Phosphate buffer saline (PBS) 1X (pH 7.4), bovine serum albumin (BSA), and TRIS Buffer (50 mM Tris–HCl, 2 mM MgCl2, 150 mM NaCl pH 7.5) were purchased from Sigma-Aldrich (Milan, Italy).

The aptamer AP_7462DNA, 5′-GCGGCGCGGTATGGAATTAGTGACCTTCCGCGCGCCCCATTTTTTATAGGGGCCGC-3′, was synthetized by Metabiom international AG (Planegg, Germany) and used for the electrochemical tests.

The aptamer with the addition at 5′ end of C6 S–S-polyT and a 6-Carboxytetramethylrhodamine (TAMRA) at 3′ end (named aptamer-TAMRA) was purchased from GenScript Biotech (Leiden GenScript Biotech, Leiden) and used for the contact angle and fluorescence experiments.

A solution of potassium ferrocyanide (K4[Fe(CN)6 × 3H2O]) 10 mM prepared using sterilized PBS buffer 1X (AnalytiCals, Milan, Italy) was used as redox probe in the electrochemical tests.

The electrochemical analysis was conducted using an Potentiostat/Galvanostat/Impedance Analyzer (Metrohm Italia, VA, Italy) and screen-printed carbon electrodes (110, Metrohm Italia, VA, Italy).

The design of the aptamer was obtained by an appropriate bioinformatic platform (APTERION), property of Arta Peptidion (Parma, Italy), which generates (design unit) and select (classification unit) the best aptamer sequences for target proteins (any screening analyses of aptamer sequences will be provided, upon (motivated) request, by Arta Peptidion (info@artapeptidion.it)).

The training dataset of complexes aptamer-protein was built over experimentally validated complexes, including Aptamer Base and Protein Data Bank (PDB). These are separated into training and test datasets. To feed aptamer and protein sequences into our classification model, both sequences were encoded into a numerical representation with an optimal encoding function for protein and aptamer sequences. The encoding method was implemented in Python scripts.

The aptamer sequences were designed by using Arta Peptidion’s personal implementation of well-known search technique in the field of artificial intelligence (AI) called Monte Carlo tree search (MCTS) (Browne et al. 2012).

It is a probabilistic and heuristic driven search algorithm that combines the classic tree search implementations alongside machine learning principles of reinforcement learning. The top 30 sequences are used for docking and molecular dynamics. Among the final pool of five aptamer sequences, one was selected for the best features, converted into DNA, synthesized and experimentally tested to verify the interaction with the protein target.

The static contact angle was measured using a home-made system. Before incubation electrodes were washed in the buffer solution to prepare the surface. For each measurement, 2 µL of deionized water was placed on the electrode without modifications or after aptamer-TAMRA deposition at different concentrations and washing steps. The images were acquired with a CMOS camera and analyzed by Drop-Analysis (Stalder et al. 2006). For each sample, due to the small size of sensing area, one single drop was taken. The results were reported as average value of left and right angle, and the errors were estimated as the standard deviations.

Fluorescence microscopy images were taken using a Leica DMLA microscope (Leica Microsystems), equipped with a mercury lamp and fluorescence filter N2.1 (Leica Microsystems, Germany). All samples were observed with a × 10 magnification objective and measured via cooled CCD camera (DFC 420C, Leica Microsystems, Germany). Images were analyzed with Fiji software (Schindelin et al. 2012). The fluorescence of the solutions containing the aptamer-TAMRA was measured with a SPEX FluoroMax spectrofluorometer (SPEX Industries, Edison, NJ) using an excitation wavelength of 550 nm and recording the emission spectrum from 560 to 700 nm. The calibration curve was obtained measuring known amount of fluorescent aptamer-TAMRA, integrating the signal within the range of wavelength 560–580 nm. For unknown solutions, the integrated area value was converted to nanograms (ng) using the calibration curve reported in Fig. S1. The interaction aptamer–protein was also tested by using the fluorescence signal of TAMRA bounded to aptamer chain by incubation in the presence of the spike protein at increasing concentrations. The aptamer-TAMRA was deposed at 1 µM concentration in TRIS buffer on three different electrodes, in two independent experiments. The fluorescence signal was measured in solution after the addition of the spike protein (in PBS buffer) and 30 min incubation. The values were converted in nanograms of released aptamer using the calibration curves (Fig. S1).

A total of 12 µL aptamer solution (AP_7462DNA without fluorophore 0.5 µM) in PBS 1X with 0.55 mM MgCl2 was drop casted on the carbon working electrode (CWE) of the screen-printed and incubated for 45 min at room temperature (RT) (25 °C). After washing twice with 500 µL PBS 1X containing 0.55 mM MgCl2, the screen-printed were dried under biological hood and used for tests using a Metrohm Potentiostat/Galvanostat/Impedance Analyzer (Origgio, VA, Italy).

As positive controls, the spike protein at 0.001, 0.01, and 0.1 µM was used. A total of 12 µL of each spike concentration was added to the carbon electrodes on which it was previously deposed the aptamer at 0.5 µ M and incubated for 3 min at RT, washed with PBS buffer pH 7.4 1X and subjected to DPV analyses using 10 mM potassium hexacyanoferrate (II) (K4[Fe CN]6) in PBS 1X as chemical probe.

As negative controls were used: (i) bovine serum albumin (BSA) at 0.15 µM, 1.5 µM, 3 µM and 6 µM in PBS 1X; (ii) lysozyme at of 0.15 µM, 1.5 µM, 2 µM, 3 µM and 10 µM in PBS 1X.

A total of 12 µL of a solution of BSA or lysozyme was drop casted on the carbon WE of the screen-printed after the deposition of the aptamer at 0.5 µM and tested with differential pulse voltammetry (DPV).

A solution obtained by mixing the spike protein at 0.05 µM (positive sample) and BSA at 1.5 µM (negative sample) was also used. A total of 12 µL was deposed on the carbon screen-printed functionalized with the aptamer at 0.5 µM and tested using DPV. The same procedure was carried out using 12 µL of a solution containing the spike protein at 0.05 µM and lysozyme at 1.5 µM on the CWE with the aptamer at 0.5 µM.

The DPV measurements were carried out using a 10 mM K4[Fe (CN)6] (Sigma-Aldrich) PBS 1X solution with a scan potential from − 0.2 to + 0.4 V at 0.01 V/s. The data were collected with NOVA 2.1.2 software and the anodic peak current was measured with the same software. For each measurement, the anodic peak current was recorded from the carbon electrode after deposition of the aptamer and considered the blank (ipa Blank). The difference (Δipa) between ipa Sample and ipa Blank obtained for each sample was calculated.