The evolution of contemporary medicine prompts researchers to seek new approaches to drug design and administration, especially when it comes to highly toxic tumoricidal agents whose delivery to a biological target must be strictly precise. Attention is increasingly paid to monoclonal antibodies [1], biological transporters [2], nanostructures and nanoplatforms. Especially worth mentioning are such carriers as fullerenes [3], nanodiamonds [4], inorganic nanoparticles of different nature [5, 6], liposomes [7], nanoporous silicon [8], and hydroxyapatites (HAP) [9]. There are a few important goals that novel drug delivery systems are intended to achieve. First, they are expected to reduce the toxic effect of a carried drug on healthy organs and tissues. A good example here is doxil, the liposomal formulation of doxorubicin [10], Second, they can enhance the tumoricidal effect through the synergy of the drug/carrier complex [11]. For example, synthetic Buckminsterfullerene (C₆₀)-2-(butadiene- 1-yl)-tetra(о-γ-aminobutyryl-o-phthalyl) porphyrin (BFNP) nanoparticles can be used to deliver magnetic isotopes ²⁵Mg and ⁶⁷Zn, which have a biological activity of their own. Such complexes can alleviate metabolic acidosis induced by post-chemotherapy hypoxia [12]. Third, drug delivery systems aim to increase the bioavailability of the drug (which can be done by using, say, albumin as a carrier), ensure its sustained release (nanofullerenes, HAP) and enhance the tumoricidal effect by binding to the target specifically (monoclonal antibodies). Fourth, a carrier and a drug constituting a delivery system can have different tumoricidal effects. For example, fullerene derivatives capable of killing cancer cells are used as carriers of magnetic isotopes or radionuclides [13]. This principle is employed in designing radiopharmaceuticals for anticancer radiation therapy.
HAP nanoparticles are attractive drug and radionuclide carriers due to their biodegradability, biocompatibility and bioresorbability. These qualities are determined by their calcium-phosphate origin, which mimics the chemical composition of the human bone mineral fraction [14, 15].
The aim of this work was to study the properties of НАР obtained through different methods of synthesis (and therefore exhibiting different properties) and BFNP in relation to their use as bioactive metal ion carriers, including radionuclides, and ligands with tumoricidal activity.

METHODS

Zinc complexes

The N(5,6-dihydro-4Н-1,3-thiazine-2-yl)benzamide (L¹) ligand in the form of L1HBr was synthesized following a previously described technique [16]. To obtain the L1ZnCl₂ (С¹) complex, L¹HBr in the aqueous solution was converted to its basic form in the presence of NaOH (fig. 1). A solution of zinc chloride in diethyl ether at a ratio of 1 : 1 was gradually added to the L¹ base in diethyl ether. The solution was stirred for 0.5 h. The white crystalline precipitate was separated and washed with ether. The reaction yield was 64%. Ligand L2 was 2-aminopyrimidine (Sigma; USA), ligand L³ was 2-aminopyrimidine salicylate. Their complexes with zinc Zn(L²)₂Cl₂ (C²) and Zn(L³)2 (C³) were obtained as described in [17]. The composition of all complexes was characterized by element analysis and ¹H-NMR (Bruker CXP-200 spectrometer; Germany).

Spectrophotometry

Spectrophotometry was performed using the UV-1280 spectrophotometer (Shimadzu; Japan). Calibration curves were constructed for the ligands and the complexes in the aqueous, physiological saline and alcohol solutions.

Protonation and stability constants

The constants were determined by potentiometry using the automatic titrator Metrohm 848 Titrino plus (Metrohm AG; Switzerland). The stability constant was measured using the glass electrode. Computations were done in Hyperquad 2013.

HAP synthesis

Coprecipitation. There are different methods of HAP synthesis [18–20] producing end products with different parameters. We synthesized HAP₁ at room temperature using the stoichiometric Са/Р molar ratio of 1.67 [21]. All reagents were taken in the amount sufficient to produce a 5% (solid mass content) suspension [22]. Trace amounts of calcium and zinc in the residual liquid were determined by inductively coupled plasma atomic emission spectroscopy (ICP-AES) on Optima 100 DV (Perkin Elmer; USA). To obtain HAP₂, the suspension was continuously heated to 90 °С without boiling for 4 h.
Enzymatic hydrolysis. HAP_E was synthesized through the enzymatic hydrolysis of calcium glycerophosphate in the presence of alkaline phosphatase (Merck; Germany). Reaction conditions were varied, including the enzyme concertation, the medium (water, glycerol buffer, tris buffer), pH (7 to 10), calcium concentration and the reaction time. The glycerophosphate hydrolysis reaction was triggered by adding the enzyme solution to the calcium glycerophosphate solution taken at a concentration of 0.02 mol/l. Working concentrations of the enzyme ranged from 0.02 to 0.4 μg/ml. Depending on the conditions, the reaction time was varied between a few hours and a few days.
Introduction of zinc and copper ions during HAP synthesis. Weighted amounts of zinc or copper oxides (chemically pure, C. P.) corresponding to the anticipated Zn(Сu)/Ca molar ratio of 5 (15 mol%) were dissolved in H₃PO₄. The obtained solution was introduced dropwise to the suspension of Са(ОН)₂ under continuous stirring. The samples of HAP_Zn1, HAP_Zn2 and HAP_Cu were obtained. Some of those samples were dried, annealed in a muffle furnace in air at 900 °С for 3 h and subsequently studied with XPA (see below).

Electron microscopy of HAP samples

Some of the samples were sonicated in a sonic bath for 30 s at 22 kHz frequency and 50 W power. The samples were prepared for microscopy using a standard technique and then inspected under the JSM-6380LA microscope (JEOL; Japan) at the accelerating voltage of 20 kV.

Metal nuclides

^69mZn (Т_1/2 = 13.7 h; Е_γ = 438.7 keV) was produced by the photonuclear reaction ⁷¹Ga(γ, np)^69mZn as described in [21]. To obtain a labeled compound, a reaction of isotope exchange was carried out between the С¹ complex in physiological saline and ethanol solutions (1 : 1) and a concentrated solution of carrier-free ^69mZn eluted from a chromatography column. To obtain labeled С² and С³ complexes, the ligands in the basic form were treated with ^69mZnCl₂ solution under continuous stirring and heating.
^64,67Cu (Т_1/2 = 12.7 h and 61.8 h, respectively; Е_γ(⁶⁴Сu) = 1345.8 keV, Е_γ(⁶⁷Cu) = 184.5 keV) was produced by the reaction ^natZn(γ,np)⁶⁴Cu; ^natZn(n,β)⁶⁷Cu. Copper was separated from zinc by extraction in the 0.001% dithizone solution in CCl₄ followed by re-execration in 6М HCl and ion-exchange chromatography in the column with Cu-Resin (Triskem; France). Radioactivity of the reaction products was measured using a gamma-ray spectrometer with the GC 3020 HPGE detector (Canberra; USA). ⁶⁷Zn and ²⁵Mg (Isotope; Russia) are stable isotopes with s of –5/2 and +5/2, respectively (enrichment in ⁶⁷Zn is 94.5%, isotope ²⁵Mg frequency is 99.9%).

Size of nanoparticle

Nanoparticle sizes were measured by dynamic light scattering.

Thin-layer chromatography (TLC)

For TLC we used Silufol plates with different eluents. The plates were developed with iodine vapors.

Autoradiography (ARG)

After TLC, the plates with the obtained radioactive compounds were analyzed using Cyclone Plus, the storage phosphor system for digital ARG (Perkin Elmer; USA), and storage phosphor screens coated with BaFBr:Eu by the same manufacturer.

X-ray diffraction phase analysis (XPA)

The analysis was performed on the automated X-Ray diffractometer DRON-3 (Innoscope; Russia) with a Cu-K_α anode using the software supplied by the manufacturer. The size D of crystallites was determined by the Sherrer equation:

formula

Measurements of adsorption

Measurements were done using the ASAP 2000 analyzer (Micromeritics; USA). The specific surface area was calculated in the Micromeritics software supplied by the manufacturer.

MTT-assay

The protocol of the assay is described in [13]. The cell lines used in the present study included HL-60, K-562, MOLT-4, and MOLT-4 (res) (res. means the cells were resistant to asparaginase, one of the main drugs used in children with acute leukemia). The cells were cultured following the standard protocol. Statistical analysis included the Mann Whitney U test. Each series of measurements consisted of at least 5 tests carried out in 3 replicates.

Isolation of mononuclear cells from bone marrow and peripheral blood

Bone marrow (BM) cells collected from patients with B-cell acute lymphoblastic leukemia (B-ALL), acute myeloid leukemia (AML) and T-cell acute lymphoblastic leukemia (TALL) were kindly provided by D.Rogachev National Research Center of Pediatric Hematology, Oncology and Immunology. Blood tests demonstrated that the proportion of blast cells in the peripheral blood mononuclear fraction was at least 80%.
Cell morphology was inspected under the fluorescence microscope LEICA DM6000B (Leica Microsystems; Germany) or the confocal laser scanning microscope LSM 710 (Carl Zeiss; Germany). The images were captured and saved by a digital camera.

RESULTS

Relationship between the size of porphyrin-fullerene nanoparticles and their cytotoxicity

Some findings suggest that increased aggregation of nanoparticles (in particular, fullerene-based) promotes survival leading to a rise in LC₅₀, i.e. attenuates the cytotoxic effect of a drug [23–26]. However, there is counterevidence to that, mainly for nanoС₆₀. So, we decided to study the effect of aggregation in BFNP.
tab. 1 presents data on the survival of leukemia cell lines and BM cells of untreated patients with acute leukemias in the presence of BFNP loaded with a naturally occurring zinc isotope or enriched in magnetic ⁶⁷Zn. The tab. 1 demonstrates the relationship between LС₅₀ and the nanoparticle size. In all cases, cell survival improved with increased nanoparticle aggregation, but the specificity of the studied complexes did not change.
Sonication caused the average nanoparticle size to shrink to 15–20 nm. However, a week after the average nanoparticle size reached 20–30 nm; 2 weeks after, 30–40 nm; one month after it was as big as 50–60 nm.
Morphological examination revealed that not only nanoparticles but also cells treated with ⁶⁷Zn-loaded BFNP tended to aggregate increasingly (these data are not provided in the table).

Effect of ligands and HAP-ligand compositions on the cells

Both ligands and their complexes with metal ions, such as zinc or copper (which can be substituted with radionuclides to produce a radiopharmaceutical), can be regarded as potential tumoricidal chelating agents and used for HAP doping. Ligands and metal ions can be loaded to HAP together as complexes or separately. In both cases, the complex and its components will produce either synergistic or independent effects.

tab. 2 compares the survival of K-562 cells and lymphocytes of healthy donors (HD) in the presence of L¹ and complexes. Both healthy lymphocytes and MOLT-4 и MOLT-4 (res) cells demonstrated very high survival rates when treated with different HAP samples, including those that contained zinc and copper ions.

The studied chelators and complexes displayed a tendency to increased (decreased) toxicity against healthy lymphocytes and cancer cells. The lymphocytes of healthy donors demonstrated the following pattern of survival rate decline:
L¹ > C¹ > ZnCl₂ > C³ > ZnSal2(H₂O)₂ > C². With chronic myeloid leukemia cells (K-562), the pattern changed:
С¹ > L¹ > ZnCl₂ > ZnSal₂(H₂O)₂ > C³ > C².
The therapeutic window (or the therapeutic index TI = LC₅₀ (HD)/LC₅₀ (leukemia cells)) declined in the following fashion:
L¹ > C³ ~ C¹ > ZnCl₂ > C² > ZnSal₂(H₂O)₂.
This leads us to conclude that all 3 studied complexes have a potential to be considered as tumoricidal agents.

When comparing the toxic effects of С² on leukemia cell lines and BM cells obtained from patients with B-ALL, we observed the following pattern: HD cells > K-562 > HL-60 ~ MOLT-4 > MOLT-4(res) ~ BM (B-ALL). Importantly, С² exhibited higher toxicity towards MOLT-4 (res) than to MOLT-4 cells, meaning it has specificity to the cells with the most resistance to chemotherapy. The complexes tended to exhibit higher cytotoxicity than the ligands, perhaps due to the development of necrosis in addition to apoptosis (fig. 2).

Stability of ligands and their complexes in aqueous/ physiological saline solutions is their important property. Unfortunately, much more common are bioactive chelating ligands and complexes that are poorly soluble. They need a special shell or a carrier to be delivered to a target. HAP meets this requirement only partly.
Spectrophotometric measurements of stability of the studied complexes in water, ethanol and physiological saline solutions revealed that С¹ was the least stable complex: it tended to hydrolyze over time producing thiazine that, however, has mild tumoricidal and strong radioprotective properties. Other complexes were quite stable when dissolved, which makes them more suitable for clinical purposes.
As determined by the potentiometric titration of the ligand L¹ performed in aqueous and physiological saline solutions with varying рН (at С (L¹) = 1 • 10^–3 mol/l, I = 0.15 NaCl (0.1 mol/l KNO₃)), the values of protonation constants (log K) were 5.1 ± 0.1 (aqueous solutions) and 5.3 ± 0.2 (physiological solutions). Attempts to potentiometrically determine the stability constant of С¹ failed. This might have been due to the production of zinc hydroxide that interfered with titration. For С² the stability constant logK (С²) was 10.4 ± 0.5.

Behavior of nanoHAP doped with ligands, metal ions and/or complexes

fig. 3 (А, B) shows sorption of the ligand L¹ by HAP₁ and HAP₂. Changes occur when sorption starts and are probably due to the interaction with calcium released as nanoHAP dissolves. It means that the ligand L¹ does not bind to hydroxyapatite. The same behavior was observed for L² and L³.
The morphology of nanocrystals can be visualized using electron microscopy, while the composition of the solid phase (biohydroxyapatite) can be inferred from the results of X-ray diffraction phase analysis. For HAP₁, results of these analytical modalities are presented in fig. 4. When subjected to heating, both zinc-loaded and “pure” nanoparticles shrank in size. This transformation was considerable for “pure” samples and almost insignificant for those doped with zinc (tab. 3).

The major phase of HAP_E (fig. 5 А) was hydroxyapatite (fig. 5 В); HAP_E nanoparticles tended to have a spherical shape and formed aggregates. The specific surface area of the studied sample determined by nitrogen absorption was 300 m²/g. Pore sizes varied considerably (2–300 nm). The size of crystallites in the samples with crystal structure was calculated using the Sherrer equation (12–14 nm). Crystallites and aggregates significantly varied in size; their average dimensions are shown in fig. 6.
We synthesized 6 different types of HAP: 3 without metal ions (HAP₁, HAP₂, HAP_E) and 3 with metal ions (HAP_Zn1, HAP_Zn2, HAP_Cu). HAP_E were big-sized hollow spheres that exhibited increased sorption capacity. Averages sizes of other nanoparticles were as follows: HAP₁ > HAP_Zn1 > HAP_Zn2 ~ HAP_Cu > HAP₂.

We measured specific adsorption of zinc/copper ions by HAP₁ and HAP_E and calculated adsorption isotherm constants using Langmuir and Freundlich methods. Adsorption of copper ions corresponded to the Langmuir model, while adsorption of zinc ions (with equal correlation coefficients) could be described by both Langmuir and Freundlich models.
The sorption capacity of HAP towards copper ions was significantly higher than towards zinc ions. Introduction of zinc into the reaction during HAP synthesis did not seriously affect the size of nanoparticles. However, average sizes of a thermally processed HAP₂ were smaller than those of HAP₁. HAP₂ particles acquired a more regular isometric shape growing in thickness, as compared to HAP₁, and therefore had a smaller specific surface area. As a result, the maximum sorption capacity of HAP₂ decreased.

Complexes with zinc and copper radionuclides

Complexes ⁶⁹mZnC₁ (I), ⁶⁹mZnC₃ (II) and [(L²)₂^64,67CuСl₄] (III) were obtained through isotope exchange and analyzed by TLC, ARG and
γ-spectroscopy. The time of isotope exchange was selected experimentally to be sure that the R_f values of radioactive complexes and those that did not contain a radionuclide were the same and also to prevent formation of side products. The obtained compounds were sorbed onto HAP and physical and chemical measurements were then repeated. The ready pharmaceuticals had similar properties, but the copper complex disintegrated during HAP₁ doping. HAP₁ absorbed only copper ions. Perhaps, this problem can be solved by loading HAP with the copper complex during HAP synthesis.

DISCUSSION

Previously, it was demonstrated that the magnetic isotope ²⁵Mg (amounting to 11% in the natural isotope mixture) hyper-activates magnesium-dependent regulation of ATP synthesis, which makes the delivery of this isotope to hypoxic tissues/ cells an ambitious pharmacological task [12]. Here, BFNP nanoparticles capable of sustained release of Mg²⁺ and Zn²⁺ ions can be used as carriers. However, the major constraint for the clinical application of fullerene derivatives is their possible uncontrolled aggregation accompanied by changes in their initial cytotoxicity and physiological properties. Our study demonstrates that size transformations do not lead to critical changes in the properties of nanoparticles but attenuate their cytotoxic effect. The fundamental finding of this study is that survival of BM cells of untreated patients with B-ALL decreases 6–7-fold in the presence of ⁶⁷Zn-BFNP, as compared to ^natZn- BFNP. Aggregation of nanoparticles can be slowed down by sonication.

HAP nanoparticles have good prospects as drug delivery systems. There are a lot of methods for their synthesis that produce particles of different sizes, shapes and sorption capacity. HAP nanoparticles are also lowly toxic. Introduction of ligands and complexes into the reaction during HAP synthesis is a clever trick that helps the components of the complex to retain their properties. Enzymatic synthesis seems to be especially promising as it allows varying reaction conditions and, therefore, the parameters of the end product. This significantly expands the area of the potential application of nanoparticles.

Zinc and copper are essential micronutrients [27, 28]. They are cofactors for the majority of regulatory and antioxidant enzymes; they also are involved in DNA repair and the work of transcription factors. Complexes of short-lived zinc and copper radionuclides can be used to solve a number of diagnostic and therapeutic tasks, either simultaneously or consecutively. HAP can serve as a scaffold for a drug ensuring a synergistic sustained effect. To achieve it, a vector is needed, monoclonal antibodies being the most optimal. Importantly, HAP itself has an ability to integrate into bone tissue and become a vector. This matters in the therapy of bone and blood cancers aimed at eliminating malignant stem cells.

In our experiments, the specific activity of the obtained radioactive agents was not high, because the experiments were preliminary and aimed to demonstrate that such complexes can be synthesized in reasonable time for the use in the clinical setting. We have sufficient resources to improve their specific activity.

CONCLUSIONS

Transformation of ⁶⁷Zn-BFNP sizes following their aggregation does not affect their function but attenuates their cytotoxicity against leukemic cells. By applying different methods of HAP synthesis and processing, one can alter the sorption capacity of HAP towards metal ions, ligands and complexes. HAP_E with preprogrammed properties can be synthesized by varying reaction conditions. We have synthesized HAP nanoparticles containing short-lived zinc/copper isotopes in the form of ions and compounds. These nanostructures have a good potential to solve a number of diagnostic and therapeutic tasks in patients with cancer.