As our quest for understanding the molecular mechanisms underpinning endocrine pathways and pathologies intensifies, the requirement for visualising cellular processes in high definition is ever-pressing. As such, the last 10–15 years have witnessed a technological explosion in the field of microscopy, resulting in the advent of super-resolution imaging.

Indeed, the importance of this fast-moving field was recently recognised by the Nobel Committee for Chemistry, as three pioneers of super-resolution microscopy – Eric Betzig, William Moerner and Stefan Hell – were awarded the 2014 prize for localisation microscopy and stimulated emission depletion (STED) respectively.

 

THE SIGNIFICANCE OF SUPER-RESOLUTION IMAGING

‘But what exactly is super-resolution imaging?’ I hear you ask. The exciting development it brings is the ability to break the diffraction limit. Most of the commonly used microscopy techniques, such as confocal and widefield microscopy, have a maximum (and often theoretical) resolution of 200nm. Super-resolution imaging provides enhancements in resolution of between 2- and 20-fold, depending on the sample used and technique employed.

So, for the first time, we can resolve true single molecules which, for those interested in receptor-mediated processes, or protein–protein and protein– DNA interactions, provides the possibility of imaging these processes in high definition.

 

QUESTIONS WITH G PROTEIN-COUPLED RECEPTORS

For my own research, super-resolution imaging has been transformative. A long-debated question in the G protein-coupled receptor (GPCR) field is the functional significance of GPCR di/oligomerisation, particularly for family A receptors. Our previous studies had shown that the luteinising hormone receptor (LHR), a receptor essential for reproduction, can act via di/oligomerisation in vivo, using a functional complementation approach. This was the first study to present conclusive evidence for the physiological relevance of family A receptor di/oligomerisation.

However, despite this leap in understanding, we still lacked knowledge about the cell surface landscape of the LHR. We therefore wanted to know whether the LHR formed dimers or oligomers. Additionally, if oligomers were formed, what types of oligomeric forms could be observed and were they ligand-regulated?

To answer these questions, we went back to an in vitro model, using HEK293 cells stably expressing either the LHR or the functional complementation LHR mutants employed in our ^{in vivo} study: a ligand-binding deficient LHR (LHR^B–) and a signalling deficient LHR (LHR^S–) which, when co-expressed, recapitulated LHR function via di/oligomerisation (Figure 1).

 

ANSWERS FROM SUPER-RESOLUTION IMAGING

To image the cell surface landscape of the LHR, we employed the super-resolution imaging technique of photoactivatable localisation microscopy (PALM). PALM was used as it afforded a resolution of <10nm, meaning that we could image individual LHR molecules at the cell surface and thus identify the monomer/dimer/oligomer populations of LHRs.

PALM is based upon the use of photoactivatable fluorophores, which remain in the dark state until activated by ultraviolet light. The fluorophores are activated in a stochastic fashion, emitting fluorescence in the fluorophore-defined wavelength, and are subsequently photo-bleached into the dark state (Figure 2). The important aspect for identifying single molecules (and the premise of the technique) is the stochastic nature by which fluorophores are activated, allowing spatially separate detection of activated fluorophores, and thus single molecule detection. Samples are subjected to repeated cycles of activation and photo-bleaching to ensure all fluorophores are imaged.

The fluorophores we employed were CAGE dyes – CAGE 500 and 552 – which are brighter than most photoactivatable proteins available. They could also be directly conjugated with primary antibodies to the small peptide-tagged LHRs at a 1:1 dye:antibody ratio, allowing for direct quantification of LHR protomers.

Due to the high laser power and repeated cycles of activation and photo-bleaching that are required to image samples, PALM was conducted on fixed cells using a custom-adapted Inverted Axiovert 200 wide-field fluorescent microscope (Zeiss, Germany) fitted with a commercial TIRF (total internal reflection fluorescence) condenser and a polychrome light source.

 

INTERPRETING THE DATA

Once imaging has been carried out, a large (and time consuming!) part of localisation microscopy is the post-acquisition data analysis to ‘localise’ the molecules imaged. To resolve the x and y co-ordinates of the LHRs, we used the free ImageJ plugin, QuickPALM. This software identified individual LHRs from image files, using a set of in-built stringency filters to ensure the integrity of the data generated.

Once the co-ordinates were obtained, we interrogated our study questions. To identify LHR participating in dimers and oligomers, we employed a mathematical model adapted from the Getis–Franklin nearest neighbourhood approach (Figure 3).

Our analysis revealed that approximately 40% of LHRs were in di/ oligomers in both the wild type LHR and functional complementation LHR^B–/LHR^S–. Interestingly, LHR^B–/LHR^S– favoured the formation of oligomers. No ligand-dependent changes in the number of di/oligomers were observed. However, modulating the ratio of LHR^B–:LHR^S– within lower order oligomers regulated the magnitude of G protein-dependent signal output.

Our studies show how super-resolution imaging can be used to provide such levels of unprecedented molecular detail, revealing how receptors within an oligomer can fine-tune signal responses.

Kim Jonas

Lecturer in Reproductive Physiology, St George's University of London

 

FURTHER READING

1. Jonas KC et al. 2016 Methods in Cell Biology 132 55–72.

2. Jonas KC et al. 2015 Journal of Biological Chemistry 290 3875–3892.