Single-molecule techniques can extract information about individual molecules from complex and heterogeneous mixtures, which are usually obscured in ensemble averaging measurements. Currently, single-molecule optical characterization methods rely on ultra-dilution (concentrations below nM) and/or molecular immobilization to address the challenge of reducing the diffraction-limited detection volume. However, these strategies may not be applicable to certain biological systems, where the biomolecules are engaged in crucial interactions with surrounding molecules and environments at relatively high concentrations (μM-mM for physiological protein assembly and enzyme activity). For example, as an intrinsically disordered protein (IDP) associated with type II diabetes, human Islet Amyloid Polypeptide (amylin, hIAPP) exhibits structural flexibility and heterogeneity. It undergoes dynamic structural conversions and aggregation/dissociation, depending on environmental conditions such as concentration and pH. Hence, we are developing new approaches that facilitate single-molecule characterizations while maintaining the influences of surrounding environments, in order to understand the behaviors and mechanisms of these molecular systems in solution.

In this study, we utilized surface-enhanced Raman spectroscopy (SERS) due to its unique nanometer-scale properties. As a surface-sensitive technique, SERS probes the vibrational signals of molecules present near the plasmonic nanostructures with electromagnetic field enhancements, known as “hotspots”. Since this SERS-active effect is spatially confined and rapidly diminishes in nanometer ranges, it can overcome the optical diffraction limit and reduce the detection volume to study individual molecules. However, creating a well-defined SERS-detection volume with precise spatial control for reliable and efficient molecular characterizations in solution is a challenge.

To tackle this challenge, our team developed a novel platform that combines optical manipulations with SERS characterizations. This gives us the ability to actively control substrates and construct a dynamic plasmonic nanocavity in a solution. Our previous study demonstrates the formation and adjustment of a SERS-active hotspot between two Ag nanoparticle-coated silica microbeads through the manipulation of optical tweezers. In this study, to achieve a more precise and stable position control, we designed a new approach for exerting plasmonic trapping effect to create a dynamic hotspot among three Ag nanoparticles at a confined region, which further reduced detection volume, boosted SERS enhancement, and increased detection turnover efficiency all at once.

We constructed a plasmonic junction between two Ag nanoparticle-coated silica microbeads to trap a freely diffusing Ag nanoparticle under the focused laser light, forming a highly confined hotspot with enhanced SERS activity. It is worth noting that we coated micrometer-sized silica beads with Ag nanoparticles and assembled them into dimers using DNA linkers, making it easier to observe and locate the nanoscale inter-Ag nanoparticle junction under regular microscopes. Moreover, the Ag nanoparticle-coated silica microbead dimer is more stable than the conventional Ag nanoparticle assembles in solutions, which can improve the efficiency and reproducibility for continuous Ag nanoparticle manipulations and simultaneous SERS measurements.

One practical challenge we encountered was finding the ideal laser power settings to ensure effective trapping and sufficient SERS enhancement, while minimizing the photothermal effect. In our setup, a 1064 nm laser beam and a 532 nm laser beam were collinearly focused at the junction of the Ag nanoparticle-coated silica microbead dimer to excite surface plasmons for nanoscale trapping and SERS measurements. We designed the junction to be 20 nm in size by engineering DNA linkers and synthesized Ag nanoparticles in 70 nm by adjusting experimental conditions. Such that a freely diffusing Ag nanoparticle could be attracted towards the center of the plasmonic junction and wedged against the Ag nanoparticles coated at the junction to form a trimer-like nanocavity with the highly confined detection volume and the enhanced local field to provide single-molecule sensitivity. More importantly, we conducted both theoretical simulations and experimental trials to determine the optimal power level of the 532 nm and the 1064 nm lasers at 2.7 mW and 8.9 mW, respectively. Figure 1 shows that the calculated trapping force and potential at the equilibrium position can suppress Brownian motion to achieve stable trapping, which is essential for reproducible SERS measurements at the nanocavity. Besides generating surface plasmons, the focused 1064 nm laser light might also contribute to recruiting the freely diffusing Ag nanoparticle to the vicinity of the nanoscale plasmonic junction with high trapping efficiency. Hence, by switching the focused laser light on and off, we could control the plasmonic trapping process to form this dynamic nanocavity and facilitate high-throughput analysis.

Another challenge was the data analysis and interpretation on the structural heterogeneity of hIAPP under different physiological pH conditions. To address this, we followed the constructive suggestions from the reviewers and combined experimental and computational approaches. Our platform has the ability to detect single molecules, distinguish low-populated species in mixtures, and track structural changes in response to real-time pH transitions. To better characterize the structures of various hIAPP species, we conducted parallel SERS measurements with a large sample size of 10,000 spectra at each time point during the incubation of hIAPP at pH 5.5 (pancreatic β-cells) and 7.4 (extracellular compartments), respectively. Our results show that hIAPP maintained ensemble helix-coil structures throughout the 2-hour incubation period, regardless of the pH conditions. However, two types of low-populated transient species of hIAPP containing either turn or β-sheet structure among its predominant helix-coil monomers are characterized during the early-stage incubation at the neutral environment, which is supported and elaborated by Molecular Dynamic (MD) simulations with possible structural details. Such a slight shift in the equilibrium between different hIAPP species could drive the irreversible amyloid developments even after the post-adjustment of pH from 7.4 to 5.5. Thus, the direct structural characterization of these hIAPP transient species among the heterogeneous mixture reveals the effect of pH on their intra- and inter- molecular interactions and provides novel insight to reveal the mechanism on pH-regulated amyloid aggregation for better treatments of type 2 diabetes in the future.

In this work, we offer an easy-to-use strategy to utilize optical plasmonic trapping and control a dynamic SERS-active nanocavity, which reduces detection volume, boosts SERS enhancement, and enlarges high-throughput sampling capacity for efficient single-molecule SERS characterizations. Our results highlight the technical effectiveness of integrating active substrate control and passive spectroscopic detections in optical tweezers-facilitated SERS platforms to inspire new single-molecule biophysical method developments. By combining experiments and simulations, the statistical analysis of the single-molecule level structural details reconstructs its bulk properties and brings the unique information on population and probability of specific type of molecules in heterogeneous mixtures. It holds great potential to remove ensemble averaging to address the challenge of characterizing individual molecules in physiological milieus and unveil the molecular mechanisms of amyloid aggregations involving in complex biological processes associated with neurodegenerative diseases.

Link to the full article: https://doi.org/10.1038/s41467-023-42812-3