An early challenge in labeling was to find fluorescence dyes suitable for STED microscopy that were excited by the excitation laser and de-excited by the stimulated-emission laser. By now chemists have produced a sufficient number of suitable compounds to consider this problem solved. But a second labeling problem still exists: the total size of the labeling probes.
An antibody has a molecular weight of 150 kDa and a length of 10–15 nm, and the combination of a primary and a secondary antibody is up to 30 nm long. This is not a problem when using the traditional resolution of 300 nm, but with a resolution of 20–40 nm, labeling probes of this size certainly do pose a problem: First, the fluorophores can be up to 30 nm away from their targets and second, because of spatial constraints, the probes will not bind to every target molecule, and the images tend to become spotty (Figure 1).
Right: Fig. 1: STED image showing the artifact typically generated by antibody stainings in super-resolution microscopy. Neurofilaments – known to be continuous structures – in a human neuroblastoma cell line were immunostained with mouse anti-O-glycosylated neurofilament M IgG. Source: Detail from Wildanger et al. (2009).
Felipe Opazo, a postdoctoral fellow at Silvio Rizzoli's group found in cooperation with international collaborators that small single-stranded oligonucleotides, called aptamers (from Latin aptus – fit and Greek meros – part), are well suited for this purpose (Opazo et al. ). These oligonucleotides with lengths of 15–100 nucleotides are produced in vitro – synthesized much like primers for PCR – and assume a specific tertiary structure in solution with binding affinity to other structures (Figure 2). Because of this property, aptamers were and still are investigated as potential drug substances, for instance to inhibit disease-related proteins.
Aptamers are much lighter and smaller than antibodies or antibody fragments (Fab fragments: ca. 50 kDa, length ca. 9 nm), with a molecular weight of 10–15 kDa and a length down to 2–3 nm. This is small enough to pose no problems at resolutions down to 20 nm or less in super-resolution microscopy. And their small size makes it possible to label almost all antigens, even in cases where only few antigens are present. Some synaptic vesicles, for instance, may contain only 10 antigens per vesicle. With a vesicle size of 45 nm, there is no space for 10 antibodies, but 10 aptamers are easily accommodated. This leads to a higher marker density in the organelle (Figure 3).
To find an aptamer specific for a target structure, aptamers from an oligonucleotide library are selected for binding to the target structure. The aptamers found are then amplified by PCR and examined for specificity.
One advantage of aptamers is already apparent: They can be selected and produced quite easily without animal or cell-culture facilities. The development costs are estimated to range between 2,000 and 4,000 €. Once the sequence is established, large milligram amounts of aptamers can be produced in vitro for a few hundred Euros.
Also, aptamers can be selected to bind specific cell types, e.g. specific cancer cells, even without knowing their actual binding partners. This is not possible with antibodies, since animals cannot be immunized with whole cells.
Similar to monoclonal antibodies, aptamers usually only recognize either the fixed or the native protein, depending on which they were selected against.
The main challenge with using aptamers as labeling probes is to ensure specificity. It is easy to find aptamers that bind to the target, but testing for specificity takes time and money.
The aptamers used by Silvio Rizzoli’s group to test the feasibility of using those molecules in STED microscopy were developed and used in collaboration with two US research groups at the Albert Einstein College of Medicine in New York and the University of Texas in Austin, which have decades of experience in using and developing aptamers. They used aptamers for three proteins of the endosome – an intracellular organelle involved in the uptake of substances into the cell (endocytosis) – and compared the acquired images with results obtained with antibodies against the same proteins and with labeled natural ligands.
When using aptamers against proteins of the endosome, the typical, circular endosomal structures with diameters between 100 and 500 nm could be visualized (Figure 4). With monoclonal antibodies directed against the same epitopes, such structures were hard to detect: While the numbers of endosome-like structures found with aptamers against TfnR and epidermal growth factor receptor (EGFR) corresponded closely to the numbers found with the labeled natural ligands transferrin and epidermal growth factor (EGF), the numbers of such structures found with monoclonal antibodies were much lower.
With aptamers, even details such as thin endosomal tubules enriched in transferrin receptor (TfnR) could be seen (Figure 5).
With small labeling probes like aptamers, much smaller structures can be visualized. And because of their smaller size, the probes can also be taken up more easily by living cells, making live imaging possible. Rizzoli and colleagues incubated cultured cells with the aptamers against endosomal proteins and were able to show that they co-localized with the endosomal labels Alexa488-transferrin and rhodamine-EGF.
The high number of epitopes recognized by aptamers resulted in very bright stainings, permitting live time-lapse super-resolution imaging to show endosomal dynamics.
Aptamer stainings were also often brighter than antibody stainings. By comparing the overall staining intensity of the images acquired with aptamer or antibody stainings with the fluorescence intensity of single aptamers or antibodies, Rizzoli and colleagues were able to estimate that the tested aptamers recognized many more epitopes than the antibodies. The EGFR aptamer recognized about twice as many epitopes, while an aptamer against the prostate-specific membrane antigen PSMA recognized 100 times the number of epitopes compared with the PSMA antibody. This corroborates the theory that the smaller size of aptamers enables them to access more epitopes.
Using smaller labeling probes showed the expected benefits regarding accuracy and sensitivity. But just how small should the labels be? The smaller, the better? With aptamers it was possible to easily modify the size of the labeling probes to tackle this question. This was done by coupling additional nucleotides to the structure. Doubling the original PSMA aptamer size of 15 kDa to ca. 31 kDa substantially reduced image quality, and tripling it to ca. 58 kDa led to a further drop in quality. So it seems that only the smallest probes deliver sufficiently accurate images.
Another class of small markers tested by Silvio Rizzoli and other groups are the so-called nanobodies. Nanobodies are single-chain antibodies obtained from camelids (llamas and camels) with a molecular weight of ca. 13 kDa (length ca. 2 to 4 nm), i.e. in the same size range as aptamers. Nanobody production starts like normal antibody production with immunization of an animal (usually a llama) with the antigen. After a few weeks, blood is taken from the animal, B cells are isolated, the specific nanobodies are selected and their sequence is determined. Then, the nanobodies can be produced recombinantly in bacteria.
The development costs for a nanobody are estimated to range between 8,000 and 10,000 €, in the range of the costs of a monoclonal antibody. Once the nanobody sequence is known, they can be produced by bacteria, leaving only the labeling costs. Nanobodies seem to be better suited for fixed samples, while aptamers might be better suited for protein complexes or complicated proteins.
Many more microscopy applications for aptamers are possible: According to Silvio Rizzoli, they could be used for instance as sensors for ATP or other compounds, for different types of FRET experiments, and even their use in electron microscopy, coupled to gold particles, is being tested.
Aptamers are currently being investigated as potential tools for cancer therapy. For instance, aptamers can be selected against specific cancer cells or against receptors expressed specifically on cancer cells. If those aptamers are coupled with a toxin or a radionuclide, they could be injected into the circulation, bind specifically to cancer cells and kill those cells.
The second possible application of aptamers in oncology is their use for diagnostic purposes: Since they deliver a more sensitive staining, this could lead to better cancer cell detection in biopsies and might also permit the use of smaller biopsies.
Silvio Rizzoli is one the very first scientists to have worked with STED microscopes. Asked about his personal highlights of his work with STED microscopy in the past years, he mentions three milestones:
In the first biological application of STED microscopy (Willig et al. ), Rizzoli, Stefan Hell and colleagues from Göttingen used STED to show that synaptotagmin I, a membrane protein from synaptic vesicles, remained clustered after fusion of the vesicles with the synaptic membrane. Because of the small size of synaptic vesicles (ca. 40 nm in diameter) and the dense packing of molecules inside the vesicles, this would not have been possible with the use of conventional fluorescence microscopy.
Second, to obtain the first video of living cells made with STED microscopy (Westphal et al. ), Rizzoli, Hell and colleagues from Göttingen acquired a time-lapse series of STED images in video rate showing the movement of synaptic vesicles in living cells.
And Silvio Rizzoli’s third personal highlight is the publication of the first STED images close to biological reality, acquired by using aptamers as labeling probes (Opazo et al. ), as described above.
Now, with new, small labeling probes optimally suited for the high resolution of super-resolution microscopy, the full potential of STED microscopy for biological applications can finally be realized.
- Opazo F, Levy M, Byrom M, Schäfer C, Geisler C, Groemer TW, Ellington AD and Rizzoli SO: Aptamers as potential tools for super-resolution microscopy. Nature Methods 9: 938–39 (2012); doi:10.1038/nmeth.2179.
- Westphal V, Rizzoli SO, Lauterbach M, Kamin D, Jahn R and Hell SW: Video-rate far-field optical nanoscopy dissects synaptic vesicle movement. Science 320: 246–49 (2008); doi: 10.1126/science.1154228.
- Wildanger D, Medda R, Kastrup I and Hell SW: A compact STED microscope providing 3D nanoscale resolution. Journal of Microscopy 236 (1): 35–43 (2009); doi: 10.1111/j.1365-2818.2009.03188.x.