Figure 1.
Schematic depicting optical setup for ratiometric localisation imaging.
A. A single laser at 671 nm is used to provide excitation and the collected light is split into two bands using a dichroic mirror. The bands are imaged side by side on an electron multiplying CCD (EMCCD). An optional cylindrical lens allows astigmatism based 3D localisation. B. A vector field that shows the distribution of lateral chromatic shifts between the channels measured with a bead calibration sample. This vector field was used for chromatic shift compensation during fitting of the single molecule events as detailed in the Methods. The longest arrows shown correspond to a shift magnitude of ∼90 nm. C. Single molecule events are observed as flashes with an intensity component in each channel. When these intensities are plotted against each other, discreet populations emerge corresponding to each fluorochrome in the sample. One such plot, obtained from a sample in which neurons had been transfected with GFP-alpha-sap97 and subsequently labeled with antibodies against GFP (Alexa 647 secondary) and synapsin (Alexa 750 secondary) is shown. Inset: Recorded emission spectra of Alexa 647 (green) and Alexa 750 (red), the black trace is the transmission curve of the dichroic mirror in the splitter device.
Figure 2.
Two-dimensional super-resolution imaging of the distribution of Ryanodine receptors (red) and Caveolin (green), using Alexa 680 and Alexa 750 secondaries, in the periphery of isolated rat cardiac myocytes and overview of dye properties.
Panel A shows the sample at conventional resolution, panel B the super-resolved image. Comparison of enlarged detail (C & D) shows that apparent overlap in the diffraction-limited images is not seen in the corresponding super-resolution image. E. Histogram of mean photon number per event of a dataset of ∼400 ratiometric super-resolution images. The mean photon numbers were calculated for each image in the dataset, the histogram of actual photon numbers per single molecule event are shown in panel F. Scale bars B: 1 µm, D: 200 nm.
Figure 3.
Correlative confocal and super-resolution imaging of a human cardiac tissue section.
The section was ∼10 µm thick and was labeled with phalloidin for f-Actin (Alexa 488), WGA for the cell membrane and extracellular matrix (Alexa 594), along with antibodies for the ryanodine receptor (RyR, Alexa 680) and calsequestrin (CSQ, Alexa 750). In addition to the applied labelling, a strong endogenous fluorescence signal from lipofuscin was recorded. The shorter wavelength labels (Actin, WGA, and RyR) were imaged on a confocal microscope, and the sample then taken to the localisation microscope where super-resolution imaging of the longer wavelength labels (RyR, lipofuscin, CSQ) was performed. Panel A shows an overview of the cellular structure across a large tissue area that is indicated by the actin labeling (largely muscle cell contractile protein). Scale bar 100 µm. Panel B is a projection of a confocal stack taken of the region indicated in A. Scale bar 10 µm. Panel C shows a confocal stack of a small detail area from B and panel D shows an optically sectioned super-resolution stack, within the region covered by the confocal stack in C. Panels E & F compare corresponding confocal (F) and super-resolution (E) images both using the RyR-Alexa 647 signal. Note the good correlation between the data and the improvement in resolution in E. Scale bar 1 µm. G: 3-colour super-resolution image of a small area in the tissue sample, note the improved resolution as compared to the conventional resolution image section. Since the ratios of Alexa 647 and lipofuscein were relatively close some crosstalk did occur. Scale bar 500 nm.
Figure 4.
3D super-resolution imaging of GFP-alpha-SAP97 (antibody labelled with an Alexa 647 secondary, green) and Synapsin (with an Alexa 750 secondary, red) in a primary hippocampal culture.
4D imaging was performed using dual-colour 3D localisation based on astigmatism in conjunction with the ratiometric multi-colour approach. A. Comparison between detail in a super-resolution image (left) with the corresponding conventional diffraction limited image (right) of the two proteins. B & C. 3D rendering of the region indicated in A using super-resolution (B) and conventional resolution (C). Note that in the high resolution data (B) ‘lateral’ as well as ‘axial’ synapses can be distinguished. The 2D images shown on the axes are average projections along the respective directions. Scale bars A: 1 µm, B: 500 nm.