Supramolecular Sensing Platforms: Techniques for In Vitro Biosensing

Abstract

Supramolecular chemistry is a relatively new field of study that utilizes conventional chemical knowledge to produce new edges of smart materials. One such material use of supramolecular chemistry is the development of sensing platforms. Biologically relevant molecules need frequent assessment both qualitatively and quantitatively to explore several biological processes. In this review, we have discussed supramolecular sensing techniques with key examples of sensing several kinds of bio-analytes and tried to cast light on how molecular design can help in making smart materials. Moreover, how these smart materials have been finally used as sensing platforms has been discussed as well. Several useful spectroscopic, microscopic, visible, and electronic outcomes of sensor materials have been discussed, with a special emphasis on device-based applications. This kind of comprehensive discussion is necessary to widen the scope of sensing technology.

1. Introduction

Biosensing platforms deal with the detection of different biologically relevant molecules related to disease, toxicity, pollutants, and biohazards. These detections can be based on chemical, immunological, or enzyme-based sensing methods. In each case, the methods should be precise, selective, sensitive, and user-friendly. Clinicians in day-to-day life encounter several problems due to dependency on several factors that hamper precision sensitivity and human efficiency [1]. To make sensing robust and more ‘care-free’, precisely interacting agents are required that can ‘hold’ and ‘see’ the analyte like a robotic system and tell us its quantity and nature at the same time. The key process for this type of analysis can be viewed as molecular recognition, a key process that prevails in nature from the shortest nucleosides to the largest proteins [2]. Detecting a biologically relevant molecule or ion using another molecule is the heart of supramolecular biosensing. Supramolecular chemistry is viewed as a key process of understanding and constructing robust, superstructures with the aid of this molecular recognition. Enzyme–substrate recognition was first simplified by Hermann Emil Fischer in the 19th century by comparing it with the precision of lock and key [3]. Moreover, after the discovery of antigen-antibody binding and deciphering of protein and nucleic acid structures, scientists started dreaming about making synthetic molecular machines with the aid of non-covalent interactions. These non-covalent interactions seemed to be coded within the conformational space and functional design of a molecule. Supramolecular artificial superstructures received much attention when Donald J. Cram, Jean-Marie Lehn, and Charles J. Pedersen won the Nobel Prize in Chemistry for their research [4,5,6]. Initially, it was termed host–guest chemistry, and later received the name ’supramolecular chemistry’ courtesy of Professor Lehn. Another leap in the field was achieved when Jean-Pierre Sauvage, Sir J. Fraser Stoddart, and Bernard L. Feringa won the Nobel Prize in Chemistry for their works demonstrating molecular machines, the tiniest robots in the world [7,8,9,10]. A molecule now can glide through the surface of another macromolecule like a minivan, and can be ’seen’ by spectroscopic and microscopic tools [11]. Coming back to our topic, these recognitions and robust three-dimensional interactions of molecules can act like a bench for the analysis of clinically important molecules. In this regard, it ought to be mentioned that if we want to make analytical methodology useful in daily clinical practice, it is necessary to also make it device-friendly and simple. Interestingly, supramolecular systems are built by rational manipulation of simple molecular building blocks. Therefore, they can be connected easily with optical, spectroscopic, or electronic devices. Moreover, a thorough knowledge of the relevant engineering is expanding the utility of these molecular assemblies. This review is intended to present a comprehensive and conceptual ground of in vitro sensing of biologically and pathologically important molecules using cutting-edge supramolecular chemistry. We have tried to present to the reader how various supramolecular interaction forces can be utilized to construct a toolbox for sensing. Moreover, it is worth noting that we present selected examples as a way of conceptualizing the topics, rather than presenting a rigorous literature survey on each topic. There are several review articles already present in scientific journals that encompass rigorous literature discussions of each section presented in this review. Before moving on to the main discussion, some background on the topic is discussed in the next two subsections.

1.1. Supramolecular Assemblies and Their Scope for Sensing

Supramolecular systems are formed by weak chemical interactions such as H-bonding, $\pi$-interactions, van der Waals force, the hydrophobic effect, and other non-covalent interactions. The most interesting aspect associated with supramolecular assemblies is that their properties can be tailored by changing their chemical recognition modes and non-covalent forces.

When closely observing supramolecular interactions and assembly, a resemblance with biomolecular interactions is apparent. Scientists from various fields have manipulated or utilized supramolecular chemistry to produce biologically relevant outcomes. G-Quadruplex [12] is one such celebrated example, wherein a natural folding process of nucleic acid has been used by several groups in both biological and non-biological research fields [13,14]. Sensing platforms have a common configuration, as shown in Figure 1a. This includes a component that interacts with the analyte, that is, a sensing site. When the analyte binds to the sensing site, the signal is transduced to the amplification site, which produces a visible outcome or macroscopic effect that is detectable by devices [15]. For supramolecular systems, such interaction can happen in a single molecule or can generate a cascade of processes within the vicinity of the molecular constitution to show some detectable outcome. Supramolecular fluorescent sensors are a perfect example of such systems [16]. Gianneschi et al. designed an allosteric supramolecular catalyst that transduces the signal to a pH-responsive fluorescent probe (Figure 1b). Such an Rh-based complex compound achieves cavity opening by the interaction of nM-scale concentration of Cl⁻ ions, catalyzing a special type of chemical reaction that lowers the pH of the system and triggers the fluorescence of anthracene fluorophore to generate a detectable signal [17]. Elkema and co-workers devised a disulfide hydrogel scaffold with allylic phosphonium salt which, upon reaction with thiol analyte, triggers gel breaking and visible detection of thiols (Figure 1c). They developed a numerical model to predict the amplification cycle’s response to different concentrations of thiol triggers and validated it with experimental data. The system is capable of detecting various thiol analyses, e.g., small molecular probes, DNA, glutathione, protein, etc., at a detection range of 132 to 0.132 μM. The researchers found that the amplification cycle could be initiated by force-generated molecular scission, allowing for damage-triggered destruction of hydrogels [18]. The supramolecular systems described in the last two examples rely on the supramolecular forces described earlier in this section. A fine balance of energy between different supramolecular interactions triggers a transformation from one configuration to the other. In the former example of a cavity-based allosteric fluorescent sensor, Cl⁻ ions open the cavity by coordinating Rh centers of the complex, then catalyze an acid-generating chemical reaction. Such cavity opening exposes reaction sites, and a shift in the chemical equilibrium of an acid-generating reaction triggers the fluorescence of a third fluorescent moiety. Therefore, complex processes such as analyte binding, allosteric activation of catalyst, and signal transduction coupled with one another are regulated by minute chemical equilibrium shifts and supramolecular forces. Goswami et al. have extensively described such coordinate complex-driven catalytic machinery in their review [19].

Cooperativity is a key point in the assembly of small molecular building blocks to form a larger superstructure. The binding kinetics of the smaller similar or dissimilar blocks dictate the stability and overall energetics of the resulting macro-assemblies [20]. A sensing platform endowed with cooperative interaction with its analyte can be viewed as an ideal sensor. Hembury et al. discussed the role of configurational and conformational chirality in sensing asymmetric molecules in terms of cooperativity factors [21]. Host–guest interaction is another facet of supramolecular chemistry; it exploits cooperative hydrophobic forces to sense an analyte in solution or vapor phase [22]. Calixarene is an example of a host molecule that swallows its guest analyte in a 1:1 ratio solely by van der Waals forces [23]. Ionic interactions, a lesser-understood phenomenon, have also been exploited for sensing ions in solutions [24]. Supramolecular chemistry is endowed with an interesting feature, namely, its dynamic nature [25,26,27,28,29], which makes it a perfect platform for sensing applications. Many of the interactions that prevail in supramolecular chemistry rest on the saddle of one or several kinetic states [30], and their shift from one point to the other opens up scope for sensing devices [31]. Waters and co-workers have reported dynamic combinatorial libraries of molecules that rely on the dynamic exchange of several supramolecular binders as excellent platforms for sensing [32]. Moreover, the interactions stated above can also be coupled with cutting-edge nanotechnology to increase their biosensing scope [33,34].

With the development of cutting-edge technologies, supramolecular assembly has found its way into applications in the world of material science. An array forming organic dyes assembled by π–π interaction when sandwiched between two electrodes can serve as an excellent semiconductor device [35]. If such an assembly is responsive to external perturbation, it can serve as an electronic sensor [36]. Therefore, coupling supramolecular chemistry with the device is a key process when making a sensing platform. Recently, O’Donnell et al. have provided a review on this topic [37]. The further inclusion of several engineering technologies is required in order to develop robust sensing platforms for future applications; thus, their review places an emphasis on device-based sensing methods developed within the past few decades that can be used to produce such robust technical advances.

1.2. Importance of In Vitro Sensing

In vitro sensing deals with both sensing substrates and interpreting the activities of living organisms. Unsurprisingly, a complex situation arises when a single component of a single biochemical cycle starts to creating problem(s). It becomes important to specify and quantify which component is causing the problem and in which way. From the point of this problem to the ultimate end of treatment, one or several sensing techniques become inevitable. Clinical detection of sugar may be the basic and starting point of in vitro biosensing [38]. This review is not the place to discuss the details of in vitro sensing protocols that are required in day-to-day clinical practice (for more discussion on this topic, readers are referred to other the recent literature [39,40,41]); however, for a brief discussion we can think about which analytes need to be considered while speaking of in vitro sensing. Macro- or micronutrients [42] and metabolites [43] are the prime molecules that are available from body fluids that require sensing for diagnostic and investigatory purposes. On a more complex level, proteins may need to be sensed as effective biomarkers of various ailments [44]. There are also a number of essential enzymes for which activity should be estimated clinically [45]. Likewise, nucleic acids are related to several different genetic diseases. Pathogenic microbes that attack their host can be detected from their genetic markings. Sensing viral/bacterial DNA or RNA using a device requires state-of-the-art sensing methodologies [46].

2. Small Molecule Sensing

Biologically relevant small molecules are highly diverse, and their sensing is not easy to sum up in a single article. Devising a sensing platform or detection kit relies on the chemistry of the corresponding molecule and knowledge of molecular interactions that can be successfully exploited.

2.1. Glucose and Saccharide Sensing

Glucose is a prime analyte that needs to be monitored on a day-to-day basis. Concentrations of plasma glucose and glycated hemoglobin are both important for the wellbeing of diabetes patients [47]. Fluorescent and colorimetric assays of plasma glucose are effective and cheap methods. Fluorophores coupled with a moiety that can interact with sugar can serve as an in vitro sensing platform [48]. Macromolecular cavities are well known hosts for glucose molecules; some, called synthetic lectins (Figure 2a), have very high binding affinities towards sugar molecules [49,50]. Davis and co-workers have extensively studied these compounds and developed an anthracene-based cavity that can bind and sense plasma glucose selectively and efficiently using conventional fluorimetric analysis [51]. On the other hand, Peng et al. devised a complex system that can exploit fluorescent resonance energy transfer as a quantifier of glucose in human serum. They initially coupled fluorescent up-conversion phosphor nanoparticles (UPCs) decorated with Concanavalin A (ConA) (glucose binding), followed by immobilization with β-cyclodextrin coated (SH-β-CD) Au nanoparticles using ConA SH-β-CD host–guest interaction. This brings the UPCs close to the quencher (the Au nanoparticles) and keeps them “off”. Upon interaction with glucose, which binds specifically at the SH-β-CD interacting pocket of ConA, the whole adduct falls apart and UPCs get back their fluorescence as a quantifier of in vivo glucose concentration [52]. Boronic acid organic derivatives can easily form covalent adducts with cis-diol moiety molecules [53], meaning that they can serve as efficient analysts of saccharides when designed meticulously [54,55,56,57]. Shinkai and coworkers coupled boronic acid systems with photoinduced electron transfer (PET) systems to obtain wonderfully engineered molecules that can bind different saccharide molecules and sense them effectively [58,59]. Boronic acid attached to fluorophores that can be encapsulated within host molecules such as β-cyclodextrin are able to provide better selectivity and sensitivity [60]. Several fluorophores have a strange property of showing aggregation-induced enhancement of emission (AIEE) [61]. This phenomenon stands at the crossroads of supramolecular chemistry and optoelectronic device manufacturing [62]. Aggregation of a fluorophore can be engineered by changing H-bonding, ionic, or other non-covalent bonding motifs [63,64]. Pyrene, an excellent AIEE active fluorophore, shows (Figure 2b) excellent selectivity towards glucose when coupled with a boronic acid derivative [65]. Aggregation also changes the absorption pattern of pyrene, which has been used as a marker of saccharide sensing. Zhang et al. recently presented an analytical platform for sensing six monosaccharides through a distinctive and accurate method using pyrene boronic acid derivatives (shown in Figure 2c) [66]. Xu et al. have expanded the scope of boronic acids even further by detecting cell surface glycans of mammalian cancer cells [67]. The biocompatibility of large organo-boronic acid derivatives makes them usable even for in vivo determination of sugar levels [68]. Glucose concentrations can also be determined on solid surfaces. Sensor molecules attached to the surface can transduce various signals [69,70]. Here, the fascinating chemistry of boronic acid is again very useful. Kong et al. made intelligent use of surface boron chemistry by preparing an organo-boronic acid functionalized surface that can capture sugar molecules. On the functionalized surface, they selectively sensed glucose using a SERS active boronic acid conjugated triosmium carbonyl secondary probe, as shown in Figure 2d [71]. Extending the same perspective of functionalized surfaces, Kim et al. recently developed boronic acid functionalized gold electrodes for electrochemical sensing of glucose [72]. Due to their size effect, nanomaterials show higher efficacy and selectivity towards sensing [73]. Wang et al. coupled boronic acid with PtAu/CNT nanocomposites to make nanoenzymes. They fabricated a potentiometric device that can sense glucose concentration to $70\times {10}^{-4}$ M concentration of the lowest limit (Figure 2e) [74]. Surface plasmon resonance is another fascinating feature of nanomaterials that has been developed for glucose sensing. Self-assembled monolayers containing a boronic acid pyrene derivative on a gold surface has been reported as a good sensing platform for glucose (Figure 2f) [75]. More sophisticated techniques, such as contact lens-based continuous monitoring of tear glucose, have been achieved using boronic acid chemistry as well [76]. Using the phenomena of hydrogel swelling, Butt and co-workers fabricated a holographic optical diffusing microstructures device with boronic acid containing a glucose-responsive hydrogel which can change the dimension of the optical microstructure upon swelling in the presence of glucose [77].

2.2. Hydrogen Peroxide (H₂O₂) Sensing

H₂O₂ is an important redox analyte that is related to human aging and disease [78]. H₂O₂ is produced by the pathophysiological action of vascular and inflammatory cells, and causes huge oxidative stress [79,80,81]. Redox-responsive motifs conjugated with supramolecular scaffolds [82,83,84,85] can act as smart redox sensing devices. Murthy and co-workers reported dye-encapsulated redox-responsive organic nanoparticles that can respond to local H₂O₂ concentration to produce light-emitting nanoparticles [86]. Ma and co-workers have devised a high throughput plate-reader-based screening method for sensing H₂O₂ in vitro and in vivo by exploiting host–guest chemistry. They developed a library of nine commercially available fluorescent dyes along with six hosts, namely, four cucurbit[n]urils, one pillar[n]arene, and one macrocyclic cyclobenzene[n]. They used a pro-guest molecule which can serve as a competitor to the above-stated dyes upon oxidation by acting as a host molecule. A reactive oxygen species, here H₂O₂, oxidizes the pro-gest to a guest molecule to create an inclusion complex and displace a dye molecule from the host (Figure 3a) [87]. Ferrocene is another example of a redox motif that has been exploited in myriad ways to make stimuli-responsive smart supramolecular materials [88,89,90]. Ferrocene can form an excellent host–guest complex with β-cyclodextrin; this interaction [91,92] has been utilized to make redox-responsive smart supramolecular materials [93,94,95]. Kivrak and co-workers developed a nonenzymatic platform for electrochemical sensing of H₂O₂ using ferrocene naphthoquinon conjugates. They utilized this sensing material in cyclic voltammetry (CV) and differential pulse voltammetry (DPV) to achieve high sensitivity in detection [96]. Porphyrins are also biologically relevant redox motifs, and serve as physiological redox carriers. A host–guest complex of Co-prot-porphyrin with a cyclodextrin dimer has been designed to sense H₂O₂ to $2.47\times {10}^{-7}$ M concentration (Figure 3b) [97].

Molecular organic frameworks (MOFs) are formed by the interaction of specially designed ligands and selected metal ions to produce rigid three-dimensional frameworks with multiple functional applications. Both the metal ion and the ligand serve as electrochemical platforms for sensing redox-active molecules. Liu et al. have reported a phthalocyanine-based conjugated coordination polymer that can effectively sense H₂O₂ secreted from A549 living cells [98]. On the other hand, Hou and co-workers devised an Au–Pd@UiO-66-on-ZIF-L/CC-based system for the same kind of detection [99]. Recently, Haynes and Dey have presented a nice review on redox-responsive MOFs, which readers can consult for further insight [100].

Supramolecular aggregates can trigger a reaction between a redox-responsive dye and H₂O₂ using its peroxidase-like catalytic activity [101]. As discussed in the previous subsection, aggregation-induced enhancement of emission (AIEE) is a special feature of organic dyes. In this phenomenon, aggregation causes the freezing of certain non-radiating processes of energy dissipation from the corresponding electronically excited states of the molecules. Tetraphenylethylene (TPE) moiety is a celebrated example of organic dye showing AIEE. Redox-responsive hydrophilic moieties such as phenylboronic esters attached to a TPE system can help it to solubilize in aqueous media, which upon the action of H₂O₂ can break into the hydrophobic molecule. Upon aggregation, this causes fluorescence (Figure 3c) [102]. Maitra and Dutta recently reported a boronic ester containing a Tb(III)-based hydrogel that can be used as a paper-based device [103]. Peptides containing thiazolidine groups can also form hydrogels that can be disrupted upon interaction with H₂O₂ [104]. On the other hand, Chan et al. recently reported a reverse self-assembly phenomenon, i.e., disassembly triggered by H₂O₂, which can serve as a sensing platform for the latter [105].

Figure 3. (a) H₂O₂ sensing exploiting host–guest interaction. Thiol pro-guest forms a guest upon oxidation, resulting in a dithiol guest which competitively displaces one or more fluorescent dye(s) to produce a fluorescent outcome measurable by a plate reader [87] (reproduced with permission from the American Chemical Society). (b) Co(III)-prot-porphyrin acts as a guest for cyclodextrin dimer, and can sense H₂O₂ electrochemically [97] (reproduced with permission from Elsevier). (c) Tetraphenylethylene (TPE) conjugated to phenylboronic ester shows H₂O₂ mediated oxidation to produce fluorescent aggregate as a visible sensor [102] (reproduced with permission from Elsevier).

Figure 3. (a) H₂O₂ sensing exploiting host–guest interaction. Thiol pro-guest forms a guest upon oxidation, resulting in a dithiol guest which competitively displaces one or more fluorescent dye(s) to produce a fluorescent outcome measurable by a plate reader [87] (reproduced with permission from the American Chemical Society). (b) Co(III)-prot-porphyrin acts as a guest for cyclodextrin dimer, and can sense H₂O₂ electrochemically [97] (reproduced with permission from Elsevier). (c) Tetraphenylethylene (TPE) conjugated to phenylboronic ester shows H₂O₂ mediated oxidation to produce fluorescent aggregate as a visible sensor [102] (reproduced with permission from Elsevier).

2.3. Metal Ions

Metal ions play a vital role in physiological activities, from oxygen carriers to neurotransmission [106,107,108]. Excess or deficiency of metal ions triggers several ailments in the human body, and their detection holds clinical importance in diagnosis [109]. Natural redox behavior [110] and binding with biological ligands [111] sometimes make them a very crucial component. Without Mg(II), which stabilizes the nucleic acid backbones, we could hardly imagine life [112]. Ten essential metal ions can be listed that are related to crucial physiological activity. Shi et al. have provided a nice tabulation of these metal ions along with their proper concentration limits in body fluid [113] (see

Table 1. Metal ions relevant to biological systems and their normal levels (adopted from [113]).

Metal Ions	Optimum Level in Physiological System
Na+	135–145 mM (serum)
K+	3.5–5.4 mM (serum), 19–66 nM (urea)
Ca2+	10–6 M (intracellular), 10–3 M (extracellular fluid)
Mg2+	0.65–1.05 mM (serum)
Cu2+	1.4–2.1 mg/kg (adult human body)
Zn2+	12–16 μM (serum)
Fe3+	14–32 μM (serum)

).

Several physical methods are available for the detection of metal ions obtained from biological samples [114], although most of them are not very cost-effective and require training. Therefore, easy, reliable, and highly sensitive detection probes and devices are needed in order to expand the scope of sensing. Supramolecular chemistry has always exploited ion-to-ligand binding as a key event of molecular recognition, such as in crown ethers [115,116,117]. From these early developments until the most recent achievements of molecular machinery, ligand metal binding has been exploited in myriad ways [118,119]. The size of a crown ether dictates which metal ions it binds to [120], and this binding is highly specific [121]. Unio and co-workers first discussed crown ether-based colorimetric sensors in the 1970s. In this approach, a chromophoric dye is attached to the crown ether scaffold, working as a sensor end, while the dye itself works as a signal transducer end in a simple yet elegant use of a molecular device [122,123]. Stubing et al. used spiropyran coupled with crown ether to make a platform for Li⁺ ion sensing by visible detection [124]. Recently, Knag et al. modified this into a spiropyran functionalized aza-crown ether that can selectively bind Li⁺ ion and shows fluorescenct emission at 550 nm. Upon binding, Li⁺ triggers isomerization of the spiropyran and produces a detectable fluorescent outcome, allowing Li⁺ to be sensed in both in vitro and in vivo systems (see Figure 4a) [125]. Expanding the horizon of crown ether even further, Olsen et al. fabricated device with 18-crown[6] ether covalently connected to a reduced graphene oxide (RGO) surface. This material can be fabricated as a device with glassy carbon electrodes (GCEs) and screen-printed carbon electrodes (SPCEs) for the selective potentiometric detection of potassium ions of biologically relevant samples, something that was never achievable with previous materials [126].

Figure 4. (a) Spyropyran appended crown ether for detection of Li⁺ [125] (reproduced with permission from Elsevier). (b) G-quadruplex for detection of Na⁺ ion, a specially designed G-quadruplex that changes conformation upon binding with Na⁺ ion and reflects a change in heme activity of hemin DNAzyme activity [127] (reproduced with permission from Elsevier). (c) Paper strip device containing metal coordinating dye for sensing of Hg²⁺ ions. A coordination polymer forms upon the addition of Hg²⁺, which changes the fluorescence outcome [128] (reproduced with permission from Elsevier). (d) BODIPY ligand for detection of both Al(III) and Sn(II) ions [129] (reproduced with permission from Elsevier).

Figure 4. (a) Spyropyran appended crown ether for detection of Li⁺ [125] (reproduced with permission from Elsevier). (b) G-quadruplex for detection of Na⁺ ion, a specially designed G-quadruplex that changes conformation upon binding with Na⁺ ion and reflects a change in heme activity of hemin DNAzyme activity [127] (reproduced with permission from Elsevier). (c) Paper strip device containing metal coordinating dye for sensing of Hg²⁺ ions. A coordination polymer forms upon the addition of Hg²⁺, which changes the fluorescence outcome [128] (reproduced with permission from Elsevier). (d) BODIPY ligand for detection of both Al(III) and Sn(II) ions [129] (reproduced with permission from Elsevier).

Nucleic acids can bind metal ions with or without biological significance [130]. This can be seen in naturally occurring G-Quadruplex, which is nothing but the binding of the metal ion to a nucleic acid oligomer chain with a certain repetition of guanine (G) residue. DNAzymes and aptamers contain “loop” site(s) that can selectively engulf a particular metal ion. A play manipulation of structures of these structures as sensor probes can produce metal ion sensors with significant proficiency. Xiong et al. recently exploited a DNAzyme, CRISPR-Cas12a, for point-of-care sensing of Na⁺ in human plasma [131]. Guanosine has selectivity for K⁺ ion [132], and G-quadruplex, which is known to bind K⁺ has been used as a sensing platform for the said metal ion when combined with thioflavin T, itself a competitive fluorescent binder of G-quadruplex. The turn-off label-free fluorescent platform provides a detection limit of $21.87\pm 0.59$ nM in body fluid [133]. On the other hand, Sun et al. exploited a special type of G-quadruplex (p25), which can replace K⁺ in the presence of Na⁺ and trigger a conformational switch which perturbs its hemin DNAzyme activity. Using this model, they developed a sensor for serum Na⁺ concentration up to 0.6 μM concentration (see Figure 4b) [127]. Enzyme activity coupled with fluorescent outcome has also been used on the above-stated G-quadruplex to reach higher sensitivity [134].

Nucleic acid aptamers are oligomeric nucleic acids that can bind to a specific target for a biological response. However, they can be manipulated to inculcate sensing properties due to their tunability in size and functionality. Hairpin-like Pb²⁺ binding DNA have been decorated with fluorophore and quencher, which in a non-bounded state get close enough to quench fluorescence. However, upon binding with Pb²⁺ it shows an elongated structure that becomes fluorescent [135]. Chen et al. designed a Pb²⁺-aptamer that can detect such ions in serum samples in the 100–1000 nM range [136]. Metal aptamers [137] have spacial selectivity towards their metal ion counters, and are made by a meticulous design strategy. An Na-aptamer can selectively encapsulate Na⁺ ion, throwing out any other ion previously bound to it. If the latter ion is a fluorescent metal ion such as Tb³⁺, then a fluorescence turn-off can be the sensing parameter of Na⁺ [138]. On the other hand, a fluorescent tag on the aptamer itself can be helpful as well [139,140]. Lu and co-workers used fluorescent tagged aptamers to determine in vivo concentrations of Na⁺ [141]. They also found a very efficient way to determine Li⁺ ions in biological samples using fluorophore-tagged Li-aptamers with more than 100-fold specificity, which is difficult to achieve by other conventional methods [142]. The choice of fluorophore is important in this regard. Yang et al. synthesized a two-photon absorbing dye and attached it with DNAzyem to sense Zn²⁺ in vivo [143]. Therapeutically important RNA-aptamers have also been found to be useful in sensing and imaging Ag⁺ ions in biological systems [144]. Further extension has been made with these DNAzymes to sense other divalent heavy metal ions such as Cu²⁺ [145,146], Pb²⁺ [147], Ni²⁺ [148], and more. A detailed discussion on this topic is beyond the scope of the current review; interested readers may consult the recently published review by Xu et al. [130].

When co-ordinated with metal ions, dyes and their supramolecular assemblies can change their conformations and mode of assembly, which in turn can dictate their fluorescent outcome [149,150]. Recently, Dey has reported a heterocyclic organic dye which changes its conformation from non-planar to planar in the presence of Hg²⁺ ions in a biological sample. This changeover triggers coordination polymer formation in the dye assembly and shows observable visible and fluorescent changes. Dey used this highly selective and sensible material in making a paper-based sensor device (Figure 4c) [128]. Madhu et al. reported a similar kind of dimer formation for benzimidazole-substituted BODIPY (4,4-difluoro-4-bora-3a,4a-diaza-s-indacene) dye in the presence of Hg²⁺ ions. This also leads to a fluorescence enhancement, which is useful for detection of Hg(II) in the human body [151]. BODIPY has been found to be an excellent biocompatible organic dye endowed with the ability to sense several physiologically relevant metal ions. It has high quantum yield, good photostability, and can show several kinds of stimuli-responsive fluorescence behavior depending on the environmental constraints. A metal ion acceptor (or co-ordinator) attached to BODIPY can serve as an excellent supramolecular platform for metal ion sensing. Recently, Sprenger et al. have demonstrated that a benzo-15-crown-5 system attached to BODIPY shows fluorescence enhancement in the presence of lysosomal Na⁺ ion [152]. On the other hand, o-aminophenol-N, N, O-triacetic acid (APTRA) metal-recognition moiety attached to BODIPY shows selective Mg²⁺ sensing in body fluid [153]. For the detection of trivalent p-block metal cations, more meticulous designing methods are needed. A quinoline or pyrazine moiety is useful for capturing such ions. Kursunlu designed a chelator molecule with two BODIPY pendants on two sides of anthraquinone. This molecule can detect a trace amount of Al(III) due to the formation of intramolecular charge transfer [154]. Two BODIPYs, appended to a phenolphthalein have also shown remarkable results for sensing both Al(III) and Sn(II) (see Figure 4d) [129]. Synthetic strategies at both simple and complex levels help to build supramolecular receptors. BODIPY chromophores with attached N/O/S moieties can sense Ni²⁺ present in intracellular fluid [155]. With the aid of modern synthetic strategies, Belfield and co-workers produced a Fe³⁺-sensing BODIPY molecule that can efficiently quantify the subject metal ion at the cellular level with little or no cytotoxicity [156]. Electronic conjugation of BODIPY dye dictates its absorption and emission wavelengths. A change in the pi-cloud electronic arrangement can generate an NIR fluorescent BODIPY dye sensor [157]. Metallic toxins are also clinically important analytes. Modification of BODIPY dye to form a selective bond with toxic metal ions has found special interest in sensor development. Zhang et al. coupled a meticulously designed BODIPY dye with a solid phase extraction (SPE) technique to fabricate a low cost paper-based sensor device. The SPE column helps to remove most of the heavy atoms and separates the metal ions, resulting in fluorimetric discrimination during detection [158]. Xu et al. recently determined the role of a rigid C-N-N pincer ligand, rationally designed by modifying the BODIPY chromophoric center, in selective detection of Pd²⁺ in cellular fluid. The covalent ligation of the C-Pd bond dictates the metal specificity and observable fluorescence change of the sensing ligand [159]. On the other hand, coordination of two toxic Hg²⁺ ions on extended conjugation of BODIPY dye can produce a visible change in fluorescence in aqueous solutions [160].

Schiff bases are a well known variety of organic dyes used in a variety of industrial products. They contain nitrogen as a heteroatom, which is useful in coordinating various metal ions. Their meticulous design and synthesis is useful in formulating metal ion sensor platforms [161]. Patra and co-workers have reported an azino bis-Schiff base for colorimetric detection of Pb²⁺ in the 8 nM concentration range [162]. Fluorophores may stay in a turned-off state due to electron transfer from the N-atom lone pair if they undergo photoinduced electron transfer (PET) in a Schiff base [163]. Das and co-workers used one such system based on naphthalene and found that Al³⁺ ion can switch off PET by engaging the lone electron pair, showing turn-on in the fluorescence of naphthalene. They proposed a sensing platform for Al³⁺ with a detection limit of $5\times {10}^{-5}$ M [164]. Other trivalent metal ions such as Fe³⁺ have also been detected using this kind of fluorescence enhancement strategy [165]. When attached to nitrogen heterocycles, the 5-(dimethylamino)naphthalene-1-sulfonyl or dansyl group shows PET-mediated fluorescence quenching [166]. Using this concept, Wang et al. used dansyl-attached Asp-His dipeptide for measuring concentration of Zn²⁺ ions in cellular fluids [167]. Unlike PET, fluorescence resonance energy transfer (FRET) is a process of exchange of excited state energy between a donor and an acceptor fluorophores. This process sometimes leads to the enhancement or extinction of fluorescence outcomes. Any FRET coupled with a metal coordination center can lead to FRET-on or -off upon metal–ligand interaction. A recent report has discussed FRET-based devices for the detection of Fe³⁺ in human serum samples [168]. Zheng et al. exploited the click chemistry of Cu ions to ligate modified rhodamine dye and reduced graphene oxide nanosheets. This process in turn causes quenching of the dye due to FRET. This method has been used to detect the Cu concentration of urine samples from patients suffering from Wilson’s Disease [169]. Kim et al. used another type of carbon-based material consisting of carbon dots functionalized with rhodamine to sense Al³⁺ on a paper-based sensing platform [170]. Metal ion coordination with proteins may lead to pathogenic outcomes such as amyloidogenic diseases. Using this as a strategy, Lee et al. designed a Zn²⁺-coordinating amyloidogenic peptide appended with a FRET couple at two ends. Upon coordination, when the two ends of the peptide come close to each other, FRET causes the fluorescence of the acceptor chromophore to decrease. The authors used this method for in vitro detection of Zn²⁺ along with a screening platform for several amyloid inhibitors [171].

2.4. Neurotransmitters

Neurotransmitters remain stored within the presynaptic knob until required, when they are released and provide neuronal impulse. Various physiological conditions cause a deficit or excess in the concentration of these neurotransmitters. The oldest known neurotransmitter is acetylcholine, found in 1921; subsequently, around 100 neurotransmitters have been found, some of which are shown in Figure 5 [172]. Although several traditional analytical techniques are available for detecting neurotransmitters in clinical samples, many of those are not cheap and require high levels of skill [173]. The intrinsic structural chemistry of neurotransmitters can be utilized for sensing; for example, dopamine has a catechol moiety that can strongly coordinate with Fe²⁺ ions. Seto et al. reported a calcein blue–Fe²⁺ complex which becomes fluorescent upon removal of Fe²⁺ ion by dopamine. This competitive binding assay can detect dopamine up to 50 μM concentration [174]. Later, Suzuki developed a better fluorescent ligand for Fe²⁺ which shows stronger fluorescence upon de-complexation in the presence of dopamine [175]. On the other hand, Pal and co-workers developed an Ag-complex with reversible redox and fluorescent properties that can be used for the detection of dopamine [176]. Dopamine is endowed with an electron-rich phenyl ring due to the presence of two hydroxyl groups in the corresponding catechol moiety. Positively charged chromophores such as 2,7-diazapyrenium can form charge transfer interaction with dopamine, and can be a platform for the detection of such neurotransmitters [177]. Similar to saccharides, dopamine has a cis-diol motif in the catechol group which can form covalent bonds with boronic acid [178]. Therefore, the vast supramolecular chemistry of boronic acids has been utilized in designing platforms for dopamine sensing. Coskun and Akkaya functionalized a well-known dye called Lucifer yellow with boronic acid in such a way that it can selectively recognize L-DOPA [179]. Li et al. used BODIPY for selective detection of dopamine by photoinduced fluorescence quenching from dopamine boronic acid adduct [180]. Recently, Kim and co-workers decorated fluorescent carbon dots with boronic acid to sense dopamine by fluorescence quenching assay up to a lower limit of $4.25$ nM concentration in real pathological samples [181]. Boronic acid–catechol adducts have interesting electrochemical properties [182], based on which researchers have devised several sensing techniques. Recently, Ali and co-workers reported a calix[4] pyrrole-appended boronic acid molecule that can sense dopamine electrochemically. They functionalized a glassy carbon electrode with this molecule and found a linear dependency between the strength of Faradaic signals and electro-oxidation of dopamine, along with high electrochemical stability and sensitivity of the device [183].

Epinephrine (adrenaline) is another important neurotransmitter and hormone that controls the muscular contraction and relaxation process. Guo et al. described how the quantitative reaction between epinephrine and formaldehyde can convert the former into a fluorescent heterocyclic dye, which can potentially be a way of determining the concentration of the molecule [184]. Epinephrine has a similar structure to dopamine (Figure 5). Less invasive and more direct ways of sensing epinephrine involve estimation methods using rare earth metals [185]. Recently, Zhang and Yan have integrated the chemistry of a terbium-molecular organic framework (Tb-MOF) with the host–guest interaction of cucurbit[7]uril (CB) to achieve a molecular robot for sensing epinephrine in serum. In the robotic system, the authors describe the CB moiety as a “functional hand” which holds epinephrine using H-bonding and ion–dipole interactions, while the Tb³⁺ present in the environment can sense the neurotransmitter by the fluorescence output [186]. Norepinephrine, on the other hand, is a non-N-methylated analog of epinephrine, and is difficult to distinguish from epinephrine and dopamine using trivial sensing techniques. However, Tian et al. have recently designed two phenyl pyridiniums containing two responsive photons in the fluorescent BPS3 probe (Figure 6a), which can selectively react within a sequential nucleophilic C=O group transfer reaction. The reaction happens on a 100 ms scale, which is very fast, allowing for high temporal resolution in the detection of norepinephrine [187].

Unlike dopamine, epinephrine, and norepinephrine, serotonin (also known as 5-hydroxytryptamine or 5-HT) has only one hydroxyl group in the aromatic ring; however, its intrinsic fluorescence enables sensing using cutting edge fluorescence techniques such as two-photon excitation fluorescence spectroscopy [191]. Moreover, easy derivatization can generate more intense and specific fluorescence outcomes [192,193]. Using the concept of the Hantzsch reaction, Peng and Jiang have reported trace detection of 5-HT in human body fluid [194]. Yoshida et al. used 4-(1-pyrene)butanoyl chloride as a derivatizing agent for serotonin, where both phenolic and amino groups bond with pyrene to produce fluorescent excimers. They demonstrated this protocol for the separation and detection of serotonin from a mixture of other similar neurotransmitters [195]. Serotonin receptors have been investigated in detail for several years for drug design. Recently, Shi and co-workers exploited the idea of serotonin receptors to build a cage-based metal–organic framework (NKU-67-Eu) (Figure 6b) that can detect 5-HT at the 36 nM range in human plasma. In this complex, 5-HT binds to produce an energy transfer to the 1,2,4,5-benzenetetracarboxylate moiety, which in turn results in fluorescence enhancement [188]. Serotonin can form an inclusion complex with cucurbituril, which has been a proposed way of serotonin delivery [196]. Most of its detection deals with host–guest interactions coupled with electrochemistry at the electrode surface [197]. Nanomaterials show particularly impressive sensing ability in this regard [198]. Coupling host–guest interaction with electrochemistry can be an effective way of sensing this neurotransmitter. Abbaspour and Noori used β-cyclodextrin on a carbon nanotube electrode as a host for serotonin and reported an efficient way of sensing the molecule [199]. More recently, Liang et al. used reduced graphene oxide containing Fe₃O₄ nanoparticles decorated with hydroxypropyl-β-cyclodextrin as a composite that can be cast on glassy carbon electrode (GCE), making for an efficient platform for serotonin detection [200].

Histamine, another important small molecule in this class, is a decarboxylated product of histidine related to functions such as appetite, memory, hormonal balance, temperature regulation, and more. Similar to serotonin, chemical modification of histamine produces detectable fluorescent outcomes. Yoshitake et al. used pyrene as a ligator to histamine –NH₂ and the imidazole ring (–NH–) group to produce an intramolecular excimer [201]. Nakano et al. used a similar method to detect histidine in Japanese soy sauces [202]. It is worth noting that histamine has both imidazole and amine groups, which can easily form complexes with several metal ions. This metal–ligand complex formation can be exploited for the detection of histamine in biological samples [203]. Histamine forms a strong complex with several metal ions, including Ni²⁺, Zn²⁺, Cu²⁺, and more [204]. Ali and co-workers used a Ni²⁺ replacement reaction in a nanopore decorated with nitrilotriacetic for the detection of histamine in a micro-fluidic label-free setup [205]. Karim and co-workers designed a Zn²⁺–Schiff base complex salphen for fluorimetric detection of histamine [206]. Crossley and co-workers, on the other hand, developed a Zn-tetraphenylporphyrin for selective detection of histamine from its other analogs (i.e., histidine, nicotine). They presented a two-site binding model for the interaction of histamine, with the complex having an association constant value of $(2.32\pm 0.57)\times {10}^6$ [207].

Acetylcholine (Ach) is a cationic neurotransmitter that transduces signals in both the central and peripheral nervous systems. Much effort has been made to detect Ach and its non-acetylated analog choline (Ch) [208,209,210]. Yitzchaik and co-workers demonstrated a calixarene (host) and aromatic dye (guest) duo. In normal conditions, it remains as a host–guest inclusion complex; however, in the presence of Ach the dye is displaced due to the better binding ability of Ach in the cavity. This results in a blue-shifted enhanced fluorescence (20–60-fold) outcome and acts as a sensor platform. They also immobilized the calixarene on a silica surface to make a solid-state platform for Ach sensing [211]. Recently, Chen et al. used the same host-dye interaction strategy to detect Ch and butyrylcholine within a cellular environment [212]. Dutasta and Martinez with co-workers reported a series of cage molecules that can selectively encapsulate and sense zwitterionic neurotransmitters [213]. The size of these cages dictates the selective inclusion of guests. They synthesized a hemicryptophane cage molecule that can encapsulate neurotransmitters such as choline, taurine, and betaine [214]. Martinez and co-workers developed a fluorescent hemicryptophane cage to selectively distinguish Ach from Ch by attaching a naphthol unit in the cage [215]. However, as the moiety does not have a very high quantum yield, they incorporated more conjugated aromatic systems (Figure 6c) within the cage, resulting in better quantum yields without hampering the binding ability for an improved sensing platform [189]. Incorporation of cyclotriveratrylene and sucrose moieties connected via the naphthalene linkers has also been reported, resulting in better ability to distinguish between Ch and Ach (Figure 6d) [190].

Apart from the neurotransmitters discussed above, there are several other classes of similar molecules, including amino acid neurotransmitters such as GABA, glutamic acid, glycine, etc., which are discussed later in this review.

2.5. Amino Acids

Amino acids, the main constituents of proteins, act in different physiological processes, including metabolism (Lys in the Krebs–Henseleit cycle), tissue repair, generation of neurotransmitters (His), neurotransmission (GABA, glycine), protein biosynthesis (Trp), etc. [216,217,218]. Amino acids fashion all kinds of functionalities that make a protein functional as a molecular machine. First, they have –NH₂ and –COOH groups that can form amide-bonded polymers; moreover, they hold aromatic, aliphatic, charged, redox-active, and coordinating functional groups in their side chains. A common way to sense amino acids in this regard is to catch them by –NH₂ and –COOH. Notably, these two functionalities can coordinate with metal ions [219], which can be a tool for sensing them. Kim and co-workers made a colorimetric probe that can chelate copper (II) ion and show a visible color change in water. Upon the addition of amino acids such as Cys and His, Cu²⁺ ions are removed from the dye, making the solution colorless again [220]. These kinds of platforms can be a medium for sequential sensing platforms using both metal ions and amino acids. Buryak and Severin demonstrated a similar kind of indicator displacement-based assay for the 20 natural amino acids. They used Rh-complex bound with dye molecules as an analyst molecule, which in array format in a plate reader can detect amino acids in a multivariate fashion (Figure 7a) [221]. More recently, Smith et al. designed a boronic ester-containing dye (NS560) (Figure 7b) that can form covalent bonds with both the –NH₂ and –COOH ends of amino acids and act as a turn-on sensor for amino acids. They found good in vitro fluorescence enhancement of the dye molecule upon binding with all 20 natural amino acids [222]. BODIPY probes can also sense amino acids in direct and indirect methods [223]. Styryl boron-dipyrromethene (BODIPY)/2,4-dinitrobenzenesulfonyl (DNBS) is a BODIPY-containing red-emitting molecule which selectively displaces its 2,4-dinitrobenzenesulfonyl group and becomes fluorescent in the presence of Cys [224]. Wang et al. synthesized a BODIPY dye that can distinguish between cystine and homo-cystine in biological samples [225]. Porphyrin molecules are also excellent probes in detecting amino acids. They are endowed with a flat π-surface which makes them in advantageous sensor platforms in many ways. Ogoshi and co-workers reported several Zn(II)-coordinated porphyrin derivatives capable of binding different amino acids with different binding efficiencies depending on their order of hydrophobicity [226]. They also designed chiral bridge isomers of Zn(II)-porphyrin for distinguishing amino acids based on their chirality [227,228]. Several other research groups have subsequently introduced methodologies for chiral distinction of amino acids by varying side chain substitution as well as central metal ions of porphyrins [229,230]. Recently, Fu et al. designed a complex molecular and nanomaterial assembly for the detection of selective chiral amino acids in biological matrix. They employed an assembly of pyridine-substituted Zn(II)–porphyrin molecules with surfactants and CdTe quantum dots (QDs) (Figure 7c). In this assembly, the dye loses its fluorescence intensity; however, addition of dextro- and levo-amino acids restores the fluorescence to a varying extent [231].

Amino acids are endowed with both charged and hydrophobic groups, and variation of side chains makes an amino acid functional in different protein structures. Depending upon the side chain functionality, they can form stable inclusion complexes within a host molecule. In this regard, the size and nature of host molecule(s) dictate their mode and stability of binding [234]. Sulfonatocalix[n]arenes (n = 4, 6, and 8), having overall negative charge, have been found to selectively bind basic amino acids such as arginine and lysine [235]. On the other hand, cucurbit[6]uril has been found to have a certain propensity to form inclusion complexes with amino acid side chains. Interestingly, amino acids can displace a pre-included dye from this host by competitive binding, and this binding ability is reflected in the fluorescence enhancement of the solution [236]. By the choice of guest, the selection of amino acid can also be accomplished. For example, fluorescent pillar[5]arene has been found to selectively encapsulate and sense L-tryptophan within the cavity by π–π interaction [237]. Werner and co-workers have used this host–guest interaction-induced dye displacement at a new level for assaying amino acids and enantiomeric excess by employing tandem assay under a plate reader. They took a dye named Dapoxyl, which shows 200 times higher fluorescence when encapsulated within its host cucurbit[7]uril than in a free state; enzymatic decarboxylation of amino acid leads to the formation of a charged diamine, which is a potential competitor for Dapoxyl and eventually knocks down the fluorescence in the solution phase. At the same time, in a mixture of D- and L-amino acids, only L-amino acid undergoes decarboxylation with natural enzyme and causes the corresponding fluorescence loss of the solution. Thus, it is easy to quantify the enantiomeric excess of the particular amino acid in the mixture [238]. Due to the chiral feature of the host molecules, a bias towards the binding of one enantiomer of amino acids has been found. Jing et al. reported the bias of carbon nanotubes (MWCNTs)-decorated β-cyclodextrin (β-CD) on the L-Pro over its D-enantiomer during electrochemical sensing [239]. Recently, Wang et al. designed a cyclo[6]aramide (Figure 7d) that can form an inclusion complex with amino acids. The resulting inclusion complex responds strongly in circular dichroism spectroscopy (a key method for determining the chirality of aggregates) depending on the side chain of the amino acid [232]. Based on this host–guest chemistry, expansion to other known supramolecular cavities has expanded the scope of amino acid detection related to severe pathogenicity. Cavitands are another kind of rigid cavity-shaped molecules that can engulf amino acids depending on their structural features [240]. Biavardi et al. made a sensing platform based on the covalently bound cavitand Tiiii[C₃H₇,CH₃,Ph] on a silicon surface (Figure 7e). They found that the cavity can encapsulate sarcosine (N-methyl glycine), a marker for prostate cancer, selectively from its un-methylated analog glycine. They quantified sarcosine in patients’ urine samples using X-ray photoelectron spectroscopy (XPS) and luminescence assays [233].

Rational designs of chained macromolecular moieties have further enabled the detection of amino acids. Meticulous tailoring of H-bond and π–π interactions can enable cyclophanes to act as hosts to amino acids [241]. Recently, Cheng et al. reported a fluorescent tetraphenylethene containing cationic cyclophane that can detect amino acids such as Trp using both hydrophobic and ionic interactions. The reported molecule has the right cavity size and shows high selectivity towards Trp [242]. On the other hand, Wang et al. demonstrated exclusive detection of acidic amino acids such as Glu and Asp using calixpyridinium macrocycle [243]. When designed with rational ordering of oxygen and nitrogen atoms, crown ether and its aza derivatives have been found to be effective in the detection of amino acids and their primary metabolites. Alfonso et al. synthesized two chiral crown ethers with S and R chiral centers, with both showing bias towards a particular enantiomer of cationic or anionic amino acids [244]. Crown ethers with attached fluorescent probes have been helpful in detecting lysine in the N-terminals of proteins [245], while open chain cavities are useful for detecting amino acids in aqueous media. Phosphate-appended molecular tweezer has been found to be effective in recognizing Lys, Arg, and His with different binding efficiencies at physiological pH [246]. Schmuk and Geiger developed a tris-cationic pyrrol-containing receptor that can recognize glutamate while not recognizing aspartate thanks to its high specificity towards the former [247].

3. Protein Sensing

Detection of various proteins is a central point of interest in conventional clinical practice. Proteins have different physiological properties related to their structure and disposition of non-proteinaceous functional groups. Detection of proteins using conventional immunoassay encounters with several limitations, including lesser specificity in binding [248]. Supramolecular chemistry has utilized several molecular probe-based approaches to detect them. Ping Wang and co-workers made a protein sensor based on the pattern recognition property of a microcantilever. With the help of simple mercapto-compounds, proteins were detected by the interaction of the –COOH group with the –Si–OCH₃ group, resulting in increased recognition efficiency [249]. Yuan Cao and co-workers utilized a fluorescence sensor array to detect and discriminate metalloproteins. Two bispyrene fluorophores, one with cholesterol and another devoid of it, were specially designed to form an ensemble with CTAB, a cationic surfactant. This ensemble provides distinctive features for each metalloprotein by varying monomeric and excimer emission intensities, which serve as fingerprint features for a particular metalloprotein [250]. Another study showed how a cholesterol-derivatized pyrene with the help of DTAB (dodecyltrimethylammonium bromide) can tune fluorescence emission to detect pepsin and ovalbumin by the ratiometric response [251]. Non-metalloproteins were discriminated from metalloproteins by a derivative of bispyrene-modified perylene (PEPBI) which was assembled with CTAB. Moreover, the sensor was capable of discriminating multiple proteins in serum and urine from mixed samples (Figure 8a) [252]. Li et al. used a porphyrin-based metal-organic framework (PCN-222) for electrochemiluminescence sensing of thrombin ranging from 50 fg/mL to 100 pg/mL [253]. Trypsin detection was performed successfully by supramolecular micellar assemblies. Protamine, a substrate of trypsin, was designed as a supramolecular building block to form a micellar-type assembly along with sodium dodecyl sulfate (an anionic surfactant). A fluorescent signal is detected when the binding event takes place with reporter hydrophobic dye. The detection limit for this assay was 0.044 ng/mL [254]. Discriminating proteins from biological samples has always been a headache; therefore, the development of different new approaches is always desirable. Zhou et al. proposed a fluorescence-based method in which ssDNA is used to induce the aggregation of perylene. Proteins interact with an ssDNA perylene probe assembly to obtain differentiated results, then positively charged proteins trigger a fluorescence turn-on, whereas negatively charged ones turn off the fluorescence. An array-based platform was developed by varying the pH of the assay buffer, allowing nine different proteins to be recognized (Figure 8b) [255]. DNA aptamers are also useful in detecting various proteins in vitro and in vivo. Aptamers, as discussed earlier, have specific binding ability to a particular ligand; therefore, their attachment to a signal transduction moiety leads to efficient sensing. Deng et al. reported a bifunctional platform for the detection of lysozyme and adenosine within an electrochemical setup by using a DNA-aptamer of the corresponding two analytes at the same time. Initially, the two aptamers, one at the electrode surface and another with some complementarity to the previous one, bind together and form a duplex. Then, addition of either analyte leads to the opening of the duplex. The duplex unbinding event can be monitored by the cyclic voltammetric process when the surface unbound aptamer is tagged with an Au nanoparticle [256]. More recently, Ma et al. coupled liquid crystal with aptamer to devise an optical sensing platform for assaying thrombin. The liquid crystal is doped with octadecyl-trimethylammonium bromide(OTAB) resulting in a dark optical image. Upon addition of DNA aptamer (which binds specifically to thrombin) and electrostatic interaction of the OTAB with aptamer, the liquid crystal assembly adopts a bright appearance. Finally, upon addition of thrombin analyte, the aptamer is bound with protein and the liquid crystal regains the dark state [55].

Early diagnosis of diseases by different laboratories and research and development sectors has a huge impact on the healthcare sector. This trend continues for the detection and diagnosis of proteins that are related to human diseases. Detection techniques range from crude analysis, such as with urine samples, to sophisticated bulk scale detection using PCR, which today is being extended to nanoscale detection. Diseases that are caused by aberrant proteins are collectively termed proteinopathies. Mainly, such diseases are due to protein misfolding. Neurodegenerative proteinopathies involve the gradual loss of neurons structure and function. There are commonly two types of neurodegenerative proteinopathy, namely, infectious and non-infectious. Examples of non-infectious proteinopathies include Alzheimer’s disease, Parkinson’s disease, and Huntington’s disease, while infectious proteinopathies include prion diseases. Apart from these, several membrane protein folding and misfolding issues can lead to major pathological insights. Addressing these MPs folding and misfolding by proper diagnosis can allow practitioners to treat the related diseases. A few trans-membrane proteins show physiologically relevant unfolded states. Peripheral myelin protein 22 (PMP22) shows L16P disease-related mutation, which causes misfolding of proteins. Several mutations in the transmembrane protein cause many different diseases, including cancer and neurodegenerative diseases. A missense mutation in the TM-alpha helix of the membrane protein causes protein misassembly by increasing or decreasing helix–helix interactions in the TM domain. The G163R and G167R mutations in the TM alpha helix of myelin protein zero cause Charcot–Marie–Tooth disease and Dejerine–Sottas syndrome [257]. Mutation T135A in connexin 26 protein destabilizes the hexameric oligomer, which causes hereditary deafness [258]. The V232D mutation in the hairpin construct containing the TM3 and 4 -helices of CFTR causes lock-in by non-native hydrogen bonds in two helices into a compact conformation [259].

Several methods have been used for sensing proteins by in vitro approaches. In this article, we discuss them along with their efficiency in disease studies. In vitro methods can be divided into two categories: label-dependent and label-independent. Label-dependent detection technique depends on a specific property of the label, which could be chemical labeling, fluorescence labeling, etc. Label-independent techniques are suitable for those targets which are not labeled or cannot be labeled. The most common label-dependent techniques include gel electrophoresis and chromatography for protein separation. The conventionally used approach for protein detection in gel electrophoresis depends on protein staining in the gels followed by separation [260]. Post-staining methods exist as well, such as fluorescence staining, silver staining, etc. Nanoscale detection of protein pairs by proximity-dependent superresolution microscopy has been introduced as one of the finest techniques for detecting protein–protein interactions by high-resolution in situ imaging. Most superresolution fluorescence microscopy provides false positive results for protein colocalization, as it depends on the local optical resolution. Clowsley et al. derived a new method for circumventing this problem called proximity-dependent point accumulation imaging in nanoscale topography (PD-PAINT). DNA origami has been used to introduce distance-dependent fluorescent imaging of protein targets [261]. The most common non-infectious neurodegenerative disorder is Alzheimer’s disease, for which detection methods are primarily based on histological staining, computed tomographic imaging, and/or MRI. There are other options as well, such as the detection of biomarkers, especially CSF and plasma biomarkers. However, there remains a need for fast and accurate detection in the early stages of this disease. Recently, Dey et al. showed the rapid detection of the biomarker beta-secretase 1 (BACE1) antigen with the help of reduced graphene oxide (rGO) conjugated with the BACE1 antibody and immobilized on fluorine-doped tin oxide electrode. This immunosensing technique reduces detection time and increases the detection limit of BACE1 Ag from 1 fM to 1 μM [262]. Tau protein is another major protein related to neurodegenerative disorders. Apart from conventional methods, several new methods have been developed for fast and accurate detection of this protein. A highly sensitive technique based on sandwich immunoassaying and amperometric detection has been developed in which p-aminobenzoic acid is immobilized on the surface and the antibody is crosslinked with poly(amidoamine) (PAMAM) dendrimer–gold nanoparticle. A secondary captured Tau protein antibody is then bound to the assembly. The detection limit of the device is 1.7 pg/mL, which is very low [263]. Xueli Zhang and co-workers successfully demonstrated near-infrared (NIR) fluorescence imaging for monitoring of Alzheimer’s disease. They showed that CRANAD-3, a curcumin analog, is highly capable of detecting soluble Aβ plaques both by in vitro and in vivo techniques. They also observed that when the Aβ lowering drug LY2811376 was used in vivo, CRANAD-3 was capable of monitoring the decreased Aβ after drug treatment (Figure 9a) [264]. In their previous study, the same group elaborated the effect of other CRANAD molecules (curcumin derivatives) in the detection of soluble and insoluble Aβ plaques by NIR fluorescence imaging. Here, they showed the effect of CRANAD-58 and CRANAD-17 by significant changes in fluorescence properties while mixing them with soluble and insoluble peptides under in vitro conditions [265]. Another interesting study by Law et al. showed the detection of amyloid fibrillation by a luminescence assay. Here the advantage of Pt(II)-Pt(II) interaction to form supramolecular self-assembly was taken into account for the detection of amyloid fibrils. This complex is more advantageous than Thioflavin T due to its much lower interference from autofluorescence [266]. Apart from above above-stated label-dependent detection of non-pathogenic proteins or peptides, several label-independent approaches have been developed over the past few years. One such study demonstrated the use of a graphene oxide–gold nanoparticle-based hydrogel as a biosensor and platform for detecting amyloid beta oligos (ABOs) [267]. It was observed that this impedimetric label-free technique can detect Aβ ABO from artificial cerebrospinal fluid (CSF) while leaving out the monomers and fibrils. α-Synuclein is another important aggregation-forming peptide. It is responsible for Parkinson’s disease, and can be detected by label-independent optical biosensing techniques. Khatri and co-workers showed that a chitosan film in direct contact with a gold nanoparticle array can act as a polysaccharide bioreceptor to detect α-synuclein by the interaction of glucosamine groups of polysaccharide. With the help of a localized surface plasmon (LSPR) detector, this sensing platform can detect up to 70 nM monomers as well as fibrils. Most significantly, this technique can distinguish between monomers and fibrils due to the change in absorbance. It was found that when the sensor detects monomers, the binding kinetics show a sharp rise and then reach saturation. On the other hand, when the sensor detects fibrils it shows a gradual rise in the binding kinetics (see Figure 9b) [268]. Another interesting work demonstrates the identification of small molecule inhibitors of amyloid beta fibril aggregation by label-free surface plasmon resonance (SPR) assay. This methodology relies on SPR peak shifts due to the binding of small molecules to amyloid Aβ-gold nanoparticles (AβGNP). This supramolecular approach detects fibril formation and identifies the blockers of fibril aggregation with the help of SPR [269]. Antman-Passig et al. developed an interesting optical nanosensing detection method to detect amyloid-beta fibrils in the early stages from intra- and intercellular sources. Single-walled carbon nanotubes (SWCNTs) can be transformed to Aβfunctionalized SWNT by π–π and hydrophobic interactions between Aβ and SWCNT. When SWCNT-Aβ42 interacts with the new Aβ the binding favors the Aβ conformation, which ultimately decreases the peptide’s structural stability for increased hydrophobic and pi stacking interactions with SWCNT surface. This disrupts the tertiary conformation of AβẆhile incubated with fresh low molecular weight Aβ species, the observed increase in blue-shifting and the high surface coverage of ABCNTs leads to disaggregation of fibrils. Therefore, this technique can be specifically useful for targeting low molecular weight Aβ42 intermediates and is capable of detecting using live specimens (Figure 9c) [270]. Another study developed a new method using a surface plasmon resonance (SPR) biosensor and sandwich assay for the detection of the complex between tau and Aβ. This approach can identify a potential Aβ–τ complex biomarker from cerebrospinal fluid [271].

Thus far, we discussed the detection of non-infectious proteins, mainly those causing neurodegenerative disorders. We now move on to discussing cancer biomarkers. Protein biomarkers are considered one of the best targets for anti-cancer medicine development [272,273]. One such important biomarker consists of cancer stem cell (CSC) surface biomarkers. Jing Zhao and co-workers developed a strategy to detect breast cancer biomarkers in CD44 stem cells. They employed multiple signal amplification with the help of supramolecular nanocomposites, which contain a binding peptide that is capable of binding to CD44 (CD44BP) and self-assembled FF (diphenylalanine). FF helps to find the AuNP aggregation, which is placed on an electrode surface by a specific linkage. When this assembly detects CD44, it produces a higher electrochemical signal, which is then amplified as this assembly is capable of protecting the target (CD44) from trypsin digestion. When there is no CD44, the whole assembly of architecture is destroyed, and the signal is much lower. This method is very sensitive, able to detect CD44 protein as a cancer stem cells biomarker at a linear range of 2.17 pg/mL (Figure 10a) [274].

A ligand that binds to a specific protein when attached to a fluorescent probe can serve as a simple supramolecular sensor platform. Hamachi and co-workers reported an MTX ligand tagged with rhodamine dye. A phenylalanine hydrophobicity promoter senses intercellular pathogenic and nonpathogenic proteins. Moreover, it undergoes fluorescence turn-off inside the cellular environment and in turn serves to detect protein activity (Figure 10b) [275]. On the other hand, Xu et al. designed an aptamer–graphene oxide (GO) composite probe where the fluorescent tag of the aptamer is quenched in the probe due to the conjugation of GO. Upon binding, several pathogenetic cancer-probing proteins the aptamer leave the GO surface and undergo fluorescence turn-on [276]. Qiu et al. proposed a method based on a bifunctional oligonucleotide probe for detecting protein and nucleic acids by a parallel fluorescence detection technique. Two disease-specific markers for cancer, namely, platelet-derived growth factor BB (PDGF-BB) and the p53 gene, were detected using an aptamer probe combined with a hairpin DNA probe by circular common target molecule (non-nucleic acid strand)-displacement polymerization (CCDP) signal amplification. This highly sensitive and selective procedure can detect PDGF up to $1.8\times {10}^{-10}M$ [277]. Supramolecular techniques are becoming useful for the detection of other proteins, even those that do not cause major health issues such as neurodegeneration or tumor-specific proteins. Examples include the detection of biomarkers produced due to chronic wounds. The recent emergence of the COVID-19 pandemic led to supramolecular protein sensing approaches capable of detecting pathogenic proteins. One such interesting study showed that a novel aptasensor-based technique can be used to target heat shock protein 70 (HSP70) as a biomarker for COVID-19 patients. A sandwich-like system was used to improve the sensitivity of the detection. This sandwich-like system was made up of aptamer conjugated with gold nanoparticle (AuNP) along with glassy carbon electrode (GCE). The GCE surface consisted of reduced graphene oxide (rGO) along with a Acropora-like gold (ALG) nanostructure. The aptamer sequence was covalently bound as a primary bioreceptor to the GCE surface. The analyte (HSP70) was added after that as a second bioreceptor, with the aptamer conjugated AuNP bound to electrode surface, improving the subsequent diagnosis. The detection limit with this technique was 5 pg/mL to 75 ng/mL. This sandwich-like system’s sensitivity compared to well-established RT-PCR was around 90% [278].

4. Nucleic Acid Sensing

Nucleic acid sensing has become a major part of detection of disease-specific gene mutations for pathogenic and non-pathogenic diseases. Detection of nucleic acids by supramolecular sensing platforms is necessary for genetic disease, viral infection, and diseases related to immune response. Recently, with the help of supramolecular chemistry the detection of DNA/RNA for disease detection, food technology, and drug discovery has become a popular and efficient tool. The self-assembling procedure and higher signal recognition efficiency of synthetic molecular patterns make the detection range for sensing much wider. Signal amplification is caused due to the low energy barrier for supramolecular structure reassembly [279]. Organic and inorganic structures, hydrogels, polymers, nanoparticles, and more are used as supramolecular materials to build sensing platforms.

A unique approach has been demonstrated by Law et al. for detecting RNA with the help of supramolecular self-assembly. They used a luminescence assay to detect RNA and image nucleolus with the help of a supramolecular self-assembly of a water-soluble platinum (II) complex. Here, two complexes, 1 and 2, are used; complex 1 has a specific feature, termed the guanidium pendant, which detects RNA via hydrogen bonding, whereas complex 2 lacks this feature. When the targeted RNA binds with the platinum(II) complex due to the aggregation, the complex shows luminescence turn-on [280]. Fluorescence biosensing techniques are very effective among other nucleic acid detection methods. In recent studies, supramolecular constructs of pyrene-modified oligonucleotides were used as fluorescence probes, as they have unique fluorescence properties such as long fluorescence lifetime, intercalating nature within nucleic acid duplexes, and high quantum yields, which ultimately lead to enhanced sequence-specific detection and determination of single nucleotide polymorphism (SNP), among other effects. Krasheninina et al. used dual probes that formed pyrene excimer and detected intracellular RNA by visualizing it. This special probe has higher binding affinity towards its complementary RNA. The pyrene moiety was attached with linkers of varied lengths and structures, which influences the excimer formation. Shorter linkers lead to more excimer emissions after target hybridization. In addition, the length of the 3′ component’s linker arm plays a crucial role in pyrene excimer formation (Figure 11a) [281]. These kinds of two-component pyrene probes based on 2′-O-methylribonucleotides were used in another study to detect miRNAs. In presence of target miRNAs, these two-component probes form a three-way junction (3WJ) structure; one of the probe components contains a penta-adenosine fragment with modified adenine and deoxyriboadenosine in the center. With this supramolecular assembly, a cancer-specific miRNA biomarker was detected by [282]. In another study, one excimer forming pyrene derivative 2′ phenylethynylpyrene was designed to detect three single nucleotide polymorphisms in the Helicobacter pylori 23S RNA gene and differentiate it from the wild type [283]. Another derivative with 5′-Bispyrene molecular beacon was used to detect RNA. This formed excimer when the BHQ1 fluorescence quencher at 3′-end and at 5′ end 2′-O-methyl RNAs containing 5′-bispyrenylmethylphosphorodiamidate formed complex upon binding of the target RNA [284].

Nanopores are another fascinating example of supramolecular detection of nucleic acids. Solid-state nanopores are stable and have higher affinity for the detection of specific nucleic acid sequences (Figure 11b) [285]. With the help of nanopores, DNA detection can be done at a single molecule level; most importantly, the technique is label-free. Nanopores are formed by supramolecular assembly of biomolecules such as proteins [286], 2D materials such as graphene [287], silicon [288], metals [289], quartz [290], etc. Nanopores were initially used for metal ion detection, but recently many developments have taken place in which they are also used for DNA sequencing. In the very early stage of nanopore-based detection, the α-hemolysin channel is used; by applying an electric field at a certain time, single-stranded DNA purine and pyrimidine can be differentiated based on their size difference [291]. Apart from hemolysin, a variety of different nanopores can be used, including aerolysin, porin A, OmpF, etc. Another such engineered nanopore is the protein nanopore MspA, which can be used to detect DNA at single-molecule resolution. MspA, a porin from Mycobacterium smegmatis, contains a narrow channel that gives the advantage to pass a short single-stranded DNA segment through it [292]. Most recently, a high-class cutting-edge detection method was been developed by Oxford nanopore technology to sequence DNA/RNA for the first time [293].

Recently, an interesting study utilized nanopore technology to detect virus sequences and nucleic acid motifs. Biotinylated DNA oligonucleotide probes were designed with respect to the complementary target sequence from the genome. When annealed, the portion of the target which is unpaired to the probe is digested by endonuclease, and as the biotin moiety is present with the probe, these duplexes respond to SS-nanopore assay. This approach was successfully used to detect the HIV-1B genome and highly conserved DNA sequences from the genome. The sensitivity was 10 fM [285]. Asandei et al. used Peptide Nucleic Acid (PNA) as beacons to detect nucleic acid using the nanopore approach. They utilized α-hemolysin along with PNA, which was polycationic peptide-functionalized. Due to the greater binding affinity of PNA and its resistance towards nuclease/protease digestion, detection of target DNA was highly specific. With the help of electrophysiological recording, the single-channel properties of the nanopore were utilized to detect various ssDNA and differentiate among the ssDNA containing mismatched bases [294]. Single-molecule sequencing method from Oxford nanopore was used to detect the plant pathogen. DNA and RNA sequences were isolated from the diseased tissues from plants of several families infected with pathogens such as bacteria, viruses, phytoplasma, and fungi. Then, they were sequenced using standard nanopore sequencing and the data were analyzed. From all the samples, the pathogens were identified within 1 to 2 h with good specificity [295]. Mereuta et al. utilized a protein nanopore containing gold nanoparticles to detect ssDNA with good sequence specificity. Citrate anion-coated gold nanoparticle was used with α-hemolysin (α-HL) nanopore for detecting specific ssDNA at nanomolar concentration. In this novel approach, charge-neutral peptide nucleic acids (PNA) were used as hybridization probes for its complementary ssDNA. The aggregation propensity of the gold nanoparticles (AuNP) was recorded at single-molecule level when PNA-DNA duplex was formed. This method provided a clear readout signal due to different-sized PNA-induced AuNP aggregates when bound to its complementary ssDNA. Overall, this method is a very sensitive and selective option for label-free nanopore-based ssDNA detection [296]. Hairpin probes, especially molecular beacons, have had a huge impact on nucleic acid sensing over the past decade. In one such study to detect nucleic acid at high sensitivity and selectivity, Liu et al. used isothermal amplification with a hairpin probe (HP) to detect miRNA. Here, they used exponential isothermal amplification (EXPAR) along with two hairpin probes. At first, targeted miRNAs were hybridized to DNA template and extended with polymerase, then nicking enzyme was used to detect dsDNA and cut it out. The short DNA generated afterwards, termed the trigger, was first hybridized with HP1, resulting in a DNA construct with a single-stranded tail. HP2 was then bound to this construct as a catalyzed hairpin assembly; due to its extension and replacement, the DNA trigger was displaced and the new construct contained fluorophores. These freed trigger DNA can regulate new cycles again and again. With this technique, $3.0\times {10}^{-15}$ M miRNA could be detected by measuring fluorescence intensity (Figure 12a) [297]. Wang et al. showed that ferrocene-labeled molecular beacons can be used to detect DNA on the surface (AuNP electrode) electroluminescence-based biosensing at lower concentration ranges upto 1 fM from 10 pM (Figure 12b) [298].

Apart from the hairpin probe, aptamers are important tools for biomolecular sensing. In one such interesting work, Wilner and co-workers showed a fluorescence-based DNA detection using a thrombin/aptamer complex. Initially, these thrombin-specific aptamers were used to detect specific peptide sequences; in this study, the authors showed that DNA could be detected using this assembly at a lower concentration [299]. Differentiating between dsDNA and ssDNA was done by Ma et al. using fluorescence technique along with a AuNP-based probe. The method is based on the quenching mechanism of AuNP for fluorescent DNA probes, where two DNA probes are labeled with two different fluorophores, FAM (6-carboxyfluorescein) and ROX (carboxy-X-rhodamine); according to their different emissions, ssDNA and dsDNA can be detected simultaneously [300]. Xiang and Lu combined sandwich hybridization assay with a personal glucose meter (PGM) to detect target DNA from the hepatitis B virus. They captured DNA-coated magnetic beads along with DNA invertase, which was also partially complementary to the target DNA. When the target DNA binds to the MB-capture DNA, invertase DNA is attached to the counterpart and the personal glucose meter produces a signal proportional to the concentration of target DNA by conventional glucose assay (Figure 12c) [301].

Figure 12. (a) Schematic representation of HP1 and HP2 binding with trigger DNA by exponential isothermal amplification to detect miRNA by monitoring fluorescence intensity of attached fluorophore [297] (reproduced with permission from Elsevier). (b) Schematic representation of DNA detection by on-surface electroluminescence assay with Fc-MB on AuNP electrode [298] (reproduced with permission from Elsevier). (c) A sandwich assay with a dual probe to detect target DNA by using a personalized glucose meter (PGM) [301] (reproduced with permission from the American Chemical Society).

Figure 12. (a) Schematic representation of HP1 and HP2 binding with trigger DNA by exponential isothermal amplification to detect miRNA by monitoring fluorescence intensity of attached fluorophore [297] (reproduced with permission from Elsevier). (b) Schematic representation of DNA detection by on-surface electroluminescence assay with Fc-MB on AuNP electrode [298] (reproduced with permission from Elsevier). (c) A sandwich assay with a dual probe to detect target DNA by using a personalized glucose meter (PGM) [301] (reproduced with permission from the American Chemical Society).

5. Comparison of Different Methodologies

So far we have discussed sensing of different biomolecules using supramolecular sensing platforms. However, it is worth mentioning how different sensing techniques vary due to their intrinsic drawbacks and advantages. Sensings are not devoid of physical techniques (e.g. spectroscopic, microscopic, etc.) for detecting changes in bulk or molecular level. Moreover, the use of different types of supramolecular systems needs different kinds of chemical and physical methods to be prepared. Synthesis of large supramolecular structures sometimes needs a big synthetic effort. Supramolecular cages reported in reference [189,190], Figure 6d need multi-step synthesis using complex and costly starting materials and catalysts. However, the detection of neurotransmitters using cage-like structures shown in Figure 6b [188], needs less synthetic endeavor and a higher sensing limit. However, this does not make the former approach inferior to the latter, but a “stitching” is needed to utilize both parallelly. For example, in glucose sensing, earlier reports of sensors were synthetic macromolecules with lower sensitivity [48], however, with time coupling of synthetic chemistry with probe-based sensing methods [71] has made considerable advancement in sensing limit. Combining nanochemistry with supramolecular chemistry is another worthy example of combining two fields of study to synergize sensor technology. Utilization of conventional nanomaterials such as nanoparticles, nanopores, and nano-surface chemistry has revolutionized the sensing ability of pathologically relevant nucleic acids and proteins. Apart from strategic differences, the utilization of assaying techniques also needed to be considered. Many of the above-discussed methodologies used in supramolecular sensing are fluorescence. However, fluorescence is not a noninvasive method. It depends highly on local environments and the temperature of analytical conditions. Tagging biomolecules with fluorescence molecules may alter their active structure and can therefore impact detection. Label-free approaches [133,205] are therefore much less invasive and efficient. High throughput methods are also useful alternatives to label-based approaches. The use of surface-enhanced Raman spectroscopy is one step further in the detection of trace amounts of analytes and can be used as a high throughput method [221] in the future. Cost is another problem in sensing technology. Utilization of cheaper sensing platforms therefore needs ingenious effort. A visible outcome of sensing is sometimes the cheapest way of detection (Figure 1c) [18]. However, this is not achievable in many cases, therefore, utilization of cheap platforms such as paper-based devices is very useful to cut down the cost [103].

6. Scope and Future Prospect

Intelligent use of chemical tools for developing in vitro biosensors has shown a scope of developing platforms that can easily be handled to produce precise qualitative and quantitative outcomes. In a few cases discussed in this current article, the experimentally verified sensing techniques can be extended for real-time diagnostic detections. For example glucose sensing approach used in Reference [52] deals with in vivo measurement of glucose concentration. On the other hand, a multi-calibration potentiometric detection of glucose has been done in urine samples at home [74]. Neurotransmitters like dopamine and 5-HT are involved in brain diseases such as Parkinson’s disease, schizophrenia, drug dependence, alcoholism, etc. Dopamine and 5-HT have been detected from real pathological samples. To eradicate the interference of proteins in serum the samples were treated with acetonitrile which helps the proteins to precipitate [181,194]. Detection of endogenous bio-thiols has also been reported with fluorescent probes under microscopic studies [225]. On the other hand, nanopore sensing for nucleic acids is a very celebrated example where experimental findings have been extended to market available devices [286]. These above-stated examples can easily be converted to real-time clinical devices. More effort should be given to make them market available.

Chemistry with engineering science has already produced many technological devices that we use in our daily lives [302]. Liquid crystals [303] and organic light emitting diodes [304] are such important examples where molecular science has met engineering to produce technological marvels. In our current review, we have discussed some examples that got coupled with conventional engineering counterparts to produce nice sensing outcomes [74,158]. Whereas in some cases the use of data analysis and science has proven useful in tandem detection of bio-analytes [87,221]. We also discussed some examples of how surface chemistry coupled with spectroscopic and microscopic techniques have been useful for analyzing trace biological samples [233,250,270]. Nano-science, a bigger assembly of molecules, when coupled with supramolecular chemistry produced efficient sensing platforms [169,231]. Recent amalgamation of artificial intelligence in sensor devices is showing new hope for more efficient and reliable biosensing platform in the near future [305]. Computer-based chemistry [306] is now a new field that can also find ways towards sensing devices in the future. Cost is another problem in biosensing and efficient molecular design should cut down the cost by employing a tiny amount of sensor molecules to sense a fairly larger number of samples. Lastly, it is also very important to make a multidisciplinary approach to a particular sensing platform. The involvement of high-throughput techniques to make parallel detection of several categorically similar but intrinsically different analytes is only possible by the involvement of cutting-edge engineering techniques such as microfluidics, nano-lithography, 3D printing, and many others.

7. Conclusions

Sensing at the clinical level is challenging if the sensing methods are rigorous and time-consuming. Supramolecular sensing technologies with high sensing ability and low sensing limit can replace the setting of a large-scale sensing system with conventional methods. During the last few decades, a paradigm shift in supramolecular chemistry has made in vitro biosensing much easier and accessible to clinicians. Using molecular assemblies or large synthetic receptor-like structures exploit conventional chemistry to detect and transduce signals received from pathologically important molecules. They act like molecular machines that can hold a small molecule and make it visible to the “beholder”. Continuous effort is needed to develop and promote supramolecular sensing platforms. Incorporation of knowledge from cutting-edge engineering and biomedical research into supramolecular chemistry is necessary to shape realistic sensing platforms for pathologically related analytes.