National Physical Laboratory

Super-resolution fluorescence imaging of cells

Overcoming transparency is a necessary first step in applying optical techniques to the understanding of cells. An early approach to the identification of different cytological regions was to use stains to enhance the absorption or reflection of light. Later it became possible to improve contrast without staining by combining the light passing through the sample with a reference beam and measuring the phase change from the resulting interference (i.e. phase contrast microscopy). A third and more widely used approach is to carry out selective labelling of structures and molecules within cells with fluorescent dyes. Fluorescence affords good levels of discrimination between labelled and unlabelled regions of the cell because it relies on emitted light of longer wavelength than the light used for excitation. Rejection of the latter with dichroic filters allows just the fluorophore-labelled molecules or structures of interest to be imaged. Further improvements in signal-to-noise may be achieved by allowing the excitation light to pass through the sample and away from the detector (see epifluorescence microscopy) or by using total internal reflection to excite fluorophore-labelled molecules in a restricted region of the specimen (within the evanescent field ~100 nanometres).

A second challenge in light microscopy is to improve image resolution, i.e. the ability to distinguish features of the cell that are close together. Clearer images may be generated by using point source illumination (cf. wide-field illumination) and rastering the point source across the sample while imaging in-focus light through a pinhole, i.e. the confocal method. Improvements in resolution may also be achieved by using adaptive optics (spatial light modulators and liquid lenses) to correct for imperfections in the shape of the light beam and/or inhomogeneities in the sample, as well as obtain a narrower focal spot deeper into the sample. Even with these improvements the best achievable resolution of an optical imaging system due to diffraction is around 250 nanometres, which is too great to directly follow the interplay of molecules and structures within cells. New optical imaging methods, however, are being developed which circumvent the diffraction limit and may allow fluorescently-labelled molecules and structures located less than 10-20 nanometres apart to be distinguished. These techniques, which use either the non-linear response of fluorophores to excitation, or stochastically illuminate the labelled molecules so their emission is resolvable with time, are more amenable to widespread use than the electron microscopes used typically by cell biologists and their adoption will greatly aid understanding of important molecular processes underlying cell behaviour, e.g. protein biosysnthesis, endocytosis.


Direct stochastic optical reconstruction microscopy (dSTORM) and photoactived localisation microscopy (PALM)

The image of a single molecule (A) can be analysed and fitted with a 2D gaussian function (B) to determine the actual localisation with nanometer precision (C). This is the principle behind the localisation-based super-resolution approaches, such as dSTORM (direct stochastic optical reconstruction microscopy) and PALM (photoactivated localisation microscopy).

The figure above illustrates one of the central principles behind direct stochastic optical reconstruction microscopy (dSTORM) and photoactived localisation microscopy (PALM), which are methods in development within the Group. A first step is to ensure only a subset of spatially separated molecules are excited. These molecules are imaged with a CCD camera and the pixels fitted with Gaussian functions, from which the exact positions of the molecules may be localised to 10-20 nm accuracy. The fluorescence from a second subset of molecules is then localised in a similar manner and so on and so forth to build up a complete picture with the precise locations of all the molecules in the sample, i.e. a super-resolution image. Working with collaborators, the group is developing and refining these techniques to help realise their potential and transfer the knowledge gained to stakeholders in academia, government and industry.

Selected publications


Last Updated: 15 Oct 2012
Created: 1 Nov 2011