org/10.1038/s41388-024-03076-3https://doi.org/10.1038/s41388-024-03076-3https://doi.org/10.1017/erm.2025.2https://doi.org/10.1017/erm.2025.2https://doi.org/10.1017/erm.2025.2https://doi.org/10.1111/jop.13099https://doi.org/10.1111/jop.13099https://doi.org/10.1111/jop.13099https://doi.org/10.1186/1471-2199-9-34https://doi.org/10.1186/1471-2199-9-34https://doi.org/10.1186/1471-2199-9-34https://doi.org/10.1371/journal.pbio.3000896https://doi.org/10.1371/journal.pbio.3000896https://doi.org/10.1371/journal.pbio.3000896https://doi.org/10.1093/nar/25.10.1999https://doi.org/10.1093/nar/25.10.1999https://doi.org/10.1073/pnas.96.16.9236https://doi.org/10.1016/j.bios.2021.113282https://doi.org/10.1016/j.bios.2021.113282https://doi.org/10.1039/D1CC01326Bhttps://doi.org/10.1039/D1CC01326Bhttps://doi.org/10.1093/nar/28.12.e63https://doi.org/10.1093/nar/28.12.e63https://doi.org/10.1038/nprot.2008.57https://doi.org/10.1038/nprot.2008.57https://doi.org/10.1038/nprot.2008.57https://doi.org/10.1038/s41598-019-43817-zhttps://doi.org/10.1038/s41598-019-43817-zhttps://doi.org/10.1038/s41598-019-43817-zhttps://doi.org/10.1038/s41598-020-57473-1https://doi.org/10.1038/s41598-020-57473-1https://doi.org/10.1038/s41598-020-57473-1https://doi.org/10.1038/s41598-020-57473-1https://doi.org/10.1021/acs.analchem.9b04967?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.analchem.9b04967?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1063/5.0253333https://doi.org/10.1063/5.0253333https://doi.org/10.3389/fbioe.2024.1484638https://doi.org/10.3389/fbioe.2024.1484638www.acsnano.org?ref=pdfhttps://doi.org/10.1021/acsnano.5c15853?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-as(23) Iwanaga, M.; Hironaka, T.; Ikeda, N.; Sugasawa, T.; Takekoshi,K. Metasurface Biosensors Enabling Single-Molecule Sensing of Cell-Free DNA. Nano Lett. 2023, 23, 5755−5761.(24) Iwanaga, M. All-Dielectric Metasurfaces with High-Fluores-cence-Enhancing Capability. Appl. Sci. 2018, 8, 1328.(25) Yang, Y.; et al. Nanofabrication for Nanophotonics. ACS Nano2025, 19, 12491−12605.(26) Iwanaga, M. All-Dielectric Metasurface Fluorescence Biosensorsfor High-Sensitivity Antibody/Antigen Detection. ACS Nano 2020, 14,17458−17467.(27) Iwanaga, M. High-Sensitivity High-Throughput Detection ofNucleic-Acid Targets on Metasurface Fluorescence Biosensors.Biosensors 2021, 11, 33.(28) Iwanaga, M.; Tangkawsakul, W. Two-Way Detection of COVID-19 Spike Protein and Antibody Using All-Dielectric MetasurfaceFluorescence Sensors. Biosensors 2022, 12, 981.(29) Iwanaga, M. Rapid Detection of Attomolar SARS-CoV-2 NucleicAcids in All-Dielectric Metasurface Biosensors. Biosensors 2022, 12,987.(30) Iwanaga, M. Robust Detection of Cancer Markers in HumanSerums Using All-Dielectric Metasurface Biosensors. Biosensors 2023,13, 377.(31) Hill, A. V.; Hill, A. L.; Paganini-Hill, A. The possible effects of theaggregation of the molecules of hæmoglobin on its dissociation curves.J. Physiol. 1910, 40, 4−7.(32) Gesztelyi, R.; Zsuga, J.; Kemeny-Beke, A.; Varga, B.; Juhasz, B.;Tosaki, A. The Hill equation and the origin of quantitativepharmacology. Arch. Hist. Exact Sci. 2012, 66, 427−438.(33) Iwanaga, M. Highly sensitive wide-range target fluorescencebiosensors of high-emittance metasurfaces. Biosens. Bioelectron. 2021,190, 113423.(34) Kim, I.; Kim, H.; Go, M.; Lee, S.; Nguyen, D. D.; Kim, S.;Shrestha, K.; Alsaadi, A.; Jeon, Y.; Jeong, S.; Cho, G.; Kim, J. K.; Rho, J.;Lee, L. P. Ultrafast Metaphotonic PCR Chip with Near-PerfectAbsorber. Adv. Mater. 2024, 36, 2311931.(35) Joshi, G. K.; Deitz-McElyea, S.; Liyanage, T.; Lawrence, K.; Mali,S.; Sardar, R.; Korc, M. Label-Free Nanoplasmonic-Based ShortNoncoding RNA Sensing at Attomolar Concentrations Allows forQuantitative and Highly Specific Assay of MicroRNA-10b in BiologicalFluids and Circulating Exosomes. ACS Nano 2015, 9, 11075−11089.(36) Miti, A.; Thamm, S.; Muller, P.; Csaki, A.; Fritzsche, W.;Zuccheri, G. A miRNA biosensor based on localized surface plasmonresonance enhanced by surface-bound hybridization chain reaction.Biosens. Bioelectron. 2020, 167, 112465.(37) Nagatake, Y.; Sato, M.; Mouri, Y.; Tomita, N. Fully automatedmicroRNA quantification technique based on bioluminescent enzymeimmunoassay. Anal. Biochem. 2022, 656, 114880.(38) Sigma-Aldrich. 5-Carboxy-tetramethylrhodamine, 2025, Websiteat https://www.sigmaaldrich.com/US/en/product/mm/851026 (ac-cessed September 2, 2025).(39) Hu, Q.; Iwanaga, M.; Tang, Y. Metasurface PlatformIncorporating Aggregation Induced Emission Based Biosensor forEnhanced Human Serum Albumin Detection. Adv. Opt. Mater. 2024,12, 2400868.(40) Choi, B.; Iwanaga, M.; Miyazaki, H. T.; Sugimoto, Y.; Ohtake, A.;Sakoda, K. Overcoming metal-induced fluorescence quenching onplasmo-photonic metasurfaces coated by a self-assembled monolayer.Chem. Commun. 2015, 51, 11470−11473.(41) Iwanaga, M. Plasmonic Resonators: Fundamentals, Advances, andApplications; Pan Stanford Publishing: Singapore, 2016.DOI: 10.1201/9781315364711.ACS Nano www.acsnano.org Articlehttps://doi.org/10.1021/acsnano.5c15853ACS Nano 2025, 19, 38841−3884838848https://doi.org/10.1021/acs.nanolett.3c01527?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.nanolett.3c01527?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.3390/app8081328https://doi.org/10.3390/app8081328https://doi.org/10.1021/acsnano.4c10964?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.0c07722?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.0c07722?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.3390/bios11020033https://doi.org/10.3390/bios11020033https://doi.org/10.3390/bios12110981https://doi.org/10.3390/bios12110981https://doi.org/10.3390/bios12110981https://doi.org/10.3390/bios12110987https://doi.org/10.3390/bios12110987https://doi.org/10.3390/bios13030377https://doi.org/10.3390/bios13030377https://doi.org/10.1007/s00407-012-0098-5https://doi.org/10.1007/s00407-012-0098-5https://doi.org/10.1016/j.bios.2021.113423https://doi.org/10.1016/j.bios.2021.113423https://doi.org/10.1002/adma.202311931https://doi.org/10.1002/adma.202311931https://doi.org/10.1021/acsnano.5b04527?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.5b04527?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.5b04527?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.5b04527?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1016/j.bios.2020.112465https://doi.org/10.1016/j.bios.2020.112465https://doi.org/10.1016/j.ab.2022.114880https://doi.org/10.1016/j.ab.2022.114880https://doi.org/10.1016/j.ab.2022.114880https://www.sigmaaldrich.com/US/en/product/mm/851026https://doi.org/10.1002/adom.202400868https://doi.org/10.1002/adom.202400868https://doi.org/10.1002/adom.202400868https://doi.org/10.1039/C5CC04426Jhttps://doi.org/10.1039/C5CC04426Jhttps://doi.org/10.1201/9781315364711?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-aswww.acsnano.org?ref=pdfhttps://doi.org/10.1021/acsnano.5c15853?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://www.cas.org/solutions/biofinder-discovery-platform?utm_campaign=GLO_ACD_STH_BDP_AWS&utm_medium=DSP_CAS_PAD&utm_source=Publication_ACSPubs