With a growing interest in biology and the composition of living biological entities as well as a good understanding about the fact that biological entities were composed of extremely small complexes, it was essential to come up with an instrument which would help in viewing objects that could not be seen with an unaided eye. The earliest development of the microscope can be traced back to the use of a magnifying glass even though it wasn’t until the 16th century when the earliest simple microscope was developed by inverting a telescope and this was further modified and improved in the 17th century. Ever since then, new techniques have been developed in order to gain a better understanding of biological entities and presently, the world has reached an era of ‘super-resolution microscopy’ which helps I surpassing ‘Abbe’s resolution limit’. These techniques have helped in imaging nanoscopic molecules that play an essential role in different biological processes and has improved the understanding of the structural and functional properties of subcellular components. Although these techniques have been developed to provide a wide range of properties like 3-dimensional imaging and live imaging, each of them still has its advantages and pitfalls and this essay discussed a few of these techniques in detail.
Introduction
The history of discoveries in cell biology and its related fields is mirrored with the advancements made with the microscope over the past five centuries. Although the simplest microscope was first known to be made and used by Robert Hooke, It was Antonie van Leeuwenhoek who earned the title of “Father of the Microscope” for building the first microscope in 1674, and pioneering discoveries concerning bacterial cells and erythrocytes. The nineteenth century was marked with improvements in microscopes and staining methods, which further led to scientists establishing the cell theory and viewing the key cell components, understanding cell division and differentiation and the discovery of mitochondria. With the breakthrough of research in the biological field, Ernst Abbe, a mathematician formulated the “Abbe Sine Condition” which enabled calculations that allowed the maximum resolution in microscopes possible . However, this also meant that that the resolution of optical microscopes was limited by diffraction, which would reach a peak and limit the ability of seeing molecules closely located to one another.
Cell biology revolutionised in the mid-twentieth century with the advances in fluorescent-labelling techniques, which proved to be important tools in biological research, and advances in microscope design and technology. Since then, more specifically in the past decade, there has been an outbreak in the practical implementation of microscopic techniques, with the emergence of super-resolution microscopy that can overcome Abbe’s limit of resolution , hence converting fluorescence microscopy into an effective 3D visualization tool . This enables scientists to view single nanoscopic molecules of 10-20nm, not only in all three dimensions, but also trace these molecules in cellular processes. These techniques, as seen in fig.1, follow one of the two approaches; the first is based on spatial patterning of excited light (illumination-based) and this is used in stimulated emission depletion (STED) microscopy and structured illumination microscopy (SIM). The other approach is based on the localization of single molecules (probe-based) and this is used in (fluorescence) photoactivation localization microscopy/Stochastic optical reconstruction microscopy [(f) PALM/STORM] .
STED
SSIM
TIRFM
PALM/STORM
Best resolution (nm)
~20(lateral)
~50 (spatial)
~50 (lateral)
~90 (spatial)
~230(lateral) ~100 (spatial)
~ 20-30 (lateral)
~60-70 (spatial)
Principle approach
Patterning of excited light using two laser beams
multiple interfering light beams to form moiré patterns
Evanescent field produced by total internal reflection of light
single-molecule localization of photoswitchable fluorophores
Limitations
Photobleaching can occur due to limited light wavelengths
Technical faults
High-intensity pulsed lasers can cause damage to the sample
Photobleaching can occur
Very sensitive to even the smallest changes in sample position
For 3D SSIM, a large number of images need to be taken per wavelength; this takes a long time.
Only 1 plane can be imaged
3D imaging is not possible unless TIRFM is combines with another technique
Require large number of raw images
Data acquisition speed is very slow, bearing a direct effect on imaging live samples
Live imaging
TIRFM-SSIM
3D imaging
TIRFM-SSIM
Multicolour imaging
Table 1: Comparison of the distinct features of the different super-resolution techniques.
Two different approaches to breaking the diffraction limit. A. STED microscopy uses two different lasers- an excitation laser (left) and a doughnut-shaped STED laser (middle- this laser deactivates the fluorophores molecules). Using these 2 lasers, the effective excitation area is limited to a small central zone (right). B. Single molecule localization microscopy methods such as PALM and STORM use photoactivatable fluorophores which can switch between their excited state and ground state to successively image the localization of a small number of molecules at a time at high precision by finding the molecule’s centroid. The many ‘raw’ images are then reconstructed to generate the final super-resolution image.
The emergence of super-resolution microscopy has opened many doors in the field of modern biology and medicine, giving an insight on processes that were unable to be followed using conventional microscopy. It is important to understand that every protein found in living cells has a specific function and is a part of a much larger molecular network. In order to understand the functioning of these large networks, it is important to track the movement and interactions of these proteins within the cell . Super-resolution microscopy, aids in visualising the 3D-structure and accurate location of single protein molecules on distinct organelles and on structures like lysosomes and microtubules, helping in understanding protein interactions and providing a better understanding of the molecular-scale architecture of cells .
Three dimensional STORM image of the mitochondria network in a mammalian BS-C-1 cell. The z-position is colour-coded according to the colour scale bar.
In the past decade, super-resolution microscopy has been used to map the 3D-organization of distinct components of the nuclear pore complex; the polygonal network that makes up the endoplasmic reticulum in cells was imaged, as seen in fig.3, in living PtK2-cells of the kidney; the movement of synaptic vesicles have been traced inside living neurons by tagging the vesicle protein synaptotagmin with antibodies and these techniques have been used to study the co-localization of two mitochondrial proteins by labelling them with different fluorophores. These studies would not have been possible without nanoscale-resolution provided by these techniques since all these structures are extremely small in size .
Super-resolution imaging of the endoplasmic reticulum in living PtK2-cells of the kidney cell. (A) Shows the confocal image and (B) shows the simultaneously recorded STED (x, y) images from the ER marked by the fluorescent protein Citrine targeted to the ER. The arrows point out rings formed by the tubular network of the ER, which are clearly visible only in the STED image (B).
The emergence of super-resolution microscopy has put light on important details of cell biology, holding great importance for research in the future. This essay discussed the different techniques of super-resolution microscopy, its application in cell biology, and its limitations as an instrument.
The first super-resolution microscopy technique, STED microscopy’s concept was introduced a decade ago and has advanced within the past few years. It is based on patterning the excited light in such a way that the volume of light in the excited-state is extremely small, hence maintaining the amount of light that emits fluorescence to small volumes .
This is achieved by using two pulsed laser beams of different wavelengths; the wavelength of light from the first laser beam excites the fluorescent marker and the second laser beam illuminates the sample with a doughnut-shaped beam (called the STED-beam) as seen in fig.4. The wavelength of light from the STED-beam is such that it causes the excited fluorescent molecules to de-excite, bringing them back to the ground-state via stimulated emission. The doughnut-shaped beam from the second laser ensures that the molecules of the centre-most part of the labelled sample are in the excited state, and fluorescence is detectable .
Schematic diagram showing the use of the excitation and deexcitation (STED) beams for 3D-STED imaging inside a living cell. (A) An objective lens focuses the excitation (blue) beam and deexcitation (orange) beam into the ER while also collecting the resulting beam from the fluorescence photons. (B) xy-axes imaging: excitation spot (blue) and doughnut-shaped focal spot (orange) for stimulated emission (C) xz-axes imaging: excitation spot (blue) and STED spot composition consisting of a spot featuring a maximum above and below the focal plane along the z- axis, referred to as STEDz, and an enlarged doughnut-shaped beam called STEDr.
The lateral resolution of STED microscopy has been pushed to below 20nm and has been successful in imaging the synaptic vesicle movement in live neurons after neurotransmitter release during an impulse. In the past, synaptic vesicle exocytosis was suggested and confirmed by using electron microscopy, where ‘pockets’ in the pre-synaptic membrane terminals of chemically-fixed nerve cells were seen, hinting on exocytosis as the process of neurotransmitter release . Further, fluorescence microscopy was used to study the vesicular movement after neurotransmitter release by using FM-dyes . Even though it was known that vesicles are recycled via endocytosis, the fate of its components after fusion with the membrane was still unclear since the vesicles were too small to be resolved by available microscopes.
To solve this problem, STED microscopy was used to determine the entire process of vesicle endocytosis. Monoclonal antibodies against the intravesicular membrane protein synaptotagmin was used for imaging purposes; these antibodies only bound to those protein molecules that were exposed during vesicle exocytosis and were internalised when the vesicle was endocytosed. Fluorescent-labelled secondary antibodies were attached after membrane fixation and permeabilization and were used for visualisation of these vesicles. Images showed synaptotagmin molecules clustered on the pre-synaptic membrane, suggesting that vesicle components remain together on the pre-synaptic membrane during recycling by endocytosis. Each synaptic vesicle is 40-50nm in size and they usually occur in groups of 100-300 vesicles. Therefore, fig.5 shows that using STED microscopy was essential for localising individual vesicles, and contrary to previous beliefs that vesicles hardly move, STED revolutionised the understanding of vesicle-recycling by showing that vesicles constantly move rapidly and randomly .
Comparison of confocal (left) and STED (right) counterpart images of a small region of a neuron terminal labelled with an anti-synaptotagmin antibody, ?xed, permeabilized and visualized using Atto532-labelled secondary antibodies. The STED image reveals a marked increase in resolution and also shows the accurate location of individual vesicle components on the neuron membrane.
However, STED microscopy is limited by wavelength. The absence of sufficient tuneable pulsed light sources in the visible range of light which de-excite the already excited fluorescent-labelled molecules has limited STED microscopy to only a small fraction of fluorophores, which causes bleaching and phototoxicity . STED also requires the use of high intensity pulsed lasers which can cause significant damage to the samples. Furthermore, there are technical limitations set by the laser power required for this technique and the very often, mechanical drift of the optical instruments causes imperfection of the doughnut-shaped beam around the sample, limiting the spatial resolution.
Another example of an illumination-based technique, SSIM follows the approach of illuminating the sample with multiple interfering light beams in order to break the resolution barrier . When multiple beams of mutually coherent light are allowed to interfere, they form a structured pattern, like that of Moiré fringes seen in fig.6. When focussed on the labelled sample, the illumination pattern further interacts with the fluorescent probes. The emitted light contains image details of higher resolution, including details that cannot be resolved using a normal microscope. The illumination patterns are modulated by changing the orientation of light on the sample and high-resolution images are captured within the illumination from different patterns.
The approach of resolution enhancement followed by structured illumination. (a) If two line patterns are superposed in each other, moiré fringes will be formed as a product (seen here as the apparent vertical stripes in the overlap region). (b) A conventional microscope is limited by diffraction to a circular ‘observable region’ of reciprocal space. (c) A sinusoidally striped illumination pattern-the possible positions of the two side components (light beams) are limited by the same circle that defines the observable region (dashed). If the sample is illuminated with structured light, moiré fringes which represent information that has changed position in reciprocal space will appear. The observable region will contain normal information and moved information that originates in two offset regions (d). From a series of images with different orientation and pattern phase, it is possible to recover information from a region that is twice the size of the normally observable region can be obtained, corresponding to twice the normal resolution (e).
The images are collected and reconstructed using computer software which extracts the details from the moiré images, reconstructing them into 3-dimensional images with doubled resolution. The original 2D-SIM involved using two beams of light which interacted with the sample probe to increase its resolution and form 2D images. However, this technology was extended by using 3 light beams, generating resolved images with finer details of the sample in the axial and lateral directions, resulting in a 3-dimensional image of the sample.
Using 3D-SIM in comparison with conventional wide-field epifluorescence-microscopy, experiments to better the understanding of higher order chromatin and to study the accurate localizations of other nuclear components like the nuclear pore complexes (NPCs) and nuclear lamina were performed. The chromatin of formaldehyde-preserved mouse C2C12-myoblast cells were stained with 4',6-diamidino-2-phenylindole (DAPI) and they were observed using 3D-SSIM. The Images obtained from this technique showed a large number of ‘holes’ within the area of the stained chromatin as in fig.7, a feature that could not be observed in the images obtained by wide-field epifluorescence-microscopy.
Comparison of 3D-images obtained from conventional wide-field microscopy (left) and 3D-SIM (right) used in order to resolve interphase chromatin structure of the same DAPI-stained C2C12 cell nucleus. Deconvolution was applied to the wide-field data set (middle). (A) Mid cross-section shows brightly stained clusters of centromeric heterochromatin. Inset shows higher-detail information of chromatin substructures when recorded with 3D-SIM. Arrow in 3D-SIM inset points to a less-bright chromatin structure that has been spuriously eroded by the deconvolution procedure. (B) Apical sections (corresponding to a thickness of 0.5 µm) taken from the surface of the nuclear envelope closest to the coverslip. The raw image shows diffuse DAPI-staining, the deconvolved image shows more pronounced variations in fluorescence intensities and the image obtained from 3D-SIM shows extended resolution and reveals a punctuated pattern of regions without DAPI-staining.
Further taking advantage of SSIM’s multicolour and 3D-imaging properties, the same cells were immunostained with antibodies specific to the nuclear pore complexes (NPC), which detect the NPC proteins; and antibodies against lamin-B, a major component of the nuclear lamina (intermediate filament protein). Hence, the 3D-SSIM images showed the chromatin on the nucleoplasmic side, followed by nuclear lamina and then the nuclear pore complexes on the cytoplasmic side forming a triple-layered organization as in fig.8. Not only was the heterochromatin distinguished from the euchromatin, but at every ‘hole’ where DAPI-labelled chromatin was absent, some amount of NPC-staining was present, suggesting that chromatin was absent within close proximity of the NPCs. Even though all three sub-nuclear structures were obtained using conventional fluorescence microscopy, the spatial organization of these structures was obtained only by using 3D-SSIM.
multicolour imaging of DNA, nuclear lamina, and NPC structures in C2C12 cells by 3D-SIM. The cells are immunostained with antibodies against lamin B (green) and antibodies that recognize different NPC epitopes (red). DNA is counterstained with DAPI (blue). The image on the top left shows the same sample imaged using confocal laser scanning microscopy (CLSM) and the image on the top right shows the images obtained using 3D-SSIM which are better resolved and more clearer. The bottom picture clearly shows the triple layered organization of the three structures.
Therefore, 3D-SSIM has proved to be essential in understanding the spatial organization and interactions of sub-cellular structures that were unable to be studied before. Even though some of the initial limitations of SSIM like the time required to reconstruct and analyse the images have been overcome, SSIM is still restricted by the photostability of the fluorophores used since photobleaching leads to a less intensive image.
Even though total internal reflection fluorescent microscopy (TIRFM) was first used in 1981, it’s still a very important technique and has been used extensively since it allows selective excitation of labelled molecules in a cellular/aqueous environment which are near the surface only. This is not only beneficial because of its ability to view labelled molecules, but also because the region of interest is thin enough to obtain the highest frame-rates. TIRFM combined with structured illumination microscopy (SIM) has developed into a super-resolution technique which can break the resolution barrier and improve resolution of the region of interest.
The conventional-TIRFM is based on the diffraction properties of a light beam when incident onto a surface separating two media with different (high and low) refractive indexes. At a high incident angle (greater that the critical-angle), all the incident light is ‘totally reflected’ as long as it is coming from the medium with a high refractive index through the medium with a low refractive index . At this surface, an ‘evanescent field’ is produced. This field is considered to be an electromagnetic field capable of exciting fluorophore molecules present on the surface. This field rises from the surface into the medium of lower refractive index . The depth of fluorophore excitation is minimised in this phenomenon because as the evanescent field rises parallel to the surface and the distance between the field and surface increases, its strength decreases exponentially, limiting the fluorescent region. TIRFM not only provides a very thin, sectioned layer of excited fluorophores which helps in minimising the background noise caused by water molecules, it also omits unwanted fluorescence of molecules that are out of focus. However, the major drawback of this technique is that only one plane (z-plane) can be imaged, limiting its use to study cell surface events. Therefore, to obtain limit-breaking resolutions and to view multiple planes of the sample region, TIRFM is used in combination with SIM.
The TIRFM-based SIM was used to image EGFP-labelled a-tubulin of living S2-cells of Drosophila. a-tubulin is a protein present in microtubules. Comparing the images of the same sample region obtained by using conventional-TIRFM and TIRFM-SIM, the latter showed a significant improvement in the resolution of the image as seen in fig.9(a,b).
Comparison of conventional TIRF (a) and TIRF-SIM (b) images of the microtubule cytoskeleton in a single S2 cell. The image obtained after combining TIRF and SIM shows better resolution hence giving a clearer image.
Live imaging using TIRFM-SIM was applied to image polymerisation-depolymerisation of microtubules located near the centrosome of a Drosophila S2-cell which was in its mitotic state. Since the length of microtubules was constantly changing due to its polymerisation and depolymerisation, kymographs were used to process images and to determine the spatial-position of the microtubules over time by determining the difference in GFP-labelling density along the microtubule length at different times. Combing SIM with TIRFM helped in imaging the GFP-labelled a-tubulin with enhanced clarity and allowed accurate localization of the end of the microtubule, hence being able to follow it through the process. The images obtained from live-SSIM showed distinct transformation between the microtubule’s polymerisation state, depolymerisation state and its steady state, hence being able to track the dynamics of the microtubules (fig.10), a phenomenon which was not possible to understand properly using conventional-TIRFM.
TIRF-SIM images at different time frames of EGFP-a-tubulin in a S2 cell. (a) 95th image from a 180-frame sequence. Each frame was acquired in 270ms. (b) The green-boxed area of (a) shown at selected times as indicated on the individual images, using conventional TIRF (left) and TIRF-SIM (right). Green arrows follow the end of a single microtubule, which can be seen elongating until approximately the 100 s time point, and then rapidly shrinking. These changes are much easier to follow in the TIRFM-SIM images which are much clearer compared to the TIRFM images obtained.
In contrast to STED and SIM-microscopy (based on the spatial patterning of excited light), STORM and photoactivated localization microscopy (PALM) are probe-based methods principled on single-molecule localization and were developed recently in 2006. These techniques combine 3D and multicolour-imaging and obtain images with a spatial-resolution of 20-30nm and an axial and lateral-resolution of 60nm and 70nm respectively . Keeping in mind that single molecule localization is made difficult in fluorescently-labelled biological samples because it contains millions of fluorophore molecules in a large density , PALM/STORM use photoswitchable probes which can be switched between its visible (fluorescent, excited) and invisible (nonfluorescent, de-excited) state by using light of different wavelengths. Therefore, this approach consists of repeated cycles of sample imaging. In each cycle, different fluorophore-molecules within a diffraction-limited region are excited, such that each excited molecule can be individually imaged without overlapping (due to the images of closely located fluorophore molecules which are invisible in this cycle) and subsequently deactivated to the ground-state as seen in fig.11. In following cycles, a stochastically different set of fluorophore-molecules are excited, determining the accurate coordinates of different molecules in each image. Using these individual images, an overall image is constructed and the position of each molecule in the sample is determined. The PALM/STORM techniques and based on the same concept of single-molecule localization, the only difference being the fluorescent probes that each of them uses. While PALM originally used photoactivable fluorescent proteins that are attached to sub-cellular structures, STORM used synthetic photoswitchable cyanin dyes that carried out the same function.
Schematic diagram showing the basic principle followed by STORM imaging. (a) Shows the microtubules within a cell. (b) shows a distinct set of fluorophore molecules in its excited state. (c), (d) and (e) show different set of fluorophore molecules that are excited while the other closely situated molecules are in the ground-state by their photoswitchable property. (f) shows the complete reconstructed image formed by compiling all the raw images into one image.
Further, STORM is developed to provide multi-coloured imaging by using combinatorial pairs of “reporter” dyes which cause the fluorescence and “activator” dyes which can reactivate the ‘switched-off’ reporter dyes when placed in close proximity to the reporter. Thus, each pair has a different colour of emitted light, determined by the reporter dye and a different colour light that activates the reporter, determined by the activator dye . This technique, therefore, allows the study of molecular interactions between different sub-cellular structures by co-localizing them within a cell.
Comparison between images of microtubules in a mammalian cell obtained from conventional microscopy and 3D-STORM (A) Conventional immunofluorescence image of microtubules in an area of a BS-C-1 cell. (B) The 3D-STORM image of the same area with the z-position of the microtubules colour-coded according to the colored scale bar. (C-E) Show the x-y, x-z and y-z cross-sections of a small region of the BS-C-1cell outlined by the white box in (B), showing 5 microtubule filaments.
To understand the interaction and spatial relation between mitochondria and microtubules within a cell, a two-colour 3D-STORM was performed which proved to be fundamental towards the understanding of mitochondrial-microtubule interactions. It is certain that mitochondria are the “power houses” of a cell and hence, to maintain its dynamic morphology , these organelles are constantly moving about a cell with help from motor-proteins which attach particularly to microtubules within a cell. For this experiment, fixed monkey kidney BSC-1-cells were used and two different sets of reporter-activator dyes were used to stain Tom20, part of the translocase outer mitochondrial membrane complex (used as an outer membrane marker for mitochondria) and ß-tubulin, a protein present in microtubules; the reporter dyes were attached to secondary antibodies.
Comparison of images of microtubule-mitochondrial interactions in mammalian cells as obtained from conventional and 3D-STORM microscopy. (a) A conventional fluorescence image of mitochondria (magenta) and microtubules (green). The image is slightly blurred and the distance between the mitochondria and microtubules, if any, is not visible since a single mitochondrion is seen to touch multiple microtubules. (b) STORM image of the same area with all localizations at different z positions stacked. The image is acquired in aqueous media and reconstructed from 500,000 localization points. This image, contrary to the conventional image (a), clearly shows a 150nm separation between the mitochondrion and one microtubule, whereas the same mitochondrion was in much closer proximity to another microtubule.
The STORM-image provided a clear picture of mitochondria and microtubules, allowing a better understanding of the spatial relation between them as compared in fig.13. STORM is not limited to imaging the interactions of only two sub-cellular structures and can be used to image multiple structures by differentially labelling them. STORM can be extended to imaging motor-proteins, the main complexes which facilitate the mitochondrial movement along microtubules, further illustrating their interactions and providing a better understanding of the regulation of morphology of these “power houses” within a cell, thereby having much potential for future nanoscale-research.
However, STORM requires large numbers of raw images of localised-molecules to be taken from different imaging-frames so that the entire super-resolution image can be constructed, and this limits the speed of this technique and the acquisition time required to construct the highest-resolution image requires a few minutes .
Conclusion
It is safe to conclude by saying that microscopy has come a long way since its first discovery in the late 16th century and reached an era when the diffraction limit is being surpassed so that individual nanoscale molecules can be observed. In the past decade, super-resolution microscopy has taken a big jump and techniques like STED, SSIM, TIRF-SSIM and PALM/STORM have been developed. Even though each of these techniques accommodates features like greatly improved image-resolution, 3-dimensional imaging, live-sample imaging and multi-coloured imaging, each of these has its own limitations. In the ideal world, STED microscopy would be expected to work independent of light-wavelength and unaffected by the high-intensity lasers. Similarly, PALM/STORM would be expected to be faster techniques requiring lesser raw-images and SSIM would be expected to be unaffected by photobleaching and sample-positioning. SIM can be applied for live-imaging, 3D-imaging and multicolour-imaging; however, its resolution is still not as good as that provided by STED microscopy (live-imaging, 3D-imaging) and PALM/STORM (3D-imaging, multicolour-imaging). Therefore, at this point, it is hard to tell as to which of the above explained techniques is the best since each of them have their advantages and pitfalls and each has significant potential in different areas of biological research. As of now, considerable progress has been made in microscopy, hence opening many doors in cell biology and it is safe to say that in the future, technology will improve, and new imaging techniques will be developed
This is Preview only. If you need the solution of this assignment, please send us email with the complete assignment title: ProfessorKamranA@gmail.com
