A Modular Approach for the Synthesis of Diverse Heterobifunctional Cyanine Dyes

Abstract

Herein, we present a straightforward synthetic route for the design and synthesis of diverse heterobifunctional cyanine 5 dyes. We optimized the workup by harnessing the pH- and functional group-dependent solubility of the asymmetric cyanine 5 dyes. Therefore, purification through chromatography is deferred until the last synthesis step. Demonstrating successful large-scale synthesis, our modular approach prevents functional group degradation by introducing them in the last synthesis step. These modifiable heterobifunctional dyes offer significant utility in advancing biological studies.

Introduction

Cyanine dyes are renowned for their excellent spectral properties including broad wavelength tunability, sharp fluorescence bands, and high sensitivity. These attributes render them indispensable in various biomedical applications.^1,2 Due to their versatility compared to other organic fluorophores, they are especially utilized as photocages,^3,4 in biomedical imaging and biomolecule labeling, for example in single-molecule fluorescence microscopy (SMFM)⁵⁻⁷ or fluorescence resonance energy transfer (FRET),^8,9 nucleic acid detection,¹⁰ and cancer imaging,¹¹ as well as biomedical screening techniques¹² and analyte-responsive fluorescent probes in optoelectronic applications.¹³⁻¹⁵ Furthermore, their ability to photoswitch makes them well-suited for stochastic super-resolution microscopy techniques such as STORM and dSTORM.¹⁶

Despite numerous synthetic strategies for symmetric or monofunctional dyes published in recent years,¹⁷⁻²⁰ the synthesis and purification of cyanine dyes with modifiable functional groups remains a major challenge.²¹ Moreover, there are only a few reported examples of asymmetric pentamethine cyanine dyes with modifiable functional groups and less than 10 examples of heterobifunctional cyanine 5.^14,16,22 Furthermore, they involve the use of high boiling point solvents, as well as reversed-phase chromatography (RPC) and/or preparative high-performance liquid chromatography (HPLC) purification steps, resulting in low yields and a limited synthesis scale. Also, methodologies including solid-phase approaches or the application of microwave assistance have been explored.^{2,14−16,23,24}

Heterobifunctional fluorescent dyes have garnered interest in recent years due to their utility in studying interactions between biomolecules and small molecules. By incorporating chemically compatible reactive functional groups, they can be used as cross-linking systems, labels, tags, and probes to study and quantify biological systems.^25,26 A comprehensive understanding of protein functions and interactions is critical for deciphering the role of protein networks in human physiology and pathology. For example, protein tagging systems, such as the HaloTag or SNAP-Tag technology, facilitate the study of biological systems at the protein level.²⁷ Comprising two elements, which are a protein tag that can be fused to any protein of interest (POI) and a ligand, the technology enables irreversible binding of the ligand to the protein tag. This ligand can be further modified to introduce various entities including fluorescent dyes, affinity handles, or solid supports.²⁸ This diversification of the ligand extends the range of applications, encompassing in vivo molecular imaging, in vitro cellular imaging, evaluation of protein function, analysis of molecular interactions, and protein assays.²⁷

In light of this, we report a simple and efficient synthetic strategy for heterobifunctional cyanine 5 dyes (Scheme 1). These dyes incorporate a ligand capable of covalently binding to a protein tag (haloalkane for HaloTag, benzylguanine for SNAP-tag), along with another reactive functional group. This second functional group enables the attachment to a desired molecule of interest through amide bond formation, bioorthogonal click chemistry, or esterification. These dyes can be used as versatile tools for linking biomolecules of interest, facilitating in-depth studies. Within this work, we provide the detailed synthetic strategy of these dyes as well as their spectroscopic characterization.

Scheme 1. General Synthesis Pathway for the Formation of Heterobifunctional Cyanine 5 Dyes.

(i) Indolium derivative formation. (ii) Hemicyanine formation. (iii) Cyanine 5 dye formation. (iv) Functional group attachment.

Synthesis

The conventional method for cyanine dye preparation involves a stepwise condensation reaction of two nucleophilic aza-heterocycles with a polyene-chain precursor.²⁹ The formation of the aza-heterocycle salt via N-alkylation necessitates high temperatures, risking the decomposition of temperature-sensitive functional groups.¹⁶ To address this challenge, we chose a modular approach where the delicate functional groups were attached in the last synthesis step. Commencing with a commercially available indolenine derivative, we synthesized the carboxy-indolium precursor 5 (Scheme 2). We tested the reaction in various solvents such as toluene³⁰ or acetonitrile, as well as in solvent-free conditions, with the highest yield and minimal side product formation observed in acetonitrile. The highly water-soluble precursor was purified by uptake in water, followed by washing the aqueous phase with different organic solvents of increasing polarity. After lyophilization, the resulting product exhibited sufficient purity for the subsequent synthesis step.

Scheme 2. Synthesis of the Indolium Precursors 5, 6a, and 6b.

Reagents and conditions: (a) 2.00 eq. bromopropionic acid, ACN, reflux, 16 h (88%). (b) (6a): 1.00 eq. EDC*HCl, 1.00 eq. HOBt, 1.00 eq. Et₃N, 1.00 eq. of the HaloTag ligand 12, dry DCM, rt, 1h (57% crude product). (6b): 1.40 eq. EDC*HCl, 1.40 eq. HOBt, 1.40 eq. NMM, 1.40 eq. of indolium derivative 5 and 1.00 eq. of 6-((4-(aminomethyl)benzyl)oxy)-9H-purin-2-amine, dry DMF, RT, 7.5 h (62% crude product).

The methyl group in position 2 exhibited increased reactivity, which can be attributed to the positive charge at the quaternary nitrogen atom.^31,32 This reactivity was evidenced by proton exchange with deuterium ions, observed through the disappearance of the signal corresponding to the methyl group at position 2 of the carboxy-indolium precursor 5 when studied in polar, protic NMR solvents like methanol-d₄.

To achieve heterobifunctionality, we further modified the carboxy-indolium derivative 5 by attaching a haloalkane (Scheme 2), which constitutes the ligand of the HaloTag. The haloalkane 12 (Scheme S.1) featuring a reactive amine group at its distal end was synthesized following standard literature procedures with minor adjustments.^33,34 However, the positive charge introduces instability to the carboxy-indolium derivative 5 under overly harsh conditions, which could lead to the potential cleavage of the side chains at the nitrogen centers of the indole ring.¹ To mitigate this risk, we optimized the condensation reaction between the carboxy-indolium derivative 5 and the amine group of the haloalkane 12. The conversion of the carboxylic acid to its amide was achieved using activating agents, additives, and non-nucleophilic bases. We tested several combinations of activating agents and non-nucleophilic bases in different solvents and monitored the stability of the carboxy-indolium derivative 5 or the asymmetric cyanine dye 1 (Table 1). Stronger activating agents like DCC, HATU, HBTU, or TSTU as well as the combination with stronger non-nucleophilic bases used in excess resulted in the cleavage of the side chain. Solvent selection showed minimal impact on the reaction outcome, prompting us to choose DCM over DMF for its ease of removal. The best results were obtained using EDC*HCl with HOBt and DIPEA or Et₃N in DCM. It is noteworthy that the reaction with the carboxylic acid formed a highly reactive O-acylisourea intermediate. Without the addition of HOBt, we observed the conversion of this active O-acylisourea intermediate to the corresponding inert N-acylurea via an intramolecular acyl O → N migration, which is a common side reaction in peptide couplings.³⁵ The addition of HOBt yielded an active ester that is less reactive than the urea derivative, preventing unwanted side product formation.³⁶ Following the workup, the HaloTag-ligand-bearing indolium derivative 6a was used without further purification and analysis.

Table 1. Tested Reaction Conditions for Amide Coupling and Esterification Reactions^a.

entry	reagents	base	conditions	result
1	1.0 eq. DCC	3 mol %DMAP	0 °C–RT (ice bath), 2 h	cleavage of carboxy group
2	1.0 eq. DCC 1.0 eq. NHS		RT, 17 h	no full conversion to NHS ester
3	1.5 eq. DCC 1.5 eq. NHS	1.0 eq. NMM	RT, 16 h	cleavage of carboxy group
4	1.2 eq. EDC		0 °C–RT (ice bath), 16 h	N-acylurea derivative
5	1.0 eq. EDC*HCl 1.0 eq. HOBt	1.1 eq. DIPEA	RT, 5 h	trace amounts of cleavage product, efficient conversion
6	1.0 eq. EDC*HCl 1.0 eq. HOBt	2.0 eq. NMM	RT, 5h	trace amounts of cleavage product, slow conversion
7	1.0 eq. EDC*HCl 1.1 eq. HOBt	1.0 eq. TEA	RT, 16 h	cleavage of carboxy group for asymmetric dye 1, efficient for indolium derivatives 6a and 6b
8	1.0 eq. HATU 1.0 eq. HOAt	1.0 eq. NMM	RT, 16 h	cleavage of carboxy group
9	1.0 eq. HBTU	0.5 eq. NMM	RT, 4 h	cleavage of carboxy group
10	1.0 eq. PyBOP	1.0 eq. DIPEA	RT, 90 min	cleavage of carboxy group
11	1.0 eq. PyBOP	1.0 eq. NMM	RT, 90 min	cleavage of carboxy group
12	1.0 eq. PyBOP	0.5 eq. NMM	RT, 4 h	no conversion, cleavage of carboxy group
13	1.1 eq. TSTU	1.1 eq. DIPEA	RT, 20 min	cleavage of carboxy group
14	1.0 eq. TSTU	2.0 eq. NMM	RT, 16 h	slow conversion, cleavage of carboxy group overnight

^aStability and conversion were monitored via HPLC-MS. The most efficient conditions are highlighted with black boxes (entries 5 and 7).

Subsequently, malonaldehyde dianilide 7 (Scheme 3 and Scheme S.1) was synthesized following standard literature procedures via condensation of malondialdehyde bis(dimethyl acetal) with aniline under acidic conditions in high yield.³⁰ In some literature protocols, the crude product is subjected to multiple washing steps with water, which is designed to eliminate residual acid without compromising purity. However, we discovered that malonaldehyde dianilide exhibited partial solubility in water. To address this, we opted for a single wash with ice-cold water.

Scheme 3. Synthesis of Hemicyanine 8 and Asymmetric Cyanine Dye 1.

Reagents and conditions: (a) 1.10 eq. of aniline derivative 7, 15 eq. acetic anhydride in acetic acid, 110°C oil bath, 5 h (used without further purification). (b) 1.00 eq. of the indolium derivative 5, 2.40 eq. sodium acetate in MeOH, RT, 3 h (81%).

In the next reaction step, the HaloTag-ligand-bearing indolium derivative 6a attacks the imine of malonaldehyde dianilide 7 under relatively harsh conditions (Scheme 3), in that the synthesis of the hemicyanine requires elevated temperatures using a mixture of acetic acid and acetic anhydride,^2,15 where the latter serves to generate a good leaving group through the acetylation of the imine in the subsequent reaction step.¹⁶ Due to its susceptibility to decomposition, especially in protic solvents,¹⁶ hemicyanine 8 was promptly utilized in the following synthesis step without undergoing further purification. Maintaining an excess of the indolenine derivative in this synthesis step can lead to the formation of the symmetric dye in the next reaction step. Separation of this symmetric dye can be challenging and would require an additional purification step. However, excess malonaldehyde dianilide 7 can be removed easily by aqueous workup and was used therefore in slight excess (1.1 equiv). Reaction monitoring via HPLC-MS is therefore crucial in this synthesis step.

The formation of the asymmetric cyanine dye 1 is known to occur readily in the subsequent reaction under relatively mild conditions. The activation of the carboxy-indolium derivative 5 with sodium acetate induces nucleophilic substitution of acetophenone at room temperature. For the detailed aspects of this process, we refer to the analysis conducted by Wolf et al.¹⁶ Moreover, an earlier publication offers valuable structural insights³⁷ while another study delves into the optical properties of these dyes.³⁸ Together, these works contribute a multifaceted understanding of cyanine 5 dyes, encompassing their synthesis mechanisms, structure, and optical characteristics.

The presence of a positive charge introduces complexity to the purification process, potentially leading to a reduction in yield. Additionally, side and cleavage products (such as asymmetric dye 1 without the carboxylic acid group) exhibit similar retention times to the main product. However, their solubility in organic solvents differs slightly due to the different polarities of the side chains. This allowed us to streamline the workup steps, capitalizing on their pH- and functional group-dependent solubility. Following the workup, the formed acetophenone byproduct was removed through washing with ice-cold diethyl ether. The resulting blue solid was obtained in 81% yield and 88% purity, as determined by HPLC.

The entire synthesis strategy leading to cyanine 5 dye 1 has been systematically optimized here for larger scales, reaching up to low-gram quantities. Usually, the synthesis of dyes or fluorophores, particularly cyanine 5 dyes, is conducted in relatively small scales^14,16,17,19 due to factors such as the increase in side product formation during upscaling and the high polarity. These characteristics contribute to the complexity of purification under standard column conditions, with a notable challenge being the irreversible adsorption of substantial amounts of the dye in classical silica gel chromatography. While reversed-phase chromatography, utilizing commercial C-18 absorbents, is often the method of choice for purification,^22,39 it proves relatively expensive, especially for larger scales, and is not commonplace in a standard laboratory. Alternatively, preparative HPLC is commonly employed, resulting in highly pure compounds. Nevertheless, drawbacks include the limited amount of crude product applicable for HPLC purification and the irreversible adsorption of significant dye quantities on the column.

Contrary to previous methods, our approach involves the use of activated basic aluminum oxide for low-pressure column chromatography. This alternative is cost-effective compared to C-18 adsorbents and less polar than silica gel, which is advantageous for purifying highly polar compounds. Furthermore, due to the weakly acidic nature of silica gel, there is a tendency toward preferential adsorption of basic compounds relative to their adsorption on neutral and basic adsorbents. Here, as eluent, a mixture of DCM/MeOH proved optimal, as the adsorption of polar compounds is reduced in a polar medium.⁴⁰ We tested the purification conditions for cyanine dye 1. Impurities were removed by using a gradient of DCM/MeOH and the product was obtained by eluting with the addition of 1% triethylamine to the eluent mixture (a purity of >99% can be achieved for the cyanine dye 1; see Figure S.5).

To further demonstrate the versatility of our synthetic strategy, our objective was to introduce different functional groups in the final synthesis step. The desired groups were incorporated through amide coupling or esterification, yielding azido-homoalanine derivative 2 and propargyl ester derivative 3 (Scheme 4).

Scheme 4. Further Modification of the Asymmetric Cyanine 5 Dye 1.

Reagents and conditions: (a) (2): 1.00 eq. of the azido-homoalanine 14, 1.20 eq. EDC*HCl, 1.20 eq. HOBt, 1.20 eq. DIPEA in DCM, RT, 16 h (30%), R₂=azido-homoalanine (-Homoala-(OtBu)-N₃). (3): 2.00 eq. of propargyl alcohol, 1.00 eq. EDC*HCl, 1.00 eq. HOBt, 1.00 eq. DIPEA in dry DCM, RT, 16 h (34%), R₃= propargyl ester (−OCH₂C≡C).

To incorporate azido-homoalanine (compound 2), we introduced a tert-butyl group to protect the carboxylic acid and removed the Fmoc group from commercially available Fmoc-γ-azido-homoalanine, yielding the amino acid bearing a free amine group (14, Scheme S.1). As previously mentioned, the coupling conditions to cyanine 5 derivative 1 required optimization (Table 1) due to the instability of the dye toward harsh reaction conditions. For the synthesis of the propargyl ester derivative 3, we employed conditions similar to those used for the synthesis of 6a, since the standard Steglich esterification conditions⁴¹ resulted in cleavage of the carboxylic acid of the asymmetric dye. Nevertheless, in the final reaction step, side reactions for compounds 2 and 3 could not be entirely prevented. Consequently, the crude products were directly transferred to a column, as there was no discernible advantage in conducting prior workup. Azido-homoalanine derivative 2 was purified using conditions as described for compound 1, employing a less polar eluent mixture, and a yield of 30% with 91% purity was achieved.

Propargyl ester derivative 3 turned out to be less stable. Utilizing activated basic or neutral aluminum oxide for the purification of the dye led to the cleavage of the ester group. Consequently, column chromatography was conducted with standard silica gel, employing a gradient of DCM/MeOH, and resulted in 34% yield with 96% purity.

In addition to the HaloTag, other protein self-labeling systems such as the SNAP-Tag or a modified version of it, the CLIP-Tag, are widely employed in biochemical and cellular contexts.⁴² We were interested in exploring whether our synthetic strategy could be extended to introduce an O⁶-benzylguanine derivative, which is commonly used as a SNAP-Tag ligand.

The modification of the carboxy-indolium derivative 5 involved the attachment of the O⁶-benzylguanine derivative, employing the same coupling reagents as those used for the HaloTag ligand (Scheme 2). Instead of Et₃N, a milder base (NMM) was necessary to prevent degradation of the O⁶-benzylguanine derivative. Given the poor solubility of the O⁶-benzylguanine derivative in DCM, THF, or a mixture of DCM/THF, DMF was used as the solvent. The resulting SNAP-Tag-ligand-bearing indolium derivative (6b) was utilized without undergoing further purification and analysis.

During the formation of hemicyanine 9a, we observed its degradation (Scheme 5A). The harsh acidic conditions employed may have led to the cleavage of the otherwise relatively stable ether bond of O⁶-benzylguanine. Accordingly, we decided to modify our approach by attaching the SNAP-Tag-ligand-bearing indolium derivative 6b in the second synthesis step and forming the hemicyanine with the carboxy-indolium precursor 5 (Scheme 5B).

Scheme 5. Synthesis of Hemicyanine 9a and b and the Asymmetric Cyanine Dye 4.

Reagents and conditions: (A, a) 1.00 eq. of indolium derivative 6b, 1.10 eq. of the aniline derivative 7, 12/1 acetic anhydride/acetyl chloride, 110°C oil bath, 4 h. (B, a) 1.00 eq. of indolium derivative 5, 1.10 eq. of the aniline derivative 7, 12/1 acetic anhydride/acetyl chloride, 110°C oil bath, 4 h (used without further purification); (b) 1.00 eq. of the indolium derivative 6b, 2.40 eq. sodium acetate in MeOH, RT, 21 h (6.9%).

Purification of the SNAP-Tag-ligand bearing asymmetric cyanine 5 dye 4 proved to be very intricate. The high polarity of the O⁶-benzylguanine derivative and its limited solubility in organic solvents immiscible with water rendered an aqueous workup not feasible. Consequently, we opted for column chromatography on basic aluminum oxide using a gradient of DCM/MeOH as the eluent. Even with pure MeOH, we were unable to elute the product. Therefore, we changed the eluent system by using MeOH to remove main impurities and then switched to isopropanol/water/ethyl acetate. Following lyophilization, we obtained the SNAP-Tag-ligand-bearing asymmetric cyanine 5 dye 4 in 6.9% yield and 91% purity. The low yield might be attributed to the small scale and high polarity of the synthesis reagents and product. The dark blueish color of the aluminum oxide, where the crude product was applied, led us to hypothesize that some of the product might have been irreversibly adsorbed to the aluminum oxide. However, we were unable to validate this hypothesis due to the inability to dissolve the blue residue in any solvent. Given the challenging synthesis and purification, further derivatization of the carboxylic acid was not attempted.

Photophysical Characterization

The photophysical characterization of the heterobifunctional cyanine 5 dyes 1–4 was performed in dimethylformamide (DMF), dimethyl sulfoxide (DMSO), ethanol (EtOH), H₂O, and phosphate-buffered saline (PBS). To the cyanine dyes in aqueous solution was added 0.1% of DMSO due to the slightly poorer solubility. The relevant photophysical data are summarized in Table 2. UV–vis absorption band maxima were observed around 640–653 nm, with molar extinction coefficient values consistent with literature reports^15,43,44 for other cyanine dyes. Owing to the relatively small Stokes shifts of around 20–30 nm, fluorescence emission spectra were obtained using 620 nm, not the wavelength of the absorption maximum of each dye, as the excitation wavelength. Fluorescence emission band maxima were found around 662–675 nm. Quantum yields were determined as well and are higher in organic nonprotic solvents than in protic solvents like EtOH or under aqueous conditions, aligning with existing literature.¹⁵

Table 2. Photophysical Data for Cyanine Dyes 1–4^a.

dye	solvent	λAbs	λEm	ε	ΔλST	ΦFl	brightness
1	DMF	646	670	1.31	24/555	0.36	4716
	DMSO	647	674	1.41	27/619	0.44	6204
	EtOH	644	670	1.59	26/603	0.27	4293
	H2O	640	664	1.39	24/565	0.18	2502
	PBS	640	663	1.29	23/542	0.18	2322
2	DMF	646	671	2.43	25/577	0.41	9963
	DMSO	648	673	2.33	25/574	0.48	11184
	EtOH	645	670	2.45	25/579	0.29	7105
	H2O	643	662	1.14	19/446	0.21	2394
	PBS	644	674	0.96	30/691	0.22	2112
3	DMF	650	670	2.82	20/459	0.43	12126
	DMSO	653	674	3.51	21/477	0.49	17199
	EtOH	647	670	2.76	23/531	0.32	8832
	H2O	642	664	0.90	22/516	0.21	1890
	PBS	643	663	1.10	20/469	0.21	2310
4	DMF	648	673	2.76	25/573	0.34	9384
	DMSO	649	675	2.98	26/594	0.45	13410
	EtOH	645	669	2.76	24/556	0.28	7728
	H2O	645	666	0.81	21/489	0.22	1782
	PBS	645	666	1.11	21/489	0.18	1998

^aλ_Abs and λ_Em are the absorption and fluorescence emission maxima in nanometers. ε is the molar extinction coefficient in ×10⁵ L mol^–1 cm^–1. Δλ_ST is the Stokes shift in nm/cm^–1, Φ_Fl is the quantum yield in fluorescence, and the brightness is given by units of [M^–1 cm^–1].

Figure 1 displays the normalized absorption and fluorescence emission spectra of cyanine dyes 1–4. The location of maxima and curve shapes exhibited minimal impact from the substituents or the solvents used. A general trend, characterized by a slight bathochromic effect, is observed transitioning from aqueous solution to organic protic and then to organic nonprotic solvents.

Figure 1.

Normalized absorption and fluorescence emission spectra of cyanine dyes 1–4 in solution with a dye concentration of 10^–7–10^–10 M and a cell path length of 1 cm.

Discussion

This study presents the design of heterobifunctional cyanine 5 dyes, achieving satisfactory overall yields from cost-effective starting materials (Figure 2). Notably, for the HaloTag-ligand derivatives, we reduced the workup steps and optimized the purification. The synthesis, devoid of high boiling point solvents, utilized a modular approach to circumvent functional group decomposition. Of note, we demonstrated the feasibility of the synthesis in gram scale. In the last synthesis step of the bifunctional cyanine 5 dyes, we showed the successful coupling of a compound with an amine group, such as azido-homoalanine, or the attachment of an alkyne in an esterification reaction suitable for click chemistry. Moreover, we showcased the versatility of our method by synthesizing the dye bearing a SNAP-Tag ligand instead of the HaloTag ligand. It is noteworthy that an alternative route proposed by Brun et al.⁴⁵ involves starting with a symmetric cyanine 5 dye, followed by coupling of a linker containing an azide and a SNAP-Tag ligand via click chemistry. Despite its comparable overall yields of around 2% (7% for our strategy), this approach necessitates two additional synthesis steps and relies on small-scale and preparative HPLC for purification (see Table 3). While our method offers a promising foundation for further exploration and refinement, future optimization of the synthesis and purification for the SNAP-Tag derivative is essential.

Figure 2.

Synthesized asymmetric cyanine 5 dyes (1, 2, 3, 4).

Table 3. Comparative Analysis of Heterobifunctional Dyes Synthesized in This Study and Those Reported in the Literature^a.

ref.	ligands forcyanine 5 dye	scale	purification method	overall yield [%]	synthesis steps
(45)	SNAP-Tag ligand and l-glutamic acid	one digit μmol	prep. HPLC	0.6	7
(30)	SNAP-Tag ligand and carboxylic acid	one digit μmol	prep. HPLC	1.4	5
	SNAP-Tag ligand and sulfonamide derivative	one digit μmol	prep. HPLC	0.5	5
(46)	SNAP-Tag ligand and CLIP-Tag ligand	one digit μmol	prep. HPLC	2	5
(47)	amine and hydroxy group	three digit μmol	prep. HPLC	5	7
(48)	phthalimide and carboxylic acid	?	DCVC and prep. HPLC	?	?
(22)	amine and carboxylic acid	?	prep. HPLC	?	?
(49)	DTPA derivative and RGD-based hybrid tracer	two digit nmol	prep. HPLC	0.05	7
(14)	amine and carboxylic acid	three digit μmol	prep. HPLC	33	2
	amine and carboxylic acid	three digit μmol	automated column chromatography	7.6	3
This work	HaloTag ligand and carboxylic acid	one digit mmol	workup	66	5
	HaloTag ligand and azido-homoalanine	two digit μmol	column chromatography	20	6
	HaloTag ligand and propargyl ester	two digit μmol	column chromatography	22	6
	SNAP-Tag ligand and carboxylic acid	one digit μmol	column chromatography	6	5

^aQuestion marks indicate instances where we were unable to find the required information.

As previously indicated, the existing literature offers only a few examples of heterobifunctional cyanine 5 dyes. A majority of those involve synthesis on a limited scale, subsequently purified through preparative HPLC. The overall yields, when assessed against equivalent synthetic steps, fall below 10% for the majority. Notably, no example has been documented to date of heterobifunctional cyanine 5 dyes featuring a HaloTag ligand. Conversely, a limited number do incorporate a SNAP-Tag ligand, with one example utilizing a CLIP-Tag ligand as well. The comprehensive characterization of these dyes is impeded by their synthesis on diminutive scales, and, in some instances, literature lacks essential details such as synthesis descriptions, scale, and yield. In contrast, our research provides a detailed account of the synthesis, purification, and thorough characterization of the compounds. Moreover, our yields surpass those reported in the literature, and the synthesis has been optimized for larger scales. A comprehensive comparison of our work with existing literature is presented in Table 3.

Also, it has to be noted that due to the instability of the dyes toward harsh conditions, applying strong nucleophilic bases or strong acids leads to the decomposition of the asymmetric dyes. However, under biological and physiological conditions, the dyes demonstrate exceptional stability. Extreme pH values, such as 1–2 or 13–14, lead to the degradation of the formed dyes. Thus, under moderate reaction conditions, the stability of the dyes poses no concerns and examples from the literature where similar dyes underwent nucleophilic substitution and/or copper(I)-catalyzed azide alkyne cycloaddition reaction are referred to in the conclusion below.

Lastly, the photophysical data obtained for the synthesized cyanine dyes 1–4 fall within the range observed for other asymmetric cyanine 5 dyes. Notably, we did not observe a substantial influence of the substituents on these data.

Conclusions

The modular approach in this synthetic strategy enables users to conjugate different blocks together. The synthesized dyes are equipped with a ligand, which can be fused to a self-labeling protein tag. Schnermann et al.⁵⁰ and Martin and Rivera-Fuentes⁵¹ recently demonstrated the labeling efficiency of asymmetric cyanine dyes in cells using the HaloTag or SNAP-Tag system, highlighting the excellent brightness and good photostability exhibited by the dyes when bound to a Tag.

As our aim was to induce bifunctionality, we included a functional group, which can be coupled or attached via biorthogonal click chemistry to any other molecule, peptide, or protein. This work has successfully evidenced the formation of stable ester or amide bonds within the synthesized compounds. Moreover, prior research has successfully demonstrated the conversion of cyanine 5 dyes through nucleophilic substitution.^{1,22,23,45,48,49} The coupling of cyanine 5 dyes with ligands containing either an azide or an alkyne moiety has been documented in the literature, employing the copper(I)-catalyzed azide alkyne cycloaddition reaction both within the compound and in cells.^{31,2,45,49,51−53}

Given its functional group tolerance, this approach allows for the design of asymmetrically modified dyes without limitations, especially when utilizing the HaloTag derivative. This is attributed to its better synthetic feasibility, particularly in terms of challenges in purification, compared to the SNAP-tag derivative.

Current topics using heterobifunctional molecules include PROTACs (Proteolysis Targeting Chimeras) and PhosTACs (Phosphorylation Targeting Chimeras). Both approaches leverage bifunctional small molecules capable of binding simultaneously to a target protein and either an E3-ubiquitin ligase, in the case of PROTACs (for inducing protein degradation), or a phosphatase, in the case of PhosTACs (for inducing dephosphorylation).^54,55 The heterobifunctional dyes synthesized in this study present themselves as promising candidates for use as PROTACs or PhosTACs, offering the advantage of being easily monitored with confocal microscopy. Consequently, these heterobifunctional molecules can serve as valuable tools for exploring potential interactors and understanding the consequences of their proximity in biological systems.

Experimental Section

Materials and Methods

All solvents and chemicals were commercially purchased from BLDpharm, Iris Biotech, Novabiochem, Roth, Sigma-Aldrich/Merck, or TCI and were used as received unless specified otherwise. Chemical reactions were carried out under an ambient atmosphere, except where noted. TLC plates were purchased from Merck (TLC aluminum foil, silica gel 60 F254, 20 × 20 cm), and compounds were detected by visualizing under UV light (254 nm and/or 366 nm) and/or by staining. Preparative column chromatography was performed with either silica gel (60, particle size 40–63 μm) or basic aluminum oxide (90, particle size 63–200 μm), both purchased from Roth.

Analytical measurements were conducted using an Agilent Technologies 1200 series HPLC system with an MWD SL UV detector and MS 6120 Single Quadrupole with an electrospray ionization (ESI) source. The separation was performed on a reversed-phase column (Macherey-Nagel EC 250/4 Nucleodur 100-5 C18ec) with a mobile phase comprising ACN (B) and water (A) with 0.1% FA or 0.05% TFA as an additive. Signals were detected at 210, 254, and 600 nm (UV detector G7117C 1260DADHS). The following standard method was used: 0.0–1.0 min 10% B in A, 1.0 min −15 min linear increase to 90% B, 15.0–18.0 min staying at 90% B, 18.0–18.2 min linear decrease to 10% B, 18.2–20.0 min staying at 10%. Flow rate: 0.8 mL/min.

Cyanine dye 2 was also purified through preparative HPLC using an Agilent Technologies 1260 Infinity I/II HPLC system equipped with a G7165A detector. Purification was carried out on a Macherey-Nagel 125/21 Nucleodur 100-5 C18EC (125 × 21 mm, 5.0 μm) column with eluents consisting of 95% ACN in water (B) and water (A) with 0.1% FA as an additive. Signals were detected at 230 and 600 nm. The following method was applied: 0.0–1.0 min 50% B in A, 1.0 min −20 min linear increase to 90% B, 20.0–29.0 min staying at 90% B, 29.0–29.5 min linear decrease to 50% B, 29.5–30.0 min staying at 50%. Flow rate: 20 mL/min. Cyanine dye 4 was purified using the same instruments conducting a slightly different method: 0.0–1.0 min 30% B in A, 1.0 min −20 min linear increase to 70% B, 20.0–29.0 min staying at 70% B, 29.0–29.5 min linear decrease to 30% B, 29.5–30.0 min staying at 30%. Flow rate: 20 mL/min.

NMR spectra were acquired on a Bruker AVANCE NEO 400 MHz NMR spectrometer with an autosampler (¹H: 400 MHz; ¹³C: 101 MHz). 2D NMR spectra (¹H–¹H COSY, HSQC, and/or HMBC) were recorded when necessary for peak assignment. Supporting Information includes C- and H-numbering. The residual protonated solvent signals serve as an internal standard for interpretation of the chemical shifts (δ) in ppm. Deuterated solvents were purchased from DEUTERO.

High-resolution mass spectrometry was conducted on a CE-ESI-MS-System (MS: 6520 qTOF-MS, Agilent).

The UV–vis absorption spectra were measured using a UV-1900i spectrophotometer from Shimadzu. The fluorescence emission spectra were obtained with an FP-8300 spectrofluorometer from Jasco. The UV–vis absorption spectra for calculating extinction coefficients (ε) were measured on a PerkinElmer Lambda 35 UV–vis spectrometer. The quantum yields of fluorescence (Φ_Fl) were determined using a Quantaurus-QY Absolute PL quantum yield spectrometer C11347 from Hamamatsu using a sample concertation of 10^–6 M.

HaloTag Substrate

tert-Butyl (2-(2-Hydroxyethoxy)ethyl)carbamate (10)

The reaction was performed under an argon atmosphere. A solution of 0.943 mL of 2-(2-aminoethoxy)ethanol (9.51 mmol, 1.00 equiv) in 15 mL of anhydrous DCM was treated with 2.21 g of Boc-anhydride (10.1 mmol, 1.06 equiv). After stirring for 3 h at room temperature, TLC showed full conversion of the starting material and the solvent was subsequently evaporated.

The resulting product was extracted 3× with 50 mL of DCM from 75 mL of H₂O. The combined organic layers were dried over Na₂SO₄, and the solvent was evaporated. The crude product was used without further purification in the next synthesis step (1.95 g, quant. yield, white powder, R_f = 0.27 DCM/MeOH 50/1 KMnO₄ stained). The obtained characterization data are in accordance with the literature.³³

¹H NMR (400 MHz, CDCl₃): δ = 3.69–3.77 (m, 2H, H-4), 3.50–3.60 (m, 2H, H-2, H-3), 3.32 (t, J = 5.13 Hz, 2H, H-1), 1.44 (s, 9H, H-7–9) ppm.

¹³C{¹H} NMR (101 MHz, CDCl₃): δ = 156.3 (C-5), 79.6 (C-6), 72.3 (C-3), 70.4 (C-2), 61.9 (C-4), 40.6 (C-1), 28.5 (C-7–9) ppm.

tert-Butyl (2-(2-((6-Chlorohexyl)oxy)ethoxyethyl))carbamate (11)

A solution of 1.50 g of 10 (7.31 mmol, 1.00 equiv) in 21 mL of THF/DMF 2/1 was treated with 0.438 g of sodium hydride (11.0 mmol, 1.50 eq., 60% in mineral oil) at 0 °C in an ice bath. After stirring for 1 h, 1.55 mL of 1-chloro-6-iodohexane (10.2 mmol, 1.40 equiv) was added and the reaction mixture was stirred at room temperature for 16 h. The reaction mixture was quenched by the addition of 10 mL of aq. sat. NH₄Cl. The aqueous phase was extracted 3× with 20 mL of EA, and the combined organic layers were washed with water and brine. The organic layer was dried over Na₂SO₄, and the solvent was evaporated. The crude product was purified by column chromatography using silica gel and CH/EA 4/1 as the eluent mixture. The protected HaloTag substrate was obtained in good yield (1.42 g, 60.2% yield, colorless oil, R_f = 0.24 DCM/MeOH 50/1 KMnO₄ stained). The obtained characterization data are in accordance with the literature.³⁴

¹H NMR (400 MHz, CDCl₃): δ = 5.04 (br. s, 0.78 H, NH), 3.46–3.51 (m, 2H, H-9), 3.38–3.51 (m, 6 H, H-1, H-7, H-8), 3.34 (t, J = 6.7 Hz, 2H, H-6), 3.34 (t, J = 6.6 Hz, 2H, H-10), 3.14–3.22 (m, 2H, H-10), 1.65 (quin, J = 6.7 Hz, 2H, H-2), 1.49 (quin, J = 6.8 Hz, 2H, H-5), 1.21–1.39 (m+s, 13H, H-3, H-4, H-13–15) ppm.

¹³C{¹H} NMR (101 MHz, CDCl₃): δ = 155.8 (C-11), 78.8 (C-12), 71.0 (C-6), 70.1 (C-7/8/9), 70.0 (C-7/8/9), 69.9 (C-7/8/9), 44.8 (C-1), 40.2 (C-10), 32.4 (C-3), 29.3 (C-4), 28.3 (C-13–15), 26.8 (C-2), 25.6 (C-5) ppm.

2-(2-((6-Chlorohexyl)oxy)ethoxy)ethan-1-amine) (12)

0.548 g (1.69 mmol, 1.00 equiv) of 2 was dissolved in 55 mL of DCM and cooled to 0 °C in an ice bath. Then, 11.1 mL of TFA (20 vol %) was added and the reaction mixture was stirred for 2 h. After full conversion, DCM was removed under reduced pressure and TFA was coevaporated 3× with 15 mL of toluene. The deprotected HaloTag substrate 11 was obtained as a yellow oil (0.377 g, quant. yield). The obtained characterization data are in accordance with the literature.³³

HRMS (ESI, m/z): [M + H]⁺ calcd for C₁₀H₂₂ClNO₂, 224.14; found: 224.1.

¹H NMR (400 MHz, CDCl₃): δ = 8.00 (bs, 2.74 H, NH_2/3), 3.68–3.78 (m, 2H, H-9), 3.55–3.67 (m, 4H, H-7, H-8), 3.52 (t, J = 6.7 Hz, 2H, H-1), 3.45 (t, J = 6.9 Hz, 2H, H-6), 3.16 (bs, 2H, H-10), 1.76 (quin, J = 6.9 Hz, 2H, H-2), 1.58 (quin, J = 6.9 Hz, 2H, H-5), 1.44 (m, 2H, H-3), 1.28–1.3 (m, 2H, H-4) ppm.

¹³C{¹H} NMR (101 MHz, CDCl₃): δ = 71.7 (C-6), 70.6 (C-7/8), 70.2 (C-7/8), 67.0 (C-9), 45.5 (C-1), 40.1 (C-10), 32.9 (C-2), 29.6 (C-5), 27.0 (C-3), 25.7 (C-4) ppm.

Cy5 Precursors

1-(2-Carboxyethyl)-2,3,3-trimethyl-3H-indol-1-ium (5)

To a solution of 13.3 mL (83.0 mmol, 1.00 equiv) of 2,3,3-trimethylindolenine in 90 mL of ACN, 25.4 g (166 mmol, 2.00 equiv) of bromopropionic acid was added. The reaction mixture was refluxed for 16 h. After complete conversion, the reaction mixture was precipitated in ice-cold diethyl ether. The crude product was dissolved in 250 mL of water and washed 3× with 250 mL of CH, EA, and DCM. After lyophilization, the carboxy-indolium derivative 5 was obtained as a peach solid (16.9 g, 88%).

¹H NMR (400 MHz, DMSO-d₆): δ = 7.95–8.03 (m, 1H, H-15), 7.81–7.87 (m, 1H, H-13), 7.59–7.65 (m, 2H, H-12, H-14), 4.65 (t, J = 7.0 Hz, 2H, H-4), 2.98 (t, J = 7.0 Hz, 2H, H-3), 2.86 (s, 3H, H-6), 1.52 (s, 6H, H-8, H-9) ppm.

¹³C{¹H} NMR (101 MHz, DMSO-d₆): δ = 197.9 (C-5), 171.5 (C-2), 141.8 (C-10), 140.0 (C-11), 129.4 (C-12/14), 128.9 (C-12/14), 123.5 (C-13), 115.0 (C-15), 54.3 (C-7), 43.5 (C-4), 31.10 (C-3), 21.9 (C-8/9), 14.3 (C-6) ppm.

1-(3-((2-(2-((6-Chlorohexyl)oxy)ethoxy)ethyl)amino)-3-oxopropyl)-2,3,3-trimethyl-3H-indol-1-ium (6a)

The reaction was performed under an argon atmosphere. In 150 mL of dry DCM, 1.45 g (6.24 mmol, 1.00 equiv) of the carboxy-Indolium derivative 5 was combined with 0.845 g (6.25 mmol, 1.00 equiv) of HOBt and 1.20 g (6.26 mmol, 1.00 equiv) of EDC*HCl. The reaction mixture was stirred for 5 min, followed by the addition of 1.40 g (6.26 mmol, 1.00 equiv) of deprotected HaloTag substrate 12 and 871 μL (6.24 mmol, 1.00 equiv) of Et₃N. The conversion was monitored by HPLC-MS. After 1 h, the solvent was evaporated. The crude product was dissolved in 250 mL of H₂O with the assistance of ultrasonication. The aqueous phase was washed 2× with 160 mL of CH and then extracted 3× with 200 mL of EA. The combined organic layers were dried over Na₂SO₄, and the solvent was evaporated. The purity of the crude pinkish product was sufficient for the next reaction step, and it was used without further analysis (1.50 g, pink dense oil).

1-(3-((4-(((2-Amino-9H-purin-6-yl)oxy)methyl)benzyl)amino)-3-oxopropyl)-2,3,3-trimethyl-3H-indol-1-ium (6b)

The reaction was carried out under an argon atmosphere. In 2 mL of dry DCM, a mixture of 0.0600 g (0.222 mmol, 1.00 equiv) of 6-((4-(aminomethyl)benzyl)oxy)-9H-purin-2-amine, 0.0360 g of HOBt (0.266 mmol, 1.20 equiv), and 0.0511 g EDC*HCl (0.266 mmol, 1.20 equiv) was combined, followed by the addition of 29.6 μL (0.266 mmol, 1.20 eq.) of NMM. After stirring at room temperature for 2 min, 0.0525 g (0.266 mmol, 1.20 equiv) of the indolium derivative 5 was added and stirred for 4.5 h. HPLC-MS analysis indicated remaining starting material; therefore, 0.20 eq. of HOBt, EDC*HCl, NMM, and carboxy-indolium derivative 5 was added. The reaction was stirred for 3 h, and HPLC-MS showed full conversion. The solvent was evaporated and dried under high vacuum. The residue was dissolved in 150 mL of aq. sat. NaHCO₃ using ultrasonication. The aqueous phase was extracted 3× with 150 mL of EA. The combined organic layers were concentrated under reduced pressure and dried under high vacuum. The residue was dissolved in 150 mL of aq. sat. NaHCO₃ solution using ultrasonication and extracted 3× with 150 mL of EA. After concentration under reduced pressure, the crude product (122.4 mg, pink dense oil) was obtained.

N-((1E,3E)-3-(Phenylimino)prop-1-en-1-yl)aniline (7)

The reaction was performed under an argon atmosphere. To 15 mL of EtOH, 1.00 mL (6.06 mmol, 1.00 equiv) of 1,1,3,3-tetramethoxypropane and 1.10 mL (12.0 mmol, 2 equiv) of aniline were added. The reaction mixture was cooled to 0 °C in an ice bath, and then 1.20 mL of concentrated HCl was added dropwise. The reaction mixture was stirred at room temperature for 3 h. After complete conversion, the solvent was removed and the orange precipitate was washed once with ice-cold water and dried under high vacuum (1.22 g, 92%). The obtained characterization data are in accordance with the literature.³⁰

¹H NMR (400 MHz, MeOD): δ = 8.70 (d, 2 H, H-7, H-9), 7.45–7.51 (m, 4 H, H-2, H-6, H-12, H-16), 7.37–7.42 (m, 4H, H-3, H-5, H-13, H-15), 7.26–7.35 (m, 2H, H-4, H-14), 6.27 (t, J = 11.6 Hz, 1H, H-8) ppm.

¹³C{¹H} NMR (101 MHz, MeOD): δ = 159.9 (C-7, C-9), 139.8 (C-1, C-11), 131.2 (C-2, C-6, C-12, C-16,), 127.7 (C-4, C-14), 118.8 (C-3, C-5, C-13, C-15), 99.4 (C-8) ppm.

Amino Acid

tert-Butyl 2-((((9H-Fluoren-9-yl)methoxy)carbonyl)amino)-4-azidobutanoate (13)

0.400 g (1.09 mmol, 1.00 equiv) of N-α-(9-fluorenylmethoxycarbonyl)-γ-azido-l-homoalanine was dissolved in 8 mL of 3/1 EA/CH. After addition of 0.950 g (4.35 mmol, 4.00 equiv) of tert-butyl 2,2,2-trichloroacetimidate, the reaction mixture was stirred at 40 °C in an oil bath for 30 h and the reaction progress was monitored by TLC. The solvent was evaporated, and the crude mixture was purified by column chromatography using silica gel and a gradient of 20/1–5/1 of CH/EA. The product fractions were collected, and the solvent evaporated. The tert-butyl protected amino acid 13 was obtained as a yellow oil in good yield (346 mg, 75%, R_f = 0.69 CH/EA 2/1 stained with bromocresol green).

HRMS (ESI, m/z): [M + H]⁺ calcd for C₂₃H₂₇N₄O₄⁺ 423.2027; found: 423.2060.

¹H NMR (400 MHz, CDCl₃): δ = 7.77 (d, J = 7.5 Hz, 2H, H-21, H-22), 7.60 (d, J = 7.4 Hz, 2H, H-15, H-16), 7.41 (t, J = 7.4 Hz, 2H, H-19, H-20), 7.32 (t, J = 6.9 Hz, 2H, H-17, H-18), 5.44 (d, J = 7.4 Hz, 1H, H-1), 4.21–4.47 (m, 3H, H-2, H-11), 4.23 (t, J = 6.8 Hz, 1H, H-12), 3.37–3.38 (m, 2H, H-4), 2.09–2.16 (m,1H, H-3), 1.89–1.97 (m, 1H, H-3), 1.49 (s, 9H, H-7, H-8, H-9) ppm.

¹³C{¹H} NMR (101 MHz, CDCl₃): δ = 170.8 (C-5), 156.0 (C-10), 143.8 (C-13, C-14/C-23, C-24), 141.5 (C-13, C-14/C-23, C-24), 127.9 (C-19, C-20), 127.2 (C-17, C-18), 125.2 (C-15, C-16), 120.2 (C-21, C-22), 83.0 (C-6), 67.1 (C-11), 52.4 (C-2), 47.8 (C-4), 47.3 (C-12), 30.1 (C-3), 28.1 (C-7, C-8, C-9) ppm.

tert-Butyl 2-Amino-4-azidobutanoate (14)

0.187 g (0.443 mmol, 1.00 equiv) of the tert-butyl protected amino acid 13 was dissolved in 6 mL of 1/1 diethylamine/ACN. The reaction was stirred for 90 min until full conversion was observed by TLC. The solvent was evaporated, and the crude product was purified by column chromatography using silica gel and 2/1 CH/EA as the eluent mixture. The product fractions were collected, and the solvent was evaporated. The Fmoc-deprotected amino acid 14 was obtained as a yellow oil in moderate yield (43.8 mg, 49%, R_f = 0.8 CH/EA 1/1 KMNO₄ stained) and used immediately in the next synthesis step. HRMS data could not be obtained due to the instability of the compound, that is why the mass from LCMS is shown. However, HRMS data could be obtained for the final product.

LCMS (ESI, m/z): [M]⁺ calcd for C₈H₁₇N₄O₂ ; 201.25; found 201.2.

¹H NMR (400 MHz, CDCl₃): δ = 3.42–3.50 (m, 3H, H-2, H-4), 2.13 (bs, 2H, H-1), 1.96–2.02 (m, 1H, H-3), 1.96–1.78 (m, 1H, H-3), 1.47 (s, 9H, H-7, H-8, H-9) ppm.

¹³C{¹H} NMR (101 MHz, CDCl₃): δ = 174.6 (C-5), 81.8 (C-6), 52.6 (C-2), 48.4 (C-4), 33.7 (C-3), 28.2 (C-7, C-8, C-9) ppm.

Final Cy5 Constructs

1-(2-Carboxyethyl)-2-((1E,3E)-5-((E)-1-(3-((2-(2-((6-chlorohexyl)oxy)ethoxy)ethyl)amino)-3-oxopropyl)-3,3-dimethylindolin-2-ylidene)penta-1,3-dien-1-yl)-3,3-dimethyl-3H-indol-1-ium (1)

To 0.950 g (2.17 mmol, 1.00 equiv) of HaloTag-indolium derivative 6a in 14 mL of acetic acid, 0.530 g (2.39 mmol, 1.10 equiv) of aniline derivative 7 was added. Subsequently, 3 mL (32.0 mmol, 15.0 equiv) of acetic anhydride was added dropwise. The reaction mixture was slowly heated to 110 °C in an oil bath and stirred for 5 h. Full conversion was observed by HPLC-MS. The solvent was evaporated, and the crude mixture was dried under high vacuum. The crude dark-green hemicyanine 8 was used without further analysis in the next synthesis step due to its instability.

To a mixture of 0.520 g (2.24 mmol, 1.00 equiv) of carboxy-Indolium derivative 5 and 0.440 g (5.36 mmol, 2.40 equiv) of sodium acetate in 100 mL of MeOH, all of the crude hemicyanine 8 was added in three portions over 1 h. The conversion was monitored by HPLC-MS. After stirring for an additional 2 h, the reaction was completed and the solvent was evaporated. The crude product was dissolved in 500 mL of sat. aq. NaHCO₃ with ultrasonication. The aqueous phase was washed 3× with 400 mL of CH and extracted 3× with 400 mL of EA. The combined organic layers were washed with brine, dried over Na₂SO₄, and the solvent was evaporated. The remaining residue, which contained mostly the product since the side and cleavage products could be dissolved in the aqueous phase, could not be dissolved in sat. aq. NaHCO₃ and was dissolved in 50 mL of DCM. This solution was washed with sat. aq. NaHCO₃ and brine and dried over Na₂SO₄, and the solvent was evaporated. The crude products were combined and washed 3× with ice-cold diethyl ether to remove the acetophenone byproduct. The purity of the dark-blue asymmetric cyanine 5 dye was sufficient for the next reaction step (1.27g, 81%, 88% purity by HPLC at 600 nm) (it should be noted that the product adhered to the drying agent and the frit, which is why a minimal amount of drying agent was used, and the remaining product was eluted from the frit by adding a few drops of MeOH). For purities >99%, compound 1 can be dry-loaded on a 400 mL basic Al₂O₃ column. For this, the crude product was loaded onto 10 mL of Al₂O₃ and placed on top of the column. The solvent was carefully layered on top (1 cm) and runoff. This was repeated until the layered solvent remained colorless. Sequentially, sand was added and the column continued normally. A mixture of DCM/MeOH 15:1 was used to elute the impurities. Then, 1% of TEA was added to the solvent mixture and the eluent mixture was gradually changed to DCM/MeOH 8/1.)

HRMS (ESI, m/z): [M]⁺ calcd for C₄₁H₅₅ClN₃O₅⁺ 704.3825; found: 704.3815.

¹H NMR (400 MHz, MeOD): δ = 8.26 (dd, J = 12.4 Hz, J = 12.4 Hz, 2 H, H-11, H-15), 7.21–7.50 (m, 9 H, H-2–5, H-21–24, NH), 6.65 (t, J = 12.4 Hz, 1 H, H-13), 6.44 (d, J = 13.8 Hz, 1 H, H-12), 6.29 (d, J = 13.7 Hz, 1 H, H-14), 4.35–4.43 (m, 4 H, H-26, H-41), 3.47–3.55 (m, 6 H, H-29, H-35, H-36), 3.44 (t, J = 6.5 Hz, 2 H, H-34), 3.38 (t, J = 5.7 Hz, 2 H, H-37), 3.26–3.29 (m, 2 H, H-38), 2.68 (t, J = 6.5 Hz, 2 H, H-40), 2.62 (t, J = 7.5 Hz, 2 H, H-27), 1.76–1.69 (m, 14 H, H-8, H-9, H-18, H-19, H-30), 1.55 (quint, J = 7.1 Hz, 2 H, H-33), 1.36–1.46 (m, 4 H, H-31, H-32) ppm.

¹³C{¹H} NMR (101 MHz, MeOD): δ = 177.5 (C-28), 175.4 (C-10/16), 174.0 (C-10/16), 172.4 (C-39), 156.1 (C-12/14), 155.2 (C-12/14), 143.4 (C-6/20), 143.3 (C-6/20), 142.8 (C-1/25), 142.4 (C-1/25), 129.8 (C-4/22), 129.6 (C-4/22), 127.1 (C-13), 126.5 (C-3/23), 125.9 (C-3/23), 123.4 (C-5/21), 112.5 (C-2/24), 111.9 (C-2/24), 105.2 (C-11/15), 104.1 (C-11/15), 72.2 (C-34), 71.2 (C-35/36), 71.1 (C-35/36), 70.3 (C-37), 50.8 (C-7/17), 50.4 (C-7/17), 45.7 (C-29), 42.9 (C-26), 41.6 (C-41), 40.5 (C-38), 36.2 (C-27), 34.8 (C-40), 33.7 (C-30), 30.5 (C-33), 28.1 (C-8/9/18/19), 27.8 (C-8/9/18/19), 27.7 (C-31/32), 26.4 (C-31/32) ppm.

1-(3-((4-Azido-1-(tert-butoxy)-1-oxobutan-2-yl)amino)-3-oxopropyl)-2-((1E,3E)-5-((E)-1-(3-((2-(2-((6-chlorohexyl)oxy)ethoxy)ethyl)amino)-3-oxopropyl)-3,3-dimethylindolin-2-ylidene)penta-1,3-dien-1-yl)-3,3-dimethyl-3H-indol-1-ium (2)

To 0.0420 g (0.0594 mmol, 1.00 equiv) of asymmetric cyanine 5 derivative 1 in 10 mL of DCM, 0.0114 g (0.0594 mmol, 1.20 equiv) of EDC*HCl, 0.008 g (0.059 mmol, 1.20 equiv) of HOBt, and 10.4 μL (0.059 mmol, 1.20 equiv) DIPEA were added. The reaction mixture was stirred for 5 min. Subsequently, 0.00990 g (0.0495 mmol, 1.00 equiv) of the Fmoc-deprotected amino acid 14 was added and the reaction mixture was stirred for 16 h until full conversion was observed by HPLC-MS. The solvent was evaporated, and the crude product was dissolved in DCM for purification by column chromatography using a 100 mL basic Al₂O₃ column. The dissolved product was carefully layered on top and runoff. This was repeated until the layered solvent remained colorless. Sequentially, sand was added and the column continued normally. For the purification, a gradient of 0–1% MeOH in DCM (removal of impurities from 0 to 0.6% MeOH in DCM, elution of product 0.8–1% MeOH in DCM) was used. The blue powder was obtained in moderate yield (13.0 mg, 30%, 91% purity by HPLC at 600 nm). To enable a comparison, NMR data for 2 purified by column chromatography and preparative HPLC are shown in the Supporting Information.

HRMS (ESI, m/z): [M]⁺ calcd for C₄₉H₆₉ClN₇O₆⁺ 886.4992; found: 886.5013.

¹H NMR (400 MHz, CDCl₃): δ = 8.42 (d, J = 8.4 Hz, 1H, H-43), 7.82–7.89 (m, 2H, H-11, H-15), 7.72 (bs, 1H, H-43), 7.30–7.40 (m, 6H, H-2, H-4, H-5, H-21, H-22, H-24), 719–7.24 (m, 2H, H-3, H-23), 7.06–7.12 (m, 1H, H-13), 6.52 and 6.76 (2xd, J = 12.9 and 13.5 Hz, 2H, H-12, H-14), 4.46–4.54 (m, 1H, H-44), 4.37–4.46 (m, 4H, H-26, H-41), 3.48–3.58 (m, 8H, H-29, H-35, H-36, H-37), 3.42–3.47 (m, 6H, H-34, H-38, H-46), 2.77–2.80 and 2.89–2.93 (2x t, J = 7.4 Hz and 7.0 Hz, 4H, H-27, H-40), 2.08–2.13 (m, 2H, H-45), 1.73–1.76 (m, 2H, H-30), 1.68 (s, 12H, H-8, H-9, H-18, H-19), 1.53–1.60 (m, 2H, H-34), 1.45 (s, 9H, H-49, H-50, H-51), 1.34–1.38 (m, 4H, H-31, H-32) ppm.

¹³C{¹H} NMR (101 MHz, CDCl₃): δ = 170.7 (C-10, C-16, C-47), 169.4 and 169.7 (C-28, C-39), 153.6 (C-12), 142.0 and 142.1 (C-6, C-20), 140.7 and 140.8 (C-1, C-25), 129.01 and 129.05 (C-4, C-22), 125.3 and 125.4 (C-3, C-23), 122.1 (C-5, C-21), 111.3 and 111.4 (C-2, C-24), 104.44 and 104.45 (C-11, C-15), 82.1 (C-48), 71.4 (C-34), 69.5, 70.1, and 70.4 (C-35, C-36, C-37), 51.4 (C-44), 49.2 and 49.3 (C-7, C-17), 48.3 (C-46), 45.3 (C-29), 41.7 and 41.8 (C-26, C-41), 39.5 (C-38), 34.2 and 34.3 (C-27, C-40), 32.7 (C-30), 30.8 (C-45), 29.6 (C-33), 28.02, 28.14, 28.22, and 28.27 (C-8, C-9, C-18, C-19, C-49, C-50, C-51), 25.5 and 26.8 (C-31, C-32) ppm.

2-((1E,3E)-5-((E)-1-(3-((2-(2-((6-Chlorohexyl)oxy)ethoxy)ethyl)amino)-3-oxopropyl)-3,3-dimethylindolin-2-ylidene)penta-1,3-dien-1-yl)-3,3-dimethyl-1-(3-oxo-3-(prop-2-yn-1-yloxy)propyl)-3H-indol-1-ium (3)

The reaction was performed under an argon atmosphere. To 0.0278 g (0.0394 mmol, 1.00 equiv) of asymmetric cyanine 5 derivative 1 in 3 mL of dry DCM, 0.00600 g (0.0394 mmol, 1.00 equiv) of HOBt, 0.00756 g (0.0394 mmol, 1.00 equiv) of EDC*HCl and 6.71 μL (0.0394 mmol, 1.00 equiv) of DIPEA were added. The reaction mixture was stirred for 10 min at room temperature. Then, 4.55 μL (0.0789 mmol, 2.00 equiv) of propargyl alcohol was added and the reaction mixture was stirred for 16 h until full conversion was observed by HPLC-MS. The solvent was evaporated, and the crude product was dissolved in DCM for purification by column chromatography using 180 mL of silica gel. The dissolved product was carefully layered on top and runoff. This was repeated until the layered solvent remained colorless. Sequentially, sand was added and the column continued normally. For the purification, a gradient of 0–10% MeOH in DCM was used (removal of impurities 0–8% MeOH, elution of product 9–10% MeOH in DCM). The blue powder was obtained in moderate yield (9.91 mg, 34%, 96% purity by HPLC at 600 nm).

HRMS (ESI, m/z): [M]⁺ calcd for C₄₄H₅₇ClN₃O₅⁺ 742.3981; found: 742.4011

¹H NMR (400 MHz, MeOD): δ = 8.25–8.33 (m, 2H, H-12, H-14), 7.48–7.52 (m, 2H, H-5, H-21), 7.39–7.44 (m, 2H, H-4, H-22), 7.25–7.36 (m, 5H, H-2, H-3, H-23, H-24, NH), 6.65 (t, J = 12.5 Hz, 1H, H-13), 6.37&6.33 (2xd, J = 13.9 Hz, J = 13.8 Hz, 2H, H-12, H-14), 4.66 (d, J = 2.5 Hz, 2H, H-43), 4.40–4.45 (m, 4H, H-26, H-41), 3.64 (bs, 1H, H-45), 3.49–3.55 (m, 6H, H-29, H-35, H-36), 3.44 (t, J = 6.5 Hz, 2H, H-34), 3.39 (t, J = 5.4 Hz, 2 H, H-37), 3.26–3.29 (m, 2H, H-38), 2.95 (t, J = 2.6 Hz, 1H, H-45), 2.90 (t, J = 7.0 Hz, 2H, H-27), 2.69 (t, J = 6.5 Hz, 2H, H-40), 1.74 + 1.73 (2xs, 12H, H-8, H-9, H-18, H-19), 1.55 (quint., J = 7.3 Hz, 2 H, H-33), 1.35–1.47 (m, 6 H, H-31, H-32,H-33) ppm.

¹³C{¹H} NMR (101 MHz, MeOD): δ = 175.4 (C-10/16), 174.6 (C-10/16), 172.3 (C-39), 171.2 (C-28), 156.1 (C-12/14), 155.8 (C-12/14), 143.2 (C-6/20), 143.1 (C-6/20), 142.7 (C-1/25), 142.5 (C-1/25), 129.8 (C-4/22), 127.3 (C-13), 126.5 (C-3/23), 126.3 (C-3/23), 123.4 (C-5,21), 112.3 (C-2/24), 112.0 (C-2/24), 105.1 (C-11/15), 104.5 (C-11/15), 76.8 (C-45), 72.2 (C-44), 71.2 (C-35/36/37), 71.1 (C-35/36/37), 70.3 (C-35/36/37), 53.4 (C-43), 50.8 (C-7/17), 50.6 (C-7/17), 45.7 (C-29), 41.8 (C-26/41), 40.7 (C-26/41), 40.6 (C-38), 33.7 (C-27/40), 32.6 (C-30), 30.8 (C-33), 30.5 (C-27/40), 28.0 (C-18/19), 27.7 (C-31/32), 26.5 (31/32) ppm.

2-((1E,3E)-5-((E)-1-(3-((4-(((2-Amino-9H-purin-6-yl)oxy)methyl)benzyl)amino)-3-oxopropyl)-3,3-dimethylindolin-2-ylidene)penta-1,3-dien-1-yl)-1-(2-carboxyethyl)-3,3-dimethyl-3H-indol-1-ium (4)

To a mixture of 0.0357 g (0.154 mmol, 1.00 equiv) of carboxy-indolium derivative 5 and 0.0376 g (0.169 mmol, 1.10 equiv) of the aniline derivative 7, a mixture of 47.4 μL of acetyl chloride and 569 μL of acetic anhydride was added. The suspension was stirred and heated to 110 °C in an oil bath. After 4 h, HPLC-MS showed full conversion and the solvent was evaporated. The residue was dissolved in DCM and precipitated in ice-cold diethyl ether and filtered off. The residue was redissolved in DCM and concentrated under reduced pressure.

To this residue, 0.0612 g (0.126 mmol, 1.00 equiv) of indolium derivative 6b and 0.0249 g (0.303 mmol, 2.40 equiv) of sodium acetate were added and dissolved in 9 mL of MeOH. The reaction was monitored by HPLC-MS and stirred for 21 h. After full conversion, the solvent was evaporated and the residue purified by column chromatography using 60 mL of basic Al₂O₃. For this, the crude product was loaded onto 1.5 mL of Al₂O₃ and put on top of the column. Solvent was carefully layered on top (1 cm) and runoff. This process was repeated until the layered solvent remained colorless. Sequentially, sand was added and the column continued normally.

First, major impurities were eluted using 100% MeOH, and then the solvent was changed to a mixture of water/IP/EA 1:2:2 to elute the product. The azure blue solid 4 was obtained in low yields (5.72 mg, 6.9%, 91% purity by HPLC at 600 nm) after solvent evaporation.

Impure fractions were concentrated and further purified by preparative HPLC yielding 2.27 mg (2.8%) of azure blue solid 4.

HRMS (ESI, m/z): C₄₄H₄₇N₈O₄⁺ [M]⁺ calcd 751.3715; found: 751.3734.

¹H NMR (400 MHz, DMSO-d₆): δ = 8.56 (t, J = 5.8 Hz, 1H, H-46), 8.31–8.37 (m, 3H, H-12, H-14, H-43), 7.56–7.59 (m, 2H, H-5, H-21), 7.41–7.49 (m, 2H, NH₂⁺-47), 7.37–7.39 (m, 6H, H-2, H-3, H-23, H-24, H-35, H-37), 7.21–7.29 (m, 4H, H-4, H-22, H-45), 7.13 (d, J = 8.1 Hz, 2H, H-34/H-38), 6.56 (t, J = 12.4 Hz, 1H, H-13), 6.35 (t, J = 14.2 Hz, 2H, H-11, H-15), 5.48 (s, 2H, H-39), 4.31 and 4.36 (dt, J = 22.5 and 6.9 Hz, 4H, H-26 & H-29), 4.21 (d, J = 5.7 Hz, 2H, H-32), 2.63 and 2.70 (dt, J = 12.8 and 6.8 Hz, 4H, H-30 and H-27), 1.61 and 1.65 (2xs, 12H, H-8, H-9, H-18, H-19) ppm.

¹³C{¹H} NMR (101 MHz, DMSO-d₆): δ = 173.0 (C-10/C-16), 172.6 (C-10/C-16), 171.9 (C-28), 169.2 (C-31), 159.1 (C-40), 158.1 (C-41), 154.4 (C-12/C-14), 154.2 (C-12/C-14), 154.0 (C-42), 141.79 and 141.77 (C-6/C-20), 141.1 (C-1/C-25), 141.0 (C-1/C-25), 139.7 (C-36), 139.3 (C-43), 134.4 (C-33), 129.4 (C-3, C-4, C-22, C-23), 129.0 (C-3, C-4, C-22, C-23), 128.8 (C- 35/C-37), 128.5 (C-35/C-37), 127.5 (C-34, C-38), 126.0 (C-13), 124.8 (C-4/C-22), 124.6 (C-4/C-22), 122.4 (C-5, C-21), 117.9 (C-44), 111.4 (C-2/-C-24), 111.1 (C-2/C-24), 103.8 (C-11/C-15), 103.4 (C-11/C-15), 67.7 (C-39), 49.0 (C-7/C-17), 48.8 (C-7/C-17), 42.1 (C-32), 40.4 (C-29), 39.3 (C-26), 33.2 (C-30), 31.5 (C-27), 27.12 and 27.05 (C-8/C-9/C-18/C-19) ppm.

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.joc.3c02673.

• General synthesis pathway; spectroscopical characterization; and additional photophysical data (PDF)

The authors declare no competing financial interest.