A Highly Selective Biosensor Based on Peptide Directly Derived from the HarmOBP7 Aldehyde Binding Site

This paper presents the results of research on determining the optimal length of a peptide chain to effectively bind octanal molecules. Peptides that map the aldehyde binding site in HarmOBP7 were immobilized on piezoelectric transducers. Based on computational studies, four Odorant Binding Protein-derived Peptides (OBPPs) with different sequences were selected. Molecular modelling results of ligand docking with selected peptides were correlated with experimental results. The use of low-molecular synthetic peptides, instead of the whole protein, enabled the construction OBPPs-based biosensors. This work aims at developing a biomimetic piezoelectric OBPPs sensor for selective detection of octanal. Moreover, the research is concerned with the ligand binding affinity depending on different peptides’ chain lengths. The authors believe that the chain length can have a substantial influence on the type and effectiveness of peptide–ligand interaction. A confirmation of in silico investigation results is the correlation with the experimental results, which shows that the highest affinity to octanal is exhibited by the longest peptide (OBPP4 – KLLFDSLTDLKKKMSEC-NH2). We hypothesized that the binding of long chain aldehydes to the peptide, mimicking the binding site of HarmOBP7, induced a conformational change in the peptide deposited on a selected transducer. The constructed OBPP4-based biosensors were able to selectively bind octanal in the gas phase. It was also shown that the sensors were characterized by high selectivity with respect to octanal, as well as to acetaldehyde and benzaldehyde. The results indicate that the OBPP4 peptide, mimicking the binding domain in the Odorant Binding Protein, can provide new opportunities for the development of biomimicking materials in the field of odor biosensors.

Pure ractions (> 95%, by HPLC analysis) were collected and lyophilized. The molecular weights were confirmed by ESI-MS (Waters Acquity SQD, Milford, MA)( Figure S1.b). Schematic presentation of synthetic peptides based sensor fabrication was presented in Figure  S2. Because of Cys-terminated sequence, the peptides possess an ability to form self-assembling layers on the gold electrode surface. Deposition process occurs spontaneously due to chemisorption of the modifying compound on metal. The sensor surface was automatically dried and rinsed between the immersion cycles. Modification of the sensing surface is an important process in QCM biosensor studies. Peptide deposition generates a suitable recognition layer with a specific property or function. The surface of QCM crystals was rinsed with acetone, methanol and deionized water, then dried with nitrogen until a final solvent evaporation from the surface. A strong oxidizer (Pyranha solution, 30 % H2O2:H2SO4, 1:3, v/v) was placed on the gold electrode to remove organic materials. Next step of the QCM preparation was rinsing with deionized water and ethanol and drying with nitrogen. Every deposition process was carried out in ambient temperature in the dark to avoid possible peptide degradation.
Deposition of a mass on a thin gold surface induces a decrease in the resonant frequency for a rigid substance. A relationship between frequency change and peptide deposition efficiency is expressed by the Sauerbrey [1] (1): where, and µ are the density (2.648 g·cm -3 ) and shear modulus of quartz (2.947×10 11 g·cm -1 ·s 2 ), respectively is the crystal fundamental frequency of the piezoelectric quartz crystal, is the crystal piezoelectrically active geometrical area which is defined by the area of the deposited metallic film on the crystal, and are the mass and system frequency changes. According to this equation, the mass of a thin layer deposited on the surface can be calculated through measuring the changes in the resonant frequency. For a typical piezoelectric sensor with a 10 MHz frequency, a change in mass of 4.4 ng results in a frequency change of around 1 Hz·cm -2 .
The AFM images were acquired to examine the morphology of the film surface on the peptidebased biosensors. An AFM Ntegra Prima device manufactured by NT-MDT (Moscow, Russia) was used. In the topographic measurements, the tapping mode with the set-point equal to half-value of free oscillations amplitude was applied. The measurements were carried out using conductive probes of the NSG 01 type, manufactured by NT-MDT. The geometric dimensions of the probe lever were 125 × 30 × 2 (L × W × T/mm), while other parameters were as follows: resonance frequency: 150 kHz, spring constant: 5.1 N/m, radius of tip curvature: 10 nm. In all studies on biosensors, regardless of the deposition method, a peptide monolayer was observed (10 µm × 10 µm). To show the uniformity of the layers an AFM images at 10 µm × 10 µm with basic parameters and 1 µm × 1 µm are presented in Figure S3. Average depositions were calculated for three sensors of each dip-coating. Saturation occurred at 20 mg·mL −1 , this concentration was chosen for peptide deposition using dip-coating technique. In Figure S4, shifts in frequency during deposition cycles in different concentrations are presented. The correlation between calculated and measured concentrations of gas mixtures are presented on Figure S4.
According to previously reported evaluation of most effective peptide deposition method, dipcoating was selected to obtain homogenous biosensors receptor layers. The first peptide (LEKKKKDC) was prepared in water/acetonitrile (1:1, v/v,), whereas the remaining three LFDSLTDLKC, LFDSLTDLKKKMSEC, KLLFDSLTDLKKKMSEC in acetic acid/acetonitrile (9:1, v/v) allowing complete dissolution of the compounds. Additionally, the compounds were degassed with helium (10 min) before use in order to prevent oxidation of thiol group (−SH) and formation of sulphur bridges (−S-S). Dip-coating processes were conducted at room temperature, in darkness.
Average deposition on electrode surface for four peptide-based sensors, acquired in three repetitions, is shown in Figure S5.

Gas Mixtures and Measurements Setup
Correctness of the prepared reference solutions was verified using gas chromatography coupled with thermal conductivity detector (AutoSystem XL, PerkinElmer) and flame ionization detector (430-GC, Bruker ® , Bremen, Germany), which is confirmed by presented correlation plots ( Figure S5) and high coefficients of determination R 2 > 0.97 ( Figure S5).  Table S1 presents the values of selectivity coefficients of all sensors. The selectivity coefficients were calculated using Formula (1):