The characteristic signature of Raman labels arises on the surface of noble metal nanoparticles as SERS substrates. Therefore for quantitative SERS detection, sensitive SERS labels with high degree of scattering efficiency are required [11]. Optimal SERS nanotags generate maximum signal strength combined with high chemical and mechanical stability. Nobel metals such as gold and silver NPs are the most commonly used SERS substrates. Another example, Xia and co-workers developed gold/silver nanoshells (Au@Ag) as an efficient SERS substrate with tunable plasmon resonances in the red to near-infrared [14]. Au@Ag increase SERS sensitivity roughly eightfold more than GNP with the same size. Figure 1A shows a TEM image of ca. 50–60 nm Au@Ag nanoshells (bottom) and the corresponding extinction spectrum (top). Gold nanostars (GNS) is a widely used material due to its excellent SERS sensitivity in SERS based assays. Similar to Au@Ag, the plasmon band of GNS occurs in the red to near-infrared, causing only minimal autofluorescence from the biological specimen upon red laser excitation. On the other hand, GNS exhibit high SERS enhancement factors due to their sharp tips, enabling single particle detection (Figure 1B). Previous syntheses of GNS were either performed in toxic Dimethylformamide (DMF) as an organic solvent or with cetyltrimethylammonium bromide (CTAB) as a capping agent, which is difficult to completely remove from the particle surface [15, 16]. Schütz et al. developed a biocompatible route for the synthesis of monodisperse GNS in water, with sodium citrate and hydroquinone as reducing/capping agents, in order to avoid both CTAB and toxic DMF, which can denature proteins [17]. Both Au@Ag and GNS are widely used as single solid SERS substrates due to their excellent SERS sensitivity and relatively small size (ca. 50 nm, see Figure 2). Clusters of solid GNPs such as dimers and trimers exhibit a significantly larger SERS signal strength than individual solid GNPs or Au@Ag due to the presence of “hot spots” in the particle junctions (Figure 1C) [12, 13]. GNS and clusters of GNPs are just two optimal examples for highly sensitive plasmonic SERS nanotags, which have been characterized at the single-particle level [12, 16].

Figure 1:

Normalized extinction spectra of Au@Ag (A top), GNS (B top) and silica-encapsulated gold clusters (C top). The corresponding TEM images demonstrated under each spectrum. Inset: high-resolution TEM images of nanomaterials. Red dashed line shows laser radiation at 632.8 nm (from a HeNe laser) [9, 13, 17].

Figure 2:

Uncontrolled binding of the IgG onto the surface of the solid SERS nanotags (A). Ability of Stabilizers (red) to bind with NPs covalently onto the surface of nanomaterial (B). Immunohistochemistry staining of prostate tissue using Au@Ag conjugated to an antibody against p63 proteins, demonstrating the selective abundance of p63 only in the basal cells (b), of the epithelium (e), stroma (s) and non-basal epithelial cells of prostatic gland; (l). 2C shows the non-specific binding of the antibody- conjugated nanoparticles, caused by centrifugation steps. 2D, labeling of p63-proteins and PSA (2F) with nanoparticles, which were prepared in reaction/separation chamber. Negative control experiments in immuno-SERS microscopy using SERS nanotags conjugated to BSA (2E) [9, 13].

Silica-encapsulated dimers and trimers of GNPs (small clusters) have been demonstrated to exhibit single-particle SERS sensitivity. Encapsulation of nanomaterials would minimize aggregation and desorption of Raman label molecules from the nanoparticle surface during preparation and biofunctionalization. Encapsulation as a protective shell can be accomplished with different materials such as proteins, silica and polymers [18]. Schlücker and co-workers have previously developed a method to obtain large quantities of a purified colloid containing silica-encapsulated dimers of GNPs via density gradient centrifugation [19]. Glass-encapsulated GNP enables rapid immuno-SERS imaging with low millisecond integration time per point [12].

Most of manufacturers use bacteriostatic agents (e.g. Sodium azide) and proteins such as BSA, gelatin, or sugars as stabilizers to protect their products (e.g. IgG) [20]. Such stabilizers are able to bind with NPs covalently, like peptides or IgGs due to their free amine residues [9, 13]. Figure 2B shows the competitive binding of IgGs and stabilizers with NP. In practice, the conventional bioconjugation technique does not provide site-specificity and there is no control over targeting a particular amino acid residue. This uncontrolled binding in terms of missing site-specificity may result in a “scrambled” orientation of the antibodies on the surface of the SERS nanotags (Figure 2A) [13]. Thus the antigen recognition sites of the antibody (FAB) cannot recognize its target perfectly. However, the conventional conjugation method has been successfully applied for the detection in association with labeling of solid nanotags. As shown before, the binding affinity of IgG coupled NPs depends especially on the degree of NP aggregation and purity of antibodies [9]. This investigation shows the binding affinity of solid nanotags depends more on the degree of NP aggregation and less on lack of missing site-specificity of IgG. Hamann and co-workers reported that the aggregation of nanoparticles can be minimized by using the optimal amount of crosslinker and avoiding centrifugation steps [9]. The authors established a simple method to avoid multi step centrifugation of NPs for removing excess of chemicals or proteins [21]. The method consists of one or more simple reaction chambers with lateral filter membranes, which are permeable to molecules with defined sizes. At the bottom of each reaction chambers is a small well in the form of a pocket to collect modified NPs. The same method was used to separate small “native proteins” such as bovine serum albumin (BSA) from IgGs.

Various cross-linking methods have been developed for labeling of proteins (e.g. IgG) with fluorophores or enzyme (e.g. HRP). Cross-linking is the implementation of covalent bonds between functional groups of amino acids within a protein (intramolecular), or between interacting proteins (intermolecular) using chemical reagents.

Since the SERS-nanotags are indeed several thousand fold heavier than IgG, fluorophores or enzymes, particular attention must be paid to the chosen buffers and bioconjugation technique, which allows the linkage of bio- and nanomaterials [9]. Thus, for bioconjugation of SERS labels, gentle modification and adjustments are required. As it has been shown before, the stability and biofunctionality of SERS nanotags depends on the amount of crosslinker and the avoidance of centrifugation steps [9]. Figure 2C and D demonstrate two different preparations of anti p63 coupled SERS probes with and without the use of a reaction/separation chamber [9, 21]. The p63 IgG labeled SERS probe in Figure 2C has been prepared by the use of the centrifuge. In contrast, SERS probe in Figure 2D was prepared with the filtration/reaction chamber with lateral filter. The nonspecific binding of nanoparticles between the nuclei is clearly visible in Figure 2C. Evidence to the contrary, Figure 2D, shows only the labeled nuclei of cells with a diameter of about 6 μM. The spaces between the nuclei of cells are not marked with SERS label [22].

The same reaction/separation chamber is used for continuous modification and covalent coupling of anti PSA antibody molecules onto Au@Ag. The in situ detection of PSA in epithelial tissue of prostate section by SERS microscopy is shown in Figure 2E. In contrast, no contributions from the SERS label are observed in the lumen and stroma, respectively. Authors performed negative control experiments employing BSA–SERS nanotags conjugates, i.e. without antibodies [9, 12, 13]. The SERS false color images in Figure 2C, D, F and E are based on the intensity of the Raman marker band of 4-nitrothiobenzoic acid (4-NTB) at about 1340 cm⁻¹ indicating the presence of PSA or p63. Negative control experiments with SERS labeled BSA indicate no or only minimal non-specific binding. (Figure 2F).

In comparison to single solid SERS nanotags such as Au@Ag and GNS, the glass encapsulated gold cluster shows worse target accuracy in our laboratory after conventional covalent bioconjugation. Schlücker and co-workers have introduced the proof of concept of bioconjugation as an alternative design for SERS label–antibody conjugates, which minimizes nonspecific binding [13]. Silica-coated gold nanoparticle clusters as SERS substrates were non-covalently conjugated to primary IgGs using the chimeric protein A/G, which selectively recognizes the fragment crystallizable region (Fc region) of IgGs. In contrast to conventional bioconjugation, the presented design avoids the problem of uncontrolled binding due the lack of site-specificity, by using the chimeric protein A/G, which exhibits multiple bindings to the Fc region of the antibody. Figure 3A shows nanotags covered with a self-assembled monolayer (SAM) of Raman reporter molecules, which are encapsulated with silica [18]. The silica coating as a protective shell, holds the solid GNPs together, a close spacing between the GNPs is important for efficient generation of SERS via “hot spots” [12, 13].

Figure 3:

(A), Controlled binding and directed orientation of the IgG onto the silica surface of the SERS NP cluster via coating with protein A/G (top, orange), which exhibits multiple binding sites for the Fc region of the IgG (bottom). In this case, all antigen binding sites are accessible. (B) Localization of PSA by using SERS-labeled PSA antibodies and p63 labeled nanotags on non-neoplastic prostate tissue (C). (D) demonstrates two-color immuno-SERS microscopy for the co-localization of p63 (green) and PSA (red). p63 false-color SERS image overlaid with Raman reporter: 4-MBA (green) and PSA false-color SERS image with Raman reporter: 4-NTB (red) [12, 13].

The silica-encapsulated NPs were labeled covalently with protein A/G before conjugation to monoclonal antibodies for target recognition. Controlled binding and directed orientation of the antibodies onto the silica surface of the SERS cluster via coating with protein A/G, exhibits multiple binding sites for Fc region of IgG (Shown orange). Highly purified silica-encapsulated clusters (dimers and trimers) of 60 nm GNPs were used for single and two-color staining of PSA and p63 proteins on prostate tissue. Figure 3B shows the localization of PSA in epithelial tissue of the prostate by immuno-SERS microscopy. In contrast, no contributions from the SERS nanotags are observed in the stroma and lumen, respectively. The selective abundance of p63 in the basal cells of the epithelium is clearly observable in Figure 3B. As shown, the tumor suppressor p63 is solely abundant in the basal cells of healthy donors but not in the stroma, secretory epithelium or lumen [22]. The SERS false color images in Figure 3B and C is based on the intensity of the Raman marker band of 4-NTB at about 1340 cm⁻¹. The two-color experiment in Figure 3D confirms the findings from the one-color experiments: p63 is selectively observed in the nuclei of the basal cells (green), while PSA is abundant in the entire epithelium, but not in the stroma (red). The tissue was incubated separately with different SERS nanotags for the two color experiments.

The loss of target accuracy and appearance of nonspecific binding are most important disadvantages of SERS nanoparticles [9, 21]. Apart from diagnostics, the metallic nanoparticles could although be used for therapeutic purposes due to their photothermal effect [7]. Potential limitations, such as water solubility and dose dependent toxicity of the NPs have to be investigated in future studies [23, 24]. The synthesis of SERS labels together with new conjugation protocols can play an important role in the future. Since the focus on SERS applications rather than on the synthesis or characterization of SERS nanoparticles.