In the past decade or so, great progress has been made in developing new imaging techniques that can overcome the 'diffraction barrier', which limits the resolution in light-microscopy techniques to the micrometer range. Two groups have now independently developed fluorescence-microscopy techniques that allow nanometer resolution.

Stefan Hell and colleagues invented stimulated emission depletion (STED) microscopy more than a decade ago, and subsequently showed that this technique can be used to study biological samples. STED microscopy uses a scanning excitation spot that overlaps with a doughnut-shaped counterpart for the de-excitation of fluorophores by light. Oversaturating the de-excitation reduces the fluorescence spot to sub-diffraction-barrier dimensions, resulting in super-resolved images. This technique has now been tweaked so that a major source of photobleaching — known as molecular triplet-state excitation — of fluorescent markers is eliminated. This has resulted in a ∼30-fold increase in total fluorescence signal and enabled a significant reduction in the focal spot area. So-called triplet relaxation (T-Rex) STED microscopy can achieve a 15–20-nm resolution in the focal plane.

This resolution was demonstrated by various applications of T-Rex STED microscopy in immunofluorescence imaging. Fluorescently labelled transmembrane synaptic vesicle proteins, such as synaptotagmin I and synaptophysin, could be identified as distinct 25–40-nm spots on purified endosomes. And, nuclear 'speckles', which are enriched in pre-mRNA splicing factors, could be separated into distinct particles — this technique might therefore be useful for studying other aspects of nuclear organization. Hell and colleagues have already started to explore the possibility of improving the axial resolution by combining STED with 4Pi confocal microscopy.

Taking a different approach, Eric Betzig, Harald Hess and co-workers developed a method for the isolation of single molecules at high density. The technique, which they called photoactivated localization microscopy (PALM), is based on the idea that the photoactivation of one fluorescent molecule at a time allows for a more precise molecular separation than when all molecules glow at once. Cultured cells that express a photoactivatable fluorescent protein that is attached to the protein of interest can be fixed and imaged as whole cells or processed in thin sections. The sample is photoactivated and subsequently photobleached, and this process is repeated many times until sparse fields of individually resolvable molecules are obtained. The location of each molecule is determined, and a super-resolution PALM image is acquired by adding up the locations of the other molecules in the image.

...open the door to near-molecular-resolution imaging of fluorescent targets...

Using PALM in thin sections, Betzig, Hess and colleagues visualized specific target proteins in lysosomes and mitochondria with 10–20-nm resolution. PALM images of whole fixed cells revealed the partial assembly of a vinculin network at focal adhesion regions, an increased amount of actin at the leading edge of lamellipodia and the heterogeneous distribution of the retroviral protein Gag at the plasma membrane.

T-Rex STED microscopy and PALM provide better resolution than total internal reflection microscopy and confocal imaging, as well as comparable resolution to transmission electron microscopy, without the disadvantage of cryo-preparation. These sophisticated techniques open the door to near-molecular-resolution imaging of fluorescent targets in thin sections and whole fixed cells.