Single‐Molecule Ex Situ Atomic Force Microscopy Allows Detection of Individual Antibody–Antigen Interactions on a Semiconductor Chip Surface

Although in situ atomic force microscopy (AFM) can allow single‐molecule detection of antibody–antigen binding, the practical applications of in situ AFM for disease diagnosis are greatly limited, due to its operational complexity and long operational times, including the execution time for the surface chemical/biological treatments in the equipped glass liquid cell. In this report, we present a method of graphically superimposed alignment that enables ex situ AFM analysis of an immobilized antibody at the same location on a semiconductor chip surface before and after incubation with its antigen. All of the required chemical/biological treatments can be executed feasibly using standard laboratory containers, allowing single‐molecule ex situ AFM detection to be performed with great practicality, flexibility, and versatility. As an example, we describe the analysis of hepatitis B virus X protein (HBx) and its IgG antibody. Using ex situ AFM, we extracted individual information about the topographical characteristics of the immobilized single and aggregated IgG antibodies on the chip surface and analyzed the data statistically. Furthermore, we investigated, in a statistical manner, the changes in AFM‐measured heights of the individual and aggregated IgG antibodies that occurred as a result of changes in conformation upon formation of IgG–HBx complexes. This article is protected by copyright. All rights reserved.

probe the changes in contacting force that occur when molecules from the tip specifically bind to molecules on the surface [3,11,12] and (ii) characterizing variations in AFM topography mapping profiles that occur upon the conformational changes arising from specific binding between antigens and antibodies. [14][15][16] For example, in situ AFM analysis has allowed such detection at the single-molecule level. In this approach, AFM topographical analysis at the same location, before and after incubation with the antigen, can be used to identify variations in the topographical characterization of an individual antibody. To do so, the AFM system requires integration of a glass liquid cell to function as solution transport system for the chemical/biological treatments required for detection. [15] The analysis chip must be placed in firm contact with the glass liquid cell during the whole detection procedure to ensure that the same area is maintained for AFM analysis. The practical applications of in situ AFM for disease diagnosis are greatly limited, however, due to its operational complexity and long operational times, including the time wasted when executing the required chemical/biological treatments through the glass liquid cell system.
In this paper, we present an ex situ AFM-molecule chip technique integrated with a standard manufactured semiconductor chip and a graphically superimposed alignment technique for the ex situ AFM analysis detection of antibody-antigen interactions at the single-molecule level. This approach is superior to conventional in situ AFM analysis detection in terms of its practicality, flexibility, and versatility. The graphically superimposed alignment technique ensures that the ex situ AFM analysis occurs at the same location on a semiconductor chip surface before and after incubation with the antigen; therefore, all of the required chemical and biological treatments can be executed in standard laboratory containers-and not in the glass liquid cells required for in situ AFM analysis. A designed pattern on the chip surface functions as a graphically superimposed mark for precise alignment of the AFM mapping images recorded at each stage of the detection procedure, thereby enabling stepwise exploration of the evolution of the topographical characteristics at

Accepted Article
This article is protected by copyright. All rights reserved every location on the same AFM analysis area. To test the feasibility of this ex situ AFMmolecule chip technique at detecting the specific interactions between an antibody and a small-sized protein at the single-molecule level, we employed the hepatitis B virus X protein (HBx) as our antigen. Experimental observations and statistical analysis revealed that an individual IgG antibody covalently immobilized on an aldehyde-terminated chip surface, presumably in a predominantly head-on orientation, possessed an AFM height of 5.416 nm.
Furthermore, we also used a statistical approach to obtain information about the change in AFM height (an average increase of ca. 1 nm) of an individual IgG antibody/aggregate that arose, presumably, from the conformational change that occurred upon formation of the IgG-HBx complex. We believe that this ex situ AFM-molecule chip technique based on graphically superimposed alignment will offer opportunities for next-generation singlemolecule disease diagnosis while also providing useful information regarding the behavior of individual antibodies and antigens from the single-molecule perspective.

Accepted Article
This article is protected by copyright. All rights reserved reloading of the chip onto the AFM sample stage for the third AFM analysis. From the viewpoint of applications in practical diagnosis, when compared with the typical in situ AFM analysis, [15,17] this ex situ AFM-molecule chip technique is more flexible in terms of implementing the chemical and biological treatments required for detection, while also requiring less total time for AFM usage for each detection, thereby enhancing the detection capacity.
The main challenge when developing this ex situ detection procedure was to determine how to probe the topographical characteristics of the same analysis area on the chip at different stages. We investigated the concept of graphically superimposed alignment to ensure precise alignment of designed patterns in the ex situ AFM topographical mappings at each stage of the detection procedure. We wanted the alignment pattern on the chip surface to be readily identified by the optical microscopy component of the AFM system to facilitate the movement of the AFM tip to the analysis area. Furthermore, because the designed pattern was necessary for precisely aligning the ex situ AFM topographical mappings at each stage of detection, we wanted the pattern to be located within the AFM analysis area. Using this approach, we expected that any topographical differences at the same surface location could be identified between the various stages of the detection procedure. To test the feasibility of the ex situ AFM-molecule chip for single-molecule disease diagnosis, we used HBx as a model small-molecule antigen. The HBx protein plays an important role in hepatitis B virus (HBV) infection, one of the leading causes of hepatocellular carcinoma (HCC). [18][19][20] Therefore, we suspected that our approach would also be helpful for understanding the interactions of the anti-HBx IgG and HBx from the single-molecule perspective. The Experimental Section provides details of the chip fabrication, surface chemical modification, antibody immobilization, antigen incubation, and AFM analysis; the Supporting Information provides a detailed description of the graphically superimposed alignment procedure of the ex situ AFM topographical mapping. Figure 2a displays the 3D AFM topographical mapping of

Accepted Article
This article is protected by copyright. All rights reserved the analysis area of the chip after immobilization of the anti-HBx IgG antibody. It is evident that several peaks appeared randomly on the SiO 2 surface, as revealed in the AFM image of the analysis area ( Figure 2a). By using the graphically superimposed alignment technique to align the AFM topographical mapping images to the same analysis area before and after immobilization of anti-HBx IgG, we could reasonably attribute the peaks that appeared only after the anti-HBx IgG immobilization in the AFM topographical mapping, as displayed in  Interestingly, we could observe magnitudes of the AFM peak heights ranging from few to tens of nanometers through a single analysis on the same anti-HBx immobilized chip. The topographical profile of peak 2 was approximately the same before and after incubation with HBx; thus, not all of the peak profiles were modified after incubation. These images confirm that the ex situ AFM-molecule chip technique can, indeed, be a precise and reliable experimental tool for probing changes in the topographical profiles of antibodies after incubation with their antigens. Notably, any interference arising from physical adsorption of the antigen (or other molecule) on the chip surface can be excluded when using the ex situ AFM-molecule chip technique because of the graphically superimposed alignment; thereby,

Accepted Article
This article is protected by copyright. All rights reserved we can focus directly on the AFM height variations of the immobilized antibodies on the analytical surface-an essential feature for the purpose of single-molecule disease detection.
To obtain greater molecular-level insight into the AFM peaks that appeared in the 3D AFM topographical mapping, we examined the statistical distribution of the magnitudes of the individual peak heights measured from several chips. Figure 3a reveals two main peaks in the histogram. We used Gaussian distribution functions to fit those two experimental peak profiles as red (first peak) and blue (second peak) lines. The fitted parameters for the heights at the centers (and standard deviations, σ) for the first and second peaks were 5.416 nm (1.12 nm) and 9.914 nm (1.25 nm), respectively [inset table to Figure 3a]. The peak center of 5.416 nm for the first peak is consistent with previously reported thicknesses of an IgG antibody monolayer immobilized on the aldehyde-terminated substrate [21] and heights of individual IgG antibodies extracted using AFM. [16,[22][23][24][25][26][27][28] Thus, we attribute the first peak in the statistical distribution of the experimentally determined AFM peak heights to single anti-HBx IgG antibodies immobilized on the chip surface. Furthermore, by comparing the cross-sectional AFM profiles of various orientations of individual IgG antibodies on the chip surface, [24][25][26][27][28] we suggest a predominantly head-on orientation for the single immobilized IgG antibodies on our surface-consistent with previously reported observations of predominantly head-on orientation for IgG antibodies immobilized on aldehyde-terminated chip surfaces, resulting from the reactivity of the amino group at the N-terminus of the IgG being higher than that of its lysine residues. [26,29] In addition, we attribute the second peak centered at 9.914 nm in and after performing three HBx incubations (right-hand side). We selected a single IgG (No. 2) and two IgG aggregates ( Nos. 1 and 3), as denoted in Figure 3b, for further investigation.

Accepted Article
This article is protected by copyright. All rights reserved

Experimental Section
Chip Fabrication: An SOI wafer was adopted to simplify the semiconductor process flow for fabricating the semiconductor chip with the graphically superimposed alignment mark. After etching to remove the top-layer silicon area excluding the area for the alignment mark, a silicon dioxide (SiO 2 ) layer (thickness: 150 nm) was exposed, acting as a chemical reactive surface for the chemical/biological treatment processes.

Surface Chemical Modification:
To clean its surface, the chip was immersed sequentially in acetone and EtOH, each for 10 min of ultrasonic treatment. The chip surface was then treated with oxygen plasma (power: 18 W; pressure: 550-650 mtorr; 3 min) to remove any organic contaminants and enhance the hydrophilicity of the SiO 2 surface. Next, the chip was immersed in an EtOH solution containing 2% (3-aminopropyl)triethoxysilane (APTES) for 30 min for surface silanization. The chip was rinsed with EtOH several times to remove any APTES residue from its surface, followed by drying for 10 min on a hot plate at a temperature of 120 °C. Subsequently, to introduce aldehyde functional groups onto the surface, the APTES-modified chip was immersed in 10 mM bis-tris propane (BTP) buffer solution (pH 7) containing 2.5% glutaraldehyde (GA) for 1 h at room temperature, followed by several washes with BTP buffer. The chemical modification process is described elsewhere. [31] Accepted Article This article is protected by copyright. All rights reserved Antibody Immobilization: The hepatitis B virus X protein (HBx; no. GTX17526-pro) and the mouse IgG monoclonal antibody of the HBx protein (anti-HBx; no. GTX22741) were purchased from GeneTex. The GA-terminated chip was placed in a standard laboratory container along with 10 mM BTP buffer containing 10 μg/mL anti-HBx and left at a temperature of 4 °C for 16 h. Subsequently, the chip was rinsed several times with 10 mM BTP buffer (pH 7) and then subjected to chemical blocking of the unreacted aldehyde units through exposure to 10 mM tris(hydroxymethyl)aminomethane hydrochloride (Tris-HCl) buffer containing sodium cyanoborohydride (NaBH 3 CN). Finally, the antibody-modified chip was dried with N 2 gas and stored in a vacuum bag at 4 °C until required for further experiments.
Antigen Incubation: After AFM analysis of the antibody-immobilized chip, incubation with the antigen was executed immediately by placing the chip in a standard laboratory container along with 10 mM BTP buffer (pH 7) containing 5 μg/mL HBx and leaving for 30 min at room temperature. After incubation, the chip was rinsed several times with BTP buffer and then dried with N 2 gas. AFM analysis of the chip surface was performed immediately.
AFM Analysis: For ex situ AFM analysis, a Bruker Dimension Icon atomic force microscope was operated in tapping mode to analyze the topographical characteristics of the semiconductor chip surface under ambient conditions. The AFM tip having a nanoscale tip radius of curvature was purchased from NANOSENSORS TM (PPP-NCSTR). NanoScope Analysis software was employed for data analysis of the AFM topographical mappings.
Statistical Analysis: Independent experimental sets (N > 5) were used for analysis. Statistical information, including the average value and standard deviation (σ) of the analysis data, is provided in the main text. The sample size of the data for each statistical analysis is described in the relevant figure captions.

Supporting Information
Accepted Article Figure 3. (a) Histogram of the statistical distribution of AFM heights of the individual peaks in the 3D AFM topographical mappings of the IgG-immobilized chip surface, obtained from 139 data points. The red and blue lines centered at 5.546 and 9.914 nm, fitted using Gaussian distributions, appear to correspond to single IgG antibodies and IgG aggregates of a few molecules, respectively. Inset table: Fitting parameters of the two main peaks. (b) 3D AFM topographical mapping of the same area of the antibody-immobilized surface prior to HBx incubation (left-hand side) and after three successive HBx incubations (right-hand side). The three selected AFM peaks are denoted as Nos. 1-3. (c) Evolution of the cross-sectional AFM peak profiles of the three selected species after antibody immobilization (black line) and one (red line), two (green line), and three (blue line) HBx incubations. The xand y-axes represent the relative lateral and vertical displacements, respectively, of the AFM scans.

Accepted Article
This article is protected by copyright. All rights reserved Figure 4. Histogram of the probability distribution of the change in amplitude (ΔH HBx ) of the AFM-measured anti-HBx height after incubation with HBx. Inset: Plot of 1/ΔH HBx with respect to the corresponding individual anti-HBx AFM-determined peak height, obtained from 137 data points; a rug plot of all of the data points for 1/ΔH HBx is displayed along the yaxis on the right-hand side, revealing statistically that the maximum density distribution of the values of ΔH HBx appeared at approximately 1 nm.

Table of contents
An ex situ AFM-molecule chip technique integrated with a standard manufactured semiconductor chip and a graphically superimposed alignment technique allows the ex situ AFM analysis detection of antibody-antigen interactions at the single-molecule level. This approach is significantly superior to conventional in situ AFM analysis detection in terms of