【作者】 黄伟涛；

【导师】 李念兵；

【作者基本信息】 西南大学 ， 分析化学， 2014， 博士

【摘要】 分子逻辑与计算装置是一种基于物理、化学、生理过程中分子砌块间的相互作用,实现信息处理过程的分子级装置。它有望解决传统电子计算机由于晶体管电路尺寸和性能极限所带来的量子效应问题,电子线路间信号串扰问题,不可逆计算所带来的巨大能量耗散等问题。并且通过分子逻辑与计算装置构建出具有数字化、多组分组合分析、智能分析、“探测并执行”能力的生化传感装置,将在纳米科学、生化分析、生物医药、环境检测等领域具有广泛的应用前景。但是,分子逻辑与计算研究仍面临诸多问题,如：逻辑输入输出非同一性问题、逻辑门连接性问题、输入或副产物累积问题、二态性难以区分问题等。而且“从逻辑到传感”构建实用的逻辑传感分析系统的研究还刚刚起步,仍面临数字化响应不理想、信号无法甄选和利用、输入-输出编码模式匮乏等难题。因此,发展新的分子逻辑与计算模型,提出新的逻辑与计算理论,应用新型多功能性材料,开发新的模块化系统,对于解决上述问题并实现分子逻辑与计算装置的实用化具有重要意义。本论文围绕如何构建分子逻辑门与实现分子逻辑系统的有机连接,如何构建人工神经元与解决信号模糊化等问题,针对以石墨烯纳米材料为平台,结合核酸适配体、Fenton反应、金属离子、有机染料等物质建立逻辑智能分析系统进行初步研究与探索。1.基于双功能还原态石墨烯氧化物的荧光纳米开关用于检测汞离子和逻辑门操作廉价、稳定的有机染料与纳米材料(如石墨烯、碳纳米管、金纳米等)相结合,有利于设计和构建多功能传感器。本章中,通过简单混合还原态石墨烯氧化物(rGO)和有机染料吖啶橙(AO)制得AO-rGO荧光纳米开关,并利用rGO对Hg2+的选择性吸附和其高效荧光猝灭能力,建立一种免标记、简单、灵敏测定Hg2+的新方法。此外,半胱氨酸(Cys)能与Hg2+高效结合,使Hg2+脱离rGO,导致AO荧光再次猝灭。因此,可以构建Cys-Hg2+驱动荧光可逆开关,用于禁止逻辑门操作。该有机染料-碳纳米材料通用策略可潜在应用于其他免标记生物纳米工程、分子级化学传感与计算以及环境健康科学等领域。2.基于Fenton反应诱导DNA裂解实现荧光增强检测羟基自由基和Fe2+近年来,有许多利用外部竞争物质调控石墨烯的荧光共振能量转移(FRET)作用,从而构建光学传感器的报道。但是,通过割裂探针直接切断FRET的研究还很少。本章报道了一种新型DNA-石墨烯氧化物(GO)-Fenton混合系统,利用了GO的荧光猝灭能力和羟基自由基(HO·)对DNA探针的选择性割裂作用,实现FRET作用调控,从而建立了一种高灵敏度、高选择性测定Fe2+和HO·的新方法。由于7π-π堆积作用,荧光团标记DNA与GO自组装形成复合物,导致荧光团荧光几乎完全猝灭。Fenton试剂生成的HO·可诱导割裂被吸附DNA链,使标记荧光团的DNA碎片从GO表面释放出来,从而恢复荧光。因此,荧光恢复依赖于HO·和Fe2+的浓度。实验表明,增加的荧光强度与Fe2+和HO·的浓度呈正比,线性范围从10nM到1μM,检测限为2.4nM。此外,Fenton-DNA割裂开关还可以用于区分铁元素的价态(如Fe2+和Fe3+)。与传统方法相比,所用DNA不需要修饰巯基、氨基等共价作用基团,从而简化了功能化探针的步骤,并且GO还大大降低了背景信号。该通用策略可潜在扩展应用于检测自由基清除剂,研究新的类Fenton反应,体内ROS的检测和成像,监测割裂酶活性和环境污染物分析等。3.基于还原态石墨烯氧化物模糊逻辑检测G四联体DNA及其裂解试剂G四联体(G4)结构的检测可以有力推动G4结构动态特性的研究,并且有利于发现新型DNA割裂抗癌试剂,构建基于G4构型变化的生物传感器和分子开关。但是,现有的G4结构检测方法,还存在费时、费力、相对昂贵、需要大型仪器等不足。本章结合纳米技术和模糊逻辑理论的优点,发展了一种基于有机染料-rGO复合物的简单、无标记、通用的策略,用于荧光智能检测G4DNA、HO·和Fe2+。利用AO-rGO复合物作为纳米过滤器、纳米开关,结合rGO与不同构型DNA作用力的差异,实现了目标分析物的有效区分和定量检测。实验表明,AO-rGO复合物的荧光强度会随G4DNA的浓度增加而增强,两者呈线性关系,线性范围为16-338nM,检测限为2.0nM。而Fenton试剂会重新猝灭G4DNA-AO-rGO复合物的荧光,Fenton试剂浓度与被猝灭荧光的强度呈非线性关系。但是,Fenton试剂浓度以10为底的对数与猝灭的荧光强度呈正比关系,线性范围为两段,分别为0.1-100μM和100-2000μM。此外,还发展了一种基于模糊逻辑的新型智能检测方法,主要是利用了模糊逻辑可以模拟人的推理、解决复杂和非线性问题、以及将数字信号转换为语言描述等特点。该方法未来可应用于生物化学系统、环境检测系统以及分子级模糊逻辑计算系统。4.基于石墨烯化学系统的布尔逻辑树用于分子计算和智能分子搜索查询近年来,使用生物化学分子或者生物元件作为砌块,构建新型人造计算装置的研究引起了人们的广泛兴趣。目前,分子计算装置的构建面临的最严峻问题是如何连接分子事件成为有用的装置。本章中,利用布尔逻辑分析并组织连接石墨烯、有机染料、凝血酶适配体以及Fenton反应等化学事件形成布尔逻辑树。并利用这些化学事件网络执行荧光组合逻辑(包括基本逻辑门和复杂的联合逻辑环路)和模糊逻辑计算。基于布尔逻辑树分析和逻辑计算,这些基本的化学事件可以作为可编程“字符”,化学相互作用作为“语义”逻辑规则,从而可构建分子搜索引擎用于智能分子搜索查询。该方法有利于发展分子级高级逻辑“程序”,未来有望应用于生化传感、纳米技术、载药系统等。5.pH感应分子神经元用于逻辑计算、信息编码和加密脑环路具有强大记忆、联想、推理等功能。为此,许多研究者致力于使用人工分子系统或电子装置来设计神经元类似物。但是,目前还没有提出一个合理的分子神经元模型。本章中,通过使用紫外-可见光吸收、荧光、共振光散射光谱技术研究常用的酸碱指示剂刚果红(CR)染料的pH依赖响应,展示了利用CR如何构建分子神经元以及表现出类神经元行为。所提出的分子神经元模型使用分子基团作为“树突”以接受环境刺激输入(如pH),分子母体作为“胞体”以执行整合功能,光谱性质的变化作为“轴突以产生输出。通过将简单的分子看作McCulloch-Pitts神经元(线性阈值门),可设计实验用于执行大批量并行逻辑计算,并且借助分子神经元内在的超高信息密度进行字符信息编码以及分子加密。实验结果表明,分子可以用作通用人工神经元,并具备环境刺激应答、分子事件模式识别、推理判决等能力。更多还原

【Abstract】 Molecular Boolean Logic and Computing (MBLC) is becoming increasingly popular, because a chemist’s bottom-up approach toward for information processing might be an attractive alternative, which is potential to overcome the development bottleneck of information technology:the chip size, cross-talk, and heat dissipation. And, construction of digital, intelligent,"sensing and acting" biochemical sensors based on MBLC will be helpful in development of nanoscience, biochemical sensing, biomedical engineering, environmental monitoring, etc. However, MBLC is suffering from several well-identified problems:1) lack of multi-functionality, for example, most previous logic gates usually perform only simple operations;2) poor gate connectivity and scale integration;3) buildups of inputs and waste products which will complicate logic operations;4) ignoring dynamic characteristics of MBLC operation;5) difficulty in distinguishing logical0and1. Moreover, Design and applications of logic-based sensing systems are still at a very early stage, and face lots of challenges, such as unideal digital responses, difficulty of differentiating and utilizing fuzzy signals, and lack of input-output codes, etc. Thus, it is very important for solving above problems and achieving practical applications of logic-based sensor to develop novel models of MBLC, provide new theory on MBLC, and combine new materials with new systems. In this dissertation, valuable explorations have been carried out on how to construct molecular logic gates, conncet logic systems, design artificial molecular neuron, solve and utilize fuzzy signals, and construct graphene-based intelligent logic analysis systems for detection of Hg2+, G-quadruplex (G4) DNA, Fe2+, hydroxyl radical (HO·).1. A Reversible Fluorescence Nanoswitch Based on Bifunctional Reduced Graphene Oxide:Use for Detection of Hg2+and Molecular Logic Gate OperationThe marriage of nanomaterials (including graphene, carbon nanotubes, and gold nanoparticles, et al.) with inexpensive, label-free, common, and stable organic dyes to develop novel sensors with versatile features possesses more great potential. Herein, an organic dye-reduced graphene oxide (rGO) nanoswitch, which is constructed by simply mixing the diluted aqueous solutions of both components, is successfully and efficiently utilized for the label-free, simple, sensitive, and selective determination of Hg2+. Our approach uses the highly selective Hg2+adsorption capacity and effective fluorescence quenching capacity of bifunctional rGO. Moreover, Based on the fact that the addition of cysteine (Cys) into the Hg2+-restored fluorescence system can lead to the release of Hg2+from rGO platform, resulting in the reformation of organic dye-rGO and requenching of dye fluorescence, a reversible on-off INHIBIT rGO logic gate based on the Cys-Hg2+system has also been designed. Molecular logic-gate operation which exhibits intrinsic attractive properties can expand the application of graphene. This proposed nanoswitch design using carbon nanomaterials-common organic dye complex provides a general design strategy, which may expand application potential of various nanomaterials with fluorescence resonance energy transfer (FRET) in label-free optical bionanotechnology, molecular-level chemical sensing, and environmental health science.2. A Simple and Facile Strategy Based on Fenton-Induced DNA Cleavage for Fluorescent Turn-On Detection of Hydroxyl Radicals and Fe2+Recently many graphene-based FRET optical sensors are well-known for external competition sensing approaches. However, the directly switching off FRET based on the probe cleavage remains rarely reported. Herein, a novel DNA-GO-Fenton hybrid system was constructed for sensitive and selective determination of Fe2+and HO·. This strategy is based on exceptional fluorescence quenching ability of GO towards the proximate dye and the switching off of FRET through the highly selective HO-induced cleavage of DNA strands. Due to the π-stacking interactions, a mixture of dye-labeled single-stranded DNA and GO forms a self-assembly of two components which almost completes fluorescence quenching of the dye and needs neither specific design of the probe DNA structure nor additional modification of the DNA termini. In comparison with traditional covalent assembly to fabricate sensing probe, the proposed approach reduces background signal and simplifies procedures for the preparation of functional probe. The GO as efficient nanoquencher modules incorporate in fluorescent probes for DNA damage caused by HO·-generating Fenton reagent. Therefore, switching off FRET depends on the amounts of Fe2+and HO·. In vitro assays with Fe2+and HO· demonstrated increases in fluorescence intensity with a linear range from10nM to1μM and a detection limit as low as2.4nM. The approach based on Fenton-DNA cleavage switch expands Fenton reaction for discriminating various chemical species of an element (such as Fe2+and Fe3+). Besides, the general strategy can be shown potential applications in sensing radical scavengers and investigating new Fenton-like reactions. Our proposed general strategy with its simplicity, sensitivity, and specificity will hold great promise in applications such as probing and bioimaging ROS in vivo, monitoring activity of cleavage enzymes and environmental contaminations, and open new opportunities for the development and design of other extended nanomaterials-DNA systems for oxidative DNA damage study, molecular engineering and sensing applications in the future.3. Fuzzy Logic Sensing of G-quadruplex DNA and Its Cleavage Reagents Based on Reduced Graphene OxideThe specific detection of G4formation can be a great help in studying the dynamic nature of G4structures and designing biosensor and molecular switches based on configurational change. The probing of G4formation also facilitates the discovery of G4DNA cleavage agents which can be used DNA cleaving anti-cancer agents. However, the existing methods for identifying G4structures are not only time-consuming, laborious and comparatively expensive but also require specialized equipment. Herein, by combining the merits of nanotechnology and fuzzy logic theory, we develop a simple, label-free, and general strategy based on an organic dye-graphene hybrid system for fluorescence intelligent sensing of G4formation, HO·, and Fe2+in vitro. By exploiting AO dyes-graphene as a nanofilter and nanoswitch and the ability of graphene to interact with DNA with different structures, our approach can efficiently distinguish, quantitatively detect target analytes. In vitro assays with G4DNA demonstrated increases in fluorescence intensity of the AO-rGO system with a linear range of16-338nM and a detection limit as low as2.0nM. The requenched fluorescence of the G4TBA-AO-rGO system has a non-linear response to Fenton reagent. But this requenching reduces the fluorescence intensity in a manner proportional to the logarithm to the base10of the concentration of Fenton reagent in the range of0.1-100μM and100-2000μM, respectively. Furthermore, we develop a novel and intelligent sensing method based on fuzzy logic which mimics human reasoning, solves complex and non-linear problems, and transforms the numerical output into the language description output for potential application in biochemical systems, environmental monitoring systems, and molecular-level fuzzy logic computing system.4. Boolean Logic Tree of Graphene-Based Chemical System for Molecular Computation and Intelligent Molecular Search QueryThere is a great deal of interest and excitement recently in building and designing of novel artificial computational devices using biochemical molecular events or engineered biological units as building blocks. The most serious, and yet unsolved, problem of constructing molecular computing devices consists in connecting all of these molecular events into a usable device. This report demonstrates the use of Boolean logic tree for analysing the chemical event network based on graphene, organic dye, thrombin aptamer, and Fenton reaction, organizing and connecting these basic chemical events. And this chemical event network can be utilized to implement fluorescent combinatorial logic (including basic logic gates and complex integrated logic circuits) and fuzzy logic computing. Based on the Boolean logic tree analysis and logic computing, these basic chemical events can be considered as programmable’words’and chemical interactions as’syntax’ logic rules to construct molecular search engine for performing intelligent molecular search query. Our approach is helpful in developing the advanced logic program based on molecules for application in bio-sensing, nanotechnology, and drug delivery. 5. Molecular Neuron:From Sensing to Logic Computation, Information Encoding, and EncryptionThe impressive functions of brain circuits have inspired many scientists to attempt in designing neuron analogues by using artificial molecular systems or electronic devices. However, the study of molecular neuron has not produced an equal variety of models. Here, using UV-Vis absorption, fluorescence, and resonance light scattering spectroscopies for recording of the pH-dependent graded responses of common indicators—Congo red (CR) dyes, we show how CR molecules can be used to construct molecular neuron and exhibit neuron-like behavior. Our molecular neural model uses molecular groups as’dendrites’which receive the environmental stimuli inputs (pH), molecular matrix as’soma’which acts as the summation function, and the change in optical characteristics as’axon’which represents outputs. Our approach allows us to utilize simple molecules as McCulloch-Pitts neuron (the linear threshold gate) for experimental implementation of large-scale logic computation in batch mode and to use extraordinary information density inherent in molecular neuron for alphanumeric information encoding and molecular cryptography. Our results suggest that molecules could be used as universal artificial neurons with the capability of responding to the environmental stimuli, remembering patterns of molecular events, and making decisions.更多还原