APTES monolayer coverage on self-assembled magnetic nanospheres for controlled release of anticancer drug Nintedanib

The use of an appropriate delivery system capable of protecting, translocating, and selectively releasing therapeutic moieties to desired sites can promote the efficacy of an active compound. In this work, we have developed a nanoformulation which preserves its magnetization to load a model anticancerous drug and to explore the controlled release of the drug in a cancerous environment. For the preparation of the nanoformulation, self-assembled magnetic nanospheres (MNS) made of superparamagnetic iron oxide nanoparticles were grafted with a monolayer of (3-aminopropyl)triethoxysilane (APTES). A direct functionalization strategy was used to avoid the loss of the MNS magnetization. The successful preparation of the nanoformulation was validated by structural, microstructural, and magnetic investigations. X-ray photoelectron spectroscopy (XPS) and Fourier transform infrared spectroscopy (FTIR) were used to establish the presence of APTES on the MNS surface. The amine content quantified by a ninhydrin assay revealed the monolayer coverage of APTES over MNS. The monolayer coverage of APTES reduced only negligibly the saturation magnetization from 77 emu/g (for MNS) to 74 emu/g (for MNS-APTES). Detailed investigations of the thermoremanent magnetization were carried out to assess the superparamagnetism in the MNS. To make the nanoformulation pH-responsive, the anticancerous drug Nintedanib (NTD) was conjugated with MNS-APTES through the acid liable imine bond. At pH 5.5, which mimics a cancerous environment, a controlled release of 85% in 48 h was observed. On the other hand, prolonged release of NTD was found at physiological conditions (i.e., pH 7.4). In vitro cytotoxicity study showed dose-dependent activity of MNS-APTES-NTD for human lung cancer cells L-132. About 75% reduction in cellular viability for a 100 μg/mL concentration of nanoformulation was observed. The nanoformulation designed using MNS and monolayer coverage of APTES has potential in cancer therapy as well as in other nanobiological applications.


Scientific Reports
| (2021) 11:5674 | https://doi.org/10.1038/s41598-021-84770-0 www.nature.com/scientificreports/ coloured precipitated solution, which was stirred at 50 °C for 30 min. Afterwards, the solution was transferred into a Teflon liner, fitted in a stainless-steel mould, and baked for 6 h at 200 °C in a furnace. The prepared MNS were removed from the Teflon liner and thoroughly rinsed with methanol and water in the presence of a permanent magnet and dried in a vacuum desiccator. For the functionalization of MNS by APTES, the MNS (1.0 gm) was dispersed in 100 mL methanol/toluene mixture 1:1 (V/V) by sonicating for 30 min. The resultant solution was heated at 95 °C until half of the solution was evaporated. The volume of the reaction mixture was adjusted to 100 mL by adding methanol. This procedure was repeated three times so that the solution becomes anhydrous. Then, 2% (V/V) APTES was added to the solution and kept in the shaking incubator for 24 h at 70 °C 49 . A washing step similar to that of MNS was applied to rinse the MNS-APTES.
Ninhydrin assay. Different concentrations of leucine (20 − 100 µg/mL) in 50% ethanol (1 mL) were pipetted into a series of eppendorf tubes containing 1 mL ninhydrin solution in 2% ethanol. The above mixture was sonicated and then heated in a water bath at 100 °C for 5 min and the colour change (formation of Ruhemann's purple) was measured at 570 nm. A suspension of MNS in 50% ethanol (4 mg/mL, stock) was prepared and used for the ninhydrin assay, as described above. The concentration of the amino groups (from APTES) on the MNS was estimated using the standard calibration curve for leucine. The assay was performed in three replicate experiments.
Drug loading and in vitro release. The NTD loading was performed by stirring 10 mg of MNS-APTES in 10 ml ethanol with different concentrations of NTD. Details regarding the optimization of drug loading are discussed in the supplementary information. In vitro drug release study of optimum NTD-conjugated MNS was carried out at 37 °C in a 75 mL dissolution medium (pH 5.5 and pH 7.4 phosphate buffers (PBS)) in a shaking incubator for 42 h.
Cytotoxicity assay. Early passage human lung cancer cell line L-132 was procured from National Centre for Cell Sciences (NCCS, Pune, India). The cytotoxicity assays were performed as per our previous report 50 and briefly discussed in the supplementary information.
Structural, elemental and magnetic characterizations. X-ray diffraction (XRD) measurements were performed on a PANalytical Empyrean diffractometer using Cu-K α radiation (λ = 1.54184 Å) in the 2θ range from 10° to 80° with a step size of 0.02°. XRD patterns were analyzed by the Rietveld refinement method using the FullProf.2 k program. During refinement, the shape of the peaks was assumed to be a pseudo-Voigt function with asymmetry. The backgrounds of the patterns are fitted to a fourth-degree polynomial function. The size and shape of MNS were evaluated by a JEOL field emission scanning electron microscope (FE-SEM) JSM-7600F. A CM 20 FEG transmission electron microscope (TEM) from Philips was employed to record micrographs and energy-dispersive X-ray spectroscopy (EDX) patterns. Fourier transform infrared (FTIR) spectra were recorded at room temperature in the range of 400-4000 cm −1 using a Shimadzu FTIR spectrophotometer. The magnetic properties of the MNS and MNS-APTES were investigated using superconducting quantum interference devicevibrating sample magnetometry (SQUID-VSM) from Quantum Design. The magnetic hysteresis loop, zero-field cooled (ZFC), and field cooled (FC) measurements were performed over a temperature range of 2-400 K and a magnetic field up to 60 kOe. The experimental procedure for thermoremanent magnetization (TRM) meas- www.nature.com/scientificreports/ urements consisted of cooling the sample down to 2 K in the presence of a 40 kOe magnetic field, followed by magnetization measurements at increasing temperatures. For each data point at a given temperature, the sample was magnetized in the magnetic field of 40 kOe for 5 min. Afterward, the field was switched off, and the sample was allowed to relax for 1 min, followed by a magnetization measurement. Thus, the recorded magnetization is the measure of remanent magnetization of all the MNS in the blocked state at a given temperature. The X-ray photoelectron spectroscopy (XPS) analysis was performed with an ESCALAB 250Xi photoelectron spectrometer from Thermo Scientific in an ultra-high vacuum (UHV) using a monochromatic Al-K α (1486.6 eV) X-ray source and a beam diameter of 300 µm. The hemispherical electron deflection analyzer operated in the constant analyzer energy (CAE) mode at pass energy of 200 eV for survey spectra and 20 eV for high-resolution spectra.
The binding energies of all spectra were referenced to the binding energy of C1s (284.4 eV).

Results and discussion
Size and phase analysis. The structural and morphological investigations of MNS were performed using XRD, FE-SEM, and HR-TEM. The XRD pattern with Rietveld refinement of an MNS is shown in Fig. 1a. According to the refinement data, the MNS exhibits pure magnetite (Fe 3 O 4 ) phase. The characteristic diffraction peaks for the lattice planes (220), (311), (400), (422), (511), and (440) of a cubic structure with space group Fd-3m were observed (JCPDS card No. 01-088-0866). The detailed parameters extracted after the refinement of diffraction spectra are tabulated in Table S1. The lattice parameter 8.3852 Å and the unit cell volume 589.58 Å 3 are comparable with the reference value for stoichiometric magnetite. The average crystallite size determined by the Debye-Scherrer formula for the most intense peak (311) was about 23 nm. Representative FE-SEM and TEM micrographs displaying the spherical shape of MNS are presented in Fig. 1b,c, respectively. It can be seen from these micrographs that the spheres consist of smaller particles. The size distribution histograms are drawn to estimate the average diameter of MNS (  www.nature.com/scientificreports/ its polynanocrystalline nature. All diffraction rings correspond to the magnetite phase, which corroborates well the XRD results. It has previously been suggested that the size and shape of nanostructures may determine the biodistribution of particles 51 . Decuzzi et al. 52 studied NPs in the range of 50 nm-10 μm and observed that particles having a size less than 500 nm support the Brownian motion and can have colloidal stability. Depending on application perspectives, different sizes and shapes of NPs are recommended. The particles having spherical morphology and critical size less than 200 nm can be used for TDD applications [53][54][55][56] . Monolayer coverage. The quantification of the amine groups existing on the MNS was evaluated by ninhydrin colourimetric assay. The nucleophilic displacement reaction among the primary amines and ninhydrin occurs, which includes the displacement of a hydroxyl group from ninhydrin by amine. It produces a colour complex known as Ruhemann's purple. The free − NH 2 groups estimated were found to be around 41.5 ± 3 µmol/g of MNS. The calculation of fractional monolayer coverage was performed as reported by Sun et al. 65 , assuming four aminopropylsilanes/nm 2 for monolayer coverage of APTES and using the surface area (6.44 m 2 /g) calculated from the mean MNS diameter (180 nm). The details regarding the calculation of surface area (m 2 /g) of MNS is provided in SI.
These results reveal monolayer coverage of APTES over MNS. It was reported that primary amine-modified monolayers on the NPs surface significantly improved the cellular delivery 66 . Besides, the monolayer coverage can overcome the problems associated with polymerization of the initial silane, possible formation of heterogeneous multilayers on the surface in the presence of excess silane and the lack of stability of the aminated surfaces 67 .
To get the information on the magnetic phases present in the samples by thermoremanent magnetization, several hysteresis loops were recorded at different temperatures. The normalized thermoremanent magnetization (TRM) values measured for MNS are shown in Fig. 5a, which helps to find the magnetic phase transition temperature for the smaller MNPs constituting the MNS. As anticipated for NPs, a gradual decrease in remanent magnetization was observed with increasing temperature for both MNS and MNS-APTES (not shown here). The thermoremanent magnetization curve reveals a subtle step at around 10-40 K, which is similar to the feature observed below 50 K in the field cooled curve (Fig. 4)). To highlight the changes in TRM variation, the first-order derivative of the TRM curve with respect to temperature is plotted, which reflects the distribution of www.nature.com/scientificreports/ the blocking temperature 78 . The derivative curve can be divided into three regions depending on the observed change in slopes, namely region-1 (T < 10 K), region-2 (10 K < T < 50 K), and region-3 (T > 50 K). Region-1 for the range T < 10 K corresponds to extremely fine MNPs with blocking temperatures below 10 K still being in the paramagnetic state. Region-2, between 10 and 50 K, represents particles with a moderate size, which constitute the majority of the sample. The third region T > 50 K is for the MNPs in the samples with blocking temperature considerably above room temperature. Figure 5b presents the magnetization (M S ) and coercive field (H C ) response of MNS and MNS-APTES. MNS has a higher saturation magnetization compared to MNS-APTES, which scales linearly with increasing temperature. The scaling of M S with temperature can be explained using Bloch's law 79 , which is related to low energy collective spin wave excitations (magnons).
Here, M S (T) and M S0 are the saturation magnetization at any given temperature (T) and absolute zero temperature (0 K), respectively. "β" is the Bloch constant, which depends on the Curie temperature (β ∝ 1/T C ). The values of M S0 of 90.2 emu/g and 89.7 emu/g were obtained for MNS and MNS-APTES, respectively.
In general, MNS exhibits hysteretic behaviour for the temperature below the blocking temperature and nonhysteretic behaviour above the blocking temperature. The coercive field (H C ) is observed to decrease monotonously with the square root of temperature and reaches zero at blocking temperature. The temperature dependence of H C can be expressed with Kneller's law 80,81 ,   www.nature.com/scientificreports/ Here, T B is the blocking temperature, and H C0 is the coercive field at 0 K. The value for H C0 was obtained by linear fitting the H C0 vs T 1/2 plot in the temperature range between 5 and 400 K. The values of H C0 obtained for MNS and MNS-APTES were 92.5 Oe and 211.6 Oe, respectively. Here it is worth mentioning that only a limited temperature range was chosen for the linear fitting, as the accuracy of H C is drastically lower above the blocking temperature. The estimated blocking temperatures for MNS and MNS-APTES were 126 K and 346 K, respectively. A probable reason for the higher value of T B estimated from the Eq. 5 is because the investigated samples are conglomerate of various sized interacting NPs. This assumption is further supported by the ZFC-FC (cf. Fig. 4b) and TRM results (cf. Fig. 5a). Nevertheless, the linear fit to H C vs. T 1/2 and a negligibly small value of H C below 200 K (14 K 1/2 ) indicates superparamagnetic response.
NTD loading and release. To achieve optimum drug loading on MNS, the ratio of MNS:NTD was varied.
The drug loading efficiency and loading capacity for different MNS:NTD ratios are summarized in Table S2 and shown in Fig. 6a. With a fixed MNS weight, initially, the NTD loading increased linearly with an increase in the amount of NTD and saturated for higher NTD amount. For MNS:NTD ratio (1:3), a maximum loading efficiency (79%) and loading capacity (23%) corresponding to greater conjugation of NTD was observed. This optimum nanoformulation (MNS:NTD, 1:3) was used for further studies.
The drug release from the MNS-APTES-NTD nanoformulation at 37 °C at pH 7.4 and 5.5 is shown in Fig. 6b. The pH values used to investigate the drug release mimics the physiological and the endosomal pH value of cancer cells. At physiological pH 7.4, the nanoformulation was quite stable, and only 28% cumulative drug release was observed after 48 h. The low drug release profile at neutral pH is observed due to the stable imine bond between MNS-APTES and NTD at physiological conditions. At acidic pH 5.5, the cumulative NTD release was about 50% within the first few hours, and release was up to 85% in 48 h. The mechanism of the controlled drug release can be attributed to the cleavage of the pH-responsive imine bond between NTD and functionalized MNS. Scheme 2 shows the possible mechanism of drug release in which imine groups get protonated under acidic conditions.
The functional polymers which exhibit ionizable groups (i.e., − NH 2 , − COOH, − SO 3 H, − PO 3 H 2 , − B(OH) 2 ) at support may get converted into charged moieties depending on the different pH conditions 82 . The linkages like imine bond are mostly stable under physiological conditions (pH ≈ 7.4), whereas they get hydrolyzed in acidic surroundings. The observed different pH response is ascribed to the protonation state of terminal primary amino groups under different pH conditions 83 . In vitro cytotoxicity study. The MTT assay was performed to examine the in vitro cytotoxicity of MNS-APTES, MNS-APTES-NTD, and free NTD. Figure 7 shows the cell viability of L-132 cancer cells after exposure to all the samples at various concentrations (20-100 μg/mL). Even at higher levels of MNS-APTES, the survival rate of L-132 cells does not grieve due to the biocompatibility of surface-modified MNS. The viability of L-132 cells decreased with an increase in concentrations of free NTD and MNS-APTES-NTD, indicating dose-dependent cytotoxicity of the formulations. The free NTD and MNS-APTES-NTD showed cellular toxicity of 57% and 47% at 40 μg/ml, respectively. For 100 μg/ml of free NTD and MNS-APTES-NTD, cellular death reached 85% and 75%, respectively. The difference in cell viability of MNS-APTES-NTD and free NTD can be attributed to the lower amount of NTD in the nanoformulation compared to free NTD. Similar cytotoxicity observed for NTD loaded nanoformulation, and free NTD suggests that NTD released from the nanoformulation has the same anticancer activity as free NTD.  Table S3, Fig S3 and Fig S4.

Conclusions
This work demonstrates the pH-responsive, controlled release of poorly water-soluble model drug NTD from the nanoformulation prepared by monolayer coverage of APTES over MNS. The use of a direct functionalization strategy and monolayer coverage of APTES prevents the loss in the magnetization of MNS reported for several functionalization approaches [30][31][32]42,84 . Thermoremanent magnetization studies reveal the superparamagnetism and qualitative distribution of magnetic properties in the MNS. Functionalization of APTES and loading of the anticancerous drug NTD is confirmed by FTIR. A drug loading capacity of 23.6% is obtained. Monolayer coverage of aminosilane is established by quantifying the amine groups with the ninhydrin assay. NTD is conjugated with MNS-APTES through the acid liable imine bond. At physiological conditions, the MNS-APTES-NTD nanoformulation maintains high stability and inhibits drug release. In contrast, the cleavage of imine bonds in an acidic environment would lead to the drug release on demand. The MNS-APTES-NTD nanoformulation exhibits dose-dependent cytotoxicity for the L-132 cell line. Further magnetic studies of the MNS with a different size distribution of the constituent particles are underway to investigate the interparticle interaction and to explore them for various applications. The higher magnetization of functionalized MNS owing to monolayer coverage can be useful in many nanobiotechnology applications such as magnetoresistive biosensors, nanobiocatalysis, magnetic cell separation, etc.