2.1.1. Anatoxin-A (ATX)
Anatoxin-a (ATX-A) is a neurotoxic alkaloid secondary amine produced by some strains of freshwater cyanobacteria of the genera Anabaena
, and Raphidiopsis
. ATX-A and homoanatoxin-a (HATX) are the most commonly occurring anatoxins, with a molecular weight of 165.232 and 293.282 g/mol, respectively. These are strong nicotinic acetylcholine receptor antagonists mimicking acetylcholine, binding to acetylcholine receptor from where the cell machinery is unable to remove it, bringing nerve and muscles to exhaustion. In high doses, they may eventually result in respiratory muscles paralysis and death [64
Jackson et al. [61
] developed a sensor to detect ATX-A based on responsive fluorescent resonance energy transfer (FRET). The feasibility of this optical sensor was tested with thyroxin, a similar size molecule, but not really with ATX-A, due to its high toxicity. Thus, it cannot be considered an ATX-A biosensor. The aptamer sequence was not disclosed, and the sensor performance data were not shown. Aptamers conjugated to terminal quencher fluorophore dyes are the molecular recognition elements, in a competitive-binding assay (Figure 1
). ATX-A molecules were immobilized on the surface of highly stable and fluorescent quantum dot (QD) semiconductor nanocrystal nanoshells (~2–100 nm) by covalent bonds, covered by a polyethylene glycol (PEG) coating. Once bound to the ATX-A immobilized on the QD surface, aptamers quenched QDs fluorescence. Upon exposure of this reagent mixture to free ATX, the quenching aptamers are released from the QDs surface and the fluorescence is restored, producing a signal proportional to the ATX concentration in the measured sample. The primary amine of thyroxin (T4) was covalently immobilized on the carboxyl terminated QDs that fluoresce at 655 nm (QD655, Invitrogen, Carlsbad, CA, USA) using 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride/N
-hydroxysuccinimide (EDC/NHS) chemistry. A carboxyl version of the quenching chromophore QSY21 (Invitrogen) was conjugated to the DNA aptamers synthesized with a 5′-amine termination. A Qubit™ portable fluorometer (Invitrogen) was used to acquire data, a system that suits data collection in the field or remote low-tech laboratories. Authors highlighted significant advantages provided by QDs over organic fluorophores, such as size-tunable photoluminescence spectra, higher quantum yields, broad absorption, narrow emission wavelengths [70
], and the much-increased resistance of QDs to photobleaching, that allows more reproducible and stable assays.
To produce an impedance aptasensor for ATX-A detection, Elshafey et al. [31
] used a monolayer of aptamers self-assembled on a gold electrode (Figure 2
). Aptamers were disulfide modified [HO−(CH2
−DNA]). The self-assembled monolayer (SAM) was modified by mercaptohexanol (a short alkanethiol) to displace those aptamers weakly adsorbed from the gold surface, and backfill, leaving only aptamers that were covalently bound via their thiol moieties. Aptamer binding was monitored through the [Fe(CN)6
standard redox probe. In the absence of ATX-A, the negatively-charged redox probe was repelled from the electrode’s surface and its redox reaction is hampered, thereby increasing the resistance to charge-transfer on the gold support. Upon ATX-A recognition, aptamers switched into more compact structures, enabling the contact of the redox probe to the surface, thereby decreasing the resistance to electron transfer, which was used as the sensor signal. The sensor showed a LOD of 0.5 nM, below the guideline value of 1 μg/L for water safety [72
]. The linear range of detection (LRD) was between 1 nM and 100 nM, with a dissociation constant, Kd
, of 27.14 ± 75.38 nM.
Microcystins (MCs) are monocyclic heptapeptides consisting of five constant amino acids and two variable ones e.g., MC-LR contains leucine (L) and arginine (R) at the two variable positions, and a MW of 995.171 g/mol. Four MCs (MC-LR, -RR, -LA and -YR) are on US Environmental Protection Agency (US EPA) Contaminant Candidate List III [74
] and are of special concern to the US EPA. MCs are mostly produced by cyanobacteria of the genera Microcystis
, and Planktothrix
. There are more than two hundred known MC congeners [75
] which exhibit structure variations and differ in toxicity, being the most prevalent MC-LR, also the first MC chemically identified. Apart from varying in the amino acids of the two variable positions, MCs differ on the degree of methylation, hydroxylation, epimerization, and consequently, on toxicity [76
]. MCs are hepatotoxic upon drinking contaminated water and foodstuffs, or skin contact, being also a potential tumor promoter and carcinogen. MCs inhibit protein phosphatase 1A and 2A (two key enzymes in cellular processes) to varying degrees, causing high levels of protein phosphorylation and cytoskeletal collapse. The World Health Organization (WHO) guidelines indicate 1 μg/L in drinking water and a Tolerable Daily Intake (TDI) of 0.04 μg/kg body weight per day for MC-LR [78
]. MC-LR levels lower than 1 μg/L can significantly interrupt cellular processes [80
]. These regulations and data call for simple, rapid, sensitive and reliable MC detection methods, suitable for field use.
Nakamura et al. [48
], in 2001, selected the first aptamers targeting MC-LR. It was used in an optical aptasensor based on surface plasmon resonance (SPR). Biotin-modified aptamers were attached on a sensor chip (Biacore International AB, Uppsala, Sweden) through affinity of biotin-avidin, and an SPR apparatus, BIACORE X (Biacore International AB), was used to detect the MC-LR. Despite the target molecule being MC-LR, the aptamer had higher specificity to MC-YR (Y-tyrosine; R-arginine) than to MC-LR. Authors concluded that the aromatic ring of tyrosine (YR) was more responsible for binding than the leucine’s (LR) alkyl chain. The detection range was 50 to 1000 mg/L and the affinity was low, with Kd
. The parameters obtained were lower than those of other methodologies in use at that time, namely those using antibodies and protein phosphatases; however, this work was an early demonstration of the potential use of aptamers on toxin detection.
Quite later, in 2012, Ng et al. [28
] reported the development of novel MC-targeting aptamers. One aptamer displayed high affinity and specificity to MC-LR, -YR, and -LA, simultaneously, with Kd
ranging from 28 to 60 nM. Two other aptamers were selected, one specific to MC-LR, and another simultaneously specific to MC-LA and -LR. These selectivity results implied that the selected aptamers were targeting variable regions from the MCs. Authors underline the importance of letting the variable regions of the target molecules exposed after immobilization to increase selectivity. The three aptasensors, each constructed with a different aptamer, showed differences in response for different MC analogs on a voltammetric electrochemical platform. Aptamers were self-assembled on a gold electrode via thiol chemistry and exposed to the redox marker cations [Ru-(NH3
. The cations bind electrostatically to the negatively charged phosphate backbone of the DNA of the aptamer modified surface, resulting in a reduction on the peak current in the square wave voltammetry (SWV) measurements. The decrease in current is possibly due to a change of the aptamers conformation, resulting in less [Ru-(NH3
bound to the electrode. The LRD of the aptasensors ranged from 7.5 to 12.8 pM, below the WHO (1998) drinking water maximal concentration of 1 μg/L for MC-LR, indicating the potential use of this aptasensor for water analysis. LRD was between 0.01 nM and 10 nM for the 3 aptamers.
Hu et al. [23
] characterized the adsorption capacity and selectivity of covalently immobilized RNA aptamers [81
] on graphene oxide (GO), to procedure polydisperse and stable RNA–graphene oxide (RNA-GO) nanosheets that specifically recognized and adsorbed trace MC-LR in drinking water. The nanosheets had a large surface area (2630 m2
) and very low toxicity, important features to water purification. The maximum adsorption capacity of RNA–GO was 1.44 mg/g, decreasing at extreme pH, temperature, ionic strength, and in the presence of natural organic matter. A specificity experiment using MC-LR, MC-RR, MC-LW, and nodularin, showed that more than 95% of MC-LR was absorbed by the smart RNA–GO and, less than 12% of the other toxins. Regeneration of the MC-LR loaded RNA–GO was achieved in hot pure water at 50 °C for 10 min. The adsorption capacity was reduced after five regeneration cycles by less than 10%. No sensor was developed by these authors; however, the potential of RNA-GO to purify water, concentrate toxins in biological or environmental samples and, possibly as recognition element in biosensors was shown. Covalent immobilization of RNA on GO nanosheets increases its stability. Authors proposed two possible mechanisms for this increased stability: (1) the steric hindrance could protect RNA from nucleases owing the covalent immobilization and the noncovalent π-π stacking; and (2) GO could inactivate nucleases. They concluded that the interactions of RNA–GO with RNase needed further studying.
Lin et al. [29
] have also developed an electrochemical impedance aptasensor for MC-LR detection, by immobilizing covalently the aptamers on the gold electrode through Au–S interaction. Binding of the toxin to aptamers resulted in impedance decreasing that presented a linear trend with the logarithm of the MC-LR concentration, in the range of 0.05–100 nM. LOD was 0.018 nM. The sensor showed good selectivity, tested against the lower concentration of MC-RR. High concentration of MC-RR could cause interference; however, little MC-RR interference is expected in real samples since MC-RR concentration tends to be far lower than MC-LR.
Eissa et al. [36
] designed a very sensitive and selective aptasensor for MC-LR detection which utilizes label-free DNA aptamers noncovalently assembled on graphene-modified screen-printed carbon electrodes (SPEs). The DNA aptamers assemble on the electrode giving rise to a marked reduction in the square wave voltammetric signal of the [Fe(CN)6
redox couple. In the presence of MC-LR, there was a dose-responsive rise in peak current, allowing the quantification of MC-LR. The LOD for this method was 1.9 pM. Excellent selectivity was observed when the sensor was tested against okadaic acid (OA), MC-LA and -YR, despite the aromatic rings in OA and MCs structures that could be adsorbed by π-π stacking on the graphene surface. It was demonstrated that part of the aptamer had changed conformation after coupling to the toxin, without the complete release of the bound pair from the graphene surface. Authors highlighted that the eventual release of the aptamer from graphene surface upon binding would not lead to universal detection architectures of small target molecules, such as some aquatic toxins, which are less likely to cause a change in the entire aptamer’s sequence conformation, and cause the release of the bound unit from the graphene surface. The stability of the aptamer-graphene electrodes was studied over a period of 2 weeks and no loss of signal was observed. A 2.9% decrease in response was observed after one month.
A colorimetric sensor for detection of MC-LR was developed by Wang et al. [58
] using specific aptamer sequences as linkers to prepare gold nanoparticle (AuNP) dimers (Figure 3
). AuNPs have oligonucleotides immobilized on their surface, which have complementary sequences to a part of the linker aptamer. The two DNA probes and the linker aptamer form a Y-shaped DNA duplex in the presence of the aptamer, which keeps the DNA probes linked and the AuNPs adjacent. Then, in the presence of MC-LR, the aptamers change structure causing the pre-formed AuNP dimers to disassemble into monomers and, solution color changing from blue to red. Absorbance measurements were conducted in a portable spectrometer. Nearly 5% of AuNPs surface was modified with DNA probe 1 or probe 2, and ~95% with PEG. This asymmetrical modification was to preclude the formation of large AuNP aggregates after DNA aptamers addition. Controlled formation of AuNP dimers, rather than aggregates, greatly benefited the stability of this type of sensors, which have an overwhelming sensitivity and stability. The LOD was 0.05 nM and the LRD between 0.1 nM and 250 nM. Sensitivity was investigated using MC-LA and -YR, and only MC-LR could cause a significant change in absorbance.
A dual FRET aptasensor was developed by Wu et al. [59
] to monitor simultaneously MC-LR and OA, employing green and red upconversion nanoparticles (UCNP) luminescence as the donors, and two quenchers as the acceptors (Figure 4
). Lanthanide-doped near-infrared (NIR)-to-visible UCNPs can emit strong visible luminescence with the excitation light typically at 980 nm [82
]. Biotinylated aptamers against each toxin were conjugated to different UCNPs. NaYF4: Yb, Ho UCNPs were linked with aptamers against MC-LR and NaYF4: Yb, Er/Mn UCNPs with aptamers against OA. UCNPs have many advantages over other organic fluorophores due to their optical and chemical features, such as tunability, the absence of auto-fluorescence, and the light-scattering background that can be induced by samples. Multicolor UCNP can be used to build multiplex biosensors for the detection of various analytes simultaneously [83
]. The two donor-acceptor pairs were produced by hybridizing the aptamers with their complementary DNA. The upconversion luminescence was quenched by quenchers (BHQ1 and BHQ3 from SIGMA, St. Louis, MO, USA) of high overlapping spectrum, attached to oligonucleotide probes (AATGAGGTGGTATGGGTAATTGTCATGGTGGTCCTGTTTG-BHQ3 for MC-LR and TCGTCACATTGACCCTTCGCGTAGCGCTCCCTGTTGTTGG-BHQ1 for OA). When analytes are added, the corresponding aptamers de-hybridize from the complementary DNA to bind to them, preventing the quenching of green and red luminescence. The relative luminescence intensity increased proportionally with increasing toxins concentration, allowing their quantification. The LODs for MC-LR and OA were 0.025 and 0.05 μg/L, respectively. The LRD concentration ranged from 0.1 to 50 μg/L for both toxins. Good recovery and repeatability rates (RSD) were found, 6.47% for MC-LR and 6.24% for OA. This dual FRET assay is robust to interference and has excellent specificity, as tested with DTX-1 and DTX-2 for OA aptamer, and MC-LA and MC-YR for MR-LR aptamer.
Du et al. [32
] conceived an aptasensor for quantitative MC-LR detection based on bismuth oxybromide nanoflakes (BiOBr-NF) and nitrogen co-doped graphene (NG), as a photoelectrochemical (PEC) transducer platform, in which the aptamers were immobilized on an indium tin oxide electrode (ITO) support. BiOBr (photocatalyst) promotes the photocatalytic degradation of MC-LR [58
]. In the presence of MC-LR, the aptamers immobilized on the BiOBrNFs-NG ITO electrode capture the toxin on the surface of the sensor, increasing the photocurrent response proportionally to its concentration. The observed photocurrent increase was attributed to a greater amount of MC-LR molecules captured and involved the PEC process, which were quickly oxidized by the photoinduced holes of BiOBr. The current was amplified by the retarded recombination of photoinduced electrons and holes [32
]. The LRD of the sensor was from 0.1 pM to 100 nM, with a LOD of 0.03 pM. Specificity was studied against MC-LA and MC-YR, being obtained significant photocurrent changes mostly in the presence of MC-LR. Recovery of MC-LR in fish spiked matrices ranged from 97.8 to 101.6%, with RSD of 2.52–5.14%.
The same research group [34
], has changed the detection methodology of the aptasensor developed previously [32
], to validate the effect of the aptamer-MC-LR assembled to a quartz crystal microbalance (QCM); the 2D graphene layer was replaced by a three-dimensional (3D) graphene hydrogel, co-doped with boron and nitrogen (BN-GH), self-assembled on a tris
(2,2′-bipyridine)ruthenium (II)-luminophore immobilization platform, to increase aptamers loading due to the nanoporous structure and large specific surface area of the new sensor. Three-D graphene hydrogels (GHs) can supply multidimensional electron transport pathways and be very suitable to assemble more electrochemiluminescent (ECL) molecules on its surface for enhancing its ECL intensity [86
]. The sensor is a “signal-off” switch system to the Ru(bpy)32+
−TPrA (tripropylamide) (Figure 5
). When MC-LR was incubated with the aptamers, the ECL signals detected by an electrochemiluminescence analyzer decayed intensely as the toxin/aptamer couples blocked the approaching of the coreactant TPrA to the Ru(bpy)32+
on the electrode interface. This novel ECL biosensor displayed high sensitivity and selectivity, with a LOD of 0.03 pM and a LRD ranging from 0.1 to 1000 pM.
Another electrochemical aptasensor to target MC-LR was built by Bilibana et al. [38
] research group, using a GCE with a surface of cobalt (II) salicylaldimine metallodendrimer (SDD–Co(II)) doped with silver nanoparticles (AgNPs), where 5′-thiolated DNA aptamers self-assembled (Figure 6
). Dendrimers are a class of polymers consisting of a core, self-replicating branching units, and peripheral surface groups [88
]. Dendrimers can be entirely modified with functional groups at different positions. In the case of metallodendrimers, the structure is connected to transition metal complexes. The redox reactions of the metal complex and the organic dendrimeric structure organize a molecular electronic communication, on a conductive platform that supports the transport of electrons and increases the sensitivity of the biosensor [89
]. AgNPs have a large specific surface area and the capacity to transfer photoinduced electrons at the surfaces of colloidal particles rapidly, assisting on direct electron transfer [92
]. The decrease in the peak current indicates aptamer-MC-LR complexes formation. The aptasensor showed a linear response for MC-LR detection between 0.1 μg/L and 1.1 μg/L and the LOD was 0.04 μg/L. The great sensitivity of the sensor was attributed to the AuNPs and to the nanocomposite on the surface of the GCE, which significantly increased aptamers loading. The analysis of spiked samples showed average recoveries that ranged from 94 to 115% with RSD less than 5% (n
= 3). The aptasensor was very efficient in distinguishing between MC-LR and other toxins, and between MC-LR and MC congeners. Therefore, the alkyl chain of MC-LR’s leucine seems to play a greater role on affinity than the heteroatom chain of MC-RR and the aromatic ring of tyrosine of MC-YR [38
A simple and highly sensitive colorimetric aptasensor was developed by Li et al. [49
] for the selective detection of MC-LR. In this, the plasma resonance absorption of AuNPs red peak shifts to blue upon binding of aptamers to MC-LR, at a high concentration of NaCl (Figure 7
). The random coiling of aptamers alters into a regulated structure, to form MC-LR-aptamer complexes that release AuNPs, leading to NPs aggregation and subsequent color change. This sensor exhibited a LRD ranging from 0.5 nM to 7.5 μM, and a LOD of 0.37 nM, presenting a highly selective performance in the presence of interfering substances (acetaminprid, glyphosate, dylox, atrazine, and clofentezine).
A label-free visible-light driven photoelectrochemical (PEC) aptasensor was developed by Liu et al. [52
] with high sensitivity and selectivity for MC-LR (Figure 8
). PEC sensors are ultrasensitive because of the different forms of energy for excitation (light) and detection (current). Aptamers were attached through π-π stacking interactions between the graphene hexagonal cells and the DNA nucleobases. Functionalization of vertically-aligned titanium dioxide nanotubes (TiO2
NTs) by graphene enhanced the aptasensors visible-light response activity. This sensor combined favorable characteristics of both TiO2
and graphene to produce a very high sensitivity aptasensor with a LOD as low as 0.5 fM and the LRD of the photocurrent increment from 1.0 to 500 fM. Upon MC-LR binding, there is a dissociation of the aptamer from the surface of the electrode that leads to increased photocurrent. Selectivity was tested against contaminants that may interfere with MC-LR in field samples and it was proved to be very high. Another important characteristic of this MC-LR PEC aptasensor was its photocurrent stability. The RSD of consecutive readings, as the excitation light was turned on and off repeatedly, was only 0.3% (n
= 6). The outstanding stability and repeatability are largely related to the steady loading of aptamers on the functionalized electrode surface and the potential confining effects provided by the special 3-D nanotubular structure of TiO2
Lv et al. [53
] build another ultrasensitive optical aptasensor, this time using the enhanced fluorescence of lanthanide ions doped core/shell upconversion nanoparticles (CS-UCNPs) and MoS2
). NaYF4 nanoparticles, co-doped with Yb and Tm, were first produced and used as seeds for the epitaxial growth of NaYF4: Yb shells to form NaYF4: Yb,[email protected]
: Yb core/shell NPs. Two-dimension MoS2
nanosheets have a structure analogous to that of graphene with exceptional nanoelectronics, optoelectronics, and photovoltaics characteristics [84
], showing high fluorescence quenching capacity and specific affinity to ssDNA [97
], characteristics that are very convenient to biosensors. Polyacrylic acid functionalized CS-UCNPs were activated via EDC/sulfo-NHS chemistry, then avidin was added to form avidin-conjugated CS-UCNPs, and at continuation combined with biotin-aptamers. MoS2
adsorbed the aptamer-modified CS-UCNPs through van der Waals forces between the nucleobases and the basal plane of MoS2
, causing energy transfer to the MoS2
, and quenching of the fluorescence. After adding MC-LR, the aptamer conjugated preferentially with it, changing conformation, detaching from MoS2
, and allowing for fluorescence recover. Upconversion fluorescence spectra of NaYF4: Yb,[email protected]
: Yb core/shell NPs occurred at 980 nm, and MC-LR quantitative analysis was performed by measuring fluorescence response at 361 nm. LRD ranged between 0.01 μg/L and 50 μg/L, and the LOD was 0.002 μg/L.
A novel ultrasensitive fluorescent aptasensor was developed by Taghdisi et al. [35
] (Figure 10
) for detection of MC-LR, using single-walled carbon nanotubes (SWNTs), dapoxyl dye and an aptamer (DAP-10) with affinity to dapoxyl dye. SWNTs were used to immobilize DAP-10 (5′-CAATTACGGGGGAGGGTGTGTGGTCTTGCTTGGTTCGTATTG-3′) and an unmodified MC-LR aptamer (Apt) that was used as a sensing ligand. Significantly different fluorescence intensity was measured in the presence and absence MC-LR. Without MC-LR, DAP-10 could not adsorb to the surface of SWNTs because it was loaded with Apt and no space was available. After adding dapoxyl to the supernatant and have the sample centrifuged, it becomes strongly fluorescent. When dapoxyl is attached to DAP-10 its fluorescence increases significantly because the surrounding gets less polar as compared to water. When the Apt binds to MC-LR it leaves free space on SWNTs surface for DAP-10 aptamer to attach, resulting in very weak fluorescence intensity of the sample after addition of dapoxyl. Apt was used without any modification, which increased the sensitivity of the sensor. The LOD was 138 pM (0.137 μg/L) and the LRD was from 0.4 to 1200 nM.