Keep PALM and STORM on

13 February 2018

MT-Feb-18-STORM-fluorescence-microscope.jpg

Scientists have long depended on microscopes in order to observe and understand biology, particularly when the organisms under investigation are smaller than the human eye can see. The earliest observations using microscopes date to the 17th century with Antonie van Leeuwenhoek and Robert Hooke, who were the first to observe micro-organisms. 

Technological advances have continued to improve how much detail can be seen, one of the most powerful being fluorescence labelling of pre-defined targets for the direct visualisation of cellular processes. However, higher image magnification does not equate to an improved image resolution at small length scales because of the interactions of light as it passes through the microscope optics. As a wave, light is subject to diffraction – the bending of a wave after it interacts with an obstacle or object – and this causes the image of a point object to blur into a finite-sized spot. 

MT-Feb-18-STORM-fig-1.jpg
Fig. 1. Single molecule localisation microscopy.

Seeing is believing

This was first described by Ernst Abbe in 1873, and its relation to image resolution defined by Lord Rayleigh in 1879. In a fluorescence microscope using visible light (λ = 550 nm) and a high numerical aperture (NA = 1.4) lens, for two objects to be resolved as separate they must be positioned ~240 nm apart. This is considerably larger than much of the subcellular architecture that constitutes a cell. Therefore, the molecular details of objects are obscured by this blurring as the signals of thousands of fluorescently labelled targets overlap and merge together.

In 2014, Eric Betzig, Stefan W. Hell and William E. Moerner jointly shared the Nobel Prize in Chemistry for the development of super-resolved fluorescence microscopy. These advances in optical engineering and photochemistry led to the development of a number of fluorescence microscopy techniques which bypass the diffraction barrier. These techniques can resolve biological samples using conventional fluorescence labelling strategies with up to a 10-fold improvement in image resolution.

Bypassing the barrier

One type of super-resolution microscopy employs mathematical modelling to localise single fluorescent events with a precision better than the original diffraction-limited volume. Termed single molecule localisation microscopy (SMLM), these approaches rely on the ability to separate fluorescent events for individual labelled molecules such that only a subset are imaged at any one time (Fig. 1). This is achieved using different photochemical approaches but results in the successive blinking of fluorophores ‘on’ and ‘off’. Sequential imaging and localisation of thousands of fluorophores that are spatially distinct can then be used to reconstruct a super-resolved image. 

Since the initial development of these techniques, the field has become swamped in acronyms which each detail how the particular flavour of SMLM causes fluorophore blinking, or how fluorophores are localised. However, all approaches can ultimately be grouped into two main categories. 

Photo-activated localisation microscopy (PALM) uses photo-switchable fluorescent proteins that are genetically tagged to the biological protein of interest. These proteins undergo a photoconversion, either by switching from one fluorescent state to another (e.g. from green fluorescence to red), or by activation from a dark state to a fluorescent ‘on’ state.

Stochastic optical reconstruction microscopy (STORM) on the other hand uses organic fluorescent dyes. These are placed into a ‘dark’ state under certain experimental conditions, and stochastic activation into the ‘on’ state is controlled by careful laser excitation (Fig. 2).

Fig 2. Lasers supplied to a fluorescence microscope.
MT-Feb-18-STORM-fig-2.jpg

SMLM microscopy: bringing viruses into focus

Considering the similarities in size between virus particles and the resolution of a light microscope, it is surprising that there are but a few virological studies exploiting the resolution improvements of SMLM. However, the application of these specialised imaging techniques has begun to gain traction in the microbiology community and is providing novel insights into virus biology.

One such study directly visualised the precise organisation of tegument and envelope proteins within herpes simplex virus (HSV) particles. The authors combined the molecular identification strategies of fluorescence labelling, the improved resolution of SMLM, and particle averaging techniques used in electron microscopy to distinguish individual protein layers within the HSV virion. Similar approaches have been applied to human immunodeficiency virus (HIV) and have been of sufficiently high resolution to distinguish the conical morphology of the mature capsid and the internal architecture of HIV particles. 

We at the University of Leeds have been using SMLM to investigate the molecular architecture of hepatitis C virus (HCV) replication factories. These are specialised sites for virus genome replication that are formed in a coordinated process that requires a number of viral and cellular proteins. Our investigations have resolved differences in size as small as 10–30 nm in the distributions of two key proteins known to be involved in the establishment and maintenance of replication factories. This novel information is at a length scale far below what is achievable using conventional microscopy approaches.

SMLM as a quantitative tool

The power of SMLM, however, is not just in the ability to reconstruct high- resolution images from the coordinates of localised fluorophores. Information about the molecular stoichiometry, clustering behaviour and geometric features can be extracted from the mathematical relationship between the positions of localisation coordinates in space and time. 

However, with great power comes great responsibility. The pointillist nature of SMLM data requires that sufficient testing of fluorophores from the sample is conducted such that the underlying biological structure is accurately revealed. For example, are gaps in a filamentous structure biologically relevant or a consequence of undersampling? 

Conclusion

This is an exciting time to be a virology researcher; the advancements in fluorescence imaging now allow a direct visualisation of viruses and their intricate interplay with the host cell using conventional fluorescence labelling strategies. The meeting point between disciplines frequently proves fruitful for developing ever more impressive technical feats. The application of these methodologies will continue to offer novel insights into often well-trodden topics. Looking forward, the next biggest challenge for virologists seeking to exploit SMLM is incorporating robust image analysis methods in order to correctly interpret the observations and ensure that seeing is believing. 

Further reading

Grove, J. (2014). Super-resolution microscopy: a virus’ eye view of the cell. Viruses 6, 1365–1378.

Laine, R. F. & others (2015). Structural analysis of herpes simplex virus by optical super-resolution imaging. Nat Commun 6, 5980.

Schermelleh, L. & others (2010). A guide to super-resolution fluorescence microscopy. J Cell Biol 190, 165.

MT-Feb-18-STORM-chris.jpg
Christopher Bartlett

School of Molecular and Cellular Biology, University of Leeds, Leeds, LS2 9JT

[email protected] 
Facebook: bartlett.christopher

Christopher Bartlett is a molecular virologist with a specialism in advanced microscopy technologies. He achieved a first-class degree in Microbiology with Virology at the University of Leeds in 2012 before completing a PhD in Molecular Biology in 2017. Software engineering and image analysis methods are Christopher’s current interests.

How did you enter this field?

I developed an interest in virology through my undergraduate studies in Microbiology at the University of Leeds. I then completed a PhD in Molecular Biology in the laboratories of Mark Harris and Michelle Peckham. My research focused on using PALM and STORM to study hepatitis C virus replication. I continued this project for a further year as a research fellow, where I established a greater interest in software engineering and image processing technologies. 

What parts of your job do you find most challenging?

The super-resolution microscopy techniques produce fantastic, high-resolution images. The biggest challenge is the subsequent analysis and careful interpretation of these large data sets.


Images: Figs. 1 and 2. C. Bartlett.