Unraveling the geometry of complex protein organizations by polarized fluorescence imaging

Main text

Molecular interactions, which are at the core of biological phenomena, lead to complex and heterogeneous molecular organization that is challenging to decipher in living samples. Spatiotemporal live organization can be imaged by optical imaging microscopy for which increasing spatial resolution is reached nowadays, but in dense sample, organization is rather dominated by orientational patterns, which are the signature of specific ordered or disordered interactions. Orientational organization is crucial to access to understand key processes such as phase formations in cells and plasma membranes, protein clustering involved in immunology, cytoskeleton filaments alignments in cell and tissue mechanics, protein aggregation involved in neurodegenerative diseases, or DNA organization involved in repair processes in the nucleus. Light polarization was introduced several decades ago in fluorescence microscopy to decipher orientation properties of proteins and lipids in biological samples.

Linear incident light polarization interacts with fluorescent labels in a way that it maximizes absorption (and thus fluorescence) when its direction lies along the fluorophore direction defined by its absorption transition dipole moment, which is along the main conjugation direction of the label. This approach has been successful in lipid probe orientational imaging, providing sensitivity to lipid phases [1] and membrane nanoscale morphological changes such as ruffling [2]. Using this approach to study proteins organization requires, however, labeling protein domains in a rigid way that minimizes orientational flexibility. Increasing numbers of examples of rigid labeling are introduced, as illustrated in studies such as septin in yeast [3], nuclear pore complexes [4], integrin [5], actin [6], or amyloids [7].

The paper by Dean et al. pushes the capability of the methods for the determination of complex geometries of the organization of proteins at the adhesive interface between cells, targeting the desmosomal cadherins (DCs) in cell-cell adhesion regions [8]. DC architecture is expected to be related to the adhesive state of two plasma membranes, but it is currently poorly known. By combining together knowledge gained from both fluorescence polarized microscopy (FPM) and electron tomography (ET), the authors probe different organization scenarios and exclude geometries that are still at the center of controversies. To do so, a careful engineering of location of the fluorescent probe at specific domain locations is performed, where the orientational flexibility is seen to be reduced compared with control locations at the C terminal of the proteins. This study shows the great advantage of combining two very complementary methods: on the one hand, ET brings ultimate spatial resolution level of protein organization in 3D, but it is not specific, requires averaging, and might lead to ambiguities in the reconstructed structural information. On the other hand, FPM is applicable to live samples with high labeling specificity in precise domain locations, and it gives access to spatiotemporal statistical heterogeneities, but it is limited to spatial resolutions of hundreds of nanometers, requires rigid labeling, and measures only in-plane projected information.

In [8], combined information from ET and FPM are used to quantify the orientational order of the extracellular and intracellular domains of three DC isoforms (Desmoglein 2, Desmocollin 2, and Desmoglein 3), whose interactions with binding partners involve intrinsically disordered domains that only become structured upon binding. The authors have chosen to target the close proximity to the membrane by a strategic labeling, a region delicate to access with cryo-ET. DC ectodomains are found to be significantly more ordered than the cytoplasmic domain, which is conserved over all DCs tested. This drastic difference in DC order between opposing sides of the plasma membrane suggests that extracellular adhesive interactions induce significant ordering of the desmosomal architecture. Simulations confirm the possibility of a recently proposed “antiparallel” arrangement of DC ectodomains [9].

This study illustrates the gain brought by constructive knowledge between different methods, but it also highlights the progresses still to be done: future developments in polarized optical imaging are required, not only to increase time resolution, but also for the determination of 3D orientations. This parameter is indeed currently missing and might lead to ambiguous interpretation of projected structures of complex geometries. At last, remaining challenges remain in the access to labels of low-orientational flexibility, and independent characterization of this flexibility is necessary to infer nonambiguous models for orientational patterns in complex architectures. These considerations emphasize possible advantages of the use of polarized single molecule localization microscopy in the field of structural imaging. Recent progresses in single molecule orientation imaging indeed have demonstrated an advantageous combination of high spatial 3D localization with tens of nanometer resolution, with 3D orientational measurements [10,11]. These approaches could bring additional knowledge on 3D single labels orientation and rotational mobility and the nanoscale, as well as bridge the gap to optical structural imaging of protein organizations in cells and tissues.

Editor: Anne Kenworthy