Silicon‐Based Sensing Surface for Alzheimer's Disease Diagnosis by Phages Probes

Alzheimer's disease (AD) is a diffused neurodegenerative disorder affecting people in advanced age causing loss of memory and dementia. Nowadays, diagnosis and treatment of AD are still challenging due to the lack of diagnostic systems that allow for an early and reliable diagnosis and therapy monitoring. Moreover, conventional strategies for AD diagnosis are based on brain imaging techniques that are invasive and expensive for early and massive screening. Phage display approach, using engineered phage probe for direct amyloid‐β (Aβ)‐autoantibodies detection, overcome these limitations leading to the possibility of safe and low‐cost screening. Moreover, the combination with silicon technology further improves the easiness of diagnosis due to the portability of devices and the integration of sensitive transduction signals. In this work, an innovative silicon‐based sensing technology is reported detecting Aβ‐autoantibodies, specifically Immunoglobulin G (IgG), in human sera by engineered M13‐phage probes (ADPP). The strategy hinges on a bio‐surface that is integrated on top of a silicon biosensor. Thanks to phages probes exposing Aβ‐mimic peptides, this chip can capture and reveal Aβ‐autoantibodies, discriminating between healthy and AD conditions. The surface chemistry is morphologically and chemically characterized and the phage‐based biosensor ability to recognise Aβ‐autoantibodies is proved by transduction with enzyme‐linked anti‐M13 antibodies.


Introduction
Alzheimer's disease (AD) is the most prevalent age-dependent neurodegenerative disorder.It is characterized by a slow, disabling, and irreversible progression. [1]urrently, worldwide >50 million people are affected by AD, which number is expected to increase to over 113 million by 2050. [2]The research community has made significant progress in identifying the risk factors associated with AD, some of which can be modified by simple lifestyle changes while others are non-modifiable since linked to age, sex, and genetics. [3,4]elying on its pathological advancement, AD brings symptoms that are mainly expressed as short-term memory, mood, visuospatial, and language deficits that are initially mild but as the disease progresses become severe, leading to a total loss of basic mental abilities.This degeneration is, mainly, the consequence of senile plaques formation by small peptides, called amyloid- (A), [5] which accumulate in medial temporal lobe and neocortical section of brain. [6]These deposits contribute to alterations in neuronal activity and immune response disruption. [7]owadays, an early reliable diagnosis of AD would help to increase the effectiveness of cures against the neurological disorder. [8,9]Moreover, due to the complexity of the diagnosis, mostly based on the correlation between the build-up of amyloid plaques and cognitive functions degeneration, there is a lack of systems able to make accurate, low-cost, and non-invasive screening of the neurological disease.Conventional strategies for AD diagnosis are mainly based on imaging techniques and cerebrospinal fluid (CSF) analysis.The imaging can be performed by magnetic resonance (MRI), assessing hippocampal atrophy, an invariable neuropathological feature of AD, and positron emission tomography (PET), detecting A plaques in the brain. [10,11]hese technologies, although consolidated, are invasive, since involve hard sampling procedures like lumbar punctures, and are limited to advance stage AD patients and too expensive for early and massive screening.On average, in fact, a PET scan needs to be performed in replicas and costs between 4000 € and 5000 €.This represents also a big drawback for large clinical trial (new drug discovering) that can have thousands of people enrolled making this PET methodology not sustainable.
Therefore, there is a growing need to develop alternative lowcost and non-invasive strategies for AD diagnosis, especially in the preclinical stages.In this scenario, several biosensors using different transduction methods have been reported for AD diagnosis.Most of them screen specific biomarkers that have shown great potential in AD screening [12,13] and are based on detection techniques like lossy mode resonance (LMR) technology implemented in D-shaped single mode fibers (reporting high sensitivities and detection limit), [14] surface plasmon resonance (SPR), [15,16] localized SPR (LSPR), [17,18] electrochemistry [19,20] and immunochemistry.23] The traditional immunological sensing is evolving toward the use of phage display to select specific probes.This is a biotechnology that allows to specifically recognize different molecular targets (proteins or antibodies) through engineered phage probes. [24,25]These phages express on their capsid a large number of specific peptides to bind and identify a target biomolecule. [26]y this approach, starting from libraries of phage probes expressing high number of random peptides and throughput a biopanning process, it is possible to select specific peptides that can detect targets involved in many types of human syndromes.33] For AD filamentous phage families (M13, Fd, f1) [34] genetically modified to identify AD biomarkers, such as the A and other peptides involved in amyloid fiber formation. [35]][38] In this respect, the combination with silicon technology is still unexplored.Silicon materials allow the integration of many technological advances such as high sensitive transduction, fluidic movement, and electronic reading. [39]42][43][44] In this work, we presented an innovative silicon-based sensing biointerface that uses phage probes for AD detection in human sera.The technology relies on a bio-surface that has been opportunely derivatized on two silicon substrates: i) a flat silicon slide, as reference model for characterization; ii) a miniaturized chip, as first attempt for future PoC development.The sensing surface was structurally and chemically characterized by atomic force and scanning electron Microscopy (AFM and SEM), X-ray photoelectron spectroscopy (XPS), and contact angle measurements (CA).The molecular recognition of AD A-autoantibodies in human sera by phage probes was ascertained by transduction with enzyme-linked anti-M13 antibodies.

Human Sera Preparation
Sera used in this study were provided by the Neurologic Unit of the University Hospital "Policlinico Vittorio Emanuele" of Catania, Italy, and diagnosticated as AD by the Mini Mental State Examination (MMSE).The study was approved by the Ethics Committee of the "Policlinico Vittorio Emanuele" of Catania.For the testing experiments, 3 sera with MMSE between 12 and 15 (severe AD) of a 75-year-old male patient and 3 sera with MMSE of ≈29 (healthy) were used.It was diluted 1:50 in a buffer, prepared by mixing 50 mL PBS with 500 mg of 1% non-fat dried milk powder and 50 μL of 0.1% Tween 20.

Probe Phage Preparation
The ADPP phage probe (named 12III1) was previously identified [38] and displays a capsidic peptide, RWPPHFE-WHFDD.It was prepared by infecting 90 μL of exponential broth culture of E. coli strain TG1 (Kan-, Amp-, lacZ-) with 10 μL of ADPP, then incubated at 37 •C in static condition for 15 min, followed by shaking for 20 min.After incubation, culture was plated onto Luria-Bertani modified with bacteriological agar (Condalab) plates containing 50 μg mL −1 of ampicillin (Merck) and incubated at 37 •C in static condition.One colony of transformed E. coli was inoculated into 10 mL of LB medium containing ampicillin and incubated at 37 •C with shaking until reaching Optical density (OD) 600 nm = 0.2.Then, the culture was added with isopropylthio--galactoside (IPTG, 40 μg mL −1 ), from Merck, and helper phage M13K07 (10 9 TU mL −1 ), incubated at 37 •C in static condition for 30 min, and gently shaken for 30 min.The cells were harvested by centrifugation at 8000× g, transferred to 500 mL of LB medium containing ampicillin and kanamycin (50 μg mL −1 ), from Merck, and incubated overnight with shaking at 37 •C.The infected culture was centrifuged 8000× g for 20 min at 25 •C, the supernatant was then mixed with 25% (v/v) of PEG/NaCl solution (Merck), cooled on ice for 4 h, and precipitated by centrifugation at 15000× g for 45 min at 4 •C.The pellet was resuspended in 10% (v/v) of TBS (Merck), mixed again with 25% (v/v) of PEG/NaCl, cooled in ice for 4 h, and the solution was centrifuged as above.The pellet containing the phages were suspended in 10% (v/v) of TBS, filtered through 0.22 μm-pore size membrane (GVS), and stored at 4 •C.

Sensing Surface Functionalization Based on Silicon Substrate
The first surface used for sensing was constituted by a 3 × 3 mm silicon flat substrate (silicon biointerface).This was functionalized as schemed in Figure 1 using a Teflon beaker as processing holder.First, the surface was prepared by a cleaning step performed by adding 20 μL of a H 2 O 2 :NH 3 :H 2 O (1:1:4) oxidizing solution at 80 °C for 25 min.Then, a treatment with an acid solution (HCl:H 2 O, 1:7.5) at room temperature for 10 min was carried out.Lastly, the surface was rinsed with deionized water.After that, the surface was silanized by chemical vapor deposition (CVD) with GOPS silane at 125 °C for 4 h.Then, the surface was rinsed with a washing buffer composed of 50 mL PBS + 25 μL 0.05% Tween 20 and used for the subsequent functionalization step.The protein G was anchored to the silanized surface by soaking the silicon substrate overnight at 4 °C with 200 μL of 0.5 μg mL −1 of protein G solution diluted in PBS.Then the substrate was rinsed three times with PBS.

Sensing Surface Preparation of Silicon Chip
The second substrate was a miniaturized 1 cm × 2 cm silicon chip composed of six circular wells of 3 mm diameter that have been realized by lithography of a 15.24 cm diameter silicon wafer (Figure 7a).On top of the silicon chip a polycarbonate mask was attached creating six microchambers of 25 μL capacity.To avoid fluids leakages during the functionalization and testing treatments, the silicon wells and the plastic mask were aligned and glued via a conductive resin.The sensing surface was treated according to the protocols above described using a volume of 20 μL for each treatment.

Testing
The testing was performed in three steps: a) antibodies binding.The protein G coated surface was used to bind the Fc domain of the IgG antibodies.The IgG binding step was performed by adding both healthy and AD human sera.As negative control, PBS 10 mM was used.First, 200 μL of 1:50 diluted sera and PBS 10 mm (negative control) were added in each Teflon beaker containing the silicon surface and incubated for 1 h at 37 °C with 80 rpm of agitation.For the silicon chip, 20 μL of diluted sera was used.Lastly, the unbound IgG were removed by three cleaning steps of washing buffer; b) ADPP phage recognition.This step consisted in the capturing of immobilized A auto-antibodies by the ADPP (12III1 probe) phage.To block the uncovered silicon spots in the surface, a preliminary passivation of the silicon surface was performed by incubating the silicon surface with 200 μL of a blocking solution prepared by mixing 5 mL of 0.01 m PBS with 250 mg of non-fat dried milk powder and 2.5 μL of Tween 20 at 0.05% for 2 h at 37 °C.For the silicon chip, 20 μL of blocking solution was used.Once the surface was passivated, the recognition of A auto-antibodies was performed by spotting 10 μL of 10 11 copies of ADPP probe diluted in 0.01 m PBS for 1 h at 37 °C with 80 rpm of agitation.The unbound phages were removed by five cleaning steps of washing buffer; c) transduction with enzyme-linked anti-M13 antibodies test.A volume of 200 μL of a 1:750 solution of the anti-M13 HRP-conjugated IgG, suspended in the dilution buffer, was added to the silicon substrates previously treated with steps a) and b), and incubated at 37 °C for 1 h with 80 rpm of agitation.For the silicon chip, 20 μL of anti-M13 HRP-conjugated IgG was used.Subsequently, the chambers were rinsed three times with washing buffer and used for the final immunochemical analysis.For the immunological detection, a volume of 100 μL, for the silicon substrates of TMB was added to the sensing surface and incubated for ≈15 min at room temperature in the dark.For the silicon chip, 10 μL of TMB was used.When a blue color appeared in the positive wells, because of the TMB oxidation by HRP, the reaction was stopped by adding 100 μL, and 10 μL for the chip, of 0.6 N H 2 SO 4 , causing a color change from blue to yellow.The color intensity of was, then, measured by reading the samples absorbance at 450 nm using a VICTOR 3 V Multilabel Plate Reader spectrophotometer.In the case of silicon chip, the volume of each well was transferred into a 96-well plate and tenfold diluted by adding of 180 μL of PBS, in order to achieve the right volume for the absorbance reading; subsequently, the absorbance values had been corrected by multiplying for the dilution factor.Statistical analysis was performed by one-way ANOVA test using GraphPad InStat version 3.1.

AFM Analysis
The surface topography of step-by-step surface assembly (GOPS silanization, protein G anchoring, IgG binding, and probe phage capturing) was investigated by atomic force microscopy (AFM), performed by a SolverP47 NT-MDT instrument, in semicontact mode, using a scan window of 5 μm × 5 μm and 2 μm × 2 μm.Silicon probes (NT-MDT) with constant force in the range of 0.01-0.5 Nm have been used for the analysis.All the data for each AFM image were evaluated by Gwyddion software.

SEM Analysis
The SEM images were obtained using a field emission scanning electron microscope (FESEM) ZEISS VP 55 (Oberkochen, Germany).

XPS Analysis
X-ray photoelectron spectroscopy (XPS) was carried out with a with a PHI 5000 Versa Probe Instrument using a monochromatic Al K X-ray source excited with a micro-focused electron beam.All the analyses were performed with a photoelectron take-off angle of 45 (relative to the sample Surface).The XPS binding energy (B.E.) scale was calibrated on the C 1s peak of adventitious carbon at 285.0 eV.

CA Measurements
The contact angle (CA) was measured by the OCA25 System from DataPhysics.The CA, as function of the surface wettability after each functionalization step, was assayed in drop casting mode using 1 μL of distilled water and analyzing the drop age at 4 time points (0, 68, 134, and 203 s).In parallel, CA was also measured using a drop of buffers citrate at pH 4.40 and 6.02, buffers carbonate/bicarbonate at pH 9.36 and 10.2, to evaluate the isoelectric point (IP), and the surface charge profile of the biomolecules anchored in step-by-step assembled sensing surface.

Results and Discussion
The silicon-integrated detection strategy for capturing and detecting A-autoantibodies in human sera by ADPP probes is reported in Figure 1.The principle of the method is based on a molecular recognition of AD A-autoantibodies by ADPP probe, exposing the capsidic RWPPHFEWHFDD peptide, attracted to the silicon surface by the AD A-autoantibody and detected by transduction with enzyme-linked anti-M13 antibodies.To this end, as reported in Figure 1-functionalization, the silicon surface was first functionalized with GOPS by covalent grafting between its Si-OCH 3 groups and the exposed surface OH groups (previously formed by activation treatment with piranha solution).The next step of functionalization was the anchoring of the protein G trough a reaction between its NH 2 terminal groups and the epoxy group of GOPS.The protein G is a streptococcal immunoglobulin-binding protein that is well known to bind specifically the Fc domain of human IgG 1-4, avoiding the unspecific interaction with IgMs and IgAs. [45]Therefore, the protein G layer is fundamental for our detection strategy to capture all the IgGs present in sera and to immobilize them in the proper orientation to expose the antigen binding sites (F ab domain) toward the probe phage. [46]The improvement of the detection efficiency due to the IgGs orientation has been preliminary proved by testing in triplicate an AD serum in ELISA performed on a plastic 96-well plate functionalized with and without the protein G layer.Results, reported in Figure S1 (Supporting Information), shown four times enhanced absorbance values of wells functionalized with protein G.
After functionalization of the surface with G protein, the specific identification of A-autoantibodies was achieved by a molecular recognition with ADPP probe, forming IgG-ADPP molecular complex (Figure 1 -testing).In fact, this phage brings on its pVIII capsid hundreds of copies of a A-mimics peptide (RW-PPHFEWHFDD) that molecularly recognize specific regions of A-autoantibodies, thereby discriminating between healthy and AD sera. [38]In other words, while the Protein G layer immobilize all the IgGs present in the serum, ADPP will bind only the A auto-antibodies, thus proving selectivity to the assay.The IgG-ADPP complex is subsequently transduced with a secondary probe constituted by the enzyme-linked anti-M13 antibodies conjugated with HRP that produce a colorimetric signal (Figure 1 testing scheme).
We performed a detailed morphologically and chemically characterization by AFM and XPS analysis of all functionalization steps, in order to characterize in deep, both the morphological and chemical features of the bio-surface.The AFM image of the GOPS sample is reported in Figure 2a and shows flat surface due to the formation of a homogeneous layer of silane on the silicon surface.This was confirmed by the RMS data, 0.267 nm, and the heights distribution profile, in Figure S2a (Supporting Information), that reveal an average height peak at 0.94 nm while the height profile in Figure 2b does not show any structure on the surface, reporting a homogeneous background.AFM images of GOPS-functionalized SiO 2 surface after the anchoring of the protein G (GOPS-PG) do not show the presence of hole or noteworthy rough element (Figure 2c).However, the RMS data show an increased surface roughness of 0.33 nm, 25% higher than the one of GOPS sample, which could be attributed to the additional protein G layer anchored to the silane one.This assumption was supported by the distribution profile of the heights in which the deconvolution of the average peak, obtained by fitting two gaussian functions, reveals the presence of two dispersed objects on the GOPS-PG surface.As described in Figure S2b (Supporting Information), the Fit Peak 1 (red line) reports an average height of 1.1 nm, that could be related to the GOPS distribution, while the Fit Peak 2 (green line) gives an average of 1.4 nm, attributable to the protein G layer.This structurization of the surface is also confirmed by the background reported in the height profile shown in Figure 2d.
The proposed assembly was also confirmed through the XPS analysis reported in Figure 3.In particular, the C1s band of GOPS can be deconvoluted into three components: the primary component at 285.0 eV corresponds to the hydrocarbon backbone of GOPS and adventitious carbon.A second component at 285.8 eV can be attributed to oxidized carbon in the ether group, while the third component at 287 eV represents the carbon of the epoxy groups.Following the protein G anchoring, the bands shape changes significantly.Besides the typical C1s signal at 285 eV, the spectrum is characterized by the prominent signal of oxidized carbon at 286.4 eV that arises from the carbon atoms of the protein's carboxylic groups.
Lastly, N1s peak was found at 399.8 eV and was detected only in the sample containing the protein G, due to its amino acid com-position, confirming the binding of the molecule to the GOPS layer.
To assess the effectiveness of the IgG binding to the GOPS-PG biointerface, AFM analysis was carried out after the first step of the testing that included the IgG anchoring.Data are reported in Figure 4a-d.The images show a topographic variation as consequence of the presence of widely distributed circular features.In the 5 μm × 5 μm and 2 μm × 2 μm AFM images reported in Figure 4a,d the presence of big structures related to an accumulation of organic materials.The height profile in Figure 4b, referenced to the line 1 in Figure 4a, exhibits a feature with a height ranging from 2 to 15 nm, in which the few larger structures in the 5-15 nm range can be identify as the IgG aggregation.Additionally, the configurations of the structures shown in Figure 4b and Figure S3a,b (Supporting Information) reveal a distinct conformation with two peaks at the bodies' tips, possibly associated with the characteristic "Y" shape structure commonly observed in IgG.
A height profile along line 2 in Figure 4a, obtained without intersecting the prominent features, and the total height distribution are presented in Figure 4c and Figure S4 (Supporting Information) to illustrate the morphology of the background outside the large IgG assemblies.The height distribution indicates that most features are in the 3.8-5.5 nm range, confirming variations in morphology compared to GOPS and GOPS-PG samples where almost all features were in the 1-2 nm range.These data are consistent with previous works in literature, that indicates a mean of heights for antibodies in the range of 3-8 nm. [47,48]he RMS analysis confirmed the topographic evidence by reporting an increased surface roughness of 3.11 nm, compared to the 0.33 nm of the GOPS-PG sample.
Further evidence of the crucial role of protein G layer in the binding of IgGs was highlighted by SEM inspection.The comparison of GOPS-PG-IgG and GOPS-IgG samples reported in Figure S5 (Supporting Information).In the first sample (Figure S5a, Supporting Information) the antibodies appeared widely dispersed with a homogeneous surface distribution due to the strong interaction of Fc domain of IgG with the protein G.In the second sample (Figure S5b, Supporting Information), instead, the absence of this protein led to the formation of IgG aggregates as consequence of unspecific lateral Fab-Fab interactions [49,50] between physiosorbed antibodies.To determine the binding between the IgGs and ADPP, we inspected the morphology of the GOPS-PG-IgG-Phage biointerface by AFM and SEM.AFM topographic images of this surface (Figure 4e) show the presence of larger features than those observed in the GOPS-PG-IgG surface (Figure 4a).The height profiles related to the lines 1 and 2 of Figure 4e reported in Figure 4f,g ,respectively, show big structures of ≈50 nm height (Figures S6 and S7, Supporting Information) and a background outside these structures (line 2) with a morphology that exhibits greater height and less ruggedness compared GOPS-PG-IgG samples due to the presence of ADPP.It is notable that, as the height of the structures increases it is not possible to observe their "Y" configuration (Figure S3c,d, Supporting Information), possibly due to the aggregation of ADPP at their tips.This is consistent with the observed roughness, which changes from 3.11 to 10.9 nm in the absence and presence of ADPP on the surface, respectively.The magnification (Figure 4h) of a 2 μm × 2 μm zone without big structures shows a distribution of object all over the IgG surface with ≈0.25 μm length and with a height distribution ≈5.8 nm (Figures S6 and S8, Supporting Information).These objects can be associated to the captured probe phage.The SEM image confirms a topographic profile exposing blurred fractal bodies (Figure S5c, Supporting Information).
CA measurements for each step of surface functionalization have been carried out to confirm the effectiveness of interface modification.The results, reported in Table 1, show that the CA  decreased of ≈40°after the silicon surface cleaning step.With the GOPS silanization, instead, the CA increases >70°as consequence of the epoxy-layer hydrophobic coating.Then, the wettability gradually increased again after each step of biomolecular assembly due to the addition of the polar protein G, IgG, and phage structures.
The charge surface characteristics of the anchored biomolecules have been also investigated by CA analysis performed at various pH in the range of 4-10.The results reported in Figure 5 (the dashed lines have been added just as a guide for a better data visualization) show a qualitative change of the surface wettability as function of the pH for the several levels of functionalization.It can be noticed that the GOPS surface (black symbol in Figure 5) was inert to the pH variations, with no significative changes of the wettability, may be due to the non-polar nature of the silane.On the contrary, both the PG and PG-IgG derivatized surfaces (red and blue symbols, respectively), show a wettability change linked to pH with a first increase of CA at pH 4 for the PG and a second common peak at pH 7 for both PG and IgG molecules.This behavior may be due to the similar biostructures of protein G and IgG, both composed of zwitterionic amino acids, suggesting an isoelectric point (pI) of about pH 7. Lastly, the complete GOPS-PG-IgG-Phage surface (green symbols) gave a different wettability behavior, probably, as consequence of the presence of protein capsids and exposed peptides at pVIII moiety of the anchored phages.The observed behavior can be attributed to specific conformational changes of the ADPP structures that could occur as consequence of their charge arrangement, molecular interactions in combination with the ionic strength of the buffer used for the analysis.To finally assess the validity of our detection strategy, we tested three healthy and 3 AD sera, compared to negative controls made of PBS 10 mM.The presence of the ADPP in all samples was measured by performing an immune recognition with anti-M13 antibodies conjugated with HRP that produces a color absorbance @450 nm.The results reported in Figure 6, indicate that the biointerface can detect autoantibodies in both healthy and AD sera.Data show a good trend in discriminating AD from healthy sera (absorbance equal to 0.18 ± 0.06 for healthy sera and 0.85 ± 0.3 for AD sera), highlighting that our biointerface is suitable to be used in an integrated into a biosensor device.
To this end, we transferred the above-described concept into a silicon biochip, as first attempt, to demonstrate the possibility to develop miniaturized biosensor for future exploitation in PoC device for AD early diagnosis and screening.Therefore, we developed the biochip schematized in Figure 7a, that was chemically functionalized with protein G according to the procedure described above.The chip is an array of six reaction chambers where we carried out the detection test using two replicas of three negative controls and three type of healthy and AD sera.Results are reported in Figure 7b, showing a statistically significative change in the absorbance between healthy sera (Abs 450nm = 0.42 ± 0.08) and the AD sera (Abs 450nm = 0.78 ± 0.22).
All data from both silicon flat substrate and chip testing were analyzed by one-way ANOVA test, which indicated a significant difference among the samples (negative control, healthy, and AD) group of values (p < 0.0001 for the Si flat and p = 0.0002 for the chip).Moreover, for both series of testing, post-hoc analysis with Tukey's correction indicated that the negative control group was not statistically significant from the healthy group.In contrast, the AD group was statistically significant from the other two groups (p < 0.001 for both comparisons).
These findings were the consequence of the specific interaction between the anchored A-autoantibodies and the phage mimics peptides, that could lead to a higher number of captured phages, thus, a more intense detection by the secondary anti-M13 antibody conjugated with HRP.

Conclusion
A silicon-based biosensing technology for AD diagnosis in human sera has been described.The technology relies on a biointerface comprising of a protein G layer and ADPP phage.The former is chemically linked to a silicon surface via silane coupling and molecularly recognizes IgGs.The latter acts as transducer by selectively recognizing A-autoantibodies present in sera samples.Silicon technology has been used since is a stable material and is easily arranged in variegated innovative devices thanks to its consolidated technology.Both AFM and XPS analyses confirmed the functionalization of the silicon surface with GOPS silane and protein G.The binding of the IgG antibodies was evident by The SEM and AFM inspections of both GOPS-PG-IgG and GOPS-PG-IgG-Phage biointerfaces by SEM and AFM inspections highlighting circular fractal bodies.
The charge characteristics of the surface were also studied by CA measurements at various pH, demonstrating the effectiveness of each step of functionalization together with a dependence of surface wettability upon pH that can be attributed to structural arrangement of the anchored biomolecules.
The sensing ability of the system was first validated with a silicon biointerface by testing both healthy and AD human sera.Results indicate a good discrimination between AD and healthy sera, validating our biointerface for sensing purpose.To this end, we finally developed a biochip constituted by six microchambers that was used as first proof of concept for AD diagnosis and screening.Data show a statistically significative discrimination between healthy and the AD sera, paving the way to future development of devices in PoC format for AD early diagnosis and screening.Thanks to the combination with the miniaturized architecture of the silicon chip, the technology opens to the possibility of a very large-scale integration toward a new frontier of decentralized biomedical diagnostics.

Figure 2 .
Figure 2. AFM analysis of GOPS and GOPS-PG surfaces: a,c) surface topography, and b,d) heights distribution profile.

Figure 3 .
Figure 3. XPS analysis of C1s and N1s spectral regions of GOPS and GOPS-PG surface.

Figure 6 .
Figure 6.ADPP immune recognition of the silicon bio-surface of negative control, healthy, and AD sera.Data were analyzed by one-way ANOVA test (p < 0.0001).

Figure 7 .
Figure 7. a) Silicon biochip; b) immune recognition of negative control, healthy and AD sera.Data were analyzed by one-way ANOVA test (p = 0.0002).

Table 1 .
CA analysis of sensing surface step-by-step assembly.