Super-resolution microscopy, which can be used to visualize objects as close together as a few tens of nanometers, has recently been much celebrated, with the 2014 Nobel Prize in Chemistry awarded for its development. Although it is exciting and useful to see a labeled protein in super-resolution, the resulting fluorescence data still lack cellular context. Where, in other words, is the labeled protein relative to the ultrastructure of the cell? Conversely, electron microscopy (EM) can give exquisite structural information, but identifying specific proteins with EM is relatively laborious and typically not quantitative. Correlation of electron micrographs with diffraction-limited fluorescence images, though informative, cannot be performed at the scale of a few tens of nanometers.

Correlated PALM and EM localize proteins within ultrastructure. Credit: Image adapted from Watanabe, S. et al. Nat. Methods 8, 80–84 (2011), Nature Publishing Group.

But super-resolution fluorescence imaging is now approaching the scale of an electron microscope (though there is still at least an order-of-magnitude difference in resolution between the two techniques). In the imaging of fixed samples in particular, correlating electron and super-resolution fluorescence microscopy offers exciting prospects.

Putting these two methods together is not necessarily a simple matter. The sample must be prepared in such a way that it can be imaged at high quality by both methods, with minimal distortion. Images obtained in both modes must be aligned precisely if they are to genuinely provide complementary information. For correlating transmission EM and interferometric photoactivated localization microscopy (PALM) images of cell surface structures, for instance, Taraska, Hess and colleagues recently developed sample preparation and alignment methods based on embedded gold nanorods, achieving correlation at 20-nanometer resolution (Nat. Methods 11, 305–308, 2014). CLEM also requires fluorescent probes compatible with sample preparation for EM. To correlate PALM with cryo-electron tomography of bacterial cells, for example, Jensen and colleagues identified fluorescent proteins that could photoswitch in frozen conditions (Nat. Methods 11, 737–739, 2014). They also needed to find ways to prevent ice-crystal formation due to warming of the sample during PALM.

Methods development for precise super-resolution CLEM is likely to continue and to result in an increased interest in correlative imaging more generally.