A Novel Fluorogenic Probe Reveals Lipid Droplet Dynamics in ME/CFS Fibroblasts

Lipid droplets (LDs) are dynamic cellular organelles that play an essential role in lipid metabolism and storage. LD dysregulation has been implicated in various diseases. However, investigations into the cellular LD dynamics under disease conditions have been rarely reported, possibly due to the absence of high performing LD imaging agents. Here a novel fluorogenic probe, AM‐QTPA, is reported for specific LD imaging. AM‐QTPA demonstrates viscosity sensitivity and aggregation‐induced emission enhancement characteristics. It is live cell permeable and can specifically light up LDs in cells, with low background noise and superior signals that can be quantified. After validation in cell model with LD accumulation induced by oleic acid treatment, AM‐QTPA is applied in a small proof‐of‐concept number of human fibroblast samples derived from people diagnosed with myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS), a complex and debilitating disease with unknown cause. The results indicate the presence of larger but fewer LDs in ME/CFS fibroblasts compared to the healthy counterparts, accompanying with frequent LD‐mitochondria contacts, suggesting potential upregulation of lipolysis in ME/CFS connective tissue like fibroblasts. Overall, AM‐QTPA provides new understanding of the anomalous LD dynamics in disease status, which, potentially, will facilitate in‐depth investigation of the pathogenesis of ME/CFS.


Introduction
Lipid droplets (LDs), also known as lipid bodies or adiposomes, are dynamic cellular organelles that play a crucial role in lipid metabolism and storage. [1,2]These structures, often spherical DOI: 10.1002/adsr.202300178 in shape, consist of a hydrophobic core of neutral lipids, primarily triglycerides, and sterol esters, surrounded by a phospholipid monolayer embedded with various proteins. [1]Originally perceived as passive lipid reservoirs, LDs are now recognized as highly dynamic and functionally versatile entities involved in a range of cellular processes, [3] encompassing key roles in energy metabolism, [1] cellular stress response, [4,5] as well as disease progressions. [6,7]LDs have been implicated in various diseases, including obesity [8] and metabolic disorders, [9,10] fatty liver disease, [11] cardiovascular diseases, [12,13] infections [14][15][16] and neurodegenerative disorders. [17,18]However, the linkage between fluctuations of LD dynamics and disease progressions remains missing.
Hence, precisely monitoring the dynamic changes of LDs is of particular importance in biomedical research and for early diagnosis of the associated diseases, which includes not only the localization of LDs but also quantitative information such as number and size of LDs in cells and other physiochemical properties such as viscosity, pH, and polarity of LDs.[21] SRS microscopy allows for the highly specific detection of vibrations of lipid molecules and does not perturb the samples without the requirement of pre-labeling.However, SRS microscopy has limit accessibility for some laboratories due to the complexity and high cost of the equipment.In addition, there are some interferences from other components in the complex biological system, making it challenging to isolate the signal of interest.On the other hand, fluorescence imaging techniques represent one of the most common approaches to visualize LDs in real-time within cells, owing to their high spa-tial resolution, real-time, and non-invasive advantages.Several commercially available LD dyes have been widely used in biological research, including Nile red, BODIPY and Oil Red O.[24][25] Meanwhile, BODIPY has a small Stokes shift which delimits its applicability, [26] whilst Oil Red O is impermeable for live contexts and hence requires tedious sample processing. [27,28]The most prominent drawback of conventional fluorophores is the concentration-caused quenching effect, [29] particularly in scenarios where high concentrations of molecular dyes accumulated in LDs.This may cause inaccuracy in using fluorescence intensity for quantitative analysis and misinterpretation in practical applications, especially under a complicated disease condition.
[53] While there are observations and pilot studies highlighting abnormal lipid metabolism in samples from ME/CFS patients, [54][55][56][57][58] it remains unclear whether this phenomenon is a cause or consequence of the pathology.57 ] Studies of lipids or LDs in clinical samples from ME/CFS patients have been mostly restricted to biofluids or immune cells but yet expanded to connective tissue, for example, fibroblasts.In this work, we design and synthesize an AIEE-active fluorophore, AM-QTPA, composing of a diphenylamine as the electron donor (D) group and a quinoline as the electron acceptor (A) group (Figure 1A).AM-QTPA exhibits AIEE properties with a decent Stokes shift and can penetrate into live cells within a 30 min staining period.Using the oleic acid (OA) treatment as a model condition, we validate the quantitative capability of AM-QTPA in LD accumulation.We further apply it in fibroblasts derived from people with ME/CFS.Aided by AM-QTPA, systematic investigation has revealed differences in LD dynamics within a small preliminary cohort of ME/CFS versus healthy control fibroblasts, including LD number and size, and spatial mitochondrial interactions, suggesting LD dysregulations in ME/CFS connective tissue.

Synthesis and Photophysical Characterization of AM-QTPA
AM-QTPA was synthesized according to the synthetic route shown in Scheme S1 (Supporting Information).Briefly, Knoevenagel condensation between 4-(diphenylamino)benzaldehyde and 2-methyl-8-nitroquinoline yielded NO 2 -QTPA, which was subsequently reduced by iron to afford the final product AM-QTPA.The chemical structures of products were confirmed by NMR spectroscopies and high resolution mass spectrometry (HRMS) (Figures S8-S13, Supporting Information).AM-QTPA is completely soluble in common organic solvents such as CH 2 Cl 2 , THF, acetonitrile, DMSO, and DMF, but poorly soluble in water.UV-vis spectra of AM-QTPA show the absorption maxima respectively at 395 and 404 nm, in DMSO and water, matching the channel of 405 nm laser on most commercial optical microscopes (Figure 1B).DFT calculation using B3LYP/6-31+g(d,p) level suggests HOMO is mainly localized on the electron donating triphenylamino group and the amine group on the quinoline while LUMO is mainly on the quinoline and central C═C bond, with an energy gap of 3.18 eV, matching the absorption maximum observed (Figure 1C).With the emission maximum at 559 nm, the Stokes shift of AM-QTPA is calculated as 165 nm in DMSO, which is much larger compared to BODIPY and Nile red.AM-QTPA shows low absorption and emission in the aqueous environment (1% DMSO in water), giving a low background that is potentially more favorable for the applications in a cellular context.AM-QTPA possesses a positive solvatochromism, with emission maximum at 519 nm in non-polar toluene to 559 nm in more polar DMSO (Figure 1D; Figure S1, Supporting Information).Furthermore, fluorescence of AM-QTPA was found to be viscosity sensitive.In the glycerol-ethylene glycol cosolvent system, increasing the viscosity by higher fraction of glycerol results in the enhancement of the fluorescence intensity of AM-QTPA (Figure 1E).Specifically, AM-QTPA in 90% glycerol emits approximately eightfold higher than in pure ethylene glycol (0% glycerol).This is ascribed to the effect of restriction of intramolecular motions (RIM) in high viscous environment.To further validate the RIM effect, we tested the fluorescence response of AM-QTPA upon aggregation (Figure 1F-H).AM-QTPA is faintly fluorescent in EtOH but becomes emissive in EtOH-H 2 O mixtures when the water fraction (f water ) is higher than 60%, demonstrating a characteristic AIEE phenomenon.This conclusion is also supported by its bright yellow fluorescence in the solid state, as demonstrated by the photo taken under UV illumination (Figure 1F).

AM-QTPA Specifically Lights up Lipid Droplets in Cells
With the photophysical data in hand, we then proceeded to explore the applications of AM-QTPA in cell imaging.The cytotoxicity assay showed that AM-QTPA posed no significant toxicity toward cells at concentrations of up to 10 μM and incubation times as long as 24 h (Figure S2, Supporting Information).Next, it was found that AM-QTPA can readily enter live cells spontaneously within a 30 min staining period, without the necessity of fixation nor permeabilization.Z stack projection of HeLa cells stained with AM-QTPA showed distinctly bright spherical objects in cytosol, which were identified as LDs, evidenced by the colocalization with signals from Nile red.To further validate this observation, we performed the same counter staining of AM-QTPA and Nile red in HeLa cells treated with OA which is known to stimulate the production of LDs in cells. [59]Figure 2C shows that upon 6-h OA treatment, the number of LDs per cell is significantly increased, accompanying with the enlargement of the LD size.Plot profiling of the linear region of interest (ROI) (Figure 2B,D) demonstrates the excellent colocalization between AM-QTPA and Nile red in cells with or without OA treatment, with the Pearson's correlation coefficients of 0.86 and 0.80, respectively.Nevertheless, Nile red staining shows relatively higher background than AM-QTPA in the situation of low LD number (Figure 2B).Whereas in some LDs from the OA-treated group, Nile red signals seem receded due to its concentration-caused quenching effect at high concentration. [60]n OA treatment, LDs are originated and form on endothelium reticulum (ER) through membrane budding.At this stage, immature LDs remain connected to the ER membrane. [1,47]Aided by super-resolution airy scan imaging, the close contact between immature LDs, stained by AM-QTPA, and ER, stained by ER-Tracker Red, can be observed (Figure S3, Supporting Information).Furthermore, counterstaining of AM-QTPA with dyes for plasma membrane (MemBright 560) and cell nucleus (DRAQ5) affords the Z stack image showing the spatial distribution of LDs in the OA-treated cell (Figure S4, Supporting Information).

Quantification of Cellular LDs Induced by OA
We next examine the quantification capability of AM-QTPA using OA-treated HeLa cells as a model.Using spinning disk confocal microscope, we can then quantify the number and size of LDs in cells treated by OA for varying durations.Note that using the high-speed Z stack acquisition enables the visualization of the total LDs in cells rather than on a single frame, which guarantees the spatial accuracy of our approach.To specifically quantify LDs in single cells, a robust plasma membrane marker, PM-ML, was used to draw an outline for each cell (Figure 3A-E). [61]After acquisition, the information of LDs in the Z stack images was extracted using FIJI.Results showed that the number of LDs increases strikingly even after only 1-h OA treatment and remains barely changed in the 1-12 h timepoints (Figure 3F).However, the size of LDs keeps increasing until 6 h to reach its maximum (Figure 3G).These data demonstrate the ability of AM-QTPA for studying and quantifying LD dynamics in cells.For the ease of operation, we fixed cells for handling of large number of samples and found the signals of AM-QTPA retained, though this phenomenon was first observed in live cells (Figure S5, Supporting Information).

Quantification of Cellular LDs in Fibroblasts Derived from ME/CFS Patients and Healthy Controls
ME/CFS is a severe, long-term disease characterized by postexertional malaise, overwhelming fatigue, sleep abnormalities, pain, and other symptoms that do not improve adequately with rest.][57][58]62] Fibroblasts derived from patients' skin biopsies have been widely used for disease studies, with the merits of convenient sample handling and culturability.Besides, the adherent nature and large cell volume make the image analysis of fibroblasts on a single cell level relatively easy compared to blood cells.Previous research has shown that fibroblasts derived from patients exhibit similar and measurable cellular abnormalities as the lesion tissue.Here we utilize fibroblasts derived from ME/CFS patients and healthy individuals as a model to study the LD dynamics in ME/CFS.Cells from healthy controls or ME/CFS patients under unstressed conditions were stained by AM-QTPA for LD visualization, PM-ML to mark the cell outlines, and DRAQ5 for cell nucleus, which were then subjected to Z stack imaging using spinning disk confocal microscope.Guided by the signals of PM-ML and DRAQ5, the LD number and size were analyzed on a single cell level in these samples.Quantification was performed with more than 30 cells per cell line.Results (Figure 4B,C, all data plot Figure S6, Supporting Information) suggests that the LDs are smaller with fewer numbers in control cells, whereas ME/CFS fibroblasts tend to form fewer but significantly larger LDs, which is consistent with the previous electron microscope observation of large lipid droplets in stimulated PBMCs from two ME/CFS patients. [57]These data demonstrate that AM-QTPA is capable to pinpoint subtle changes of LD dynamics in clinical samples such as fibroblasts.

AM-QTPA Enables Visualization of LD-Mitochondria Interaction in ME/CFS Fibroblasts
Previous research has pointed out the important roles of LDs in cellular energy supply, particularly during the lipolysis when fatty acids are released from LDs to mitochondria for energy production via -oxidation.The transport of these fatty acids likely occurs at contact sites between LDs and mitochondria.We next inspected the trafficking of LDs and LD-mitochondria in fibroblasts derived from ME/CFS patients and healthy controls.Time-lapse imaging (Figure 5A,B) of fibroblasts co-stained by AM-QTPA and MitoTrackers was acquired for a duration of 2 min with a 5 s interval, which allowed us to monitor the transient contact between LDs and mitochondria.We observed frequent interactions of these two organelles in fibroblasts from ME/CFS patients (Figure 5B).In combination with the larger size of LDs in the ME/CFS fibroblasts, this elevated frequency of interaction with the mitochondria suggests the occurance of lipolysis according to the previous finding that larger LDs are preferentially used by lipolysis. [63,64]Whereas smaller LDs are more likely to be degraded through the process of selective autophagy of LDs called lipophagy. [63,65]Looking into details, an interesting phenomenon of LD translocation between adjacent mitochondria was observed (Figure 5C; Video S1, Supporting Information), highlighting the capability of AM-QTPA for LD visualization in a highly dynamic disease-related scenario.

Conclusion
In summary, we report the synthesis, characterization, and applications of a novel fluorophore, AM-QTPA, for visualizing LD dynamics in live cells.Photophysical characterization has demonstrated that AM-QTPA is solvatochromic, viscosity sensitive and AIEE active.With the advantages of low background in aqueous environments, high selectivity, and easy usage, AM-QTPA represents an excellent probe for selective imaging and tracking of LDs in cells.Aided by AM-QTPA, the visualization and quantification of LD number and size can be easily achieved.More importantly, the applicability of AM-QTPA was further expanded into the clinical samples, which exposed complex and abnormal LD behaviors in the case of ME/CFS.Collectively, this work provides a new imaging tool for the LD research, which could potentially find a wide range of applications from fundamental biological research to disease-related pathological research and diagnosis of diseases including virus infection, cancers, neurodegeneration, and beyond.

Experimental Section
Materials and General Information: All starting materials were commercially available.Reagents and solvents were purchased from Sigma-Aldrich, AK Scientific, AJAX Chemicals, Fluorochem, and Chem-Supply, and used as received without purification. 1H-NMR and 13 C-NMR spectra were recorded using 400 MHz Bruker AV3HD-400 spectrometer with tetramethylsilane (TMS,  = 0) as the internal standard.HRMS spectra were acquired using an Agilent Technologies 6530 Accurate-Mass Q-TOF LC-MS operating in electrospray ionization (ESI) mode.
UV-vis absorption spectra were recorded at room temperature on an Agilent Cary 300 UV-vis spectrophotometer equipped with a 1.0 cm quartz cell.Fluorescence emission spectra were recorded on an Agilent Cary Eclipse Fluorescence Spectrophotometer and used 1.0 cm quartz cells.For UV-vis absorption measurement, the background of (co-)solvent alone was subtracted.For photoluminescence measurement, the emission slit was set at 5 nm, and the scan speed was set at fast.Data were plotted by using Origin 2019.
DFT calculations were carried out using B3LYP/6-31+g(d,p) level with Gaussian16, using the University of Melbourne HPC system, Spartan, for all calculations.Visualisation of structures were done using Avogadro 1.20 and GaussView 5.0.No negative frequencies for optimised structures were observed.
Synthesis-Preparation of NO 2 -QTPA. [66]: In a pressure tube filled with 2-methyl-8-nitroquinoline (1.13 g, 6 mmol), 4-(diphenylamino)benzaldehyde (1.97 g, 7.2 mmol), and Fe(OAc) 2 (52 mg, 0.3 mmol) was added toluene (9 mL) and TFA (46 μL, 0.6 mmol).The reaction was sealed under the protection of N 2 and stirred at 105 °C for overnight.Upon completion, the mixture was cooled to room temperature, and the solvent was removed under reduced pressure.The residue was purified by silica chromatography with petroleum spirit/ethyl acetate/triethylamine (90:9.5:0.5) to afford NO 2 -QTPA as an orange solid.Yield 79% (2.1 g). 1  Synthesis-Preparation of AM-QTPA. [67]: A suspension of NO 2 -QTPA (1.6 g, 3.6 mmol) in ethanol (51 mL) was heated at reflux.To this solution was added iron powder (2 g, 36 mmol) followed by a solution of NH 4 Cl (1.93 g, 36 mmol) in water (15 mL).The resulting suspension was heated at reflux for 2 h.The hot mixture was then filtered through a Celite pad, and the filtrate was evaporated under vacuum.The residue was dissolved in ethyl acetate and worked up with water, and the aqueous phase was further extracted with ethyl acetate.The combined organic extract was dried over Na 2 SO 4 , filtered, and evaporated under vacuum.The residue was washed with diethyl ether to afford the pure product as a bright yellow solid.Yield 82% (1.2 g). 1  Cell Culture and Imaging-Cell culture: HeLa cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) containing 10% FBS and antibiotics (100 units mL −1 penicillin and 100 μg mL −1 streptomycin) in a 5% CO 2 humidified incubator at 37 °C.
Fibroblast cultures were generated from skin biopsies according to the method previously described. [68]The 4 mm skin biopsy was dissected into 12 pieces with 2 pieces placed into each well of a 6-well plate containing advanced DMEM with 10% FBS.This medium favors the growth of fibroblasts but not for other cell types that require additional supplements.The fibroblast cells grow out from the skin biopsy after ≈1 week and typically reach confluence after 3-4 weeks.Fibroblasts were then maintained and cultured in the advanced DMEM supplemented with 10% FBS, Gluta-MAX (2 mm), and antibiotics (100 units mL −1 penicillin and 100 μg mL −1 streptomycin) in a 5% CO 2 humidified incubator at 37 °C.Access to clinical samples used in this study was approved by the Ethics, Integrity, and Biosafety Team of La Trobe University under human ethics HEC19316.In-formation for clinical samples was shown in Table S1 (Supporting Information).There were no significant differences in age between the two groups and gender proportions were matched.
Cell Culture and Imaging-Cytotoxicity test: Cells (6 × 10 3 ) were plated onto 96-well plates 24 h prior to dye application.Varying concentration of dye was applied for designated timepoints to test the cellular viability.CLARIOstar monochromator plate reader (BMG Labtech) in fluorescence intensity mode with excitation at 550/15 nm and emission at 600/20 nm was used to assess cell viability using AlamarBlue assay, following manufacturer's instruction.
Cell Culture and Imaging-Cell treatment with OA: The method of OA treatment was adopted from Biotium website [69] with modifications.Pure OA stored at −20 °C was warmed up to 37 °C until it was completely liquefied, and diluted in 50% EtOH/H 2 O (v/v) at 150 mm.OA-BSA complex was freshly prepared immediately before cell treatment, by mixing equal volumes of 150 mM OA in 50% EtOH/H 2 O (v/v) with 100 mg mL −1 BSA in MilliQ water.The mixture was then incubated at 37 °C for at least 1 h prior to cell experiment and subsequently used as the OA stock at 75 mm.The OA containing medium was made by diluting the OA stock (1:375) with culturing medium and used directly for cell treatment.The final working concentration for the treatment was 0.2 mM.
Cell Culture and Imaging-Cell Imaging: Cells were directly seeded on pre-sterilized Ibidi μ-Slide 8-well chambers at a density of 3 × 10 4 (for HeLa) or 2 × 10 4 (for fibroblasts) cells per well and incubated in cell culture medium for overnight prior to experiments.After treatments, cells were stained with 5 μm of AM-QTPA (1:100 dilution in FluoroBrite DMEM) at 37 °C for 45 min, and excess dye was washed away with FluoroBrite DMEM.Nile red, ER-Tracker Red, MitoTracker Red CMXRos / Deep Red FM or DRAQ5 was added into the staining medium at the concentration of 100 ng mL −1 , 1 μm, 200/100 nm, 5 μm, respectively, for counterstaining.Cells were maintained in FluoroBrite DMEM in the incubator with 5% CO 2 at 37 °C during live cell imaging.For fixed cell imaging, the stained cells were subsequently fixed in 4% (w/v) paraformaldehyde (PFA) solution in PBS.The stained and fixed cells were then rinsed with PBS and maintained in cold PBS for imaging.PM-ML staining was performed for fixed cells in PBS at the concentration of 5 μm at the room temperature.AM-QTPA was stored as stock solution at 0.5 mm (100×) in DMSO.
Imaging was performed on chambers using the Andor Dragonfly 202 spinning disk confocal microscope.AM-QTPA was excited at 405 nm (1.5% laser power; 105 ms exposure time) and fluorescence was collected with the emission filter at 520 nm.To visualize detailed structure of ER, the experiment of ER-Tracker Red counter staining was performed with the airy scan module using the Zeiss LSM 900 confocal microscope.
Statistical Analysis: All statistical analyses were performed using the unpaired t-test function in GraphPad Prism 8.0.

Figure 1 .
Figure 1.Photophysical characterization of AM-QTPA.A) Chemical structure of AM-QTPA.B) Absorption and photoluminescence (PL) spectra of AM-QTPA in pure DMSO and mixed water/DMSO (99:1).C) DFT calculation of HOMO and LUMO of AM-QTPA.D) Normalized PL spectra of AM-QTPA in different solvents.E) PL spectra of AM-QTPA in glycerol-ethylene glycol mixtures (E) with different glycerol fractions (f Gly ).F) Photographs of AM-QTPA in solid state under daylight and UV light.G) PL spectra of AM-QTPA in EtOH-H 2 O mixtures with different water fractions (f water ).H) Plot of maximum emission intensity of AM-QTPA versus the water fraction in EtOH-H 2 O mixtures.The above tests were performed with the concentration of AM-QTPA at 10 μM. ex = 395 nm.

Figure 2 .
Figure 2. AM-QTPA lights up LDs in cells.Confocal images of AM-QTPA (pseudo color: green) and Nile red (pseudo color: magenta) co-stained HeLa cells in culturing medium (CM, A) or OA (200 μm)-containing medium (C).B,D) Plot of gray value of linear ROIs in the merge images of (A,C), respectively.Scale bar, 20 μm.

Figure 3 .
Figure 3. Quantification of LD number and size in cells treated by OA.A-E) Maximum projection and Z stack orthogonal view of AM-QTPA (pseudo color: green) and PM-ML (pseudo color: red) stained HeLa cells with OA treatment at the indicated time points.Scale bar, 20 μm.Quantification of LD number per cell F) and averaged LD size G) based on signals of AM-QTPA in (A-E).

Figure 4 .
Figure 4. Visualizing LDs in fibroblasts derived from patients with ME/CFS versus healthy controls (information shown in Table S1, Supporting Information).A) Maximum projection of Z stack imaging of fibroblasts stained by AM-QTPA (pseudo color: green), PM-ML (pseudo color: red) and DRAQ5 (pseudo color: cyan).Scale bar, 20 μm.Median values of LD number per cells B) and averaged LD sizes C) quantitated from 30-40 cells per lines.* p ≤0.05.Error bars, mean ± SEM.

Figure 5 .
Figure 5. Imaging LD-mitochondria interactions in fibroblasts derived from ME/CFS patients.Representative images of live fibroblasts co-stained with AM-QTPA (pseudo color: green) and MitoTracker TM Red CMXRos (pseudo color: red), from healthy individual (A) or patients with ME/CFS (B).Scale bar, 20 μm.Lower panels: zoom-in images of ROIs (yellow solid outlines).C) Time lapse images of the ROI (white dash outline) in Figure 5B.Scale bar, 5 μm.Dash circles highlight the movement of LDs.