La³⁺, Ce³⁺, Eu³⁺, and Gd³⁺ Complex Formation with Hydrazones Derived from Pyridoxal 5'-Phosphate in a Neutral Tris–HCl Buffer

Abstract

Hydrazones derived from pyridoxal 5'-phosphate and 4-hydroxybenzohydrazide, 3-hydroxynaphthohydrazide, and L-tyrosine hydrazide have been synthesized. The complexation of La³⁺, Ce³⁺, Eu³⁺, and Gd³⁺ ions with these hydrazones in a neutral aqueous solution (0.05 M Tris–HCl buffer, pH 7.2) has been studied. The most probable stoichiometric composition and apparent stability constants of the complexes at a fixed pH value have been determined. The possibility of using pyridoxal-5-phosphate hydrazones to create fluorescent probes for bioimaging or contrasting agents for MRI has been considered.

Lanthanide(III) ions due to their electronic structure and fluorescent properties are widely used when creating biological fluorescent probes and contrast agents for magnetic resonance imaging. The choice of ion allows one to adjust the emission range of the fluorescent probe: Eu³⁺, Tb³⁺, Sm³⁺, and Ce³⁺ compounds give, respectively, red, green, orange [1], and from purple-blue to bluish-green light [2, 3], while Yb, Nd, and Er compounds emit in the near-IR region [1]. For MRI diagnostics, Gd³⁺ compounds are optimal because of the highest number of unpaired electrons (configuration f ⁷) [4]. For the synthesis of lanthanide(III) complexes, polydentate chelating or macrocyclic N,O-donor ligands are commonly used [5, 6]. In this case, it is necessary to take into account that the water molecules remaining in the coordination sphere of a metal ion deteriorate the luminescent properties of the complex, but are necessary for using it as a contrast agent in MRI [5]. The resulting complexes should have high stability to avoid the toxic effect of lanthanide cations [7].

It is of interest to investigate the stability of complexes of some lanthanides, namely, La³⁺, Ce³⁺, Eu³⁺, and Gd³⁺, with chelating N,O-donor ligands, such as hydrazones derived from pyridoxal 5-phosphate, in a neutral aqueous solution. It has been previously shown that such hydrazones are capable of binding d-metal ions to give strong coordination compounds [8–11] even in a phosphate buffer, and they can compete for metal ion even with some metal chaperone proteins [8]. However, the chelating of f-metal ions can be not as efficient, so that the pyridoxal 5-phosphate hydrazones should not be used to create new fluorescent probes or contrast agents.

EXPERIMENTAL

Hydrazones formed by pyridoxal 5-phosphate (PLP) and 4-hydroxybenzohydrazide (4HBH), 3‑hydroxynaphthohydrazide (3HNH), and L-tyrosine hydrazide (TYR) (Fig. 1) were synthesized as described in [11] using the reagents purchased from Sigma.

Fig. 1.

Structural formulas of hydrazones (a) PLP-4HBH, (b) PLP-3HNH, and (c) PLP-TYR with the atom numbering scheme used for signal assignment in NMR spectra.

A 25-mL volume of an aqueous solution of 1 mmol of pyridoxal 5-phosphate heated to 90°С was added to 25 mL of a water–ethanol (1 : 1 v/v) solution of the corresponding hydrazide also heated to 90°С. An instant formation of a crystalline product was observed, which was settled at room temperature for 1 h, filtered off, washed with ice water and acetone, and dried in air to constant weight. The yield of products was 80–85%; their structure was confirmed by ¹H and ¹³C NMR and IR spectroscopy. The assignment of NMR signals was based on 2D ¹H,¹³C HSQC experiments. The following designations were used for multiplets: s for singlet, d for doublet, t for triplet, and ddd for doublet of doublets of doublets. The following notation was used in the IR spectra: vs for very strong, s for strong, m for medium, and w for weak.

Pyridoxal 5-phosphate 4-hydroxybenzocarbohydrazone (PLP-4HBH). Pale yellow crystals.

¹H NMR (D₂O, pD ~13), δ, ppm: 8.32 (s, 1H, H₇), 7.44 (s, 1H, H₆), 7.28 (d, ³J = 8.7 Hz, 2H, H_10,14), 6.30 (d, ³J = 8.7 Hz, 2H, H_11,13), 4.60 (d, ³J = 4.0 Hz, 2H, H_5'), 2.05 (s, 3H, H_2').

¹³C NMR (D₂O, pD ~13), δ, ppm: 170.7 (C₈), 166.2 (C₃), 154.6 (C₂), 149.0 (C₁₂), 145.0 (C₇), 134.7 (C₆), 130.4 (C₅), 129.9 (C_10,14), 122.3 (C₉), 118.3 (C_11,13), 116.0 (C₄), 61.7 (C_5'), 17.8 (C_2').

IR, cm^–1: 3552 vs, 3477 vs, 3415 vs, 3236 m ν(NH, OH), 2961 w ν_as(CH), 1637 s amide-I, 1616 s, 1558 w, 1479 w, 1382 m, 1278 m, 1176 s, 1091 m, 1020 m. M [hydrazone + H] 381.92 (calcd. 382.07).

Pyridoxal 5-phosphate 3-hydroxynaphthoyl-2-hydrazone (PLP-3HNH). Bright yellow crystals.

¹H NMR (D₂O, pD ~13), δ, ppm: 8.05 (s, 1H, H₇), 7.93 (s, 1H, H₆), 7.27 (d, ³J = 7.9 Hz, 1H, H₁₆), 7.17 (s, 1H, H₁₈), 7.05 (d, ³J = 7.9 Hz, 1H, H₁₃), 6.89 (t, ³J = 7.3 Hz, 1H, H₁₅), 6.71 (t, ³J = 7.3 Hz, 1H, H₁₄), 6.62 (s, 1H, H₁₁), 4.59 (d, ³J = 5.0 Hz, 2H, H_5'), 1.93 (s, 3H, H_2').

¹³C NMR (D₂O, pD ~13), δ, ppm: 162.8 (C₈), 157.8 (C₁₀), 149.12 (C₃), 147.6 (C₇), 137.3 (C₂), 137.2 (C₁₈), 132.0 (C₁₂), 130.5 (C₁₃), 130.7 (C₁₇), 128.4 (C₁₆), 127.5 (C₁₅), 120.6 (C₁₄), 112.3 (C₁₁), 121.3 (C₉), 124.2 (C₄), 124.4 (C₅), 130.4 (C₆).

IR, cm^–1: 3550 vs, 3477 vs, 3411 vs ν(NH, OH), 2924 w ν_as(CH), 1637 s amide-I, 1616 vs, 1530 w, 1458 w, 1383 m, 1189 w, 1093 w, 1052 w. M [hydrazone – H] 430.03 (calcd. 430.09).

Pyridoxal 5-phosphate L-tyrazinoylhydrazone (PLP-TYR). Pale yellow crystals.

¹H NMR (D₂O, pD ~13), δ, ppm: 8.30 (s, 1H, H₇), 7.50 (s, 1H, H₆), 6.90 (d, ³J = 8.5 Hz, 2H, H_12,16), 6.53 (d, ³J = 8.5 Hz, 2H, H_13,15), 4.09 (t, ³J = 7.1 Hz, 1H, H₉), 2.75 (ddd, ³J = 21.8, ⁴J = 13.9, 7.1 Hz, 2H, H₁₀), 2.17 (s, 3H, H_2').

¹³C NMR (D₂O, pD ~13), δ, ppm: 172.9 (C₈), 163.4 (C₁₄), 160.5 (C₃), 159.9 (C₂), 149.1 (C₇), 151.3, 130.9 (C₆), 130.5 (C_12,16), 131.4 (C₁₁), 131.3 (C₄), 117.2 (C_{13, 15}), 125.2 (C₅), 62.3 (C_5'), 56.7 (C₉), 37.2 (C₁₀), 18.7 (C_2').

IR, cm^–1: 3550 vs, 3474 vs, 3412 vs ν(NH, OH), 2929 w ν_as(CH), 1638 s amide-I, 1515 m, 1382 m, 1263 m, 1069 w. M [hydrazone + H] 424.85 (calcd. 425.11).

In addition, the Gd³⁺ complex with PLP-4HBH was synthesized. To 25 mL of an ethanol–dimethylsulfoxide solution (1 : 1 v/v) of hydrazone PLP-4HBH (0.3 mmol, 0.1144 g), 5 mL of a GdCl₃ · 6H₂O solution in DMSO (0.15 mmol, 0.0558 g) was added at room temperature. The resulting solution was stirred for 3 h, and then the reaction mixture was poured into a beaker with cold distilled water. The deposited yellow-orange precipitate was filtered off, washed with distilled water, ethanol, and acetone, and dried to constant weight at 50°C. The product was studied by IR spectroscopy, cm^–1: 3415 vs, 3236 m ν(OH), 2919 s ν_as(CH), 2851 s ν_s(CH), 1616 s, 1510 w, 1454 w, 1384 m, 1272 m, 1177 w, 1081 w.

As compared with the spectrum of the free ligand, the spectrum of the complex shows a smaller number of absorption bands, especially in the range of OH and NH stretching vibrations and the amide-I band, and the bands arising from vibrations of the C=C and C=N bonds of aromatic rings and the bridging moiety of the hydrazone molecule are shifted. In addition, the spectrum of the complex shows a smaller number of bands corresponding to P=O and P–O–C (P–O–H) stretching vibrations at 1250–1280 and 1000–1100 cm^–1, respectively. The electronic absorption spectrum of the complex dissolved in a 0.05 M Tris–HCl buffer (pH 7.2) shows a new band at 395 nm. The major light absorption of the free ligand is observed at 300–310 nm where transitions occur in a single π–π–p–π-electron system. The mass spectrum does not show peaks corresponding to the complex of 1 : 1 or 1 : 2 stoichiometry, but there is a set of peaks m/z 719–725, the intensity of which is proportional to the natural abundance of stable gadolinium isotopes (¹⁵⁴Gd, ¹⁵⁵Gd, ¹⁵⁶Gd, ¹⁵⁷Gd, ¹⁵⁸Gd, and ¹⁶⁰Gd). This set of lines probably corresponds to the product of partial decomposition of the Gd(PLP-4HBH)₂ complex.

LaCl₃·7H₂O, Ce(NO₃)₃·6H₂O, Eu(NO₃)₃·6H₂O, and GdCl₃·6H₂O of chemically pure grade (redkiimetall.rf, Russia) were used as purchased. The content of lanthanide crystal hydrate varied within 99.0–99.4 wt %. Buffer solutions with pH 7.2 were prepared from tris(hydroxymethyl)aminomethane (Tris, Sigma-Aldrich). The buffer was brought to the desired value of the acidity of the medium, which was controlled potentiometrically, with standardized 1 M HCl.

All solutions were prepared in bidistilled water (κ = 1.7 μS/cm, pH 6.6).

Note that the reaction between aldehyde and hydrazide is reversible, although it is shifted towards the formation of hydrazone, and pyridoxal 5-phosphate forms stable Schiff bases with a buffering agent (Tris) [12]. A day later, changes were observed in the spectra of buffer solutions of hydrazones, indicating their decomposition and the formation of products with Tris. For this reason, only freshly prepared solutions were used.

The electronic absorption spectra of solutions of PLP-4HBH, PLP-3HNH, and PLP-TYR hydrazones, as well as mixtures of hydrazones and metal cations (C_hydr = (4.0 ± 0.7) × 10^–5 mol/L, ${{C}_{{{\text{L}}{{{\text{n}}}^{{3 + }}}}}}$ = (1–27) × 10^–6 mol/L) in Tris–HCl buffer solutions (pH 7.2) were recorded on a Shimadzu UV 1800 spectrophotometer in the wavelength range 210–500 nm and optical densities 0–1. Quartz cells with an optical pathlength of 1 cm were used. The error in determining the wavelength did not exceed 0.5 nm, the accuracy of measuring the optical density was ± 0.003. The temperature was maintained at 298.2 ± 0.1 K using an external temperature control unit.

The most probable stoichiometric composition of lanthanide(III) complexes with hydrazones and apparent stability constants were calculated from spectrophotometric titration data using the KEV software [13].

RESULTS AND DISCUSSION

Titration of solutions of PLP-4HBH, PLP-3HNH, and PLP-TYR hydrazones with solutions of lanthanide(III) salts in Tris–HCl is accompanied by typical changes in electronic absorption spectra. The short-wavelength absorption band caused by the transition in the π–π–p–π electron system combining the hydrazone molecule becomes weaker, and a new band appears in the long-wavelength range of the spectrum (Fig. 2). Inasmuch as both the metal-to-ligand and ligand-to-metal charge transfers are high-energy ones for f metals [14], it is unlikely that they are responsible for the new long-wavelength absorption band. It is more likely that the observed spectral changes are due to that the hydrazones adopt a definite conformation favorable for chelation and to that a highly charged ion appears near the ligand molecule.

Fig. 2.

Examples of spectrophotometric titration of solutions: (a) PLP-4HBH (3.39 × 10^–5 mol/L)–La³⁺ ((0–2.63) × 10^–5 mol/L), (b) PLP-3HNH (4.02 × 10^–5 mol/L)–Ce³⁺ ((0–1.09) × 10^–5 mol/L), (c) PLP-TYR (3.77 × 10^–5 mol/L)–Eu³⁺ ((0–9.95) × 10^–6 mol/L), and PLP-4HBH (3.93 × 10^–5 mol/L)–Gd³⁺ ((0–1.02) × 10^–5 mol/L) in 0.05 M Tris–HCl buffer (pH 7.2).

Similar spectral changes have been observed upon the complexation of d-metal ions with hydrazones derived from pyridoxal 5-phosphate [8–11].

The first task is to determine the stoichiometric ratio in which cations and ligands interact. To do this, it is necessary to know exactly how the coordination of the metal occurs.

It is known that the hydration shell of early lanthanide(III) ions includes nine water molecules, while the late lanthanides are characterized by a hydration number of 8 (ions of intermediate elements exist in an aqueous solution in the form of a mixture of nona- and octahydrates) [15]. In d-metal complexes, the studied hydrazones are tridentate ligands, as follows from the results of X-ray diffraction analysis of complexes with similar ligands [16–18]. They coordinate to the transition metal cation through the oxygen atom in 3-position of the pyridoxal 5-phosphate residue, the imine nitrogen atom, and the carbonyl oxygen atom. Lanthanide complexes with hydrazones or Schiff bases of a similar structure are also characterized by the coordination of the metal through the carbonyl (carboxyl) group [19–22], the imine nitrogen atom [19–24], and the side hydroxyl group [19, 21–23] or a nitrogen heteroatom [23, 24]. However, the ligands studied in this work, in contrast to those described in the literature [19–24], contain a phosphate group, which is also capable of complexing lanthanides. Its participation in the complex formation can be indirectly judged from the changes in the range 1000–1400 cm^–1 of the IR spectrum of the free ligand and its complex with Gd³⁺ (Fig. 3).

Fig. 3.

IR spectrum of the PLP-4HBH hydrazone and its complex with Gd³⁺ in KBr.

In particular, the P=O stretching vibrations of the free ligand [25] are observed as two bands at 1251 and 1278 cm^–1, whereas the spectrum of the Gd³⁺ complex shows only one band at 1272 cm^–1. The bands at 1020 and 1091 cm^–1, which can be assigned to P–O–C (P–O–H) stretches of the free ligand [25], also merge into one band at 1081 cm^–1 upon complexation. In addition, in the ³¹P NMR spectra of the PLP-4HBH hydrazone, the resonance is shifted upfield upon the complexation with La³⁺ (from 3.35 to –2.54 ppm). For comparison, in the ³¹P NMR spectrum of the zinc complex with PLP-4HBH, this shift is insignificant (to 3.17 ppm).

Hence, taking into account the high coordination number of f-metal cations and ligand denticity, one can hypothetically assume the formation of LnL', ${\text{LnL}}_{2}^{'}$, and ${\text{LnL}}_{3}^{'}$ complexes.

Because of the strong hydrolysis of lanthanide(III) ions even in a neutral medium (except La³⁺) and the formation of an insoluble Ce(III) compound with Tris observed by us previously [26], it was necessary to maintain a three- or fourfold excess of the ligand with respect to metal ions. This means that the graphical method is inapplicable for determining the stoichiometry of a complex [27]. In addition, in some cases [28, 29], such methods can give incorrect results. For this reason, in order to determine the most probable stoichiometric composition of the complex using the KEV software [13], the experimental spectral data were described using three models assuming the formation of LnL', ${\text{LnL}}_{2}^{'}$, and ${\text{LnL}}_{3}^{'}$ species. Among these models, the most plausible one was recognized as having a physical meaning and providing satisfactory statistical criteria. For example, when describing the experimental data in Fig. 2a, the assumption about the formation of an LaL' complex led to $\log \beta _{1}^{'}$ = 5.41 with a standard error of 5.70 log units, which is unsatisfactory. The assumption on ${\text{LaL}}_{2}^{'}$ complex formation of stability $\log \beta _{2}^{'}$ = 10.58 ± 0.32 shows much better results. The model according to which LaL₃ is formed gives an even lower value of the standard deviation ($\log \beta _{3}^{'}$ = 13.39 ± 0.15); however, to provide a good fit of the experimental optical densities with calculation, the molar absorptivities of a complex species in the wavelength ranges 280–316 and 338–354 nm should be negative, which has no physical meaning. Consequently, the most adequate model implies the formation of a bis-ligand lanthanum(III) complex. Similar calculations were performed for all metal–ligand combinations. Note that stoichiometric models implying the simultaneous formation of ML and ML₂ complexes, as well as ML₂ and ML₃ complexes, were also tested. In this case, the calculation gave a $\log \beta _{2}^{'}$ value close to that calculated in the model that took into account the formation of only ML₂, and the yield of the other complex was negligible ($\log \beta _{1}^{'}$ or $\log \beta _{3}^{'}$ were also small). Thus, the assumption of the formation of a single ${\text{LaL}}_{2}^{'}$ complex is the optimal way to describe the experimental data under the chosen concentrational conditions. The remaining places in the coordination sphere of cations are occupied by water molecules, which has detrimental effect on the fluorescent properties of the complexes, but is favorable for using the Gd³⁺ complex as a contrast agent for MRI [5]. The formation of a complex with three hydrazone molecules is likely hindered due to the steric factor associated with the need to place three bulky molecules around the central ion.

As previously [8–11], we determined the apparent stability constants of complexes of f-metal ions with hydrazones at a constant pH value provided by the use of a buffer solution. In this case, the protolytic equilibria of the ligand and the metal complex can be neglected. In addition to the target formation of ${\text{LnL}}_{2}^{'}$ (LnL' and ${\text{LnL}}_{3}^{'}$when testing other models), the stoichiometric scheme included the processes presented in Table 1.

Table 1. side processes that occur in the system containing 0.05 M Tris–HCl buffer, La³⁺, Ce³⁺, Eu³⁺, or Gd³⁺ ions, and PLP-4HBH, PLP-3HNH, and PLP-TYR hydrazones

The hydrolysis constants (K_h) of the second and third stages are given in the literature only for La³⁺; therefore, for Ce³⁺, Eu³⁺, and Gd³⁺, we took K_h1 = K_h2 = K_h3 by analogy with [26]. At the end of the titration, the yield of hydrolyzed forms of europium cations, most susceptible to hydrolysis, was ~4% (including 0.5% of Eu(OH)₃), and further titration of the ligand with a solution of a metal salt led to the turbidity of the mixture noticeable with the naked eye after one or two additions. The turbidity of the solution also limited the maximum total metal concentration during titration with a cerium(III) solution, but due to the formation of a poorly soluble complex with Tris, the yield of which did not exceed 20%. The yield of the complex with Tris for Gd³⁺, which reacted most strongly with the component of the buffer solution, was ~26% at the end of the titration. The yield of metal complexes of hydrazones was 40–70% of the total cation concentration at the end of titration.

In addition to the processes indicated in Table 1, we took into account the possibility of the formation of lanthanide(III) chloride or nitrate complexes. However, the stability constants of these coordination compounds are small [31–33], and the yield under experimental conditions was vanishingly small; therefore, the processes of LnCl²⁺ and ${\text{LnNO}}_{3}^{{2 + }}$ formation were excluded from the calculation scheme.

The calculated apparent stability constants of the complexes are given in Table 2. The errors in Table 2 represent the half-width of the confidence interval at a confidence level of 0.95 and a sample size of three or four experiments.

Table 2. Apparent stability constants ${\text{lg}}\beta _{2}^{'}$ of La³⁺, Ce³⁺, Eu³⁺, and Gd³⁺ complexes with PLP-4HBH, PLP-3HNH, and PLP-TYR hydrazones in an aqueous solution at T = 298.2 K and рН 7.2 (0.05 M Tris–HCl)

As follows from the data in Table 2, the nature of the ligand has practically no effect on the stability of the complex, except for the Eu³⁺ compound with PLP-TYR, which is ~0.5log units more stable than complexes with other hydrazones. The cerium(III) ion forms the least stable complexes with all chelating hydrazones.

From the dependence of the apparent stability constants on the radius of the lanthanide ion (Fig. 4), it can be seen that the Ln³⁺ complexes with PLP-4HBH and PLP-3HNH obey a pattern inverse to that noted in [34] for the La³⁺–Ce³⁺–Eu³⁺–Gd³⁺ series upon complexation with cyclic triamine in a DMSO medium. In this case, the cerium complex is the most stable, and the equilibrium constant decreases on going to europium and gadolinium.

Fig. 4.

Apparent stability constant of La³⁺, Ce³⁺, Eu³⁺, and Gd³⁺ bis-complexes with hydrazones PLP-4HBH, PLP-3HNH, and PLP-TYR as a function of the cation radius.

The abnormally high value of the stability constant of the Eu³⁺–PLP-TYR complex violates this pattern.

The apparent stability constants are relatively low, which significantly limits the possibilities of using lanthanide(III) complexes with hydrazones in living tissues as fluorescent sensors or contrast agents for MRI. The reason lies in the phosphate ions contained in the blood plasma (at a physiological pH of 7.4, they are predominantly in the form of ${\text{HPO}}_{4}^{{2 - }}$ and ${{{\text{H}}}_{2}}{\text{PO}}_{4}^{ - },$ and their total concentration in a healthy human is 0.8–1.4 mmol/L [35]). Lanthanides form not only complexes with hydrogen phosphate and dihydrogen phosphate ions [36], but also a poorly soluble compound LnPO₄ with a solubility product of the order of 10^–25–10^–26 [36]. This, in particular, is the basis for the medical use of lanthanum(III) carbonate to lower the level of phosphates in the blood in case of impaired renal function [4].

Calculations show that at a physiological concentration of phosphate ions of 1 mmol/L and pH 7.4, the decomposition of the studied lanthanide complexes with hydrazones with a concentration of 0.05 mmol/L to form LnPO₄ is inevitable. To maintain at least a 10% yield of a complex species, a stability constant on the order of ${\text{log}}\beta _{2}^{'}$ ~ 24.5 is required, ${\text{log}}\beta _{2}^{'}$ ~ 26 guarantees the retention of half of the complex species, and ${\text{log}}\beta _{2}^{'}$ ~ 28.5 corresponds to the binding of 90% of lanthanide(III) ions into a complex. Thus, we can conclude that pyridoxal 5-phosphate hydrazones are unsuitable ligands for Ln³⁺ ions from the point of view of application in biology or medicine.

The calculation results agree with the experimental data. The addition of a small amount of phosphate buffer (0.09 M, pH 7.4) to a solution of the Gd(III)–PLP-4HBH complex in 0.05 M Tris–HCl (pH 7.2) leads to immediate deposition of a yellowish white precipitate, and the spectrum of the filtered solution coincides with the spectrum of free ligand in a 0.05 M Tris–HCl medium (pH 7.2) (Fig. 5).

Fig. 5.

Electronic absorption spectra of (1) hydrazone PLP-4HBH in 0.05 M Tris–HCl buffer (pH 7.2), (2) a solution of the Gd³⁺–PLP-4HBH complex filtered off after the addition of phosphate ions, and (3) the Gd³⁺–PLP-4HBH complex in 0.05 M Tris–HCl buffer (pH 7.2).

CONCLUSIONS

The complexation of La³⁺, Ce³⁺, Eu³⁺, and Gd³⁺ ions with hydrazones derived from pyridoxal 5-phosphate and 4-hydroxybenzohydrazide, 3-hydroxynaphthohydrazide, and L-tyrosine hydrazide in an aqueous solution at pH 7.2 (0.05 M Tris–HCl buffer) and T = 298.2 K has been studied. All coordination compounds have a 1 : 2 stoichiometry. From spectrophotometric titration data, apparent stability constants of the complexes have been determined at a fixed pH of the medium. The nature of the hydrazone has no significant effect on the stability of the complexes, except for the Eu³⁺–PLP-TYR compound, the stability of which is higher than that of europium(III) complexes with other ligands by 0.5 log units. Among the studied cations, the least stable complexes are formed by Ce³⁺. The lanthanides(III) complexes with pyridoxal 5‑phosphate hydrazones are unsuitable for use as fluorescent sensors for studying living tissues, since they contain water molecules, which impairs luminescent properties. In addition, the relatively low stability of the complexes leads to their decomposition in the presence of phosphate ions in a physiological concentration range.