1 Introduction

Technological advances, especially in material chemistry, provided an increase in human life expectancy and, consequently, an increase in age-related diseases has been observed [1, 2]. In this context, materials that seek to replace or complement any tissues or organs of the body are called biomaterials [3,4,5,6,7]. They should be capable of guiding the functionality restoration of the tissue and be harmless to the organism [7, 8]. Among these biomaterials, temporary implants made of biocompatible polymers gained importance [5]. For this application, the most studied biomaterials are the polycaprolactone, the poly(lactic acid) [9], and the polyvinylpyrrolidone (PVP). Recently, the biocompatible and thermosensitive poly(N-vinyl caprolactam) (PNVCL) has gained interest due to its lower critical solution temperature (LCST) be near to the physiological temperature (~ 32–37 °C) [10, 11]. PNVCL has been used in several biomedical applications, such as transport and controlled release of drugs, microencapsulation of enzymes, support for cell growth [12], and as artificial cartilage and tendons [13]. This polymer is soluble below the LCST due to hydrogen bonding between some groups of the polymeric chain and water molecules.

The PNVCL contains large number of hydrophilic groups, but also hydrophobic ones. However, when the temperature of PNVCL aqueous solution is increased, the transition from hydrophilic to hydrophobic occurs at LCST, and the polymer phase separates [13]. This increase in temperature causes a partial displacement of water from the polymeric coil, weakening the hydrogen bonds, allowing intramolecular interactions between the hydrophobic segments of the polymer macromolecules [10]. Consequently, the polymer collapses, forming hydrophobic aggregates. This hydrophilic-hydrophobic transition can be reversed upon cooling the thermosensitive polymers [14].

Usually, implants are exposed to complex loads, which demands biomaterials of superior mechanical and chemical stabilities. In this sense, the use of nanocomposite allows the combination of the flexibility of the polymeric materials with the mechanical strength of ceramics. Bone tissue is a natural hybrid material composed of an inorganic phase of hydroxyapatite nanocrystals immersed in an organic phase, predominantly made of collagen type II, forming a hierarchical structure ranging from micro to nanoscale. In this sense, the incorporation of hydroxyapatite nanoparticles (instead of micrometric particles) in nanocomposites can have the capacity to recapitulate the organization of the extracellular matrix [15] and induce a faster and natural bone generation. It also has similar chemical structure to the mineral portion of bone, favoring osteoconductibility and osteoinductibility [16,17,18]. the presence of hydroxyapatite can improve the bioactivity due to the presence of surface hydroxyl groups that facilitate intermolecular interactions and the cell adhesion on the surface of the substrate [19, 20].

Thus, in this work, we propose the preparation of PNVCL-based nanocomposites with nanoparticles of hydroxyapatite (HA) as potential “scaffolds” in tissue engineering. The advantages of this type of material are innumerable, since PNVCL is biocompatible and HA nanoparticles are bioresorbable and can induce the osseointegration process and can improve the thermomechanical properties of the polymer.

2 Experimental

Hydroxyapatite nanoparticles (HA) were prepared by the hydrothermal processing of an aqueous solution of phosphate and calcium ions. The ammonium hydrogen phosphate solution (0.10 mol L−1) was added dropwise into the solution of calcium chloride (also at 0.10 mol L−1) under continuous and gentle stirring, and the pH was adjusted to 10 by adding aqueous ammonia solution (28%). Then, this solution of phosphate and calcium was transferred to an autoclave bottle and placed in an oven for 24 h at 140 °C. After this period, HA nanoparticles were centrifuged and dried at 60 °C. The nanocomposites were prepared by in situ polymerization following similar procedures from literature [14, 21,22,23]. In brief, HA nanoparticles were incorporated at concentration of 1% (by mass) into a solution of PNVCL monomer (1% w/w of HA nanoparticles) in dimethyl sulfoxide and azobisisobutyronitrile as a radical initiator. The system temperature was increased to 70 °C, 2% AIBN (mass of initiator/mass of monomer) was added to the system, and the reaction proceeded for 4 h under a nitrogen atmosphere. The nanocomposites were purified after being centrifuged at 10,000 rpm for 1 min and resuspended in hot distilled water four times. Then, the nanocomposites were dried at 50 °C in a laboratory oven with forced air circulation [14, 21,22,23].

These nanoparticles were evaluated by scanning (SEM) and transmission (TEM) electron microscopies using a Philips CM-120, and the powder X-ray diffraction (XRD) patterns were recorded at room temperature using a Rigaku DMax 2500PC with Cu Kα radiation, in the 2θ range of 10–90°, in step scan mode with step width 0.02° and step time of 1 s. Raman spectra of all materials were collected using a Renishaw microscope system 2000, with the 514.5 nm line of an argon ion laser, covering the range from 300 to 3500 cm−1. The measures of differential scanning calorimetric (DSC) were held using a thermogravimetric analyzer model Netzsch Phoenix 204, with heating rate of 20 °C min−1 in the range of temperature of − 100 to 200 °C, in a nitrogen atmosphere with a flow rate of 40 cm3 min−1. NMR spectra were obtained using a Bruker AVANCE III spectrometer, operating at 100.4 MHz for 13C. All of the NMR experiments were done with the probe at ambient temperature and were performed using gated high decoupling, using DMSO-d as solvent.

The lower critical solution temperature (LCST) of PNVCL solution and its nanocomposites (1 wt %) were evaluated by ultraviolet–visible transmission spectrophotometer in the temperature range 25–35 °C. Dynamic light scattering (DLS) analyses were performed on Malvern Zetasizer Nano ZS instrument at an angle of 173°, with a He − Ne 4.0 mW power laser operating at 633 nm. Samples were prepared by dispersing PNVCL and its nanocomposites in deionized water (1 wt %), and the hydrodynamic diameter was analyzed in triplicate at 25 °C, therefore at a temperature below the LCST.

All procedures using zebrafish (Danio rerio) embryos were performed in accordance with Animal Care and Use Committee protocols at Federal University of Goiás (#0017/2018). Adults zebrafish were housed at circulating aquarium system with controlled environment (27.5 ºC, 14 h light/10 h dark cycle). All embryos of the same spawns were randomly assigned and placed into a 96-well tissue culture plate (one per well) with 200 µL of base medium (14.6 g NaCl, 0.65 g KCl, 2.2 g CaCl2, and 4.05 g MgSO4) without any material (control) and exposed to pure PNVCL and nanocomposites of PVNCL with 1% w/w of HA nanoparticles. These embryos were evaluated throughout 168 hpf (hour post-fertilization), and four replicates were performed for each treatment. Embryos were evaluated under stereomicroscope and classified as viable (embryos with homogeneous and translucent blastomeres) or degenerated (embryos whit the presence of small, extruded, and heterogeneous blastomeres). Heart rate analyzes were assessed with the aid of microscopy connected to image system. Video clips (~ 15 s) were recorded and then evaluated to determine the heart rate (beats per minute).

3 Results and Discussion

The synthesized HA nanoparticles presented pure hexagonal phase, as observed in the X-ray diffraction (XRD) pattern (Fig. 1 left) and in the Raman spectrum (Fig. 1 right) [20]. Crystallographic coherence domains (crystallite size) estimated using Scherrer equation were about 5 nm. The main peaks observed in the Raman spectrum are typical of hydroxyapatite, where it is possible to observe the phosphate ν2 vibrations (431 and 450 cm−1), phosphate ν4 vibrations (585 and 610 cm−1), and phosphate ν1 PO43− vibrations (960 cm−1) [24, 25]. Small nanoparticles are commonly found as single phase. While catalytic applications usually employ a mixture of one or more HA crystalline phases to result in better catalysts, for biomedical applications could be recommend, but not mandatory, the application of nanometric nanoparticles of single phase.

Fig. 1
figure 1

XRD pattern (left) and Raman spectrum (right) of hydroxyapatite nanoparticles

Full size image

Well-defined nanorods of 170 nm in average length were observed in scanning and transmission electron microscopies images shown in Fig. 2 (SEM at left and TEM at right). This value compared with the smaller crystallite size calculated from the XRD pattern (5 nm) suggests the formation of polycrystalline nanorods by the union of several small HA nanoparticles during the hydrothermal processing. These SEM images also suggest dense HA nanorods, without any signal of porosity as usually occuring in synthesis of nanoparticles in the presence of surfactants. The histograms of diameter calculated from the SEM image (Fig. 3) show relatively narrow size distribution.

Fig. 2
figure 2

SEM (left) and TEM (right) imagens of hydroxyapatite nanoparticles

Full size image
Fig. 3
figure 3

Histograms of diameter (left) and length (right) of hydroxyapatite nanoparticles

Full size image

Although some polymeric nanocomposites can exhibit excellent mechanical and functional properties, one of the most serious problems for their use as biomaterial is the presence of residual monomers, which can be sufficiently toxic and make any biological application unfeasible. For this reason, it is of fundamental importance to verify the polymerization degree before any application, which can be done through nuclear magnetic resonance (NMR) spectroscopy. Figure 4 (left) shows the 13C NMR spectrum of PNVCL nanocomposite (upper) and the spectrum of pure PNVCL (bottom), without any signal of free monomers. Figure 4 (right) shows typical PNVCL Raman spectra, with absorption bands at 1980–1990 cm−1 of C-H stretch; C = O at 1635 cm−1 and amide band (C-N) at 1435 cm−1.

Fig. 4
figure 4

13C NMR spectra (left) and Raman spectra (right) of the PNVCL and of the PNVCL-HA nanocomposite

Full size image

The differential scanning calorimeter curves of pure PNVCL and its respective nanocomposite show the same glass transition temperature (Tg) at 147 °C, which were obtained by the first derivative of DSC curves. These values are similar to that published in previous studies [13, 15]. Interestingly, the Tg was unaffected by the presence of HA nanoparticles, at least at concentrations of 1%. The broad band observed in DSC curve of pure PNVCL reflects the presence of water molecules between the polymer chains. [19]

The transition temperature of PNVCL and its PNVCL nanocomposites could be determined through the cloud point temperatures, measured by ultraviolet–visible transmission spectroscopy. However, there is a fundamental difference between the pure PNVCL (Fig. 5 left), which shows a sharp transition and its nanocomposite (Fig. 5 right), which exhibits a diffuse transition. This diffuse profile was also observed in composites of PNVCL and silica nanoparticles due to modifications on the molecular motion of polymeric chains by the presence of nanoparticles [21].

Fig. 5
figure 5

UV–VIS spectra as function of the temperature for PNVCL (left) and PNVCL-HA with 1% wt (right) showed LCST temperature transition

Full size image

The hydrodynamic diameters, measured by dynamic light scattering (DLS), showed different values for pure PNVCL (32 nm) and for the PNVCL/HA nanocomposite (350 nm) [10], which indicates the influence of the HA nanoparticles on polymeric chains conformation in aqueous solution, as previously observed in the profile of phase transition. These two experimental results strongly suggest that the presence of HA nanoparticles modifies the globalization of the PNVCL chains [21, 26]. In fact, size distribution of hydrodynamic size is quite narrow for pure PNVCL and usually occurs in the range from 10 to 35 nm [23]; however the presence of nanoparticles modifies this distribution, increasing the average hydrodynamic size due the particle size [14]. The larger the nanoparticle size, the larger the hydrodynamic diameter at temperatures below the LCST. Evidently, the hydrodynamic will reflect the presence of hydrophobic globules formed at temperatures above the LCST [14, 22. 23].

New biomaterials, including drug delivery systems, need animal experimentation to demonstrate biocompatibility, which can be effectively evaluated through zebrafish embryos, a promising bridge model between in vitro and in vivo research. Fish embryo toxicity (FET) tests utilize zebrafish embryos to assess the potential human toxicity of new biomaterials since 70% of human genes have an orthologue in zebrafish genome. Although obvious differences exist between mammalians and fishes, zebrafish is recognized as an excellent model for pre-screening studies of new biomaterials and drugs [23].

In general, the interaction between biomaterials and zebrafish embryos results in the disruption of embryonic development. There are several visible indicators that can be used to evaluate the toxicity of a material, which includes viability, altered growth, abnormalities in brain morphology, pharyngeal arches, and jaw, fins, notochord, tail, body shape, cardiovascular function, and locomotor and sensory function [26, 27]. In this study, embryos were classified between viable and degenerated. Average survival ratio of 95% of embryos after exposition to different amounts of pure PNVCL (from 0.02 to 50 wt %) during 96 h (Fig. 1) revealed a polymer of low toxicity to zebrafish embryos. The embryo eclosion was another parameter used to measure cytocompatibility, which also confirmed the low toxicity of pure PNVCL.

Nanocomposites of HA/PNVCL at different concentrations (0.02 to 50 wt %) showed an average survival ratio of 93% after 96 h (Fig. 2); however only 10% of survival ratio when exposed to 50% of HA/PNVCL nanocomposites.

Small numbers of teratogenic embryos were identified (Fig. 6). The most common alterations were spinal alterations, and 3.33% of embryos are exposed to PNVCL. However, 0.83% of embryos had this alteration after ZET with HA/PNVCL. This can indicate an influence of HA in the formation of spinal cord. Muto et al. (2011) [28] showed an influence of calcium in motor neurons during embryo development that coordinate muscle movements and normal spinal cord development. The percentage of malformations that stood out was pericardial and yolk sac edema with 5% in the treatment exposed to PNVCL-HA. A delay in hatching was observed in 2.50% of embryos, even so, a value that did not influence the embryos that, with 96hpf, there was 100% hatching, both in embryos exposed to PNVCL and PNVCL-HA. It can be noted that PNVCL and PNVCL-HA treatments are harmful to tested embryos (Fig. 3).

Fig. 6
figure 6

Images of different times of exposure to HA/PNVCL or PNVCL embryos. (A, D, G and J) control embryos without abnormalities, (B, E, H, and K) wmbryos exposed to HA/PNVCL, and (C, F, I and L) embryos exposed to ★ PNVCL. Indicates delay on embryo development (B, E and K); arrow ● indicates edema of pericard (E, I) and alteration of spinal cord (F, K); indicates embryos that did not ecloded after 96 hpf

Full size image

Hydrogels for biomedical application mimic the complex native tissue microenvironment due to their porous and hydrated structure. Although new applications have been emerged, most of them demand properties that are absent in ordinary polymers used in hydrogels but are possible to be introduced through the insertion of functional nanoparticles, such as carbon-based nanomaterials, inorganic ceramic, or oxide/metallic nanoparticles. These nanoparticles, which can be covalently bonded to the polymer or simply mixed with the hydrogel, opened a new field of biomaterials [29]. Nanometric hydroxyapatite has been extensively studied as local source of calcium or phosphates for bone or teeth regeneration together with some natural or synthetic polymers but only recently with intelligent polymers [30]. Here we presented a new biomaterial that combines biocompatibility with a thermoresponsive behavior that can be potentially used as injectable material for bone regeneration.

4 Conclusions

PNVCL-HA nanocomposites were prepared by in situ polymerization of PNVCL monomer in the presence of HA nanoparticles. The complete polymerization was showed by typical 13C NMR spectrum. The differential scanning calorimeter analysis showed the same glass transition temperature (Tg) at 147 °C for the PNVCL and its respective nanocomposite, which is the typical value for the dry polymer. The PNVCL and the nanocomposites presented LCST with similar values around 34 °C, though the PNVCL/Ha nanocomposites exhibited a diffusion phase transition between 30 and 34 °C. The dynamic light scattering (DLS) resulted in an increase of the hydrodynamic diameters of 32 nm for pure PNVCL polymer to 350 nm for the PNVCL/HA nanocomposite, indicating that the HA nanoparticles influenced in polymeric chain conformational in aqueous solution. The average survival and eclosion of embryos reveled low acute cytotoxicity to zebrafish embryos.