Design and Synthesis of Porphyrin–Nitrilotriacetic Acid Dyads with Potential Applications in Peptide Labeling through Metallochelate Coupling

Abstract

The need to detect and monitor biomolecules, especially within cells, has led to the emerging growth of fluorescent probes. One of the most commonly used labeling techniques for this purpose is reversible metallochelate coupling via a nitrilotriacetic acid (NTA) moiety. In this study, we focus on the synthesis and characterization of three new porphyrin–NTA dyads, TPP-Lys-NTA, TPP-CC-Lys-NTA, and Py₃P-Lys-NTA composed of a porphyrin derivative covalently connected with a modified nitrilotriacetic acid chelate ligand (NTA), for possible metallochelate coupling with Ni²⁺ ions and histidine sequences. Emission spectroscopy studies revealed that all of the probes are able to coordinate with Ni²⁺ ions and consequently can be applied as fluorophores in protein/peptide labeling applications. Using two different histidine-containing peptides as His₆-tag mimic, we demonstrated that the porphyrin–NTA hybrids are able to coordinate efficiently with the peptides through the metallochelate coupling process. Moving one step forward, we examined the ability of these porphyrin–peptide complexes to penetrate and accumulate in cancer cells, exploring the potential utilization of our system as anticancer agents.

Introduction

Proteins play a vital role in numerous biological events such as biocatalysis, molecular recognition, metabolism, and cell signaling.¹ While proteins have been extensively analyzed in vitro in their purified form, recent studies have been focused on studying them under more biologically relevant conditions. Protein analysis under such conditions is expected to provide insight into real protein functions and structures in the presence of various naturally occurring substances. Thus, labeling and tracking of biomolecules (proteins, antibodies, amino acids, and peptides) inside living cells are of tremendous importance in the determination of dynamics, mobility, localization, interactions, and functions.² At present, several labeling methods such as isotope markers,³ radioactive tracers,⁴ colorimetric biosensors,⁵ photoswitchable biomaterials,⁶ photochromic compounds,⁷ electrochemical sensors,⁸ and fluorescent labels⁹ are available for this purpose.

Among labeling techniques, fluorescent labeling is the most promising one due to its nondestructive nature and the high sensitivity of fluorescence spectroscopy. Moreover, it fulfills the requirements of a small measurement volume and low concentration of fluorescent material.¹⁰ Detection based on fluorescence techniques has gained great attention, and significant progress has been made in both fluorescence instrumentation and synthesis of new fluorophores. The fluorophores possess a reacting moiety, which may be bound covalently or noncovalently to the target biomolecules. Various types of derivatives have been employed as fluorophores, such as semiconductor nanocrystals, fluorescent proteins, or organic molecules.¹¹

The most used approach includes the fusion of fluorescent proteins such as green fluorescent protein (GFP) to the protein of interest (POI), enabling the fluorescent visualization of the POI inside living cells.^12,13 Although labeling with GFP is powerful and widely used in various biological research areas, the large size of GFP may interfere with the natural function and the localization of the POIs. To overcome these limitations and provide a more reliable method for protein labeling, many research groups have developed new approaches for selective labeling of POIs with appropriate small organic molecular probes.¹⁴

The attachment of the fluorescent derivative to the desired biomolecule may be accomplished chemically or biologically. Tag-labeling is an alternative type of labeling that can be performed both chemically and biologically.¹⁵ In the tag-based approach, the POI is genetically fused with a short peptide that binds site-specifically to a designed synthetic fluorescent probe.¹⁶ Protein labeling through the incorporation of a short tag holds several advantages over the other techniques. The introduced tag must not disrupt protein folding or function and usually presents high labeling specificity. In the last few decades, remarkable progress has been made in utilizing small molecule-based probes to visualize cellular events.¹⁷

Metallochelate coupling is one of the most common approaches that exploits the ability of nitrilotriacetic (NTA) complexes with transition metal ions (Zn²⁺, Ni²⁺, Co²⁺, or Cu²⁺) to coordinate certain amino acid residues, mainly oligo-histidine sequences, such as the hexahistidine tag (His₆-tag).¹⁸ The His₆-tag is an oligo-histidine sequence where six histidine residues are able to interact noncovalently with a transitional metal complex of NTA. The advantage is its specificity and the tag can either be expressed at the C-termini or N-termini during protein expression.¹⁸ The His₆–Ni²⁺–NTA system has been used extensively in molecular biology and biotechnology for affinity chromatography-based protein purification,¹⁹⁻²³ and large numbers of His-tagged protein libraries exist worldwide.¹⁶ Thus, the application of this tagging technique to image proteins in live cells may offer the opportunity to track various cellular events with minimal functional and spatial perturbation on a POI. A large variety of organic derivatives such as fluorescein,²⁴ rhodamine,²⁵ cyanine,²⁶ dibromobimane,²⁷ perylene,²⁸ and coumarin¹⁶ have been connected with NTA and used as labels targeting the His₆-tag. Moreover, in various NTA-based fluorescent probes, an increased number of NTA derivatives was introduced (mono-, di-, tri-, or tetra-NTA) to overcome the weak binding nature of His-tag with Ni²⁺-NTA.²⁹

Porphyrins and metalloporphyrins represent a class of compounds, which appears to be a very promising labeling tool due to their unique spectroscopic and luminescent properties. The electronic absorption spectra of porphyrins consist of two distinct regions: the strong intensity Soret or B band at 380–500 nm and a weaker set of Q bands in the range of 500–750 nm.³⁰ Additionally, their emission, depending on the metal inside the porphyrin center and the nature of the peripheral substituents, can be extended from 600 to 800 nm. Thus, porphyrins offer excellent potential as imaging agents since they are biologically compatible, their metal complexes are both thermodynamically and kinetically stable, and they exhibit high intrinsic specificity for tumors.³¹ To date, porphyrins are widely used for the photodynamic therapy (PDT) of tumors³² and as labels for cancer detection.³³ Moreover, there are several publications where porphyrin-based fluorophores were employed in peptide labeling.³⁴⁻⁴⁰ However, to the best of our knowledge, there is only one report in the literature where porphyrin derivatives bearing NTA groups have been employed in the labeling of various peptides and proteins that possess the His₆-tag.⁴¹ More specific, metallochelate coupling was applied to produce several oligopeptide- and polypeptide-based phosphorescent probes. Notably, the coordination of the porphyrin derivative with the polypeptide had minimal effect on the photophysics of the porphyrin moiety.

Although known organic fluorophore probes include compounds with emission from ultraviolet to near-infrared, there are still limitations with probes emitting in the longer wavelengths of the electromagnetic spectrum.⁸ Additionally, the coordination of the probe with the oligopeptide tag often quenches its emission and decreases fluorescence quantum yield up to 70–80%.¹⁸ Thus, there is a strong need for the development of new NTA-based fluorophores, where their emission maxima lie beyond about 600 nm and their fluorescence quantum yields remain high after their coordination with the His₆-tag.

Here, we report the preparation of three porphyrin-based fluorescent probes (Figure 1) that contain lysine-NTA moieties and could be applied in metallochelate coupling chemistry to label a target biomolecule. The only difference between the first two probes is in the linker that connects the porphyrin ring with the lysine-NTA part. In TPP-Lys-NTA, the connection was achieved through an amide bond, while in TPP-CC-Lys-NTA, a triazine ring was selected as the linker. Finally, in dyad Py₃P-Lys-NTA, the connection of the two components was also accomplished through an amide linkage and the presence of the three pyridyl groups is expected to increase the hydrophilicity of the final dyad.

Figure 1.

Structures of the three porphyrin-based probes synthesized in this study.

Moreover, as a proof of concept, we used two different peptides (Fmoc-FH and RGDSGAITIGH, Figure 2) to verify that our porphyrin–NTA hybrids have the ability to coordinate with them through the NTA part and can be used as fluorescent probes. These two peptides have the ability to self-assemble, forming various nanostructures, and due to the presence of one histidine residue in their backbone, they can be considered as models to mimic the His₆-tag.

Figure 2.

Structure of the three peptides utilized as His₆-tag mimic.

In both peptides, the histidine in their C-terminus site is the key residue for the metallochelate coupling via the nitrilotriacetic acid (NTA) moiety. However, the structure of the peptides is dissimilar due to the different hydrophilicity of the porphyrin–NTA hybrids. The Fmoc-FH dipeptide was selected for the hydrophobic dyad (TPP-Lys-NTA). The fluorenylmethoxycarbonyl (Fmoc) protecting group, apart from its widespread application in peptide chemistry, also presents interesting anti-inflammatory properties.⁴² Moreover, the Fmoc dipeptides can self-assemble and form well-ordered architectures, owing to the p–p interactions between the aromatic electrons of the fluorenyl rings and the formation of hydrogen bonds. Based on numerous studies that have demonstrated the capability of the Fmoc-FF dipeptide to self-assemble into well-defined nanostructures,⁴³ it is expected that the similar Fmoc-FH derivative will also possess the ability to self-assemble and then through the histidine residue it would be possible to coordinate with the porphyrin via the NTA moiety.

On the other hand, for the studies of the water-soluble Py₃P-Lys-NTA dyad, the RGDSGAITIGH (RGD_H) undecapeptide was selected as the probe. The RGD (Arg–Gly–Asp) tripeptide is one of the most preserved motifs throughout the evolution, as well as the most commonly present on the surface of various proteins of the extracellular matrix (ECM) such as fibronectin, vitronectin, osteopontin, and fibrinogen.⁴⁴ In particular, the RGD motif is recognized by, and then bound to, a major class of transmembrane glycoproteins called integrins. The main role of integrins is to facilitate the cell–extracellular matrix adhesion and regulate cell aggregation.⁴⁵ Computational studies have shown that the RGD motif can potentially possess cell adhesion properties.⁴⁶ Moreover, according to the literature, the GAITIG sequence is a β-amyloid-forming motif of the adenovirus fiber shaft, and it represents an innovative tool for the synthesis of new multipurpose biomaterials with a large variety of applications, either biomedical or technological.⁴⁷ The RGDSGAITIGF (RGD_F) oligopeptide (Figure 2), which lacks the histidine residue, was used as a control peptide. Both peptides RGD_H and RGD_F are highly ordered and form well-aligned β-sheet states. The residues in their C-terminus site (histidine and phenylalanine, respectively) are exposed outside the self-assembling GAITIG core and therefore are prone to coordination.

Results and Discussion

Synthesis and Characterization

The synthesis procedures that were followed for the preparation of the compounds discussed in this paper are shown in Schemes 1–3. All dyads consist of a free base porphyrin moiety covalently linked with a modified nitrilotriacetic acid (NTA) chelate ligand. In the case of TPP-Lys-NTA and Py₃P-Lys-NTA, the lysine-NTA moiety was connected with the porphyrin macrocycle through an amide bond, while in TPP-CC-Lys-NTA, a 1,3,5-triazine cyanuric chloride (CC) bridge was employed as a linker. The synthesis of dyad TPP-Lys-NTA is presented in Scheme 1.

Scheme 1. Synthesis Procedures for the Preparation of TPP-Lys-NTA.

Experimental conditions and reagents: (i) 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), N-hydroxysuccinimide (NHS), dimethyl formamide (DMF), room-temperature (rt), 24 h; (ii) DMF, Et₃N, rt, 96 h (43% for two steps); (iii) LiOH·H₂O, MeOH/tetrahydrofuran (THF) (1:1), rt, 96 h (96%).

Scheme 3. Synthesis Procedures for the Preparation of Py₃P-Lys-NTA.

Experimental conditions and reagents: (i) dry DMF, hexafluorophosphate azabenzotriazole tetramethyl uronium (HATU), DIPEA, rt, 72 h (60%); (ii) LiOH, MeOH/THF, (1:1), 40 °C, 96 h (96%).

The first step involves the quantitative conversion of mono-carboxyl porphyrin 1(48) to the corresponding succinimidyl ester intermediate 2 with the use of 1.6 equivalents of N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC·HCl) and 2 equivalents of N-hydroxysuccinimide (NHS).⁴⁹ Porphyrin 2 was not isolated (since succinimidyl ester group is a better leaving group than −OH and can be easily hydrolyzed) and was used in the next reaction without any other purification. In the second step, the activated succinimidyl ester 2 reacted with the NTA derivative 3(50−52) in the presence of triethylamine (Et₃N). Preparation of the modified nitrilotriacetic acid ligand 3 was achieved through the commercially available precursor compound N(epsilon)-benzyloxycarbonyl-l-lysine.⁵³ After purification with column chromatography, porphyrin 4 was isolated as red-purple solid in moderate yield. For amide coupling between porphyrin 1 and derivative 3, alternative coupling reagents such as thionyl chloride (SOCl₂), oxalyl chloride, N,N′-dicyclohexylcarbodiimide (DCC), and 1-hydroxybenzotriazole hydrate (HOBt) were also examined. However, all of the above-mentioned reagents failed to accomplish the formation of the desired amide bond. The desired product TPP-Lys-NTA was obtained in high yield after the basic hydrolysis of compound 4 with an excess amount of LiOH.^54,55

The synthesis of TPP-CC-Lys-NTA is outlined in Scheme 2. In this dyad, the NTA ligand was connected with the porphyrin through a triazine linker based on cyanuric chloride (CC). Cyanuric chloride was chosen as the bridging ligand because its reactivity is temperature-dependent and allows the introduction of up to three different nucleophiles on the s-triazine unit through sequential substitution of its three chlorine atoms. As a result, it allows the use of one-pot protocol reactions, providing a modular synthetic route for the preparation of unsymmetrical multiporphyrin arrays bearing different chromophores, in good yields and avoiding extensive preparation and purification procedures.⁵⁶⁻⁶²

Scheme 2. Synthesis Procedures for the Preparation of TPP-CC-Lys-NTA.

Experimental conditions and reagents: (i) THF, N,N-diisopropylethylamine (DIPEA), rt, 96 h; (ii) piperidine, DIPEA, 65 °C, 24 h (25% for two steps); (iii) LiOH·H₂O MeOH/THF (1:1), rt, 96 h (98%).

The initial step for the synthesis of TPP-CC-Lys-NTA involved the reaction of triazine porphyrin 5(63) with 1.2 equivalents of NTA derivative 3 in the presence of N,N-diisopropylethylamine (DIPEA) at room temperature in THF. This resulted in the formation of an intermediate porphyrin adduct 6, bridged to a modified NTA ligand, which was neither isolated nor characterized but was further reacted at 65 °C, with an excess amount of piperidine. After purification with column chromatography, compound 7 was obtained. The introduction of the piperidine was performed to increase the stability of the final dyad by eliminating any undesired side reactions that can take place due to the presence of the −Cl group in the triazine linker. In the present form, the TPP-CC-Lys-NTA dyad is very hydrophobic and is difficult to disperse or solubilize in water, limiting its practical applications in metallochelate coupling. However, instead of piperidine, any other appropriately substituted derivative can be connected to enhance the desired properties (i.e., increase the binding with the His-tag or improve its solubility in water) of our probe. Finally, basic hydrolysis with an excess amount of LiOH afforded the desired dyad TPP-CC-Lys-NTA in almost quantitative yield. The presence of the 1,3,5-triazine bridge increases the distance between the porphyrin chromophore and the NTA group in TPP-CC-Lys-NTA compared to that in TPP-Lys-NTA. This is expected to modify the photophysical properties of the porphyrin after complexation of NTA with a paramagnetic metal like Ni²⁺. Usually, after the formation of the NTA–Ni²⁺–(His₆-tag) complex, the emission of the fluorophore is significantly quenched.⁴¹ The increased distance between the porphyrin and NTA is expected to reduce the undesired quenching of porphyrin’s emission. Moreover, the s-triazine linker offers the possibility to attach a second NTA group. This will enhance the binding constant of the label with the His₆-tag since multivalent complexes carrying multiple NTA moieties present more stable labeling, as reported in the literature.²⁹ Finally, in the third available position of the s-triazine bridge, we can also add another functional moiety that may increase the solubility of our derivative in water or enhance its membrane permeability to label intracellular proteins in living cells.

The synthetic approach for the preparation of the porphyrin dyad Py₃P-Lys-NTA is illustrated in Scheme 3. The pyridyl groups of the porphyrin ring were chosen to provide our fluorophore with high hydrophilicity, which is desired for labeling studies with water-soluble peptides.

Initially, we performed amide coupling of porphyrin 9(48) with NTA derivative 3 using hexafluorophosphate azabenzotriazole tetramethyl uronium (HATU) as the coupling reagent, yielding porphyrin 10. Subsequently, basic hydrolysis of the methyl esters was performed via LiOH, providing the desired product in a high yield.

All intermediates and final products were fully characterized by matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry, and in all cases, the desired molecular ion was observed (Figures S28–S33). The low solubility of the dyads TPP-Lys-NTA and TPP-CC-Lys-NTA prevented their characterization through NMR spectroscopy. Therefore, we fully characterized their ester analogues 4 and 7 with ¹H and ¹³C NMR spectroscopy (Figures S1–S10). The hydrophilic dyad Py₃P-Lys-NTA presented good solubility in dimethyl sulfoxide (DMSO), enabling its characterization through NMR spectroscopy (Figures S11–S20). In all cases, for the complete assignment of the peaks, it was necessary to record two-dimensional (2D) NMR spectra (COSY, HSQC, and HMBC). The successful formation of the amide bond between lysine-NTA and porphyrin in compounds 4 and Py₃P-Lys-NTA was confirmed from the peaks at 6.93 and 8.83 ppm, which correspond to the NH amide protons of derivatives 4 and Py₃P-Lys-NTA, respectively. Additionally, in the ¹³C spectra, the peaks at 168.0 and 166.3 ppm correspond to the carbonyl carbon of the amide bond of derivatives 4 and Py₃P-Lys-NTA, respectively. The peaks at around 174–172 ppm were assigned to the carbonyl carbons of the NTA moiety. Accordingly, in adduct 7, all of the expected signals were observed and fully assigned.

Theoretical Measurements

The molecular structures of the hydrophobic porphyrin-based fluorescent probes were explored through density functional theory (DFT) calculations to study the influence of the linkage on the properties of the two dyads. The gas-phase-optimized structures of TPP-Lys-NTA and TPP-CC-Lys-NTA are shown in Figure 3, while the corresponding coordinates are provided in Tables S1 and S2, respectively. In both dyads, the meso-phenyls are oriented almost perpendicular to the porphyrin macrocycle. Furthermore, in both cases, we observe that the lysine-NTA unit is extended linearly away from the porphyrin plane.

Figure 3.

Gas-phase geometry-optimized structures of TPP-Lys-NTA (left) and TPP-CC-Lys-NTA (right). Carbon, nitrogen, hydrogen, and oxygen are shown as gray, blue, white, and red spheres, respectively.

The optimized minimum-energy structures were obtained as stationary points, and Figures S21 and S22 illustrate the frontier molecular orbitals (FMOs) of both porphyrin-based fluorescent probes, displaying the corresponding energy contributions. Moreover, the energies of the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) and the band gaps (ΔE_HL) of TPP-Lys-NTA and TPP-CC-Lys-NTA were calculated and compared to the already-known values of [Ni(II)(H₂O)₆]²⁺ (Table 1).⁶⁴ Interestingly, TPP-CC-Lys-NTA possesses higher HOMO energy than TPP-Lys-NTA, thus leading to a smaller HOMO–LUMO gap between TPP-CC-Lys-NTA (HOMO) and [Ni(II)(H₂O)₆]²⁺ (LUMO). This finding suggests that Ni(II) and TPP-CC-Lys-NTA ligand present the most noticeable electrophilic and nucleophilic characteristics, respectively. More importantly, based on the fact that the E_HOMO’s of TPP-Lys-NTA and TPP-Lys-NTA are higher than the E_LUMO of [Ni(II)(H₂O)₆]²⁺, the chelation of the metal ion with the NTA ligand is expected to occur easily in both cases. Additionally, to explore the chemical reactivity of the two porphyrin-based fluorescent probes and Ni(II), three important DFT descriptors (namely, chemical potential (μ), global hardness (η), and electrophilicity (ω)) were calculated through the following equations and are listed in Table 1.^65,66

1

2

3

As clearly evident, the μ value of the Ni(II) complex is more negative than those calculated for both TPP-Lys-NTA and TPP-CC-Lys-NTA derivatives, indicating once again that [Ni(II)(H₂O)₆]²⁺ acts as an electrophile, while both porphyrins serve as nucleophiles. Among these two nucleophiles, TPP-CC-Lys-NTA presents a slightly decreased electronegativity (χ = −μ), revealing that it possesses a stronger nucleophilic ability. Moreover, the smaller η and ω values that were calculated for TPP-CC-Lys-NTA also suggest that this derivative presents a slightly stronger affinity toward the nickel ion compared to TPP-Lys-NTA.

Table 1. Calculated E_HOMO, E_LUMO, ΔE_HL, μ, η, and ω in Hartree for both TPP-Lys-NTA and TPP-CC-Lys-NTA Ligands, as well as the Previously Reported Values of the [Ni(II)(H₂O)₆]²⁺ Complex.

derivative	EHOMO (Ha)	ELUMO (Ha)	ΔEHL (Ha)	μ (Ha)	η (Ha)	ω (Ha)
TPP-Lys-NTA	–0.1837	–0.0852	0.0985	–0.1345	0.0493	0.1835
TPP-CC-Lys-NTA	–0.1770	–0.0795	0.0975	–0.1283	0.0488	0.1687
[Ni(II)(H2O)6]2+	–0.2593a	–0.1401a	0.1192a	–0.1997a	0.0596a	0.3346a

^aThese values were taken from the literature.⁶⁴

Photophysical Studies

Absorption and emission spectroscopy experiments were performed to examine the ability of TPP-Lys-NTA, TPP-CC-Lys-NTA, and Py₃P-Lys-NTA to coordinate with Ni²⁺ ions. For these studies, a small amount from a concentrated aqueous solution of NiSO₄ was added to a solution of the corresponding porphyrin–NTA dyad in DMSO (detailed experimental details are included in the Experimental Procedures section). UV–vis absorption and emission spectra were recorded before and after the addition of the NiSO₄ solution.

In all cases, the initial absorption spectra of the dyads (Figures S23 and S24) presented typical bands that correspond to a free-base porphyrin, with an intense Soret band at around 420 nm and four weaker Q bands in area 500–650 nm. After the addition of the Ni²⁺ ions, the absorption spectra were remained identical (Figures S23 and S24). These results suggest that in the ground state there are no significant electronic interactions between the porphyrin chromophore and the Ni²⁺ ions. Moreover, since the number of the Q bands remains identical, we can easily suggest that under these conditions metallation of the porphyrin ring cannot take place.

The emission spectra of TPP-Lys-NTA, TPP-CC-Lys-NTA, and Py₃P-Lys-NTA (Figures 4 and S24) present two bands at ∼650 and 720 nm, which are typical for a free base porphyrin. In the case of the TPP-CC-Lys-NTA dyad, the two emission maxima are red-shifted (4 nm) compared to TPP-Lys-NTA. A similar red shift was also observed in their corresponding absorption spectra. This difference can be attributed to the presence of the cyanuric ring linker that slightly alters the photophysical properties of TPP-CC-Lys-NTA both in the ground state and excited state. Moreover, the Py₃P-Lys-NTA derivative presents a significantly lower emission intensity compared to the other two dyads. After the addition of the NiSO₄ solution, the intensity of the emission was reduced (Figures 4 and S24). The quenching is more intense in the Py₃P-Lys-NTA dyad, reaching 28%. According to the literature, when a fluorophore coordinates with Ni²⁺ ions, its emission usually is quenched due to the paramagnetic nature of the Ni²⁺ ion.^29,41 Control experiments with monoester porphyrin 8(48) and the tetrapyridyl porphyrin TPyP (Figure S25) were also performed to prove that the observed quenching is due to the coordination of the Ni²⁺ with the NTA part. As expected, the emission spectra of porphyrin 8 and TPyP are not altered after the addition of NiSO₄ (Figures S26 and S27). Thus, the observed quenching in the porphyrin–NTA dyads can be attributed to the interaction of Ni²⁺ with the NTA ligand.

Figure 4.

Emission spectra of TPP-Lys-NTA (left part) and TPP-CC-Lys-NTA (right part) in DMSO before and after the addition of Ni²⁺ ions. The excitation wavelength was 505 nm.

Moreover, quenching in the case of TPP-CC-Lys-NTA was slightly smaller compared to that in TPP-Lys-NTA (9 vs 12%). This can be ascribed to the increased distance between the porphyrin and the NTA part due to the presence of the s-triazine bridge. One of the main drawbacks of the most reported fluorophores is the significantly quenched emission (70–80%) after their coordination with the transition metal.⁴¹ This unfavorable process limits the detection limit of the fluorophores after their complexation with the target biomolecule. It is noteworthy that in our probes after coordination with the Ni²⁺ ion their emission is not significantly quenched. Finally, the ability of the reported porphyrin–NTA dyads to emit photons up to 720 nm is also a desired property since it broadens the spectral window and there are few examples in the literature that fulfill this requirement.

Metallochelate Coupling Studies

Solid-state absorption spectroscopy studies were performed to investigate the metallochelate coupling between porphyrin–NTA hybrids, nickel (Ni²⁺) ions, and the histidine-bearing peptides. The spectra were recorded in the solid state since the peptides that were used have the ability to self-assemble and form water-insoluble aggregates. Among the two hydrophobic derivatives, only TPP-Lys-NTA was used in these studies since the TPP-CC-Lys-NTA derivative was not dispersible enough under the selected experimental conditions. Moreover, in the case of TPP-Lys-NTA, the Fmoc-FH dipeptide was employed, while for the water-soluble Py₃P-Lys-NTA dyad, the RGDSGAITIGH oligopeptide was employed as a mimic of the His₆-tag. In all of these experiments, the ratio of the porphyrin hybrid relative to the corresponding peptide was 1:2.

The solid-state UV–vis spectrum of TPP-Lys-NTA (solution 3, see the Experimental Procedures section) displayed the characteristic porphyrin features with the Soret band at 430 nm (Figure 5a and Table 2). Compared to the corresponding solution spectrum in DMSO (Figure S23), all of the absorption maxima were red-shifted and broadened. After the addition of Ni²⁺ ions (solution 4), in accordance to the corresponding studies in DMSO solution, the spectrum remains intact without any significant shift of the peaks. This result suggests that there are no noteworthy electronic interactions between the porphyrin chromophore and the Ni²⁺ ions. Additionally, the control sample that contained only TPP-Lys-NTA and Fmoc-FH (solution 5) did not present any significant differences. However, the solid-state spectrum that contains all of the three components (TPP-Lys-NTA, Ni²⁺ and Fmoc-FH) presented several differences (Figure 5a). More specific, the characteristic porphyrin maxima were red-shifted and significantly broadened. These changes can be attributed to the binding of the TPP-Lys-NTA hybrid with the Fmoc-FH peptide via the metallochelate coupling with the NTA moiety.

Figure 5.

(a) Solid-state absorption spectra of TPP-Lys-NTA (solution 3, blue line), the complex of TPP-Lys-NTA + Ni²⁺ (solution 4, red line), the control sample TPP-Lys-NTA + Fmoc-FH (solution 5, olive line), and the complex TPP-Lys-NTA + Ni²⁺ + Fmoc-FH (solution 1, black line). (b) Solid-state absorption spectra of Py₃P-Lys-NTA + Ni²⁺ (solution 8, black line), after the addition of RGDSGAITIGH (solution 6, blue line), after adding RGDSGAITIGF (solution 7, red line), the control sample RGDSGAITIGH + Py₃P-Lys-NTA (solution 9, olive line), and the control sample RGDSGAITIGF + Py₃P-Lys-NTA (solution 10, magenta line).

Table 2. Absorption Peak Values of All of the Studied Compounds in Solution and in the Solid State.

compound	soret band λmax (nm)	Q bands λmax (nm)
TPP-Lys-NTA solution	418	514, 548, 590, 646
Py3P-Lys-NTA solution	416	512, 546, 587, 643
TPP-Lys-NTA	430	519, 553, 594, 648
TPP-Lys-NTA + Ni2+	431	521, 553, 593, 649
Fmoc-FH + TPP-Lys-NTA	432	518, 553, 592, 645
Fmoc-FH + TPP-Lys-NTA + Ni2+	435	524, 555, 594, 651
Py3P-Lys-NTA + Ni2+	434	530, 568, 601, 658
RGDSGAITIGF + Py3P-Lys-NTA	422	521, 557, 599, 652
RGDSGAITIGH + Py3P-Lys-NTA	422	522, 558, 597, 650
RGDSGAITIGF + Py3P-Lys-NTA + Ni2+	431	527, 562, 600, 654
RGDSGAITIGH + Py3P-Lys-NTA + Ni2+	420

Concerning the hydrophilic Py₃P-Lys-NTA derivative, the histidine-containing peptide RGDSGAITIGH was applied as the His₆-tag mimic. Moreover, the RGDSGAITIGF peptide, in which the histidine residue is replaced by phenylalanine, was employed as a “control” molecule.

As illustrated in Figure 5b, the spectrum of Py₃P-Lys-NTA-Ni²⁺ (solution 8) changes significantly after the addition of the RGDSGAITIGH peptide (solution 6). More specifically, all of the porphyrin absorption peaks were blue-shifted up to 14 nm (Table 2). On the other hand, the introduction of the control RGDSGAITIGF oligopeptide (solution 7) did not alter significantly the absorption features of the porphyrin (the Soret band was blue-shifted by only 4 nm). Interestingly, the control samples that contained the Py₃P-Lys-NTA porphyrin and the oligopeptides RGDSGAITIGH or RGDSGAITIGF without the Ni²⁺ ions (solutions 9 and 10, respectively) presented also significantly blue-shifted spectra (Table 2). The recorded absorption maxima were similar to the sample containing RGDSGAITIGH + Py₃P-Lys-NTA + Ni²⁺ (solution 6). These unexpected results indicate that the Py₃P-Lys-NTA porphyrin interacts strongly with both oligopeptides, even without the contribution of Ni²⁺ ions. Most likely, the carboxylic groups of the NTA moiety can form hydrogen bonds with the amino groups of the oligopeptides. However, this interaction is possible only in the absence of Ni²⁺ ions since in the sample with RGDSGAITIGF + Py₃P-Lys-NTA + Ni²⁺ (solution 7) we did not observe this process. Thus, in the presence of Ni²⁺ ions, the NTA part coordinates with the metal ions and then it cannot form any additional hydrogen bonds. All of the above results indicate that in the sample with RGDSGAITIGH + Py₃P-Lys-NTA + Ni²⁺ (solution 6) the observed interaction is ascribed to the concomitant metallochelate coupling of the NTA moiety with the Ni²⁺ ions and the histidine residues.

Solid-state emission spectroscopy studies were also performed (Figure S28) to investigate metallochelate coupling. In the case of TPP-Lys-NTA porphyrin, the emission spectrum of the complex Fmoc-FH + TPP-Lys-NTA + Ni²⁺ was red-shifted compared to those of the samples where one of the three components was not present (Table S1). For the pyridyl-substituted derivative Py₃P-Lys-NTA, the emission maxima were blue-shifted in the samples where the porphyrin interacts strongly with the oligopeptides (Table S1). In all cases, the emission maxima were shifted following the same trend as in the absorption spectroscopy studies, verifying metallochelate coupling.

Field-Emission Scanning Electron Microscopy (FE-SEM) Studies

To provide additional evidence for the ability of porphyrin–NTA hybrids to coordinate with histidine-bearing peptides in the presence of Ni²⁺ ions via metallochelate coupling, field-emission scanning electron microscopy (FE-SEM) studies were performed. TPP-Lys-NTA derivative was able to self-assemble and formed flake-shaped nanostructures, while the addition of Ni²⁺ ions altered the self-assembly mode of the porphyrin and resulted in spherical architectures with a size of 100–200 nm (Figure 6). This observation verifies the ability of the NTA part to coordinate with the Ni²⁺ ions. On the other hand, the Fmoc-FH dipeptide could assemble into a fibrillar network. Interestingly, in the sample containing both Fmoc-FH and TPP-Lys-NTA, the two components self-assemble independently and form fibrillar nanostructures covered with spheres, verifying that there is no significant interaction between them. Finally, the combination of Fmoc-FH with TPP-Lys-NTA-Ni²⁺ resulted also in the formation of more well-defined spherical nanostructures with significantly increased diameters (100–500 nm). These modifications can be attributed to the coordination of the TPP-Lys-NTA hybrid with the Fmoc-FH through metallochelate coupling

Figure 6.

Field-emission SEM (FE-SEM) pictures of (a) TPP-Lys-NTA (solution 3), (b) TPP-Lys-NTA and Ni²⁺ (solution 4), (c) Fmoc-FH and TPP-Lys-NTA with Ni²⁺ (solution 1), (d) Fmoc-FH (solution 2), and (e) Fmoc-FH and TPP-Lys-NTA (solution 5). All of the above samples were prepared in hexafluoroisopropanol/4-(2-hydroxyethyl)-1-piperazine ethanesulphonic acid (HFIP/HEPES) 50 mM pH 7.4 (3:7, v/v) solvent mixture.

Similar observations by FE-SEM studies were also detected for the water-soluble Py₃P-Lys-NTA derivative. The Py₃P-Lys-NTA hybrid after the addition of Ni²⁺ ions showed undefined structures (Figure 7c). However, after the addition of the RGDSGAITIGH peptide, the self-assembly mode was modified completely and the formation of a porous fibrillar network was observed. On the other hand, the introduction of the control RGDSGAITIGF peptide resulted in the formation of thinner fibrils and smaller pores were detected between the nanostructures (Figure 7b). The divergences that were noticed in these images before and after the addition of the oligopeptides indicate that metallochelate coupling takes place. Remarkably, RGDSGAITIGH and the control RGDSGAITIGF, after the addition of the Py₃P-Lys-NTA hybrid without Ni²⁺ ions, could also assemble into well-defined fibrillar networks. These results are in accordance with the absorption studies where we observed that Py₃P-Lys-NTA interacts with both oligopeptides even in the absence of Ni²⁺ ions.

Figure 7.

Field-emission SEM (FE-SEM) pictures of (a) RGDSGAITIGH + Py₃P-Lys-NTA + Ni²⁺ (solution 6), (b) RGDSGAITIGF + Py₃P-Lys-NTA + Ni²⁺ (solution 7), (c) Py₃P-Lys-NTA + Ni²⁺ (solution 8), (d) RGDSGAITIGH + Py₃P-Lys-NTA (solution 9), and (e) RGDSGAITIGF + Py₃P-Lys-NTA (solution 10). Time of incubation: 7 days.

Confocal Microscopy Investigation with HeLa Cells

Based on the above encouraging results, which verified the ability of the porphyrin–NTA hybrids to coordinate with the histidine-containing peptides, we additionally performed internalization experiments into mammalian cells to study the capability of the final complexes to penetrate the cells. For these studies, solutions 1, 4, 6, and 8 were employed.

Confocal microscopy experiments demonstrated that the porphyrin TPP-Lys-NTA alone and the complex of TPP-Lys-NTA with the peptide Fmoc-FH entered efficiently into the HeLa cells and accumulated into the cytoplasm. There is no evidence of accumulation in the nucleus (Figure 8). The fluorescence signal was calculated and compared for both the porphyrin and the peptide–porphyrin complex with ImageJ software. The emitted fluorescence was similar for each sample with no statistical difference between them (Figure 8).

Figure 8.

Confocal microscopy pictures correspond to cellular uptake of (A) TPP-Lys-NTA + Ni²⁺ and (B) TPP-Lys-NTA + Ni²⁺ + Fmoc-FH. HeLa cells were exposed to 5 μL of solutions 1 and 3, followed by 24 h incubation. The bright-field illumination form was applied additionally to circumscribe the limits of the cell membranes.

Similar studies were performed with the hydrophilic porphyrin Py₃P-Lys-NTA before and after its complexation with the peptide sequence RGDSGAITIGH. Moreover, to facilitate the examination of the cell internalization propensity of the complexes, the 4′,6-diamidino-2-phenylindole (DAPI) nuclear staining assay was employed. DAPI is a blue-fluorescent DNA staining dye commonly used to distinguish the cell nucleus location. The porphyrin gives a strong fluorescent signal around the nucleus, as shown in Figure 9. In both cases (with or without the RGDSGAITIGH peptide), the porphyrin seems to internalize in the cell and localize in the cytoplasm.

Figure 9.

Confocal microscopy pictures correspond to the cellular uptake of (A) Py₃P-Lys-NTA + Ni²⁺ and (B) Py₃P-Lys-NTA + Ni²⁺ + RGDSGAITIGH. HeLa cells were exposed to 5 μL of solutions 5 and 6 followed by 24 h incubation. For confocal microscopy observations, cells were washed and stained with DAPI nuclear staining.

This observation can be rationalized based on the work of Hiyama and his group,⁶⁷ which demonstrated that exogenous porphyrin could be recognized by the heme carrier protein receptors HCP-1 since the porphyrin structure is virtually the same as heme. These receptors are overexpressed into the cancer cells, and as a result, the uptake of porphyrin increases.⁶⁷ Moreover, the RGD motif can be recognized by the integrin receptors on the cell membranes, so this can further facilitate the penetration of the bound porphyrin complex into the cell matrix.⁴⁵

Small therapeutic molecules like porphyrins usually exhibit a lack of tissue targeting ability, poor pharmacokinetics, and poor stability. Although porphyrin molecules alone can independently internalize the cells, the complexation with a peptide containing the RGD motif could greatly enhance the internalization ability and porphyrin accumulation into the cells. RGD is widely used in drug delivery applications and especially for tumor-targeted purposes due to the increased overexpression of integrins on cancerous cells, as described before.⁶⁸ It is known that the RGD motif located in the N-terminus of the chelator can be recognized specifically by 7–8 integrins, such as the a5b1, avb5, and αvβ3.⁶⁹ Moreover, in various tumor cells, there is overexpression of some RGD-specific integrin receptors.⁷⁰ The HeLa cells could overexpress at least 2–3 receptors, including the integrin avb3.⁷¹

Since porphyrins are also commonly used for photodynamic therapy (PDT) for anticancer treatment, this combination of the RGD peptide bound with a porphyrin molecule could hold great potential for antitumor applications.

Furthermore, it should be noted that for this kind of theranostic molecules, sustaining blood retention is of great importance. Previous studies demonstrated the importance of the RGD motif as a blood circulation-prolonging (BCP) peptide.⁷² Usually, the endogenous ligands in the bloodstream compete with the drug delivery vectors for binding to the tissue target. The presence of a target motif, like the RGD motif for tumor cells, greatly enhances the receptor uptake of cells in contrast to the nontargeted vectors.⁷³

Based on our observations and having in mind the advantages of the RGD motif, one could consider that the coordinated Py₃P-Lys-NTA with the RGDSGAITIGH peptide could extend the retention time and the accumulation quantity of the porphyrin into the cells. More experiments in vivo are needed to testify this potential use of porphyrin.

Conclusions

Exploitation of functional organic molecules for the decoding of biological functions within organisms has been of great importance in the past few decades. Complexation of appropriately bound chromophores through chelate bonding with metals and amino acids, known as metal–chelate binding, can boost the research in the field of labeling and imaging.

In this work, three new porphyrin–NTA dyads, TPP-Lys-NTA, TPP-CC-Lys-NTA, and Py₃P-Lys-NTA connected to a nitrilo triacetate metal–chelate ligand through different bridging groups, were successfully synthesized and spectroscopically analyzed. According to fluorescence measurements, complexation of each dyad with Ni²⁺ ions is possible, making them good candidates as potential probes. Alterations in the periphery groups of the porphyrin ring as well as the distance between the chromophore and the chelator can overcome the solubility issues and the quenched emission of the fluorophores after their interaction with Ni²⁺ ions. The synthesized hybrids have the ability to coordinate with histidine-containing oligopeptides through metallochelate coupling. Histidine-containing peptides are particularly appropriate for intracellular delivery of labeling and imaging chromophores or other types of cargo. The imidazole group of histidine, with a pK of around 6, protonates in intracellular conditions, especially within the endosomes in increasingly acidic conditions. This results in endosomal lysis and release of the chromophore or other types of cargo in the cytoplasm.⁷⁴ Consequently, chemically differentiated porphyrin derivatives may contribute to the increase of new and efficient probes for labeling and other biological applications, both in vitro and in vivo. Toward this direction, further experiments with peptides and the aforementioned porphyrin–NTA compounds are foreseen.

Experimental Procedures

Materials

Porphyrins 1,⁴⁸5,⁶³8,⁴⁸ and 9(48) and TPyP(75) as well as NTA derivative 3(53) were prepared following procedures already reported in the literature. DMF was freshly distilled from MgSO₄. The oligopeptides Fmoc-FH, RGDSGAITIGH, and RGDSGAITIGF were obtained from Genecust, France, while all other chemicals and solvents were purchased from commercial sources and used as received.

NMR Spectra

NMR spectra were recorded on Bruker AVANCE III-500 MHz and Bruker DPX-300 MHz spectrometers using solutions in deuterated solvents, and the solvent peak was chosen as the internal standard.

Mass Spectra

Mass spectra were obtained on a Bruker UltrafleXtreme matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) spectrometer using trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylidene] malononitrile (DCTB) or α-cyano-4-hydroxycinnamic acid (CHCA) as matrices.

Spectroscopy

Absorption spectra were recorded on a Shimadzu UV-1700 spectrophotometer, and steady-state emission spectra were obtained using a JASCO FP-6500 fluorescence spectrophotometer.

Density Functional Theory (DFT) Calculations

All theoretical calculations were performed following previously published procedures.⁷⁶⁻⁷⁸

Synthesis of Porphyrin 4

To a solution of mono-carboxyl porphyrin 1 (22 mg, 0.033 mmol) in dry dimethyl formamide (DMF) (3 mL), which was previously degassed with a stream of Ar, were added N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride and EDC·HCl (10.3 mg, 0.053 mmol) followed by N-hydroxysuccinimide, NHS (7.7 mg, 0.067 mmol). The reaction mixture was left under an Ar atmosphere and stirred at room temperature for 24 h, shielded from ambient light. The formation of the succinimidyl ester intermediate 2 was monitored by thin-layer chromatography (TLC) (CH₂Cl₂ was used as the mobile phase). Thereafter, in the same flask were added triethylamine, Et₃N (100 μL), and an excess amount of lysine trimethyl ester 3 (52 mg, 0.171 mmol) dissolved in DMF solution (2 mL). The reaction mixture was stirred at room temperature for 96 h. The progress of the reaction was monitored by TLC (CH₂Cl₂/MeOH, 95:5). After evaporation of the solvent under vacuum, the crude solid was dissolved in CH₂Cl₂ and extracted with water (3 × 30 mL). The organic layer was collected, and the solvent was evaporated. The crude mixture was purified by column chromatography (silica gel, CH₂Cl₂/MeOH, 98:2, then gradually increasing polarity to CH₂Cl₂/MeOH, 90:10) to afford the desired porphyrin 4 as a red-purple solid. Yield: 13.6 mg (43%).

¹H NMR (500 MHz, CDCl₃): δ = 8.87 (m, 6H), 8.81 (d, J = 5 Hz, 2H), 8.29 (d, J = 8.5 Hz, 2H), 8.23 (m, 8H), 7.77 (m, 9H), 6.93 (t, J = 5.5 Hz, 1H,), 3.69 (m, 15H), 3.55 (t, J = 6.5 Hz, 1H), 1.67–1.86 (m, 6H), −2.75 (s, 2H) ppm. ¹³C NMR (75 MHz, CDCl₃): δ = 173.4, 172.1, 168.0, 145.4, 142.2, 134.7, 134.3, 131.4, 127.9, 126.8, 125.7, 120.6, 120.4, 118.9, 64.4, 52.8, 51.9, 51.7, 40.1, 29.8, 28.7, 23.2 ppm. UV–vis (CH₂Cl₂): λ_max, nm (ε, mM^–1 cm^–1) 418 (373.8), 514 (14.8), 550 (6.2), 590 (4.2), 644 (3.2). HRMS (MALDI-TOF): m/z calcd for C₅₈H₅₃N₆O₇ [M + H]⁺: 945.3976, found: 945.3968.

Synthesis of Dyad TPP-Lys-NTA

Porphyrin 4 (13.6 mg, 0.014 mmol) was dissolved in 1:1 MeOH/THF solvent mixture (4 mL) and an excess amount of LiOH (47 mg, 1.12 mmol) was added at 0 °C. Then, the ice bath was removed and the reaction mixture was stirred at room temperature for 96 h. The organic solvents were evaporated under vacuum, and the obtained solid was dissolved in water (20 mL). The solution was acidified until pH ≈ 4 with the dropwise addition of HCl (0.1 M), and the formation of the precipitate was observed. The precipitate was filtered and washed with water to afford the dyad TPP-Lys-NTA as a red-purple solid. Yield: 12.5 mg (96%). ¹H and ¹³C NMR spectra could not be obtained due to solubility issues. UV–vis (DMF): λ_max, nm (ε, mM^–1 cm^–1) 418 (359.5), 514 (16.3), 548 (8.2), 590 (5.4), 646 (5.1). HRMS (MALDI-TOF): m/z calcd for C₅₅H₄₆N₆O₇ [M]⁺: 902.3428, found: 902.3437.

Synthesis of Porphyrin 7

To a solution of porphyrin 5 (42 mg, 0.054 mmol) in dry THF (9 mL), lysine trimethyl ester 3 (33 mg, 0.107 mmol) was added, followed by the catalytic amount of DIPEA (25 μL). The reaction mixture was stirred under an Ar atmosphere and at room temperature for 96 h, shielded from ambient light. The formation of intermediate porphyrin 6 was observed by thin-layer chromatography. Thereafter, in the same flask, an excess amount of piperidine (70 μL) and the catalytic amount of DIPEA (185 μL) were added and the reaction was allowed to stir for 24 h at 65 °C under Ar. The progress of the reaction was monitored by TLC (CH₂Cl₂/EtOH, 98:2). The solvent was evaporated under vacuum, and the crude product was purified by column chromatography (silica gel, CH₂Cl₂/EtOH, 98:2) to afford porphyrin 7 as a red-purple solid. Yield: 14.9 mg (25%).

¹H NMR (500 MHz, CDCl₃): δ 8.96 (d, J = 4.5 Hz, 2H), 8.86 (m, 6H), 8.23 (m, 6H), 8.16 (d, J = 8.5 Hz, 2H), 8.04 (br s, 2H), 7.77 (m, 9H), 3.86 (br s, 4H), 3.67 (m, 13H), 3.48 (m, 3H), 1.47–1.8 (m, 12H), −2.74 (s, 2H) ppm. ¹³C NMR (75 MHz, CDCl₃): δ 173.3, 172.0, 164,6, 142.3, 139.5, 136.1, 135.2, 134.7, 131.2, 130.6, 127.8, 126.8, 120.4, 120.2, 120.1, 117.9, 64.8, 52.5, 51.8, 51.6, 44.7, 40.5, 30.3, 29.5, 26.0, 25.0, 23.4 ppm. UV–vis (CH₂Cl₂): λ_max, nm (ε, mM^–1 cm^–1) 420 (321.9), 516 (13.8), 552 (7.3), 591 (4.0), 647 (3.3). HRMS (MALDI-TOF): m/z calcd for C₆₅H₆₄N₁₁O₆ [M + H]⁺: 1094.5041, found: 1094.5053.

Synthesis of dyad TPP-CC-Lys-NTA

In a solution of porphyrin 7 (14.9 mg, 0.014 mmol) in 1:1 MeOH/THF solvent mixture (4 mL), an excess amount of LiOH (39 mg, 0.93 mmol) was added at 0 °C. The reaction mixture was stirred at room temperature for 96 h. The progress of the reaction was monitored by TLC (CH₂Cl₂/MeOH, 95:5). The solvents were evaporated under vacuum, and the obtained solid was dissolved in water (20 mL). The solution was acidified until pH ≈ 4 with the dropwise addition of HCl (0.1 M). The formed precipitate was filtered and washed with water to afford the dyad TPP-CC-Lys-NTA as a red-purple solid. Yield: 13.9 mg (98%). ¹H and ¹³C NMR spectra could not be obtained due to solubility issues. UV–vis (DMF): λ_max, nm (ε, mM^–1 cm^–1) 420 (157.9), 516 (9.6), 553 (6.5), 592 (3.8), 648 (4.0). HRMS (MALDI-TOF): m/z calcd for C₆₂H₅₈N₁₁O₆ [M + H]⁺: 1052.4572, found: 1052.4584.

Synthesis of Porphyrin 10

Porphyrin derivative 9 (80 mg, 0.12 mmol) was dissolved in dry DMF (12 mL) under an Ar atmosphere. The solution was cooled to 0 °C, and HATU (184 mg, 0.48 mmol) was added. The reaction mixture was stirred at 0 °C for 1 h to achieve the activation of the carboxylic acid. Subsequently, lysine trimethyl ester 3 (74.5 mg, 0.25 mmol) and DIPEA (0.34 mL) were added and the reaction mixture was stirred at room temperature for 72 h. After the completion of the reaction, DMF was evaporated under vacuum; the crude product was dissolved in ethyl acetate (30 mL) and washed with H₂O (2 × 30 mL). Finally, the desired product was isolated through silica gel column of chromatography (CH₂Cl₂/MeOH (95:5) as eluent solvents) as a purple solid (Yield: 60%).

¹H NMR (300 MHz, CDCl₃): δ 9.05 (m, 6H), 8.85 (m, 8H), 8.27 (m, 4H), 8.16 (m, 6H), 6.99 (t, J = 5.27 Hz, 1H), 3.73 (s, 3H), 3.69 (s, 6H), 3.68 (s, 4H), 3.63 (m, 2H), 3.54 (t, 1H), 1.80 (m, 6H), −2.89 (s, 2H) ppm. ¹³C NMR (75 MHz, CDCl₃): δ 173.41, 172.15, 167.74, 150.03, 148.51, 144.60, 134.74, 134.64, 131.35, 129.48, 125.83, 120.51, 117.71, 117.47, 64.44, 52.79, 51.90, 51.71, 40.13, 29.84, 28.66, 23.16 ppm. HRMS (MALDI-TOF): m/z calcd for C₅₅H₅₀N₉O₇ [M + H]⁺: 948.3833 found: 948.3842.

Synthesis of Dyad Py₃P-Lys-NTA

To a MeOH:THF (1:1) solution (12.4 mL) of the porphyrin ester (10) (42 mg, 0.044 mmol), an excess amount of LiOH (149 mg, 3.55 mmol) was added and the mixture was stirred for 10 min at 0 °C. Then, the reaction mixture was stirred under mild temperature (∼40 °C) for 96 h and the progress of the reaction was monitored with TLC (DCM/MeOH, 95:5). The organic solvents were evaporated under vacuum, and the obtained solid was dissolved in water (15 mL). The aqueous mixture was acidified upon dropwise addition of HCl (3 M) solution until pH ≈ 4. The precipitate was filtered and washed with water to afford the final product as a purple solid. Yield: 40 mg (96%).

¹H NMR (500 MHz, DMSO-d₆): δ 9.04 (d, J = 4.35 Hz, 6H), 8.90 (m, 8H), 8.84 (t, J = 5.42 Hz, 1H), 8.27 (m, 10H), 3.43 (m, 4H), 1.60 (m, 6H), −3.03 (s, 2H) ppm. ¹³C NMR (75 MHz, DMSO-d₆): δ 174.26, 173.13, 166.25, 148.90, 148.41, 143.54, 134.21, 131.75, 129.23, 125.97, 120.29, 117.70, 117.46, 69.35, 62.68, 39.51, 28.75, 24.50, 23.53 ppm. UV–vis (DMF): λ_max, nm (ε, mM^–1 cm^–1) 416 (386.0), 512 (18.4), 546 (6.3), 587 (5.8), 643 (3.2) nm. HRMS (MALDI-TOF): m/z calcd for C₅₂H₄₃N₉O₇ [M + H]⁺: 906.3364 found: 906.3352.

Photophysical Studies with NiSO₄

Initially, we prepared the stock solutions of TPP-Lys-NTA, TPP-CC-Lys-NTA, and the reference porphyrin 8 in DMSO. The concentration in all cases was 4 × 10^–6 M. Then, an aqueous solution of NiSO₄ 3.8 × 10^–3 M was also prepared. For the absorption and emission experiments, 3 mL from the porphyrin stock solution was transferred to a quartz cuvette and the respective spectra were recorded. Then, 8 μL from the NiSO₄ solution was added to the cuvette, where the final concentration of NiSO₄ was 1 × 10^–5 M. Finally, the corresponding spectra were recorded.

Water Solubilization Protocol

One milligram of Py₃P-Lys-NTA was initially dispersed in 2 mL of deionized water with the support of an ultrasonic bath. Then, a few drops (approximately 5–6) of 25% aqueous NH₃ solution were added to the flask until the complete dissolution of Py₃P-Lys-NTA in water. Finally, the solvent was evaporated under reduced pressure and a solubility test was conducted with deionized water at neutral pH to confirm the hydrophilicity of the porphyrin after treatment with an aqueous NH₃ solution.

Preparation of the Solutions for the Metallochelate Coupling Studies

For the water-insoluble hybrid TPP-Lys-NTA, the following procedure was used: A stock solution (1.9 × 10^–2 M) of TPP-Lys-NTA was prepared by dissolving 0.86 mg in 50 μL of HFIP, while for the NiSO₄ stock solution (4 × 10^–3 M), we dissolved 0.5 mg in 800 μL of HEPES pH 7.4. Then, to a solution of the Fmoc-FH peptide (0.3 mg) in 15 μL of HFIP, we added 15 μL from TPP-Lys-NTA and 70 μL from NiSO₄ stock solutions. The final concentration of both TPP-Lys-NTA and NiSO₄ was 2.85 × 10^–3 M in this solution (solution 1). Moreover, the following control samples were prepared:

Solution 2: 0.3 mg of Fmoc-FH dissolved in 30 μL of HFIP and then diluted with 70 μL of HEPES pH 7.4.

Solution 3: 15 μL from the stock solution of TPP-Lys-NTA porphyrin was diluted with 15 μL of HFIP and 70 μL of HEPES pH 7.4.

Solution 4: 15 μL from the stock solution of TPP-Lys-NTA porphyrin was diluted with 15 μL of HFIP and then 70 μL from the NiSO₄ stock solution was introduced.

Solution 5: 0.3 mg of Fmoc-FH dissolved in 15 μL of HFIP and then 15 μL from the stock solution of TPP-Lys-NTA porphyrin and 70 μL of HEPES pH 7.4 were introduced.

For the water-soluble Py₃P-Lys-NTA derivative, the following procedure was applied: A stock solution (4.6 × 10^–3 M) of Py₃P-Lys-NTA was prepared by dissolving 0.5 mg in 120 μL of HEPES pH 7.4, while for NiSO₄, a stock solution of 4.6 × 10^–3 M concentration was also prepared. Peptides RGDSGAITIGH (4.6 × 10^–3 M) and RGDSGAITIGF (4.6 × 10^–3 M) were dissolved in HEPES pH 7.4 and incubated for 3 days at room temperature to self-assemble. Then, equimolar quantities of Py₃P-Lys-NTA and NiSO₄ were mixed and left at room temperature for 30 min. Twenty-five microliters from the above Py₃P-Lys-NTA-Ni²⁺ solution was added to 25 μL of the RGDSGAITIGH (solution 6) and RGDSGAITIGF (solution 7) peptide stock solutions. The final concentration of the porphyrin was 1.15 × 10^–3 M and that of the peptides was 2.3 × 10^–3 M in both cases. Moreover, the following control samples were prepared:

Solution 8: 25 μL from the Py₃P-Lys-NTA-Ni²⁺ stock solution was diluted with 25 μL of HEPES pH 7.4.

Solution 9: 25 μL from the RGDSGAITIGH stock solution was mixed with 12.5 μL from the Py₃P-Lys-NTA stock solution and then 12.5 μL of HEPES pH 7.4 was introduced.

Solution 10: 25 μL from the RGDSGAITIGF stock solution was mixed with 12.5 μL from the Py₃P-Lys-NTA stock solution and then 12.5 μL of HEPES pH 7.4 was introduced.

For all of the solid-state absorption experiments, 10 μL from each solution was transferred to quartz slides 2 × 2 cm² and left to dry overnight.

Field-Emission Scanning Electron Microscopy (FE-SEM)

Samples for FE-SEM analysis were prepared by depositing 10 μL from each peptide solution on a 12 mm coverglass and leaving to dry overnight. Then, the samples were covered with 10 nm Au sputtering and observed directly. FE-SEM experiments were performed using a JEOL JSM 7000F (FE-SEM) operating at 15 kV.

Cell Lines and Culture Conditions

Human epithelial cervical carcinoma cells (HeLa) were grown in Dulbecco’s modified Eagle’s medium (DMEM) growth medium (pH 7.4) supplemented with 10% fetal bovine serum (FBS) and 50 μg·mL^–1 gentamycin at 37 °C in a 5% humidified CO₂ incubator.

Confocal Microscopy Studies with Cells

In total, 8 × 10⁴ HeLa cervical carcinoma cells were seeded for 24 h in a 24-well plate after the addition of a 13 mm tissue culture coverslip at the bottom of the well and allowed to attach overnight. The following day, the culture medium was removed, 5 μL from solutions 1, 3, 5, and 7 was added to 495 μL of fresh DMEM and the mixture was subsequently added to the cell culture. Cells were incubated with the added solutions for 4 h at 37 °C. Afterward, the cells were washed two times with 1× PBS for 5 min, fixed with 4% formaldehyde for 15 min, and washed twice with 1× PBS. Cells were further treated with a drop of the DAPI nuclear staining dye. After that, a 13 nm coverslip was placed on top, and the internalization of the metallochelate conjugation system was assessed using a Leica SP8 inverted confocal microscope. The excitation wavelength for porphyrin Py₃P-Lys-NTA was 514 nm and that for DAPI was 405 nm. To avoid unspecific excitation of the porphyrin, the TPP-Lys-NTA hybrid was excited at 405 nm, but without DAPI nuclear staining. Considering these limitations in the latter case, pictures with no fluorescence filter were also taken to delimit the cells.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.1c05013.

• ¹H NMR and ¹³C NMR spectra of all of the synthesized compounds, frontier molecular orbitals of TPP-Lys-NTA and TPP-CC-Lys-NTA, absorption and emission spectra in solution and solid state, MALDI-TOF spectra of the porphyrin derivatives, and tables with the coordinates of the optimized gas-phase geometry structures (PDF)

Author Contributions

^∥ E.G., M.K., and C.P.A. contributed equally to this work.

The authors declare no competing financial interest.