Construction of a Homogeneous Enzyme-Free Autocatalytic Nucleic Acid Machinery for High-Performance Intracellular Imaging of MicroRNA

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE7 Nov 2022Construction of a Homogeneous Enzyme-Free Autocatalytic Nucleic Acid Machinery for High-Performance Intracellular Imaging of MicroRNA Jie Wei†, Jinhua Shang†, Shizhen He, Yu Ouyang, Itamar Willner and Fuan Wang Jie Wei† College of Chemistry and Molecular Sciences, Wuhan University, Wuhan 430072 , Jinhua Shang† College of Chemistry and Molecular Sciences, Wuhan University, Wuhan 430072 , Shizhen He College of Chemistry and Molecular Sciences, Wuhan University, Wuhan 430072 , Yu Ouyang Institute of Chemistry, The Hebrew University of Jerusalem, Jerusalem 91904 , Itamar Willner *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Institute of Chemistry, The Hebrew University of Jerusalem, Jerusalem 91904 and Fuan Wang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] College of Chemistry and Molecular Sciences, Wuhan University, Wuhan 430072 https://doi.org/10.31635/ccschem.021.202101545 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Enzyme-free autocatalytic nucleic acid machinery has been widely used to engineer various self-assembled nanostructures as well as high-performance bioanalysis, yet has rarely been realized in live cells for complicated design, low robustness, and limited reliability. Herein, we constructed simple yet versatile enthalpy-driven autocatalytic hybridization chain reaction (AHCR) machinery with high reliability and robustness for in situ microRNA analysis in live cells. The homogeneous AHCR machine was composed of two differently designed HCR modules, the lead-in HCR-1 amplification module, and the reverse HCR-2 feedback module. After the AHCR amplification system was delivered into live cells, target microRNA stimulated the autonomous cross-invasion of the HCR-1 module and the HCR-2 module for assembling hyperbranched dsDNA nanostructures with synergistically amplified Förster resonance energy transfer readout, thus enabling accurate intracellular microRNA imaging. The synergistic AHCR execution was systematically investigated by a series of experimental studies and computer-aided theoretical simulations. The multiple recognition capacity of HCR constituents and the successive signal amplification of the AHCR machine enabled the accurate intracellular microRNA imaging with precise signal localization inside living cells. Based on its intriguing and modular design, the AHCR machinery can be extended for analyzing diverse biomarkers, thus supplementing a powerful toolbox for clinical diagnosis and therapeutic assessment. Download figure Download PowerPoint Introduction Dynamic DNA nanotechnology has emerged as a powerful tool for programming the assembly of diverse and ingenious DNA nanostructures through the autonomously successive hybridization-based strategies,1–3 including hybridization chain reaction (HCR),4–9 entropy-driven catalysis,10–12 and catalyzed hairpin assembly.13–15 HCR represents a general principle to initiate the autonomous cross-opening or polymerization of two hairpin reactants to generate long dsDNA copolymers upon the introduction of triggering nucleic acids or small molecules.4 HCR is driven by the potential energy change between hairpin-structured reactants and the hybridization chain products, which have a low background and remarkable signal gains.16 Furthermore, HCR has been extended with improved amplification by integrating additional DNAzymes17–19 or enzymes.20–22 Upon the introduction of analyte, the coupled HCR-DNAzyme machinery is motivated to autonomously assemble dsDNA nanowires carrying numerous DNAzyme units that produce significant readout signal. However, most of the current HCR systems are still limited to the conventional linear polymerization of DNA hairpins, with significant challenges remaining for fabricating nonlinear HCR systems.23 That is, the translation of conventional HCR systems into programmed systems generating branched DNA nanostructures, particularly dendritic assemblies, via self-sustained exponential polymerization of DNA hairpins, is still a challenge. Indeed, branched and dendritic DNA nanostructures have been assembled through reprogrammed DNA self-assembly pathways; however, these processes were not self-sustainable since the different hairpin reactants were step-wise supplemented to the systems or the surrounding temperature of the system was cyclically altered during the assembly of the nanostructures.24 The self-sustained clamped HCR was developed for assembling branched DNA nanostructures via the cross-opening of hairpin-dimer reactants,25–27 yet the branched HCR is, in fact, an interconnection of HCR polymer product with limited branching degree, and the designed assembly process was still constrained within a linear polymerization process instead of the augmented exponential growth. More recently, a target-triggered exponential HCR has been designed for assembling dendritic DNA nanostructures from hybridized dsDNA substrates instead of the conventional hairpin reactants previously reported.28 Nonetheless, these dsDNA reactants required careful purification to remove unhybridized reactants, which eventually introduce perturbing spurious initiation (spontaneous hybridization) events and subsequent exponentially amplified system signal leakage. Furthermore, the morphological monitoring of these exponential HCR-assembled nanostructures has not been well supported.29 It is thus still challenging to devise an isothermal self-sustained autocatalytic HCR amplifier with high reliability and signal gain. In principle, the autocatalytic HCR machine, could detect the target at the single-molecule level since the autocatalytic amplifier autonomously and progressively generates numerous triggers in a successive round of invading executions.30 Besides the in vitro applications of DNA machines, their practical operation within living cells remains a significant challenge.16,31,32 Fabricating DNA nanodevices that perform specific functions inside living cells to execute sophisticated biological tasks, for example, in vivo bioimaging, has been a long-term goal.33–35 For example, an in-depth comprehensive understanding of the intracellular differentiation process requires a robust intracellular visualization technique for accurately localizing endogenous biomolecules, for example, microRNAs under in vivo conditions.36–39 Thus, there is an imperative requirement to develop versatile imaging tools for the visualization of minimally expressed microRNA in living cells.40 First, the short nature of microRNAs needs a significant signal amplification strategy to provide adequate signals.41 Second, the intrinsic composition of the complex cytoplasmic milieu requires multiple guided recognition events for enhanced specificity.42 Third, the tedious operations of intracellular imaging require a reliable and compact signal acquisition method for the accurate localization of microRNAs.43 It is noteworthy that most of the molecularly engineered DNA machines are designed with sophisticated DNA-origami scaffolds, large-scale nanomaterials, or complicated motion patterns.44–47 The tedious preparation and complex functionalization of the DNA origami or inorganic nanomaterial carriers are inevitable, thus limiting their potential applications inside living cells.48–50 Therefore, there is an urgent demand to explore more robust DNA-integrated nanomachines with enormous signal amplification to execute diverse in vivo biological sensing tasks in a rational and effective manner. In this study, we introduce an isothermal autocatalytic HCR (AHCR) paradigm that executes autonomous enzyme-free DNA amplification in a homogeneous condition. The AHCR machine is composed of the lead-in HCR-1 module and the feedback HCR-2 module. The HCR-1-produced DNA nanowire is composed of the tandem trigger strand for stimulating the feedback HCR-2 transducer. Meanwhile, the amplicon product of transducer HCR-2 is utilized as a tandem transmission initiator chain for lead-in HCR-1. The parallel and simultaneous execution of these two HCRs leads to the assembly of hyperbranched dsDNA nanostructures and affords an ultrasensitive fluorescence detection of analyte. With a programmable and modular design, the AHCR amplifier can be easily adapted as a universal amplification strategy for monitoring microRNAs of living cells. Furthermore, the high molecular weight of hyperbranched dsDNA copolymer products, as compared to the hairpin monomers, facilitates the amplified intracellular imaging of analyte with precise and accurate localization in living cells. The modular and scalable design of the AHCR amplifier holds great promise for clinical diagnostics and prognosis monitoring. Experimental Method Materials High-performance liquid chromatography (HPLC)-purified oligonucleotides were purchased from Sangon Biotech Co., Ltd. (Shanghai, China) and are listed in Supporting Information Tables S1 and S2. All reagents were supplied by Sigma-Aldrich (St. Louis, MO). A Millipore water purification system-purified ultrapure water was used in all experiments to avoid undesired interference. Fluorescence experiment The fluorescence experiment was carried out in 10 mM 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid sodium salt (HEPES) reaction buffer (containing 1 M NaCl, 50 mM MgCl2, pH 7.2) at 25 °C, and the spectra were collected by a Cary Eclipse spectrometer (Agilent Technologies, Santa Clara, CA). For target DNA assay, the hairpin mixture (200 nM each) of different DNA circuits was incubated with analyte in reaction buffer for 5 h. For microRNA analysis, H7 (50 nM) and AHCR hairpins (200 nM each) were reacted with various concentrations of target microRNA in the reaction buffer. Here, the fluorescence spectra were recorded at 550–700 nm with the excitation of 532 nm, and the Förster resonance energy transfer (FRET) signal was calculated by the acceptor-to-donor emission ratio (FA(670 nm)/FD(565 nm)) of the corresponding DNA machine. The AHCR-mediated intracellular microRNA imaging by using confocal laser scanning microscopy The transfection of the AHCR imaging system (50 nM H7 and 200 nM H1– H6) was carried out according to the manufacturer's protocol of lipofectamine 3000 (Invitrogen, Carlsbad, CA). These cells were incubated with the AHCR-containing transfection mixture for 5 h, washed three times with 1 mL phosphate-buffered saline (PBS), and finally added to 1 mL Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS) for confocal laser scanning microscopy (CLSM) imaging. For the control experiment, the transfection operation of the corresponding amplifier was the same as described above. For the microRNA-inhibiting experiment, the cells were incubated with 100 nM anti-microRNA transfection solution for 1 h before being transfected with the AHCR imaging mixture. Fluorescence imaging of cells was performed by using confocal microscopy (Leica, Weztlar, Germany) with the 63.0× objective lens. For the Cy3 channel, the excitation was set at 553 nm to collect the fluorescence emission from 563 to 623 nm. For the Cy5 channel, excitation was set at 645 nm to collect the fluorescence emission from 655 to 715 nm. For the FRET channel, excitation was set at 553 nm, and emission collection was from 655 to 715 nm. And the background FRET signal caused by the individual Cy3 or Cy5 was subtracted in each sample. The FA/FD channel was calculated by the ImageJ and Fiji software. Results and Discussion Principle of the AHCR system The AHCR amplification system consists of two major components: a lead-in HCR-1 module and a reverse feedback HCR-2 module. Both of the two circuitry constitutes are composed of metastable hairpin reactants ( Supporting Information Tables S1 and S2), which can only be activated and propagated by their respective initiators to generate dsDNA nanowires with a linear growth format. The hierarchical reaction scheme of the AHCR machine is shown in Figure 1A. The operation of transducer HCR-2 was guided by the initiator-assembled HCR-1 nanowire, and the as-generated HCR-2 nanowire product then conversely activated the lead-in HCR-1. The continuous cross-stimulation of the two intercommunicating HCR units led to a progressively accelerated reaction, resulting in the highly amplified readout of the AHCR system. Figure 1 | (A) Scheme illustration and the corresponding reaction graph of the isothermal AHCR machinery involving the autonomous cross-invasion of lead-in HCR-1 and transducer HCR-2 processes. (B) Detailed reaction procedure of the AHCR system. Download figure Download PowerPoint As shown in Figure 1B, the initiator I stimulated the cross-hybridization between hairpin H1 and H2 of HCR-1 to generate the dsDNA nanowires (for details, see Supporting Information Figures S1 and S2). Then the tandem T-analog transmission triggers (newly exposed domain d-e) of the I-assembled lead-in HCR-1 product autonomously activated the HCR-2 system via the successive and sequential hybridization among hairpins H3– H6. Each of the transmission triggers T produced the copolymeric dsDNA nanowire containing numerous adjacent Cy3 and Cy5 fluorophores with an amplified FRET signal (for details, see Supporting Information Figure S3). Meanwhile, the split domains a* and b* of HCR-2 reactants were simultaneously connected to generate tandem transmission initiator I1 ( I-analog with colocalized b*-a* domain), which reversibly motivated the HCR-1 machinery. Thus, the transducer HCR-2 reversibly stimulated the lead-in HCR-1 machinery that led to the generation of multiple HCR-1 nanowires from the tandem initiator sites of the HCR-2 backbone. Thus, the cross-invasion of HCR-1 and HCR-2 systems led to the mutual germination of multiple HCR nanowires on their underlying dsDNA nanowire scaffolds, resulting in the formation of hyperbranched dsDNA nanostructures with an exponentially amplified FRET signal for the AHCR machine (for details, see Supporting Information Figure S4). To avoid the undesired signal leakage, all the cross-interaction of the hairpin components of the AHCR machine should be eliminated in the absence of the analyte, yet activated to perform the cascaded transformation with the target. The secondary structures of all reactants are theoretically designed by using NUPACK software (for details, see Supporting Information Figure S5).51 The thermal stability of hairpin conformation is critical to eliminate the undesired hybridizations of HCR reactants without the corresponding trigger. Furthermore, the released region of lead-in HCR-1 (reassembled transmission triggers T) needs to be partially caged into the stem domain of the hairpin reactant in order to avoid the cross-talk between HCR-1 and HCR-2 units ( Supporting Information Figure S6). Demonstration of the AHCR system After the theoretical and experimental optimization of AHCR probes, the underlying working mechanism of the proposed AHCR machinery was verified. Here, the FRET readout process was presented by the reduction in the fluorescence emission intensity of the Cy3 fluorophore donor (λ = 565 nm) and the simultaneous promotion in the emission intensity of the Cy5 fluorophore acceptor (λ = 670 nm) upon excitation of the donor unit associated with the AHCR system at λex = 532 nm. The ratio of the emission intensities of two fluorophore labels provided a relative FRET signal readout (FA/FD or F670/F565), where the higher ratios indicate higher energy transfer efficiency. The AHCR mixture was rather stable without initiator I, as revealed by its negligible fluorescence variation (curve a, Figure 2B). In contrast, the initiator I-motivated AHCR led to a significant FRET signal response that reached a plateau after ca. 5 h (curve b, Figure 2B). This intense fluorescence change was ascribed to the feedback-driven coupling of the HCR-1 and HCR-2 modules that generated a dramatically amplified FRET signal. The FRET-based transduction approach needed the simultaneous acquisition of fluorophore donor and acceptor emission, thus providing a more reliable and precise analyte detection method in living cells. Figure 2 | (A) Scheme for the comparison of different HCR amplifiers. (B) The FRET (F670/F565) monitoring of the AHCR machine without (a) or with (b) 50 nM I. (C) Fluorescence intensity-wavelength curves of the HCR-2 system (a), HCR-1 system (b), HCR-1/HCR2T cascade (c), and the integrated AHCR amplifier (d) upon analyzing their corresponding analytes (T or I, 1 nM). Inset: Variation of the fluorescence intensity in Figure C. Error bars are obtained from five independent experiments. (D) The electrophoretic experiment of the AHCR machinery and its component modules: HCR-1 mixture without (a) and with (b) its initiator I; HCR-2 mixture without (c) and with (d) its trigger T; AHCR machinery without (e) and with (f) its initiator I. AFM images of the AHCR-generated hyperbranched dsDNA nanostructures (E), the traditional HCR-1-driven linear dsDNA nanowires (F), and the two-layered CHCR-activated brush-like branched dsDNA nanowires (G). Scale bars correspond to 500 nm. Download figure Download PowerPoint The individual HCR-1, HCR-2, and the two-layered cascade HCR (CHCR or HCR-1/HCR2T, for details, see Supporting Information Figure S7) was then introduced and examined to evaluate the amplification capacity of our AHCR machinery (Figure 2A). Here, the I-triggered H6-expelled AHCR system was adapted as the HCR-1 system, and the T-triggered H1-expelled AHCR system was adapted as the HCR-2 system (for details, see Supporting Information Figure S7). For a clear comparison of these different systems, their corresponding fluorescence spectra were acquired after 5 h (Figure 2C). Using a fixed concentration of the corresponding trigger T or initiator I (1 nM), the individual I-triggered HCR-1 system and T-triggered HCR-2 systems showed substantially lower FRET responses (curves b and a of Figure 2C, respectively), while the two-layered CHCR showed a higher FRET response, yet was still significantly lower (curve c of Figure 2C) than the AHCR amplification system (curve d of Figure 2C). The AHCR amplifier showed a 36-fold FRET enhancement as compared to the conventional HCR-1 or HCR-2 system, and a 4-fold FRET enhancement as compared to the two-layered CHCR. The variable FRET responses were rationalized by the autonomously and progressively hierarchical cross-HCR reaction accelerations guided by the AHCR amplifier, where the autocatalytic activation of HCR-1 and HCR-2 constitutes led to synergistically enhanced FRET signal generation. The linear signal amplification (1:N) and quadratic amplification (1:N2) were, respectively, achieved in conventional HCR (HCR-1 and HCR-2) and two-layered HCR (HCR-1/HCR2T cascade), while a substantially higher signal amplification (1:NN) was obtained by the newly introduced AHCR amplification process ( Supporting Information Tables S3 and S4). As expected, a pseudosigmoidal profile of the FRET response was observed for the AHCR machinery ( Supporting Information Figure S7), suggesting an autonomous and synergistic cross-activation, feedback-driven, lead-in HCR-1, and transducer HCR-2 processes, and a powerful exponential amplification format. The detailed working principles of the AHCR machinery were further elaborated by additional control experiments (for detailed descriptions and discussions, see Supporting Information Figure S8). Native electrophoretic experiments were further applied to confirm the entire AHCR working principle. To verify whether the AHCR machine proceeds as engineered, each reaction module of the AHCR amplifier was investigated. Upon introducing the trigger into the corresponding HCR-1 (lane b, Figure 2D) or transducer HCR-2 (lane d, Figure 2D) system, lots of supramolecular dsDNA nanoassemblies emerged with the simultaneous exhaustion of the hairpin reactants. The integrated AHCR amplifier was also investigated by gel electrophoresis. No new bands were observed for the AHCR reactants without the initiator (lane e, Figure 2D), yet many bright bands of substantially higher molecular weights were formed for the initiator-activated AHCR system (lane f, Figure 2D). These results reveal that the reactants of AHCR and its HCR-1 or HCR-2 constitute are stable enough without undesired signal leakage and are assembled into supramolecular nanostructures upon the introduction of their respective triggers. The morphological features of the AHCR-generated supramolecular copolymers were evaluated by atomic force microscope (AFM). Figure 2E shows the AFM image of the AHCR-assembled hyperbranched dsDNA nanocomposites (for additional images, see Supporting Information Figures S9–S11). As expected, the height of the micrometer-sized hyperbranched dsDNA nanostructures was observed to be ∼1.5 nm ( Supporting Information Figure S9B), a characteristic height of dsDNA. To further evaluate the reaction scheme of the AHCR machinery, the morphological features of the copolymer product at the earlier stage of AHCR assembly were probed by AFM while allowing the AHCR to propagate for different reaction time-intervals ( Supporting Information Figure S11). Interestingly, a dendrimer-like dsDNA nanostructure was revealed by AFM even at the early stages of the AHCR machinery. This is consistent with the fact that the lead-in HCR-1 and transducer HCR-2 are autonomously cross-initiated and cross-amplified during the isothermal AHCR process. Moreover, the size (diameter) of the hyperbranched dsDNA assembly increased with prolonged reaction duration ( Supporting Information Figure S11), demonstrating the correctness of the autonomous reaction pathway of the AHCR machine. Compared to the tiny spots of the AHCR mixture without initiator I ( Supporting Information Figure S9C), the I-stimulated AHCR amplifier generated an increasing number of HCR-1 transmission initiators and HCR-2 transmission triggers with prolonged reaction time-intervals, leading to the generation of hyperbranched dendrimer-like dsDNA nanostructures (Figure 2E). Furthermore, AFM images of traditional HCR and CHCR systems (as two important positive controls) revealed long linear DNA structures of traditional HCR-1 system (Figure 2F and Supporting Information Figure S10) and a brush-like branched dsDNA structure of the two-staged CHCR system (Figure 2G and Supporting Information Figure S10), respectively, the correctness of our AHCR The of the highly amplified AHCR machinery was then applied for the analysis of target I. Figure represents the fluorescence monitoring in the of of the AHCR amplifier upon analyzing variable concentrations of the initiator I. fluorescence were an by the concentration of the initiator I. A of the was observed with the increased initiator The experimental fluorescence at variable concentrations of I were stimulated (for a comprehensive on the and the see Supporting The of was utilized for the of the And the of the of each reaction of the AHCR amplifier was listed in Supporting Information 1 and 2 the of HCR-1 reactants and HCR-2 reactants, And the fluorescence change of the AHCR machinery. These provided the for the of two different HCRs (HCR-1 and HCR-2) of the AHCR amplifier, and were in a The fluorescence readout of the AHCR machine upon analyzing different concentrations of analyte were collected and further to nonlinear procedure in the by using the AHCR A between the experimental results in Figure and results Figure was Figure | (A) The FRET (F670/F565) monitoring and of the AHCR machinery incubated with different concentrations of initiator I. (B) Fluorescence intensity-wavelength curves of the AHCR amplifier upon different concentrations of initiator I. The concentration of initiator I from a to h is 1 10 1 10 and 50 Inset: of the AHCR system in Figure (C) of the curves of AHCR amplifier, two-staged and traditional HCR upon the same initiator I. (D) Fluorescence intensity-wavelength curves obtained from the AHCR amplifier upon 10 nM corresponding I, analyte. Inset: Variation of FRET in Figure Error bars were obtained from five independent experiments. Download figure Download PowerPoint d M d = d M d = 1 2 I ( I M 1 M M 2 1 M d d = d d = d d = d d = 1 T ( M M 1 2 1 M d d = d d different concentrations of initiator I were to evaluate the sensing of the AHCR system after 5 h of the mixture (Figure It was that the FRET signal significantly increased with the concentration of the initiator I, the enhanced formation of hyperbranched dsDNA And the detection of initiator was acquired to be for the AHCR machinery (Figure The of the two-staged CHCR amplifier (HCR-1/HCR2T and the conventional HCR amplifier ( AHCR were also examined to variable concentrations of I (Figure and Supporting Information Figure lower FRET responses were observed for the two-staged CHCR amplifier and conventional HCR system, which were consistent with the control experiments in Supporting Information Figure and Figure The results that the signal of the AHCR machinery was significantly improved as compared to the two-staged CHCR amplifier and conventional HCR amplifier Besides the high signal amplification the sensing is also To further verify the of the sensing of initiator I were Figure the fluorescence intensity-wavelength curves of the AHCR amplifier upon with I and its It was that the analyte (curve Figure and analyte (curve Figure showed FRET as compared with the control (curve Figure the in a FRET response (curve Figure that was lower than that of the initiator I (curve I, Figure Thus, the AHCR amplifier has for its and can Intracellular microRNA imaging by using the AHCR machinery this developed amplification the of an additional recognition module into the AHCR amplification module could easily a universal sensing for has been

Topics & Concepts