The Trap in the FRAP: A Cautionary Tale about Transport Measurements in Biomolecular Condensates

Main Text

Membraneless organelles are intracellular compartments that segregate their components from the surrounding cellular milieu via phase separation. In the last 10 years, a growing number of experiments have brought to light the many functions of these biomolecular condensates, which include transport and processing of RNA, territorial organization of the genome, and cellular response to stress factors (1). At the same time, the dysfunction of membraneless organelles plays a role in disease, including neuropathies and cancer (2). Together, these observations have boosted systematic in vitro and in cell investigations of the molecular determinants required for intracellular phase separations, aiming to understand how the inherent physical properties within the single components control phase boundaries, transport, and material properties of membraneless organelles (3).

Many of these biomolecular condensates assemble through liquid-liquid demixing and therefore retain characteristics of liquid droplets (e.g., dripping, fusing, and wetting). Fluorescence recovery after photobleaching (FRAP) is the technique often chosen to test the mobility of the molecular components inside the condensates and verify their liquid character. Briefly, in these FRAP experiments, the fluorescent emission in a region of interest within a fluorescently doped condensate is “switched off” (photobleached) by using a sufficiently high-power laser; subsequently, the exchange of fluorescent molecules with the surrounding environment is visualized by monitoring the recovery of the fluorescence emission over time (see Fig. 1). The amplitude of the recovery emission relates to the fraction of molecules in the mobile phase within the condensate, whereas the recovery time reports on the diffusion coefficient of these molecules. Absence of recovery indicates that the components inside the condensate do not exchange within the probed timescale.

Figure 1.

Fluorescence recovery after photobleaching in membraneless organelles. At time t_PB, a region of interest (ROI) of the condensate is exposed to a high intensity beam that photo-bleaches all the fluorescent molecules within this area; the rapid exchange of molecules within the dense phase and between the dense and the light phase contributes to the fluorescence recovery after photobleaching. The amplitude and characteristic time of the recovery provide information on the transport properties of the condensate. However, the quantitative interpretation of FRAP data requires a careful testing of the adopted model, including contributions due to 1) the ratio between the photobleached volume/overall volume of the droplet, 2) the shape of the excitation and detection beam, 3) assumptions on fixed or infinite boundaries, and 4) diffusion- or reaction-dominated interactions between the molecules (4). To see this figure in color, go online.

FRAP is not a new technique; on the contrary, it is a well-established approach, of easy implementation, with a long list of diverse applications in literature. However, easy does not mean simple, and the work of Taylor et al. (4) is a timely cautionary tale about the quantitation of transport properties in biomolecular condensates using FRAP. Currently, there is no unique strategy for the analysis of such data, either in selecting the model (simple exponential, fixed, or infinite boundaries) or in the choice of the relevant observable (the half-time, the exponential lifetime, the diffusion coefficient). The lack of having an agreed-upon analytical approach significantly hinders the comparison of measurements across different condensates both in vitro and in vivo, limiting the understanding of how the molecular components modulate the material properties of the dense phase.

Taylor et al. (4) set out to explore this question by comparing how different assumptions in the models affect the interpretation of FRAP data. For example, the typical single-exponential fit of fluorescence recovery can be derived by invoking reaction-limited transport (i.e., the recovery is dominated by the interactions between components) or assuming constant concentration at the photobleached boundary (a condition that is commonly not met because of the exchange between the photobleached area and the surrounding milieu). An alternative strategy is offered by infinite boundary models, which go beyond the single-exponential fit and allow for the concentration at the boundary of the photobleached region to vary. To test the implications of these approaches, Taylor et al. (4) makes use of an ingenious application of microfluidics, which enables generating experimentally “infinite boundary” conditions. The direct comparison of these models reveals hidden “traps” in their applicability to the study of biomolecular condensates, which requires taking in consideration the shape of the photobleached profile and the ratio of the photobleached area versus the overall size of the condensate. Notably, the application of “incorrect” theories can lead to orders of magnitude of discrepancy in the estimate of diffusion coefficients! These results parallel recent studies that pointed to the importance of geometrical and optical factors in the quantitative analysis of cellular FRAP data (5). Though a unique approach may not exist and more detailed treatments can be applied, the work of Taylor et al. (4) calls for a careful consideration of the models used in the analysis of FRAP data and delineates a strategy to test their applicability to biomolecular condensates.

FRAP is one method in a growing list of approaches for the quantification of the transport properties inside condensates, which includes fluorescence correlation spectroscopy (6,7), raster image correlation spectroscopy (8), and microrheology particle tracking (9,10). The development of new models that account for the different length- and timescales probed by these techniques is essential for decoding how the physicochemical properties of the single constituents affect the collective transport properties of dense phases.

The work of Taylor et al. (4) is a step in the right direction, pointing to the importance of determining physical observables that can be compared across different systems and techniques and providing guidelines for the correct determination of such quantities.

Editor: Rohit Pappu.