Effect of Antigen Retrieval Methods on Nonspecific ... - ACS Publications

Nov 17, 2017 - In this study, we found that the antigen retrieval method strongly influences the extent of nonspecific ... Alternatively, for antigens...
0 downloads 10 Views 668KB Size
Subscriber access provided by READING UNIV

Article

Effect of antigen retrieval methods on non-specific binding of anti-body-metal nanoparticle conjugates on FFPE tissue Yuying Zhang, Xinping Wang, Sven Perner, Ágnes Bánkfalvi, and Sebastian Schlücker Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b03144 • Publication Date (Web): 17 Nov 2017 Downloaded from http://pubs.acs.org on November 17, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Effect of antigen retrieval methods on non-specific binding of antibody-metal nanoparticle conjugates on FFPE tissue Yuying Zhang1,4*, Xinping Wang1, Sven Perner2, Agnes Bankfalvi3, Sebastian Schlücker1* 1

Physical Chemistry I, Department of Chemistry and Center for Nanointegration Duisburg-Essen (CENIDE), University of Duisburg-Essen, Universitätsstr. 5, Essen 45141, Germany. 2 Pathology of the University Medical Center Schleswig-Holstein, Campus Lübeck and the Research Center Borstel, Leibniz Center for Medicine and Biosciences, Lübeck, Germany. 3 Institute of Pathology, University Hospital Essen, Hufelandstrasse 55, 45122 Essen, Germany 4 Medical School of Nankai University, 94 Weijin Road, Tianjin 300071, China ABSTRACT: Immunohistochemical analysis of FFPE tissues provides important diagnostic and prognostic information in pathology. Metal nanoparticles (NPs) and in particular surface-enhanced Raman scattering (SERS) nanotags as a new class of labelling reagents are promising to be used for multiplexed protein profiling on tissue sections. However, non-specific binding of NPs onto the tissue specimens greatly hampers their clinical applications. In this study, we found that the antigen retrieval method strongly influences the extent of non-specific binding of the antibody-SERS NP conjugates to the tissue. Our SERS labels comprised ca. 70 nm Au nanostars coated with ethylene glycol-modified Raman reporter molecules for hydrophilic stabilization and subsequent covalent bioconjugation to antibodies. We systematically investigated the influence of heat- and protease-induced epitope retrieval (HIER and PIER, respectively) on the immuno-staining quality of prostate-specific antigen (PSA) on human prostate tissue sections. The best staining results were obtained with PIER. Pretreatment of the tissue sections by HIER led to selective but nonspecific adsorption of the antibody-Au nanostar conjugates onto epithelial cells, while enzymatic treatment within PIER did not. In addition to gold nanostars also other types of metal NPs with different shapes and sizes (incl. ca. 20 nm quasi-spherical Au NPs and ca. 60 nm quasi-spherical Au/Ag nanoshells) as well as tissue sections from different organs (incl. prostate and breast) were tested; in each case the same tendency was observed, i.e., PIER yielded better results than HIER. Therefore, we recommend PIER for future NP-based tissue immuno-staining such as immuno-SERS microscopy. Alternatively, for antigens which can only be unmasked by heating, PEGylation of the NPs is recommended to avoid non-specific binding.

Introduction Immunohistochemistry (IHC) is an important technique in daily diagnostic practice that helps pathologists to elucidate differential diagnoses which are difficult or not possible by conventional morphological analysis1-3. Classic IHC employs enzyme-labelled (immunoperoxidase) or fluorophore-labelled (immunofluorescence, IF) antibodies to identify proteins and other molecules in tissue specimens4. Nanoparticles (NPs) are a new class of labelling reagents which hold great promise in a broad range of biomedical applications5-9. For example, quantum dots are bright and stable labels for optical detection via fluorescence10, while colloidal gold (immunogold) is normally utilized for immuno-detection by electron microscopy13. More recently, surface-enhanced Raman scattering (SERS) labels, that combine metallic NPs with organic Raman reporter molecules, have emerged as new promising candidates for protein localization on tissue14-17 Compared with existing labeling approaches, SERS labels offer significant advantages such as the potential for a multiplexed detection of more than ten targets simultaneously requiring only a single laser excitation line for all colors, quantification based on spectral intensities, minimal photo-bleaching and improved image contrast by red to near-infrared laser excitation17,18,20,21.

In the past decade, tremendous effort has been made to generate optimal SERS labels/nanotags as well as other NP labels in order to promote their application to clinical tissue specimen10,19,21. However, a major obstacle is that the NP-based probes are ‘sticky’, tend to aggregate in bio-fluids and bind nonspecifically to various cells or tissue samples12,24,25. Nonspecific binding of NPs increases background signals, limits tagging specificity and induces false staining results. To resolve this problem, NP probes are usually encapsulated with poly(ethylene glycol) (PEG)26, bovine serum albumin (BSA)15 or silica or stabilized by small hydrophilic molecules29-31. On the other hand, biological samples are often pre-incubated with inert proteins, such as BSA, non-fat milk or goat serum to block reactive groups on tissue specimen32. PEG is a biocompatible polymer with a low affinity to proteins which is a requirement for minimizing non-specific binding. PEGylated NPs have therefore been successfully applied for both in vitro and in vivo studies10,26. However, the increased hydrodynamic size may prevent the probes from accessing targets deeply within complex tissue materials29. Instead, coating the NPs with small hydrophilic molecules might help to stabilize the NP probes, reduce non-specific binding and simultaneously overcome problems associated with bulky NPs29-31,33. In our recent studies, we have developed hydrophilically stabilized SERS labels by coating metal NPs with a self-assembled

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

monolayer (SAM) of Raman reporter molecules, covalently linked to either a non-reactive monoethylene glycol (MEGOH) or a carboxy-reactive triethylene glycol terminus (TEGCOOH)33. The SAM ensures maximum loading of the Raman reporter molecules on the metal NPs for maximum SERS signals, the MEG units with terminal OH groups stabilize the NPs while the TEG units with terminal COOH moieties additionally allow for subsequent conjugation to antibodies. The antibody-metal NP conjugates have been successfully applied to detect cytokines in a direct dot-blot assay with a femtogram detection limit34. In the present study, Au nanostars were synthesized as metal substrates since they are highly plasmonically active and enable the detection of single particles30. The Au nanostars were coated with the above mentioned ethylene glycol-modified Raman reporter molecules and conjugated with antibodies, and subsequently applied for protein localization on formalin-fixed paraffin-embedded (FFPE) tissue slides by immuno-SERS (iSERS) microscopy. FFPE tissues are the most widely used specimen for histopathological diagnosis in hospitals world-wide since formalin fixation results in cross-linking of macromolecules and keeps the tissue in an excellent morphological condition for histopathological analysis3,36. An antigen retrieval step is usually required before IHC staining of FFPE sections to unmask hidden or denatured epitopes. Two frequently employed approaches of antigen retrieval are heat-induced epitope retrieval (HIER) and protease-induced epitope retrieval (PIER)37,38. Although the complex mechanisms of formalin fixation and antigen retrieval still remain unclear, it is speculated that enzyme digestion cleaves peptide bonds while heat breaks some of the formalin-induced cross-linkages, both allowing hidden determinants to be exposed4. In the classic IHC or IF staining, it has been found that some antigens prefer enzymatic over heat-mediated antigen retrieval and vice versa. The extent of epitope exposure might be varied and thereby the signal brightness could be different4,37. In this study, prostate specific antigen (PSA) on human FFPE prostate tissue sections was demasked using either HIER or PIER for a direct comparison of the corresponding staining quality. iSERS and IF as a reference method were performed in parallel to evaluate the influence of antigen retrieval methods on the staining qualities. PSA was chosen here as a model antigen because it is abundantly expressed in the epithelium of prostate tissue and has already served as a test system in the very first as well as subsequent iSERS studies14,15. For conventional IF staining comparable fluorescence signals were observed in tissue sections treated with either HIER or PIER. However, in the case of iSERS staining using SERS-labelled antibodies combined with fluorophore-labelled secondary antibodies, we clearly obtained different results: Using PIER both IF and iSERS microcopy showed very bright and specific staining of PSA on epithelial cells, without any non-specific binding of the NPs. Using HIER, also bright iSERS staining was observed, however, in the negative control, non-specific binding of the SERS probes was detected. To the best of our knowledge, this is the first time that the antigen retrieval method is reported to have an influence on non-specific binding of NPs on tissue sections. In order to investigate whether this phenomenon is also observed for different antibodies and tissues, we tested staining of PSA on prostate tissue section by SERS NPs conjugated with either primary anti-PSA antibodies or secondary antibodies, and immuno-staining of human epidermal growth factor receptor 2 (HER2) on breast tissue sections with SERS-labelled second-

Page 2 of 10

ary antibodies. In all cases, non-specific binding of NPs was only observed in tissue samples pretreated with HIER, but not in the case of PIER. To investigate the effect of incubation buffers and physicochemical properties of NPs, tissue samples were treated in different buffers with or without heating, and then incubated with SERS probes with different particle sizes, shapes and/or surface modifications. The results indicate that non-specific binding occurs in all cases when the tissue was pre-treated by heat, and the NPs with different sizes (20/60/70nm) or shapes (spheres or stars) show a similar binding behavior. In contrast, when the NPs were encapsulated with a thick PEG layer (ca. 12 nm), they did not bind nonspecifically to HIER-treated tissues, but the staining accuracy and signal intensity decreased compared to the non-PEGylated ones. Based on these results, we speculate that heating pretreatment induces charge and/or structural changes in the tissue specimens. Therefore, an enzymatic pre-treatment of the tissue samples is superior with respect to minimizing nonspecific binding of NP-based probes. Experimental Section Reagents and Instruments Tetrachloroauric acid (HAuCl4·3H2O), silver nitrate, 1ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC), Nhydroxysulfosuccinimide sodium salt (s-NHS), bovine serum albumin (BSA), 20 nm Au NPs, ethanolamine, 5,5’-dithiobis (2-nitrobenzoic acid) (DTNB) and 4-(2-hydroxyethyl)-1piperazine ethane sulfonic acid (HEPES) were purchased from Sigma Aldrich. L(+)-ascorbic acid was obtained from AppliChem. HS-PEG-COOH (Mw~3 kDa) was obtained from Rapp Polymers. Sodium chloride and hydrochloric acid were obtained from Bernd Kraft, Germany. Disodium monohydrogen phosphate (Na2HPO4) and potassium dihydrogen phosphate (KH2PO4) were obtained from Carl Roth, Germany. Monoclonal anti-PSA antibody (colon 8301) was obtained from Abcam. Monoclonal anti-HER2 antibody (colon TAB250) and a fast enzyme for antigen retrieval were purchased from Zytomed, Germany. Alexa647-conjugated goat anti-mouse (G@M) antibody was purchased from Thermofisher. Human prostate FFPE tissue blocks were provided by the Institute of Pathology, University Hospital Bonn, Germany (former affiliation of S.P.). Human breast FFPE tissue blocks were provided by the Institute of Pathology, University Hospital Essen, Germany. Ultrapure water (18.2 MΩ cm, Millipore) was used in the experiments. SERS and fluorescence images were recorded by an in house modified correlative confocal Raman-fluorescence microscope (WITec Alpha 300R, 30 cm focal length and 600 grooves per mm grating spectrometer) equipped with an EMCCD (Andor Newton DU970N-BV-353). A high-pressure mercury lamp (LEJ LQ-HXP120 VIS) was used for wide field excitation and coupled by a liquid waveguide to the illumination port of the microscope. Bandpass filters for detection of FITC and Alexa647 were placed in the excitation and emission path. The fiber coupler to the grating spectrometer in the original configuration (Raman only) was replaced by a Tcoupler (Zeiss) including a transfer optic, which was used for simultaneously coupling to the grating spectrometer for Raman microscopy and to a CCD (Zeiss Axio am ICm1) for fluorescence microscopy. Extinction spectra were recorded using a Jasco V600 UV/vis absorption spectrophotometer. The concentrations of SERS probes were determined using a Nanosight nanoparticle

ACS Paragon Plus Environment

Page 3 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

tracking analysis (NTA) system. Transmission electron microscopy (TEM) measurements were performed with a Zeiss EM910 transmission electron microscope operating at 100 kV. Dynamic light scattering (DLS) and zeta potential were measured using a Malvern Zetasizer Nano ZS90. Preparation of SERS Probes Au nanostars and Au/Ag nanoshells (TEM images are shown in Figure S1) were synthesized following protocols developed previously 16,35. The extinction peaks of the NPs were adjusted to be at ca. 640 nm (Figure S2), which is in between the laser excitation line (632.8 nm) and the characteristic Stokes-Raman wavelengths of the Raman reporter, in order to obtain strong SERS signals. Ethylene glycol-modified Raman reporter molecules 4-nitrothiobenzoate-monoethylene glykole-hydroxy (4-NTB-MEG-OH) and 4-nitrothiobenzoatetriethylene-glykole-carboxy (4-NTB-TEG-COOH) were synthesized and used to stabilize the NPs (Scheme 1A), and simultaneously provide functional groups for bioconjugation33. In detail, a mixture of 5 µl 4-NTB-MEG-OH (10 mM in ethanol) and 5 µl 4-NTB-TEG-COOH (10 mM in ethanol) was added into 1 ml of metal NPs and incubated overnight. The NPs were centrifuged (Au stars: 800 g, 15 min; AuAg nanoshells: 6000 g, 15 min; 20 nm AuNPs: 9000 g, 15 min), washed one time with HEPES and re-suspended in 500 µl HEPES. For preparation of PEGylated SERS probes (Scheme 1B), the heterofunctional linker HS-PEG-COOH (500 µl, 50 µM) was added dropwise into 1 ml of Au nanostars under sonication and incubated overnight. Then 5 µl of 4-NTB (10 mM in ethanol) was added into the NPs suspension and incubated for 2.5 h. The NPs were centrifuged (800 g, 15 min), washed one time with HEPES and re-suspended in 500 µl HEPES.

Scheme 1 A) SERS probes prepared by coating Au nanostars with a dual SAM of 4-NTB-MEG-OH and 4-NTB-TEGCOOH, and then covalently linked with antibodies (Au starantibody); B) PEGylated SERS probes prepared by coating Au nanostars with SH-PEG-COOH and supplemented with 4NTB, and then covalently linked with antibodies (Au starPEG-antibody). Bio-conjugation of the SERS labels was achieved by covalent coupling of the carboxyl groups from 4-NTB-TEGCOOH or HS-PEG-COOH with primary amines (-NH2) from the antibody. For activation of carboxyl groups, 10 µl of freshly prepared EDC (6 mM in HEPES) and sulfo-NHS (15 mM in HEPES) were added to the suspension of SERS labels. Af-

ter shaking at RT for 25 min, excess EDC and sNHS were removed by centrifugation and the NPs were re-suspended in 500 µl HEPES buffer. Subsequently, antibodies were added to the NPs at 500 molar equivalents per NP and incubated at RT for 2.5 hours. Then, ethanolamine (volume ratio 1%) was added to the particle suspension and shaken for 30 min to block the remaining activated sites. Finally, the reaction mixture was diluted at a volume ratio of 1:1 with 2 % BSA/PBS, and purified four times via centrifugation, and then the supernatant was removed and replaced with 2 % BSA/PBS. After determination of the concentration by UV-Vis extinction spectroscopy, the SERS probes were used immediately for staining or stored at 4 °C. Pretreatment of tissue sections Human prostate and breast FFPE tissue blocks were cut into 4-µm-thick sections, mounted on Histobond®+ glass slides (Carl Roth, Germany) and dried overnight at 37 °C. Before immunostaining, the sections were deparaffinized in xylol for 30 min, rehydrated in ethanol and a series of ethanol/water mixtures, and then rinsed with water and PBS. The tissue sections were then treated with either a heat induced antigen retrieval step or an enzyme induced antigen retrieval step. Heat induced antigen retrieval was accomplished by heating in citrate (pH=6.0) or EDTA (pH=8.5) buffer for 20 minutes at 96 °C in a water bath, followed by cooling in the same buffer for 20 min. Enzyme induced antigen retrieval was accomplished by adding a drop of fast enzyme onto the tissue specimen and incubation at RT for 5 minutes. After antigen retrieval, the slides were washed thoroughly with PBS and incubated with 2 % BSA/PBS for 1 h to block non-specific binding sites. Immuno-staining with SERS probes The SERS probes were dispersed in 2 % BSA/PBS (OD=1) for tissue staining. In case of SERS labelled-primary antibody, the tissue sections were incubated with SERS probes for 1 h, then washed with PBS to remove the unbound NPs, subsequently the Alexa647-conjugated G@M secondary antibody was added onto the slides and incubated for 30 min. In case of the SERS-labelled secondary antibody, the tissue sections were firstly incubated with primary antibodies for 1 h, washed with PBS, and then SERS probes were added and incubated for another 40 min. After staining, the tissue slides were washed thoroughly with PBS to remove unbound NPs, and mounted with Fluoromount G (Southern Biotech) and a coverslip. Correlative IF-iSERS microscopy Localization of SERS-labelled antibodies was achieved by imaging with a correlative confocal Raman-wide field fluorescence microscope. Autofluorescence images recorded with the FITC channel were used for guidance to find the same position on adjacent tissue slides. Then fluorescence images in the Alexa647 channel were recorded to get a global overview of the spatial distribution of the SERS probes, and regions of interest were chosen for subsequent Raman mapping. The 632.8 nm radiation from a HeNe laser was focused onto the sample using a 40x objective (Olympus) with a numerical aperture of 0.6. The integration time per pixel in the Raman mapping experiments was 50 ms and the laser power at the sample was 18.9 mW.

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Results and discussion IF staining of PSA with HIER versus PIER Most epitopes in FFPE tissues are masked and a proper pretreatment of the sections is necessary before the antigen can be detected with labelled antibodies1. HIER, which was first established in 199139, and PIER, which started in 197040, are currently the two most widely used techniques for antigen retrieval37. Both methods serve to break the crosslinking bridges and expose the antigenic sites in order to allow the antibodies to bind41. In this study, two adjacent prostate tissue slides were deparaffinized and then either boiled in citrate buffer for 20 min (HIER) or incubated with enzyme for 5 min (PIER) at RT. Then the tissue sections were washed and incubated with anti-PSA primary antibodies and Alexa647-G@M secondary antibodies. The IF staining results are presented in Figure 1. Although HIER is recommended in the product manual of the anti-PSA primary antibody, we found that both HIER and PIER show very specific staining patterns with comparable signal brightness (Fig. 1 left, 10x objective). The staining after PIER treatment is more uniform when we look at the images with higher magnification (Fig. 1 B right, 40x objective), while in the sample with HIER, the staining is uneven with some brighter clots and other weaker regions (Fig. 1 A right).

Figure 1: Comparison of two different antigen retrieval methods with respect to their influence on the staining quality using conventional immunofluorescence (IF) for PSA localization in FFPE prostate tissue. A) heat-induced epitope retrieval, HIER, versus B) protein-induced epitope retrieval, PIER. Scale bar: 50 µm.

Page 4 of 10

Scheme 2 In iSERS microscopy either A) primary antibodies are conjugated with SERS NPs to perform direct staining, or B) secondary antibodies are conjugated with SERS NPs to perform indirect staining. In both case, negative control experiments should be included (A and B right). iSERS staining of PSA with HIER versus PIER SERS probes comprise metal NPs, Raman reporter molecules, and target-specific binding molecules such as antibodies or peptides17. Both primary or secondary antibodies can be conjugated to the SERS NPs and subsequently used for direct or indirect immuno-staining (Scheme 2A and B), respectively. For multiplexed protein localization, the conjugation of SERS NPs to primary antibodies is favored. However, some antibodies may not be suitable for chemical modification and then SERS-labelled secondary antibodies can be used instead 10. In both cases careful control studies need to be planned and included for data validation and interpretation. For SERSlabelled primary antibodies, a negative control experiment is to evaluate the non-specific binding of SERS NPs which are conjugated with the isotype- and species-matched immunoglobin (IgG); For SERS-labelled secondary antibodies the critical control is to determine the level of non-specific binding when the primary antibody is omitted in the staining process10. In this study, Au nanostars were hydrophilically stabilized with ethylene glycol-modified Raman reporter molecules (Scheme 1A) and then conjugated with either anti-PSA primary antibodies (Au star-anti-PSA) or Alexa647-labelled G@M secondary antibodies (Au star-A647G@M). The SERSlabelled antibodies were then applied for direct/indirect immuno-staining of PSA on FFPE human prostate tissue and imaged by iSERS microscopy. Fluorescence-labelled secondary antibodies are additionally used here (Scheme 2) for two reasons: firstly, in order to compare the global staining quality of IF (fluorescence-labelled antibodies) and iSERS (NPlabelled antibodies). Secondly, this correlative Raman/fluorescence approach allows us to obtain a global overview using wide-field fluorescence microscopy for guiding subsequent SERS mapping experiments on regions of interest35. Correlative IF-iSERS images obtained from direct staining (SERS labelled primary antibody, Scheme 2 top) are shown in Figure 2. Four adjacent tissue sections were treated with either HIER (Fig. 2A and B) or PIER (Fig. 2C and D), and then incubated with SERS-labelled anti-PSA primary antibodies (Fig. 2A and C) or SERS-labelled isotype control antibodies (Fig. 2 B and D). Fluorescence images from each section were first recorded and then two small regions were randomly selected for SERS mapping. The false-color iSERS images were obtained by least square fitting of the detected Raman spectra with the Raman spectrum of Au star-4NTB (Figure S3A). The bright autofluorescence emitted from the FFPE tissues indicates the outline structure of the prostate glands (l0x objective views shown in Figure S4), and therefore green autofluorescence images were recorded for convenience to identify the same region on different slides (Figure 2, first column). Note that the autofluorescence as well as fluorescence from the fluorophore Alexa 647 can also be detected in the SERS spectra (Figure S3B). Therefore a background subtraction was performed before least square fitting. In Fig. 2 A and B, the tissue specimens were treated with HIER. Both IF and iSERS results show that the SERS-labelled anti-PSA primary antibodies bind to epithelial cells (Fig. 2 A), but in the negative

ACS Paragon Plus Environment

Page 5 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 2: Comparison of two different antigen retrieval methods with respect to their influence on the staining quality (direct staining) in immuno-SERS (iSERS) microscopy for PSA localization in FFPE prostate tissue. Correlative IF-iSERS images were obtained by using a SERS-labelled anti-PSA primary antibody and a fluorophore-labelled secondary antibody. A) HIER, and B) negative control with isotope antibody. C) PIER, and D) negative control with isotope antibody. E) left: typical SERS spectra of A) and B); E) right: typical SERS spectra of C) and D); red and blue arrows indicate the location of the SERS spectra. Scale bar in fluorescence images: 50 µm; Scale bar in SERS images: 10 µm.

Figure 3: Comparison of two different antigen retrieval methods with respect to their influence on the staining quality (indirect staining) in immuno-SERS (iSERS) microscopy for PSA localization in FFPE prostate tissue. Correlative IFiSERS images were obtained by using an unlabelled primary anti-PSA antibody and a SERS- and fluorophore-dual-labelled secondary antibody. A) HIER, and B) negative control without primary antibody. C) PIER, and D) negative control without primary antibody. E) left: typical SERS spectra of A) and B); E) right: typical SERS spectra of C) and D); red and blue arrows indicate the location of the SERS spectra. Scale bar in fluorescence images: 50 µm; Scale bar in SERS images: 10 µm.

control (SERS-labelled isotope antibody), non-specific signals were also observed in the epithelium (Fig. 2 B). In contrast, for tissue pre-treated with PIER, the staining of PSA is more specific (Fig. 2 C), with a much higher signal brightness in both IF and iSERS images compared to Fig. 2 A, and almost no signals could be detected in the negative control (Fig.2 D). The good correlation between the IF and iSERS signals (merged images are shown in Figure S5) confirms that the antibodies were successfully conjugated to the SERS NPs. One typical SERS spectrum from each group is shown in Fig. 2 E, the strong peak at 1341 cm-1 is assigned to the symmetric nitro stretching vibration of 4-NTB (identical to the SERS spectrum of Au star-4NTB, Figure S3). It is obvious that larger spectral differences between the positive (red) and negative (blue) group exist in the tissue samples treated with PIER (the peak intensity at 1341 cm-1 is 86 and 0 CCD counts, respectively) than HIER (the peak intensity at 1341 cm-1 is 34 and 18 CCD counts, respectively). Overall, antigen retrieval with PIER offers a more specific staining with higher SERS signal levels and better image contrast. Correlative IF-iSERS images obtained from indirect staining (SERS-labelled secondary antibody, Scheme 2 bottom) are shown in Figure 3. The staining performance of SERSlabelled secondary antibodies is similar to that of the SERSlabelled primary antibodies. Both fluorescence and SERS sig nals are much stronger and more specific in the PIER-treated tissue specimens (Fig. 3 C and D) than the HIER- treated

ones (Fig. 3 A and B). In a typical SERS spectrum from each section (Fig. 3E), the peak intensity at 1341 cm-1 is 17 and 15 CCD counts for HIER treated sample and negative control, respectively, while the intensity for PIER treated sample is 45 CCD counts, much higher than the negative control (1 CCD count). Overall, these observations are in agreement with the results obtained by direct iSERS staining (Figure 2). The histograms of the IF and iSERS images in Fig 1-3 are analyzed and presented in Figure S6. The average fluorescence intensity of the slides stained by conventional IF increased only slightly (1.1 times) when the tissues were treated with PIER instead of HIER. However, for both direct and indirect iSERS staining using SERS labelled antibodies, the fluorescence intensity increased more obviously (2.6 and 1.5 times for direct and indirect iSERS, respectively). SERS signals increased in accordance with fluorescence signals (Figure 2-3, and Figure S5), however, the average SERS intensity increased by an even larger extent (4.5 times and 3.2 times for direct and indirect iSERS staining, respectively). The above results indicate that iSERS staining of PSA using SERS-labelled antibodies is more sensitive to antigen retrieval methods compared with conventional IF: the Raman signals increased significantly and the non-specific binding almost vanished when PIER was adopted instead of HIER. Although the major focus of this contribution is not to investigate the influence of the different antigen retrieval methods on tissue morphology and antigen accessibility, we at least at-

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

tempt to presume some possible mechanisms from our observations. Firstly, we speculate that PIER leads to a better steric accessibility of the antigen within the tissue for binding to the antibody-metal NP conjugates (iSERS), which are significantly larger than the fluorophore-labelled antibodies (IF). Increased steric accessibility not only implies better antigenantibody binding but also more space for a larger number of NPs per unit volume. Additionally, it is known that very high electric field enhancements occur in the junction between two plasmonic NPs42, which are usually called as “hot spots”18,21. We assume that on tissue samples treated with PIER, a larger number of SERS-labelled antibodies bind to the antigens and locate in proximity to each other, therefore more hotspots are generated and the SERS signals are additionally enhanced. Secondly, we speculate that HIER using citrate or EDTA buffer leads to increased non-specific binding because of increased electrostatic interactions between the tissue and the antibodymetal nanoparticle conjugates. Additionally, we also hypothesize that this increased charge may also lead to particle aggregation on the tissue since experiments with more stable, PEGylated SERS nanoparticles exhibit less non-specific binding (vide infra). iSERS staining of HER2 in breast tissue with HIER versus PIER The apparent influence of antigen retrieval approaches for FFPE tissues on binding behavior of SERS probes is interesting and of central importance for diagnostic applications of metal NP-based probes. To test whether this phenomenon is only specific for PSA and prostate tissue or a more general phenomenon also observable for different tissues, human breast FFPE tissues were treated with either HIER or PIER and incubated with SERS-labelled antibodies. HER2 is a critical biomarker in clinical prognosis of breast cancer and predicting response to therapy with trastuzumab43. Fig. 4 A and B display the results from correlative IF-iSERS microscopy for the localization of HER2 on breast cancer and normal breast tissue, respectively. The tissue specimens were pretreated with an enzymatic antigen retrieval method, then incubated with anti-HER2 primary antibodies and subsequently with Au starA647G@M secondary antibodies. As shown in Fig. 4 A, both fluorescence and SERS signals could be observed on the cell membranes of cancer tissue, while on the normal breast tissue (Fig. 4 B), neither fluorescence nor SERS signals could be detected. However, when the tissue specimens were pretreated with HIER, the Au star-A647G@M secondary antibodies bind non-specifically onto the tissues (Fig. 4 C), with a completely different pattern compared to the specific staining in Fig. 4 A. The results indicate that iSERS staining of HER2 on breast tissue was also severely affected by antigen retrieval methods: HIER treatment induces non-specific adsorption of the antibody-metal NP conjugates onto the breast tissues while PIER does not.

Page 6 of 10

Figure 4: Localization of HER2 on breast tissue using correlative bright field (left), immunofluorescence (middle), and iSERS (right) microscopy. Indirect iSERS staining of PIERtreated A) cancerous tissue and B) normal tissue; C) Nonspecific adsorption of Au star-A647G@M secondary antibodies on HIER-treated cancerous tissue. Scale bar in fluorescence and BF images: 30 µm; Scale bar in SERS images: 4 µm, 8 µm and 9 µm for A), B) and C), respectively. Influence of particle size/shape on HIER induced nonspecific binding In iSERS staining of PSA/HER2 using Au star-antibodies, HIER treatment of the tissue slides induced obvious nonspecific adsorption of antibody-metal NP conjugates (vide supra), however, in case of conventional IF using fluorescence-labelled antibodies, non-specific binding was not observed (Figure S7). Relevant parameters are both the geometry of the NPs (anisotropic nanostars) and the surface chemistry (Raman reporters stabilized only by short ethylene glycol moieties). Two possible reasons that induce non-specific binding of SERS-labelled antibodies might be: 1) the heating process induces structural changes of the tissue, where the Au nanostars with a relatively large size (~70 nm, Figure S1) and multibranched shape can be trapped inside; 2) heating in citrate buffer breaks the crosslinking bonds and induces charged groups on the tissue slides, which leads to electrostatic adsorption of NPs onto the sample. To investigate the influence of NPs with different sizes or shapes, we used 60 nm quasi-spherical Au/Ag nanoshells and 20 nm quasi-spherical Au NPs for performing staining under otherwise same conditions. The correlative IF-iSERS staining results using AuAg nanoshell-A647G@M is shown in Figure S8. In accordance with the results of the Au nanostars, superior staining quality was observed when the tissue was treated with PIER compared to HIER, and AuAg nanoshells-labelled antibodies also adsorbed non-specifically to the HIER treated tissue specimens (Figure S8 B). A comparison between 70 nm Au nanostars and 20 nm Au nanospheres is presented in Figure 5. Adjacent tissue sections were heated in citrate buffer/EDTA buffer for 20 min, or in buffer without heating, then incubated with either Au star-A647G@M antibodies or 20 nm Au NP-A647G@M antibodies for 40 min, respectively. Fig.5 A-C shows the fluorescence and SERS images of tissue sections incubated with Au star-A647G@M antibodies. Apparently, the Au nanostars selectively bind to epithelium of the

ACS Paragon Plus Environment

Page 7 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

tissue samples which were heated in citrate buffer (Fig. 5 A) and EDTA buffer (Fig. 5 B), and even with a larger amount for the latter one. In contrast, almost no particles could be detected either by fluorescence or by SERS on the tissue slides treated without heating (Fig. 5 C). 20 nm Au NPs with spherical shapes were used to perform the same staining process, and the results are shown in Fig. 5 D-F. With a pretreatment by heating in citrate buffer (Fig. 5 D) or EDTA buffer (Fig. 5 E), the 20 nm Au NPs show similar binding pattern as the Au nanostars (Fig. 5 A and B), while no NPs were detected in the sample treated without heating (Fig. 5 F). The above results demonstrate that the heating process induces non-specific binding of NPs onto the tissue slides, independent of their size or shape.

Figure 5: Non-specific adsorption of NP-labelled secondary antibody on prostate biopsies. Influence of particle size and shape: gold nanostars (A-C) versus 20 nm gold nanospheres (D-F). Influence of heat plus buffer (A, B and D, E) and no heat (C and F). Scale bar in fluorescence and BF images: 30 µm; Scale bar in SERS images: 20 µm. It is interesting to note that, although non-specifically, most of the NPs are bound selectively to the epithelium but not to the stromal region of the prostate glands, with a layer of cells attracting apparently more NPs than the others (indicated by white arrows in Fig. 5). This peculiar phenomenon triggered us to perform a literature research on staining with colloidal gold since immuno-gold has been applied in tissue staining for several decades44. We found that selective binding pattern of NPs has been reported 30 years ago in two separate studies. H. Jessen and his coworker found that colloidal gold-protein conjugates selectively bind to epidermal phosphorus-rich keratohyaline granules and cornified cells45. Another one is that Asada-Kubota observed selective labeling over secretory granules of granular skin glands and myelin sheaths with colloidal gold alone and with gold-protein conjugates46. The authors could not explain the mechanism, but they speculated that the negatively charged gold NPs were not completely coated by the stabilizing protein, leading to electrostatic interactions with positively charged proteins in sections of cells. However, in these two studies, no antigen retrieval step was involved. On the other hand, although the mechanism involved

in HIER is not fully understood yet, it’s known that heating induced hydrolysis of methylene crosslinks and break some of the calcium coordinate bonds4. Unexpected staining pattern induced by antigen retrieval has been reported by Guiter et al.47, who found that HIER treatment induced unwanted immuno-reactivities of antibodies49. In our study, the IF staining was not obviously influenced by antigen retrieval methods, while for iSERS staining, the NPs only bind to the tissue sample pretreated by heating (Figure 5 A, B, D, E), but did not adhere to the tissue without heating treatment (Figure 5 C and F). The zeta potential of the Au star-A647G@M was determined to be -13.9±1.2 mV, we therefore speculate that the heating process disrupts the crosslinking bridges and induces positively charged groups in certain cell structures. Nevertheless, the exact mechanism and detailed process requires further studies. Influence of PEGylation on HIER induced non-specific binding HIER and PIER are currently the two most developed antigen retrieval techniques for IHC staining on FFPE tissue sections41. It has been reported that some antigens can be unmasked by either a HIER or PIER treatment, while the else benefit selectively from one treatment but not from the other50. The above iSERS staining results indicate that PIER is superior to HIER in terms of reducing non-specific binding of NP labels, however, in case of antigens which can only be unmasked by HIER, additional strategies need be adopted to reduce non-specific binging caused by HIER. Efficient blocking of tissue sections is an essential aspect for preventing non-specific interactions of NP-based probes with tissue sections32. BSA, nonfat dry milk, and normal serum are commonly used reagents that bind to non-specific epitopes and reduce non-specific binding. We have tested blocking the tissue specimens with 2 % BSA for longer time (3 h), or blocking with 0.4 % milk + 5% goat serum, but nonspecific binding effects were not reduced for the SERSlabelled antibodies (Figure S9). PEGylation of the NPs is a commonly used method for minimizing non-specific binding due to resistance to dehydration and steric confinement of the swollen polymer51. For metallic NPs, heterobifunctional PEG monothiols can be used to form a monolayer on the NP surface, while the antibodies can be then covalently attached to the outer ends (usually via amide bonds)27. We fabricated PEGylated SERS probes with a PEG layer (MW~3 kDa) on the NP surface supplemented by 4-NTB to provide SERS signals. The antibodies were then covalently attached to the NPs and applied for PSA staining (Au starPEG-antibodies, scheme 1 B). For direct iSERS staining, four adjacent tissue sections were pretreated with HIER in citrate buffer, and then incubated with Au star-anti-PSA antibody and Au star-PEG-anti-PSA antibody, respectively. The correlative IF-iSERS images are shown in Figure 6. In contrast to the Au star-isotype control (Fig. 6B and Fig. 2B), SERS NPs with a PEG layer did not exhibit non-specific binding to the tissues (Fig. 6D). However, the fluorescence as well as SERS signals are apparently weaker in case of Au star-PEG-anti-PSA (Fig. 6C) compared with Au star-anti-PSA (Fig. 6A). The average fluorescence intensity and average SERS intensity decreased about 28% and 39%, respectively). Indirect iSERS staining of PSA using Au star-A647G@M and Au star-PEG-A647G@M was also performed and similar results were obtained (Figure.

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6: Direct iSERS staining of PSA in HIER-treated prostate biopsies using SERS probes and PEGylated SERS probes. A) Au star-anti-PSA with B) the negative control. C) Austar-PEG-anti-PSA with D) the negative control. Scale bar in fluorescence images: 50 µm; Scale bar in SERS images: 10 µm. 7): PEGylation of the SERS probes minimized the nonspecific binding but the staining signals also decreased The SERS intensity and hydrodynamic size of SERS probes and PEGylated SERS probes were then determined. Figure S10 presents the SERS spectra of Au star-4NTBM/T (red line) and Au star-PEG/4NTB (black line) suspensions dispersed in 25% ethanol with the same particle density. By normalizing the Raman intensity at 1341 cm-1 (SERS, 4-NTB) to the one at 880 cm-1 (normal Raman, from suspension medium ethanol), the SERS signal intensity of Au star-PEG/4NTB was found to be about 85% of the Au star-4NTBM/T. This is because the dual-SAM design of 4NTBM/T (Scheme 1A) offers maximum loading of Raman reporter molecules on the NP surface, while the amount of Raman reporter molecules on the surface of Au star-PEG/4NTB is smaller (Scheme 1 B). In addition, the diameter of the PEGylated SERS probes (98.6 nm determined by DLS) was about 12 nm larger than the Au star-antibodies (86.8 nm). Because of the increased particle size as well as decreased SERS signal intensity, the iSERS staining of PSA using PEGylated probes was weaker than using the Au starantibodies. To combine advantages of dual-SAM design and PEG, Raman reporter molecules can be linked with a PEG spacer, for example, 4-NTB-PEG-COOH could be synthesized and used instead of 4-NTB-TEG-COOH. By this way, the SERS signals will be maximized and the non-specific binding of NPs will be minimized.

Page 8 of 10

Figure 7: Indirect iSERS staining of PSA in HIER-treated prostate biopsies using SERS probes and PEGylated SERS probes. A) Au star-A647G@M with B) the negative control. C) Austar-PEG-A647G@M with D) the negative control. Scale bar in fluorescence images: 50 µm; Scale bar in SERS images: 10 µm. the binding behavior of the SERS-labelled antibodies, that is, NPs tend to bind non-specifically to tissue specimens treated with HIER but not PIER, independent of NP size and shape. Therefore, PIER is recommended for future iSERS staining for antigens which can be retrieved with enzymatic methods, but for antigens which can only be unmasked by HIER, PEGylation is an efficient way to avoid non-specific binding, although the particle size will increase and the signal intensity decreased slightly.

Supporting Information The Supporting Information is available free of charge on the ACS Publications website. TEM images, UV-Vis extinction spectra of Au nanostars and AuAg nanoshells; SERS spectra of Au star-4NTB dispersed in water, Au star-4NTBM/T and Au star-PEG/4NTB dispersed in 25% ethanol; Typical SERS spectra recorded on prostate tissue specimens with direct/indirect iSERS staining before and after background subtracting; Merged IF-iSERS images; Histogram analysis of IF and iSERS images; Non-specific adsorption of NPlabelled secondary antibody on prostate biopsies with different parameters (blocking conditions, AuAg nanoshells).

AUTHOR INFORMATION Corresponding Authors

Conclusion Immuno-staining using SERS-labelled antibodies enables multiplexed and quantitative profiling of protein biomarkers in tissue samples, offering more information about disease heterogeneity as well as helping to conserve limited tissues52. However, non-specific binding of NPs onto the tissue specimens greatly hinders clinic applications of SERS labels. In this study, SERS labels were fabricated by coating gold nanostars with a mixture of 4-NTB-MEG-OH and 4-NTB-TEG-COOH, followed by covalent conjugation to antibodies. Surprisingly we find that antigen retrieval methods show a strong impact on

* [email protected] [email protected]

Author Contributions The manuscript was written through contributions of all authors.

Notes The authors declare no competing financial interest.

ACS Paragon Plus Environment

Page 9 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

ACKNOWLEDGMENT We greatly appreciate the technical support by Bernd Walkenfort for building up and maintaining the correlative IF-iSERS microscopy system. We thank Dr. Martin Braun from the Institute of Pathology at the University Hospital Bonn for providing the FFPE prostate tissue blocks. We thank Ms. Andrea Kutritz from the Institute of Pathology at the University Hospital Essen for the preparation of sections from breast tissue. Thanks also to Vi Tran for synthesizing the ethylene glycol-modified Raman reporter molecules and Aleksandar Radojcic for the synthesis of the Au nanostars.

REFERENCES (1) Buchwalow, I. B.; Böcker, W. Immunohistochemistry: Basics and Methods; Springer-Verlag Berlin Heidelberg, 2010. (2) Varma, M.; Jasani, B. Histopathology 2005, 47, 1-16. (3) Matos, L. L. d.; Trufelli, D. C.; Matos, M. G. L. d.; Pinhal, M. A. d. S. Biomark Insights 2010, 5, 9-20. (4) Kabiraj, A.; Gupta, J.; Khaitan, T.; Bhattacharya, P. T. Int J Biol Med Res 2015, 6, 5204-5210. (5) Azzazy, H. M.; Mansour, M. M. Clin Chim Acta 2009, 403, 1-8. (6) Fu, X.; Chen, L.; Choo, J. Anal Chem 2017, 89, 124-137. (7) Anselmo, A. C.; Mitragotri, S. Bioengineering & Translational Medicine 2016, 1, 10-29. (8) Zhou, W.; Gao, X.; Liu, D.; Chen, X. Chemical reviews 2015, 115, 10575-10636. (9) Zhou, W.; Gao, X.; Liu, D.; Chen, X. Chemical reviews 2015. (10) Xing, Y.; Chaudry, Q.; Shen, C.; Kong, K. Y.; Zhau, H. E.; Chung, L. W.; Petros, J. A.; O'Regan, R. M.; Yezhelyev, M. V.; Simons, J. W.; Wang, M. D.; Nie, S. Nat Protoc 2007, 2, 1152-1165. (11) Liu, J.; Lau, S. K.; Varma, V. A.; Moffitt, R. A.; Caldwell, M.; Liu, T.; Young, A. N.; Petros, J. A.; Osunkoya, A. O.; Krogstad, T.; Leyland-Jones, B.; Wang, M. D.; Nie, S. ACS Nano 2010, 4, 2755–2765. (12) Kairdolf, B. A.; Smith, A. M.; Stokes, T. H.; Wang, M. D.; Young, A. N.; Nie, S. Annu rev anal chem 2013, 6, 143462. (13) Hainfeld, J. F. Science 1987, 236, 450-453. (14) Schlücker, S.; Küstner, B.; Punge, A.; Bonfig, R.; Marx, A.; Ströbel, P. J Raman Spectrosc 2006, 37, 719-721. (15) Sun, L.; Sung, K. B.; Dentinger, C.; Lutz, B.; Nguyen, L.; Zhang, J.; Qin, H.; Yamakawa, M.; Cao, M.; Lu, Y.; Chmura, A.; Zhu, J.; Su, X.; Berlin, A. A.; Chan, S.; Knudsen, B. Nano Lett 2007, 7, 351-356. (16) Küstner, B.; Gellner, M.; Schütz, M.; Schöppler, F.; Marx, A.; Ströbel, P.; Adam, P.; Schmuck, C.; Schlücker, S. Angew Chem Int Ed Engl 2009, 48, 1950-1953. (17) Schlücker, S. Chemphyschem 2009, 10, 1344-1354. (18) Wang, Y.; Yan, B.; Chen, L. Chem Rev 2013, 113, 13911428. (19) Lane, L. A.; Qian, X.; Nie, S. Chem Rev 2015, 115, 10489-10529. (20) Jamieson, L. E.; Asiala, S. M.; Gracie, K.; Faulds, K.; Graham, D. Annual review of analytical chemistry 2017, 10, 415-437. (21) Wang, Y.; Schücker, S. Analyst 2013, 138, 2224-2238. (22) Schlücker, S. Chemphyschem 2009, 10, 1344-1354. (23) Xie, W.; Schlücker, S. Physical chemistry chemical physics : PCCP 2013, 15, 5329-5344.

(24) Schneider, C. S.; Perez, J. G.; Cheng, E.; Zhang, C.; Mastorakos, P.; Hanes, J.; Winkles, J. A.; Woodworth, G. F.; Kim, A. J. Biomaterials 2015, 42, 42-51. (25) Bentzen, E. L.; Tomlinson, I. D.; Mason, J.; Gresch, P.; Warnement, M. R.; Wright, D.; Sanders-Bush, E.; Blakely, R.; Rosenthal, S. J. Bioconjug Chem 2005, 16, 1488-1494. (26) Jokerst, J. V.; Lobovkina, T.; Zare, R. N.; Gambhir, S. S. Nanomedicine 2011, 6, 715-728. (27) Eck, W.; Craig, G.; Sigdel, A.; Ritter, G.; Old, L. J.; Tang, L.; Brennan, M. F.; Allen, P. J.; Mason, M. D. ACS nano 2008, 2, 2263–2272. (28) Salehi, M.; Schneider, L.; Ströbel, P.; Marx, A.; Packeisen, J.; Schlücker, S. Nanoscale 2014, 6, 2361-2367. (29) Kairdolf, B. A.; Mancini, M. C.; Smith, A. M.; Nie, S. Anal Chem 2008, 80, 3029-3034. (30) Schütz, M.; Steinigeweg, D.; Salehi, M.; Kömpe, K.; Schlücker, S. Chem Commun 2011, 47, 4216-4218. (31) Wei, H.; Insin, N.; Lee, J.; Han, H.-S.; Cordero, J. M.; Liu, W.; Bawendi, M. G. Nanoletters 2012, 12, 22-25. (32) Roth, J.; Taatjes, D. J.; Warhol, M. J. Histochemistry 1989, 92, 47-56. (33) Jehn, C.; Küstner, B.; Adam, P.; Marx, A.; Ströbel, P.; Schmuck, C.; Schlücker, S. Phys Chem Chem Phys 2009, 11, 7499-7504. (34) Wang, Y.; Salehi, M.; Schutz, M.; Schlücker, S. Chem Commun (Camb) 2014, 50, 2711-2714. (35) Wang, X. P.; Zhang, Y. Y.; König, M.; Papadopoulou, E.; Walkenfort, B.; Kasimir-Bauer, S.; Bankfalvi, A.; Schlücker, S. Analyst 2016, 141, 5113-5119. (36) Berg, D.; Hipp, S.; Malinowsky, K.; Bollner, C.; Becker, K. F. Eur J Cancer 2010, 46, 47-55. (37) Denda, T.; Kamoshida, S.; Kawamura, J.; Harada, K.; Kawai, K.; Kuwao, S. Cancer Cytopathol 2012, 120, 167-176. (38) Shi, S. R.; Shi, Y.; Taylor, C. R. J Histochem Cytochem 2011, 59, 13-32. (39) Shi, S. R.; Key, M. E.; Kalra, K. L. J Histochem Cytochem 1991, 39, 741-748. (40) Sternberger, L. A.; Hardy, P. H., Jr.; Cuculis, J. J.; Meyer, H. G. J Histochem Cytochem 1970, 18, 315-333. (41) Leong, T. Y.; Leong, A. S. Advances in Anatomic Pathology 2007, 14, 129-131. (42) Aravind, P. K.; Abraham, N.; Horia, M. Surface Science 1981, 110, 189-204. (43) Slamon, D. J.; Godolphin, W.; Jones, L. A.; Holt, J. A.; Wong, S. G.; Keith, D. E.; Levin, W. J.; Stuart, S. G.; Udove, J.; Ullrich, A. Science 1989, 244, 707-712. (44) Faulk, W. P.; Taylor, G. M. Immunochemistry 1971, 8, 1081-1083. (45) Jessen, H.; Behnke, O. J Invest Dermatol 1986, 87, 737740. (46) Asada-Kubota, M. J Ultrastruct Mol Struct Res 1988, 98, 147-157. (47) Guiter, G. E.; Kwan, P. W.; DeLellis, R. A. Laboratory Investigation 1995, 72, 165A. (48) Vanstapel, M. J.; Gatter, K. C.; de Wolf-Peeters, C.; Mason, D. Y.; Desmet, V. D. American journal of clinical pathology 1986, 85, 160-168. (49) Norton, A. J. The Journal of pathology 1993, 171, 79-80. (50) Cattoretti, G.; Pileri, S.; Parravicini, C.; Becker, M. H. G.; Poggi, S.; Bifulco, C.; Key, G.; D'Amato, L.; Sabattini, E.; Feudale, E.; Reynolds, F.; Gerdes, J.; Rilke, F. The Journal of pathology 1993, 171, 83-98.

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 10

(51) S. Herrwerth , T. R., C. Feng , J. Fick , W. Eck , M. Himmelhaus , R. Dahint , M. Grunze. Langmuir 2003, 19, 1880-1887. (52) Stack, E. C.; Wang, C.; Roman, K. A.; Hoyt, C. C. Methods 2014, 70, 46–58.

Insert Table of Contents artwork here Non-specific binding of nanoparticles onto cell/tissue samples greatly inhibits their clinical applications. A systematic investigation using antibody-metal nanoparticle conjugates for correlative IFimmuno SERS staining on human prostate/breast FFPE tissue specimens has been performed, indicating that non-specific adsorption of the nanoparticle conjugates onto the tissue samples can be avoided by using protease-induced antigen retrieval method instead of using heating treatment.

Title: Effect of antigen retrieval methods on non-specific binding of antibody-metal nanoparticle conjugates on FFPE tissue

ACS Paragon Plus Environment