Tip-enhanced Raman scattering of DNA aptamers for Listeria monocytogenes Free

Nucleic acid oligomers (aptamers) are oligonucleotides that bind to a specific target molecule.¹ For example, DNA aptamer selective to Listeria monocytogenes is a short strand of DNA that binds to a specific molecule at the surface of L. monocytogenes,² a Gram-positive intracellular foodborne pathogen often found in food and causing a serious disease called listeriosis.³ Previous work has shown a great potential of aptamers for applications in gene therapy, cancer detection, and drug delivery.^4–8 Therefore, the study of morphological and functional properties of DNA aptamers, more specifically, their rapid and label-free sequencing, surface distribution, and configuration could help to understand the related mechanisms of binding between aptamers and target molecules [in particular, protein Internalin A (InlA) on L. monocytogenes cell membrane], their binding affinity, and to better visualize the binding locations in a cell membrane, for example; and ultimately to develop new types of biosensors and delivery systems.

Surface- and tip-enhanced Raman scattering (TERS) have attracted wide interest in life sciences as useful tools to reveal the molecular composition and functional components of biomaterials, which was applied to molecular sensing.^9–17 Surface-enhanced Raman scattering (SERS) utilizes plasmonic resonances of metallic nanostructures to enhance Raman signals of biomolecules.¹⁸ Weak Raman signals of complex biomolecules could be challenging for measurement at the few molecule level,^18,19 but SERS can be used to enhance the Raman signals under the plasmonic resonance conditions.

TERS is the combination of SERS and scanning-probe microscopy, and it provides the ability of label-free nanoscale high-sensitivity detection and chemical imaging.²⁰ TERS has been demonstrated as a powerful tool for the detection of the characteristic functional components and chemical bonds of biomaterials, such as DNA. Conventional DNA sensing methods often involve staining as well as enzyme processing, which modify the sample molecules. Recently, several studies of optical DNA sensing have been reported.^21–24 TERS is a noninvasive and a highly sensitive spectroscopic imaging technique, and it could be a promising way for DNA aptamer sensing.

Here, we report the nanoscale TERS detection and mapping of DNA aptamer molecules for L. monocytogenes immobilized on gold substrates and the corresponding Raman spectral variations. We compare topographic and nano-optical signals and analyze the characteristic Raman peaks of different chemical components. We investigate the limits of the TERS spatial resolution by obtaining TERS maps with 3-nm step size, which is an order of magnitude smaller than the typical size of the scanning tip. Our results show that the TERS mapping provides valuable information about the chemical composition and conformation of aptamers, and paves the way to future single-molecule label-free biosensing.

Thiol-functionalized DNA aptamers for L. monocytogenes (from Aptagen, LLC), that were used in the TERS experiments, consisted of 47-unit DNA oligomers and acted as targets for the InlA protein.²⁵ The sequence of the DNA aptamer is shown in Fig. 1(a). These aptamers have both single and double stranded DNA parts in one molecule. In our experiments, the 3′ end was modified with a terminal thiol group, which works as a constraint for the aptamer motion on the gold surface. This functionalization scheme helped the purification and immobilization of DNA by adsorption of the terminal thiol group to a metallic substrate.² The thiol functionalization of DNA was performed following previously reported protocols.²⁶ 

Figure 1(b) shows the sketch of the conformational structures of DNA aptamers for L. monocytogenes on the gold substrate. Each functionalized molecule consists of a DNA aptamer, and a terminal thiol group [-SH-linker-(CH₂)₆OH]. We used a 1 × 1 cm atomically flat gold substrate due to its suitability for single molecule topographic imaging and optical field gap mode enhancement. The gold substrate was cleaned with piranha solution that included the ratio of 3:1 concentrated sulfuric acid to hydrogen peroxide. One minute exposure was used to remove all the organic residues from the surface of the gold substrate. Then, the gold substrate was washed using deionized water for 1 min to ensure the residues and piranha solution were removed. After that, the gold substrate was air dried. The stock solution of 13 μM L. monocytogenes DNA aptamers was diluted to 100 nM in water. Then, 65 $μl$ of this solution was used to functionalize the gold substrate by drop coating and air drying for 12 h inside a biosafety cabinet. Next, the gold substrate was washed using deionized water three times and air dried for 30 min.

TERS imaging experiments were performed using our previously described microscope (Omega Scope-R, formerly AIST-NT, now Horiba Scientific, coupled with LabRAM Evolution microscope, Horiba Scientific),²⁷ using a gold-coated nanotip in the contact atomic force microscopy (AFM) mode [the sketch of our experimental setup is shown in Fig. 1(c)]. Topographic imaging was performed using the tapping AFM mode with 10 nm amplitude. The laser excitation wavelength was 660 nm. The TERS signals were collected in the back-scattering geometry using a 100× long-distance objective. Gold-coated TERS tips had a radius of ∼10–20 nm. The tip size has been commonly believed to determine the spatial resolution in TERS; however, TERS tips also often have nanoscale surface roughness with features smaller than the tip size. Therefore, higher spatial resolution can also, in principle, be achieved down to the subnanometer scale, as it was recently reported in the cases of the ultrahigh vacuum UHV-TERS of single porphyrin molecules and of the ambient TERS of ss-DNA oligomers.²⁸ TERS spectral acquisition time at each spatial location in the TERS maps was about 5 s.

We obtained the AFM topography of a large area (1 × 1 μm), which contained many isolated aptamers [Fig. 2(a)] and its corresponding TERS map [Fig. 2(b)]. Previous work reported similar AFM images of aptamers.²⁹ The TERS maps were obtained by integrating the full spectral range shown in Figs. 2(d) and 2(f). Figures 2(c) and 2(e) show the magnified views of the two selected smaller areas highlighted by two white squares 1 and 2 in Fig. 2(b), respectively. The scan step size of 10 nm is of the same order of magnitude as the size of the folded aptamer (further discussed below). Therefore, it was not possible to obtain a detailed molecular conformation and sequence from the TERS map with this limited spatial resolution. However, we obtained TERS signals from aptamer molecules, which showed a strong correlation with the topography (Fig. 4). The enlarged views of the selected regions of interest in Figs. 2(c) and 2(e) show several isolated spots with strong Raman signals labeled a–h and a–f, respectively. The corresponding spectra are shown in Figs. 2(d) and 2(f), respectively; and are labeled according to the literature-based band assignments.

Figure 3 shows a comparison of the TERS signals from two closely spaced locations labeled a and b in Fig. 2(c). Both spectra are very similar [Fig. 3(a)], which confirms reproducibility of the TERS signals. As Fig. 3(a) shows, blue and green lines are the original spectra from the locations a and b in Fig. 2(d). Lorentz fitting curves for those spectra are plotted on the same graph using dashed lines. The total combinations of fitted peaks for both spectra are depicted by two red solid lines. The PO₂ stretching vibrations of DNA molecules are often considered to have strong and repeatable Raman signals. Therefore, we performed the normalization of both spectra in Fig. 3(a) by calculating the ratio of each peak height divided by the intensity of the PO₂ stretching signal at 1107 cm⁻¹. Figure 3(b) shows the normalized results based on the intensity of the PO₂ signal in Fig. 3(a). As the corresponding ratio of intensity [Fig. 3(b)] shows, the three positions of the PO₂ stretching vibrational bands have the same intensities for the two spatial locations. The green and orange lines in Fig. 3(b) correspond to the TERS spectra of the locations a and b in Fig. 2(c), respectively. The intensities of the DNA nucleobases A and T decreased in the location a compared to the location b. All the signals for A/G in the two TERS spectra have the same intensities, which indicates that the intensity of G in the spectra of a is higher than in b. The equal intensity of C/G also indicates that the intensity of G in the spectra of b is higher than a. These differences allow for the first steps in the conformational analysis of the DNA aptamers. The differences in the TERS spectra indicate the relative concentration of different nucleobases in the different regions of space. Our data provide a first qualitative picture. More quantitative analysis may be performed by obtaining TERS maps with higher spatial resolution and by using molecular modeling, which is beyond the scope of the proof-of-principle experiments in this work and will be considered in the future.

We observed a strong spatial correlation between the sample height topography measured by AFM [Fig. 4(a)] and the corresponding nanoscale optical signals measured by TERS [Fig. 4(b)]. The inset in Fig. 4(a) shows the width and height of the aptamers obtained from the AFM height profile. The typical nucleobase size is ∼0.7 nm.³⁰ Therefore, the expected spatial dimensions of our 47-unit aptamers are approximately 5–20 nm, based on how the aptamer tertiary structure lays on the gold surface. In Fig. 4(a), the measured AFM diameters of the four aptamers are ∼20 nm and their heights are in the range of 1.5–3 nm, which corresponds well to the expected size of the aptamers, which varies due to the differences in the conformational configurations of the individual molecules, their dynamics and varying substrate environment. The spatial resolution of the AFM and TERS images is limited by the size of the tip, whose diameter varies in the range of 20–40 nm. The TERS mapping of aptamers that were located using AFM in Fig. 4(a) is shown in Fig. 4(b), which strongly correlates with the shape and contours of the four aptamers in the AFM image. Therefore, in the subsequent TERS analysis, we selected the spots with 20–40 nm width and ∼2 nm height.

Finally, we obtained the high spatial resolution AFM and TERS maps of DNA aptamers with a step size of 3 nm for TERS (Fig. 5). Figure 5(a) shows the AFM topography of a DNA aptamer with a width of ∼20 nm and a height of ∼3 nm. These size estimates correspond to the typical sizes of the other measurements described above. The zoomed-in area of the TERS map in Fig. 5(b) with marked pixels is shown in Fig. 5(c). Figures 5(d)–5(g) show the corresponding spectra in rows A–D of Fig. 5(c). The spatial locations in the row below D were not considered due to the strong background, which probably originated from the plasmonic substrate photoluminescence signal at the beginning of the scan or it could also be some impurity, which was initially present on the tip and later got “self-cleaned” during the scan. The surprising result is the strong dependence of the TERS spectra on the spatial location. Figures 5(d)–5(g) show strong variations in the TERS spectra from pixel to pixel with the step size of 3 nm. This gives a strong indication of the 3 nm spatial resolution obtained under the ambient conditions in air at room temperature AFM-based TERS. Such high spatial resolution, in principle, allows for the possibility of mapping the structural conformations of small DNA aptamers. Currently, no other method is capable of doing this under similar conditions. However, it also cannot be excluded that the apparent high spatial resolution of TERS may reflect the complexity of the experiment and the insufficient signal-to-noise ratio. The band assignment is shown in Table I based on the previous literature of Raman spectra of DNA.

Our results provide the first application of TERS spectroscopic imaging to DNA aptamers. We demonstrate the localization of aptamers on a large scale of a few micrometers with a 10 nm step size. We also showed a strong correlation of the TERS signals with AFM topography, and finally we showed a possibility to obtain detailed TERS maps of aptamers with 3 nm spatial resolution. Further experimental work and simulations are needed to perform a more precise analysis of the aptamer conformations. Our work shows feasibility of such studies, which may provide new insights into the structure-functional relations in the performance of aptamer-based medical and nanobiosensing applications.

C.L.G. acknowledges the support of NSF, Grant No. CBET-1511953/1512659, Nano-biosensing. S.H. and H.L. acknowledge the support from the Education Program for Talented Students of Xi'an Jiaotong University. The authors thank Marlan Scully from Texas A&M University for the use of his facilities.