Microscopy Bypasses Light’s Limits to Reveal Finer Biological Details

A new technique is pushing the boundaries of super-resolution microscopy by circumventing traditional limitations imposed by saturation and stochastic switching of emitters. Larnii Booth and colleagues at the The University of Queensland in collaboration with RMIT University and University of Otago, demonstrate a Structured Detection Microscope (SDM) capable of achieving deep super-resolution by employing spatial mode demultiplexing. The approach enhances sensitivity to sub-diffraction emitter separations in two-dimensions, achieving resolutions surpassing 40nm, fivefold below the diffraction limit, as evidenced by imaging fluorophores attached to DNA nanorulers with 50nm separations. By avoiding phototoxicity and increasing imaging speed, it sharply advances the study of biological structure, function and dynamics. Conventional optical microscopy is limited by the diffraction of light, restricting resolution to approximately 200nm. This poses a significant challenge when investigating biological structures, many of which are considerably smaller, such as proteins, viruses, and individual molecules. Super-resolution microscopy techniques were developed to overcome this barrier, but many rely on inducing stochastic behaviour in fluorescent molecules or utilising high light intensities, both of which can damage sensitive biological samples and limit imaging duration.

Reshaping light’s point-spread function for nanoscale fluorophore localisation

Structured Detection Microscopy (SDM) carefully reshapes the point-spread function, which is the blurry spot created when imaging a single point of light, emitted from a sample. This function fundamentally limits the ability to distinguish between closely spaced objects. Instead of increasing brightness, which often leads to saturation and photobleaching, SDM intelligently redistributes the information within the light; this is similar to sharpening a blurry image to distinguish closely spaced objects. Spatial mode demultiplexing, a key SDM component, achieves this redistribution by separating emitted light into different spatial modes, distinct patterns of light propagation. These modes are then processed to enhance visibility of fine details, effectively encoding information that would otherwise be lost within the diffraction limit. Images, consisting of 512x 512 pixels, were captured with 500ms exposure times to minimise sample drift when imaging fluorophores attached to DNA nanorulers separated by as little as 50nm. Minimising drift is crucial for accurate localisation of fluorophores at this scale. Bayesian analysis modelled images as random photon arrivals weighted by the microscope’s point-spread function, incorporating background noise. This statistical approach allows for precise estimation of fluorophore positions even in the presence of noise. The technique achieved resolutions surpassing 40nm, a fivefold improvement over the diffraction limit, without saturating or switching emitters. The Bayesian analysis incorporates a prior probability distribution reflecting the expected distribution of fluorophores, further refining the localisation accuracy.

Spatial mode demultiplexing achieves sub-diffraction imaging of nanoscale DNA structures

Currently, resolutions surpass 40nm, representing a fivefold improvement over the 200nm diffraction limit that previously hindered imaging of sub-wavelength biological structures. SDM employs spatial mode demultiplexing to enhance sensitivity to separations below the diffraction limit. The principle relies on the fact that different spatial modes carry unique information about the sample, and by separating and analysing these modes, the effective resolution can be significantly improved. Predictions suggest even greater precision, with potential resolutions dipping below 5nm achievable by optimising illumination intensity and acquisition time, though this requires careful balancing against background noise. Achieving such resolutions would necessitate extremely high signal-to-noise ratios and sophisticated data processing algorithms. The system utilises a total internal reflection fluorescence (TIRF) microscope, manipulating the point-spread function to redistribute information away from areas prone to interference. TIRF microscopy selectively illuminates a thin layer near the sample surface, reducing background noise and enhancing signal from fluorophores in that region. A linear decrease in the Fisher information, a measure of data quality, was observed with decreasing separation using SDM, unlike the quadratic degradation seen in conventional microscopy; this indicates a more stable and precise measurement. The Fisher information quantifies the amount of information that can be extracted from the data, and a slower decrease with decreasing separation signifies improved precision. However, these results were obtained with idealised samples, and translating this performance to complex, densely packed biological environments remains a significant challenge. Biological samples exhibit significantly more complexity than DNA nanorulers, including variations in refractive index, autofluorescence, and uneven fluorophore distribution.

Gentle super-resolution imaging using rearranged light signals avoids sample damage

Super-resolution microscopy is rapidly becoming indispensable for visualising the intricate details of life at the molecular level. Understanding the organisation and dynamics of biomolecules is crucial for deciphering cellular processes and developing new therapies. This new SDM technique avoids a common problem with existing methods, which often force a trade-off between image clarity and damage to delicate biological samples. Techniques like stimulated emission depletion (STED) and single-molecule localisation microscopy (SMLM) often require high light intensities or complex molecular tagging, which can perturb biological processes or even destroy the sample. While this approach holds promise, the authors acknowledge that current validation relies on precisely manufactured DNA nanorulers. Using artificial DNA structures for initial validation is a limitation, as biological samples present far greater complexity. DNA nanorulers provide a well-defined and controlled testbed for evaluating the performance of the SDM technique, but their behaviour may not fully reflect the characteristics of biological systems.

Demonstrating over forty-nanometre resolution, five times beyond the usual diffraction limit, establishes the potential of spatial mode demultiplexing as a viable route to clearer images. This technique avoids damaging samples through intense light or complex molecular tagging, offering a gentler approach to biomolecular observation. The new microscopy technique offers clearer views of biological structures without damaging delicate samples, and will begin to unlock deeper understanding of biomolecular function and dynamics. Potential applications include studying the organisation of proteins within cells, visualising the structure of viruses, and tracking the movement of molecules in real-time. By employing spatial mode demultiplexing, the Structured Detection Microscope (SDM) achieves resolutions exceeding 40nm without inducing phototoxicity or relying on random fluctuations in fluorescent molecules. Successfully imaging fluorophores attached to DNA nanorulers separated by just 50nm validates its potential for biological applications. Future work will focus on applying this technique to more complex biological samples and further optimising the system to achieve even higher resolutions.

The research successfully demonstrated super-resolution microscopy achieving resolutions exceeding 40nm, five times beyond the diffraction limit. This was achieved through spatial mode demultiplexing, a technique that enhances image clarity without relying on intense light or complex molecular tagging which can damage biological samples. By avoiding these potentially harmful methods, the Structured Detection Microscope (SDM) offers a gentler approach to visualising biomolecular structures. The authors intend to apply this technique to more complex biological samples and further optimise the system for even higher resolutions.