Introduction

Biosensors play a major role in biomedical diagnosis1. The well described applications of biosensors contribute to progress in the clinical, environmental and agricultural industries2. Biosensors are analytical devices that perform two functions: recognition of an analyte and transduction to a measurable signal3. Electrochemical biosensors are commonly used for label-free sensing of bio-recognition events4. Electrochemical impedance spectroscopy (EIS) is an electrochemical technique sensitive to changes in the sensory layer’s electronic properties5,6. EIS often used to measure a recognition event between a biomolecule and an analyte because, upon binding, the sensory layer undergoes morphological changes that result in alternations of interface properties such as charge, hydrophobicity, capacitance, and surface density. These changes affect the permeation ability of the RedOx agent through the layer, thus, results in detectable impedimetric signal5,7,8,9.

Peptides are often used as recognition entities in biosensors due to their specific affinity towards many moieties, which are crucial for their biological activity10,11,12. Peptides are specifically attractive for EIS since they undergo massive conformational and dipole changes upon analyte binding resulting in measurable signal12,13,14. Their selectivity towards analytes such as proteins15, glycans16,17, small molecules18, and ions19 can be controlled by changing the sequence of amino acids. The multi-functionality of the peptide can also enable anchoring to a surface without harming their affinity towards a specific analyte20,21. Such decoupling enables impedimetric sensing on differently modified sensing platforms22.

Oxytocin (OT) is a neuropeptide with an affinity to both Zn2+ and Cu2+ ions23. Moreover, binding these ions is essential for OT biological activities24,25. Zn2+ and Cu2+ ions bind OT via different mechanisms and do not utilize the same binding sites for the interactions23,26,27. Namely, the terminal amine is essential for activation of OT-Cu2+ and initiates binding cascade, which proceeds with deprotonation of backbone amides23,28,29. Chelation of Zn2+ in OT proceeds via interactions with amides carbonyls23,28. Each of these metal ions induces distinctive conformational, electronic, and hydrophobicity changes to the OT complex24,30. Therefore, OT selectivity and sensitivity can be tuned for biosensing applications of Zn2+ and Cu2+20,31,32.

In our previous studies, we described several types of electrochemical OT-based biosensors for Zn2+ and Cu2+ ions. In these models, oxytocin was either anchored to a gold surface through a native disulfide bond20 or attached to a glassy carbon electrode using a silane-coupling strategy33. We showed that by coupling OT through terminal amine to lipoic acid, the affinity towards Cu2+ was blocked and the selectivity towards Zn2+ was improved32. This sensor proved highly selective but suffered from low surface charge transfer resistance (RCT) alternation, which can be explained by the low impedimetric signal-to-noise ratio (SNR) in the system. Low surface coverage is usually associated with peptides assembly since they are flexible, in constant conformational equilibrium/shift and experience repulsion from neighbouring peptides34,35. Contrary, long alkyl thiols pack to a very dense monolayer on gold surfaces and can act as a host layer for hydrophobic moieties36,37. We hypothesize that a recognition of analyte by a peptides embedded within a dense host layer may lead to significant changes in the properties of entire sensory layer and result in enhancement of the impedimetric signal.

In this study we present a biosensing platform in which OT functionalized with a dodecanoic acid (Dd) was embedded in a host alkane thiol monolayer by non-covalent interactions. OT functionalization leads to both selectivity toward Zn2+ ions due to terminal amine blocking and also to interlayer affinity by introducing alkane chain. The presented strategy extrapolates the formed hydrophobic surface of OT after Zn2+ chelation to change peptide interactions with the hexadecane thiol (HDT) layer and thus produce a measurable and concentration-dependent impedimetric signal. We envisioned that the effect of the peptide-metal ion chelation on the morphological changes in the sensory layer might lead to amplification of the impedimetric RCT parameters and hence could be applied for improving electrochemical biosensing. These bioactive intercalated monolayers were characterized by atomic force microscopy (AFM), variable angle spectroscopic ellipsometry (VASE), contact angle (CA), X-ray photoelectron spectroscopy (XPS) to study the interplay between surface morphology and the impedance.

Results and discussion

Platform design

OT was synthesized by solid-phase peptide synthesis in which dodecanoic-acid was coupled to the OT-terminal amine (see SI, Figure S1). The obtained Dodecanoic-Oxytocin (Dd-OT) designed to facilitate embodiment to the intercalation to the HDT self-assembled monolayer (SAM) on gold substrate by van der Waals (vdW) interactions between alkyl groups (Fig. 1).

Figure 1
figure 1

The fabrication steps of the sensing platform. Step i, self-assembly of HDT monolayer. Step ii, intercalation of Dd-OT molecule in the HDT SAM. Step iii, OT-Zn2+ complexation.

Full size image

Characterization of the system

Au-evaporated layer on Si wafers was modified with HDT SAM following incubation with Dd-OT and further exposure to Zn2+ (Fig. 1). These modified wafers were characterized by VASE, AFM, CA, and XPS. The HDT formed homogenous typical and well described38,39. SAM with measured thickness of 1.9 ± 0.1 nm, CA of 104° ± 2°, and roughness of 1.8 ± 0.1 nm (Figure S3). These substrates were incubated with Dd-OT and analyzed by XPS. The analysis (Figure S7) confirmed the presence of amide bonds related to Dd-OT at BE of 400.2 eV (related to N1s 1/2)40 and carbonyl groups (Figure S8) by the peak of 288.5 eV (related to C1s/8). The embedding of the layer with Dd-OT resulted in a total layer thickness of 3.4 ± 0.2 nm and roughness of 1 ± 0.1 nm (Fig. 2A). The measured increase in thickness and the decrease of roughness indicate that embedding Dd-OT changes the morphological properties of the host HDT SAM on Au substrate. However, the CA with a value of 103° ± 1° indicated that the change in hydrophobicity due to the assembly of the peptide on the host layer is insignificant.

Figure 2
figure 2

Zinc ion binding influence on the morphology and the stiffens of the layer: AFM topography analysis of (A) HDT-Dd-OT (Ra = 1.05 ± 0.09 nm) and (B) HDT-Dd-OT-Zn2+ (Ra = 0.52 ± 0.05 nm) and Slope of force-distance images of (C) HDT-Dd-OT and (D) HDT-Dd-OT-Zn2+.

Full size image

Exposure of the sensory layer to Zn2+, which is the analyte of the Dd-OT recognition element, resulted in the appearance of an XPS peak (Figure S9) with BE of 1022.3 eV (related to Zn2p 3/2), which is shifted relative to free Zn2+, that indicates the presence of the chelated Zn2+20,41.  AFM analysis indicated a roughness decrease to 0.5 ± 0.1 nm (Fig. 2B), while CA of 94° ± 1° related to a decrease in hydrophobicity. No significant thickness changes were observed by VASE analyses upon exposure for Zn2+. The decrease in hydrophobicity may indicate that a more hydrophilic layer containing the charged Dd-OT-Zn2+ was formed. The roughness alternation, which was measured by AFM (Fig. 2A,B), indicates the morphological changes in the layer due to Zn2+ binding20,32,42. The fact that VASE measurements did not show any changes in the thickness of the layer indicates that there are no significant changes in the positioning of the peptide in relation to the host layer. AFM with a quantitative mode of operation (QI) was performed to calculate stiffness values and Young’s modulus of Dd-OT modified substrates before and after exposure to Zn2+ (Fig. 2C,D and Figures S4–S6). The exposure to Zn2+ caused an increase in Young’s modulus from 2 ± 1 GPa to 4.8 ± 0.5 GPa, indicating that Zn2+ results in morphological changes of the monolayer43. The changes in surface morphology and other presented parameters caused by the binding of Zn2+ on the biosensing platform, can be further applied in impedimetric biosensing, which is a sensitive technique to exploit changes in layer morphology.

Impedimetric analysis of HDT SAM and Dd-OT intercalation

Au working electrodes were modified with HDT SAM by the presented protocol and analyzed by EIS measured in a solution containing [Fe(CN)6]4−/[Fe(CN)6]3−. This modification resulted in an increase of RCT from 0.2 kΩ for bare Au electrode (Figure S2) to 89 kΩ for HDT SAM (Fig. 3) indicating the formation of a dense layer44. After the immersion of the HDT-modified electrode in Dd-OT solution, the RCT increased to 697 kΩ, which is much higher than the value observed for OT covalently attached to the surface20,32. This large increase in the RCT values can be attributed to changes in layer morphology such as density, stiffness, rigidity, etc. caused by intercalation of HDT with Dd-OT2,4,5,6,9.

Figure 3
figure 3

Nyquist plot of HDT-modified electrode before (black curve) and after (red curve) Dd-OT intercalation.

Full size image

Dose–response of Dd-OT sensor to Cu2+ and Zn2+ ions

HDT-Dd-OT sensor was exposed to Zn2+ solutions with increasing concentrations from 10–12 M to 10–6 M (Fig. 4). The exposure to Zn2+ concentrations lower than 10–10 M had no considerable effect on the RCT values, which is below the limit of detection of the presented sensor. At concentrations ranging from 10–10 M to 10–8 M, the impedimetric response gradually increased and eventually reached a value five times higher than the measured RCT values prior to Zn2+ exposure (Fig. 4B). At higher concentrations (above 10–8), the sensor reached its saturation and the signal stabilized. The exposure of the intercalated OT layer to 10–8 M of Zn2+ (HDT-Dd-OT-Zn2+) resulted in a signal enhancement that reached an RCT value which is five times higher than that of the surface prior to incubation with the ion (HDT-Dd-OT). This signal enhancement surpasses the response following exposure to Zn2+ of chemisorbed or covalently bound OT surface layers, which only reached an increase of up to 1.35 in RCT values over the same range of concentration32. We assume that the signal enhancement observed is because the OT complexation with Zn2+ influences not only the peptide conformation itself but also modifies the morphology and the net charge of the entire sensory layer. This directly influences the permeation of the redox species through the intercalated layer. To verify the selectivity of the sensor, dose–response analysis for Cu2+ ions was performed. No significant impedimetric response was observed in the relative concentration range (10–12 M to 10–6 M; see Fig. 4, Figure S10). This demonstrates that the dodecyl functionalization of OT by amidation of the terminal amine blocks its ability to complex Cu2+, as previously reported32.

Figure 4
figure 4

Nyquist plots for (A) different Zn2+ ion concentrations and (B) dose–response of Dd-OT sensor to Zn2+ (black curve) and Cu2+ (red curve) ions. The presented averaged values are based on measurements from three different electrodes (Figure S10).

Full size image

Conclusions

In this work, we designed a new Zn2+ ion-selective receptor based on an alkyl-amidated OT embedded in a host alkanethiol monolayer by vdW interactions. This intercalated peptide monolayer induces enhanced impedimetric response upon ion-binding. Its response to Zn2+ induced conformational changes that evoked morphological changes to the entire monolayer. These alterations were probed by impedimetric measurements and yielded a selective sensor for Zn2+ ions in the range of 0.1 nm up to 10 nm. By this, the monolayer serves as a host for the Dd-OT sensing peptide, which enhances the RCT multiplication for the biorecognition event. This approach is fundamentally different from sensors assembled by direct covalently bound OT monolayers. This non-covalent assembly concept can be further extended to other bio-receptors incorporated in natural and artificial membranes.

Materials and methods

Surface modifications

OT was synthesized using solid-phase peptide synthesis (SPPS) and modified by coupling dodecanoic-acid (Dd) to the N-terminal using the same method (see SI). Au layer (100 nm) was evaporated on top of a Cr layer (10 nm), which was evaporated on the substrate of n-type Si wafer (\(\left\langle {100} \right\rangle\)). The bare gold surfaces were washed with absolute ethanol, before and after 20 min cleaning using Ultra Violet Ozone Cleaning Systems (UVOCS Inc). HDT (CH3–(CH2)15–SH, Sigma-Aldrich) was freshly distilled under vacuum before use. HDT adsorption was performed by immersing bare gold surfaces in HDT solution (10 mM in absolute ethanol) for 18 h at 22 °C. After incubation, the surfaces were washed with absolute ethanol and dried under dry N2. Dd-OT was dissolved in ammonium acetate (AA) buffer (50 Mm, pH = 7) (Sigma-Aldrich). Then, HDT-modified surfaces were immersed in the Dd-OT solution (10 µM) overnight at 22 °C to allow the intercalation of Dd-OT with the HDT SAM. Afterward, the surfaces were washed by immersion in AA buffer for 10 min and dried under dry N2. Complex formation of Dd-OT with either Zn2+ or Cu2+ ions was obtained by incubating the modified surfaces in 10 µM Zn2+/Cu2+ in AA buffer for 10 min.

Ellipsometry

Thickness measurements were performed by variable angle spectroscopic ellipsometry (VASE) measurements with VB-400 ellipsometer (Woollam, Lincoln, NE, USA) at the Brewster angle of 75°.

Atomic force microscopy (AFM)

Measurements were carried out by NanoWizard3 (JPK Instruments, Berlin, Germany) in fresh AA buffer (pH = 7) at 298 K, using Aspire CT-130 (Team Nanotech, Villingen-Schwenningen, Germany) cantilevers with a spring constant of 6 N/m under QI mode—force curve-based imaging. To assure homogeneity of the surface, images were taken at different regions of the same surface and at a resolution of 500 × 500 pixels, an area of 2 × 2 μm and at a pixel rate of 0.8 Hz. Young’s modulus and stiffness values were calculated by JPK Instrument data analysis software using Derjaguin-Muller-Toporov (DMT) model and the force curve slope.

Electrochemical impedance spectroscopy (EIS)

Electrochemical analyses were conducted by Metrohm-Autolab PGSTAT-302N digital potentiostat (EcoChemie BV, Utrecht, Netherlands) operated by Nova software. The electrochemical cell contained three electrodes: Ag/AgCl/3M KCl as a reference electrode, Pt as a counter electrode, and a polycrystalline disc gold electrodes with a 2 mm diameter (CH instruments) The electrolyte solution was prepared from AA buffer (50 mM, pH = 7) which contained 0.1 M KCl as a supporting electrolyte, and 5.0 mM K3[Fe(CN)6], 5.0 mM K4[Fe(CN)6] as RedOx species. The gold electrodes were cleaned by polishing with 0.05 µm alumina suspension on micro-cloth pads (Buehler). Electrode modifications were performed following the precedent described procedures. The frequency range was 0.1 Hz–10 kHz for HDT-modified electrode measurement, and 0.01 Hz–10 kHz for Dd-OT-modified electrode measurement. After dip—coating, the HDT-modified electrode was washed with ethanol absolute and the Dd-OT-modified electrode with AA buffer, until stabilization of the impedance response. Then, dose–response of the sensor was conducted by 10 min incubation of the Dd-OT-modified gold electrode in Zn2+/Cu2+ solution (1 pM–10 µM at room temperature). After each EIS measurement, the modified electrode was washed and moved to a higher concentration. The equivalent circuit chosen to fit EIS data was RS(Q[RCT|W]), with RS value for solution's resistance, Q for layer’s capacitance, RCT for charge transfer resistance of the layer, and W for Warburg impedance. The results are presented as normalized RCT. The charge transfer resistance was normalized by dividing the RCT value after exposure to heavy metal ions by the RCT value before exposure.