SFN: Super-Resolution Synaptic Imaging minisymposium

This was my first event of the conference, so I hope, dear readers, you will keep in mind that I was still getting my science head screwed back on as I took all this down... The session was chaired by U.V. Nägerl from the University of Bordeaux and T.A. Blanpied of U. Maryland School of Medicine. They introduced the session by outlining the critical importance of understanding the structure and function of synapses and pointing out the technological gap that still exists in our ability to investigate these nanoscopic structures. However, we are embarking on a new era in synaptic imaging with the advent of a number of new techniques that bend or break the traditional diffusion limit to image structures in the range of 10s to 100s of nanometers, making synaptic imaging a reality.

The first speaker was D.A. DiGregorio, who admitted that his lab was still working within the boundaries of the diffusion limit, mainly, he joked, to find out just how bad it is for synaptic imaging. He is focused on the giant calyx of Held compound synapse as a model in order to investigate the relationship between the mutual developmental plasticity of morphology and physiology. Specifically, his lab has produced evidence that the increased precision of firing at this synapse with age may be related to refinement of spatial coupling of vesicles and presynaptic calcium channels. His technique involves a refinement of confocal imaging, using "spot imaging" to enable sub-millisecond temporal resolution rather than the slower method of line-scanning.

I'm afraid I missed some crucial parts of the next talk, but here are a few of the main points I caught: A. Triller discussed the role of syntaxin in the docking and fusion of vesicles. His group have labeled syntaxin with pHGFP, a pH sensitive fluorescent marker which localizes to the cell surface. FRAP imaging shows the molecule to diffuse rapidly within the plasma membrane, and the lab uses this diffusion to model sub-membrane events using quantum dot tracking. One principal observation utilizing this method was that syntaxin molecules pause in their free diffusion temporarily when they pass presynaptic sites.

U.V. Nägerl then spoke about imaging actin dynamics inside spines using STED. He eulogized spines as amazing signaling machines with thousands of proteins, key for brain function in health and disease, numerous, dense, dynamic, and SMALL, then hailed the new methods arising to allow the application of physiology and biochemistry to individual synapses and spines. 2-photon imaging has been extensively used to explore the structural changes accompanying synaptic plasticity, but differentiation of distinct synaptic compartments remained a challenge, as these compartments are smaller than diffraction limited resolution. This limit, however, is being broken by super-resolution imaging down to 10-50 nm. The group utilizes STED microscopy, which involves suppressing the fluorescence evoked by a primary laser beam using a secondary annular (or donut-shaped) illumination laser of a different wavelength. This produces a central spot of fluorescence much smaller than the diffraction limit.

Nägerl has been working to use STED to image spines in live cells, and has produced preliminary observations of nicely symmetrical distributions of spine neck diameters (which confocal tended to skew towards the larger diameters it could resolve). The lab has begun to investigate synaptic actin and tubulin to discover mechanisms of "morpho-functional" plasticity. Using Lifeact, a small peptide which reversibly binds to actin within synapses, the group has visualized actin cables inside spines and dendrites, some linear and some curved, presumably "handshaking" with microtubules in the dendritic shaft. In addition, different shapes of actin are observed in the spine head, whose shape changes with plasticity along a ~100nm range. Some evidence has been uncovered of changes in spine neck diameter with chemically induced LTP, which the lab intends to follow up with more refined methods. The group hopes to develop nanoscale imaging of dynamic signals such as Ca2+, to explore the possibility of nanoscale photomanipulation / uncaging, and to combine STED bulk imaging with PALM single-particle tracking methods.

The final talk I caught from this session was by E. Jorgensen of the University of Utah, who described his groups progress developing conjugate fluorescence and electron microscopy - a feat which has stymied many great scientists for decades. His group studies synaptic transmission at the 500nm diameter NMJ of the nematode worm C. elegans. Hoping to understand the arrangement of synaptic proteins, one might wish to tag specific molecules with fluorophores, but at that scale one would only see an unresolvable blob. The resolution (so to speak) of this problem involves 4 microscopy techniques: correlative fluor-EM, STED, PALM, and 'biplane'. First, correlative fluor-EM requires a delicate balance between the conflicting optimizations required for fluorescence and electron microscopy. Sections must be mounted on coverglass and not the grids required for TEM, and both florescence and EM-scale ultrastructure must be preserved. The single nanometer scale of TEM can be sacrificed for the ~5 nm resolution of back-scatter SEM, which permits the use of flat-mounted structures, but the tension between fluorescence and ultrastructure remains. EM works best with dehydrated tissue infiltrated with oxidizing agents such as osmium at an acidic pH, whereas fluorescence requires hydrated tissue, a neutral pH, and is severely disrupted by oxidization. Remarkably, Jorgensen claims to have found a compromise which satisfies both techniques, involving embedding tissue for EM in methacrylate plastics which tolerate moderate (5%) hydration, use of K2MnO4 (potassium permanganate) instead of osmium, and tight pH control using ethanolamine.

However, correlating EM ultrastructure with regular fluorescence imaging does not resolve the problem of how to image proteins at a subcellular scale (you are still correlating a coarse blob with the fine structure), though it may be very useful for identifying labeled cells and axons. Instead, the group has used STED on 100 nm sections in C. elegans to image the subcellular distributions of histones, mitochondrial protein TOM-20, and liprin. The trouble with the STED system, Jorgensen remarked, are that 1) it can only resolve down to 60 nm, 2) it's expensive, and 3) he doesn't have one. The group also successfully used PALM (comparable to STORM, but using photoconvertable markers rather than organic dyes) with a liprin-Eos conjugate which is photoconvertible from green to red by application of a little UV light. Preliminary results using this technique to explore the localization of Actin-4 in the worm nose have suggested that the molecule is to be found in the membrane of glial sheath cells, but not in the sensory cells, as some had presumed. Still, all this conjugation proved very time consuming, which led to the development of biplane microscopy by the lab, which, as the name implies, improves on the efficiency of PALM by imaging in 2 planes at once. The system is largely computerized - as Jorgensen proclaimed, "If you still have an eyepiece, you probably still have a rotary phone - give 'em up!" This system is now commercially available as the Vutura Avalanche (and is present in the vendor stands at SFN!)

All in all, these super-resolution techniques seem to be opening up a bright new future for imaging the tiny world of the synapse!