Hybrid of Indolizine and Merocyanine—A New Class of Organelle‐Specific Dyes

Abstract

Unlocking the potential of electron‐rich 2‐hydroxyindolizines led to the discovery of a new class of dyes—indolizine‐merocyanines (IndMer) and indolizine‐cyanines. Tandem Friedel–Crafts alkylation followed by intramolecular nucleophilic aromatic substitution afforded structurally diverse dyes, as both the nucleophilic and electrophilic partners can be broadly modified. A convergent fragment coupling strategy allowed rapid access to these π‐conjugated merocyanines in three steps from pyridines. Uniform distribution of the HOMO and LUMO combined with negligible change of dipole moment upon excitation is responsible for the intense orange or red emission of this new family of reasonably photostable dyes in a broad range of solvents. The new merocyanine dyes have the potential to target a variety of organelles—both uncharged and positively charged indolizine–merocyanines localize a subset of cellular lysosomes, positively charged indolizine–cyanine hybrid accumulates in mitochondria, while coumarin–merocyanine shows context‐dependent localization to mitochondria and RNA‐rich nucleoli of the living cells.

The new merocyanines and cyanines possessing an indolizine scaffold were synthesized via tandem Friedel–Crafts alkylation followed by nucleophilic aromatic substitution, by leveraging the special reactivity of 2,3,5,6‐tetrafluoro‐4‐hydroxybenzaldehyde. Indolizine scaffold as a donor enables the bathochromic shift of emission independent of solvent polarity, whereas indolizine–cyanine selectively marks mitochondria or RNA‐rich nucleoli.

Introduction

Fluorescein, rhodamine, and rhodol were all discovered in the 19th century, and over the course of 140 years, extensive effort has been devoted to advancing the photophysical properties of these ubiquitous dyes through structural modifications. In particular π‐expansion^{[ 1} , ² , ³ , ⁴ , ⁵ , ⁶ , ^{7 ]} and replacement of the xanthene oxygen atom bridge with silicon,^{[ 8} , ⁹ , ¹⁰ , ^{11 ]} phosphorus,^{[ 12} , ^{13 ]} sulfur,^{[ 14 ]} or carbon^{[ 15} , ¹⁶ , ¹⁷ , ^{18 ]} in scaffolds of rhodamine,^{[ 19} , ²⁰ , ²¹ , ^{22 ]} fluorescein,^{[ 23} , ^{24 ]} and rhodol^{[ 25 ]} have proven popular and effective (Figure 1).

Figure 1.

Examples of core‐modified rhodamines and rhodols.

Other notable modifications encompass altering the size of the middle ring^{[ 26} , ^{27 ]} and replacing dialkylamino groups as a donor with indolizine,^{[ 28} , ^{29 ]} allowing further modulation of the photophysical properties.

The key application of xanthene dyes is fluorescence imaging, which has become a powerful tool for molecular biology due to its merits of real‐time monitoring capability, non‐invasive nature, and high spatiotemporal resolution.^{[ 30} , ^{31 ]} Compared to classical fluorescence microscopy, fluorescence imaging in the near‐infrared window (NIR, 750–1700 nm) allows for both deeper tissue penetration and a higher signal‐to‐noise ratio for noninvasive in vivo imaging at subcentimeter tissue depths, due to decreased scattering.^{[ 32} , ³³ , ³⁴ , ^{35 ]} Advancements in this imaging technique have demanded more stringent and challenging properties to be included in dyes, which currently available dyes do not entirely fulfill. Indeed, NIR dyes lack a large fluorescence quantum yield (Φ_fl) and photostability.^{[ 7} , ³⁶ , ³⁷ , ^{38 ]} Some of the modern, highly modified xanthene dyes possess emission in NIR often with reasonable quantum yield.^{[ 39} , ^{40 ]}

In the context of addressing the above challenges, an interesting question emerges about possible modifications of electron‐donating moiety especially by an electron‐rich aromatic heterocycle. The hypothesis underlying our current work is that by incorporating the entire indolizine^{[ 41 ]} moiety into the molecule as an electron donor, while simultaneously fusing it with the rest of the structure, a novel conjugated π‐system could be formed, potentially exhibiting remarkable photophysical properties (Figure 1). We aim to address this general idea by constructing a merocyanine,^{[ 42} , ^{43 ]} in which the electron‐rich indolizine scaffold would form one large, planar π‐expanded system. Needless to say, accomplishing this task would require establishing a brand‐new synthetic methodology. In the process we discovered several structurally unique organelle‐specific dyes with excitation in far‐red range.

Results and Discussion

Design and Synthesis

To overcome the aforementioned challenges, we propose a molecular design strategy that, compared to previous concepts, offers significant advantages in terms of simplicity and efficiency. Our key idea was to employ the tandem Friedel–Crafts alkylation/nucleophilic aromatic substitution, which we previously discovered for the reaction of 2,3,5,6‐tetrafluoro‐4‐hydroxybenzaldehyde (1) with aminophenols.^{[ 44 ]} Instead of benzene derivatives, we decided to use 2‐hydroxyindolizine as a nucleophilic coupling partner. Highly electron‐rich 2‐hydroxyindolizines can be obtained in situ using the methodology originally developed by Kakehi's group^{[ 45 ]} and optimized by us.^{[ 46 ]} Rewardingly, 2‐hydroxyindolizine 2, possessing an additional methyl group in position 3, reacted smoothly with aldehyde 1 under neutral conditions (xylenes, 80 °C, overnight) furnishing compound 3a in 72% yield (Scheme 1). A straightforward examination of the NMR data confirmed that the product has the expected indolizine‐merocyanine (IndMer) structure. Preliminary photophysical studies revealed strong fluorescence with an emission maximum (λ _em ^max) at 577 nm. With a proof of concept thus established, we began to explore this new core and its chemistry. We found that both the π‐expansion of the 2‐hydroxyindolizine substrate and its 3‐substituted derivatives are typically compatible with the new synthetic strategy, giving rise to various IndMer dyes 3b–f (Scheme 1). Mildly electron‐donating alkyl groups and electron‐withdrawing groups, such as an ester or pyridine, were well tolerated. On the other hand, when strongly electron‐donating groups were present at the pyridine core, the corresponding 2‐hydroxyindolizines could not be generated via Dieckmann condensation.

Scheme 1.

Straightforward synthesis of the indolizine–merocyanines (IndMer) and imidazo[1,2‐a]pyridine‐merocyanines (ImPMer).

The products 3b–f were obtained in 26%–90% yields and were conveniently harnessed by precipitation owing to their poor solubility in the reaction medium. The only exception was dye 3b, which was designed specifically to improve solubility and was purified by column chromatography. Having demonstrated that the synthesis of new merocyanines is straightforward and effective, we next turned our attention to an analogue possessing an additional nitrogen atom at position 3, i.e., a derivative of imidazo[1,2‐a]pyridine.^{[ 47} , ^{48 ]} To our delight, 2‐hydroxyimidazo[1,2‐a]pyridine was transformed into the corresponding dye 3Ng (ImPMer), under the same conditions as described earlier in 62% yield (Scheme 1). Due to the low yield and poor stability of other 2‐hydroxyimidazo[1,2‐a]pyridines, we decided to carry out the reaction in situ, starting with the corresponding 2‐aminopyridines. This approach allowed us to obtain two additional ImPMer dyes, 3Nh and 3Ni, with overall yields of 24% and 44%, respectively. Finally, IndMer 3e possessing pyridyl substituent was quaternized to obtain the positively charge salt 3eMe (Scheme 1, see Supporting Information for details).

Encouraged by these results, we also explored the possibility of using different types of aromatic aldehydes sharing the same features (i.e., halogen atom in vicinal position to CHO and electron‐deficient character of the aromatic ring) but possessing different electron‐donating groups at position 4. Reaction of indolizine 2b with 2,3,5,6‐tetrafluoro‐4‐(piperidin‐1‐yl)benzaldehyde (1N)^{[ 49 ]} unexpectedly led to product 4 in 10% yield, a dye formed by the reaction of two molecules of indolizine and one molecule of aldehyde, as confirmed by NMR and HRMS (Scheme 2). This structure can be compared with that of a V‐shaped bis‐xanthene dye.^{[ 4 ]} In contrast to other above‐described dyes, compound 4 is non‐planar and has a slightly twisted geometry because of the steric hindrance between the two indolizine units in close spatial proximity. By contrast, the reaction of coumarin‐based aldehyde 1C ^{[ 50 ]} led to the formation of cyanine‐type dye 5, in 27% yield, albeit both time and the temperature had to be extended (Scheme 2).

Scheme 2.

The synthesis of two indolizine‐cyanine hybrids 4 and 5 from aldehydes 1N and 1C.

To further demonstrate the robustness and synthetic utility of this protocol, we performed the synthesis of dye 3a on a 20 mmol scale, which gave 4.6 g of product (76% yield) indicating the synthetic potential of this method for bulk‐scale synthesis.

Photophysics

With a diverse series of new polymethine dyes in hand, we performed comprehensive photophysical characterization (Table 1 and Table S1, Figures 2 and 3 and Figure S1). UV–Vis absorption and emission spectra were recorded in DCM, THF, and DMSO for all dyes. The absorption and emission of this new family of dyes exhibited only a slight bathochromic shift (λ _abs ^max ≈ 570 nm vs. 550 nm, and λ _em ^max ≈ 580 nm vs. 540 nm) when compared to typical rhodols.^{[ 25} , ^{51 ]} In comparison to the majority of merocyanine dyes, IndMer and ImPMer surprisingly showed no solvatochromism. When comparing the photophysical properties of IndMer dyes with various substituents at position 3 (Figure 2), one can notice that the influence of electron‐donating alkyls and the electron‐withdrawing ester group is rather negligible. In contrast, insertion of a nitrogen atom at position 3 (ImPMers 3Ng, 3Nh, and 3Ni) causes the significant (≈30 nm) blue‐shift for both absorption and emission. It is also important to emphasize that fluorescence quantum yields (Φ_fl) are above 50% in most cases, even in DMSO. The Stokes shifts for all these dyes are around 500 cm⁻¹, consistent with the structural rigidity of this class of molecules and the negligible influence of solvent polarity on this π‐conjugated system.

Table 1.

Photophysical properties of synthesized dyes in DCM and DMSO.

		λ abs max (nm)	ε max (M−1cm−1)	λ em max (nm)	Φ fl (%)
3a	DCM	570	87 000	577	74
3a	DMSO	567	75 000	580	53
3b	DCM	571	97 000	579	77
3b	DMSO	567	70 000	584	62
3c	DCM	570	56 000	578	67
3c	DMSO	567	44 000	583	40
3d	DCM	557	30 000	569	57
3d	DMSO	550	40 000	566	49
3e	DCM	575	14 000	593	67
3e	DMSO	569	42 000	593	56
3f	DCM	554	30 000	567	53
3f	DMSO	545	39 000	562	50
3Ng	DCM	541	10 000	565	36
3Ng	DMSO	530	30 000	558	29
3Nh	DCM	543	37 000	566	42
3Nh	DMSO	531	37 000	557	35
3Ni	DCM	541	29 000	565	42
3Ni	DMSO	533	33 000	558	35
4	DCM	537	41 000	585	32
4	DMSO	531	20 000	600	5
5	DCM	624	31 000	705	2
5	DMSO	616	8000	735	<1
3eMe	DCM	554	33 000	570	43
3eMe	DMSO	554	36 000	570	56

Figure 2.

UV–vis absorption spectra (solid lines) and fluorescence spectra (dashed lines) of compounds 3a, 3d, 3Ng, and 3f measured in DCM.

Figure 3.

UV–vis absorption spectra (solid lines) and fluorescence spectra (dashed lines) of compounds 4 and 5 measured in DCM.

Bis‐indolizine 4 (Figure 3), structurally comparable to a V‐shaped dimeric xanthene dye,^{[ 4 ]} displays a bathochromically shifted λ _em (585 nm vs. 547 nm) with a small decrease in Φ _fl (32 vs. 40%). The properties of indolizine‐cyanine 5 (Figure 3) are strikingly different—λ _abs ^max is shifted to 624 nm and λ _em ^max is beyond 700 nm, although fluorescence quantum yield is rather low (2% in DCM). For selected dyes spectra were also recorded in H₂O/DMSO to simulate biological environment (see Table S1). In the case of merocyanines 3Nh and 4 Φ _fl was comparable to results in pure DMSO, whereas for merocyanines 5 and 3eMe it was markedly larger. Only in the case of merocyanine 3b emission in H₂O/DMSO was weaker than in other solvents.

The comparison between the strong prompt fluorescence observed for compounds 3a, 3b, 3Ng, and 5, and the very weak signal (close to noise level) detected in the delayed emission clearly indicates that none of the studied dyes exhibit phosphorescence, even at 5 K (Figure S3).

To assess the photostability of compounds 3eMe, and 5 in aqueous medium, their solutions were irradiated with a Xe lamp for approximately 2 h, while monitoring the absorption maxima over time (Figure 4 and Figure S4). The photostability of Rhodamine 6G was also evaluated under the same conditions and used as a reference. In general, the new merocyanines showed slightly lower, yet comparable, photostability relative to the reference. All tested compounds exhibited a limited (less than 10%) and quasi‐linear decay with minor deviations. Notably, compound 3eMe demonstrated the highest stability. Photostability was also studied in DCM and DMF, and those cases, Bodipy was used as a reference (Figure S4).

Figure 4.

Photostability of dyes 3eMe and 5 compared to Rhodamine 6G in H₂O + 0,5% DMSO.

The limited photostability of new merocyanines prompted us to study what are the photodecomposition products. The photodegradation of xanthene dyes was investigated numerous times^{[ 52} , ⁵³ , ^{54 ]} although typically these studies focused either on improving the photostability or on halogenated derivatives.^{[ 55 ]} The solution of dye 3b was irradiated in quartz vessel using a xenon lamp. It turned out that that the number of photodecomposition products exceeded 20 species and purification of the vast majority of them proved impossible because of their small quantities and the tendency to decompose further during column chromatography. Most of them were more polar than dye 3b and displayed strong blue fluorescence. For one crude decomposition product, both the molecular mass and the number of aromatic signals suggested the dimeric structure. The chemical shift of the aromatic signals, along with the observation of strong blue fluorescence, suggests the presence of an isolated indolizine chromophore. The analysis of ¹H NMR spectrum as well as MS proved inconclusive, although several hypothetical structures were suggested (see Supporting Information).

Computational Studies

To understand how differences in the structure relate to the observed photophysics, we performed computational studies for a selection of dyes (3a, 3d, 3Ng, 5) via DFT and TDDFT O3LYP and M06/6–31G(d,p) methods.^{[ 56} , ⁵⁷ , ⁵⁸ , ^{59 ]} The calculation results are collected in Tables S3–S8 and Figures S6–S9. Table S3 provides data on the absorption and fluorescence energies and oscillator strengths characterizing the molecules optimized in the ground – S ₀ and electronically excited – S ₁ states, respectively.

The calculations were performed with two functionals. The transition energies and oscillator strengths calculated with the O3LYP functional are slightly smaller than those calculated with M06, but the results obtained with both functionals correlate with each other (Figure S6).

Moreover, calculations with both functionals consistently lead to the conclusion that the studied group of dyes are characterized by a large dipole moment in the ground S ₀ state (16–19 D), which changes little in the excited S ₁ state (see Table S3). From the point of view of the classical theory of the solvent effect on electronic spectra,^{[ 60 ]} this feature means that the solute–solvent interaction energies in both electronic states are comparable, and therefore the polarity of the solvent should not affect the absorption and fluorescence energies, as confirmed in Table 1.

Minimizing solvent impact created conditions for partial extraction of the vibrational structure of absorption and fluorescence spectra, usually blurred by polarity effects.^{[ 61 ]} Based on the calculation results for the optimized structures of the molecule in the ground and excited states, the absorption and fluorescence spectra can be simulated (see Supporting Information for details).^{[ 57} , ^{58 ]} Figure S8 shows a simulation of the absorption spectra of 3a, 3d, and 3Ng. As seen, these simulations quite accurately reflect the positions and relationships between the spectra obtained experimentally (Figure 2). The ability to reproduce the vibronic structure of dyes provides information about the structural changes in the excited molecule. For example, the Franck–Condon factors (base for simulation – Figure S7) are an important element in the expressions for the rate constants of nonradiative processes.^{[ 62 ]} Without going into details and to simplify, the higher the ratio of the intensity of the vibronic structure band relative to the first (0,0) band, the higher the FC factor. Thus, the intensity distribution in the 3a, 3d, and 3Ng spectra leads to the expectation of an increase in the rate constant of nonradiative transitions in this ordering. It should be added here that the vibronic structure in the spectra of the studied group of compounds is much less extended than in the spectra of aromatic hydrocarbons, i.e., their geometry changes are much smaller during electronic excitation.

Figure 5 shows the calculated energy diagram of molecule 3a together with the simulations of absorption and fluorescence spectra. Figure 5 shows the HOMO and LUMO orbitals of molecule 3a, as the single configuration (HOMO, LUMO) describes the electronic transition between the S ₀ and S ₁ states. This is a π–π* transition with large oscillator strengths (Table S3). As can be seen, both the HOMO and LUMO are characterized by uniform density distributions over the entire molecule.

Figure 5.

(bottom) Diagram of energy, HOMO and LUMO orbitals, and dipole moments for 3a optimized in S ₀ and S ₁ electronic states. In the center of the figure, the vibronic states of the molecule in S ₀ and S ₁ are symbolically shown. Transitions between them, with intensities described by Franck–Condon factors, build absorption and fluorescence spectra—their simulation is shown at the top of the figure. The strongest line in both spectra is the E(0,0) line, corresponding to the transition between n = o and m = 0, which means that the structure of 3a undergoes relatively small changes at the transitions between S ₀ and S ₁. Detailed data on the FC factors and spectra simulation are in the Supporting Information.

The shapes of the HOMO and LUMO orbitals for the other molecules considered here, as well as the shapes of other frontier orbitals, HOMO −1 and LUMO +1, along with data on their energies, are shown in Table S6. It allows us to claim that the shapes of the HOMO and LUMO of parent dye 3a are preserved with substitutions or π‐expansion of the IndMer chromophore; however, the effects of structure modifications are visible in changes of the shapes of HOMO −1 and LUMO +1.

Therefore, more pronounced effects of structure modification may occur in processes involving these orbitals. Such a process is intersystem crossing (ISC) S ₁→T ₂, because the T₂ state is based on the configuration of the HOMO −1 and LUMO. According to the calculation results, the T ₂ energy in molecule 3a is clearly lower than the S ₁ energy, but the energy gap decreases with modifications of the molecular structure (Table S5). The spin—orbit coupling elements between S ₁ and T ₂ are not large, but in conditions close to resonance, the intersystem crossing process can be triggered. The results of O3LYP and M06 calculations are not completely consistent, but based on the O3LYP data, in the cases of 3Ng and 5, it is likely that the reason for the experimentally observed decrease in fluorescence efficiency is the S ₁→T ₂ ISC process. In summary, the strong spatial overlap of HOMO and LUMO combined with the lack of change of dipole moment in the excited state are responsible for the strong emission of IndMer and ImPMer dyes in polar solvents. In addition, strong emission is facilitated by good separation and lack of coupling of the S ₁ state with other electronic states.

Fluorescence Imaging

The aqueous solubility of compounds 4, 5, and 3b was assessed (Table S2), with values ranging from 0.28 mg per 100 mL (3b) to 2.31 mg per 100 mL (5), corresponding to approximately 10–80 µM. These results confirm the suitability of these compounds for bioimaging experiments.

Confocal microscopy studies in HT1080 fibrosarcoma and MRC5 normal human lung fibroblast cells showed that merocyanine hybrid dyes are cell‐permeable and have the potential to target a variety of organelles in live cells. Specifically, both uncharged and positively charged indolizine‐merocyanines 3b, 3e, 3Nh, and 3eMe localize to cytoplasmic puncta consistent with localization to endosomes and lysosomes (Figures S10–S12; but not significantly to lipid droplets, see Figures S13 and S14), positively charged indolizine‐cyanine hybrid 4 accumulates in mitochondria (Figure S15), while related dye 5 shows context‐dependent localization to mitochondria and RNA‐rich nucleoli (Figures S16–S18). The latter observation aligns with recent reports of single fluorescent probes exhibiting dual localization in mitochondria and nucleoli through mechanisms such as pH‐responsiveness, concentration‐dependent redistribution, and energy‐sensitive migration, with some dyes further enabling density mapping of subcellular environments.^{[ 63} , ⁶⁴ , ⁶⁵ , ⁶⁶ , ⁶⁷ , ^{68 ]} This dual localization ability provides a valuable platform for studying mitochondrial damage and monitoring mitochondrial state in live cells under physiological and stress conditions.

Despite maximal λ _abs ≈ 530–575 nm, uncharged dyes 3b, 3e, and 3Nh displayed an unexpected increase in the brightness of green emission in live cells (Figure S10), suggesting that these dyes may be sensitive to their local environment, such as organelle pH, protein–lipid interactions, dye aggregation, or spatial confinement. Although positive charge is often associated with mitochondrial localization, in the case of our dyes this correlation appears more nuanced and is not strictly predictive. As mentioned above, 3eMe localizes to a subset of lysosomes (Figure S12), while dye 4 exclusively stains the mitochondria. The Pearson's correlation coefficient between dye 4 and MitoTracker Green FM was r = 0.94 and 0.92 for HT1080 and MRC5 cells, respectively. Despite the low quantum yield for dye 5 in vitro, we observed unexpected bright fluorescence in live cells, specifically in mitochondria and nucleoli (Figures S17 and S18). On one hand, reproducible but low‐intensity mitochondrial labeling was achieved when using low concentrations of molecule 5 in the presence of serum (FluoroBrite DMEM supplemented with 5% FBS, 1 h at 37 °C, 200 nM) (Figure 6 and Figure S16). Mitochondrial localization was confirmed by a high Pearson's correlation coefficient between the fluorescence of dye 5 and the MitoTracker Green FM for HT1080 (r = 0.89) and slightly lower values for MRC5 cells (r = 0.60). On the other hand, with higher concentrations of dye 5 in serum‐free media (FluoroBrite DMEM, 1 h at 37 °C, 1 µM), we observed specific nucleolar staining reminiscent of RNA‐selective dyes, such as SYTO RNASelect (Figures S17 and S18). Both mitochondria and nucleoli are RNA‐rich, suggesting that the localization of dye 5 is influenced not solely by its delocalized cationic charge, but also by its planar structure, which may promote selective—and possibly fluorogenic—interactions with RNA, a hypothesis to be explored in future studies. In addition, dye 5 displayed excellent photostability in live cells retaining ≈ 93% of its fluorescence after 5 min of continuous imaging, while dye 3eMe showed above‐average photostability retaining ≈ 50% of its fluorescence in similar experiments (Figures S19 and S20). Importantly, all dyes were used for organelle staining at concentrations that did not noticeably affect cell viability (Figure S21).

Figure 6.

Subcellular localization of dye 5 in living HT1080 fibrosarcoma cells. Cells were loaded with 200 nM dye 5 (magenta), along with the 50 nM MitoTracker Green FM dye (green), and images were recorded using confocal fluorescence microscopy. The fluorescence of MitoTracker Green FM (green) was recorded with excitation at 487 nm and an emission range of 500–550 nm, and the fluorescence of molecule 5 (magenta) was recorded with excitation at 638 nm and an emission range of 663–738 nm. The brightness and contrast for individual color channels were adjusted and advanced denoising filter was applied using NIS Element's software to improve visualization of colocalization. Colocalized pixels appear white in overlays, and colocalization is assessed by Pearson's correlation coefficient (r), see Figures S16–S18 for more details. Scale bars are 10 µm for the main images and 5 µm for the insets.

Conclusion

It is possible to directly condense 2,3,5,6‐tetrafluoro‐4‐hydroxybenzaldehyde with 2‐hydroxyindolizines in a tandem process embracing Friedel–Crafts alkylation and nucleophilic aromatic substitution. Highly electron‐rich 2‐hydroxyindolizines serve as analogues to 3‐dialkylaminophenols enabling the formation of heretofore unknown hybrid dyes—indolizine‐merocyanines (IndMer). This strategy can be extended to π‐expanded 2‐hydroxyindolizines and 2‐hydroxyimidazo[1,2‐a]pyridines. Replacing 2,3,5,6‐tetrafluoro‐4‐hydroxybenzaldehyde with 3‐formyl‐4‐chloro‐7‐dialkylaminocoumarin enables the synthesis of positively charged indolizine‐cyanine. Salient features of our approach include exceptionally simple synthesis, broad structural tolerance, and compatibility with a large scale. Indolizine and imidazo[1,2‐a]pyridine as donors are game changers in the design of merocyanines: in a simple manner, a diverse range of strongly orange‐red‐emitting dyes can be prepared. Their fluorescence is mostly resistant to solvent polarity, and they possess reasonable photostability. Replacing the hydroxy group with an amino group at the para position of the electrophilic partner diverges the reaction pathway and a helicene‐like bis‐indolizine is formed, which emit around 600 nm. The simultaneous manipulation of various structural details makes it possible to obtain dyes with large molar absorption coefficients, fluorescence quantum yields > 50% and good photostability. TD‐DFT calculations rationalize their strong emission in polar solvents. Positively charged indolizine–cyanine dyes selectively mark mitochondria or RNA‐rich nucleoli of the living cells, whereas neutral ones permeate the membranes or enter through endocytosis and localize in a subset of cellular endosomes and lysosomes. We expect that the introduction of this new structural concept within the merocyanine family will open up an innovative avenue for dye chemistry.

Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

Supporting Information

Dedicated to Prof. Jonathan Lindsey on his 70^th birthday

Badaro J. S. A., Koszarna B., Perkowska M., Kandere‐Grzybowska K., Kolygina D. V., Morawski O., Park J., Grzybowski B., Deperasińska I., Gryko D. T, Angew. Chem. Int. Ed.. 2025, 64, e202508044. 10.1002/anie.202508044

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.