Method for optical label-free detection of molecules at ultra-low concentrations, in particular as low as zetomolar concentrations

Abstract

Disclosed herein is a method for detecting at least one binding event of a specific binding pair-forming substance, comprising: providing a layer (2) of a first specific binding pair-forming substance on a substrate (3), the specific binding pair-forming substance having a density of 102-104 / [mu] m2 of the substrate (3). Providing a housing (4) of the first specific binding pair-forming substance layer (2) having a first dielectric comprising a first solution having a first pH value and a first ionic strength value, rinsing the housing (4) with a second dielectric comprising a second solution having a second pH value and a second ionic strength value for a first period of time, wherein at least the second pH value is different from the first pH value,-after the rinsing, recovering the first dielectric having a first pH value and a first ionic strength value in the housing (4),-feeding a solution of a second specific binding pair-forming substance into the housing (4) of the layer (2) of the first specific binding pair-forming substance, after the recovery of the first dielectric in the shell (4), causing an interaction between the first specific binding pair-forming substance and a second specific binding pair-forming substance, and culturing the shell (4) for a second period of time, the interaction comprises a binding event when the first specific binding pair forming substance and the second specific binding pair forming substance provide said specific binding pair,-detecting an offset ([Delta] [Theta]) of a parameter ([Theta]) representative of a dielectric function of the layer (2) of the first specific binding pair forming substance after said incubation shell (4) lasts for a second period of time, the offset is defined as a change in a parameter ([theta]) representing a dielectric function of a first specific binding pair forming a layer (2) of a substance from a first value occurring when the housing (4) comprises only the first dielectric after the recovery of the first dielectric to a second value occurring after the incubation of the housing (4) for a second period of time, -comparing the offset ([delta] [theta]) with a threshold value, whereby when the offset exceeds the threshold value, at least one binding event occurs between said first and second specific binding pair-forming substances.

Description

Technical Field

[0001] The present invention relates to the optical label-free determination of molecules at very low concentrations, particularly concentrations as low as 10 zeptomolar. Background Art

[0002] Near-field methods involving nanometer-sized interfaces can achieve label-free detection with single-molecule resolution. 1 That is, a transduction nanointerface biofunctionalized with a few recognition elements (e.g., antibodies) can detect a series of single antigen-antibody affinity binding events. However, the occurrence of such events is limited by the Brownian diffusion of the antigen. 2-4 Under these conditions, binding can only occur within a few minutes if the capture nanointerface and the antigen are confined to a femtoliter volume. This means that in the 0.1 ml volume to be analyzed, at least 10 11 antigens (nM-10 -9 Therefore, near-field methods driven by nanointerfaces involve studying a series of events with single-molecule resolution, but cannot achieve the measurement of single molecules in 0.1 ml, or equivalently, cannot measure the limit of detection (LOD) or limit of identification (LOI) of single molecules in 0.1 ml volumes, that is, to 10 -20 LOD in moles per liter (M).

[0003] Optical methods for single-molecule studies are no exception, as they rely heavily on near-field methods, where femto / picomolar (10 -15 / 10 -12 M) detection limit. 5-9 This low LOD, but still far from a single molecule, is achieved by using techniques such as localized surface plasmon resonance (LSPR) 7 or Plasmon Enhanced Raman Spectroscopy 9 It is well known that standard surface plasmon resonance (SPR) of antigens captured by an antibody layer deposited on a gold SPR slide 10-12 The LOD is at 10 -8 Within the range of M. 13-18

[0004] Recently, label-free biochemical detection at a single-molecule LOD of 0.1 ml has been demonstrated by using a large (millimeter-wide) detection interface. 2,19–22 and Kelvin probe force microscopy (KPFM) 23 10 years of research on bioelectronic devices 4 μm 2 In both cases, the surface is filled with 104 / μm 2 Density of capture elements. KPFM studies show how the extended surface potential change after 10 ± 3 antigen affinity bindings can be directly measured on a sample surface filled with 10 8 antibodies. It has been demonstrated that the key to overcome the diffusion limit that plagues near-field methods is the millimeter wide capture surface itself. Such a large detection interface greatly increases the probability that some of the 0.1 ml of antigens are captured by one of the 10 11 -10 12 antibodies attached to it. 24

[0005] Large-area bioelectronic approaches that are drawn include the application of an electric field, which is believed to be necessary to achieve the detection of single binding events with a sufficiently high signal-to-noise ratio. For this purpose, an electrostatic domino-like amplification process is actually assumed to enable the propagation of the electrostatic conformational change of the antibody involved in a single antigen binding event. The electrostatic change starts from this antibody and propagates to the neighboring antibodies (which can be induced by the applied electric field), successively changing their conformational state. The propagation continues until a large fraction of the trillion antibodies filling the millimeter wide gate electrode are exchanged. 19,20 The detailed mechanism of such an amplification process, which is a necessary condition to make the barely noticeable single binding electrostatic change detectable, remains difficult to understand. SUMMARY

[0006] Object of the invention

[0007] It is an object of the present invention to solve the technical problems mentioned in the prior art. In particular, it is an object of the present invention to provide a method for analyzing molecules at ultra-low concentrations, in particular at the single-molecule level of 0.1 ml (10 -20 M), with an optical label-free method.

[0008] Summary of the invention

[0009] The object of the present invention is achieved by a method having the features of the above claims, which form an integral part of the technical solution provided herein in relation to the present invention. In particular, the optical label-free method of the method of the present invention is performed by surface plasmon resonance (SPR). BRIEF DESCRIPTION OF DRAWINGS

[0010] The present invention will now be disclosed with reference to the attached drawings, which are provided by way of non-limiting example, in which:

[0011] - Figure 1 (a) is a schematic representation of a surface plasmon resonance (SPR) layer and of the surface plasmon mode propagating at the interface between the metal layer and the dielectric layer,

[0012] - Figure 1 (b) shows the structure of a SPR device in Kretschmann configuration,

[0013] - Figure 1 (c) shows the SPR angular reflectance curve measured with a 670 nm laser source at the physical adsorbed anti-IgG layer (hollow circles) at the gold surface (hollow squares) and the fitted curve;

[0014] - Figure 1 (d) shows the SPR trace line (plasmon peak angular shift vs. time) during the physical adsorption of the anti-IgG layer on the SPR gold slide;

[0015] - Figure 2 (a) shows the IgG SPR analysis using a physical adsorbed anti-IgG capture layer comprising (2.8 ± 0.4) - 10 11 antibodies;

[0016] - Figure 2 (b) shows the IgG calibration curve (hollow squares) and a negative control experiment, where bovine serum albumin was used as a non-capture layer (hollow circles); the LOD was estimated as the average of the noise in the negative control experiment plus three times its standard deviation 30 ;

[0017] - Figure 3 (a) shows the SPR transient signal recorded according to protocol A disclosed below;

[0018] - Figure 3 (b) is a zoomed view of the part of (a) identified by the dashed box and includes only steps 1-7; Figure 3

[0019] - Figure 4 (a) shows the SPR transient signal recorded according to protocol B disclosed below;

[0020] - Figure 4 (b) is a zoomed view of the part of (a) identified by the dashed box and includes only steps 1-7; Figure 4

[0021] - Figure 5 (a): SPR transient signal recorded according to protocol C disclosed below;

[0022] - Figure 5 (b): SPR transient signal recorded according to protocol D disclosed below;

[0023] - Figure 6 ​​(a): SPR transient signal recorded on an anti-IgG film physically adsorbed on a gold-coated slide according to Protocol E (disclosed below);

[0024] - Figure 6 (b): SPR transient signal recorded according to Protocol E (disclosed below) on anti-IgG physically adsorbed on a gold nanoparticle-based film deposited on a gold-coated glass slide.

[0025] - Figure 7 : SPR transient signals recorded according to Protocol F (disclosed below) on anti-IgG physically adsorbed on gold nanoparticle-based films deposited on gold-coated glass slides.

[0026] - Figure 8 : SPR theta shift at IgG 100zM, 1aM and 100nM for all protocols AF compared to the baseline given in Table 2. Results from another protocol called "Protocol F" are also added. Negative control experiments were performed with IgM at 1fM. All experiments were performed on an anti-IgG capture layer physically adsorbed on a gold-coated slide.

[0027] - Figure 9 :It is 2 of plan AD 2 Graphical representation of a factorial design.

[0028] - Figure 10 (a): Figure 9 2 2 Graphical representation of the coefficients of the SPR angle shift model at 100 zM for the factorial design.

[0029] - Figure 10 (b) shows Figure 9 2 2 At 100zM for the factorial design, the iso-response contour plot of the SPR angle shift in the entire experimental domain response is shown.

[0030] - Figure 10 (c) shows Figure 9 2 2 Graphical representation of the coefficients of the SPR angle shift model at 100 aM for the factorial design,

[0031] - Figure 10 (d) shows Figure 9 2 2 At 100aM response for the factorial design, the iso-response contour plot of the SPR angle shift in the entire experimental domain is shown.

[0032] - Figure 10 (e) and (f) are at 100 nM and exposed to 2 2Graphical representation of the coefficients of the model of the SPR angular shift at the IgM 1 fM (negative control experiment). Error bars are at the confidence level of p = 0.05, while the asterisks indicate the significance of the coefficients (* = p < 0.05, ** = p < 0.01, *** = p < 0.001).

[0033] - Figure 11 (a) shows a graphical representation of an analysis carried out under physiological conditions, which shows how to detect real markers (e.g. HIV-1 p24 (10zm and 100zM)) in 0.1 ml of physiological buffer at the single molecule level, in the presence of interfering proteins (e.g. MERS-CoV Spike S1 (1 fM)). In this case "Protocol D" is used.

[0034] - Figure 11 (b) is the same as Figure 11 (a), but, in order to better simulate the analysis of a patient fluid, only the baseline in the reference fluid (HEPES buffer 1, i s = 150 mM and pH = 7.4) is proposed, followed by a rinse with HEPES buffer 4 (i s = 150 mM and pH = 6) and the analysis is carried out at 10zM in a solution containing the MERS-CoV Spike S1 (1 fM) interfering protein and the marker to be detected, HIV-1 p24.

[0035] - Figure 11 (c) is the same as Figure 11 (b), but the rinse with HEPES buffer 4 (i s = 150 mM and pH = 6) is only carried out after proposing the detection at 10zM (carried out in a solution containing the MERS-CoV Spike S1 (1 fM) interfering protein and the marker to be detected, HIV-1 p24).

[0036] As a preliminary remark, the attached drawings of the present disclosure can contain textual content in addition to the reference numerals or letters, the purpose of which is to enhance the understanding of these drawings, especially considering the nature of the invention. The textual content included in the drawings can also be repeated in the following disclosure. DETAILED DESCRIPTION

[0037] Surface Plasmon Resonance (SPR) relies on an evanescent wave associated with the total internal reflection of a laser at the boundary between a high refractive index prism and a dielectric, also as Figure 1 (a) 12 shown. If the angle of incidence exceeds a critical value (Bruster angle), the incident light will undergo total (internal) reflection.

[0038] Surface plasmons appear at the interface between a metal with a dielectric function (real part) ε' m < 0 and a dielectric (e.g. water) with ε' d > 0.15 The strong optical field associated with the collective oscillations is generated along the metal-dielectric interface ( Figure 1 The evanescent characteristic propagates along the x-axis in (a) and decays exponentially in metallic and dielectric media. Such evanescent characteristics give the technique surface specificity and allow monitoring of layers as small as 30-300 nm deposited at metal / liquid interfaces.

[0039] SPR does not envisage an applied electric field, but rather an optical field in which charge oscillations (plasmons) at the metal / dielectric interface are excited by light. 15,12

[0040] More specifically, the present invention comprises a novel method comprising a physiological solution (i s =150 mM and pH = 7.4), the pH of the solution used to flush the detection interface was changed. This alone enables the -20 Optical label-free detection was performed at an LOI of 100 M. s ) alone will not produce a priming effect, although together with a pH change it may help to enhance the signal. Washing is preferably performed prior to detection.

[0041] refer to Figure 1 (b), reference numeral 1 denotes an analytical device (and an experimental device) configured to perform the method according to the present invention. Figure 1 (b) Angle-modulated Kretschmann configuration, using BioNavis-200 multi-parameter surface plasmon resonance (MP-SPR) Navig TM The instrument performed all the experiments mentioned in this article. Here, the evanescent wave monitored the capture anti-immunoglobulin G (anti-IgG) layer 2 (generally, according to the present invention, the layer of the first specific binding pair-forming material) deposited on a gold substrate 3. The SPR modulus was equipped with two laser sources L1 and L2 (both at λ = 670 nm), which illuminated 3 mm-spaced areas of the sample to assess its homogeneity. In all the experiments mentioned in this article, the two measurement traces were almost identical, which proves that the uniformity of the capture antibody layer performance was always very high.

[0042] Therefore, for the sake of clarity, only the tracks from one laser source are shown in the experiments presented herein. The laser beam incidence angle Θ (greater than the critical angle for total reflection) was varied in the range 50.290°-77.930° with an instrumental error of 0.002°. A flow cell 4 with an internal volume of 0.1 ml was used as fluidic. Cell injection was performed manually with a 1 ml sterile syringe and the experiment was carried out at room temperature. All the proteins (anti-IgG, IgG, IgM) used in the experiments mentioned herein are polyclonal antibodies from Sigma-Aldrich and were used without further purification.

[0043] The sensor slide (SPR Navi-200) configured in the sample holder comprises an optical glass 5 matched to the high refractive index of the prism 6, covered by a semi-transparent film of thermally evaporated gold (-50 nm) on a chromium adhesion layer (-2 nm). Before use, the slide is immersed in an aqueous solution of NH4OH / H2O2(1 : 1 :5 v / v) at 80-90 °C for 10 minutes; after this it is rinsed with water, dried with nitrogen and treated in a UV-ozone cleaner for 10 minutes. The heading 7 indicates the photodetectors configured to detect the reflected light beams from L1 and L2.

[0044] Results

[0045] The capture anti-IgG layer 2 was deposited by injecting 0.1 ml of a 50 μg / ml solution of anti-IgG in phosphate buffered saline (PBS, i s = 163 mM and pH = 7.4) in the flow cell 2. The SPR slide (glass 5) previously placed in the flow cell 2 was contacted with the anti-IgG solution for 2 hours, while the SPR signal was measured.

[0046] In Figure 1 In (c) the angular reflectivity curves containing the plasmonic peak as a function of the incidence angle Θ are shown for the bare sensor slide (hollow square) and for the slide covered by the physisorbed anti-IgG layer (hollow circle).

[0047] The experimental curves were simulated with the Winspall 3.02 software 26 The experimental curves were simulated with a multilayer model based on the Fresnel equations. The thickness and optical parameters obtained from the simulation of the SPR curves are summarized in Table 1 below.

[0048] Layer Thickness / nm Refractive index Extinction coefficient Glass BK7 - 1.518 0 Cr 2.0±0.1 3.3±0.2 2.8±0.4 Au 41±1 0.16±0.02 3.85±0.05 Anti-IgG 6.9±0.5 1.374±0.005 0 PBS - 1.330 0

[0049] The results obtained from the bare gold-coated slide are in agreement with the nominal thickness declared by the SPR slide supplier and with the related optical parameters given in the literature 27There is a perfect agreement, which demonstrates the reliability of the simulations performed.The resulting value of the simulated thickness of the anti-IgG layer is 6.9 ± 0.5 nm, where the error bars are estimated as the relative standard deviation of two different sampling areas.

[0050] exist Figure 1 In (d), the SPR plasmon peak angle shift versus time was recorded when the anti-IgG layer 2 was adsorbed on the glass slide. According to the convention in SPR experiments, the de Feijter equation 28,21 The surface coverage of anti-IgG physically adsorbed on gold surface 3 was quantitatively evaluated and the result showed that the surface coverage was (7.1±0.4)·10 11 / cm 2 Typically, according to the present invention, the density or surface coverage of the layer 2 of the first specific binding pair forming material is 10 2 -10 4 / μm 2 Base 3( Figure 1 (d) The gold surface is provided on a substrate 3.

[0051] Immunoglobulin antibodies have a Y-shaped structure of 14.5nm×8.5nm×4.0nm. 29 Therefore, the thickness of 6.9 ± 0.5 nm indicates that the 11 A single monolayer of antibodies was deposited on a 0.4 cm 2 The theoretical coverage of the antibody layer (standing or tiled) corresponds to approximately 10 12 molecules / cm 2 . 21 These data consistently support a model in which the gold surface is covered by a highly packed monolayer of anti-IgG proteins that are physically adsorbed in a disordered manner.

[0052] The second capture layer was produced according to the following procedure. The concentration of 3.45·10 9 Gold nanoparticle (Au NP) suspension of 100 particles / ml was purchased from Sigma-Aldrich (product number 753688) and used as is without further purification. These nanoparticles maintain a diameter of 100 nm and are already stable, requiring no capping agent. First, in order to surface modify Au NP with anti-IgG antibodies, 4 ml of the nanoparticle suspension was mixed with an anti-IgG solution (0.1 mg / ml) in 180 μl PBS.

[0053] Therefore, the total number of nanoparticles in the solution is about 1·10 10 Nanoparticles and 1·10 15The mixture of antibodies. This means having an excess of anti-IgG to completely cover the surface of the NPs. In fact, a 100 nm size NP can accommodate about 540 anti-IgG. The solution of NPs and anti-IgG was stirred at 525 rpm for 1 hour at 20-22 °C to achieve the physical adsorption of the antibodies. Then, the solution was centrifuged at 8000 x g for 30 minutes to remove the supernatant containing the excess of anti-IgG not physically adsorbed. The remaining pellet was resuspended in 1 ml of fresh PBS to have a final concentration of about 1 · 10 10 nanoparticles / ml of modified NPs (anti-IgG NPs). The colloidal solution was stirred with vortex (1000 rpm) for 1 minute and sonicated in an ultrasonic bath for 20 minutes. The solution thus obtained was used for the SPR gold slide modification. To this end, the anti-IgG NP solution was injected into the SPR flow cell in contact with the bare gold SPR slide. The injection of 500 μL was repeated five times by making the same solution flow through the flow cell. During each injection, the bi-functionalized NP solution was left in contact with the SPR slide for 15 minutes to completely cover the sensor area (0.42 cm 2 ) of the SPR slide. The SPR slide prepared according to this protocol can accommodate about 5 · 10 9 anti-IgG NPs.

[0054] The capture of IgG antigens on the physically adsorbed capture antibodies on gold was then investigated to measure the shift in the angle of deflection as a function of time. The data are given in Figure 2 (a), where a typical SPR analysis of IgG antigens (in PBS solution in the range 2 nM - 440 nM) incubated with the anti-IgG capture layer is shown.

[0055] In Figure 2 (b), the relative calibration curve (squares) is shown, as well as a negative control experiment (circles) in which a non-capturing physical adsorption layer of bovine serum albumin (BSA 0.1 mg / ml, 2 hours) was required. The negative control experiment (circles) allowed to estimate the noise level and its standard deviation. The LOD (Limit of Detection) level was estimated as three times the standard deviation added to the average signal of the negative control experiment, 30 corresponding to a concentration of 36 nM, in agreement with the literature data.

[0056] In addition, it is clear from the minimum concentration required to detect a significant signal that the change in the SPR signal (related to the change in the dielectric function at the metal-water interface) is due to the formation of a stack of layers of IgG antigens captured on the anti-IgG capture film 2. The IgG layer at saturation of the SPR signal contains (1.4 ± 0.8) · 10 11 nanoparticles on 0.42 cm 2antigen, meaning that there is a dense layer of IgG with approximately one antigen per two anti-IgG. The simulated thickness of the IgG layer 21 (over the top of the 6.9 ± 0.5 nm thick anti-IgG layer) is 1.2 ± 0.8 nm. Such a small thickness should be attributed to the lying-down orientation of the IgG antigens on the anti-IgG layer 2, which is also consistent with previously published studies. 31

[0057] To investigate the SPR detection signal at very low LOD, experimental protocols A-F (B-F correspond to embodiments of the method of the present invention) have been established, which will be disclosed hereinafter.

[0058] Negative control experiments include the measurement of SPR traces (plasmonic peak angle shift Δθ vs. time) on anti-IgG covered slides after injection of 1 fM (10 -15 M) of non-affine binding IgM antigen. The noise level was calculated considering all negative control experiments (performed according to different protocols, see below), which have an average Δθ shift of 0.003° and a standard deviation of 0.002°. The angle shift was continuously evaluated after a step of 45 minutes of exposure to the IgM (1 fM) in buffer and a step of 20 minutes of rinsing in buffer or water. The limit of detection (LOD) level was calculated as the average noise plus three times its standard deviation 30 : LOD level = 0.009°. The limit of identification (LOI) level was calculated as the average noise plus six times its standard deviation: 30 LOI level = 0.015°. At the LOI level, the detection is very reliable, since both false negative and false positive random errors are lower than 1%. This leads to at least 99% of diagnostic sensitivity, specificity and selectivity.

[0059] After the level of the negative control experiments was evaluated, the trace measurements were continued starting with the injection of increasing concentrations of IgG, spanning a concentration range from 10 zeptomolar (10 -21 M, zM) to 100 nM. Each incubation with the affinely binding IgG solution lasted 45 minutes, followed by a 20 minutes rinsing step.

[0060] In the first experimental protocol, referred to herein as "protocol A", the 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer containing a zwitterionic HEPES molecule and sodium chloride (i s = 150 mM, pH = 7.4) was chosen to mimic physiological conditions, which will be referred to as "HEPES buffer 1". Considering that most antibodies have more basic isoelectric points (pI > 6, and for anti-Ig in the range 6.4-7.6) 32,33The anti-IgG layer should be slightly negatively charged.

[0061] The glass slide covered with the physisorbed anti-IgG layer 2 from PBS was placed in the SPR flow cell 4. In each step of all experiments, an analysis solution providing physiological conditions (HEPES buffer 1 containing IgM (non-binding antigen for negative control experiments) or IgG (affinity binding substance to be detected)) was injected (0.1 ml) into the cell and incubated with the capture layer for 45 minutes. When IgG was present, binding actually occurred. The capture layer was then washed with 3 ml of HEPES buffer 1 without antigen. This step lasted 20 minutes. More specifically, the experiment started with the injection of 0.1 ml of IgM (1 fM) in HEPES buffer 1, which was incubated for 45 minutes. Then 3 ml of pure HEPES buffer 1 were injected to rinse the sensor surface and left for 20 minutes. The injection of a solution of IgG still in HEPES buffer 1 (0.1 ml) was continued, which spanned a concentration range from 100 zM to 100 nM. In each incubation step, IgG bound to the capture anti-IgG layer, followed by a rinsing step.

[0062] More schematically, a first experimental protocol, hereinafter referred to as "protocol A", comprises:

[0063] 0. Anti-IgG physisorption of PBS on the glass slide (2 hours);

[0064] 1. Baseline (45 minutes): injection of HEPES buffer 1 (i s = 150 mM, pH = 7.4);

[0065] 2. Rinsing (20 minutes): injection of HEPES buffer 1;

[0066] 3. Negative control experiment (45 minutes): incubation of IgM (1 fM) in HEPES buffer 1;

[0067] 4. Rinsing (20 minutes): injection of HEPES buffer 1;

[0068] 5. Detection at 100 zM (45 minutes): incubation of IgG (100 zM) in HEPES buffer 1;

[0069] 6. Rinsing (20 minutes): injection of HEPES buffer 1;

[0070] 7. Detection at 1 aM (45 minutes): incubation of IgG (1 aM) in HEPES buffer 1;

[0071] 8. Rinsing (20 minutes): injection of HEPES buffer 1;

[0072] 9. Detection at 100 aM (45 min): Incubate IgG (100 aM) in HEPES buffer 1;

[0073] 10. Wash (20 min): Inject HEPES buffer 1;

[0074] 11. Detection at 100 fM (45 min): Incubate IgG (100 fM) in HEPES buffer 1;

[0075] 12. Wash (20 min): Inject HEPES buffer 1;

[0076] 13. Detection at 100 fM (45 min): Incubate IgG (100 zM) in HEPES buffer 1;

[0077] 14. Wash (20 min): Inject HEPES buffer 1.

[0078] Figure 3 The SPR angle shift (Δθ) vs. time for protocol A is shown. Each step of the protocol is marked by the relevant number in the list above, and the baseline is the SPR signal level in HEPES buffer 1 of the stabilized original anti-IgG layer.

[0079] In Figure 3 (a), the entire sequence of steps 1-14 in the list above is shown, while in Figure 3 (b), an enlarged view of steps 1-7 is provided. During each injection, a spike in the SPR signal (change in angle θ, i.e. Δθ) can be seen, which returns to baseline within about 2-3 min. It is clear and expected that Figure 3 (a) shows how the baseline is not significantly deviated until the injection of the 100 nM IgG solution. A close-up examination in the fM range Figure 3 (b)) shows that after step 6, i.e. after injection and wash of the 100 zM IgG, the SPR angle shift is negligible with Δθ 100zM = 0.005° ± 0.002° (average of 4 replicates). This SPR angle shift is even more pronounced if compared to the signal level of the negative control (NC) experiment after washing (step 4), which is Δθ NC = 0.002° ± 0.002°. In fact, the magnitude of the instrumental error is Δθ 100zM - Δθ NC = 0.003° ± 0.003°, where the error is calculated as the propagated error.

[0080] However, as indicated by the relevant arrows, no significant further change is seen after step 6. This means that no significant change occurs when 100 zM IgG is injected in the cell compared to the signal level of the negative control experiment. Overall, the data show that not enough antigen molecules are provided in solution to build an IgG layer on top of the captured anti-IgG until the 100 nM IgG solution is injected. Only at this IgG concentration, the SPR evanescent wave reveals a significant change of the dielectric function, which is related to the deposition of an IgG layer on top of the anti-IgG capture layer 2.

[0081] A second experiment, called "protocol B", also includes a step of incubation in 0.1 ml of HEPES buffer 1, which provides physiological conditions suitable for performing IgG binding. These steps last 45 minutes. The washes are performed in a solution called "HEPES buffer 2" (i s = 150 mM, pH = 7.4) with both the pH and the ionic strength decreased. This protocol includes the injection of 3 ml of solutions characterized by different salinities (HEPES buffer 1 and HEPES buffer 2). This leads to a change in the dielectric function measured in the cell, which can be attributed to the solution and the capture layer. In order to measure the change in the delta theta shift related to the capture layer alone after washing, a step is introduced, which is called "restoring physiological conditions". Thus, protocol B starts with the injection of 0.1 ml of HEPES buffer 1 (with IgM or IgG) for 45 minutes of incubation. Then, HEPES buffer 2 (without dissolved antigen) is injected and left for 20 minutes to wash the capture layer. Here, a large shift in the delta theta angle is measured. After that, a new batch of 0.1 ml of HEPES buffer 1 is injected and left for 5 minutes to restore the physiological conditions, accompanied by a shift in the delta theta angle related to the solution change. More schematically, protocol B is performed according to the following steps:

[0082] 0. Injection of 0.4 cm 2 Physical adsorption of anti-IgG in PBS (2 hours) on the part of the slide;

[0083] 1. Baseline (45 minutes): injection of HEPES buffer 1 (i s = 150 mM, pH = 7.4);

[0084] 2. a) Rinse (20 minutes): injection of HEPES buffer 2 (i s = 5 mM, pH = 6);

[0085] b) Restore physiological conditions (5 minutes): injection of HEPES buffer 1;

[0086] 3. Negative control experiment (45 minutes): incubation of IgM (1 fM) in HEPES buffer 1;

[0087] 4. a) Wash (20 min): injection of HEPES buffer 2;

[0088] b) Restore physiological conditions (5 min): injection of HEPES buffer 1 ;

[0089] 5. Detection at 100 zM (45 min): incubation of IgG (100 zM) in HEPES buffer 1 ;

[0090] 6. a) Wash (20 min): injection of HEPES buffer 2;

[0091] b) Restore physiological conditions (5 min): injection of HEPES buffer 1 ;

[0092] 7. Detection at 1 aM (45 min): incubation of IgG (1 aM) in HEPES buffer 1 ;

[0093] 8. a) Wash (20 min): injection of HEPES buffer 2;

[0094] b) Restore physiological conditions (5 min): injection of HEPES buffer 1 ;

[0095] 9. Detection at 100 aM (45 min): incubation of IgG (100 aM) in HEPES buffer 1 ;

[0096] 10. a) Wash (20 min): injection of HEPES buffer 2;

[0097] b) Restore physiological conditions (5 min): injection of HEPES buffer 1 ;

[0098] 11. Detection at 100 fM (45 min): incubation of IgG (100 fM) in HEPES buffer 1 ;

[0099] 12. a) Wash (20 min): injection of HEPES buffer 2;

[0100] b) Restore physiological conditions (5 min): injection of HEPES buffer 1 ;

[0101] 13. Detection at 100 nM (45 min): incubation of IgG (100 zM) in HEPES buffer 1.

[0102] 14. a) Wash (20 min): injection of HEPES buffer 2;

[0103] b) Restore physiological conditions (5 min): injection of HEPES buffer 1 ;

[0104] Figure 4Data related to protocol B are provided. It is worth noting that in protocol A, both detection (affinity binding) and washing are performed in the same HEPES buffer 1 environment, while in protocol B, detection is still performed in HEPES buffer 1, but washing includes HEPES buffer 2, which results in a decrease in both pH and i s .

[0105] As it is evident from Figure 4 (a) and Figure 4 (b), after washing in HEPES buffer 2 and restoring physiological conditions in HEPES buffer 1 (step 4(b)), for the negative control experiment a negligible shift from baseline (SPR signal of the original anti-IgG layer in HEPES buffer 1) of Δθ NC = 0.002° ± 0.002° is observed. In contrast, when the anti-IgG layer is incubated with a 100 zM IgG solution, a considerable shift from baseline is observed. After washing and restoring physiological conditions (step 6(b)), Δθ 100zM = 0.019° ± 0.002° is measured, resulting in Δθ 100zM - Δθ NC = 0.017° ± 0.002°. In 0.1 ml of 100 zM, there are only 10 ± 3 IgG molecules present, which can only bind to a few of the 10 11 anti-IgG present in the capture layer. Therefore, the change in the dielectric function at the interface with the gold layer 3 is related to the changes occurring in the anti-IgG layer. When incubated with a solution of higher anti-IgG concentration up to 100 fM, the signal shift is slightly increased. At 100 nM, a much larger SPR Δθ shift is recorded. Here there is a new layer of IgG binding to the anti-IgG, which can be stated as follows:

[0106] i) Washing with HEPES buffer 2 generates a dielectric change mechanism that includes the anti-IgG layer alone, which occurs when only a small number of binding events are involved;

[0107] ii) Capturing the entire IgG layer at much higher concentrations is still possible because the anti-IgG layer does not denature, so it can still bind about 10 11 IgG.

[0108] In order to better characterize Figure 4 the invention disclosed in protocol B, two variants of protocol B were implemented, which are referred to as "protocol C" and "protocol D". Specifically, "protocol C" and "protocol D" use a "HEPES buffer 3" (i s = 5 mM, pH = 7.4) and a "HEPES buffer 4" (i sother buffer solutions (150 mM NaCI, pH = 6) were flushed. In both cases, the incubation was always performed in HEPES buffer 1. As in the case of protocol B, the incubation antigen lasted 45 minutes, again including 0.1 ml of solution, the flushing lasted 20 minutes (again including 3 ml of solution), and the restoration of the physiological conditions lasted 5 minutes (again including 3 ml of solution).

[0109] The data related to protocols C and D are shown in Figure 5 (a) and Figure 5 (b), respectively, where, for the sake of clarity, only steps 1-7 are shown. In both cases, the injection of a 100 nM solution produced a large angular shift (data are shown only in Figure 8 , see below). For Figure 5 (a) (protocol C), no significant shift with respect to the baseline (compared to the instrumental error, ΔΘ 100zM - ΔΘ NC = 0.003° ± 0.002°) was recorded until the IgG (100 nM) solution was injected into the cell 3. As expected, here, a signal comparable to step 14 of Figure 4 (a) can be seen (not shown). The signal at 100 nM can also be seen in the data collected with protocol D. Therefore, it is also demonstrated that, for protocols C and D, also the capture layer did not denature.

[0110] It can be concluded that Figure 5 the data presented in (a) show how flushing in HEPES buffer 3 (involving a decrease in ionic strength, but not in pH, compared to HEPES buffer 1) does not trigger a SPR ΔΘ shift compared to the baseline until the injection of IgG at a 100 nM concentration. This means that the single variation of the ionic strength i s produces an effect similar to that seen when HEPES buffer 1 is also used as flushing solution.

[0111] Figure 5 A different situation can be seen in protocol D, which involves flushing with HEPES buffer 4, where only the pH is lowered ( Figure 5 (b)) compared to HEPES buffer 1. Here ΔΘ 100zM - ΔΘ NC = 0.037° ± 0.011° indicates that a considerable SPR signal shift (step 4a) compared to the baseline and to the negative control experiment is recorded. In this case, the ΔΘ shift starts to increase as soon as a small amount of IgG is injected, which indicates that the effect of flushing at a lower pH is triggered by the exposure that takes place before the affinity binding.

[0112] To assess the generality of this effect, different buffer solutions and rinse solutions than HEPES were tested. For this, phosphate buffered saline (PBS, at i s = 163 mM and pH = 7.4) solution was used for antigen binding, while deionized water (i s = approximately 5 mM and pH = approximately 5.5) was used for washing. This is referred to as "protocol E". Indeed, the PBS buffer comprises sodium phosphate dibasic, sodium chloride, potassium chloride and potassium phosphate monobasic, so it is more complex than the HEPES-based buffer currently used. On the other hand, deionized water is not a buffer. However, this combination allows for more drastic changes in pH and ionic strength. This more extreme binding and washing protocol has been tested on both physisorbed anti-IgG and nanostructured anti-IgG AuNP-based capture layers.

[0113] Gold-coated glass slides covered with a physisorbed original anti-IgG layer from PBS or with anti-IgG AuNPs were placed in the SPR cell 3. As for all other protocols, the negative control experiment consisted in injecting 1 fM of non-binding IgM antigen in PBS solution for 45 minutes incubation. Then the sensor surface was rinsed by injecting 0.1 ml of deionized water and left for 20 minutes. After that, to restore physiological conditions, 0.1 ml of PBS solution was injected and left for 5 minutes. The experiment continued with the injection of IgG solutions in PBS buffer 1 (0.1 ml) over a concentration range spanning 10 zM to 100 nM. Each incubation of antigen lasted 45 minutes, followed by a rinsing step in deionized water for 20 minutes and reconstitution of physiological conditions in PBS solution for 5 minutes.

[0114] SPR data obtained according to protocol E show in Figure 6 , in this case, only steps 1-8 are shown. Similar to the previous protocols, a large signal was detected at 100 nM (not shown). Each step of this protocol is labeled with the relevant number, and the baseline is the SPR signal level in PBS of the original anti-IgG layer. In Figure 6 (a), the sequence of steps is shown for a pure physisorbed anti-IgG layer, while in Figure 6 (b), the same experiment was performed on a physisorbed anti-IgG layer on gold nanoparticles forming a nanostructured film on the SPR glass slide.

[0115] In both graphs, after injection of non-bound IgM (1 fM), a rather stable signal was recorded. Then washed with water and recovered in PBS, the signal was set at a new level below baseline. When a 10 zM solution was injected, a very correlated signal change was recorded, which was stable after washing with water and recovery in PBS. A further minor change was observed at 100 zM. Especially on the pure physically adsorbed anti-IgG layer at 10 zM, i.e. when only one IgG was present in solution, a very high signal was recorded. Here Δθ 10zM -Δθ NC = 0.016° ± 0.004°.

[0116] In "protocol F", the pH was increased to 8.2, while i s was kept constant at 150 mM ("HEPES buffer 5"). Typically, incubations were performed in HEPES buffer 1. As in the case of protocol B, the antigen was detected for 45 minutes, washed for 20 minutes, and recovered in physiological conditions for 5 minutes. Figure 7 Data related to protocol F are shown, where only steps 1-8 are shown, and a rather large Δθ 10zM -Δθ NC = 0.016° ± 0.003°. Also in this case, IgG (100 nM) returned a signal comparable to step 14 (not shown). Figure 4 (a) - step 14.

[0117] Data related to all experiments measured in protocols A, B, C, D, E, and F are shown in Figure 8 Figure 6. Δθ signals of all negative control experiments (17 replicates) are given as mean (open star), while error bars (falling within the symbol area) are one standard deviation. LOD and LOI levels are shown, and all data are averaged over at least 3 replicates, error bars as one standard deviation. Responses at the lowest IgG concentrations (10 zM, 100 zM, and 100 aM) are plotted on the left, while responses recorded at IgG (100 nM) are plotted on the right (note x and y axis breaks). Data are shown as a shift relative to the negative control experiment level. Clearly, all protocols resulted in a large shift Δθ = 0.153° ± 0.022° at 100 nM without exception. This signal originates from a change in the dielectric function resulting from the construction of an IgG layer on top of the anti-IgG. This is referred to here as the "IgG layer mechanism", and by the de Feijter equation 28,21 , the IgG coverage at 100 nM can be estimated to be (3.3 ± 0.8) · 10 11 / cm 2 This coverage increases to (3.8 ± 0.2) · 1011 / cm 2 The number of IgG molecules captured at these concentrations is comparable to the number of anti-IgG molecules captured below.

[0118] At concentrations lower than 1 fM, protocols B, D, E and F lead to a SPR Δθ shift beyond the LOI, also at 10 zM, while protocols A and C return a signal at the LOD or lower. At these very low concentrations, no IgG layer can be formed, since only 1 ± 1 IgG is present at 10 zM, while 10 ± 3 molecules are found in 0.1 ml (100 zM) (10 4 at 100 aM). These molecules are involved in the binding to an equal number of anti-IgG capture antibodies. The conformational change associated with the binding of such a small number of IgG antigens to the anti-IgG capture antibodies requires a negligible change in the dielectric function, since the number of bindings that occur is very small, involving a very small fraction of the 10 11 anti-IgG of the capture layer 2. Despite this, a significant SPR Δθ shift is detected, here referred to as "single molecule mechanism". The signal recorded with protocols B, E and F, and more significantly with protocol D, can be attributed to a change in the dielectric function that occurs in the anti-IgG layer 2, which becomes detectable due to the amplification process. Effect of buffer ionic strength and pH variation

[0119] In order to better understand the essence of the present application, which enables the reliable measurement of a single antigen in 0.1 ml, an experimental design was set up according to the protocols given in Figure 9 Table 1. The study included the results of protocols A, B, C and D, since these protocols include a HEPES-based buffer based on homology. The measured Δθ shifts, listed in Table 2, were estimated as the difference between the baseline value and the level of the signal after the rinse (protocol A) or the physiological conditions (protocols B, C and D).

[0120] According to Table 2 2 The experimental design was carried out according to a full factorial method to study the domain, in which protocols A, B, C and D represent the angles of the domain in which the SPR angle shift Δθ has been measured.2 2 The factorial design is based on two variables, namely the decrease in ionic strength (Δi s ) and the decrease in pH (ΔpH, leading to an increase in H + concentration).

[0121] HEPES buffer 1 was always used for the incubation step, while the other HEPES buffers 2, 3, 4 were used for the rinse. According to a two-level design, the two quantitative variables were coded as X1 and X2, representing Δi s and ΔpH, respectively, which were set at two levels, namely -1 (no change) and +1 (full change).

[0122]

[0123]

[0124] Table 2

[0125] From a geometrical point of view, 2 2 Factorial design investigates the corners of the square, which means that the model developed in this paper allows two variables to vary simultaneously. The output of the experiments performed on the corners of the experimental domain (which are marked by black circles (in Figure 9 ) has been used to feed the model. Moreover, two experiments have been performed at the center point, as shown by the grey circles in Figure 9 , to evaluate the predictive ability of the model.

[0126] The analysis based on multiple linear regression provides the following model relating the SPR angle shift Δθ to the detection at 100 zM and the two coded variables, as described in equation 1 below:

[0127]

[0128] Where the usual conventions for the significance levels are indicated: *= p < 0.05, **= p < 0.01, ***= p < 0.001. These p-values define the confidence interval of the researcher's t-test, which is used to assess the significance of each coefficient in the regression model.

[0129] The selected design, with 15 degrees of freedom, gives a maximum leverage of 0.33 over the whole experimental domain. The leverage value, multiplied by the experimental variance, corresponds to the variance of the estimated response at that point. A leverage value of 1 means that the model will predict the response with the same accuracy as the experiment, while a leverage value < 1 means that the response can be predicted with better accuracy than the experimental data collected under the same conditions. Therefore, it can be concluded that the proposed model has an excellent predictive ability.

[0130] Moreover, the experimental value of the SPR angle shift Δθ recorded at the center point is (0.017 ± 0.005) degrees, which is not significantly different from the predicted value (0.017 degrees). Therefore, the model is valid and acceptable over the whole experimental domain.

[0131] As Figure 10(a) reported in Figure 2(a), the coefficients of the model are all significant terms, although the linear term of X2 has a larger absolute value than the other terms. Indeed, decreasing the pH of the washing buffer from 7.4 to 6 leads to an increase of one order of magnitude of the SPR angular shift. On the other hand, the influence of the washing buffer with a lower ionic strength on the SPR angular shift is limited. The coefficient of each term represents the change of the average response associated with an increase of one coded unit of this term, while the other terms remain constant. By changing the ionic strength of the washing buffer from 150 mM to 5 mM, the SPR angular shift changes by less than the LOD, while the pH change is higher than the LOI, and is therefore significant.

[0132] Figure 10 The isoreponse contour plot reported in Figure 2(b) provides information on the interaction between the two variables X1 and X2. The geometric shape of the linear model without interaction is a plane, leading to isoresponses lines that are parallel, while if there is a relevant interaction, the contour plot shows a twisted plane, with isoresponses lines that are not parallel. This means that the influence of the lower ionic strength of the washing buffer becomes more relevant only at higher pH differences. In contrast, at much lower pH of the washing buffer, the ionic strength has almost no influence. Moreover, Figure 10 (b) confirms that the conditions that correspond to the preference (and provide the best results) for obtaining the SPR response at very low IgG concentrations (in the experimental domain of the study) correspond to a decrease of the pH of the washing buffer.

[0133] On the other hand, the change of the ionic strength of the washing buffer does not significantly improve the SPR angular shift. This is in agreement with the results already published on the influence of pH and ionic strength on protein-protein interactions, unfolding and aggregation of IgG antibodies. 34

[0134] For the SPR angular shift ΔΘ recorded at 100 aM, the analysis based on the multiple linear regression provides the following model, which relates to the two coded variables in the following equation 2:

[0135]

[0136] Equation 2 is in agreement with the main features of the model calculated for the detection at 100 zM. The selected design, calculated with 15 degrees of freedom, also leads to a maximum leverage value of 0.33. Moreover, the experimental value of the SPR angular shift recorded at the center point is (0.020 ± 0.015) degrees, which is not significantly different from the predicted value (0.024). Therefore, this model is valid and acceptable in the whole experimental domain. Figure 10 The model coefficients reported in Figure 2(c) are all significant terms, even if in this case the linear term of X2 has a larger absolute value than the other terms. Figure 10(d) the isoresponses contour plot reported in (d) shows that, even in this case, the conditions that are preferred (and that provide the best results) for enhancing the SPR angular shift in the experimental domain investigated require a decrease in the pH of the washing buffer, while a decrease in the ionic strength of the washing buffer does not improve the SPR angular shift.

[0137] The analysis of the SPR angular shifts recorded upon exposure to 100 nM IgG returns a model given by equation 3 below:

[0138]

[0139] Figure 10 (e) shows that X1, related to the ionic strength variation, is the only significant coefficient in this case, meaning that the dependence on the ionic strength of the washing buffer is the only important aspect. Also in this case, the experimental value of the SPR angular shift ΔΘ recorded at the center point is (0.118 ± 0.040) degrees, which is not significantly different from the predicted value (0.135). Notably, in this case, a decrease in the ionic strength of the washing buffer from 150 mM to 5 mM reduces the SPR angular shift by about 15%, while the effect of the pH becomes negligible.

[0140] The need for very different models at low concentrations (100 zM and 100 aM) and at 100 nM fully justifies the consideration of two completely different mechanisms in the analysis of the SPR IgG in the presence of the capture anti-IgG layer, i.e. a "single molecule mechanism" (10 zM - 10 nM) in which the change in the dielectric function involves an electrostatic rearrangement in the anti-IgG layer 2 and an "IgG layer mechanism" (> 10 nM) in which the dielectric function is changed by building a new layer on top of the capture anti-IgG layer. Finally, as shown in Figure 10 (f) shows that the SPR angular shifts recorded during the negative control experiments are negligible, independently of the measurement protocol, i.e. neither the ionic strength nor the pH of the washing buffer is important.

[0141] Propagating mechanism enabling amplification of single molecule affinity binding SPR signal

[0142] The information gathered so far allows to deepen the understanding of the propagation mechanism that enables to amplify the signal in the presence of some single molecule affinity binding events. In particular, the following evidence, summarized in figure (8), can be recalled:

[0143] - Washing the detection interface with a solution with a lower or higher pH than the physiological buffer associated with the event (HEPES-based or PBS) is essential to achieve label-free optical detection of small numbers of affinity binding events on a millimeter-wide interface. A ApH from 7.4 to 5.5-6.0 or to 8.2 is the preferred option to achieve this.

[0144] - The single-molecule mechanism in the label-free optical analysis according to the application necessarily requires a change in the dielectric function of the capture layer comprising several trillion highly packed antibodies. Their density is 10 4 / μm 2 , as high as the density of proteins / receptors on the cell surface that can be detected or tracked in single molecules. 35,36

[0145] - Bioelectronic single-molecule detection with transistors 2,19–22 or with Kelvin probe force microscopy 23 requires the same ApH between washing of the capture layer and binding under physiological conditions. Thus, the washing step at lower or higher pH and the single-molecule affinity binding under physiological conditions (schemes B, D, E and F) (rather than any applied electric or optical field) are the basis of the single-molecule label-free optical mechanism according to the application.

[0146] - Reduction of the ionic strength of the washing solution / buffer (Δi s ) cannot essentially achieve the single-molecule mechanism. However, when associated with a low pH change, it can contribute to increasing the SPR signal towards IgG single-molecule detection.

[0147] - Multiple repeated exposure of the anti-IgG capture layer to solutions that go from physiological conditions (i s = 150 mM, pH = 7.4) to lower i s and lower pH values or higher pH values does not cause denaturation of these antibodies. It is demonstrated that anti-IgG is able to capture IgG, which forms a layer as dense as anti-IgG. Moreover, the number of IgG captured in schemes B, D, E, F is comparable to the number deposited on anti-IgG in experiments that do not perform exposure to non-ideal physiological conditions (see Figure 2 and data related to scheme A).

[0148] All this evidence can be used to delve deeper into the amplification mechanism that makes a single affinity binding event detectable on a large interface. So far, a domino-like propagation process has been evoked, initially requiring a self-assembled monolayer (mixed SAM of activated and blocking 3-mercaptopropionic acid and 11-mercaptoundecanoic acid) 19Antibodies are covalently attached to the transistor Au gate electrode. The hydrogen bond network throughout this chemical SAM is believed to be responsible for the cooperative interaction between SAM chains.

[0149] It is hypothesized that defects in the hydrogen bond network are caused by local electrostatic changes resulting from the single affinity binding. Such changes are ultimately propagated due to the network provided by the applied electric field associated with the electronic detection. 2,20

[0150] Further data suggest that monomolecular detection is also possible when the capture antibody layer is physically adsorbed 21,23 on the gate electrode. Thus, the hydrogen bond network associated with the SAM is not essential for propagation. On the other hand, the data presented here demonstrate that a gating field is not required either. Under these conditions, to fully detail the actual propagation mechanism, infrared spectroscopy of the capture layer at each step of protocols B, D, E, F is required to obtain molecular level information. On the other hand, based on data published in the literature, some reasonable hypotheses can be made about the nature of the propagation / amplification process. The main experimental results are summarized as follows:

[0151] - the capture antibody (or more generally, the capture specific binding pair forming substance) layer will be exposed to a pH change,

[0152] - the layer that changes its dielectric properties includes a highly densely packed protein,

[0153] - most capture antibodies (or more generally, the capture specific binding pair forming substance) do not undergo irreversible denaturation when exposed to non-physiological conditions.

[0154] These evidences point to one of the following mechanisms or a combination thereof:

[0155] - conformational propagation related to the allosteric phenomenon - the function of many proteins is regulated by allosteric processes, by which a binding event at one site of a protein influences the behavior of another distant site. This long-range communication between protein sites represents a fundamental process of signal transduction. Allosterism is generally believed to be confined within a single protein. However, a highly packed tight antibody layer can allow allosteric interactions between adjacent proteins to occur. In fact, conformational changes also show propagation through the elongation system of protein molecules in bacterial chemotaxis receptors, muscle ryanodine receptors, and actin filaments. The statistical mechanics of idealized linear two-dimensional arrays of allosteric proteins has shown that arrays of tightly packed units can exhibit large-scale integrated behavior in similar Ising models. 37

[0156] - Protein aggregation (amyloid proteins) - Proteins form aggregates, a process that involves different stages: (i) partial monomer unfolding or misfolding; (ii) reversible self-association of folded or unfolded monomers; (iii) formation of net irreversible aggregates through strong, stable inter-protein contacts (e.g. hydrogen bonds and burial of hydrophobic amino acids) - this often leads to inter-protein beta-sheet structures; (iv) subsequent growth by aggregation of additional monomers; (v) growth by association of aggregates to form larger soluble and / or insoluble aggregates. Due to the role of different types of antibody aggregates in disease progression, there is currently less understanding of stages (iii) to (v), whereas some aspects of aggregate growth and higher order assembly have been studied for selected systems (e.g. Alzheimer's disease antibodies). Changes in protein-protein interactions, protein unfolding and non-native aggregation of a range of human IgGl antibodies were assessed as a function of pH and solution ionic strength. 34,38

[0157] - Resonance-assisted hydrogen bonds - Hydrogen bonds play a crucial role in the formation of protein secondary structures (e.g. alpha-helices and beta-sheets). A special property of these hydrogen bonds is their cooperativity, which manifests itself in an additional energy gain upon extension of the hydrogen bond unit ((N-H···O=C) n ). Some important protein behaviors, which include the aggregation of beta-amyloid peptides that can lead to Alzheimer's disease or mad cow disease, are attributed to this cooperativity. The origin of cooperativity has been studied mainly theoretically, with a typical hypothesis being that resonance-assisted hydrogen bonds are involved in the formation of protein secondary structures. Under this hypothesis, the lone pair of electrons on the nitrogen atom and the pi bond of the carbonyl group in the peptide bond resonate to form an enolic structure. 39

[0158] Based on the information collected from the literature, it can be envisaged that a network of hydrogen bonds already exists in the disordered deposit of physically adsorbed antibodies that connects the entire system. When the physiological conditions remain unchanged, the system is not triggered to change its dielectric function or its surface potential. In the presence of affinity binding, the conformational electrostatic changes of the single anti-IgG involved do not propagate to the other units. If the capture layer 2 is exposed to solutions of different pH and ionic strength, but without affinity binding occurring, no significant changes in the dielectric function of the anti-IgG layer are detected. Very surprisingly, when the capture antibody layer 2 is exposed to solutions of different pH (and ionic strength, although it has a minor critical influence) and affinity binding occurs, a considerable change in the dielectric function of the anti-IgG layer is detected. In this case, it can be assumed that the change in pH and the presence of local electrostatic changes (dielectric rearrangement of the dipoles due to conformational changes) create propagation conditions for the reorientation of the dipole moment, which can lead to significantly more interactions between proteins (possibly through their a-helices and b-sheets). This leads to phenomena such as conformational propagation, which makes the system reach a more stable conformation.

[0159] The generalization of the present invention can be referred to here as the formation of a number of specific binding pairs to substances - such as antigen / antibody pairs - which have been studied to date with bioelectronic devices, in which the following systems (only antibodies) are mentioned: - anti-human immunoglobulin G and anti-human immunoglobulin M, anti-C-reactive protein, anti-HIV1 p24, anti-MUC1, anti-CD55, anti-S1 covid19, purified IgG against Xylella fastidiosa, anti-IL6 (monoclonal cytosine). In addition, single-molecule electronic reactions against streptavidin, avidin and neutravidin have also been demonstrated. In addition, label-free electronic single-molecule detection has also been demonstrated with peptides, DNA markers (such as KRAS) and microRNAs (such as miR-182-5p and SARS-CoV-2 encoded microRNAs that regulate viral replication).

[0160] Single molecules are detected by label-free optical detection in 10 -20 M under simulated real-world conditions.

[0161] In order to demonstrate that protocol D, which provides the highest response to a single binding event, can be used in real (i.e. non-experimental) environments, the following proof-of-principle test was designed and carried out. An i sA solution of MERS-CoV spike S1 antigen at a concentration of 1 fM was added to a HEPES buffer 1 solution at a pH of 7.4, simulating the physiological conditions of human plasma. This protein acts as an interfering substance and does not bind to the anti-p24 capture layer. MERS-CoV spike S1 HEPES buffer 1 was also supplemented with 10zM and 100zM of HIV-1 p24, which binds to anti-p24 with high affinity. The HIV-1 p24 protein, which selectively binds to the anti-p24 capture antibody, is a hallmark of HIV-1 infection, and its early detection is highly relevant for screening asymptomatic patients. For this purpose, "Scheme D" was used for the analysis according to the following steps:

[0162] 0.04 cm on the slide 2 Anti-p24 physical adsorption with PBS was performed on the sections (2 h);

[0163] 1. Baseline (45 minutes): Inject 0.1 ml HEPES buffer 1 (i s =150 mM, pH=7.4);

[0164] 2.a) Wash (20 min): Inject 3 ml HEPES buffer 4 (i s =150 mM, pH=6);

[0165] b) Restoration of physiological conditions (5 min): injection of 3 ml of HEPES buffer 1;

[0166] 3. Negative control experiment with interferors (45 min): incubate MERS-CoV spike S1 (1 fm) in HEPES buffer (0.1 ml injection);

[0167] 4.a) Rinse (20 min): inject 3 ml of HEPES buffer 4;

[0168] b) Restoration of physiological conditions (5 min): injection of 3 ml of HEPES buffer 1;

[0169] 5. Detection at 10zM (45 minutes): Incubate MERS-CoV spike S1 (1fm) with HIV-1p24 (10zM) in 0.1ml HEPES buffer 1;

[0170] 6.a) Rinse (20 min): Inject 3 ml of HEPES buffer 4;

[0171] b) Restoration of physiological conditions (5 min): injection of 3 ml of HEPES buffer 1;

[0172] 7. Detection at 100 zM (45 min): incubation of MERS-CoV Spike S1 (1 fM) with HIV-1 p24 (100 zM) in 0.1 ml of HEPES buffer 1 ;

[0173] 8. a) Wash (20 min): injection of 3 ml of HEPES buffer 4;

[0174] b) Restoring physiological conditions (5 min): injection of 3 ml of HEPES buffer 1.

[0175] The results are shown in Figure 11 (a). It can be seen that detection can be performed at the single molecule level (10 zM) with an angular shift as high as ΔΘ 10zM - ΔΘ NC = 0.006° ± 0.001° and ΔΘ 100zM - ΔΘ NC = 0.010° ± 0.002°.

[0176] In order to better simulate detection under real conditions, the following protocol (modified protocol D) was performed. The method of the application performed under such conditions relies on an already deposited anti-P24 capture layer, which basically means that under real (i.e. non-experimental) conditions, the operator in the test laboratory will be provided with a detection kit comprising a delivered box already functionalized (where the box basically comprises a substrate provided with a). Therefore, the protocol becomes:

[0177] 1. Baseline (45 min): injection of 0.1 ml of HEPES buffer 1 (i s = 150 mM, pH = 7.4), which is referred to as the reference fluid;

[0178] 2. a) Wash (20 min): injection of 3 ml of HEPES buffer 4 (i s = 150 mM, pH = 6);

[0179] b) Restoring physiological conditions (5 min): injection of 3 ml of HEPES buffer 1;

[0180] 5. Detection at 10 zM (45 min): incubation of MERS-CoV Spike S1 (1 fM) with HIV-1 p24 (10 zM) in 0.1 ml of HEPES buffer 1;

[0181] 6. a) Wash (20 min): injection of HEPES buffer 4;

[0182] b) Restoring physiological conditions (5 min):

[0183] injection of 3 ml of HEPES buffer 1.

[0184] At 10 zM a very low change in signal with respect to baseline can be measured as Δθ 10zM = 0.003° ± 0.002°. This demonstrates that, unless a washing step is performed before injecting the affinity antigen, the single-molecule mechanism is not activated.

[0185] To assess whether the washing step should be performed only before detection or only after detection, the following protocol was measured:

[0186] 1. Baseline (45 min): injection of HEPES buffer 1 (i s = 150 mM, pH = 7.4), which is called the reference fluid.

[0187] 5. Detection at 10 zM (45 min): incubation of MERS-CoV Spike S1 (1 fm) with HIV-1 p24 (10 zM) in HEPES buffer 1 ;

[0188] 6. a) Wash (20 min): injection of HEPES buffer 4;

[0189] b) Restore physiological conditions (5 min):

[0190] Injection of HEPES buffer 1 ;

[0191] In this case, a very low change in signal with respect to baseline can be measured at 10 zM as Δθ 10zM = 0.003° ± 0.002°. This demonstrates that, unless a washing step is performed before injecting the affinity antigen, the single-molecule mechanism is not activated.

[0192] To improve the reliability of the binary classification, a machine learning-based classifier can be used. For this purpose, a binary classifier can be designed that is able to read the raw data as shown in Figure 11 directly and produce an output with 0 for negative samples (N) and 1 for positive samples (P). The binary classifier can be trained with a training set comprising about 100 samples, divided between about 50 negative samples and 50 positive samples. The predictive power of the binary classifier can be evaluated using an external test set of about 40 blind samples. Thus, the binary classifier can be specifically designed to reliably distinguish between negative and positive samples with a false positive and false negative error below 1%.

[0193] Therefore, in view of the above, a method for detecting a binding event of a specific binding pair forming substance is defined according to an embodiment of the application, which can be summarized as comprising the steps of:

[0194] - providing a layer of a first specific binding pair forming substance on a substrate, the density of the specific binding pair forming substance being 10 2 - 10 4 / μm 2 of the substrate.

[0195] - providing a housing of the layer of the first specific binding pair forming substance with a first dielectric, the first dielectric comprising a first solution having a first pH value and a first ionic strength value,

[0196] - flushing the housing with a second dielectric comprising a second solution having a second pH value and a second ionic strength value for a first time period, wherein at least the second pH value is different from the first pH value,

[0197] - after said flushing, restoring in the housing the first dielectric having the first pH value and the first ionic strength value,

[0198] - feeding a solution of a second specific binding pair forming substance into the housing of the layer of the first specific binding pair forming substance to cause an interaction between the first specific binding pair forming substance and the second specific binding pair forming substance after the restoration of the first dielectric in the housing, and incubating the housing for a second time period, the interaction comprising a binding event when the first specific binding pair forming substance and the second specific binding pair forming substance provide said specific binding pair,

[0199] - detecting a shift of a parameter representative of the dielectric function of the layer of the first specific binding pair forming substance after the incubation of the housing for the second time period, the shift being defined as a change of the parameter representative of the dielectric function of the layer of the first specific binding pair forming substance from a first value occurring when the housing comprises only the first dielectric after the restoration of the first dielectric to a second value occurring after the incubation of the housing for the second time period,

[0200] - comparing the shift Δθ to a threshold value, in particular a limit of identification LOI, whereby when the shift exceeds the threshold value, at least one binding event occurs between the first specific binding pair forming substance and the second specific binding pair forming substance.

[0201] As disclosed in the present application, the preferred parameter representative of the dielectric function of layer 2 of the first specific binding pair forming species is the SPR angle Q. However, since the detection of the SPR angle Q and the associated shift mainly involves optical techniques, other preferred embodiments can rely on the detection of a parameter representative of the dielectric function of layer 2 by other optical-based techniques or optical-related parameters, typically not requiring the reliance on electrical inputs as in biological assays carried out by potentiometric measurements, for example.

[0202] From the experimental protocols disclosed herein and under the real (non-experimental) detection conditions discussed above, it is evident that the rinsing can be carried out before and after the feeding of the solution of the second specific binding pair forming species to the housing 4. In such cases, the detection of the shift of the parameter representative of the dielectric function of layer 2 of the first specific binding pair forming species (SPR angle Q and shift AQ) is carried out after the rinsing, which is carried out after incubating the housing 4 for the second time period (i.e. after feeding the solution of the second specific binding pair forming species to the housing 4).

[0203] In other embodiments, the rinsing is carried out only before feeding the solution of the second specific binding pair forming species to the housing 4, so that the detection of the shift of the parameter representative of the dielectric function of layer 2 of the first specific binding pair forming species is carried out after incubating the housing 4 for the second time period.

[0204] REFERENCES

[0205] 1. Gooding, J. J. & Gaus, K. Single-Molecule Sensors: Challenges and Opportunities for Quantitative Analysis. Angew. Chemie-Int. Ed. 55, 11354-11366 (2016).

[0206] 2. Macchia, E. et al. Large-Area Interfaces for Single-Molecule Label-free Bioelectronic Detection. Chem. Rev. (2022) doi:10.1021 / acs.chemrev.lc00290.

[0207] 3. De Angelis, F. et al. Breaking the diffusion limit with super-hydrophobic delivery of molecules to plasmonic nanofocusing SERS structures. Nat. Phot onics 5, 682-687 (2011).

[0208] 4. Macchia, E., Manoli, K., Di Franco, C., Scamarcio, G. & Torsi, L. New trends in single-molecule bioanalytical detection. Anal. Bioanal. Chem. 412, 5005-5014 (2020).

[0209] 5. Wu, Y., Tilley, R. D. & Gooding, J. J. Challenges and Solutions in Developing Ultrasensitive Biosensors. J. Am. Chem. Soc. 141, 1162-1170 (2019).

[0210] 6. Mayer, K. M., Hao, F., Lee, S., Nordlander, P. & Hafner, J. H. A single molecule immunoassay by localized surface plasmon resonance. Nanotechnology 21, 255503 (2010).

[0211] 7. Anker, J. N. et al. Biosensing with plasmonic nanosensors. Nanosci. Technol. A Collect. Rev. from Nat. Journals 7, 308-319 (2009).

[0212] 8. Zijlstra, P., Paulo, P. M. R. & Orrit, M. Optical detection of single non-absorbing molecules using the surface plasmon resonance of a gold nanorod. Nat. Nanotechnol. 7, 379-382 (2012).

[0213] 9. Ding, S. Y. et al. Nanostructure-based plasmon-enhanced Raman spectroscopy for surface analysis of materials. Nat. Rev. Mater. 1, (2016).

[0214] 10. Kretschmann, E. & Raether, H. Radiative Decay of Non Radiative Surface Plasmons Excited by Light. Zeitschrift fur Naturforsch.-Sect. A J. Phys. Sci. 23, 2135-2136 (1968).

[0215] 11. Gedig, E.. Surface Chemistry in SPR Technology. in Handbook of Surface Plasmon Resonance. (Royal Society of Chemistry, 2017). doi:10.1039 / 9781788010283.

[0216] 12. Knoll, W. Interfaces and thin films as seen by bound electromagnetic waves. Annu. Rev. Phys. Chem. 49, 569-638 (1998).

[0217] 13. Hifumi, E. et al. Elimination of ingredients effect to improve the detection of anti HIV-1 p24 antibody in human serum using SPR apparatus. Anal. Sci. 18, 863-867 (2002).

[0218] 14. Malmqvist, M. Surface plasmon resonance for detection and measurement of antibody-antigen affinity and kinetics. Curr. Opin. Immunol. 5, 282-286 (1993).

[0219] 15. Neumann, T., Johansson, M. L., Kambhampati, D. & Knoll, W. Surface-plasmon fluorescence spectroscopy. Adv. Funct. Mater. 12, 575-586 (2002).

[0220] 16. Sarcina, L. et al. Surface Plasmon Resonance Assay for Label-Free and Selective Detection of HIV-1 p24 Protein. Biosensors 11, 180 (2021).

[0221] 17. Vashist, S. K., Dixit, C. K., MacCraith, B. D. & O’Kennedy, R. Effect of antibody immobilization strategies on the analytical performance of a surface plasmon resonance-based immunoassay. Analyst 136, 4431-4436 (2011).

[0222] 18. Nguyen, H. H., Park, J., Kang, S. & Kim, M. Surface plasmon resonance: A versatile technique for biosensor applications. Sensors (Switzerland) 15, 10481-10510 (2015).

[0223] 19. Macchia, E. et al. Label-Free and Selective Single-Molecule Bioelectronic Sensing with a Millimeter-Wide Self-Assembled Monolayer of Anti-Immunoglobulins. Chem. Mater. 31, 6476-6483 (2019).

[0224] 20. Macchia, E. et al. Single-molecule detection with a millimetre-sized transistor. Nat. Commun. 9, (2018).

[0225] 21. Macchia, E. et al. A handheld intelligent single-molecule binary bioelectronic system for fast and reliable immunometric point-of-care testing. Sci. Adv. 8, 1-15 (2022).

[0226] 22. Guo, K. et al. Rapid single-molecule detection of COVID-19 and MERS antigens via nanobody-functionalized organic electrochemical transistors. Nat. Biomed. Eng. 5, 666-677 (2021).

[0227] 23. Di Franco, C. et al. Extended Work Function Shift of Large-Area Biofunctionalized Surfaces Triggered by a Few Single-Molecule Affinity Binding Events. Adv. Mater. Interfaces (2022) doi:10.1002 / admi.202201829.

[0228] 24. Macchia, E. et al. Why a Diffusing Single-Molecule can be Detected in Few Minutes by a Large Capturing Bioelectronic Interface. Adv. Sci. (2022) doi:10.1002 / advs.202104381.

[0229] 25. Singh, G. P. & Sardana, N. Smartphone-based Surface Plasmon Resonance Sensors: a Review. Plasmonics 17, 1869-1888 (2022).

[0230] 26. Winspall 3.02 software. 3 http: / / www.res-tec.de / downloads.html.

[0231] 27. Refractive Index reference value. https: / / refractiveindex.info / .

[0232] 28. De Feijter, J. A., Benjamins, J. & Veer, F. A. Ellipsometry as a tool to study the adsorption behavior of synthetic and biopolymers at the air-water interface. Biopolymers 17, 1759-1772 (1978).

[0233] 29. Liu, G. Y. & Amro, N. A. Positioning protein molecules on surfaces: Ananoengineering approach to supramolecular chemistry. Proc. Natl. Acad. Sci. U. S. A. 99, 5165-5170 (2002).

[0234] 30. Thompson, M., Ellison, S. L. R. & Wood, R. Harmonized guidelines for single-laboratory validation of methods of analysis (IUPAC Technical Report). Pure Appl. Chem. 74, 835-855 (2002).

[0235] 31. Cowsill, B. J. et al. Interfacial structure of immobilized antibodies and perdeuterated HSA in model pregnancy tests measured with neutron reflectivity. Langmuir 30, 5880-5887 (2014).

[0236] 32. Geerligs, I. E. J., Beijnen, J. H., Bekers, O., and U. & M., W. J. Quality control of protein and peptide drugs: Monoclonal antibodies and some biological response modifiers derived by recombinant DNA technology. Drug De V. Ind. Pharm 19, 33-84 (1993).

[0237] 33. Boswell, C. A. et al. Effects of charge on antibody tissue distribution and pharmacokinetics. Bioconjug. Chem. 21, 2153-2163 (2010).

[0238] 34. Sahin Erinc, Grillo O. Adeola, Melissa D. Perkins, R. J. C. Comparative Effects of pH and Ionic Strength on Protein–Protein Interactions, Unfolding, and Aggregation for IgG1 Antibodies. J. Pharm. Sci. 99, 4830-4848 (2010).

[0239] 35. Pichlo, M. et al. High density and ligand affinity confer ultr asensitive signal detection by a guanylyl cyclase chemoreceptor. J. Cell Biol. 206, 541-557 (2014).

[0240] 36. Rieke, F. & Baylor, D. A. Single-photon detection by rod cells of the retina. Rev. Mod. Phys. 70, 1027-1036 (1998).

[0241] 37. Duke, T. A. J., Le Novere, N. & Bray, D. Conformational spread in a ring of proteins: A stochastic approach to allostery. J. Mol. Biol. 308, 541-553 (2001).

[0242] 38. Hartl, F. U., Bracher, A. & Hayer-Hartl, M. Molecular chaperones in protein folding and proteostasis. Nature 475, 324-332 (2011).

[0243] 39. Zhou, Y. et al. Evidences for Cooperative Resonance-Assisted Hydrogen Bonds in Protein Secondary Structure Analogs. Sci. Rep. 6, 1-8 (2016).

Claims

1. A method for detecting at least one binding event of a specific binding pair-forming substance, comprising: - providing a layer (2) of a first specific binding pair forming substance on a substrate (3), wherein the density of the specific binding pair forming substance is 10 2 -10 4 / μm 2 The substrate (3), - providing the housing (4) of the layer (2) of said first specific binding pair forming substance with a first dielectric comprising a first solution having a first pH value and a first ionic strength value, - flushing the housing (4) with a second dielectric comprising a second solution having a second pH value and a second ionic strength value for a first period of time, wherein at least the second pH value is different from the first pH value, - after said flushing, restoring said first dielectric with a first pH value and a first ionic strength value in said housing (4), - feeding a solution of a second specific binding pair forming substance into the housing (4) of the layer (2) of the first specific binding pair forming substance to cause an interaction between the first specific binding pair forming substance and the second specific binding pair forming substance after said restoring the first dielectric in said housing (4), and incubating the housing (4) for a second period of time, the interaction comprising a binding event when the first specific binding pair forming substance and the second specific binding pair forming substance provide said specific binding pair, - detecting a shift (Δθ) of a parameter (θ) representing the dielectric function of the layer (2) of the first specific binding pair forming substance after said culturing housing (4) for a second period of time, the shift being defined as a change in the parameter (θ) representing the dielectric function of the layer (2) of the first specific binding pair forming substance from a first value occurring when said housing (4) comprises only said first dielectric after said restoring of said first dielectric to a second value occurring after said culturing of said housing (4) for a second period of time, - comparing the offset (Δθ) to a threshold value (LOI), whereby when the offset exceeds the threshold value, at least one binding event has occurred between the first specific binding pair forming species and the second specific binding pair forming species.

2. The method according to claim 1 , wherein the flushing is performed before and after the feeding of the solution of the second specific binding pair forming substance, and wherein the detecting of the shift in the parameter representing the dielectric function of the layer (2) of the first specific binding pair forming substance is performed after the flushing after the culture housing (4) has been in place for the second period of time.

3. The method according to claim 1 , wherein the flushing is performed only before the feeding of the solution of the second specific binding pair forming substance, and wherein the detecting of the shift in the parameter representing the dielectric function of the layer ( 2 ) of the first specific binding pair forming substance is performed after the culture housing ( 4 ) has been in place for a second period of time.

4. The method according to any one of the preceding claims, wherein the concentration of the solution of the second specific binding pair forming substance is in the range of 10zM-10nM. The method of claim 1 , wherein the second ionic strength value is different from the first ionic strength value.

6. The method of claim 1, wherein the first specific binding pair forming agent comprises an antibody and the second specific binding pair forming agent comprises an antigen, wherein the binding event is binding of the antigen to the antibody.

7. The method according to any one of the preceding claims, wherein the parameter representing the dielectric function of the layer of first specific binding pair forming substance is the surface plasmon resonance angle (θ).

8. Method according to any one of the preceding claims, wherein the substrate (3) is a metal base plate.

9. The method according to claim 8, wherein the metal substrate (3) comprises gold.

10. The method according to claim 1, wherein the first solution has physiological pH and ionic strength values, in particular pH = 7.4 and ionic strength = 150 mM. The method according to claim 10 , wherein the second solution has a pH of 6 or 8.

2.

12. The method according to claim 10 or claim 11, wherein the second solution has an ionic strength of 5 mM or an ionic strength of 5.5 μM.

13. The method of any one of the preceding claims, wherein the first solution and the second solution are buffered solutions comprising 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid.

14. The method according to any one of the preceding claims, wherein the first solution and the second solution are buffered solutions comprising PBS.

15. The method of any one of the preceding claims, wherein the shift is the shift of a parameter representing a dielectric function of a first specific binding pair forming substance layer.

16. The method of claim 7, wherein the threshold is a threshold SPR angle corresponding to a limit of identification (LOI). The method according to claim 16 , wherein the threshold value is 0.015°.

Images