Polydopamine-based surface modification of paclitaxel nanoparticles for osteosarcoma targeted therapy

In order to achieve the purpose of targeting treatment of osteosarcoma, we developed novel paclitaxel (PTX) nanoparticles (Nps) coated with polydopamine (PDA) and grafted by alendronate (ALN) as ligand. Dopamine can be easily polymerized on various surfaces to form a thin PDA film in alkaline environment, which provided a versatile platform to perform secondary reactions for compounds without functional groups. The targeting Nps had a mean particle size of 290.6 ± 2.2 nm and a zeta potential of −13.4 ± 2.7. It was stable in phosphate buffer saline (PBS, pH 7.4), 5% glucose, plasma and displayed sustained drug release behavior. In vitro assay demonstrated the targeting Nps had stronger cytotoxicity against K7M2 wt osteosarcoma cells than the non-targeting Nps. Furthermore, in vivo distribution study indicated they could accumulate much more in tumor than non-targeting Nps. This is consistent with the in vivo antitumor study, targeting Nps achieved a better therapeutic effect than Taxol (8 mg kg−1, i.v.) (71.85% versus 66.53%) and prominently decreased the side effects of PTX. In general, the PTX-PDA-ALN-Nps may offer a feasible and effective strategy for osteosarcoma targeted therapy.


Introduction
Osteosarcoma is the most ordinary bone neoplasms, occurring predominantly in teenagers and children with practically 400 new cases every year in the United States [1][2][3][4][5]. It is characterized by high proclivity for early systemic metastases (such as lung metastasis) and local invasion [6,7]. According to the National Cancer Institute database of the USA, the five year survival rates of patients diagnosed with metastasis Nanotechnology Nanotechnology 30 (2019) 255101 (12pp) https://doi.org/10.1088/1361-6528/ab055f 5 These authors contributed equally to this work. 6 Authors to whom any correspondence should be addressed.
Original content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. cancer is only 5%-20% [8,9]. Although chemotherapy has extremely enhanced the survival in patients with osteosarcoma, some severe problems still remain, including relapse, metastatic disease and serious side effects. In addition, some patients do not respond to chemotherapy and others showed multidrug resistance [10][11][12]. Thus, improving targeted chemotherapeutic drugs is an important approach for the treatment of osteosarcoma. [13][14][15].
Paclitaxel (PTX) could congest cancer cells at the G2/M phase and stabilized microtubules to stimulate apoptosis. Furthermore, PTX exhibits significant antitumor activity against osteosarcoma by enhancing nuclear factor-B (NF-B)'s expression, a transcription factor, which adjusts various physiological processes, such as differentiation, apoptosis and inflammation [16,17]. Hence, we chose PTX as a model drug for the treatment of osteosarcoma. However, the low bioavailability, high toxicity and poor water solubility of PTX extremely hamper its therapeutic effect and limit its clinical application [18]. Among various nanoscale drug delivery systems, Nanoparticles (Nps) formulations have been studied for the transport of insoluble drugs, including chemotherapeutic drugs [19]. Nps always have high drug loading and can be straightly injected intravenously in the form of solid particles, which can be prepared by bottom-up or top-down method [8,20,21]. P188 (or Pluronic F68) is a poloxamer which can be used to as a stabilizer and improve the solubility, absorption and bioavailability of Nps [22]. In order to increase the therapeutic efficacy, we prepared PTX into Nps to improve its poor solubility. In this manuscript, Doxacurium Chloridew was used as surfactant, which could decrease surface tension and be used as stabilizer in nanoparticle system [23,24].
Targeted Nps are a feasible method to reduce the side effects and improve the locally effective drug concentration at tumor [25][26][27]. By the way of grafting a variety of active targeting agents (peptides, antibodies and nucleic acids), drugs can be delivered specifically by the formulation of Nps to the tumor cells and tissues. Tetracycline, small peptide aspartic acid, alizarin red and bisphosphonate are recognized functional ligands in drug delivery for bone-targeted [28][29][30]. Owing to high affinity to bone and curative effect on bone diseases of bisphosphonates, such as alendronate (ALN) and etidronate, which are commonly adopted as bone-targeting ligand. In particular, protein tyrosine phosphatase is one of the bisphosphonate binding receptor, which is over expressed on osteosarcoma cell [28,31]. Thus ALN can increase the accumulation of targeted Nps into the osteosarcoma cells, thereby achieving the purpose of treating osteosarcoma [32].
Actually, because of the short of activity groups, some materials can not be modified diametrically through functional ligands [33]. In addition, the polymers' chemical properties could be changed by the added ligands, the pre-functionalized polymers may lose its capacity to encapsulate and keep the drug. To resolve this problem, dopamine polymerization was used as a method to functionalize the surfaces of Nps [34]. Dopamine can self-polymerized and deposition on various materials' surface to emerge a thin polymerized dopamine (PDA) film. More importantly, PDA can easily immobilize functional ligands (nucleophiles containing amine and thiol groups) on its surface by Schiff-base reaction and Michael addition reaction [35]. All this method required was just a short incubation of Nps in weak alkaline solution with dopamine, then added ligands in the solution for secondary reaction [17,36,37]. Using this method, we have fabricated a PDA-coated PTX-Nps with ALN (PTX-PDA-ALN-Nps) through surface modification for osteosarcoma targeted therapy.

Preparation of PTX-Nps and PTX-PDA-ALN-Nps
Briefly, PTX (50 mg) was put in ethanol (1 ml) then transferred into 10 ml aqueous with poloxamer 188 (50 mg) and doxacurium chloride (20 mg) under magnetic stirring (1200 rpm) for 20 min at room temperature. Then the suspension was stirred for 20 min at the same speed and ultrasound at 100 W for 5 min. The PTX-Nps were obtained by centrifugation (13 000 rpm, 15 min) and washed three times by deionized water (preparation procedure one).
The resultant PTX Nps were immersed by 1 ml Tris buffer (10 mM, pH 8.8) containing 0.5 mg of dopamine and stirred constantly for 2 h at room temperature (preparation procedure two). Subsequently, the PDA coated PTX Nps were collected by centrifugation (13 000 r min −1 , 10 min). ALN was adopted for particle surface functionalization by conjugating with PDA coatings. Simply, 10 mg of ALN was dissolved in 10 mM Tris buffer with pH 8.8 and PDA coated PTX Nps were added in. After constant stirring for 3 h, the resultant Nps were centrifuged (13 000 rpm, 10 min) and washed three times with deionized water to obtain PTX-PDA-ALN-Nps (preparation procedure three).
To obtained Dir labeled Nps, 'PTX and Dir (w/w= 40:1)' was put together in ethanol instead of 'PTX', then 'preparation procedure one, two and three' were repeated. To obtained drug-free Nps, '1 ml ethanol (without PTX)' was transferred into 10 ml aqueous with poloxamer 188 (50 mg) and doxacurium chloride (20 mg), then 'preparation procedure one, two and three' were repeated. Instruments, UK) at room temperature. Every specimen was measured in three times and the average±standard deviation were used. The Transmission Electronic Microscopy (TEM, JEOL Ltd, Tokyo, Japan) was employed to ascertained the morphology of Nps. The Nps were suspended in 1% (w/v) double color pomelo and dropped on copper grids.

Drug loading capacity (LC) and encapsulation efficiency
(EE). The amount of PTX was detected by high-performance liquid chromatography (HPLC; Ultimate 3000, DIONEX, USA) for measuring the LC and EE of nanoparticles. The drug-loaded nanoparticles were suspended in methanol and vortexed for 3 min. After that, the supernatant was obtained by centrifugation (13 000 r min −1 , 5 min) and transferred into mobile phase, which consistent with water and acetonitrile (50:50, v/v). A C18 column (5 μm, 250×4.6 mm, Waters Symmetry) was adopted at 35°C under a UV-vis absorption at 266 nm, with a flow rate of 1.0 ml min −1 , and an injection volume of 20 μl. The following equations (1), (2) were used to calculated the LC and EE of nanoparticles, respectively

Stability of PTX-PDA-ALN-Nps in various physiological
solutions. To evaluate the suitability of PTX-PDA-ALN-Nps for intravenous injection, the size change in various physiological media including physiological saline, isotonic glucose (5% glucose), phosphate buffer saline (PBS, pH 7.4) and plasma were performed. PTX-PDA-ALN-Nps were mixed (1:1, v/v) with 2×PBS (pH 7.4), 10% glucose, 1.8% NaCl and mice plasma (1:4, v/v), then incubated at 37°C. Particle size and distribution at particular times were measured and each sample was performed for three times.

Hemolytic assay
The fresh anticoagulated blood from the health mice eye canthus was centrifuged (5000 rpm, 15 min) to eliminate the fibrous proteins, then washed by normal saline twice and diluted with the ratio of 4% (v/v). Different concentrations of PTX-PDA-ALN-Nps (0.06-1 mg ml −1 ) adjusted to isotonic by NaCl were confused with 4% red blood cell suspensions (1:1, v/v) and the suspensions were incubated for 4 h at 37°C after that centrifuged for 5 min (5000 rpm). The supernatants were collected to measure the absorbance at 540 nm by ELISA plate reader (Biotek, Winooski, VT). The negative control and positive control were 0.9% NaCl and deionized water, respectively. The hemolysis percentage (%) was counted by the following equation (3): where, A sample is the absorbance of the PTX-PDA-ALN-Nps, A negative is the negative control's absorbance, and A positive is the positive control's absorbance. Each samples were analyzed in three times.

In vitro drug release kinetics
The in vitro release characteristic of PTX from PTX-Nps, PTX-PDA-Nps, PTX-PDA-ALN-Nps were investigated by using dialysis. Briefly, Nps were suspended in regenerated cellulose dialysis bag (MWCO, 8000-14 000, Sigma, USA) with 40 ml phosphate buffer saline (PBS, pH 7.4) and 0.2% Tween 80. The tube was moved into a water bath with the temperature of 37°C and shaken at 120 rpm. Then 1 ml sampled was removed at the indicated time, and centrifuged at 10 000 rpm for 10 min. The sampled supernatant was used for HPLC analysis and replaced with fresh release medium. Each batch of experiments was performed in three times.

In vitro cytotoxicity assay
The cytotoxicity of the PTX-PDA-ALN-Nps was calculated by the MTT assay. The GraphPad Prism, Version 5 (Inc., La Jolla, CA) was used to calculated the half maximal inhibitory concentration (IC 50 ).

In vivo distribution study
The real-time fluorescent imaging was adopted to evaluated the distribution of PTX in the Balb/c mice with K 7 M 2 wt tumors (volume about 100 mm 3 ). The mice were intravenously injected with 2 mg kg −1 PTX/DiR-Nps, PTX/DiR-PDA-Nps, PTX/DiR-PDA-ALN-Nps (PTX: Dir=40:1), respectively. Living Image Software 4.4 (PerkinElmer USA) was used to detected fluorescent images at 1, 4, 8, 12, 24 h. All the mice were used in quantitative analysis. The relative intensity was calculated according to the following equation (5) Mice were given a euthanasia and dissected on the 14th day of treatment, then the tumor were weighed. The volume of tumor was calculated by equation (6): The tumor inhibition rate (TIR) was calculated using the following equation (7): Comparison of liver and spleen index effect between Taxol and normal saline solution by the independent samples T test.
Other comparisons (cell inhibitory rate, antitumor rate) were based on a one-way analysis of variance and post hoc analysis (F-test). * p<0.05 was considered statistically significant.

Preparation of nanoparticles
The preparation technique of PTX-PDA-ALN-Nps is schematized in figure 1. The PTX-PDA-ALN-Nps were prepared in three steps. Firstly, PTX Nps were synthesized according to the procedure one (prepared at 2.3). Secondly, PTX-Nps were immersed in 1 ml Tris buffer (10 mM, pH 8.8) with 0.5 mg of dopamine to coat polydopamine on the Nps. Finally, ALN was added in the weak alkaline solution to conjugate on the surface of PTX-PDA-ALN. When dopamine hydrochloride was added, the suspensions began to turn dark, which indicated the dopamine had smoothly polymerized [34]. Then, the amine group of ALN covalently conjugated to PDA via Schiff base or Michael addition reaction. Based on the above three parts, we successfully conjugated ligand (ALN) to the PDAcoated PTX-Nps. The principle of PDA was first discovered in 2007 and published in Science, which could be used on various surface of inorganic and organic materials (noble metals, oxides, polymers, semiconductors, and ceramics) [35]. The catechol of dopamine was oxidized to quinone, and then reacts with other catechols and/or quinones to form polymerized dopamine (pD) [38]. While the PDA method has not been used in the modification of polymeric nano-carriers, until 2014. Joonyoung Park et al functionalized the surface of PLGA NPs with folate, cRGD, and a stealth polymer, and observed the influence of pH, initial DA concentration and incubation time on the film formation [33]. As the concentration of dopamine increases, the increase in particle size can be interpreted as an indication of micron pD aggregates. Increased size of PTX-PDA-ALN-Nps and PTX-PDA-Nps can be used as an indicator of quantitative density. The size of PDA coated nanoparticles were stable during 14 d at room temperature (figure S1 is available online at stacks.iop. org/NANO/30/255101/mmedia), which means the PDA film was stable. However, there is some evidence that the structures of PDA were stable at physiological pH, but unstable in acidic conditions [39,40], which is which is why some PDA-coated nanoparticles are pH sensitive.

Verified the conjugation of ALN on PTX-PDA-Nps
Size and surface properties of Nps induces significant differences in drugs cellular uptake, drug release, bio-distribution and in vivo pharmacokinetics. Small size Nps with relatively high surface area can pass through the leaky tumor micro vessels into tumor site, which is known as enhanced permeability and retention effect (EPR effect) [41,42]. As shown in table 1, The particle sizes were compared between different groups with the result that PTX-Nps Zeta potential is crucial for the stability of the Nps. We can reach a conclusion that the value of zeta potential of Nps were slightly increased after coating with dopamine. This result may be caused by PDA coating, which is a kind of negative polyelectrolytes.
The morphology of Nps can be investigated through the transmission electron microscope (TEM), and showed in figures 2(D)-(F). However, there was no noticeable difference nor visible surface feature attributable to PDA aggregates between these three groups of Nps under TEM. This may be due to the low magnification of transmission electron microscopy and the concentration of dopamine [33].
To verify the surface chemical group composition of Nps and the successful modification of PDA film, FT-IR spectroscopy was adopted in this study. There are some new absorption signals emerged after surface modification ( figure 3). All of the three experimental groups exhibited a strong peak at 3477 cm −1 , indicating the presence of hydroxyl groups stretching vibration, including surface adsorbed water. At the same time, due to the N-H and O-H stretching modes of ALN and PDA, we can notice that PTX-PDA-Nps and PTX-PDA-ALN-Nps were increased than PTX-Nps at 3477 cm −1 . The 1736 cm −1 band significantly decreased in PTX-PDA-Nps that proving the PDA-coating was conjugated on Nps [44]. The absorption peaks at 1654 and 1579 cm −1 were caused by the N-H bending vibrations and C=C resonance vibrations in the aromatic ring [41]. The increased or decreased of these characteristic absorption bands was all from PDA or ALN.
High-revolution XPS spectra were used for further prove the successful modification of PDA and conjugation of ALN  PTX-PDA-ALN-Nps were stable in isotonic glucose (5% Glucose), phosphate buffer saline (PBS, pH 7.4) and plasma after incubation at 37°C for 8 h. There were no visible particle size increase or aggregation among the three physiological media (P ˃ 0.05). Many enzymes or serum albumin in plasma  can be adsorbed on the surface of Nps, which in some cases leads to aggregation or clogging of blood [47]. Besides, PTX-PDA-ALN-Nps retained their size in plasma during 8 h, suggesting its excellent plasma stability ( figure S2). Furthermore, PTX-PDA-ALN-Nps revealed no hemolysis under 1 mg ml −1 , which proved that PTX-PDA-ALN-Nps were suitable for intravenous administration (figure S3).

PTX-Nps possessed sustained release property
In vitro drug accumulation release of the PTX-Nps, PTX-PDA-Nps and PTX-PDA-ALN-Nps during the 120 h were shown in figure 5. The drug release of PTX-PDA-ALN-Nps was from 1% to 69% within 120 h. The PTX-Nps showed steady, continuous release patterns, just like the drug release profiles of PTX-Nps, PTX-PDA-Nps and PTX-PDA-ALN-Nps, which indicating that PDA coating had no effect on the drug release. Perhaps PTX was excellently wrapped in Nps, which keeping a sustained release effect [44]. The release kinetic models (zeroorder release) and t1/2 (half the time of the release of the drug) of each Nps were calculated and shown below       for 24 h or 48 h and >14 mg ml −1 for 72 h). PTX-Nps and PTX-PDA-Nps also showed similar cytotoxicity under different concentrations, which further evidenced that PDA coating had no effect on cell viability, toxicity and biocompatibility [31].
3.6. Targeting PTX-Nps exhibited higher accumulation at the tumor site Non-invasive near-infrared optical imaging technology was adopted to evaluate the bio-distribution and tumor-targeting efficiency of the Nps. The Dir was entrapped in PTX-Nps, PTX-PDA-Nps and PTX-PDA-ALN-Nps before injecting to K 7 M 2 wt cell-bearing Balb/C mice. Figure  It may be owing to protein tyrosine phosphatases, a bis-phosphonate binding receptor, over expressed in osteosarcoma [10]. Furthermore, the absorption of macrophages of the reticuloendothelial system especially the Kupffer cells in the liver could also lead to the liver's high fluorescence density for all PTX formulations. ). This means that the Nps could reduce liver and spleen toxicity of Taxol. The antitumor efficacy experiment showed that PTX-PDA-ALN-Nps have a higher tumor inhibition rate and lower toxicity, could release PTX and maintain its bioactivity for osteosarcoma therapy [48].

Conclusions
In summary, PTX-Nps with PDA for surface modification, were prepared for targeted therapy osteosarcoma. Through the novel dopamine polymerization method, PTX-PDA-Nps were simply functionalized by alendronate. The size of PTX-PDA-ALN-Nps is about 290 nm with smooth surface. In addition, XPS and FT-IR was used to testify the successful incorporation of PDA and conjugation of ALN on PTX-Nps. PTX-PDA-ALN-Nps were stable in physiological media and showed no hemolysis. In vitro cytotoxicity experiment demonstrated the PTX-PDA-ALN-Nps remarkably inhibit cell proliferation with PTX-Nps and PTX-PDA-Nps comparing. In vivo distribution essay through living imaging investigations declared that PTX-PDA-ALN-Nps could be eligible for a potential drug delivery system for malignant osteosarcoma targeting treatment. Moreover, the antitumor efficacy experiment showed that PTX-PDA-ALN-Nps have a higher TIR. All experiments and data showed that PTX-PDA-ALN-Nps is a promising drug for targeting treatment of osteosarcoma in the future.