Fluorescence based microviscosity mapping in membraneless organelles

25 26 Membraneless organelles (MLOs) are cellular biomolecular condensates formed by liquid-27 liquid phase separation. Their biological functions are intimately linked to their material 28 properties including viscosity. Condensate viscosity is determined by the size, shape, 29 concentration and molecular interactions between MLOs components. It impacts the diffusion 30 of MLOs constituents and the selective permeability of the condensate, thereby regulating the 31 rate of biochemical reactions. Viscosity modifications associated with liquid -to-gel transition 32 of the condensates are related to pathologies. 33 Current experimental approaches for characterizing the material properties of cellular 34 condensates remain limited. In this study, we report the use of BODIPY-based molecular rotors, 35 in combination with fluorescence lifetime imaging microscopy (FLIM), to mo nitor the 36 microviscosity of cellular MLOs directly in living cells. The fluorescence lifetime of BODIPY 37 derivatives increases with the viscosity of their microenvironment, enabling quantitative 38 assessment of microviscosity within condensates. HaloTag techn ology was employed to 39 specifically label MLO components. 40 Our findings reveal that the nucleolus exhibits higher viscosity than the surrounding 41 nucleoplasm and that microviscosity varies across nucleolar sub-compartments. Furthermore, 42 nucleolar reorganizati on induced by inhibition of rRNA synthesis results in a measurable 43 increase in microviscosity. Finally, we demonstrate that the microviscosity of stress granules is 44 lower than that of the nucleolus. 45 Overall, presented results demonstrate the strong potenti al of the BODIPY based molecular 46 rotors as a versatile and powerful tools for probing the material properties of cellular MLOs. 47 48 49 50 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 2

51 52 Liquid-liquid phase separation (LLPS) has recently emerged as a fundamental mechanism 53 underlying cellular orga nization, driving the formation of biomolecular condensates, lipid 54 membrane domains, and selective barriers .1 LLPS is a thermodynamically driven, reversible 55 process in which components of a homogeneous solution demix to form dense liquid 56 condensates dispersed within a dilute phase. In the cells LLPS gives rise to organelles, call ed 57 membraneless organelles (MLOs), such as stress granules, p -bodies, Cajal bodies or 58 nucleolus.2,3 59 MLOs are complex, heterogeneous condensates that behave like fluids: they fuse, coalesce and 60 drip, which are behaviors governed by their physicochemical characteristics (i.e. surface 61 tension, viscosity, etc. ). Due to their LLPS nature, MLOs represent transient, dynamic, and 62 open systems whose biological functions are tightly linked to their composition, architecture, 63 and material properties. These parameters are of crucial importance, as the internal environment 64 of condensates must be precisely regulated to fulfill their biological roles.4 65 Viscosity is one of the keys p hysicochemical properties of molecular condensates. Viscosity 66 represents the resistance of a fluid to flow. On a molecular scale this property is determined by 67 concentration, size, shape, and interactions of the fluid’s components. In MLOs, viscosity is 68 particularly important because it affects internal dynamics and diffusion of their components. 69 Condensate viscosity regulates the interfacial properties responsible for the droplet shape and 70 the selective permeability which in turn impacts the rate id biochemical reactions. Under certain 71 conditions, MLOs may undergo further phase transitions into gel-like or solid states, which are 72 often associated with patholgies.5 Therefore, understanding the relationship between biological 73 function and the viscoelastic properties of MLOs remains a major challenge. 74 Several experimental approaches have been developed to probe condensate rheology at 75 different scales. Passive microrheology, based on tracking fluorescent beads within 76 reconstituted droplets, enables viscosity quantification in controlled environments , while 77 droplet coalescence dynamics provide complementary estimates of interfacial tension and 78 viscosity.6 Active micro-rheology using optical tweezers further allows direct measurement of 79 viscoelastic properties by applying oscillatory forces to embedded particles. 7–10 80 Compared to “in vitro” systems, viscosity measurements in cellular condensates are more 81 fastidious. In living cells, fluorescence recovery after photobleaching (FRAP) has been widely 82 used to assess component mobility and infer viscosity from diffusion parameters. 11,12 More 83 recently, optical diffraction tomography , a label-free imaging technique exploiting refractive 84 index contrast, has been applied to quantify condensate density.13,14 Together, these approaches 85 have expanded our understanding of condensate material states, yet mapping viscosity at the 86 nano to micro scale within living cells remains challenging. 87 In this work, we introduce a complementary method for measuring the microviscosity within 88 cellular condensates. We designed and optimized a BODIPY-based molecular rotor compatible 89 with fluorescence lifetime imaging microscopy (FLIM), enabling direct, qu antitative readout 90 of local microviscosity. BODIPY dyes are environment -sensitive fluorophores, also called 91 “molecular rotors” whose excited-state relaxation depends on environmental viscosity. In low-92 viscosity media, energy relaxation occurs via non-radiative intramolecular rotation, whereas in 93 viscous environments, restricted motion leads to increased fluorescence quantum yield and 94 lifetime. BODIPY-based rotors have already demonstrated high sensitivity for probing 95 viscosity in cellular compartments such as the plasma membrane, mitochondria, and 96 endoplasmic reticulum, 15–19 as well as in protein aggregates .20 However, their potential for 97 investigating cellular MLOs has remained unexplored. 98 To address this gap, we developed a chemogenetic labeling strategy based on a BODIPY 99 derivative functionalized with a chloroalkane ligand for covalent conjugation to HaloTag fusion 100 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 3 proteins.21 This approach enables selective labeling of condensate components and direct 101 measurement of local viscosity in living cells. Using this tool, our experiments revealed distinct 102 viscosity distributions within nucleolar sub-compartments and dynamic changes accompanying 103 nucleolar reorganization upon inhibition of RNA transcription. Furtherm ore, we identified 104 significant viscosity differences between nucleoli and stress granules, reflecting their distinct 105 internal architectures. Together, these findings establish BODIPY-based molecular rotors as a 106 powerful tool for live-cell imaging of microviscosity in biomolecular condensates, offering new 107 insights into the material properties that underlie cellular organization. 108 109 110

111 112 Chemicals and Synthesis: 113 All starting materials were purchased from Sigma -Aldrich, BLD Pharm, or TCI Europe and 114 were used as received unless otherwise stated. 115 NMR spectra were recorded on a Bruker Avance III 400 MHz spectrometer. Chemical shifts 116 (δ) are reported in ppm relative to residual solvent signals. High -resolution mass spectra were 117 acquired using an Agilent Q -TOF 6520 mass spectrometer. Detailed synthetic protocols and 118 full characterization of all compound s are provided in the Supplementary Materials and 119

information. 120 121 Preparation of Glycerol–Water Mixtures for Rotor Calibration 122 To calibrate the viscosity sensitivity of the BODIPY -based rotor, a series of glycerol -water 123 mixtures with different viscosities were prepared. Glycerol (Euromedex, France) and ultra-pure 124 MilliQ were mixed with defined mass ratios to achieve final glycerol percentages (w/w) 0%, 125 35%, 45%, 60%, 67%, 72%, 75%, 77%, 82%, 85%, 90% and 95%. Glycerol mixtures with 10 126 mM Tris, 150 mM NaCl at pH 7.5 were prepared to assess the influence of protein binding 127 under physiological buffer conditions. The viscosity of each solution was measured at 20°C 128 using a HAAKE MARS rotational rheometer (Thermo Scientific, Waltham, MA, USA) 129 equipped with a 35 mm parallel-plate geometry (P35/Ti/SB) set to a 0.10 mm gap and operated 130 with a solvent trap to minimize evaporation. Approximately 100 µL of sample was loaded onto 131 the lower plate and allowed to equilibrate at 20°C prior to measurement. Steady -shear flow 132 curves were acquired in controlled shear -rate mode, applying shear rates from 0.1 to 1000 s⁻¹ 133 and recording the corresponding shear stress. The apparent viscosity was calculated as 𝜂 = 𝜏 𝛾̇⁄ 134 (where 𝛾̇ is shear rate, 𝜏 the shear stress), and, for such Newtonian mixtures, taken from the 135 shear-rate-independent plateau. Measurements were performed in triplicate. 136 137 Fluorescence Quantum yield measurements: 138 The quantum yield (QY) of the BODIPY acid was measured for G0%, G45%, G67%, G77%, 139 G85% and G95% using the integration sphere (SC -30) module of F S5 fluorometer from 140 Edinburgh Instruments. This module allows to determine the absolute fluorescence quantum 141 yield of the dye. A dye concentration of 500 nM was used in a total sample of 4 ml. The sample 142 was heated overnight at 50°C to allow the dye to mix properly at h igh viscosity (i.e. G77%, 143 G85% and G95%). The excitation wavelength was set to 430 nm and the emission was recorded 144 from 440 to 700 nm. 145 146 Fluorescence Lifetime Measurements 147 Time-resolved fluorescence measurements were performed with the time -correlated si ngle-148 photon counting technique. Excitation pulses at 500 nm were generated by a supercontinuum 149 laser (NKT Photonics SuperK Extreme) with 10 MHz repetition rate. The fluorescence decays 150 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 4 were collected at 520 nm using a polarizer set at magic angle and a 16 mm band -pass 151 monochromator (Jobin Yvon). The single -photon events were detected with a micro -channel 152 plate photomultiplier R3809U Hamamatsu, coupled to a pulse pre -amplifier HFAC (Becker -153 Hickl GmbH) and recorded on a time -correlated single photon counting board SPC -130 154 (Becker-Hickl GmbH). The measured decays were fitted by using a function corresponding to 155 an exponential decay convolved with a normalized Gaussian curve of standard deviation σ 156 standing for the temporal IRF and a Heavyside function. The fitt ing function was built in Igor 157 Pro (Wavemetrics). All emission decays were fitted using a weighting that corresponds to the 158 standard deviation of the photon number squared root. 159 Glycerol-water solutions were incubated at room temperature overnight to ensure homogeneity. 160 A stock solution of 200 µM BODIPY rotors was prepared in DMSO. For each glycerol -water 161 mixture, the BOD-L, BOD-PEG4-L, and BOD-PEG12-L rotors were added to achieve a final 162 concentration of 50 nM. 163 164 Cell Culture 165 Human Embryonic kidney (HEK293) cell lines and Human bone osteosarcoma epithelial cell 166 lines (U2OS) were used for live cell imaging experiments. HEK293T cells were grown in 167 Dulbecco’s Modified Eagle Medium (DMEM,Gibco) supplemented with 10% foetal bovine 168 serum (FBS), 100 μg/mL penici llin, streptomycin, 2 mM L -glutamine and 1 mM sodium 169 pyruvate. U2OS cells were grown in modified Mc’Coy media containing 10% FBS, 100 μg/mL 170 penicillin, streptomycin and 1 mM sodium pyruvate. All cell lines were cultured at 37°C in 171 humidified atmosphere containing 5% CO2. 172 For fluorescence imaging, cells were seeded onto a 35 mm Ibidi ibi-treat or glass bottom dish. 173 HEK293T cells were seeded with density of 1.5 × 105 cells while U20S were seeded with a 174 density of 5 × 104 cells/dish in 2 mL of their respective media. 24 hours post seeding; cells 175 were transfected with plasmid DNA using Jet PEI transfection kit according to supplier’s 176 protocol. 177 pcDNA-NPM-HaloTag22 and pcDNA-Fib-HaloTag were cloned from eGFP-NPM1 (Addgene 178 n°:17578) kindly provided by Dr. Wang 23 and eGFP-Fib (Addgene n°:26673) provided by Dr. 179 Chen.24 The sequence coding for HaloTag was amplified from pSEMS-Halo7Tag-hFis (111136 180 Addgene) vector provided by Dr. Karin Busch 25. After purification, the PCR products were 181 digested by BamHI/Xho restriction enzymes and inserted in pcDNA3.1 (zeo) vector. The 182 obtained pcDNA-HaloTag was further digested with HindIII/Bam-HI restriction enzymes and 183 ligated with the NPM or Fib inserts amplified from eGFP -NPM1 and eGFP -Fib plasmids 184 respectively. The final pcDNA-NPM-HaloTag was ligated and sequenced for verification. 185 Fib-SNAP plasmid was provided by Dr. Castano and Dr. Kriz 26 and HaloTag-G3BP1 by Dr. 186 Brands’ lab.27 Plasmids coding for eGFP1 and eGFP3 were provided by Dr. Nalaskowski. 28 187 Plasmids coding for free HaloTag, HaloTag -H2B, HaloTag-LifeAct were kindly provi ded by 188 Dr. Gauthier. 29 189 Inhibition of RNA Pol 1 was p erformed by incubating the cells during 2 hours in complete 190 medium containing 2µg/mL Actinomycin D. The stres s granules formation was induced by 191 addition of 500µM NaAsO2 into the complete growth medium during 1 hour. Cellular RNAs 192 were stained by Pyronin Y (1µM, 15minutes, 37°C). Actinomycin D, NaAsO2 and Pyronin Y 193 were purchased from Sigma-Aldrich. 194 Janelia fluor dyes coupled to Halo and Snap ligands were provided by Janelia materials (HHMI 195 Janelia Research Campus). 196 197 198 199 200 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 5 Fluorescence lifetime imaging 201 Approximately 24 hours post transfection, the growth media was replaced with Opti -MEM 202 (Gibco), and cells were incubated with 200 nM of BODIPY rotor dye for 15 -20 minutes at 203 37°C. Excess dye was removed by washing with Opti-MEM. 204 FLIM measurements were performed on a Leica Stellaris equipped with a Falcon module and 205 on a homemade two-photon scanning FLIM microscope. 206 Commercial inverted confocal laser scanning microscope (STELLARIS 8, L eica 207 Microsystems, Nanterre, France) was equipped with a fully fast integrated FLIM module, 208 the so-called FAst Lifetime CONtrast (FALCON, Leica Microsystems, Nanterre, France) 209 and a white light laser (WLL2 440−790 nm). Acquisitions were performed through a 210 512×512 image format, a scan speed at 400 Hz and a 63X (NA 1.4) oil immersion objective. 211 Bod-4PEG-L imaging was performed with 488 nm excitation (WLL, 8% power). In a photon 212 counting mode, hybrid PMT HyD-X detector was used to detect fluor escence emission from 213 505 to 750 nm. FLIM images were acquired with accumulation of 8 lines and 3 frames 214 repetitions to detect ~1000 photons in each pixel showing Bod-PEG4-L labelled structures. 215 A minimum of 100 photons was set to pixels represented in the phasor plot. 216 The homemade multiphoton scanning microscope is based on an inverted microscope (IX83, 217 Olympus) with a 60X 1.2 NA water immersion objective operating in the descanned 218 fluorescence collection mode. 30 The BODIPY derivatives were excited at 780 nm using a 219 femtosecond laser (Insight DeepSee, Spectra Physics). Fluorescence photons were collected at 220 a dwell time of 4 µs/pixel using a short -pass filter with a cutoff wavelength of 720 nm 221 (Semrock, FF01-720/SP-25). The fluorescence was directed to a fiber -coupled APD (SPCM-222 AQR-14-FC, Perkin Elmer), which was connected to a time-correlated single photon counting 223 module (SPC830, Becker & Hickl). The measurement was controlled by SPCM ver 9.83 224 (Becker & Hickl). The time -resolved fluorescence decay at each pixel was analyzed using a 225 commercial FLIM analysis software package Becker a nd Hickl SPCImage. The decay curves 226 were fitted using a biexponential model, convolved with the instrument response function 227 (IRF), as: 228 229 𝐼(𝑡) = 𝐼𝑅𝐹⨂(𝐴1𝑒−𝑡/𝜏1 + 𝐴2𝑒−𝑡/𝜏2) 230 231 where I(t) represent the intensity decay, IRF denotes the instrument response function of the 232 system. A1 and A2 are the amplitudes of the two decay components with 1 and τ2 fluorescence 233 lifetimes. The average fluorescence lifetime  was then calculated as: 234 235 𝜏 = 𝜏1 × 𝐴1 + 𝜏2 × 𝐴2 𝐴1 + 𝐴2 236 237 Pixel-wise fitting was carried out to generate fluorescence lifetime maps, and regions of interest 238 (ROIs) were selected to extract quantitative data for comparison across experimental 239 conditions. 240 Phasor analysis: 241 Phasor analysis is a fit -free technique in which the fluorescence decay from each pixel is 242 transformed into a point in two-dimensional (2-D) phasor space.31,32 If P(i,j) represents a pixel 243 in the FLIM image with coordinates (i,j) and Ii,j(t) is the fluorescence intensity decay at that 244 pixel, the corresponding coordinates in the phasor plot (g,s) for time-domain measurements are 245 given as: 246 247 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 6 248 𝑔𝑖,𝑗(𝜔) = ∫ 𝐼𝑖,𝑗(𝑡) cos(𝜔𝑡) 𝑑𝑡 𝑇 0 ∫ 𝐼𝑖,𝑗(𝑡)𝑑𝑡 𝑇 0 249 250 𝑠𝑖,𝑗(𝜔) = ∫ 𝐼𝑖,𝑗(𝑡) sin(𝜔𝑡) 𝑑𝑡 𝑇 0 ∫ 𝐼𝑖,𝑗(𝑡)𝑑𝑡 𝑇 0 251 252 253 where ω = 2 πf and f = 1/T is the laser repetition rate. Phasor analysis provides a visual 254 distribution of the molecular species in an image by clustering pixels with similar lifetimes. 255 256 Recombinant Protein Production: 257 258 Recombinant HaloTag protein with a N -terminal 6× -His tag was purified using the E. coli 259 expression system, BL21(DE3) cells. Competent bacteria were transformed with pET28 -Halo 260 plasmid, and protein expression was induced by adding 500µM IPTG. After 4h of culturing the 261 cells were centrifuged and the pellet was conserved at -80°C. For the purification the cells were 262 lysed by ultrasonication : 2sec ON, 2sec OFF (1 min/mg of dry pellet) , 17W/wave, in Lysis 263 buffer (20 mM Tris -HCl, pH 7.5, 150 mM NaCl, 10 mM imidazole, mM β-mercaptoethanol, 264 1mM PMSF and a protease inhibitor mixture (Roche Diagnostics), then centrifuged at 20000 265 g, 4°C during 45 minutes. The su pernatant was filtered with 0,22 µ m low binding filters and 266 loaded on Ni-NTA agarose column (Qiagen) beforehand equilibrated with Equilibration Buffer 267 (20 mM Tris-HCl, pH 7.5, 500 mM NaCl, 15mM imidazole). HaloTag protein was eluted with 268 Ni-Elution buffer (20 mM Tris, pH 7.5, 500 mM NaCl and 1000 mM imidazole). Elution from 269 Ni-NTA was concentrate d to 2mL (Amicon Ultra 4, 10K, Millipore) and loaded into size 270 exclusion chromatography column (SEC) beforehand equilibrated with SEC Equilibration 271 Buffer (10 mM Tris, pH 7.5,150mM NaCl, 2mM DTT). Fractions containing Halo protein were 272 pooled, concentrated to ~ 50 M and aliquots were flash frozen in liquid nitrogen and stored at 273 −80°C. 274 275

276 277 BODIPY derived molecular rotors are sensitive to viscosity 278 279 BODIPY-based molecular rotors display absorption and emission maxima centered at 480 nm 280 and 520 nm respectively. The de -excitation pathway of these molecules depends on 281 conformational flexibility and the rotational freedom of their phenyl moiety (Figure 1A, B). In 282 viscous environments, where molecular mobility is restricted, radiative relaxation is favored, 283 resulting in an increase in both fluorescence quantum yield and lifetime. Hence, we first 284 characterized the viscosity response with water/glycerol mixture of varying viscosity by 285 measuring the fluorescence quantum yield ( ΦF) of the B ODIPY-COOH (BOD-COOH) 286 derivative. The viscosities ( η) of the mixtures were systematically determined using an 287 oscillatory rheometer. The relationship between τ, ΦF, and η follows the Förster –Hoffmann 288 model as previously described: 289 290 𝜙 = 𝐴 ∙ 𝜂𝛼 (1) 291 292 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 7 where A is a constant and α reflects the sensitivity of the molecular rotor to viscosity. The 293 viscosity values ranged from 1 mPa.s for pure water to 506 mPa.s for 95% glycerol at 20°C. As 294 shown in figures 1C and 1E, the fluorescence quantum yield increased with environmental 295 viscosity, following the Förster–Hoffmann relationship with α values of 0.71 ± 0.02. We next 296 measured the time resolved fluorescence decays of the same water/glycerol mixtures. In line 297 with the quantum yield measurements, upon increase of the viscosity, the fluorescence lifetime 298 of the BOD-COOH derivative increased (figures 1D and 1F). To account for the variation of 299 the lifetime as a function of the viscosity we used a model previously described by Vysniauskas 300 et al.: 33 301 302 𝜏 = 1 𝐴 𝐵 ∙ 𝜂𝛼 + 𝐶 + 𝐷 (2) 303 304 where A, B, C and D are unconstrained parameters that account for intrinsic lifetime measured 305 at zero viscosity, the radiative lifetime of the dye (without non -radiative relaxation pathway) 306 and the activation energy associated to the rotation of the phenyl moiety. α reflects the 307 sensitivity of the molecular rotor to viscosity and was fixed to the value obtained from the 308 quantum yield analysis. 309 310 311 312 313 Figure 1: A). Rotational mechanism and structure of BODIPY-based molecular rotors, B) Structure of BOD-314 COOH C) The fluorescence emission spectra of BOD-COOH and D) Fluorescence decay curves of BOD-COOH 315 in water/glycerol mixtures with increasing viscosities E-F) Foster-Hoffmann plots of fluorescence quantum yield 316 and lifetime vs. viscosity (blue squares correspond to the data points and the continuous red lines to the fit 317 obtained using equations 1 and 2). 318 319 To specifically target the BODIPY rotor to components of MLOs, we employed the HaloTag 320 labeling strategy. A chloroalkane HaloTag ligand was directly conjugated to the fluorophore or 321 linked via polyethylene glycol (PEG) spacers of four or twelve units (Figure 2A). The PEG 322 linkers increase the distance between the BODIPY rotor and the binding site within the HaloTag 323 binding pocket, t hereby minimizing the influence of the protein environment on viscosity 324 sensing. The time-resolved fluorescence decays of BOD-L, BOD-PEG4-L and BOD-PEG12-L 325 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 8 were measured to retrieve their fluorescence lifetimes as a function of the viscosity of 326 water/glycerol mixtures without (figure 2B) and in the presence of a n excess of the purified 327 HaloTag protein (figure 2C). As shown in Figure 2B, all three BODIPY derivatives retained 328 viscosity sensitivity ( 𝛼𝐵𝑂𝐷−𝐿 = 0.71 ± 0.04, 𝛼𝐵𝑂𝐷−𝑃𝐸𝐺4−𝐿 = 0.64 ± 0.03, 𝛼𝐵𝑂𝐷−𝑃𝐸𝐺12−𝐿 =329 0.65 ± 0.04). Notably the viscosity sensitivity of BOD-L upon binding to HaloTag protein was 330 lost (Figure 2C), indicating that the fluorophore remained inserted within the HaloTag binding 331 pocket and was insufficiently exposed to the surrounding medium (𝛼𝐵𝑂𝐷−𝐿+𝐻𝑎𝑙𝑜𝑇𝑎𝑔 = 0.07 ±332 0.01). In contrast, both B OD-PEG4-L and B OD-PEG12-L exhibited similar viscosity -333 dependent fluorescence responses, although their sensitivity was slightly reduced compared to 334 the unbound fluorophores ( 𝛼𝐵𝑂𝐷−𝑃𝐸𝐺4−𝐿+𝐻𝑎𝑙𝑜𝑇𝑎𝑔 = 0.29 ± 0.02, 𝛼𝐵𝑂𝐷−𝑃𝐸𝐺12−𝐿+𝐻𝑎𝑙𝑜𝑇𝑎𝑔 =335 0.35 ± 0.05). This reduction suggests that rotational restriction imposed by the protein 336 environment partially affects the fluorophore’s relaxation. 337 338 339 340 341 Figure 2: (A) Structures of BOD-L, BOD-PEG4-L and BOD-PEG12-L derivatives.( B) Fluorescence lifetimes 342 measured for the BODs alone (50nM) (individual points correspond to data points and continuous line to the fit 343 obtained using equation 2) and (C) in presence of HaloTag protein (0.5µM) (individual points correspond to 344 data points and continuous line to the fit obtained using equation 2). For both conditions, the alpha values are 345 reported in the main text.( D) HEK 293T cells expressing NPM-HaloTag labelled with BOD chloralkane 346 derivatives. BOD-L labels specifically the cell nucleolus, however a slight cytoplasmic signal is also present. 347 Cells labelled with BOD-PEG4-L show only specific probe binding to the NPM-HaloTag. For BOD-PEG12-L 348 no fluorescence was detected in the cells. 349 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 9 We next evaluated the three BODIPY probes for their ability to label HaloTag fusion proteins 350 in living cells. HEK293 cells expressing the nucleolar protein Nucleophosmin -1 fused to 351 HaloTag (NPM-HaloTag) were labeled and imaged by confocal microscopy. BOD-L produced 352 a strong fluorescence signal within the nucleolus but also exhibited substantial non -specific 353 staining in the cytoplasm. In contrast, BOD -PEG4-L selectively labeled the nucleolus with 354 minimal background fluorescence. No significant intrac ellular signal was detected for the 355 BOD-PEG12-L derivative, which likely failed to penetrate the cells due to the increased 356 hydrophilicity conferred by the long PEG12 linker. In light of these results, all cellular 357 experiments in this study were performed with BOD-PEG4-L. 358 BOD-PEG4-L senses the viscosity in various cellular compartments 359 360 To assess the sensitivity of the BOD -PEG4-L rotor to microviscosity within the cellular 361 environment, different subcellular structures in U2OS cells were labeled by expressing specific 362 HaloTag-fusion proteins: free cytoplasmic HaloTag protein, HaloTag-H2B in the cell nucleus, 363 LifeAct–targeting HaloTag to actin filaments, and HaloTag –hFis1 in mitochondria. FLIM 364 microscopy revealed distinct fluorescence lifetimes for each compartment. Based on the 365 calibration curve reported in figure 2C, t he average lifetime measured in the cytosol ( τcytosol = 366 1.5 ns) revealed a low microviscosity (2.4 mPa.s). 367 Higher microviscosities were de tected in the chromatin ( τchromatin = 2.3 ns, 10 mPa .s) and the 368 actin cytoskeleton (τactin = 2.5 ns, 14 mPa.s). The longest fluorescence lifetime was observed in 369 hFis1 protein located in the mitochondrial outer membrane ( τmitochondria = 3.5 ns, 71 mPa.s), 370 indicating a more viscous environment. These values are in good agreement with previously 371 reported viscosity estimations. 34,35 372 Altogether these data confirm that the sensitivity of BOD-PEG4-L probe is conserved in cells 373 and that FLIM imaging reports on the microviscosity of the environment in the proximity of 374 the labelled proteins in various cellular organelles. 375 376 377 378 Figure 3: BOD-PEG4-L is sensitive to environment in various cellular compartments: (A) FLIM images of cells 379 expressing free Halo Tag protein, HaloTag-H2B fusion (chromatin labelling), HaloTag- Life-Act to label the actin 380 fibers and mitochondrial protein HaloTag-hFis1 (B). Corresponding Lifetime Distributions. 381 382 383 384 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 10 Mapping the microviscosity in different sub-compartments of the nucleolus. 385 386 BOD-PEG4-L tool was next used to map the viscosity properties of the cell nucleolus. 387 Nucleolus is a 1 -2 µm sized MLO ,36 it is implicated in the ribosomal biogenesis, cell cycle 388 regulation and in the cellular stress response. Nucleolus is composed of three distin ct 389 compartments: the fibrillar centers (FC) - sites of rRNA transcription by RNA Pol 1, dense 390 fibrillar component (DFC), where the rRNA are processed and the granular component (GC) 391 being a site of the pre-ribosomal sub-units assembly.37 392 393 394 395 Figure 4: BOD-PEG4-L senses the microviscosity in nucleolar sub-compartments (A) Intensity images, phasor 396 plot and phasor based ROI analysis of HEK293 cells expressing NPM-HaloTag or Fib-HaloTag labelled with 397 BOD-PEG4-L. Phasor plot displays two populations with shorter lifetimes corresponding to the dye in the 398 nucleoplasm and longer lifetimes for the dye present in the nucleolus. (B) FLIM images of HEK293 and U2OS 399 cells expressing NPM-HaloTag and Fib-HaloTag. (C) Average lifetimes  SD measured for 15-30 cells in each 400 condition measured in 3 independent experiments, *** p<0.001. 401 402 To monitor the microviscosity in the nucleolar s ub-compartments, two nucleolar protein 403 Nucleophosmin-1 (NPM) and Fibrillarin (Fib) , were fused to HaloTag and expressed in HEK 404 and U2OS cells. NPM localizes mainly in the GC, while F ib is preferentially located in the 405 DFC. FLIM images were analyzed by decay fitting and also by phasor plot analysis. The phasor 406 approach offers a fit -free, intuitive way to analyze lifetime data by converting the measured 407 fluorescence decay in each pixel of the image, into a point (vector) in a 2D plot using a Fourier 408 transform at the modulation frequency of the excitation source (see Mat and Methods 409 section).31,32 In the resulting phasor plot the pixels are distributed on the semi -circle with the 410 shortest lifetimes plotted on the right side of the plot and longer lifetimes closer the origin. 32 411 Figure 3A represents Intensity images and corresponding phasor plots for NPM-HaloTang and 412 Fib-HaloTag proteins expressed in HEK293. The phasor plots for each protein show elongated 413 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 11 shape enabling to select the pixels of two populations with different lifetimes (Figure 4A, 414 Phasor ROI and Phasor p lot insets) . When displaying corresponding pixels on the FLIM 415 images, these tw o, populations represent NPM or Fib proteins located in the nucleoplasm 416 (shorter lifetime) and in the nucleolus (longer lifetime). 417 BOD-PEG4-L bound to the nucleolar NPM-HaloTag showed a lifetime NPM=2.10± 0.06 ns, 418 corresponding to 7 mPa .s. The lifetime of the probe bound to the fraction located in the 419 nucleoplasm was NPM = 1.88±0.06 ns corresponding to the viscosity of 4.7 mPa .s indicating 420 different viscosities in both locations . Interestingly the environment within the DFC seems to 421 be more viscous than GC , lifetimes measured for Fib-HaloTag are significantly higher 422 compared to NPM ( Fib =2.60 ±0.04, 16 mPa. s in DFC and Fib =2.42 ±0.05, 12 mPa .s in the 423 nucleoplasm). 424 These trends were reproducible in two cell lines tested HEK293 and U2OS (see Table 1). 425 426 Cell line GC (ns)  mPa.s GC (ns) +ActD  mPa.s DFC (ns)  mPa.s DFC (ns) +ActD  mPa.s HEK293 1.97±0.17 5.5 2.11±0.10 7 2.43±.18 12 2.62±0.25 16 U2OS 2.14±0.14 7.5 2.28±0.12 9.5 2.46±0.14 13 2.65±0.18 18 Table 1: Mean fluorescence lifetimes ± SD measured in nucleolar sub compartments and corresponding 427 viscosity ) values. Values based on 3 independent experiments with total 15-30 cells analyzed in each 428 condition 429 430 Effect of rRNA transcription inhibition on nucleolar microviscosity 431 432 To test the sensitivity of B OD-PEG4-L probe to the viscosity modifications induced by the 433 changes of the composition of the nucleolar sub -compartments, we performed the FLIM 434 imaging of cells treated with Actinomycin D (ActD) an inhibitor of RNA polymerase I. In the 435 treated nucleoli the DFCs fused together and progressively migrated to the nucleolar periphery, 436 where they formed structures called “nucleolar caps”. Concomitantly, GC became smaller and 437 spherical showing clearly its liquid like properties (Figure 5A). RNA imaging performed after 438 2h of ActD treatment showed no presence of RNA in the nucleolar caps; however, a residual 439 RNA signal was still detectable in GCs (Figure 5B) . Therefore, the DFC changed from a 440 protein/RNA condensate to a purely protein -rich condensate, while the impa ct on the GC 441 composition was less pronounced. 442 To monitor the microviscosity changes accompanying this nucleolar reorganization, HEK293T 443 and U2OS cells expressing NPM-HaloTag or Fib-HaloTag labelled with B OD-PEG4-L were 444 treated with ActD. 445 Two-photon fluorescence lifetime imaging (FLIM) confirmed a clear morphological change in 446 the nucleolus upon drug treatment in both HEK293 AND U2OS cells. The lifetime 447 measurements in ActD treated cells revealed an increase of BOD-PEG4-L fluorescence lifetime 448 in both sub-compartments (Table 1 and Fig 5C and D). Fluorescence lifetime for NPM-HaloTag 449 increased by 0.14 ns and 0.15 ns in HEK 293 and U2OS cells respectively, corresponding to a 450 viscosity increase of around 2 mPa.s. Similarly, for the Fib-HaloTag an increase of 0.19 ns and 451 0.20 ns was measured which reflects the viscosity increase of 4 mPa.s. 452 This observation confirms that the reorganization of the condensate's internal structure from 453 RNA-containing to RNA-free domain has a direct impact on the internal environment sensed 454 by the molecular rotor. 455 456 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 12 457 458 459 Fig 5. (A) Confocal images of HEK 293T cells expressing NPM-HaloTag (JF657-Halo), Fib -SNAP (JF552 -460 SNAP). Act D treatment induces the separation of GC with DFC forming nucleolar caps on the surface. 461 (B) Confocal images of ActD treated HEK293 cells expressing NPM -HaloTag (JF657 -Halo) or Fib -HaloTag 462 (JF657-Halo) co-stained with Pyrinon Y (1 M, 15 min) (C) FLIM images HEK293 and U2OS cells expressing 463 NPM-HaloTag or Fib-HaloTag labelled with BOD-PEG4-L. (D) Mean fluorescence lifetimesSD. Values based 464 on 3 independent experiments with total 15-30 cells analyzed in each condition, ** for p<0.01. 465 466 Comparative Analysis of Nuclear and Cytoplasmic Condensates 467 468 Finally, we wanted to explore whether BOD-PEG4-L probe can sense the differences between 469 cellular condensates. To this aim we performed a FLIM imaging of BOD -PEG4-L in stress 470 granules (SG) in U2OS cells expressing Ras GTPase -activating protein-binding protein fused 471 to HaloTag (HaloTag-G3BP1). G3BP1 is a major protein orchestrating the assembly of the 472 SG.38 The latter are formed as a response of the cell to harsh environmental conditions such as 473 oxidative and thermal stress or a presence of pathogens , when the cell induces a translational 474 arrest in which cellular mRNA and RNA binding proteins concentrate in these liquid-like 475 droplets.39 476 U2OS cells expressing HaloTag -G3PB1 were treated with sodium arsenite to induce the 477 oxidative stress, labelled and imaged by FLIM. The average fluorescence lifetime measured in 478 the stress granules is 1.45 ns (2.1 mPa.s, Figure 6B), corresponding to values measured for free 479 HaloTag in the cell cytoplasm (Figure 3). The microviscosity in the SG was about 3.5 times 480 lower as compared to the cell nucleolus. This result indicates that the RNA/protein network in 481 the SG is sparse compared to the nucleolus and contains larger free spaces. 482 483 484 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 13 485 Figure 6: FLIM image of U2OS cells expressing HaloTag-G3BP1 labelled with BOD-PEG4-L. Cells were 486 treated with Sodium Arsenite (0.5mM, 1h) to induce the formation of stress granules. (B) Mean fluorescence 487 lifetimesSD measured for NPM-HaloTag and HaloTag-G3BP1. Values based on 3 independent experiments 488 with total 15-30 cells analyzed in each condition, (C) Confocal images of HEK293 cells expressing HaloTag-489 NPM labelled by JF650-Halo together with eGFP or eGFP3. U2OS cells expressing HaloTag-G3BP1 labelled 490 by JF650-Halo together with eGFP or eGFP3. Scale bars: 5 µm (D) Fluorescence intensity ratios of eGFP 491 signal inside vs. outside of both MLOs. 492 493 In order to verify this difference, U2OS cells were co -transfected with NPM-HaloTag or 494 HaloTag-G3BP1 together with plasmids coding for eGFP or an eGFP trimer (eGFP3). The 495 cells were imaged by confocal microscopy and the partition coefficient of eGFP and eGFP3 in 496 both MLOs was quantified by measuring the ratio of eGFP fluorescence intensity outside vs. 497 inside. Since eGFP monomer and trimer are not supposed to interact with the cellular structures, 498 their diffusion within the MLOs in the cell is limited only by their size relative to protein/RNA 499 meshwork density and permeability, hence the partition coefficient is a measure of permeability 500 of the MLOs for each protein. eGFP and eGFP3 we re found to be excluded from the 501 nucleolus. The partition coefficient value of eGFP in the nucleolus was 0.7 and decreased to 0.6 502 for eGFP3. This observation indicates that the nucleolus is a relatively dense condensate since 503 the monomeric eGFP with the hydrodynamic radius approaching 2.5 nm 40 is excluded. On the 504 contrary, both eGFP and eGFP3 penetrated stress granules. The partition coefficient was close 505 to unity for eGFP and decreased slightly to 0.9 for eGFP3 trimer, indicating the presence of 506 much larger accessible free spaces within the stress granules. 507 These observations are in full agreemen t with the BOD -PEG4-L based FLIM imaging, 508 indicating that compared to nucleolus, the stress granules are significantly less dense and less 509 viscous MLOs. 510 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 14

511 512 In this study, we developed a BODIPY -based molecular rotor functionalized with a Hal oTag 513 ligand to enable direct monitoring of microviscosity in various cellular environments including 514 membrane-less organelles (MLOs). To this end, we optimized the BODIPY rotor with a 515 HaloTag ligand separated by a PEG4 linker. This design provided high sen sitivity to 516 environmental viscosity, cell permeability and efficient, specific labelling of cellular targets. 517 518 Viscosity sensing using a HaloTag-Functionalized BODIPY Molecular Rotor 519 520 Using fluorescence lifetime imaging microscopy (FLIM), we first confirmed that the BOD -521 PEG4-L probe is sensitive to microviscosity across various cellular compartments. Our 522 measurements indicated relatively low viscosity in the cell cytoplasm ( ηcyt ≈ 5 mPa .s), in 523 agreement with previous fluorescence correlat ion spectroscopy (FCS) studies. 41 Interestingly, 524 this value is lower than that reported by Kuimova and collaborators us ing a BODIPY rotor 525 directly functionalized with a HaloTag ligand.15 The major difference between the two designs 526 lies in the presence of the PEG4 linker in our probe, which spatially separates the rotor from 527 the HaloTag protein. We therefore hypothesize that rotors lacking this linker may still be subject 528 to partial steric or environmental constraints imposed by the protein, leading to a n increased 529 apparent viscosity. The viscosity measured by the probe targeted to H 2B in the chromatin (= 530 10 mPa.s) and actin fibers ( =14 mPa.s) was higher than that of the cytoplasm. FCS studies 531 similarly reported reduced diffusion of monomeric eGFP in the nucleoplasm, by a factor of 532 approximately 3.2 relative to aqueous solution,35 corresponding to an estimated viscosity of 3–533 4 mPa.s. Our measurements corroborate the notion that chrom atin is embedded within a low -534 viscosity aqueous phase permissive to the diffusion of low -molecular-weight proteins. 535 Fluorescence lifetimes recorded in mitochondria ( τ ≈ 3.5 ns) were close to those reported by 536 Chambers et al.15, although viscosity values inferred from calibration curves differed. These 537 discrepancies likely arise from differences in the specific mitochondrial proteins targ eted 538 (matrix versus outer membrane), leading to distinct microenvironments. 539 540 Quantification of Microviscosity in Nucleolar Subcompartments 541 542 After validating the viscosity sensitivity of the BOD-PEG4-L probe in diverse cellular contexts, 543 we used this approach to quantify microviscosity within cellular MLOs. For both nucleolar 544 proteins examined, we found that the nucleolus exhibits significantly higher viscosity than the 545 surrounding nucleoplasm. Moreover, we observed clear differences between nucleolar 546 subcompartments, the dense fibrillar component (DFC) displayed higher viscosity than the 547 granular component (GC). Feric et al . showed by coalescen ce and FRAP measurements a 548 liquid-like behaviour for GC and more visco elastic behaviour for DFC in NPM and Fib 549 reconstituted condensates and in cell nucleoli .42 Our results confirm the distinct viscosity 550 properties among nucleolar subdomains, underscoring the high sensitivity of BOD-PEG4-L to 551 its local environment. Overall viscosity values ranged from ~5 –8 mPa.s in the GC to ~12 –14 552 mPa.s in the DFC and appeared in dependent of the cell line examined. Our findings are 553 consistent with nano scale techniques such as FCS measurements of GFPs diffusion in the 554 nucleolus, which report diffusion coefficients reflecting a viscosity range of 4–26 mPa.s.13,41,43 555 However these values are considerably lower than viscosity measurements reported for in vitro 556 reconstituted condensates ~700 mPa s.42 Such differences stem from the distinct physical scales 557 probed by each technique. Passive microrheology relies on tracking of fluorescent beads that 558 are hundreds of nanometers in diameter. In these conditions, the beads diffusion is sensitive to 559 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 15 the RNA–protein meshwork and long-range interactions that remain inaccessible to molecular-560 scale probes. 561 Finally, it should be noted that all measurements in this study were performed in asynchronous 562 cell populations. Since the cell nucleolus disassembles before the cell division and reassembles 563 at the end of the mitosis,44,45 it is highly probable that the nucleolar viscosity varies during the 564 cell cycle, this question should be considered in future work. 565 566 Transcriptional Stress and Nucleolar Remodelling 567 568 We further observed an increase in viscosity within both nucleolar compartments following 569 inhibition of rRNA transcription with ActD. ActD treatment induces marked nucleolar 570 reorganization: nucleolar volume decreases the GC adopts a rounded mo rphology and DFCs 571 coalesce into nucleolar caps.13 These changes accompany the release of nucleolar proteins from 572 their molecular complexes and alterations in their interactions with rRNA, leading to an 573 increase in protein mobility and a partial reloca lization to the nucleoplasm. 46 In RNA -rich 574 condensates, nucleic acids act as a polymer scaffold interconnecting p roteins within a 575 meshwork containing relatively large solvent-filled spaces. Reduction of RNA content leads to 576 smaller, denser condensates , 14 increasing the probability of rotor –protein collisions. 577 Consequently, viscosity measured in ActD-treated nucleoli is higher than in untreated, rRNA -578 rich nucleoli. Consistent with our observations, increases in nucleolar density following ActD 579 treatment have been reported using optical diffraction tomography (ODT) and FCS. 13,14 580 581 Microviscosity of Stress Granules Reveals a Low-Density RNA–Protein Network 582 583 Finally, we compared nucleolar viscosity wi th that o f stress granules . Surprisingly, FLIM 584 measurements indicated that SGs exhibit viscosities close to that of the cytoplasm ( ηSG ≈ 2.1 585 mPa.s). This finding suggests that SGs possess a considerably more open RNA –protein 586 meshwork than nucleolar compartments. Th is interpretation is supported by the high 587 partitioning of monomeric and trimeric eGFP into SGs. Our results align with ODT analyses 588 by Kim et al., who showed that SGs provide minimal refractive -index contrast relative to the 589 cytoplasm, indicating a low co ncentration of SG components. 14 In line, single -molecule 590 tracking of G3BP1 performed by Niewidok and co-workers, revealed the presence of immobile 591 dense nanocores (100–200 nm), surrounded by a more fluid phase in which proteins remain 592 highly mobile.27 Our results thus reinforce the view that SGs represent low-density cytoplasmic 593 condensates with a large RNA meshwork. 594 In conclusion, this study introduces a microscopy-based approach for measuring microviscosity 595 in living cells. The technique provides information co mplementary to existing methods for 596 probing the material properties of MLOs. Given the key importance of rheological properties 597 in the MLOs functions, the development of new analytical tools and experimental strategies is 598 essential for advancing our understanding of condensate biophysics. 599

600 This work was funded by the French National Research Agency (ANR) FluoLLPS ANR-23-601 CE11-0016 and IdEx Recherche Exploratoire , Unistra (R22086MM). The authors would like 602 to thank Imaging Center PIQ -QuESt ( https://piq.unistra.fr/) and Plateforme d’imagerie “In 603 Vitro”, members of the national infrastructure France -BioImaging supported by the French 604 National Research Agency (ANR-24-INBS-0005 FBI BIOGEN). 605 606 607 608 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 16

609 610 (1) Falahati, H.; Haji-Akbari, A. Thermodynamically Driven Assemblies and Liquid-611 Liquid Phase Separations in Biology. Soft Matter 2019, 15 (6), 1135–1154. 612 https://doi.org/10.1039/c8sm02285b. 613 (2) Hyman, A. A.; Weber, C. A.; Jülicher, F. Liquid-Liquid Phase Separation in Biology. 614 Annu. Rev. Cell Dev. Biol. 2014, 30 (1), 39–58. https://doi.org/10.1146/annurev-615 cellbio-100913-013325. 616 (3) Boeynaems, S.; Alberti, S.; Fawzi, N. L.; Mittag, T.; Polymenidou, M.; Rousseau, F.; 617 Schymkowitz, J.; Shorter, J.; Wolozin, B.; Van Den Bosch, L.; Tompa, P.; Fuxreiter, 618 M. Protein Phase Separation: A New Phase in Cell Biology. Trends Cell Biol. 2018, 28 619 (6), 420–435. https://doi.org/10.1016/j.tcb.2018.02.004. 620 (4) Banani, S. F.; Lee, H. O.; Hyman, A. A.; Rosen, M. K. Biomolecular Condensates: 621 Organizers of Cellular Biochemistry. Nat. Rev. Mol. Cell Biol. 2017, 18 (5), 285–298. 622 https://doi.org/10.1038/nrm.2017.7. 623 (5) Alberti, S.; Dormann, D. Liquid–Liquid Phase Separation in Disease. Annu. Rev. 624 Genet. 2019, 53 (1), 171–194. https://doi.org/10.1146/annurev-genet-112618-043527. 625 (6) Alshareedah, I.; Thurston, G. M.; Banerjee, P. R. Quantifying Viscosity and Surface 626 Tension of Multicomponent Protein-Nucleic Acid Condensates. Biophysical journal. 627 United States April 2021, pp 1161–1169. https://doi.org/10.1016/j.bpj.2021.01.005. 628 (7) Alshareedah, I.; Kaur, T.; Banerjee, P. R. Methods for Characterizing the Material 629 Properties of Biomolecular Condensates. Methods Enzymol. 2021, 646, 143–183. 630 https://doi.org/10.1016/bs.mie.2020.06.009. 631 (8) Català-Castro, F.; Ortiz-Vásquez, S.; Martínez-Fernández, C.; Pezzano, F.; Garcia-632 Cabau, C.; Fernández-Campo, M.; Sanfeliu-Cerdán, N.; Jiménez-Delgado, S.; 633 Salvatella, X.; Ruprecht, V.; Frigeri, P.-A.; Krieg, M. Measuring Age-Dependent 634 Viscoelasticity of Organelles, Cells and Organisms with Time-Shared Optical Tweezer 635 Microrheology. Nat. Nanotechnol. 2025, 20 (3), 411–420. 636 https://doi.org/10.1038/s41565-024-01830-y. 637 (9) Jawerth, L. M.; Ijavi, M.; Ruer, M.; Saha, S.; Jahnel, M.; Hyman, A. A.; Jülicher, F.; 638 Fischer-Friedrich, E. Salt-Dependent Rheology and Surface Tension of Protein 639 Condensates Using Optical Traps. Phys. Rev. Lett. 2018, 121 (25), 258101. 640 https://doi.org/10.1103/PhysRevLett.121.258101. 641 (10) Alshareedah, I.; Borcherds, W. M.; Cohen, S. R.; Singh, A.; Posey, A. E.; Farag, M.; 642 Bremer, A.; Strout, G. W.; Tomares, D. T.; Pappu, R. V; Mittag, T.; Banerjee, P. R. 643 Sequence-Specific Interactions Determine Viscoelasticity and Aging Dynamics of 644 Protein Condensates. Nat. Phys. 2024, 20 (9), 1482–1491. 645 https://doi.org/10.1038/s41567-024-02558-1. 646 (11) Taylor, N. O.; Wei, M.-T.; Stone, H. A.; Brangwynne, C. P. Quantifying Dynamics in 647 Phase-Separated Condensates Using Fluorescence Recovery after Photobleaching. 648 Biophys. J. 2019, 117 (7), 1285–1300. https://doi.org/10.1016/j.bpj.2019.08.030. 649 (12) Muzzopappa, F.; Hummert, J.; Anfossi, M.; Tashev, S. A.; Herten, D.-P.; Erdel, F. 650 Detecting and Quantifying Liquid-Liquid Phase Separation in Living Cells by Model-651 Free Calibrated Half-Bleaching. Nat. Commun. 2022, 13 (1), 7787. 652 https://doi.org/10.1038/s41467-022-35430-y. 653 (13) Kim, J. K.; Lee, B.-W.; Fujii, F.; Kim, J.; Pack, C.-G. Physicochemical Properties of 654 Nucleoli in Live Cells Analyzed by Label-Free Optical Diffraction Tomography. Cells 655 2019, 8, 699. https://doi.org/10.3390/cells8070699. 656 (14) Kim, T.; Yoo, J.; Do, S.; Hwang, D. S.; Park, Y.; Shin, Y. RNA-Mediated Demixing 657 Transition of Low-Density Condensates. Nat. Commun. 2023, 14 (1), 2425. 658 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 17 https://doi.org/10.1038/s41467-023-38118-z. 659 (15) Chambers, J. E.; Kubánková, M.; Huber, R. G.; López-Duarte, I.; Avezov, E.; Bond, P. 660 J.; Marciniak, S. J.; Kuimova, M. K. An Optical Technique for Mapping 661 Microviscosity Dynamics in Cellular Organelles. ACS Nano 2018, 12 (5), 4398–4407. 662 https://doi.org/10.1021/acsnano.8b00177. 663 (16) Ferri, G.; Nucara, L.; Biver, T.; Battisti, A.; Signore, G.; Bizzarri, R. Organization of 664 Inner Cellular Components as Reported by a Viscosity-Sensitive Fluorescent Bodipy 665 Probe Suitable for Phasor Approach to FLIM. Biophys. Chem. 2016, 208, 17–25. 666 https://doi.org/https://doi.org/10.1016/j.bpc.2015.06.009. 667 (17) Kubánková, M.; López-Duarte, I.; Kiryushko, D.; Kuimova, M. K. Molecular Rotors 668 Report on Changes in Live Cell Plasma Membrane Microviscosity upon Interaction 669 with Beta-Amyloid Aggregates. Soft Matter 2018, 14 (46), 9466–9474. 670 https://doi.org/10.1039/c8sm01633j. 671 (18) Wu, Y.; Stefl, M.; Olzyńska, A.; Hof, M.; Yahioglu, G.; Yip, P.; Casey, D. R.; Ces, O.; 672 Humpolíčková, J.; Kuimova, M. K. Molecular Rheometry: Direct Determination of 673 Viscosity in Lo and Ld Lipid Phases via Fluorescence Lifetime Imaging. Phys. Chem. 674 Chem. Phys. 2013, 15 (36), 14986–14993. https://doi.org/10.1039/c3cp51953h. 675 (19) Kashirina, A. S.; López-Duarte, I.; Kubánková, M.; Gulin, A. A.; Dudenkova, V. V; 676 Rodimova, S. A.; Torgomyan, H. G.; Zagaynova, E. V; Meleshina, A. V; Kuimova, M. 677 K. Monitoring Membrane Viscosity in Differentiating Stem Cells Using BODIPY-678 Based Molecular Rotors and FLIM. Sci. Rep. 2020, 10 (1), 14063. 679 https://doi.org/10.1038/s41598-020-70972-5. 680 (20) Wei, Y.; Gao, X.; Fang, J.; Xiao, Y.; Liu, J.; Liu, Y.; Zhang, X.; Shen, B. Tailoring the 681 PK(a) of Fluorescence Lifetime Imaging Probes to Visualize Aggrephagy and Resolve 682 Its Microenvironmental Viscosity. Anal. Chem. 2024, 96 (35), 14160–14167. 683 https://doi.org/10.1021/acs.analchem.4c02065. 684 (21) Los, G. V; Encell, L. P.; McDougall, M. G.; Hartzell, D. D.; Karassina, N.; Zimprich, 685 C.; Wood, M. G.; Learish, R.; Ohana, R. F.; Urh, M.; Simpson, D.; Mendez, J.; 686 Zimmerman, K.; Otto, P.; Vidugiris, G.; Zhu, J.; Darzins, A.; Klaubert, D. H.; Bulleit, 687 R. F.; Wood, K. V. HaloTag: A Novel Protein Labeling Technology for Cell Imaging 688 and Protein Analysis. ACS Chem. Biol. 2008, 3 (6), 373–382. 689 https://doi.org/10.1021/cb800025k. 690 (22) Saladin, L.; Breton, V.; Le Berruyer, V.; Nazac, P.; Lequeu, T.; Didier, P.; Danglot, L.; 691 Collot, M. Targeted Photoconvertible BODIPYs Based on Directed Photooxidation-692 Induced Conversion for Applications in Photoconversion and Live Super-Resolution 693 Imaging. J. Am. Chem. Soc. 2024, 146 (25), 17456–17473. 694 https://doi.org/10.1021/jacs.4c05231. 695 (23) Wang, W.; Budhu, A.; Forgues, M.; Wang, X. W. Temporal and Spatial Control of 696 Nucleophosmin by the Ran-Crm1 Complex in Centrosome Duplication. Nat. Cell Biol. 697 2005, 7 (8), 823–830. https://doi.org/10.1038/ncb1282. 698 (24) Chen, D.; Huang, S. Nucleolar Components Involved in Ribosome Biogenesis Cycle 699 between the Nucleolus and Nucleoplasm in Interphase Cells. J. Cell Biol. 2001, 153 700 (1), 169–176. https://doi.org/10.1083/jcb.153.1.169. 701 (25) Appelhans, T.; Richter, C. P.; Wilkens, V.; Hess, S. T.; Piehler, J.; Busch, K. B. 702 Nanoscale Organization of Mitochondrial Microcompartments Revealed by Combining 703 Tracking and Localization Microscopy. Nano Lett. 2012, 12 (2), 610–616. 704 https://doi.org/10.1021/nl203343a. 705 (26) Guillen-Chable, F.; Rodríguez Corona, U.; Pereira-Santana, A.; Bayona, A.; 706 Rodríguez-Zapata, L. C.; Aquino, C.; Šebestová, L.; Vitale, N.; Hozak, P.; Castano, E. 707 Fibrillarin Ribonuclease Activity Is Dependent on the GAR Domain and Modulated by 708 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 18 Phospholipids. Cells. 2020. https://doi.org/10.3390/cells9051143. 709 (27) Niewidok, B.; Igaev, M.; Pereira da Graca, A.; Strassner, A.; Lenzen, C.; Richter, C. 710 P.; Piehler, J.; Kurre, R.; Brandt, R. Single-Molecule Imaging Reveals Dynamic 711 Biphasic Partition of RNA-Binding Proteins in Stress Granules. J. Cell Biol. 2018, 217 712 (4), 1303–1318. https://doi.org/10.1083/jcb.201709007. 713 (28) Böhm, J.; Thavaraja, R.; Giehler, S.; Nalaskowski, M. M. A Set of Enhanced Green 714 Fluorescent Protein Concatemers for Quantitative Determination of Nuclear 715 Localization Signal Strength. Anal. Biochem. 2017, 533, 48–55. 716 https://doi.org/10.1016/j.ab.2017.06.015. 717 (29) Pelletier, R.; Danylchuk, D. I.; Benaissa, H.; Broch, F.; Vauchelles, R.; Gautier, A.; 718 Klymchenko, A. S. Genetic Targeting of Solvatochromic Dyes for Probing Nanoscale 719 Environments of Proteins in Organelles. Anal. Chem. 2023, 95 (22), 8512–8521. 720 https://doi.org/10.1021/acs.analchem.3c00515. 721 (30) Tomishige, N.; Bin Nasim, M.; Murate, M.; Pollet, B.; Didier, P.; Godet, J.; Richert, 722 L.; Sako, Y.; Mély, Y.; Kobayashi, T. HIV-1 Gag Targeting to the Plasma Membrane 723 Reorganizes Sphingomyelin-Rich and Cholesterol-Rich Lipid Domains. Nat. Commun. 724 2023, 14 (1), 7353. https://doi.org/10.1038/s41467-023-42994-w. 725 (31) Digman, M. A.; Caiolfa, V. R.; Zamai, M.; Gratton, E. The Phasor Approach to 726 Fluorescence Lifetime Imaging Analysis. Biophysical journal. United States January 727 2008, pp L14-6. https://doi.org/10.1529/biophysj.107.120154. 728 (32) Ranjit, S.; Malacrida, L.; Jameson, D. M.; Gratton, E. Fit-Free Analysis of 729 Fluorescence Lifetime Imaging Data Using the Phasor Approach. Nat. Protoc. 2018, 730 13 (9), 1979–2004. https://doi.org/10.1038/s41596-018-0026-5. 731 (33) Vyšniauskas, A.; López-Duarte, I.; Duchemin, N.; Vu, T.-T.; Wu, Y.; Budynina, E. M.; 732 Volkova, Y. A.; Peña Cabrera, E.; Ramírez-Ornelas, D. E.; Kuimova, M. K. Exploring 733 Viscosity, Polarity and Temperature Sensitivity of BODIPY-Based Molecular Rotors. 734 Phys. Chem. Chem. Phys. 2017, 19 (37), 25252–25259. 735 https://doi.org/10.1039/C7CP03571C. 736 (34) Yang, Z.; He, Y.; Lee, J.-H.; Park, N.; Suh, M.; Chae, W.-S.; Cao, J.; Peng, X.; Jung, 737 H.; Kang, C.; Kim, J. S. A Self-Calibrating Bipartite Viscosity Sensor for 738 Mitochondria. J. Am. Chem. Soc. 2013, 135 (24), 9181–9185. 739 https://doi.org/10.1021/ja403851p. 740 (35) Liang, L.; Wang, X.; Xing, D.; Chen, T.; Chen, W. R. Noninvasive Determination of 741 Cell Nucleoplasmic Viscosity by Fluorescence Correlation Spectroscopy. J. Biomed. 742 Opt. 2009, 14 (2), 24013. https://doi.org/10.1117/1.3088141. 743 (36) Lafontaine, D. L. J.; Riback, J. A.; Bascetin, R.; Brangwynne, C. P. The Nucleolus as a 744 Multiphase Liquid Condensate. Nat. Rev. Mol. Cell Biol. 2021, 22 (3), 165–182. 745 https://doi.org/10.1038/s41580-020-0272-6. 746 (37) Dubois, M.-L.; Boisvert, F.-M. The Nucleolus: Structure and Function. The Functional 747 Nucleus. April 2016, pp 29–49. https://doi.org/10.1007/978-3-319-38882-3_2. 748 (38) Yang, P.; Mathieu, C.; Kolaitis, R.-M.; Zhang, P.; Messing, J.; Yurtsever, U.; Yang, Z.; 749 Wu, J.; Li, Y.; Pan, Q.; Yu, J.; Martin, E. W.; Mittag, T.; Kim, H. J.; Taylor, J. P. 750 G3BP1 Is a Tunable Switch That Triggers Phase Separation to Assemble Stress 751 Granules. Cell 2020, 181 (2), 325-345.e28. https://doi.org/10.1016/j.cell.2020.03.046. 752 (39) Smith, J.; Bartel, D. P. The G3BP Stress-Granule Proteins Reinforce the Integrated 753 Stress Response Translation Programme. Nat. Cell Biol. 2025. 754 https://doi.org/10.1038/s41556-025-01834-3. 755 (40) Terry, B. R.; Robards, A. W. Hydrodynamic Radius Alone Governs the Mobility of 756 Molecules through Plasmodesmata. Planta 1987, 171 (2), 145–157. 757 https://doi.org/10.1007/BF00391090. 758 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint 19 (41) Park, H.; Han, S.-S.; Sako, Y.; Pack, C.-G. Dynamic and Unique Nucleolar 759 Microenvironment Revealed by Fluorescence Correlation Spectroscopy. FASEB J. 760 Off. Publ. Fed. Am. Soc. Exp. Biol. 2015, 29 (3), 837–848. 761 https://doi.org/10.1096/fj.14-254110. 762 (42) Feric, M.; Vaidya, N.; Harmon, T. S.; Mitrea, D. M.; Zhu, L.; Richardson, T. M.; 763 Kriwacki, R. W.; Pappu, R. V; Brangwynne, C. P. Coexisting Liquid Phases Underlie 764 Nucleolar Subcompartments. Cell 2016, 165 (7), 1686–1697. 765 https://doi.org/10.1016/j.cell.2016.04.047. 766 (43) Pack, C.; Saito, K.; Tamura, M.; Kinjo, M. Microenvironment and Effect of Energy 767 Depletion in the Nucleus Analyzed by Mobility of Multiple Oligomeric EGFPs. 768 Biophys. J. 2006, 91 (10), 3921–3936. https://doi.org/10.1529/biophysj.105.079467. 769 (44) Hernandez-Verdun, D. Assembly and Disassembly of the Nucleolus during the Cell 770 Cycle. Nucleus 2011, 2 (3), 189–194. https://doi.org/10.4161/nucl.2.3.16246. 771 (45) Leung, A. K. L.; Gerlich, D.; Miller, G.; Lyon, C.; Lam, Y. W.; Lleres, D.; Daigle, N.; 772 Zomerdijk, J.; Ellenberg, J.; Lamond, A. I. Quantitative Kinetic Analysis of Nucleolar 773 Breakdown and Reassembly during Mitosis in Live Human Cells. J. Cell Biol. 2004, 774 166 (6), 787–800. https://doi.org/10.1083/jcb.200405013. 775 (46) Phair, R. D.; Misteli, T. High Mobility of Proteins in the Mammalian Cell Nucleus. 776 Nature 2000, 404 (6778), 604–609. https://doi.org/10.1038/35007077. 777 778 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 5, 2026. ; https://doi.org/10.64898/2026.01.05.697638doi: bioRxiv preprint