Introduction

In medicolegal investigations of cyanide poisoning, accurate and reliable cyanide concentrations are rarely achieved. The half-life of cyanide and its primary metabolite, thiocyanate, increase under chronic exposures1,2,3,4,5, and lead to the depletion of sulfur donors5. The diminished sulfur donors impede the thiocyanate detoxification process, resulting in an elevated cyanide levels, and enzymatic activity that remains active after prolonged exposure6,7. Conversely, the duration of acute cyanide exposure was considerably shorter, preventing the detection of the toxic parent compound. Despite acute or chronic exposure, the detection of cyanide is challenging because of its volatile and nucleophilic nature8. Therefore, most studies have focused on thiocyanate because of its longer half-life9,10. However, the notable variability of thiocyanate in preexisting biological specimens necessitates the consideration of a stable metabolite, such as 2-amino-2-thiazoline-4-carboxylic acid (ATCA), as a potential alternative biomarker for cyanide determination, even in the absence of the parent compound11,12.

ATCA, whose concentration is directly correlated with that of the parent molecule undergoing metabolism, remained stable for several months when stored in a freezer13. The direct correlation between ATCA and cyanide may help minimize the limiting factors associated with the parent compound, such as concentration fluctuations depending on storage conditions (time and temperature) and contamination by cyanogenic bacteria. In postmortem examinations, carcasses may be exposed to environmental cyanide absorption and diffusion during sample collection and separation14, resulting in variations in the total cyanide concentration15. These factors contributing to variations in the cyanide concentration may lead to false-positive or false-negative conclusions13.

Therefore, the use of ATCA as a biomarker can minimize uncertainties and confirm the presence of cyanide16, especially in poisoning cases in which the decomposition of cadavers and the closure of detectability windows may compromise the reliability of the cyanide concentration for toxicity determination in the involved specimens17. Although the forensic scientific community has not yet embraced the use of alternative biomarkers, such as thiocyanate, ATCA, and cyanide–protein adducts, the ATCA metabolite may serve as evidence for verifying cyanide intoxication5.

Cyanide is distributed in red blood cells, serum, and plasma13. Therefore, Jackson and Logue43 developed a portable and efficient diagnostic sensor based on fluorescent detection. This sensor has been thoroughly tested and proven to have a 100% accuracy rate in detecting blood cyanide in a rabbit model18. However, ongoing debates have raised concerns about the stability of cyanide in postmortem samples19. Elevated blood cyanide levels were observed in the blood of victims who had been burned after death, as reported in homicide cases15. The instability of cyanide complicates the determination of blood cyanide levels and exposure extent17.

Further research into the applicability of ATCA in medicolegal investigations of cyanide poisoning is required because no current forensic toxicology test has been used to confirm the correct amount of cyanide in poisoning cases13. Consequently, this study introduces an aptamer with high selectivity and specificity for ATCA. The aptamer was selected as the receptor because, unlike antibodies, it is extremely stable across a broad temperature range and can revert to its original form even after several selection cycles of denaturation followed by renaturation20. In addition, aptamers are more feasible to produce than antibodies because they are chemically synthesized, whereas antibodies require colony screening using mammalian cell culture, leading to variations from batch to batch21. Furthermore, synthetic aptamers can be chemically modified to enhance their stability and make them resistant to nucleases22,23. Moreover, unlike highly immunogenic antibodies, aptamers do not provoke the human immune response and are thus not perceived as foreign materials by the human immune system24. Aptamers can recognize various targets including ions and small molecules that cannot be recognized by antibodies; thus, aptamers have wider applications as biosensors than antibodies23.

Herein, we present an unexplored strategy for detecting cyanide poisoning by combining the merits of graphene oxide (GO) as immobilization matrixes and ATCA as a specific biomarker on a disposable screen-printed carbon electrode (SPCE)-based aptasensor. In addition to the high surface area of this 2-dimensional nanomaterial, GO possesses an advantage for the attachment of biological recognition elements owing to its versatile oxygen functional groups. For this purpose, the aptamer was screened through GO-SELEX from a single-stranded DNA random library and subsequently functionalized onto GO-modified SPCE to develop a voltammetric aptasensor. In clinical and forensic diagnostics, laboratory testing entails initial screening, followed by a confirmatory testing25. The proposed ATCA aptasensor can enhance the cyanide detection accuracy by reducing the false-positive and false-negative results in cases of cyanide intoxication. Stable biomarkers are crucial for detecting cyanide, as opposed to cyanide itself16. Notably, this simple and cost-effective method is efficient for screening ATCA in real sample applications, including blood, serum, and urine, without the need for advanced equipment or user expertise in forensic analysis.

Results and discussion

Aptamer selection by GO-SELEX

A high-throughput screening of aptamers with strong binding and selectivity for ATCA was conducted using a random library containing 438933.3 ng of ssDNA. Graphene oxide (GO) was used as a ssDNA carrier for the iterative selection. Before selection, both aptamers and GO underwent mass ratio optimization to determine the appropriate volume for the random library and sub-library incubation26, as shown in Fig. 1a. The ideal mass ratio of GO to single-stranded DNA (ssDNA) was identified from the distinct binding kinetics seen with varying lengths of ssDNA and GO26. The random library consisted of n = 30 random nucleobases with 18-nt for both forward and reverse primers. Therefore, optimizing the mass ratio of GO/ssDNA was crucial to ensure the 66-nt ssDNA was completely adsorbed on the GO. The optimum mass ratio was determined from the fluorescent intensity of different mass ratios of GO/ssDNA. The fluorescent intensity decreased from 1567 to 38 dfu as the mass ratio increased from 0:1 to 25:1. Hence, the optimum GO volume for 1 µL of ssDNA was 25 µL.

Fig.1
figure 1

(a) Round selections of ssDNA aptamers for ATCA via GO-SELEX. (b) GO/ssDNA mass ratio fluorescent intensity (dfu).

Full size image

The aptamer selection cycle started with the incubation of the 1 µL of ssDNA library with 25 µL of GO, followed by the incubation with a group of counter targets to minimize the aptamer candidates. Subsequently, the random aptamers unbound to counter targets were incubated with ATCA. The aptamers that were selective to the counter targets were discarded after centrifugation by removing the supernatant. After incubation with both targets and counter targets, the purified pool of ssDNA concentration was analyzed using a nanodrop bio spectrometer. The ssDNA was amplified, forming dsDNA, and after dsDNA purification, the dsDNA concentration was analyzed via a nanodrop bio spectrometer. The dsDNA was denatured by Urea-PAGE analysis to regenerate the ssDNA for the subsequent round of selection. The concentration of purified ssDNA after Urea–PAGE was analyzed using a nanodrop bio spectrometer, and the volume of the purified ssDNA was used to determine the ssDNA amount (ng). The same steps were repeated with the pooled ssDNA before PCR reached saturation state. In this study, the percentage of ssDNA recovery decreased after the sixth round of aptamer selection. Figure 1b shows the pooled ssDNA recovery percentage obtained from the seventh round of aptamer selection. The seventh-round aptamer selection reached saturation, in which the pooled ssDNA recovery was lower (95.6%) than in the previous round (96.4%). Hence, the ssDNA from the sixth round was amplified and sequenced for DNA cloning.

Twenty colonies were selected for DNA extraction and amplification. Amplicons were analyzed using gel electrophoresis. The 17 visible DNA bands were selected and subcultured into individual broths. Plasmid DNA from the colonies was extracted and sent for sequencing. Sequence candidates were analyzed for M-fold free energies and grouped into respective phylogenetic groups. A molecular docking analysis was performed for all aptamer candidates with the target ATCA. The two aptamers with the highest binding affinity were Apt20 and Apt46. These aptamers were selected for biomolecular-binding interaction characterization using isothermal titration calorimetry (ITC). The selectivity of the two selected aptamers was further assessed by molecular docking with the counter targets. According to binding affinity in molecular docking, both aptamers showed lower affinity for the counter targets. Figure 2a illustrates the molecular docking binding affinities of Apt20 and Apt46 to ATCA and the counter targets. Figure 2b shows the Apt20 (ΔG: −7.87 kcal/mol) and Apt46 (ΔG: −10.78 kcal/mol) M-fold structures.

Fig.2
figure 2

(a) In-silico binding affinity (kcal/mol) of APT20 and APT46 for various targets. (b) M-fold structures for (i) APT20 and (ii) Apt46. (c) Isothermal titration calorimetry (ITC) measurements of (i) Apt20 binding interaction with ATCA and blank (no metabolite as a control), and (ii) Apt20-ATCA using injection method. (d) ITC measurements of (i) Apt46 binding interactions with ATCA and blank, and (ii) Apt46-ATCA using injection method.

Full size image

The ITC is a label-free method that measures the heat released or adsorbed from the direct interaction between the aptamer and the ligand27. This instrument allows precise measurements of the disassociation constant (Kd), binding stoichiometry, and thermodynamic parameters of the interactions between two molecules28. In this analysis, the ATCA was titrated with Apt20 and Apt46. According to the thermogram displayed in Fig. 2c,d, both binding aptamers exhibited favorable binding interactions27. This is supported by the observed change in enthalpy, ΔH = −100 kJ/mol. When comparing the disassociation constant (Kd) values, it was found that Apt46 (Kd: 5.02 × 10−9 M) had a lower value than Apt20 (Kd: 3.87 × 10−5 M). This indicates that Apt46 has a higher affinity for ATCA than Apt20. Thus, Apt46 was selected for further use in ATCA aptasensor development. Blank samples represent interactions between the phosphate-buffered saline solution and Apt20 and Apt46.

Characterization of ATCA aptasensor

The ATCA aptasensor utilized the aptamer sequence Apt46, which was labeled with an amino tag (5′/5AmMC6/ CGT ACG GAA TTC GCT AGC CAA CGC TAG TGT GGT GGA TTA ACG AGT ACC CAC GTG GAG CTC GGA TCC-3′). This specific sequence demonstrated the highest selectivity and affinity for the target. The functionalization of Apt46 was carried out on SPCE.

Prior to that, the surface of the SPCE was modified with GO activated with EDC − NHS crosslinker. GO has functional groups with oxygen on its surface, resulting in the reduction of Π–Π stacking and deformation of the lattice structure that allows GO to interact with solvents by forming hydrogen bonds between the hydroxyl group on the GO sheet and any compounds present within the solvents29. Exploiting the existence of oxygen functional groups (= O, − OH, − O − , and − COOH) on the bilateral side of the sheets, including the borders of the plane30, it readily interacts with EDC − NHS, as illustrated in Fig. 3. In this coupling reaction, EDC reacts with the carboxyl group of GO to form an unstable, reactive O-acylisourea ester. Sulfo-NHS reacts with the labile acylisourea ester to produce a slightly stable amine-reactive NHS ester31. When the aptamer is immobilized onto the SPCE, Apt46 reacts with the amine-reactive NHS ester. This approach successfully linked the carboxyl group of GO with the amino-labeled Apt46 using the EDC–NHS crosslinker. The assay’s subsequent steps inhibited nonspecific binding by blocking with ethanolamine. After ATCA incubation, the immobilized Apt46 captures the target analyte. The Apt46–ATCA complex conformation facilitates electron transport between the redox probe and the bulk solution.

Fig.3
figure 3

Scheme of the functionalization of graphene oxide (GO) with ssDNA: (a) surface chemistries of the GO sheet, including functional groups (= O, − OH, − O − , and − COOH), (b) formation of an O-acylisourea active intermediate via the 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) crosslinker, (c) surface stabilization of EDC–NHS/GO through the addition of sulfo–NHS, (d) immobilization of Amino-Apt46 on the EDC–NHS/GO/SPCE modified surface, (e) blocking of the unreacted active sites using ethanolamine, and (f) detection of the target ATCA on the immobilized electrode surface.

Full size image

GO has been utilized as a DNA carrier in the GO-SELEX analysis. GO was preferred for aptamer functionalization due to its ability to interact with ssDNA instead of dsDNA. The ssDNA is a non-helical structure, while the dsDNA has a double helix structure that hinders direct interaction between graphene oxide and nucleobases32,33. The SPCE modified with GO activated EDC–NHS was characterized using FESEM/EDX analysis, as illustrated in Fig. 4. The GO-modified electrode exhibited a large sheet-like morphology with crumpled features, in contrast to the image of the bare SPCE with a rough surface structure, thereby confirming the presence of EDC–NHS/GO. Additionally, EDX analysis revealed a predominant peak corresponding to carbon in the bare SPCE, whereas GO/SPCE showed a dominant oxygen peak, confirming the successful attachment of GO to the SPCE surface. The weight percentage of oxygen in the GO-modified electrode was approximately 7% higher than in the unmodified SPCE, primarily attributed to the formation of oxygen-containing functional groups via GO sheet activation with EDC–NHS. In the EDC–NHS coupling reaction, EDC reacts with carboxyl groups on the GO surface to form a partially unstable acylisourea ester. Subsequently, sulfo-NHS was combined with the ester to form a partially stable NHS ester capable of reacting with amines. The NHS ester was then reacted with the amino-labeled aptamer to form a stable amide bond, as illustrated in Fig. 3.

Fig.4
figure 4

Morphological and elemental analysis using FESEM/EDX for (a) bare SPCE and (b) GO-modified SPCE at scale of 200 nm.

Full size image

The SPCE surface modification of the aptasensor was examined using cyclic voltammetry (CV), as depicted in Fig. 5a. CV analysis (scan rate:100 mVs−1) was performed for each layer of electrode surface modification from the bare electrode to the target ATCA immobilization. The bare SPCE exhibited the minimum changes in peak current due to the disabled access of the redox probe [Fe (CN)6]4−/3− to the electrode surface. After modification with EDC − NHS-activated GO, the peak current increased as the neutrally charged NHS allowed the passage of the redox probe to the electrode surface. Subsequently, after immobilization of the aptamer, a substantial decrease in peak current was observed because of an increase in the charge transfer resistance caused by the negatively charged Apt46 backbone, which obstructs the negative redox probe from reaching the active site for electron transfer. The peak current slightly increased after the ethanolamine blocking step because of the blockage of the negatively charged and unreacted carboxyl group, which allowed the negative redox probe to access the electrode active site. After incubation with ATCA, the aptamer bound to the small-molecule analyte and formed a hairpin structure that increased the electroactive surface area of the redox probe34,35.

Fig.5
figure 5

(a) Cyclic voltammogram of bare SPCE and different modified electrodes. (b)-(c) Nyquist plots of the stepwise fabrication process of Apt46-based GO-modified SPCE for ATCA detection in 1 mM [Fe (CN)6]4−/3− solution containing 1 × binding buffer (1 × BB), pH 7.6. The Apt46 and ATCA concentrations were 0.3 µM and 0.35 µg/mL, respectively (Inset is an equivalent Randles circuit: Rs = resistance between the working and reference electrodes, Zw = Warburg impedance, Cdl = double-layer capacitance, and Rct = charge-transfer resistance). (d) ATCA-Apt46 three-dimensional molecular docking conformation.

Full size image

The Apt46 aptasensor was also developed by electrochemical impedance spectroscopy (EIS), as shown in Fig. 5b. In this characterization, the magnitude of the electron transfer resistance (Rct) was determined from the diameter of the Nyquist plot semicircle. Before the modification with activated GO, the diameter of the bare SPCE semicircle was large because of the high electron transfer resistance at the active site of the working electrode. After the electrode was modified with activated GO, the electron transfer resistance was dramatically decreased because of the activation of the carboxyl group on the surface of the GO by EDC − NHS, which induced the alleviation of electrostatic hindrance from the electronegative carboxyl group35. Besides, the decrease in Rct could be attributed to the replacement of the carboxyl group with the neutral NHS ester, which allowed the transfer of a negative probe [Fe(CN)6]4−/3− on the surface of the electrode34.

Subsequently, when the Apt46-binding aptamer was immobilized on the electrode surface, Rct increased because of the presence of negative charges on the aptamer phosphate backbone, which induced repulsion toward the negatively charged redox probe. Following the blocking of the unreacted carboxyl group with ethanolamine, the negatively charged carboxyl group on the surface of the electrode reduced the resistance of the negatively charged electrode probe from reaching the electrode surface. After incubation with the target ATCA, the value of Rct was further reduced. The reduction in electron transfer resistance was presumed to be due to the aptamer enclosing the small-molecule ATCA to form a hairpin structure36, as depicted in Fig. 5d, and opening the access for the redox probe to the electrode surface.

Randles circuit fittings were conducted for each alteration layer to further examine the actual charge transfer resistance. Bare SPCE showed high impedance in EIS (2.6 × 103 Ω) due to the low conductivity of the carbon-based material. After modification with GO, the charge transfer resistance decreased to 24.7 Ω. Upon modification with the Apt46-binding aptamer, the charge transfer resistance increased to 40.4 Ω and shifted to 32.1 Ω after the immobilization of ethanolamine. The final step modification with target ATCA reduced Rct to 27.0 Ω, as listed in Table 1. The Nyquist plot overlay for the surface modification is shown in the inset of Fig. 5c.

Table 1 Randles circuit fittings.
Full size table

Analytical performance of ATCA aptasensor

The development of the ATCA aptasensor was optimized by differential pulse voltammetry (DPV). The optimized parameters included the Apt46 ligand and ethanolamine concentrations, as well as aptamer, ethanolamine, and target ATCA incubation times (Fig. 6). The method was adapted from Aye and group37. The ATCA concentration was consistently set at 0.35 µg/mL throughout the optimization process. The optimization was performed with Apt46 concentration from 0.1 to 1.0 µM.

Fig.6
figure 6

(a) Effect of Apt46 (i) concentration and (ii) incubation time; (b) Effect of Ethanolamine (i) concentration and (ii) incubation time; (c) Effect of ATCA incubation time.

Full size image

The optimum Apt46 concentration was determined to be 0.3 µM, while 15 min was determined to be the best duration for aptamer incubation to the ligand. The electrode surface was rinsed with 100 µL of binding buffer (1 × BB) for every modification layer. The subsequent optimization was the concentration of the blocking agent, ethanolamine. The blocking agent was immobilized after aptamer modification to avoid nonspecific binding on the surface of the GO/SPCE, with other than the aptamer probe to increase the aptamer efficiency38. The optimized ethanolamine concentration and time for incubation were 0.01 M and 30 min, respectively. The ATCA incubation time was also optimized to determine the most effective duration for complete interaction between the aptamer and the analyte prior to DPV measurement.

A calibration plot was developed using the optimized parameters. The ATCA was added to the binding buffer (1 × BB) at concentrations ranging from 0.05 to 1.5 µg/mL. The changes in current (ΔI) response increased as the concentration of the spiked analyte increased, as depicted in Figs. 7a ,b. The aptamer characterisation using CV and EIS confirmed that the current increased while the charge resistance decreased when the ATCA was immobilized on the modified electrode.

Fig.7
figure 7

(a) DPV response of ATCA-Buffer (1 × BB). (b) The calibration curve of the current changes corresponding to the ATCA-Buffer concentration range of 0.05 to 1.5 µg/mL. (c) ATCA aptasensor cross-reactivity analysis against target and counter targets. (d) Apt46 DPV response stability standard deviation range.

Full size image

The performance of the ATCA aptasensor to discriminate its GO-SELEX counters targets was examined by performing a cross-reactivity analysis with 2-Aminothiazole-4-carboxylic acid (ATZA), 2-Methylthiazole-4-carboxylic acid (MTCA), (4R) -(-)-2-Thioxo-4-thiazolinecarboxylic acid (TTCA ), and N-Carbamoyl—L- Cysteine (NLCC) as shown in Fig. 7c. Based on the findings, the ATCA aptasensor could recognize the same counter-targets used in the aptamer screening cycles. The DPV response towards the counter targets was consistent with the virtual binding affinity screening against both target and counter targets, as presented in Fig. 2a.

Furthermore, the stability of the ATCA aptasensor was analyzed in detail. Five sets of Apt46/GO/SPCE triplicate samples were prepared and monitored for 50 days, as shown in Fig. 7d. The samples were deposited on days 1, 20, 30, 40, and 50. Apt46/GO/SPCE was placed in a petri dish and sealed with parafilm. All the samples were stored at a temperature of 4 ℃ from day 1 to day 50. The stability of ATCA was conducted in the binding buffer matrix. The system remained stable for 50 days with a mean recovery of spiked ATCA (0.90 µg/mL) within ± 3SD as shown in Table 2.

Table 2 Stability of developed ATCA aptasensor.
Full size table

Application of ATCA aptasensor in biological matrices

The applicability of the ATCA aptasensor to biological samples was evaluated by developing calibration plots for the individual matrices (Fig. 8), i.e., blood, serum, and urine, followed by conducting repeatability analysis. ATCA was spiked in different matrices and analyzed in triplicates (n = 3) on different days. Table 3 represents the DPV response recovery of the toxic level of ATCA (0.90 µg/mL). The percent recoveries were obtained from the ratio of the mean from inter-day responses against the DPV response of the mean of 0.9 µg/mL from the calibration curve. Based on the repeatability analysis, a good recovery percentage ranging from 92.29 to 114.22% was obtained. Moreover, the RSD values for the samples (n = 3) were below 5%.

Fig.8
figure 8

ATCA–DPV responses for (a) blood spiked, (b) urine spiked, and (c) serum spiked.

Full size image
Table 3 ATCA aptasensor spike recovery in both buffer and biological matrices.
Full size table

To date, analytical methods for detecting ATCA mainly use gas and liquid chromatography. In gas spectrometry-mass spectrometry (GCMS) analysis, the derivatization process is required because ATCA is a highly polar and non–volatile compound39,40. The derivatizing agent, MSTFA, reacts with the primary, secondary amine, and carboxylic acid moieties of ATCA. After derivatization, ATCA becomes more volatile and is readily injected into GCMS39. The polar ATCA compound is separated in liquid chromatography using a hydrophilic interaction chromatography column and identified using tandem mass spectrometry (LC/MS/MS). Derivatizing agents, such as S–NIFE and DNS, are employed in liquid chromatography to purify post-mortem blood. It was noted that the interference reduced after cleaning up with S-NIFE, and the analyte peak shape improved41. GCMS and LCMS instrumental methods for ATCA analysis need a derivatizing process, making it costly, laborious, and require a longer analysis time42,43. The available ATCA analytical instrumental methods are listed in Table 4. The detection limit for this study is comparable to that of the previous research. The baseline level of ATCA in human blood was reported to be approximately 0.15 µg/mL41.

Table 4 Methods for determination of ATCA.
Full size table

Conclusion

This research focused on screening for an aptamer with the highest affinity for 2-amino-2-thiazoline-4-carboxylic acid (ATCA), the secondary metabolite of cyanide, using graphene-oxide SELEX. To estimate the specificity of the random aptamer for ATCA, a random ssDNA library was incubated with structurally similar counter-targets before the actual target to minimize nonspecific candidates. Among these counter-targets, 2-aminothiazole-4-carboxylic acid closely resembled ATCA but differed by having two double bonds in the ring structure, compared with the ATCA's single double bond. The binding aptamer Apt46 demonstrated higher specificity toward ATCA over Apt20, with the dissociation constants of 5.02 × 10−9 M and 3.87 × 10−5 M, respectively. Consequently, amino-labeled Apt46 was functionalized on the carbon electrode previously modified with an EDC–NHS linker to develop an ATCA aptasensor. This aptasensor demonstrated exceptional affinity for ATCA compared with selected counter-targets and remained stable for up to 50 days, with deviations in the current response being limited to ± 3SD. It could detect ATCA directly in various biological samples, including urine, serum, and whole blood. Thus, this aptasensor can be used to detect the cyanide metabolite in investigations related to cyanide intoxication.

Methods

Materials and reagents

(4R)-(–)-2-Thioxo-4-thiazolinecarboxylic acid, 2-Aminothiazole-4-carboxylic acid, 2-Methylthiazole-4-carboxylic acid, acrylamide, ammonium persulfate, ampicillin, calcium chloride, ethanol, N-Carbamoyl-L-cysteine, trizma base, N′-Tetramethyl ethylenediamine (Temed), hydrochloric acid, phosphate buffered saline (PBS) tablet, and sodium acetate buffer solution (pH 5.2) were purchased from Sigma-Aldrich, USA. Formamide, potassium hexacyanoferrate (II) trihydrate were purchased from Merk, Malaysia. Graphene oxide in water suspension (4 mg/mL) was purchased from Graphenea, USA. Potassium chloride and sodium hydroxide were obtained from R&M Chemicals, India. Magnesium chloride and potassium ferricyanide were purchased from Bendosen, Malaysia. 10X Tris–Borate-EDTA (TBE) buffer (pH 8.3), PrimeWay plasmid DNA extraction kit, and agarose were purchased from 1st BASE Biochemicals, Malaysia. Glycogen Invitrogen™, O’RangeRuler 10 bp DNA ladder, TOPO ® TA cloning® kit, and UltraPure™ 0.5 M EDTA (H 8.0) were obtained from Thermofisher Scientific, USA. GelPilot DNA loading dye was purchased from Qiagen, Malaysia. APTENS Screen Printed Sensors (Model SPE-RL3-05-P, Amperometric-Type 3 Channel WE (5 mm dia.)/RE/CE was purchased from Biogenes Technologies Sdn Bhd.

Optimization of GO/ssDNA

The ssDNA library used in this study consists of n = 30 random nucleobases with 18-nt for both forward and reverse primers. The optimal mass ratio of GO/ssDNA was determined because different lengths of ssDNA and GO showed dissimilar binding kinetics characteristics. Therefore, optimizing the mass ratio of GO/ssDNA is crucial to ensure the 66-nt ssDNA is completely adsorbed on the GO. GO solutions with different amounts were added to 10.5 uL of 10 µM 66-nt ssDNA labeled with 56-FAM (equivalent to 2 µg) (the 10 µM ssDNA was prepared in sterile deionized water). The binding buffer (1 × BB) was used as a solvent to determine the mass ratio GO/ssDNA. The mixture was gently mixed and incubated at room temperature with rotation for 1 h and centrifuged at 12,000 rpm for 15 min. The supernatants were collected for fluorescence intensity measurement using a Qubit R 2.0 Fluorometer.

In vitro selection of ATCA aptamers

The GO-SELEX procedure is adapted from N.H. Ahmad Raston et al.44. Heat shock analysis was performed by heating 200 pmol of the random library (5′-C GTA CGG AAT TCG CTA GC-(N30)-CA CGT GGA GCT CGG ATC C-3′) at 95 ℃ for 15 min and then cooling on ice flakes for 5 min on ice allowing the oligonucleotide to form the best structure. For every aptamer screening cycle, the denatured ssDNA was incubated on a rotator for 30 min with the ATCA counter targets; 2-Aminothiazole-4-carboxylic acid (ATZA), 2-Methylthiazole-4-carboxylic acid (MTCA), N-Carbamoyl—L- Cysteine (NLCC), and (4R) -(-)-2-Thioxo-4-thiazolinecarboxylic acid (TTCA) in the 1 × binding buffer (1 × BB) (100 mM NaCl, 20 mM Tris–HCl, 2 mM MgCl2, 5 mM KCl, 1 mM CaCl2, pH 7.6) to remove the ssDNA that have higher affinity for the counter targets. To remove the aptamer bound to the counter targets, 25 µL of 4 mg/mL GO was added (final volume 1 mL), and the incubation continued for 2 h.

The aptamer bound explicitly to the counter targets was separated by centrifugation at 1500 g rpm for 10 min, and the supernatant was discarded. The GO pellet left was added to the ATCA. The following 2 h of incubation aimed to allow the aptamer pools to form conformation with the target, and the binding complex was separated by centrifugation at 15000 g for 15 min. In this step, the supernatant was collected and re-centrifuged at the same setting as before to ensure the final supernatant was free from traces of GO. The GO-free supernatant proceeded with ethanol precipitation, and the ssDNA recovery was analyzed using a nanodrop spectrometer. The pooled ssDNA obtained was amplified via PCR using forward primer (5′-/56-FAM/CGTACGGAATTCGCTAGC-3′) and reversed primer (5′-(A20) /iSp18/ G GAT CCG AGC TCC ACG TG-3′). The ssDNA was recovered again by Urea-PAGE separation and purification. The denatured DNA was used as a random ssDNA pool for the following cycles and continued until the plateau was reached.

Cloning and sequencing of selected DNA

The ssDNA from the sixth round was amplified for DNA cloning using TOPO® TA cloning® Kit. The transformed cells (50 µL) were spread on the pre-warmed LB plates containing 50 µL Ampicillin and 40 µL of 40 mg/mL Xgal and incubated overnight. The white colonies were determined, picked using an inoculation loop, and spread on a new LB + Amp plate. Each colony picked from the incubated plates was streaked and incubated overnight. Only 20 colonies were randomly chosen for DNA extraction and underwent colony PCR analysis. The ssDNA was amplified using M13 Forward (−20) Primer and M13 Reverse Primer. The colony PCR products were analyzed using agarose gel electrophoresis, and the sample with a visible DNA band was prepared for plasmid DNA extraction analysis.

The plasmid DNA pellets were extracted using a Prime Way Plasmid Extraction Kit. The recovered Plasmid DNA concentrations were recorded using a biospectrometer and sent for sequencing. The sequencing results were analyzed using Applied Biosystems Sequence Scanner Software 2.0. The multiple sequence results were aligned using the Clustal Omega Program, and the aptamer was synthesized using the Mfold web server for nucleic acid folding and hybridization prediction. The folded DNA structure resulted in a circular plot (2D sequence) with a Gibbs free energy value in kcal/mol.

Molecular docking analysis

The molecular docking was performed via Autodock Vina Tool 1.5.6, MGL Tool 1.5.6, CADD 1.5.6, and Vision 1.5.6. This analysis was performed in Windows operating systems. The analytical procedure was adapted from the Simplified Learning Protocol by Biogenes Technologies Sdn Bhd. In addition, PyMOL tool was used for docking data analysis. The target (ATCA) and counter targets (ATZA, MTCA, NLCC, and TTCA) 3D conformers were downloaded from the PubChem website in sdf format and converted to pdb format using the Avogadro tool. The aptamers were obtained from the previous analysis and further analyzed for the docking analysis. The docking analysis aimed to reduce the number of potential aptamers. Initially, each aptamer was docked to the target to identify the two conformations with the highest affinity. This was followed by testing the selectivity conformation against the counter-targets.

Isothermal titration calorimetry (ITC) study

The virtual selection of aptamers was conducted using molecular docking analysis. Label free molecular binding interaction analysis was performed using an isothermal titration calorimeter to further study the aptamers. Specifically, two aptamers, Apt20 and Apt46, were titrated against the ATCA for affinity and specificity analysis. Blank samples were also tested for potential heat from buffer and ligand interaction. In the previous section, these counter-targets had already been docked against Apt20 and Apt46. The ligands (target and counter targets) and the macromolecule (aptamers) were prepared at 20 µM and 4 µM concentrations, respectively. The buffer used in this analysis was Phosphate Buffer Saline (PBS) solution, PH 7.6, with additional salt, MgCl2, to facilitate aptamers folding45. The salt concentration was similar to the GO-SELEX binding buffer, 2 mM MgCl2.6H2O.

Electrochemical detection of ATCA

The methodology in this research was adapted from Aye et al.37. The oxygen functional group-rich graphene oxide was immobilized onto the SPCE to improve the electrode's conductivity. The 10 mL of 400 µM EDC–NHS was added to 0.1 g of GO powder to form a crosslinker NHS ester to react with the primary amine group of the Apt46. The mixture was mixed and stirred for 4 h at room temperature, followed by 10 min of sonication. The activated EDC–NHS/GO was extracted by centrifuging at 10,000 rpm for 25 min. The resulting supernatant was discarded, and the excess EDC–NHS was washed away by rinsing it with deionized water (DIW). The activated solid GO was left to dry at 30 ℃ for 24 h.

A working suspension with a concentration of 0.5 mg/mL was prepared by adding 0.5 mg of the dried GO into 1 mL of DIW. The 7 µL of 0.5 mg/mL GO was dropped on the electrode and dried at 50 ℃ for 15 min. The excess GO was washed with 100 µL of binding buffer (1 × BB). The activated electrode was air-dried at room temperature. The dried SPCE and the bare SPCE were characterized via FESEM/EDX. The electrodes were further characterized by CV scanning rate at 0.01 – 0.1 V. Next, the SPCE was further characterized by depositing the subsequent layers of modification, i.e., 0.3 µM Apt46 followed by 0.35 µg/mL of ATCA. Every modification layer was washed with 100 µL 1 × BB and dried at room temperature. The SPCE surface was scanned at each level of deposition to observe the effects of modification. The amount of ATCA was taken from the toxic level of the ATCA. The redox probe used in this study was 5 mM [Fe (CN)6]4−/3−and 0.1 M KCl.

The electrochemical circuits for every level of modifications were analysed with EIS to perform the electrical circuit fitting for the surface modified SPCE model. The frequency (HZ) was set in the range of 0.01 – 1E + 05, and the amplitude (VRMS) was 0.05. The DPV method was used to optimize the modification of the electrode surface. The analytes were prepared in the GO-SELEX binding buffer, and Apt46 was prepared in sterilized-filtered DIW. The potential applied was in the range of −0.1 V to + 0.3 V with a pulse amplitude of 25 mV. The optimized parameters were Apt46 ligand concentration, ethanolamine concentration, aptamer incubation time, ethanolamine incubation time, and target-ATCA incubation time. The target and counter targets were spiked into the binding buffer and biological samples. The biological samples were human urine, serum, and whole blood. The recovery analysis was performed for two days in the binding buffer and three days in the biological samples. The stability analysis was conducted at a lethal dose of ATCA (0.90 µg/mL) in the binding buffer matrix. The electrode (Ethanolamine/Apt46/GO/SPCE) was prepared on the first day and the triplicate samples were kept in the fridge at 4 ℃ until use within 50 days.