Tetramic Acid‐Derived Blue‐Emitting Fluorescent Probe for Selective Lipid Droplet Imaging in Live Cells and Zebrafish

Abstract

Lipid droplets (LDs) are dynamic intracellular organelles involved in lipid storage and metabolism, and their imaging is essential for studying metabolic processes in health and disease. Here, we describe four novel fluorescent tetramic acid derivatives as potential LD markers for live imaging in cellular and whole‐organism systems, and designate the compound with the most favorable properties, 1a, for further characterization. The probe 1a was evaluated in multiple cell lines, where it enabled clear visualization of intracellular LDs and showed strong colocalization with established lipid markers such as Nile Red and BODIPY 493/503. Photophysical characterization revealed photostability and signal‐to‐noise characteristics enabling fluorescence imaging applications. In addition, cytotoxicity assays indicated minimal acute and delayed effects at imaging concentrations. Importantly, 1a was successfully applied in a zebrafish model, enabling visualization of lipid‐rich structures in a whole‐organism context. Together, these findings identify 1a as a useful LD fluorescent marker for live imaging applications both in vitro and in vivo.

Four new fluorescent tetramic acid derivatives designed as lipid droplet (LD) imaging probes were synthesized, identifying compound 1a as the most promising candidate. Compound 1a enables straightforward, stable, and low‐toxicity visualization of LDs across multiple cell lines, showing strong correlation with established markers and compatibility with standard imaging techniques in vitro and in vivo (zebrafish model).

1. Introduction

Lipid droplets (LDs) are highly dynamic intracellular organelles that serve as the primary storage sites for neutral lipids, including triacylglycerols and sterol esters [1]. Once considered inert lipid reservoirs, LDs are now recognized as metabolically active structures involved in the regulation of lipid homeostasis, energy metabolism, membrane biosynthesis, and cellular signaling [2,‍ 3]. Structurally, LDs consist of a hydrophobic neutral lipid core surrounded by a phospholipid monolayer associated with a variety of proteins that regulate lipid synthesis, trafficking, and degradation [4]. Beyond their fundamental metabolic role, LDs participate in diverse cellular processes including response to oxidative stress, host–pathogen interactions, and regulation of cell survival pathways [5, 6]. Dysregulation of the formation, turnover, or composition of LDs has been implicated in numerous pathological conditions, including metabolic disorders, fatty liver disease, atherosclerosis, neurodegeneration, and cancer [7, 8, 9, 10]. Thus, reliable visualization and quantification of LDs are essential for understanding lipid metabolism.

Fluorescence microscopy has emerged as one of the most widely used and effective approaches for investigating LD biology, as it allows noninvasive visualization of cellular structures with high spatial and temporal resolution in living systems [11, 12, 13]. Fluorescence imaging combines rapid data acquisition with compatibility for live‐cell experiments, allowing dynamic processes to be monitored in real time. These approaches depend on fluorescent probes that selectively label LDs, which have become indispensable for studying LD biogenesis, growth, fusion, and interactions with other organelles, as well as for tracking changes in lipid metabolism under both physiological and pathological conditions [14, 15].

A wide range of fluorescent probes have been developed for LD imaging, including dyes based on Nile Red, BODIPY, coumarin, pyrene, and aggregation‐induced emission fluorophores [16, 17, 18, 19, 20, 21]. Among these, Nile Red and BODIPY 493/503 remain the most commonly used commercial LD markers due to their strong fluorescence in hydrophobic environments and relatively simple staining protocols (Figure 1) [22]. However, currently available LD probes, including the two mentioned, often present several limitations. The synthesis of LD probes is often a costly and time‐consuming synthetic process which requires several synthetic steps and expensive starting reactants and reagents. Many hydrophobic dyes exhibit limited specificity, leading to nonspecific accumulation in various organelles and background fluorescence [23, 24, 25]. Other constraints include limited photostability, restricted spectral flexibility for multicolor imaging, and incompatibility with certain experimental workflows such as long‐term imaging or fixation, to which LDs are sensitive [26, 27]. These challenges stress the continued need for new LD‐targeting probes that combine high specificity, low background signal, and compatibility with diverse imaging applications, as well as the development of efficient, straightforward, and sustainable synthetic methodologies.

FIGURE 1.

Structures of Nile Red, BODIPY 493/503, and 1a.

Short‐wavelength fluorescent probes represent an underexplored but potentially useful class of LD markers. Fluorophores emitting in the blue spectral region can provide several advantages, including reduced spectral overlap with commonly used green and red fluorescent reporters, thereby facilitating multiplex imaging experiments [22, 28]. At the same time, such probes can present challenges, including the requirement for ultraviolet (UV) excitation, reduced tissue penetration compared with longer‐wavelength fluorophores, limited availability, environment‐sensitive fluorescence, and insufficient in vivo validation [29]. However, they remain valuable imaging tools when used at low excitation intensities and short acquisition times, particularly in applications requiring spectral multiplexing or reduced interference from longer‐wavelength autofluorescence.

Here, we report the development and characterization of 1a, a simple small‐molecule fluorescent probe for selective visualization of LDs in living cells and organisms (Figure 1). Compound 1a arises from our wider work on the fluorescent derivatives of tetramic acid, a privileged motif in natural products [30, 31, 32, 33]. Tetramic acid derivatives represent a structurally versatile scaffold whose substitution pattern enables systematic modulation of their electronic and physicochemical properties, including fluorescence behavior. In addition, many members of this class display affinity for hydrophobic environments and environment‐dependent emission, which motivated us to explore these compounds as potential LD probes, particularly in the underrepresented blue‐emitting spectral region [30, 32, 34, 35, 36]. The probe 1a exhibits strong fluorescence upon excitation in the near‐UV region and displays robust accumulation within neutral lipid‐rich LDs. Colocalization studies with established LD markers indicate a high degree of specificity toward LDs, while minimal signal is observed in other intracellular compartments. The probe demonstrates low cytotoxicity, stable staining performance across multiple cell types, and compatibility with live‐cell imaging workflows. In addition, the blue‐channel emission of 1a allows spectral multiplexing with commonly used green and red fluorophores. Importantly, 1a can be used not only in cell culture but also in living organisms, such as zebrafish larvae, enabling visualization of LDs during early developmental stages.

2. Results and Discussion

2.1. Chemistry

The probes 1a–d were designed to incorporate a push–pull system akin to known fluorescent probes that exhibit intramolecular charge transfer (ICT), and their two‐step synthesis is shown in Scheme 1 [22]. In the first step, a microwave‐assisted S _N Ar reaction of commercially available 2‐fluoro‐4‐(trifluoromethyl)benzaldehyde (2) and the appropriate secondary amines 3a–d was utilized to attain aldehyde intermediates 4a–d. The obtained aldehydes underwent aldol condensation with 4‐methoxy‐1,5‐dihydro‐2H‐pyrrol‐2‐one (5) to afford the fluorescent products, with yields reported as overall yields for the two‐step process. Despite modest yields and extended purifications for compounds 1b and 1d, the overall synthetic route remains straightforward and operationally simple. In the aldol condensation step, a minor amount of the (E)‐isomer is formed, which cannot be separated from the (Z)‐isomer. The E–Z interconversion cannot be avoided and occurs in the solution as well as on the column during chromatography.

SCHEME 1.

Synthesis of compounds 1a–1d.

2.2. Optical Properties of Compounds 1a–1d and Lipophilicity Estimation

The absorption and fluorescence properties of compounds 1a–1d were systematically investigated. Absorption spectra (260–‍650 nm) were recorded in dimethyl sulfoxide (DMSO) and hexane for all compounds (Figures S6 and S7), and excitation spectra (250–510 nm) were recorded in DMSO (Figure S1). Fluorescence spectra (380–670 nm) were measured in DMSO (Figure S2), aqueous media (Figure S5), and hexane (Figure S3) for all compounds. The relevant photophysical parameters determined in DMSO and water are summarized in Table 1. The photophysical parameters measured in hexane are compiled in Table 2. All investigated compounds exhibited pronounced fluorescence in both the solid state and in solution upon excitation at 366 nm. In hexane and DMSO, all probes display pronounced positive solvatochromism: their emission maxima are red‐shifted by 27–63 nm in DMSO relative to hexane, indicating stabilization of a more polar excited state in polar aprotic media. This behavior is characteristic of push–pull fluorophores with ICT character, as commonly observed for ICT emitters in polar solvents [39].

TABLE 1.

Absorption and fluorescence properties of compounds 1a–1d in DMSO.

Compound	${\lambda }_{\mathrm{ex}}^{\max }\text{ (nm)}$	ε max (M‍–‍1 cm–‍1‍ )	Concentration (M)a	${\lambda }_{\mathrm{em}}^{\max }$ (nm)b	${\lambda }_{\mathrm{ex}}^{\max }$ (nm)c	Intensity (cps)d	Stokes shift (nm)e	Quantum yield (%)f
1a	364	11 530	1 × 10–5	523	502	1.3 × 105	159	4.28
1b	361	11 830	1 × 10–5	512	560	8.6 × 105	151	21.14
1c	362	11 600	1 × 10–5	505	540	9.07 × 105	143	24.67
1d	366	7700	1 × 10–5	533	513	6.78 × 104	167	3.04

^a For recording emission spectra.

^b Recorded in DMSO.

^c Recorded in water.

^d Counts per second.

^e In DMSO.

^f Quantum yield determined for compounds in DMSO at λ = 357 nm, relative to quinine sulfate in 1 M sulfuric acid as standard, using reported procedures [37, 38].

TABLE 2.

Absorption and fluorescence properties of compounds 1a–1d in hexane.

Compound	${\lambda }_{\mathrm{em}}^{\max }\left(\mathrm{nm}\right)$	ε max (M‍–‍1 cm‍–‍1)	Concentration (M)	Intensity (cps)	Quantum yield (%)	Brightness (M–1 cm–1)
1a	496	10,380	1 × 10–5	7.73 × 105	9.88	1025
1b	461	12,110	1 × 10–5	3.44 × 105	5.19	628
1c	459	9730	1 × 10–5	2.88 × 105	3.14	305
1d	470	8230	1 × 10–5	6.68 × 104	8.98	739

In aqueous solution, all compounds show strongly reduced fluorescence intensity and altered emission maxima compared to DMSO. Given the low solubility of the hydrophobic fluorophore core in water, these changes are most plausibly associated with aggregation‐induced effects and activation of nonradiative relaxation pathways [40, 41, 42]. The opposite directions of the spectral shifts observed for 1a/1d (blue shift) versus 1b/1c (red shift) suggest that different supramolecular arrangements may be favored for individual analogues, but a detailed mechanistic assignment of specific aggregation modes is not pursued further here, as the primary focus of this work is on the application. Probe 1a displays the highest molecular brightness in hexane, which closely mimics the hydrophobic interior of LDs; accordingly, its absorption and emission spectra in hexane are presented in Figure 2. To further evaluate the solvent‐dependent fluorescence behavior of 1a, the emission spectrum was additionally recorded in toluene (Figure S4). Compound 1a exhibited an emission maximum at 479 nm in toluene, which is red‐shifted relative to hexane (469 nm), consistent with its polarity‐sensitive fluorescence response.

FIGURE 2.

(a) Emission spectrum of 1a in hexane, c = 1 × 10⁻⁵ M; (b) absorption spectrum of 1a in hexane c = 1 × 10⁻⁵ M.

Calculated LogP values obtained using SwissADME [43] (consensus LogP) indicate that the series spans a moderate lipophilicity range (2.39–4.12), with 1d being the most lipophilic analogue (see Table S1). For compound 1a, the consensus LogP is 3.41, consistent with a hydrophobic character compatible with lipid‐droplet environments.

Overall, the four probes share a common donor–acceptor architecture consisting of an amino‐substituted aryl ring connected through a conjugated benzylidene bridge to the electron‐deficient tetramic acid core. This structural arrangement is consistent with an ICT excited state, in which photoexcitation promotes electron density transfer from the amino donor towards the tetramic acid acceptor. The photophysical properties of the probes are influenced by the nature of the amine substituent. Replacement of piperidine in 1a with morpholine (1b) or thiomorpholine (1c) introduces electronegative heteroatoms into the donor ring, altering the electron‐donating ability and excited‐state stabilization, as well as decreasing lipophilicity. On the other hand, substituting the cyclic piperidine motif with the less rigid dipropylamine (1d) introduces an additional aliphatic carbon expected to increase lipophilicity. While 1b and 1c displayed the highest quantum yields in DMSO, 1a showed the highest brightness in hexane, whereas 1d was the most lipophilic member of the series. Notably, 1a provides the best lipid‐droplet imaging performance (vide infra), despite its slightly lower lipophilicity than 1d, suggesting that an optimal balance between hydrophobicity, fluorescence response, and cellular behavior, rather than maximal lipophilicity alone, governs probe efficiency.

2.3. Measurement of Intracellular Uptake of Compounds 1a–1d

The NCI‐H460 non‐small cell lung carcinoma (NSCLC) cell line, which contains abundant LDs, was selected as a robust model for evaluating fluorescent probe uptake and imaging [44].

Intracellular uptake of compounds 1a–1d in NCI‐H460 cells after incubation with 25 µM for 30 min at 37 °C, assessed by flow cytometry. Fluorescence intensity was recorded in the blue, green, and red channels (Ex/Em: 405/450 nm, 488/525 nm, and 488/585 nm, respectively). Unstained cells served as control.

To determine whether compounds 1a–1d can enter NCI‐H460 cells, their intracellular fluorescence was analyzed using flow cytometry (Figure 3). All tested compounds generated detectable fluorescence relative to unstained control cells, confirming rapid cellular uptake. The fluorescence signal was primarily detected in the blue channel, consistent with the desired compounds’ excitation/emission properties. These results demonstrate that the tested compounds efficiently penetrate the cellular membrane within 30 min, a key requirement for their use as intracellular fluorescent probes. Compounds 1a and 1d showed the highest increase in fluorescent intensity in the blue channel, whereas 1b and 1c showed smaller responses. Although 1d has the longest emission wavelength and the largest Stokes shift, it also exhibited emission in the green and red channels, which could interfere with multiplexed analyses. In addition, its quantum yield was the lowest in the series and it proved difficult to purify. Accordingly, all subsequent experiments were performed using 1a.

FIGURE 3.

Intracellular uptake screening of compounds 1a–1d in NCI‐H460 cells.

2.4. Compound 1a Selectively Labels LDs in Live Cells

We next assessed the intracellular staining pattern of 1a. All imaging experiments were conducted in live cells under serum‐free conditions with 25 μM 1a, unless otherwise specified. Under these conditions, widefield fluorescence microscopy revealed a punctate intracellular fluorescence pattern in NCI‐H460 cells. To confirm whether these structures are LDs, cells were costained with established LD markers, Nile Red and BODIPY 493/503, and imaged (Figure 4a). Widefield fluorescence microscopy revealed pronounced spatial overlap between 1a and both Nile Red and BODIPY 493/503, as supported by line intensity profiles (Figure 4a).

FIGURE 4.

Colocalization of 1a with Nile Red and BODIPY 493/503. (a) Representative widefield images of NCI‐H460 cells costained with 1a and either Nile Red or BODIPY 493/503 (left). Line intensity profiles show fluorescence distribution along the regions indicated by the yellow lines in the merged images (right). (b) Representative maximum intensity projections of z‐stack confocal fluorescence images of NCI‐H460 cells costained with 1a and either Nile Red or BODIPY 493/503. (c) Pixel intensity correlations between the 1a channel and either the Nile Red or BODIPY 493/503 channel are presented as scatter plots from confocal images. (d) Quantitative colocalization analysis of 1a with Nile Red and BODIPY 493/503 is presented using Pearson's correlation coefficient and Manders’ overlap coefficients (M1 and M2).

To better resolve LDs and enable quantitative colocalization analysis, NCI‐H460 cells costained with 1a and either Nile Red or BODIPY 493/503 were imaged by confocal microscopy (Figure ‍4b). To avoid spectral cross‐talk, all channels were acquired using frame‐by‐frame sequential scanning. Confocal imaging showed that 1a fluorescence was distributed in discrete intracellular spots, consistent with LD morphology. These structures exhibited strong spatial overlap with Nile Red and BODIPY 493/503 fluorescent signals (Figure 4c). A high degree of colocalization between the 1a signal and Nile Red or BODIPY 493/503 was confirmed by high Pearson's correlation coefficients (0.86 and 0.87, respectively) and Manders’ overlap coefficients (M1 = 0.83 and M2 = 0.87 for Nile Red; M1 = 0.85 and M2 = 0.92 for BODIPY 493/503) (Figure 4d).

Furthermore, live‐cell confocal imaging captured the dynamic behavior of LDs labeled with 1a, including rapid movement and transient contacts (Video S1), supporting the probe's suitability for live LD tracking in cells.

Additionally, widefield fluorescence microscopy revealed consistent LD labeling by 1a in multiple cell lines from diverse tissue types costained with Nile Red, including human lung adenocarcinoma (A549), glioblastoma (A172), and osteosarcoma (SAOS‐‍2) cells (Figure S8). These observations suggest that 1a is a versatile fluorescent probe for studying LD biology across diverse cellular models.

LD probes such as Nile Red and BODIPY 493/503 are widely used, but their emission in the green/red spectral range limits their compatibility with multicolor imaging [16, 45]. Notably, Nile Red exhibits strong solvatochromism and broad emission, causing spectral overlap and nonspecific background signals, while BODIPY 493/503 may undergo photoconversion [46]. In comparison, 1a produced a distinct fluorescence signal in the blue channel, allowing clear visualization of LDs while leaving the green and red channels available for other markers.

Compared to commonly used LD probes, blue‐emitting LD dyes remain relatively rare. Among these, the DBC30 probe, based on benzimidazole‐derivatized coumarin, and probes derived from fused pyrazolopyridine and pyrrolopyridine‐dihydrochromeno scaffolds have shown selective LD staining with favorable fluorescence properties [47, 48]. However, these probes require complex chemical synthesis, are not widely commercially available, and have limited validation in live organisms, restricting their broader biological use. Another example is monodansylpentane, a structurally simple dye that stains neutral lipid‐rich structures [28]. Nonetheless, it has been reported to display environment‐dependent fluorescence and potential background signals in nonlipid hydrophobic compartments, reducing imaging contrast. Additionally, its use has mostly been limited to in vitro systems, with limited validation in in vivo models. Commercial options like Droplite Blue or Lipi Blue are primarily validated in cell culture and remain insufficiently characterized regarding photophysical properties and long‐term imaging performance [49].

2.5. Compound 1a Does Not Localize to Intracellular Organelles or Cellular Membranes

Many hydrophobic dyes can accumulate in other intracellular compartments such as mitochondria or cellular membranes, leading to nonspecific background fluorescence [23]. Weakly basic fluorophores may also become trapped within acidic organelles such as lysosomes [24]. Furthermore, because LDs originate from the endoplasmic reticulum (ER), insufficient probe specificity may result in signal overlap with ER‐associated lipid structures [25]. To confirm that the specificity of 1a is directed toward LDs rather than cellular membranes more broadly, we examined its distribution relative to markers of major intracellular organelles and the plasma membrane. NCI‐H460 cells were costained with MitoTracker Red CMXRos for mitochondria, LysoTracker Red DND‐99 for lysosomes, ER‐Tracker Red for the ER, and CM‐DiI for the plasma membrane (Figure 5). In all cases, fluorescence imaging revealed no observable spatial overlap between 1a and these markers. Representative fluorescence intensity line profiles further showed clearly separated signal peaks, indicating distinct spatial localization (Figure 5). Additional mitochondrial and ER probes, Rhodamine 123 and 3,3′‐dihexyloxacarbocyanine iodide (DiOC₆), respectively, similarly showed no detectable overlap with 1a fluorescence (Figure S9). Taken together, these results indicate that 1a does not generally partition into phospholipid bilayers or membrane‐rich organelles. Instead, its localization is highly selective for LDs, consistent with a preferential affinity for neutral lipid environments.

FIGURE 5.

Costaining of 1a with markers of intracellular organelles and the plasma membrane. Representative widefield fluorescence images of NCI‐H460 cells costained with 1a and either MitoTracker Red CMXRos, LysoTracker Red DND‐99, ER‐Tracker Red, or CM‐DiI (left). Line intensity profiles show fluorescence distribution along the regions indicated by the yellow lines in the merged images (right).

2.6. Fluorescence from 1a Responds to Cellular LD Abundance

We next investigated whether 1a labeling responds to changes in cellular LD content. Oleic acid (OA) treatment is widely used to induce LD formation and to study droplet dynamics, metabolism, and associated cellular processes [50, 51, 52]. As a monounsaturated fatty acid, OA is readily taken up by cells and converted into triacylglycerols, which are subsequently stored in LDs.

Treatment of NCI‐H460 cells with OA resulted in a clear increase in both the fluorescence intensity and abundance of punctate structures labeled by 1a, as observed by fluorescence microscopy (Figure 6a). Consistent with these observations, flow cytometry analysis showed a pronounced shift toward higher fluorescence intensity in OA‐treated cells, indicating increased 1a accumulation at the population level (Figure 6b).

FIGURE 6.

Fluorescence of 1a reports oleic acid‐induced changes in cellular lipid droplet content. (a) Representative widefield images of 1a‐labeled NCI‐H460 cells treated with 0 or 200 µM oleic acid for 24 h. (b) Flow cytometry analysis of 1a fluorescence intensity (blue channel, Ex 405 nm/Em 450 nm) in NCI‐H460 cells pretreated with 0, 200, or 400 µM oleic acid for 24 h. Representative flow cytometry plots (left) and quantification of mean fluorescence intensity (right) are shown. Data are presented as mean fluorescence intensity relative to untreated control (mean ± SD). Statistical significance was determined using one‐way ANOVA (***p < 0.001).

Together, these findings demonstrate that 1a sensitively reports changes in cellular LD content and is suitable for both imaging‐based analysis and flow cytometry‐based quantification of cellular lipid storage.

2.7. Efficient Staining of Lipid‐Rich Structures in Live Zebrafish

To validate the performance of 1a in a whole‐organism context, we evaluated its ability to label lipid stores in 3 days postfertilization (dpf) zebrafish larvae. Whole‐mount live staining with 1a produced strong fluorescence signal within the yolk sac, a lipid‐rich compartment that serves as the primary energy reservoir during early development (Figure 7a). Costaining with Nile Red showed a similar distribution pattern, with both probes enriched in the yolk and displaying increased fluorescence following OA treatment. A zoomed‐in view of the yolk region (Figure 7b) revealed a punctate fluorescence pattern for both 1a and Nile Red, consistent with LD‐like structures. Individual lipid stores were more clearly resolved with 1a in both control and OA‐treated embryos. Quantification of 1a signal confirmed ~44% higher mean fluorescence intensity following OA treatment (Figure 7c).

FIGURE 7.

Compound 1a labels yolk lipid stores in zebrafish in vivo. (a) Representative whole‐larva panorama images of control and oleic acid‐treated 3 days postfertilization (dpf) zebrafish stained with 1a and Nile Red. Z‐stack maximum intensity projections are shown. (b) A zoomed‐in view of the yolk of control and oleic acid‐treated 3 dpf zebrafish stained with 1a and Nile Red. (c) Quantification of mean fluorescence intensity of 1a after the oleic acid treatment, relative to control. Data are analyzed using GraphPad Prism 8 and presented as mean ± SD. Statistical significance was determined using a two‐tailed Welch's t‐test (***p < 0.001).

Currently, there is limited data on the in vivo applicability of blue LD probes, particularly in whole‐organism models such as zebrafish, where autofluorescence and optical constraints require careful spectral selection. In this context, the blue emission spectrum of 1a enabled visualization of lipid stores without interference from endogenous yolk autofluorescence, which is particularly prominent in the green channel and limits the utility of green‐emitting probes such as BODIPY 493/503 for live whole‐mount imaging [53].

No morphological abnormalities were observed, supporting the biocompatibility of 1a with live embryo and larval imaging. Together, these results show that 1a effectively labels lipid‐rich structures in intact zebrafish embryos and enables clear visualization of yolk lipid stores. The optical transparency of zebrafish larvae and the compatibility of 1a with live imaging further support its use in developmental studies, metabolic perturbation models, and screening applications.

Blue‐emitting LD probes remain underrepresented compared to green and red dyes, and are largely restricted to cell‐based applications, with limited evaluation in whole‐organism models. Compound 1a addresses this gap by providing a blue‐channel probe with demonstrated applicability in both live cells and zebrafish larvae.

2.8. Evaluation of Labeling Conditions and Fixation Sensitivity of 1a In Vitro

Labeling conditions for 1a were evaluated in vitro to define optimal imaging parameters. Labeling in serum‐containing and serum‐free medium, different probe concentrations, and fixation conditions were assessed (Figure S10).

Serum‐free medium staining of NCI‐H460 cells with 1a produced markedly stronger fluorescence than serum‐containing conditions, which yielded only weak intracellular signal (Figure S10a). This suggests that serum components may interfere with 1a staining, potentially by reducing the availability of free probe required for efficient cellular uptake and intracellular accumulation.

Following this observation, staining efficiency was evaluated in NCI‐H460 cells across increasing concentrations of 1a (5, 10, 25, and 50 µM) in serum‐free medium (Figure S10b). The fluorescence signal increased in a concentration‐dependent manner. While 50 µM produced the highest fluorescence, it led to signal saturation and reduced staining consistency across cells, without providing improved definition of LDs compared to 25 µM. Lower concentrations yielded weaker staining and less well‐defined signal, depending on imaging settings. In contrast, 25 µM provided strong, well‐resolved staining and was selected as the optimal working concentration.

Although the optimal working concentration of 1a is higher than that typically reported for some commercial LD dyes, it provided the best balance between signal intensity, spatial resolution, and staining consistency under live‐cell conditions, without detectable effects on cell viability. The requirement for higher concentrations may reflect the physicochemical properties of 1a, including its cellular uptake efficiency and partitioning into neutral lipid environments. Future optimization of the molecular structure could further enhance probe brightness, LD affinity, or cellular permeability, potentially enabling effective staining at lower concentrations. Nonetheless, 1a shows robust live‐cell and in vivo imaging performance, supporting its use as an LD probe. It should be noted that the optimal working concentration of 1a might depend on the cellular model and imaging platform, including differences in uptake efficiency and detector sensitivity.

Fixation of LDs remains challenging, as commonly used organic solvents such as cold methanol and acetone disrupt their structure by extracting cellular lipids, whereas paraformaldehyde (PFA) better preserves lipid content and droplet integrity, making it the preferred method for LD imaging [27, 54]. However, even under optimized fixation conditions, live‐cell imaging is often preferred for accurate visualization of LDs. To assess whether 1a labeling is compatible with fixed‐cell imaging, cells stained with 1a were imaged before and after fixation with 4% PFA for 15 min at room temperature (Figure S10c). After fixation, the fluorescence pattern of 1a was partially lost, indicating disruption of intracellular localization and sensitivity to fixation under these conditions. As a result, 1a is better suited for live‐cell imaging and is not optimal for fixed‐cell imaging.

2.9. Evaluation of the Acute and Delayed In Vitro Cytotoxicity of 1a Following Transient Exposure

To evaluate the cytotoxic effects of 1a under imaging‐relevant conditions, both acute and delayed cytotoxicity were assessed in NCI‐H460 cells following transient exposure. Acute cytotoxicity was examined immediately after 30 min incubation with 1a using propidium iodide staining, which showed few positive cells, indicating minimal cell death (Figure S11a). For long‐term assessment, cells were washed after 30 min exposure and cultured for 72 h before performing a resazurin assay. Under these conditions, metabolic activity remained comparable to untreated control, indicating that transient exposure to 1a does not affect subsequent cell viability or proliferation (Figure S11b).

Together, these results indicate that short‐term labeling with 1a is well tolerated and does not impair either immediate cell viability or longer‐term cellular function following washout. This supports its compatibility with live‐cell imaging workflows that require cells to remain viable after staining for downstream analyses or continued culture. The absence of delayed toxicity suggests that 1a is either efficiently cleared after staining or retained in LDs without perturbing cellular viability over the observed time frame.

2.10. Photostability and Signal‐to‐Noise Ratio of 1a in LD Imaging

To complete the evaluation of probe 1a for LD imaging applications, we assessed its key photophysical and imaging performance characteristics, photostability and signal‐to‐noise ratio (SNR), in comparison with established LD dyes Nile Red and BODIPY 493/503 in NCI‐H460 cells. The probe exhibited moderate photostability, with a gradual decrease in fluorescence intensity during repeated imaging cycles (Figure 8a). In comparison, Nile Red and BODIPY 493/503 showed higher resistance to photobleaching under identical imaging conditions, although 1a maintained detectable fluorescence enabling LD visualization within standard imaging timeframes.

FIGURE 8.

Photostability and signal‐to‐noise ratio of 1a in NCI‐H460 cells. (a) Photostability curves showing time‐dependent fluorescence intensity of NCI‐H460 cells labeled with 1a, Nile Red, or BODIPY 493/503 during repeated imaging. Mean fluorescence intensity was quantified from 76 timelapse images acquired over 10 min at 8 s intervals (200 ms exposure time). Data are presented as percentage fluorescence intensity relative to the initial time point (0 s) (mean ± SD) and were analyzed using GraphPad Prism 8. (b) Signal‐to‐noise ratio values for NCI‐H460 cells labeled with 1a, Nile Red, or BODIPY 493/503 (mean ± SD).

SNR measurements were acquired under identical imaging settings, enabling direct comparison of probe performance. Quantitative analysis revealed SNR values of ~21 for 1a, compared with ~32 for Nile Red and ~74 for BODIPY 493/503. While 1a exhibited lower SNR than both reference dyes, LDs remained clearly distinguishable, enabling robust detection. These differences likely reflect the higher degree of optimization of Nile Red and BODIPY 493/503 for brightness and contrast, whereas 1a was not specifically optimized for maximal signal amplification. Importantly, despite lower SNR, 1a provides contrast that enables LD imaging under the experimental conditions used. A practical advantage of 1a is its blue emission, which avoids spectral overlap with commonly used green‐ and red‐emitting probes and enables multicolor imaging, positioning it as a complementary LD probe with potential for further optimization.

3. Experimental Section

3.1. Chemistry

Unless stated otherwise, all solvents and reagents were obtained from commercial sources and used without further purification. Dry‐flash chromatography was performed on SiO₂ (0.018–0.032 mm). IR spectra were recorded on a Thermo‐Scientific Nicolet 6700 FT‐IR Diamond Crystal instrument. ¹H and ¹³C NMR spectra were recorded on a Bruker Ultrashield Avance III spectrometer (at 500 and 125 MHz, respectively) and Varian 400/54 Premium Shielded spectrometer (at 400 and 100 MHz, respectively). Chemical shifts were expressed in parts per million (ppm) on the (δ) scale. Chemical shifts were calibrated relative to those of the solvent. Minor signals in the ¹H and ¹³C NMR spectra originate from the (E)‐isomer, which cannot be separated from the (Z)‐isomer. Fluorescence spectra were recorded on Horiba Jobin Yvon Fluoromax‐4 spectrometer, equipped with Peltier element and magnetic stirrer for cuvette, using quartz cells with 1 cm path length and 4 mL total volume. The excitation wavelength was 360 nm, with 5 nm slits; emission spectra were recorded in 380–670 nm wavelength range, with 5 nm slits, and 0.1 s integration time. UV–vis spectra were recorded on an Agilent Cary 3500 UV–vis Multicell Peltier spectrophotometer, using quartz cells with 1 cm path length and 4 mL total volume. Mass spectrometric measurements high‐resolution mass spectrometry (HRMS) were performed with a Shimadzu Scientific Instruments QToF 9030 LC‐MS system, equipped with a Nexera LC‐‍40D xs UHPLC, consisting of a CBM‐40 Lite system controller, a DGU‐405 Degasser Unit, two LC‐40D XS UHPLC pumps, a SIL‐40C XS autosampler and a Column Oven CTO‐40S. UV data were collected with a Shimadzu Nexera HPLC/UHPLC Photodiode Array Detector SPD M40 in the range of 190–800 nm. Mass spectra were subsequently recorded with the quadrupole time‐of‐flight (QToF) 9030 mass spectrometer. Melting points were determined on a Boetius PMHK apparatus and were not corrected.

3.1.1. General Procedure A

A solution of aldehyde 2 (1 equiv.) in DMF was treated with the corresponding amine 3a–3d (7 equiv.) under an argon atmosphere. The vial was sealed and the reaction mixture was heated at 130 °C for 1 h in a microwave reactor. The reaction mixture was extracted with ethyl acetate three times, and the combined organic layers were dried over anhydrous Na₂SO₄. Filtration was used to remove Na₂SO₄, and the organic layers were evaporated to dryness under vacuum. The crude aldehyde 4 was purified by dry flash chromatography on a column of silica gel and used without further characterization.

3.1.2. General Procedure B

The appropriate aldehyde 4a–4d (1 equiv.; from General procedure A) and 4‐methoxytetramic acid 5 (1 equiv.) were suspended in DMSO, then sodium hydroxide (7 equiv.) was added, and the mixture was stirred at 60 °C for 16 h. The reaction mixture was extracted with ethyl acetate three times, and the combined organic layers were dried over anhydrous Na₂SO₄. Filtration was used to remove Na₂SO₄, and the organic layers were evaporated to dryness under vacuum. The crude product was purified by dry flash chromatography on a column of silica gel.

3.1.3. (Z)‐4‐Methoxy‐5‐(2‐(piperidin‐1‐yl)‐4‐(trifluoromethyl)benzylidene)−1,5‐dihydro‐2H‐pyrrol‐2‐one (1a)

General procedure A was followed using aldehyde 2 (18.2 mg, 0.0947 mmol), piperidine (65.5 μL, 0.663 mmol) and DMF (2 mL) to afford product 4a (13.0 mg). According to the general procedure B, 4a (13 mg, 0.050 mmol), 4‐methoxytetramic acid 5 (5.7 mg, 0.050 mmol), NaOH (14 mg, 0.35 mmol) and DMSO (2 mL) were used to afford a yellow solid 1a (13 mg, 39% over two steps), mp = 157–160 °C. IR (ATR): 3187, 2945, 2852, 2803, 2744, 1690, 1606, 1500, 1428, 1379, 1310, 1263, 1226, 1194, 1163, 1122, 1080, 1003, 964, 950, 875, 858, 821, 796, 749, 729, 700, 666, 654, 627 cm^–1. ¹H NMR (500 MHz, CDCl₃) δ: 8.41 (s, 1H, NH), 7.40 (d, J = 7.9 Hz, 1H), 7.28 (bs, 2H), 6.32 (s, 1H), 5.19 (d, J = 1.6 Hz, 1H), 3.92 (s, 3H), 2.97 (t, J = 5.3 Hz, 4H), 1.75 (p, J = 5.7 Hz, 4H), 1.63–1.53 (m, 2H) ppm. ¹³C NMR (100 MHz, CDCl₃) δ: 171.04, 167.26, 151.72, 133.49, 132.01, 131.71, 130.78 (q, J = 32.0 Hz), 124.14 (q, J = 272.4 Hz, ‒CF₃), 119.41, 116.09, 104.28, 93.37, 58.47, 53.48, 25.96, 24.04 ppm. HRMS (ESI/Q‐TOF) m/z: calcd. for C₁₈H₂₀F₃N₂O₂ [M+H]⁺: 353.1471, found: 353.1472.

3.1.4. (Z)‐4‐methoxy‐5‐(2‐morpholino‐4‐(trifluoromethyl)benzylidene)‐1,5‐dihydro‐2H‐pyrrol‐2‐one (1b)

General procedure A was followed using aldehyde 2 (41.2 mg, 0.214 mmol), morpholine (131.0 μL, 1.498 mmol) and DMF (2 mL) to afford product 4b (20 mg). According to the general procedure B, 4b (20 mg, 0.077 mmol), 4‐methoxytetramic acid 5 (8.7 mg, 0.077 mmol), NaOH (21.6 mg, 0.539 mmol) and DMSO (2 mL) were used to afford a yellow solid 1b (20 mg, 26% over two steps), mp = 215 °C. IR (ATR): 3209, 2959, 2894, 2850, 1702, 1609, 1502, 1454, 1442, 1427, 1374, 1338, 1313, 1254, 1225, 1169, 1118, 1083, 1052, 1003, 958, 926, 872, 796, 735, 705, 660, 635, 624 cm^–1. ¹H NMR (400 MHz, CDCl₃) δ: 8.23 (s, 1H, NH), 7.49 (d, J = 8.0 Hz, 1H), 7.36–7.29 (m, 1H), 7.26–7.23 (m, 1H), 6.39 (s, 1H), 5.20 (d, J = 1.6 Hz, 1H), 3.93 (s, 3H), 3.89–3.84 (m, 4H), 3.05–2.99 (m, 4H) ppm. ¹³C NMR (100 MHz, CDCl₃) δ: 171.30, 167.17, 150.78, 133.41, 131.60, 131.22, 130.80, 124.00 (q, J = 272.8 Hz, ‒CF₃), 120.00, 115.57, 103.68, 93.24, 66.87, 58.55, 52.22 ppm. HRMS (ESI/Q‐TOF) m/z: calcd. for C₁₇H₁₈F₃N₂O₃ [M + H]⁺: 355.1264, found: 355.1265.

3.1.5. (Z)‐4‐methoxy‐5‐(2‐thiomorpholino‐4‐(trifluoromethyl)benzylidene)‐1,5‐dihydro‐2H‐pyrrol‐2‐one (1c)

General procedure A was followed using aldehyde 2 (31.7 mg, 0.165 mmol), thiomorpholine (120.1 μL, 1.155 mmol), and DMF (2 mL) to afford product 4c (29.8 mg). According to the general procedure B, 4c (29.8 mg, 0.108 mmol), 4‐methoxytetramic acid 5 (12.2 mg, 0.108 mmol), NaOH (30.2 mg, 0.756 mmol) and DMSO (2 mL) were used to afford yellow solid 1c (36.2 mg, 59% over two steps), mp = 220 °C. IR (ATR): 3203, 3107, 3018, 2977, 2939, 2911, 2832, 1690, 1608, 1502, 1453, 1429, 1380, 1321, 1284, 1224, 1203, 1157, 1120, 1081, 1034, 1003, 967, 932, 891, 874, 845, 825, 811, 795, 725, 711, 699, 654, 626 cm^–1. ¹H NMR (400 MHz, CDCl₃) δ: 8.15 (s, 1H, NH), 7.53 (d, J = 8.0 Hz, 1H), 7.36–7.31 (m, 1H), 7.25–7.22 (m, 1H), 6.40 (s, 1H), 5.22 (d, J = 1.6 Hz, 1H), 3.94 (s, 3H), 3.28–3.21 (m, 4H), 2.84–2.76 (m, 4H) ppm. ¹³C NMR (100 MHz, CDCl₃) δ: 171.48, 167.17, 152.38, 133.15, 131.99, 130.86, 130.29, 123.86 (q, J = 272.4 Hz, ‒CF₃), 119.92, 116.46, 103.28, 93.03, 58.42, 54.50, 28.00 ppm. HRMS (ESI/Q‐TOF) m/z: calcd. for C₁₇H₁₈F₃N₂O₂S [M+H]⁺: 371.10356, found: 371.10271.

3.1.6. (Z)‐5‐(2‐(dipropylamino)‐4‐(trifluoromethyl)benzylidene)‐4‐methoxy‐1,5‐dihydro‐2H‐pyrrol‐2‐one (1d)

General procedure A was followed using aldehyde 2 (18.3 mg, 0.0953 mmol), dipropylamine (91.5 μL, 0.667 mmol) and DMF (2 mL) to afford product 4d (9.5 mg). According to the general procedure B, 4d (9.5 mg, 0.035 mmol), 4‐methoxytetramic acid 5 (4.0 mg, 0.035 mmol), NaOH (9.8 mg, 0.25 mmol) and DMSO (2 mL) were used to afford yellow solid 1d (13.0 mg, 31% over two steps), mp = 115 °C. IR (ATR): 3211, 2964, 2936, 2875, 1693, 1605, 1501, 1458, 1440, 1426, 1368, 1320, 1225, 1168, 1123, 1080, 1002, 869, 799, 746, 705, 625 cm^–1. ¹H NMR (500 MHz, CDCl₃) δ: 8.06 (s, 1H), 7.43 (d, J = 8.0 Hz, 1H), 7.28 (bs, 1H), 7.24 (bs, 1H), 6.36 (s, 1H), 5.20 (d, J = 1.7 Hz, 1H), 3.92 (s, 3H), 3.04–2.95 (m, ‍4H), 1.54–1.43 (m, 4H), 0.89–0.78 (m, 6H) ppm. ¹³C NMR (125 MHz, CDCl₃) δ: 171.13, 167.22, 149.95, 133.11, 132.52, 131.49, 130.38 (q, J = 32.4 Hz), 124.18 (q, J = 272.5 Hz, ‒CF₃), 119.22, 118.20, 104.84, 93.19, 58.45, 54.84, 19.82, 11.79 ppm. HRMS (ESI/Q‐TOF) m/z: calcd. for C₁₉H₂₄F₃N₂O₂ [M+H]⁺: 369.17844, found: 369.17739.

3.2. Biology

3.2.1. Fluorescent Probes and Stock Solutions

Compounds 1a–1d were dissolved in DMSO to prepare stock solutions and stored at −20 °C protected from light. Working solutions were freshly prepared by dilution in serum‐free or serum‐containing culture medium immediately prior to use. Nile Red, BODIPY 493/503, LysoTracker Red DND‐99, MitoTracker Red CMXRos, ER‐Tracker Red, and CellTracker CM‐DiI were purchased from Thermo Fisher Scientific, while Rhodamine 123 was obtained from Sigma–Aldrich. Stock solutions were prepared according to the manufacturers’ instructions and stored at −20 °C protected from light. All probes were diluted in serum‐free medium immediately prior to use to the final concentrations specified in the corresponding experimental sections.

3.2.2. Cell Culture

Human NSCLC NCI‐H460, human lung adenocarcinoma A549, human glioblastoma A‐172, and human osteosarcoma SAOS‐2 cell lines were obtained from the American Type Culture Collection (ATCC, Rockville, MD, USA). NCI‐H460 cells were cultured in RPMI‐1640 medium (Corning) supplemented with 10% fetal bovine serum, 2 mM L‐glutamine, 100 U/mL penicillin, 100 µg/mL streptomycin, and 2.5 µg/mL amphotericin B. A549, A‐172, and SAOS‐2 cells were maintained in Dulbecco's modified Eagle medium (Sigma–Aldrich) supplemented with 10% fetal bovine serum, 2 mM L‐glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin. All cell lines were cultured in T25 flasks (Sarstedt) at 37 °C in a humidified atmosphere with 5% CO₂.

3.2.3. Resazurin Assay

Resazurin reduction assay was used to assess cellular metabolic activity following exposure to compounds 1a–1d. NCI‐H460 cells were seeded at 2000 cells per well in 96‐well plates and allowed to adhere for 24 h. Cells were then treated with compounds at concentrations of 0, 5, 10, 25, and 50 µM for 72 h.

For delayed cytotoxicity assessment, cells were exposed to 1a for 30 min, washed, and further cultured for 72 h prior to analysis.

Following incubation, resazurin solution (1% v/v; 10 mg/mL stock, Sigma–Aldrich) was added and incubated for 4 h at 37 °C. Absorbance was measured at 570 nm with a reference wavelength of 600 nm (Multiskan Sky, Thermo Scientific). Results were expressed as percentage resazurin reduction relative to untreated controls.

3.2.4. Flow Cytometry Analysis

Flow cytometry was used to assess cellular uptake of the compounds. NCI‐H460 cells were harvested with trypsin, centrifuged, and washed with serum‐free medium. Cells (2  × 10⁵ per condition) were then incubated with 25 µM of compounds 1a–1d in serum‐free medium for 30 min at 37 °C in the dark. Following incubation, cells were washed with PBS and fluorescence intensity was measured using a CytoFLEX flow cytometer (Beckman Coulter) in V450 (405/450 nm), B525 (488/525 nm), and B585 (488/585 nm) channels. At least 10,000 events were recorded per sample, and data were analyzed using CytExpert 2.4 software.

3.2.5. Optimization of 1a Cellular Staining Conditions

NCI‐H460 cells were seeded at 2000 cells per well in 96‐well plates and allowed to grow for 48 h. Cells were then stained with 25 µM 1a either in serum‐containing or serum‐free RPMI medium for 30 min at 37 °C in the dark. Fluorescence imaging was performed using an ImageXpress Pico Automated Cell Imaging System (Molecular Devices) with a 10× objective. Fluorescence of 1a was detected using the DAPI filter set under identical exposure settings for all conditions.

For concentration optimization, cells were incubated with 1a at 5, 10, 25, or 50 µM in serum‐free medium for 30 min at 37 °C in the dark. Imaging was performed using ImageXpress Pico Automated Cell Imaging System under identical channel and exposure settings across all conditions.

3.2.6. Analysis of Fixation Stability of 1a Fluorescence Signal

NCI‐H460 cells were seeded at a density of 2000 cells per well in 96‐well plates and cultured for 48 h under standard culture conditions. Cells were then stained with 1a in serum‐free medium for 30 min at 37 °C in the dark, followed by fluorescence imaging using an ImageXpress Pico Automated Cell Imaging System (Molecular Devices) with a 10× objective, with 1a signal detected using the DAPI filter set under identical acquisition settings across all samples. After initial imaging, cells were fixed with 4% PFA for 15 min at room temperature, washed with PBS, and reimaged using the same imaging system, objective, channel, and exposure parameters to ensure direct comparability between live and fixed conditions.

3.2.7. Analysis of Intracellular Distribution of 1a Using Organelle Markers

NCI‐H460 cells were seeded at a density of 2000 cells per well in 96‐well plates and cultured for 48 h under standard culture conditions. Cells were then washed with serum‐free RPMI and incubated for 30 min at 37 °C in the dark with 25 µM 1a together with organelle‐specific fluorescent probes. LDs were labeled with 1 µM Nile Red or 5 µM BODIPY 493/503, lysosomes with 500 nM LysoTracker Red DND‐99, mitochondria with 1 µM MitoTracker Red CMXRos or 1 µM Rhodamine 123, ER with 1 µM ER‐Tracker Red or 10 µM DiOC₆, and plasma membrane with 3 µM CellTracker CM‐DiI. Rhodamine 123 was obtained from Sigma–Aldrich, while all other fluorescent probes were purchased from Thermo Fisher Scientific. After incubation, cells were washed with PBS and imaged using an ImageXpress Pico Automated Cell Imaging System (Molecular Devices) with a 10× objective. Exposure times were optimized individually for each channel prior to acquisition. Fluorescence from 1a was recorded using the DAPI filter set, BODIPY 493/503 using the FITC filter set, and all other probes using the TRITC filter set.

3.2.8. Confocal Microscopy

For high‐resolution colocalization analysis, NCI‐H460 cells were seeded on 27 mm Nunc glass‐bottom dishes (Thermo Fisher Scientific) at a density of 30,000 cells per dish and cultured for 24 h. Cells were then washed with serum‐free RPMI and stained with 25 µM 1a, 5 µM BODIPY 493/503, and 1 µM Nile Red for 30 min at 37 °C in the dark. Confocal imaging was performed on a Leica TCS SP8 laser‐scanning confocal microscope (Leica Microsystems). Laser excitation wavelengths were 405 nm for 1a, 488 nm for BODIPY 493/503, and 561 nm for Nile Red. Emission detection windows were set to 420–480 nm for 1a, 500–540 nm for BODIPY 493/503, and 580–650 nm for Nile Red. The 1a detection window was kept narrow to avoid spectral overlap with BODIPY 493/503. Frame‐by‐frame sequential scanning was used, with the 1a channel acquired first, followed by Nile Red or BODIPY 493/503. Imaging was performed using an HC PL APO CS2 63×/1.40 oil‐immersion objective, with a pinhole set to 1 Airy unit, a pixel size of 80 nm, and z‐step intervals of 0.21 µm. Single‐stain controls were performed to assess channel bleed‐through. Samples stained with 1a alone were imaged in the Nile Red and BODIPY 493/503 channels, while samples stained with Nile Red or BODIPY 493/503 alone were imaged in the 1a channel to confirm the absence of signal in noncorresponding channels.

3.2.9. Induction of LD Accumulation with OA

OA (Sisco Research Laboratories) was dissolved in DMSO to prepare a 200 mM stock solution and conjugated to 5% bovine serum albumin (BSA) in PBS at 37 °C to obtain an OA:BSA complex at an approximate molar ratio of 10:1. The resulting solution was sterile‐filtered (0.22 µm) and diluted in serum‐containing culture medium to a final OA concentration of 400 µM.

NCI‐H460 cells were seeded at a density of 2000 cells per well in 96‐well plates for imaging or 100,000 cells per well in 6‐well plates for flow cytometry analysis. Cells were allowed to adhere for 24 h under standard culture conditions and then treated with OA at final concentrations of 200 or 400 µM. After 24 h of incubation, cells were washed with serum‐free RPMI for 90 min to remove excess BSA and subsequently labeled in serum‐free medium with 25 µM 1a for 30 min at 37 °C in the dark.

For imaging, cells were rinsed with PBS and imaged using an ImageXpress Pico Automated Cell Imaging System (Molecular Devices) with a 10× objective. For flow cytometry analysis, cells were harvested using trypsin, washed with PBS, and analyzed using a CytoFLEX flow cytometer (Beckman Coulter) with excitation/emission settings of 405/450 nm (V450 channel). At least 10,000 events were recorded per sample, and data were analyzed using CytExpert 2.4 software.

3.2.10. Zebrafish Maintenance and Imaging

Wild‐type zebrafish (AB) were raised and maintained at the Institute of Molecular Genetics and Genetic Engineering (University of Belgrade) in accordance with The Zebrafish Book guidelines [55]. The fish were housed in a circulating‐water system under a 14‐h light/10‐h dark cycle at 28 °C. The afternoon before mating, two males and three females were placed in a mating tank and separated by a divider. The next morning, at the onset of light, the divider was removed to allow mating. The eggs were collected 30 min after the divider was removed. Unfertilized eggs were removed, and the spawning with more than 80% of fertilized eggs was considered successful. Eggs were then maintained in the dark at 28.5 °C in embryo water (EW) (15 mM sodium chloride, 0.5 mM potassium chloride, 1 mM calcium chloride, 1 mM magnesium sulfate, 0.15 mM monopotassium phosphate, 0.05 mM ammonium phosphate, and 0.7 mM sodium bicarbonate) in Petri dishes, at a density of 30 embryos per dish. The EW was replaced daily.

Zebrafish husbandry was conducted in accordance with institutional and national ethical and animal welfare guidelines, harmonized with EU Directive 2010/63/EU (amended by Directive 2024/1262). The experiment did not need ethical approval since it was performed on 3‐day‐old larvae.

At 3 dpf, zebrafish larvae were collected and placed in a 96‐well plate (1 larva per well) and treated with 400 µM OA in EW for 2 h at 28 °C to promote LD formation. Next, the larvae were stained with 25 µM 1a and 1 µM Nile Red in EW for 30 min at room temperature in the dark.

For imaging, the larvae were anesthetized in the 96‐well plate using a buffered solution of MS‐222 (Sigma–Aldrich) at 0.1 mg/mL. Imaging was performed using the ImageXpress Pico Automated Cell Imaging System (Molecular Devices) with a 4× objective. Z‐stacks were acquired and projected for analysis.

3.2.11. Acute and Delayed Cytotoxicity Evaluation of 1a

NCI‐H460 cells were seeded in 96‐well plates at a density of 2000 cells per well and allowed to adhere for 48 h under standard culture conditions. Cells were then washed with serum‐free medium and incubated with 25 µM 1a for 30 min at 37 °C in the dark.

For acute cytotoxicity assessment, cells were costained with 1 µM propidium iodide (Thermo Fisher Scientific) to label dead cells, then washed with PBS and imaged using ImageXpress Pico Automated Cell Imaging System (Molecular Devices) with a 10× objective.

For delayed cytotoxicity assessment, cells were incubated with 25 µM 1a in serum‐free medium for 30 min at 37 °C in the dark, washed, then returned to serum‐containing RPMI and incubated for 72 h. Cell viability was subsequently assessed using the resazurin assay as described above.

3.2.12. Photostability and SNR Measurements

NCI‐H460 cells were seeded in 96‐well plates at a density of 2000 cells per well and allowed to adhere for 48 h. Cells were then washed with serum‐free medium and incubated with either 25 µM 1a, 1 µM Nile Red, or 5 µM BODIPY 493/503 in serum‐free medium for 30 min at 37 °C in the dark. After incubation, cells were washed with PBS and imaged using an ImageXpress Pico Automated Cell Imaging System (Molecular Devices) with a 10× objective.

For photostability analysis, time‐lapse imaging was performed over 10 min with 8 s intervals and 200 ms exposure time. For SNR analysis, cells labeled with each dye were imaged using identical acquisition settings with an exposure time of 70 ms.

3.2.13. Image Analysis

All image analysis was performed in Fiji (ImageJ, National Institutes of Health; version 1.54p).

In cell‐based assays, fluorescence intensity was quantified by manually defining regions of interest (ROIs) of identical size. Three ROIs were placed over fluorescent regions to measure signal intensity, and three additional ROIs were placed in cell‐free areas to determine background. Fluorescence values were background‐corrected and normalized to the number of cells within each ROI. Measurements were obtained from three independent images per condition and averaged. In zebrafish larvae, fluorescence intensity was quantified using the same approach, with five fields per condition analyzed from two independent images.

Colocalization analysis was performed using the JACoP plugin [56]. All measurements were performed on raw single optical sections. Pearson's correlation coefficient was used to quantify the linear relationship between fluorescence intensities in the two channels, with values ranging from +1 (perfect positive correlation) to −1 (perfect inverse correlation), and 0 indicating no correlation. Manders’ overlap coefficients (M1 and M2) were used to assess the degree of spatial overlap between signals, with M1 representing the fraction of 1a signal overlapping with the reference marker and M2 representing the fraction of the reference marker overlapping with 1a signal, both ranging from 0 (no overlap) to 1 (complete overlap). For each condition, 8 images were analyzed, and Pearson's correlation coefficient and Manders’ overlap coefficients (M1 and M2) were calculated and averaged.

For photostability analysis, time‐lapse image stacks (76 frames) were analyzed by measuring mean fluorescence intensity over time. Three independent ROIs were manually selected and tracked throughout the entire image sequence, and the average fluorescence decay was determined.

For SNR analysis, ROIs were manually selected for each image. Ten ROIs were placed over fluorescent regions to measure signal intensity, and 10 ROIs were positioned in cell‐free areas to determine background intensity. The average background intensity and its standard deviation were calculated, and SNR was determined as the difference between mean signal and mean background intensity divided by the standard deviation of the background.

3.2.14. Statistical Analysis

Data were analyzed using GraphPad Prism 8 software. Statistical significance in the OA‐induced LD accumulation experiment was assessed using a two‐way ANOVA (***p < 0.001), while the zebrafish, cell fixation, and delayed cytotoxicity experiments were analyzed using a two‐tailed Welch's t‐test (*p < 0.001).

4. Conclusion

In this study, four novel potential fluorescent probes were synthesized, out of which 1a was selected as having optimal properties and comprehensively evaluated for imaging applications across multiple biological models. The probe enabled robust visualization of intracellular LDs in living cells, demonstrated compatibility with standard fluorescence imaging platforms, and showed strong colocalization with established LD markers. Importantly, 1a exhibited sufficient photostability, adequate signal‐to‐noise characteristics, and minimal acute or delayed cytotoxicity under the tested conditions, supporting its suitability for live‐cell imaging. Importantly, 1a was also successfully applied in a zebrafish model, where it enabled visualization of lipid‐rich structures in a whole‐organism context. Its blue‐emitting properties enable multicolor imaging without spectral overlap with commonly used green‐ and red‐emitting probes. These results position 1a as a useful complementary tool for LD visualization in both in vitro and in vivo systems, with potential for further optimization and broader imaging applications. Its cost‐effective, straightforward, and sustainable synthesis, stable performance, and compatibility with multiple imaging modalities make it a useful addition to the expanding toolkit of chemical probes for studying LD‐associated cellular processes.

Funding

This study was supported by Ministry of Science, Technological Development, and Innovation of the Republic of Serbia (451‐03‐33/2026‐03/200007, 451‐03‐33/2026‐03/200042, 451‐03‐33/2026‐03/200168, 451‐03‐33/2026‐03/200288).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supplementary Material

Supplementary Material

Grozdanić Marija, Dinić Jelena, Jović Miloš, Podolski‐Renić Ana, Rankov Aleksandra Divac, Ljujić Mila, Pešić Milica, Opsenica Igor M., Selaković Života, Tetramic Acid‐Derived Blue‐Emitting Fluorescent Probe for Selective Lipid Droplet Imaging in Live Cells and Zebrafish, ChemMedChem 2026, 21, e70418. 10.1002/cmdc.70418

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.