β-Cyclodextrin based magnetic nanoconjugates for targeted drug delivery in cancer therapy

Abstract

β-Cyclodextrin was conjugated with diisocyanate modified Fe₃O₄ magnetic nanoparticles through urethane linkages to obtain magnetic nanoconjugates. The nanoconjugates were characterized for thermal, magnetic and structural properties by using TGA & DSC, VSM and FTIR respectively. Scanning Electron Microscope-Energy Dispersive X-Ray Spectrometry (SEM-EDS) and XRD were used to study the morphology and changes in crystalline nature of magnetic nanoconjugates respectively. Dacarbazine, an anticancer drug, was considered as a model drug for loading and release studies. It was observed that by controlling the stoichiometry of urethane linkages, the surface accessible inclusion sites of cyclodextrin could be controlled thereby achieving command over entrapment efficiency and release behaviour of nanoconjugates. The nanoconjugates did not show significant cytotoxicity although dacarbazine loaded ones showed selective toxicity against chosen cancer cell lines with distorted cellular and nuclear morphology. The prepared nanocarriers thus can provide a strong platform for magnetic tumour targeting under guidance of magnetic field and through cyclodextrin-drug complexation.

---

Introduction

Magnetic nanoparticles (MNPs) such as iron oxide have found a broad range of applications in the field of medicine and biology such as controlled drug delivery, protein purification, specific cell targeting and medical imaging.^1–3 MNPs respond to external magnetic field for easy magnetic guided transport field^1,2 but carry no residual magnetization³ so that aggregation can be avoided to some extent. They offer exceptional combination of properties including adaptability for external manipulation,⁴ controlled size, possibility of surface modification and enhancement of contrast in magnetic resonance imaging (MRI).⁵ The agglomeration of MNPs can be further prevented by coating/covalent bonding with polymeric materials.⁶ Biomedical and bioengineering applications also require the coating of the MNPs with biocompatible materials that can act as a base for targeted drug delivery applications.

Cyclodextrins possess toroidal shape with hydrophilic exterior and lipophilic interior cavity. Due to this, they have ability to form inclusion complexes^7–9 with a wide range of drugs¹⁰ for enhanced bioavailability, solubility and stability of drug molecules. Cyclodextrins possess potential sites for chemical modification which makes them suitable as drug carriers.¹¹ β-Cyclodextrin (CD) contains 21 hydroxyl groups (7 primary and 14 secondary) which can be utilized for network formation. A system with combination of cyclodextrin and MNPs can give synergistic advantage of both enhanced bioavailability of drug and magnet responsive transport respectively. The design of such system was possible due to presence of hydroxyl groups on both the moieties, which can be linked to isocyanate molecules to form multiple urethane linkages. The added advantage of multiple urethane linkages is its biocompatibility, stability, strength and adaptability for modification which enhances applicability.¹² Copolymers of cyclodextrin have been reported to form host–guest complexes comparable to native cyclodextrins. The copolymer of epichlorohydrin with β-cyclodextrin has been recently explored for its intended application in drug delivery system¹³ and electrochemical probe in aqueous solution.¹⁴ Similarly hydroxypropyl cyclodextrin,^15,16 urethane networks based on β-CD¹⁷ and randomly methylated cyclodextrin have been reported for inclusion complex formation.¹⁸ The structure of copolymer can be systematically varied in order to tune the inclusion sites of cyclodextrin.¹⁰

Specifically, in case of urethane linkages, this task is more feasible as it can be achieved by variation of mole ratio of NCO/OH.^19,20 The cyclodextrin based urethane networks have been reported by reaction of cyclodextrin with diisocyanate for the intended applications like adsorbents, binder for active pharmaceutical ingredients and molecular imprinting,²¹ however its application as a controlled drug delivery system has been largely unexplored. Incorporation of CD in polymers^22–27 may lead to useful drug carriers containing high drug payload. There is dual advantage since both cyclodextrin cavity as well as urethane network can encapsulate drug molecules. We explored a strategy to conjugate MNPs covalently with β-cyclodextrin by multiple urethane linkages to prepare magnetic nanocarriers for anti-cancer drugs. For comparison, the conjugates were also prepared with PEG instead of β-cyclodextrin.

Dacarbazine, 5-(3,3-dimethyl-1-triazeno)-imidazole-4-carboxamide (Fig. 1) is a prodrug frequently used as cancer chemo therapeutic agent, particularly for the treatment of malignant melanoma.²⁸ The mechanism of action involves alkylation of nucleic acids through methylation leading to covalent linkages with sulfhydryl groups. This in turn inhibits DNA replication resulting in cell death.²⁹

Fig. 1 Structure of dacarbazine.

Magnetic drug targeting is one of the strategies to avoid side effects associated with chemotherapy.²⁸ Hence, we have attempted to investigate magnetic drug targeting by employing CD-based magnetic nanoconjugates using DCB as a model drug.

Results and discussion

Synthesis of nanoconjugates

The synthesis of magnetic nanoconjugates is illustrated in Scheme 1. Fe₃O₄ nanoparticles are functionalized with free hydroxyl groups on their surface^30,31 because of which they can react with an NCO group of diisocyanate like hexamethylene diisocyanate (HMDI). The free NCO end groups of diisocyanate functionalized MNPs are available for reaction with hydroxyls of β-cyclodextrin/PEG. In order to define effect of variation of surface accessible inclusion sites of cyclodextrin on drug release, nanoconjugates were synthesized with different molar ratios.

Scheme 1 Synthesis of magnetic nanoconjugates.

Higher β-cyclodextrin : HMDI mole ratio results into utilization of higher number of hydroxyl groups of β-cyclodextrin leading to formation of multiple urethane linkages with higher crosslink density. As per Mohamed et al.¹⁰ as the mole ratio of β-cyclodextrin : HMDI increases from 1 : 1 to 1 : 3, the NCO group of HMDI also gets covalently attached to the secondary hydroxyl groups of β-cyclodextrin, resulting into lesser number of accessible inclusion sites. This in turn is likely to affect the drug carrying capacity of the resulting polymer. Three CD-based nanoconjugates systems namely CD-11, CD-12 and CD-13 (with β-cyclodextrin : HMDI mole ratio of 1 : 1, 1 : 2 and 1 : 3 respectively) and one PEG-based nanoconjugates system namely PEG-12 (with PEG : HMDI mole ratio 1 : 2) were synthesized.

Characterization of nanoconjugates

Characteristic peaks obtained in IR spectra confirmed the structure of nanoconjugates. As shown in Fig. 2, the FT-IR spectrum of β-cyclodextrin was characterized by a single broad band at 3300–3500 cm⁻¹ corresponding to O–H stretching vibrations. Vibrations of the C–H and CH₂ groups appeared in the 2800–3000 cm⁻¹ region. IR of CD-nanoconjugates showed peaks at 553 cm⁻¹, 1623 cm⁻¹, and 3316 cm⁻¹, corresponding to Fe–O bonds in tetrahedral sites for Fe₃O₄-nanoparticles, N–H and C=O bond for urethane linkage respectively.³²

Fig. 2 FTIR spectra of β-cyclodextrin and magnetic nanoconjugates.

The NHCO stretching was also observed at 1560 cm⁻¹. The absorption bands at 2852 and 2863 cm⁻¹ were attributed to –CH₂– stretching vibrations. Similarly, structure of PEG-12 was verified by peaks obtained at 580 cm⁻¹, 1743 cm⁻¹ and 3312 cm⁻¹ corresponding to Fe–O bonds in tetrahedral sites for Fe₃O₄ nanoparticles, N–H and C=O bond for urethane linkage respectively. Peaks obtained at 1576 cm⁻¹ and 2863 cm⁻¹ were assigned to NHCO stretching and –CH₂– stretching vibrations.

The nanoconjugates were analyzed for thermal properties using TGA and DSC. As shown in Fig. 3(a), the TGA plots of CD-nanoconjugates suggests that it exhibits three-stage degradation pattern over the range of 30–600 °C, which is in accordance with the previous report.³² The first stage of weight loss below approximately 200 °C can be attributed to the loss of surface adsorbed water as well as dehydration of surface OH groups. The second stage of weight loss of around 10–50% was observed between temperature range of 276 to 344 °C, which can be due to cleavage of urethane linkage. This temperature range for urethane bond degradation is quite comparable to other polyurethanes with carbohydrate crosslinkers like glucose, starch, cellulose that generally start degrading at temperatures as low as 200 °C.^20,33 The last stage of degradation in the region of 344–470 °C, probably, is due to thermal decomposition of cyclodextrin.³² It is apparent from the TGA data that the PEG-nanoconjugates were observed to be thermally more stable. PEG is high molecular weight polyol, which apart from forming urethane linkage also coats the surface of MNPs. This results in enhancement of thermal properties. Amongst CD-nanoconjugates, it was observed that the degradation follows entirely different pattern during initial and later degradation stages. Before 5 wt% degradation, the degradation temperature is observed to be highest for CD-11 (Table 1), and decreases with increasing β-cyclodextrin : HMDI mole ratio. As the initial degradation is attributed to loss of both the surface adsorbed water and dehydration of surface OH groups, content of both surface adsorbed water and surface OH groups shall be in the order CD-11 > CD-12 > CD-13. The higher content of surface OH groups in case of CD-11 as compared to CD-12 and CD-13 is apparent since increasing β-cyclodextrin : HMDI mole ratio leads to increased utilization of hydroxyl groups of β-cyclodextrin due to higher content of isocyanate groups. The second stage of degradation is attributed to degradation of urethane linkage. DSC thermograms for both heating [Fig. 3b(A)] and cooling [Fig. 3b(B)] cycles of the synthesized copolymers are shown in Fig. 3(b). The urethane networks are characterized with phase segregation due to presence of dual phases namely soft and hard segments.³⁴ Due to this special characteristic, the urethane networks may show two distinct glass transition temperatures corresponding to hard and soft segment namely T_gHS and T_gSS respectively.³⁵ In addition, the glass transition temperature of urethane networks depends strongly on the molecular mobility and the urethane networks with higher crosslinked structure are characterized with higher values of glass transition temperatures.³⁵

Fig. 3 (a) Thermal degradation plots of magnetic nanoconjugates. (b) DSC thermograms of copolymers.

Table 1 Thermal degradation temperature of nanoconjugates

Wt loss (%)	Degradation temperature, °C			PEG-12
	CD-11	CD-12	CD-13
1	033.26	032.89	030.95	073.15
2	063.87	050.28	041.24	204.84
5	230.98	222.71	186.22	288.72
10	275.98	289.22	284.78	321.66
50	328.85	349.92	350.91	390.21
80	496.69	497.35	453.05	410.12

---

The values of T_gHS and T_gSS obtained for PEG-copolymer and CD-copolymer is shown in Table 2. It can be noted that the values of T_g for PEG-copolymer is lesser than that for CD-copolymer, which can be attributed to highly crosslinked structure of the later. The soft segment involved in PEG-copolymer is polyethylene glycol with functionality,² whereas for CD-copolymer its β-cyclodextrin with higher functionality due to presence of multiple hydroxyl groups. The higher functionality of β-cyclodextrin might be responsible for reaction with higher number of NCO groups of HMDI leading to high crosslink density.

Table 2 Glass transition temperature of copolymers

Polymer	TgSS	TgHS
CD-copolymer	−11.6	81
PEG-copolymer	−32.8	67.2

---

The morphology of the magnetic nanoconjugates is shown in Fig. 4(a). The pure MNPs were observed as small aggregated particles due to their nano size. It was observed that with conjugation of multiple urethane linkages on the surface of MNPs, there was increase in size as well as dispersion which might be, as reported by Akbarzadeh et al., due to the electrostatic repulsion force and steric hindrance between the urethane chains on the surface of MNPs.³⁶ Also at magnification of 100×, the pure MNPs were observed as agglomerated mass while CD-nanoconjugates were noticed to be discrete particles with higher dimension and without agglomeration.

Fig. 4 (a) SEM micrographs of pure MNPs and CD-nanoconjugates. (b) EDS patterns of pure MNPs and CD-nanoconjugates.

At higher magnification, the particles were observed to be easily distinguishable from one-another and having sharp features in case of pure MNPs; but for CD-11 the image was observed to be blurred and particles were observed to have blunt surface and homogeneity in nature. With increase in HMDI content for CD-12, the layered bruises were observed on the surface of nanoconjugates that might be assigned to urethane bonds²⁰ conjugated with nanoparticles. The bruises became more apparent for CD-13 with severely blunt surface suggesting the conjugation of highly crosslinked urethane networks on the surface of nanoparticles.

The EDS spectra showed the elemental fingerprints of nanoconjugates (Fig. 4(b)). The EDS spectra of pure MNPs showed strong peaks of Fe and O with weight percent of 71.33% and 27.95% respectively (Table 3). However, the CD-nanoconjugates showed presence of elemental carbon and higher weight percent of Fe and O as compare to pure MNPs. This result again confirms the successful conjugation of urethane network on the surface of MNPs. Also, there was increase in the percentage of C and O with increase in the β-cyclodextrin : HMDI mole ratio, which indicates that with increase in mole ratio, higher content of urethane bonds was conjugated with MNPs. This result supports the hypothesis that there was increase in crosslinking of urethane network with increase in β-cyclodextrin : HMDI mole ratio.

Table 3 Composition of components of pure MNPs and CD-nanoconjugates by EDS

Element	Weight%
	Pure MNPs	CD-11	CD-12	CD-13
C K	00.00	63.05	66.19	67.48
O K	27.95	25.96	29.45	31.48
Fe K	71.33	10.99	04.36	01.03

---

Fig. 5 shows the TEM images and size distribution of pure Fe₃O₄ and CD-13 nanoconjugates. In confirmation with SEM analysis, the pure MNPs were observed to be aggregates of small nanoparticles. On the other hand, CD-13 nanoconjugates were visualized as roughly spherical in shape and discrete. The nanoconjugates were mono dispersed with narrow size distribution. The difference in dimension of pure MNPs (20–25 nm) and CD-nanoconjugates (55–65 nm) suggests that the thickness of layer of urethane network might be approximately 35–40 nm. This, once again, suggests successful conjugation of urethane network on the surface of nanoparticles.

Fig. 5 TEM images of pure MNPs and CD-13 nanoconjugates.

The hysteresis loops of nanoconjugates before and after drug loading are shown in Fig. 6(a–c). The coercivity, retentivity and magnetic saturation values are depicted in Table 4. It is observed that the magnetic saturation value of the CD-nanoconjugates was higher than PEG-nanoconjugates, which signifies comparatively higher magnetization of CD-nanoconjugates. As discussed earlier, there is a possibility that PEG may coat the surface of MNPs, which also results in reduction of the magnetic properties of PEG-nanoconjugates. Amongst β-cyclodextrin based nanoconjugates, with increase in β-cyclodextrin : HMDI mole ratio, the magnetization decreases.

Fig. 6 (a) Room temperature magnetization curves of pure MNPs and nanoconjugates (before drug loading). (b) Room temperature magnetization curves of CD-nanoconjugates. (c) Room temperature magnetization curves of nanoconjugates after drug loading. (d) Stability and response to magnet for nanoconjugates and (e) visual appearance of pure MNPs and of CD-nanoconjugates.

Table 4 Retentivity and magnetization values of CD-nanoconjugates

Sample	Magnetization (emu g^−1)	Coercivity (G)	Retentivity (emu)
Pure MNPs	46.5681	076.12	0.1092
CD-11	11.0417	336.92	0.1380
CD-12	05.0618	346.67	0.0694
CD-13	03.2932	345.99	0.0476
PEG-12	04.6563	098.94	0.0196
Drug loaded CD-12	04.7869	111.49	0.0182
Drug loaded PEG-12	00.7241	108.07	0.0035

---

This could be attributed to higher load of urethane matrix on the MNPs with higher β-cyclodextrin : HMDI mole ratio. The magnetization value of drug loaded nanoconjugates was reduced due to presence of drug molecules on the surface, which also signifies successful loading of drug. Magnetization curves reveal the ferromagnetic behaviour of the nanoconjugates. All the nanoconjugates could be easily dispersed in distilled water and could be drawn from the solution to the wall of the vial by application of external magnet as shown in Fig. 6(d). Thus, it is clear that even after conjugation with CD by urethane linkages, the magnetic behaviour of the MNPs is preserved. As a result, the nanoconjugates have excellent magnetic response as desired for intended application in the field of magnetic tumour targeting. Also, as shown in Fig. 6(e), the visual change in colour of nanoconjugates can be seen with the change in β-cyclodextrin : HMDI mole ratio. The XRD patterns of pure MNPs and CD-nanoconjugates are shown in Fig. 7. The XRD pattern of pure MNPs exhibited the peaks at 2θ = 30.34°, 35.41°, 56.72° and 62.73°, which correspond to (220), (311), (511) and (440) planes of Fe₃O₄, respectively. This could be indexed to its inverse cubic spinal structure.³⁷ The presence of all the peaks corresponding to Fe₃O₄ in CD-nanoconjugates reveals that the conjugation of urethane network on its surface did not lead to change their crystal phase structure.³⁶ Such types of results have been reported for Fe₃O₄ bonded with hyperbranched polyurethanes and multiwalled carbon nanotubes.³⁸

Fig. 7 XRD patterns of pure Fe₃O₄ MNPs and CD-nanoconjugates.

The XRD diffractograms CD-nanoconjugates exhibited peaks of Fe₃O₄ along with the peak at 23.65°, 23.48° and 23.53° with relative intensities of 2.98%, 28.37% and 29.46% for CD-11, CD-12 and CD-13 respectively; which corresponds to cyclodextrin–urethane network.³² The increase in the relative intensity of this peak with increase in β-cyclodextrin : HMDI mole ratio is attributed to the fact that the increased crosslinking of urethane linkages resulted into higher hard segment content, resulting in the enhancement of crystallinity of the CD-nanoconjugates. The results obtained in XRD analysis thus confirm successful conjugation of CD on MNPs by multiple urethane linkages. It also signifies the changes in their crosslinking behaviour with varying β-cyclodextrin : HMDI mole ratios. The urethane copolymers are reported to have direct correlation between swelling and crosslink density. The lesser the crosslinking, the higher is the flexibility and swelling.³⁹ As shown in Table 5, the equilibrium swelling of PEG-nanoconjugates was observed to be three fold higher than that of CD-nanoconjugates. This is due to the structural flexibility and reduced crosslinking of PEG-nanoconjugates compared to CD-nanoconjugates, which is again in accordance with thermal analysis.

Table 5 Swelling coefficient (Q) and entrapment efficiency (% LE) for nanoconjugates

Sample	% LE	Q
PEG-12	05.589	8.0752
CD-11	14.028	2.6958
CD-12	11.361	2.4589
CD-13	08.021	2.2589

---

Drug loading and release studies

The nanoconjugates consist of crosslinked network made from hexyl moieties of HMDI. These hexyl moieties lead to the formation of lipophilic spaces in the network, which can accommodate the drug molecules. This is one of the reason because of which the nanoconjugates can serve as drug carriers. As shown in Table 5, the drug loading capacity of CD-nanoconjugates was greater than PEG-nanoconjugates. In the present case, β-cyclodextrin was expected to offer dual advantages for drug delivery. Firstly, being multihydroxyl compound it can lead to formation of crosslinked urethane network. Secondly, it can form inclusion complex with drug. Because of both these factors there is an increased drug loading. This kind of behaviour is very common for β-cyclodextrin functionalized copolymers as reported for inclusion of various drugs by β-cyclodextrin derivatives.^15,40–44 The drug release study was carried out in phosphate buffer saline at pH 7.4 at 37 °C under dark conditions. All the studies were conducted thrice and the mean values were used to construct plot for drug release studies. Fig. 8 shows the release profiles of DCB-loaded PEG-nanoconjugates and CD-nanoconjugates as a function of time. It was observed that PEG-nanoconjugates showed higher rate of drug release as compared to CD-nanoconjugates. The drug was released within a shorter period with a burst type of release (initial rapid release of 45% of drug release during first 40 min) in case of PEG-nanoconjugates. This is attributed to higher swelling coefficient of PEG-nanoconjugates as compared to CD-nanoconjugates. The drug release in case of PEG-nanoconjugates thus follows diffusion mechanism as observed for other reported urethane copolymers.^19,20 However, in case of CD-nanoconjugates, the drug release behaviour was entirely different as it was observed to be slower and sustained. For CD-nanoconjugates, the drug release profile was characterized by an initial rapid release of 5–15% of drug release during first 20–25 min followed by comparatively slower rate of release of remaining drug. This observation suggests that a part of the total drug incorporated in the nanoconjugates is quite labile and the rest portion is apparently stable. In case of CD-nanoconjugates, in addition to diffusion mechanism, the role of cyclodextrin as a host to form drug-cyclodextrin inclusion complex also plays an important role in release of drug. The labile portion is thus hypothesized to be released due to the swelling effect of the urethane network while later sustained and slower release is attributed to dissociation of drug-cyclodextrin complex. The hydrophobic part of DCB interacts with hydrophobic cavity of β-cyclodextrin⁴⁵ which may lead to higher stabilization of drug-cyclodextrin complex contributing in slowing the release down.

Fig. 8 Drug release profiles of magnetic nanoconjugates.

To further study the role of cyclodextrin in drug release properties, we carried out drug loading and release study for CD-11, CD-12 and CD-13. As shown in Table 5, the drug loading capacity of nanoconjugates decreased from 14.028% to 8.021% as mole ratio of β-cyclodextrin : HMDI increased from 1 : 1 to 1 : 3. This can be explained based on our hypothesis that by increasing the β-cyclodextrin : HMDI molar ratio, the number of accessible inclusion sites of CD decrease. This leads to a possibility of higher number of urethane chains surrounding the CD cavity.

Dissociation of inclusion complex is usually a process driven by large increase in the number of water molecules in the surrounding environment that can again displace the guest molecules from the cavity.⁴⁶ During the course of 430 min, the total percentage of cumulative release observed was 96%, 72% and 54% for CD-11, CD-12 and CD-13 respectively. The reduction in release rate is due to enhanced capacity of cyclodextrin to provide better hosting to encapsulated drug in case of CD-11. As the moles of HMDI increase, more urethane chains surround the CD-cavity resulting in much slower release rate.

It is noteworthy that although PEG-nanoconjugates showed higher rate of drug release, the pattern of such release rate is not favourable for controlled drug delivery applications. This is because the burst release is unpredictable, and even in the cases where burst release is desired, the amount of burst cannot be significantly controlled. Also there is danger of high concentration of drug release in short time in the amounts of the drug which can be beyond the toxic level in vivo.⁴⁷

The drug released during the burst might also be metabolized and excreted without being effectively utilized. Such wastage of drug may incur drug dosage loss and economic loss.⁴⁸ All of these disadvantages could be overcome by utilizing β-cyclodextrin in the synthesis of nanoconjugates, which offers well-controlled rate of drug release as described above.

A schematic diagram of drug loading and release for CD-nanoconjugates and PEG-nanoconjugates is shown in Scheme 2. The additional benefit of combination of urethane linkages with β-cyclodextrin is that by controlling the stoichiometry of urethane network formation, the number of hydroxyl groups on the cyclodextrin molecule could be controlled. This could give control over surface accessible inclusion sites, giving command over drug release rate.

Scheme 2 Schematic of drug loading and release from magnetic nanoconjugates.

Cytotoxicity and haemolysis assay

The results obtained in the cytotoxicity (MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide)) assay of nanoconjugates against normal lung cell line (L-132) is shown in Fig. 9(a). Also, cell viability of different concentrations of DCB loaded and pure nanoconjugates against four cell lines of prevalent types of cancer viz. lung carcinoma, breast carcinoma, cervical carcinoma and hepatoma (A549, MCF-7, HeLa and HepG2 cells respectively) are shown in Fig. 9(b). Determination of cell viability using MTT (a tetrazolium dye) assay is a popular experimental protocol for evaluating the cytotoxic potential of a test compounds wherein, MTT undergoes reduction by the mitochondrial enzymes to form a blue colored formazan.⁴⁹ Colorimetric estimation of the intensity is directly proportional to cell viability.⁵⁰ In the present study, the nanoconjugates showed >90% viability against normal lung cell line, indicating its non-toxic nature (Fig. 9(a)). Besides this, pure MNPs were significantly toxic against all the four cancer cell lines. However, the nanoconjugates (CD-11/12/13) were found to show mild toxicity at the said dose.

Fig. 9 (a) Cell viability of nanoconjugates in normal lung cell lines. (b) Effect of MNPs, nanoconjugates and DCB loaded nanoconjugates on lung cancer (A549), cervical cancer (HeLa), hepatoma (HepG2) and breast cancer (MCF 7) cell lines. Data shows percentage cell viability and the results are expressed as mean ± SEM for n = 3 (replicates) where, ns = non significant, *p < 0.05, **p < 0.01 and ***p < 0.001 compared to untreated (control) cells.

A scrutiny of the results revealed that CD-11 showed mild toxicity against MCF-7 and HepG2 cells whereas; CD-12 was moderately toxic against HepG2 cells only. A comparison of cytotoxicity of DCB drug loaded nanoconjugates and naïve DCB drug revealed that all the three varieties of DCB drug loaded nanoconjugates showed significant toxicity against A549 cells. However, cytotoxic responses of CD-11, 12 and 13 were more specific against MCF-7, HepG2 and HeLa respectively (Fig. 9(b)). Cancer cells are metabolic and porous in nature and are known to internalize solutes rapidly compared to normal cells. We hypothesized that the differential response observed in DCB loaded nanoconjugates can be attributable to two processes viz. rate of drug release and differential metabolic rates of cancer cells studied herein. In our previous study, we had reported that neither size nor zeta potential alone determines the optimal cell response induced by test nanoconjugates.⁵⁰ Also, the interactions of test drug and nanoconjugates with components of cellular growth media used for in vitro studies leading to alterations in their physiochemical properties is not ruled out and warrants further studies.

Haemolysis testing is a type of acute toxicity screening assay mainly used to evaluate the haemocompatibility of the materials when in contact with blood.⁵¹ This method is rapid test that provides a direct quantitative estimate of the sensitivity of nanoconjugates on the erythrocyte membrane. During haemolysis, rupture of erythrocyte membrane causes release of cellular contents including haemoglobin. Free haemoglobin released by materials on haemolysis in vitro, is read by spectrophotometer and extent of haemolysis can be calculated.⁵² As shown in Fig. 10, the degree of haemolysis was about 4–7% in DCB loaded and pure nanoconjugates, showing its haemocompatibility.

Fig. 10 Haemolysis of pure MNPs, nanoconjugates and drug loaded nanoconjugates.

Cell and nuclear morphology

When in contact with an anti-cancer compound, cancer cells undergo nuclear shrinkage, chromatin condensation, nuclear fragmentation and formation of apoptotic bodies.⁵³ DAPI (4′,6-diamidino-2-phenylindole) staining is an effective method to qualitatively assess nuclear condensation.⁵⁴ In our study, the cells were stained with DAPI and the subsequent nuclear condensation/fragmentation was studied and photographed on Floyid Cell imaging station (40×). All the types of cancer cells studied herein (A549, MCF-7, HeLa and HepG2) showed cell shrinkage and distorted cell morphology following 100 μ g mL⁻¹ dose of DCB loaded CD-11/12/13 as compared to the ones treated with pure nanoconjugates or the absolute control cells. For instance, the results obtained for MCF-7 cell line are shown in Fig. 11. DAPI stained cells assessed herein, for alterations in nuclear morphology showed prominent nuclear condensation and fragmentation in the above-mentioned groups as compared to nanoconjugates or the absolute control cells. These results are in agreement with the observed cytotoxicity in A549, MCF-7, HeLa and HepG2 cancer cells.

Fig. 11 Apoptotic cytotoxicity of MCF-7 and A549 cells on control, pure MNPs and drug loaded nanoconjugates by DAPI staining. The staining indicates nuclear fragmentation.

Experimental

Materials

β-Cyclodextrin (≥97% purity), polyethylene glycol (PEG 6000), Fe₃O₄ nanoparticles and hexamethylene diisocyanate (HMDI) (≥98% purity) were purchased from Sigma Aldrich, India. N,N-Dimethylformamide (DMF) was purchased from Qualigens, Bombay, India. Dacarbazine (≥98% purity) was purchased from Sigma Aldrich, India and used as received. Phosphate buffer saline tablets (for preparation of pH 7.4 buffer solution) were obtained from Sigma Aldrich, India.

Synthesis of magnetic nanoconjugates

The previously reported procedure³² was adopted for covalent conjugation of MNPs with HMDI. Briefly, MNPs (0.4 g) were dispersed in DMF (60 mL) by sonication. HMDI (1.767 g in 10 mL DMF) was added dropwise during the course of 15 min. The reaction mixture was sonicated for next 3 h at 25 °C. Next, β-cyclodextrin (2 g, dissolved in 30 mL DMF) was added dropwise during the course of 30 min. The reaction was allowed to proceed at 70 °C for next 3 h. The product was separated by magnetic decantation followed by washings of distilled water (25 mL) and acetone (25 mL). The CD-nanoconjugates were dried in vacuum for 24 h at 40 °C. Using similar procedure, we have synthesized three nanocarriers with the mole ratio of cyclodextrin : HMDI of 1 : 1, 1 : 2 and 1 : 3 namely CD-11, CD-12, and CD-13 respectively (Table 6).

Table 6 Molar composition of nanoconjugates

Code	Equivalent mole ratio
	CD : HMDI	MNPs : HMDI
CD-11	1 : 1	1 : 5
CD-12	1 : 2	1 : 5
CD-13	1 : 3	1 : 5

---

Code	Equivalent mole ratio
	CD : PEG	MNPS : HMDI
PEG-12	1 : 2	1 : 5

---

The mole ratio of MNPs : HMDI was kept 1 : 5 in all the nanoconjugates. Similar procedure was adopted for synthesis of PEG-nanoconjugates with PEG : HMDI mole ratio 1 : 2. Copolymer of HMDI and cyclodextrin with cyclodextrin : HMDI mole ratio 1 : 2 and copolymer of HMDI and PEG with PEG : HMDI mole ratio 1 : 2 were prepared by same method as described above, in order to carry out DSC analysis.

Characterization methods

The FTIR spectra were recorded as KBr discs on a PerkinElmer IR spectrophotometer at room temperature. Thermo gravimetric analysis (TGA) was carried out by using TG-DTA 6300 INCARP EXSTAR 6000 in the temperature range of 30–600 °C and heating rate of 10 °C min⁻¹. The nitrogen atmosphere was maintained throughout the measurement. Differential scanning calorimetry (DSC) thermograms were recorded on a NETZSCH DSC at a rate of 10 °C min⁻¹ under nitrogen with 30–40 mL min⁻¹ gas flow rate and temperature range of −100 to 200 °C. The DSC analysis was carried out under both cooling and heating cycles. Vibrating sample magnetometer (VSM) analysis was carried out by using Lakeshore VSM 7410 at room temperature. Jeol (Jem-2100) electron microscope was used at an acceleration voltage of 200 kV in order to carry out High-Resolution Transmission Electron Microscopy (HR-TEM) analysis. Energy dispersive X-ray (EDX) analysis of the vacuum dried CD-nanoconjugates was recorded by the model-JSM-5610 LV attached to scanning electron microscopy (SEM). X-ray diffraction (XRD) was carried out by using PANalytical ‘X’PERT-PRO XRPD of Cu Kα radiation (λ = 0.15406 nm) with a scanning rate of 2° min⁻¹ and 2θ ranging from 0 to 100°.

Preparation of DCB loaded magnetic nanoconjugates

The loading of DCB on magnetic nanoconjugates was investigated through batch technique.⁵⁵ The DCB solution (30 mg mL⁻¹) was prepared by dissolving 150 mg DCB in 5 mL ethanol followed by addition of 45 mL water. For loading of drug on the nanocarriers, 15 mg of nanoconjugates were suspended in a vial containing 5 mL of 30 mg mL⁻¹ DCB solution. The mixture thus obtained was sonicated overnight at room temperature under dark conditions. The drug-loaded nanocarriers were separated from the free drug by magnetic decantation and washed twice with ethanol. The drug-loaded nanocarriers were allowed to dry at room temperature for 24 h and then stored in desiccators. The concentration of drug in supernatant was determined by using calibration plot constructed on UV-spectrophotometer for DCB at 360 nm. The drug entrapment efficiency (% LE) of CD-nanoconjugates was estimated by using following equation.⁵⁶ Where I_c denotes initial content of DCB added and S_c represents DCB concentration in supernatant. Care was taken to maintain dark conditions throughout drug loading and release studies as DCB is reported to degrade in light.⁵⁷

Swelling studies of magnetic nanoconjugates

In order to study the swelling properties of magnetic nanoconjugates, the known weight of PEG-nanoconjugates and CD-nanoconjugates were kept in phosphate buffer saline (0.1 M, pH 7) at 37 °C for 24 h. The swelling was measured by equilibrium weight gain method.⁵⁸ Following equation was used to carry out calculations for determination of equilibrium swelling (Q), where W₁ and W₂ denote the weight of nanoconjugates before and after equilibrium swelling respectively. The swelling experiments were performed in triplicate and the mean value of Q was tabulated.

Cell culture and cell viability assay

The cell lines viz. L-132 – normal lung cell line, A549 – human lung carcinoma, breast cancer – MCF7, cervix adenocarcinoma – HeLa and human hepatoma-HepG2 cells were procured from National Centre for Cell Sciences, Pune, India. Cells were incubated at 37 °C in a water-jacketed CO₂ incubator (Thermo scientific, forma series II 3111, USA). Cells were seeded (1 × 105 cells) in a T25 flask and cultured in DMEM containing 10% FBS and 1% antibiotic–antimycotic solution with trypsinization at every third day and sub-culturing with a TPVG solution. All the solutions used were filtered through a 0.22 μ filter (Millipore Biomedical Aids Pvt. Ltd, Pune) prior to their use.

Cell viability was evaluated by MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide, Sigma-Aldrich) biochemical assay as per a reported procedure.⁵⁹ Three replicates were performed for each set and the mean values are reported. Cells (7 × 103 cells per well) were seeded in 96-well culture plates for 24 h and then treated with a test compounds in dose range of 10, 50, 100, 250, 500 and 1000 μg mL⁻¹ for 24 h. Later on, 10 μL of MTT (5 mg mL⁻¹) was added and the incubation was done for 4 h at 37 °C. The contents were discarded and wells were washed with Phosphate Buffered Saline (PBS) and the resultant formazan was dissolved in 150 μL of DMSO and absorbance was read at 540 nm in ELX800 Universal Microplate Reader (Bio-Tek Instruments, Inc., Winooski, VT) and the percentage cell viability was calculated as per reported method.⁶⁰

Cell and nuclear morphology

Cells were stained with DAPI as per standardized protocol.⁶¹ Possible distortions in gross cellular and nuclear morphology were observed and photographed on Floyid cell imaging station (400×).

Haemolysis assay

A written consent was obtained from a healthy human volunteer and whole blood was collected for haemolytic assay. A prior approval of the protocol was obtained by the ethical committee and all experiments were performed as per Indian medical association for research on human subjects at Blue cross pathology lab (IMA-BMWMC no. 1093), Vadodara, India. All the procedures were in compliance with laws and guidelines of Indian Medical Association (IMA).

The blood samples were collected in ethylenediaminetetra-acetic acid (EDTA) coated vacutainer tubes and treated with said concentrations of the compounds. Untreated sample served as the negative control (with 0% haemolysis) whereas, sample treated with 3% hydrogen peroxide (with 100% haemolysis) served as positive control. After incubation for 3 h (an adjustment of the standard ASTM F-75617) the tubes containing blood samples were centrifuged at 1500 rpm for 10 min to collect the plasma. The supernatant was detected for presence of the haemoglobin at 540 nm and percentage haemolysis calculated according to the procedure described elsewhere.⁶²

Statistical analysis

Data were analysed for statistical significance using one way analysis of variance (ANOVA) followed by Dunnett's multiple comparison test and the results were expressed as mean ± SEM using Graph Pad Prism version 6.0 for Windows, Graph Pad Software, San Diego, California, USA.

Conclusions

MNPs were conjugated with β-cyclodextrin/PEG by urethane linkage to prepare magnetic nanoconjugates. The PEG-nanoconjugates were observed to possess burst release profile, whereas CD-nanoconjugates showed controlled drug release profile. In addition to this, by controlling the molar ratio of β-cyclodextrin to HMDI, it is possible to tune the drug loading and the drug release capacity of β-cyclodextrin. This aided in achieving the control over entrapment efficiency and release behaviour of DCB from nanoconjugates. The DCB-loaded nanoconjugates showed specific toxicity against some types of cancer cells studied herein and thus hold merit for targeted drug delivery. The same will presumably reduce the serious side effects encountered during chemotherapy. The approach of utilizing synergistic advantage of urethane network, β-cyclodextrin and magnetic nanoparticles invites attractive application of prepared nanocarriers in the field of cancer therapy.

Acknowledgements

Authors are thankful to GNFC Ltd, Bharuch, Gujarat, for providing infrastructure facilities and financial assistance. We acknowledge SAIF, Chennai for providing DSC and VSM facilities. We are obliged to Dr Vandana Rao & Dr Bharti Rehani form department of metallurgical and materials engineering for SEM-EDS and XRD analysis respectively. DST funding for conducting biological studies is duly acknowledged.

References

1. J. Dobson, Drug Dev. Res., 2006, 67, 55–60.
2. B. Chertok, A. E. David and V. C. Yang, J. Controlled Release, 2011, 155, 393–399.
3. T. Neuberger, B. Schopf, H. Hofmann, M. Hofmann and B. Vonrechenberg, J. Magn. Magn. Mater., 2005, 293, 483–496.
4. A. J. Cole, A. E. David, J. Wang, C. J. Galbán, H. L. Hill and V. C. Yang, Biomaterials, 2011, 32, 2183–2193.
5. Y. Liu, K. Yang, L. Cheng, J. Zhu, X. Ma, H. Xu, Y. Li, L. Guo, H. Gu and Z. Liu, Nanomedicine, 2013, 9, 1077–1088.
6. A. Z. M. Badruddoza, M. T. Rahman, S. Ghosh, M. Z. Hossain, J. Shi, K. Hidajat and M. S. Uddin, Carbohydr. Polym., 2013, 95, 449–457.
7. T. Loftsson and M. E. Brewster, J. Pharm. Sci., 1996, 85, 1017–1025.
8. T. Loftsson and D. Duchêne, Int. J. Pharm., 2007, 329, 1–11.
9. J. Szejtli, Med. Res. Rev., 1994, 14, 353–386.
10. M. H. Mohamed, L. D. Wilson and J. V. Headley, Carbohydr. Polym., 2010, 80, 186–196.
11. K. Uekama, F. Hirayama and T. Irie, Chem. Rev., 1998, 98, 2045–2076 , ST – Cyclodextrin Drug Carrier Systems.
12. K. Y. Win and S.-S. Feng, Biomaterials, 2005, 26, 2713–2722.
13. B. Gidwani and A. Vyas, Colloids Surf., B, 2014, 114, 130–137.
14. W. Zhang, M. Chen and G. Diao, Electrochim. Acta, 2011, 56, 5129–5136.
15. E. E. M. Eid, A. B. Abdul, F. E. O. Suliman, M. A. Sukari, A. Rasedee and S. S. Fatah, Carbohydr. Polym., 2011, 83, 1707–1714.
16. T. Loftsson, M. E. Brewster, H. Derendorf and N. Bodor, PZ Wiss., 1991, 4, 5–10.
17. E. P. Akçakoca Kumbasar, Ç. Akduman and A. Çay, Carbohydr. Polym., 2014, 104, 42–49.
18. J. L. Soares-Sobrinho, F. L. A. Santos, M. A. M. Lyra, L. D. S. Alves, L. A. Rolim, A. A. N. Lima, L. C. C. Nunes, M. F. R. Soares, P. J. Rolim-Neto and J. J. Torres-Labandeira, Carbohydr. Polym., 2012, 89, 323–330.
19. A. Solanki, J. Mehta and S. Thakore, Carbohydr. Polym., 2014, 110, 338–344.
20. A. R. Solanki, B. V. Kamath and S. Thakore, J. Controlled Release, 2015, 42223, 1–12.
21. F. Zia, K. M. Zia, M. Zuber, S. Kamal and N. Aslam, Carbohydr. Polym., 2015, 134, 784–798.
22. J. Hernandez-Montelongo, N. Naveas, S. Degoutin, N. Tabary, F. Chai, V. Spampinato, G. Ceccone, F. Rossi, V. Torres-Costa, M. Manso-Silvan and B. Martel, Carbohydr. Polym., 2014, 110, 238–252.
23. K. Otta, E. Fenyvesi, B. Zsadon, J. Szejtli and F. Tudos, Cyclodextrin Polymers as Specific Sorbents, 1982.
24. I. Siró, E. Fenyvesi, L. Szente, B. de Meulenaer, F. Devlieghere, J. Orgoványi, J. Sényi and J. Barta, Food Addit. Contam., 2006, 23, 845–853.
25. N. Blanchemain, S. Haulon, B. Martel, M. Traisnel, M. Morcellet and H. F. Hildebrand, Eur. J. Vasc. Endovasc. Surg., 2005, 29, 628–632.
26. F. Trotta, M. Zanetti and R. Cavalli, Beilstein J. Org. Chem., 2012, 8, 2091–2099.
27. S. Daoud-Mahammed, P. Couvreur, K. Bouchemal, M. Chéron, G. Lebas, C. Amiel and R. Gref, Biomacromolecules, 2009, 10, 547–554.
28. S. Dandamudi and R. B. Campbell, Biomaterials, 2007, 28, 4673–4683.
29. D. C. Lev, A. Onn, V. O. Melinkova, C. Miller, V. Stone, M. Ruiz, E. C. McGary, H. N. Ananthaswamy, J. E. Price and M. Bar-Eli, J. Clin. Oncol., 2004, 22, 2092–2100.
30. K. Can, M. Ozmen and M. Ersoz, Colloids Surf., B, 2009, 71, 154–159.
31. R. Nikolaevichyastrebinsky, V. I. Pavlenko, B. Shukhov and S. Technological, Middle East J. Sci. Res., 2013, 18, 1455–1462.
32. A. R. Kiasat and S. Nazari, J. Mol. Catal. A: Chem., 2012, 365, 80–86.
33. A. Solanki and S. Thakore, Int. J. Biol. Macromol., 2015, 80, 683–691.
34. T. Tawa and S. Ito, Polym. J., 2006, 38, 686–693.
35. L. Rueda-Larraz, B. F. d'Arlas, A. Tercjak, A. Ribes, I. Mondragon and A. Eceiza, Eur. Polym. J., 2009, 45, 2096–2109.
36. A. Akbarzadeh, M. Samiei, S. W. Joo, M. Anzaby, Y. Hanifehpour, H. T. Nasrabadi and S. Davaran, J. Nanobiotechnol., 2012, 10, 46.
37. J. O. Park, K. Y. Rhee and S. J. Park, Appl. Surf. Sci., 2010, 256, 6945–6950.
38. H. Kalita and N. Karak, J. Nanoeng. Nanomanuf., 2013, 3, 194–201.
39. F. M. Pardini and J. I. Amalvy, J. Appl. Polym. Sci., 2014, 131, 39799.
40. K. R. Parmar, K. A. Patel, S. R. Shah and N. R. Sheth, J. Inclusion Phenom. Macrocyclic Chem., 2009, 65, 263–268.
41. C. Garnero, V. Aiassa and M. Longhi, J. Pharm. Biomed. Anal., 2012, 63, 74–79.
42. S. Baboota and S. P. Agarwal, J. Inclusion Phenom. Macrocyclic Chem., 2005, 51, 219–224.
43. S. Sajeesh and C. P. Sharma, Int. J. Pharm., 2006, 325, 147–154.
44. Y.-Y. Liu, Y. Yu, G.-B. Zhang and M.-F. Tang, Macromol. Biosci., 2007, 7, 1250–1257.
45. R. Baktiar, Eur. J. Mass Spectrom., 1999, 5, 1.
46. E. M. M. Del Valle, Process Biochem., 2004, 39, 1033–1046.
47. M. L. Shively, B. A. Coonts, W. D. Renner, J. L. Southard and A. T. Bennett, J. Controlled Release, 1995, 33, 237–243.
48. X. Huang and C. S. Brazel, J. Controlled Release, 2001, 73, 121–136.
49. J. Carmichael, W. G. Degraff, A. F. Gazdar, J. D. Minna and J. B. Mitchell, Cancer Res., 1987, 936–942.
50. S. Thakore, P. S. Rathore, R. N. Jadeja, M. Thounaojam and R. V. Devkar, Mater. Sci. Eng., C, 2014, 44, 209–215.
51. P. Alves, R. Cardoso, T. R. Correia, B. P. Antunes, I. J. Correia and P. Ferreira, Colloids Surf., B, 2014, 113, 25–32.
52. D. Fischer, Y. Li, B. Ahlemeyer, J. Krieglstein and T. Kissel, Biomaterials, 2003, 24, 1121–1131.
53. D. Zink, A. H. Fischer and J. A. Nickerson, Nat. Rev. Cancer, 2004, 4, 677–687.
54. M. Carbonari, T. Tedesco and M. Fiorilli, Cytometry, 2001, 44, 113–119.
55. A. Z. M. Badruddoza, M. T. Rahman, S. Ghosh, M. Z. Hossain, J. Shi, K. Hidajat and M. S. Uddin, Carbohydr. Polym., 2013, 95, 449–457.
56. P. Pradhan, J. Giri, F. Rieken, C. Koch, O. Mykhaylyk, M. Döblinger, R. Banerjee, D. Bahadur and C. Plank, J. Controlled Release, 2010, 142, 108–121.
57. B. V. Shetty, R. L. Schowen, M. Slavik and C. M. Riley, J. Pharm. Biomed. Anal., 1992, 10, 675–683.
58. S. Shah, A. Pal, T. Rajyaguru, R. S. R. Murthy and S. Devi, J. Appl. Polym. Sci., 2008, 107, 3221–3229.
59. M. Zanetta, N. Quirici, F. Demarosi, M. C. Tanzi, L. Rimondini and S. Farè, Acta Biomater., 2009, 5, 1126–1136.
60. M. Valodkar, P. S. Nagar, R. N. Jadeja, M. C. Thounaojam, R. V. Devkar and S. Thakore, Colloids Surf., A, 2011, 384, 337–344.
61. S. Aryal, J. J. Grailer, S. Pilla, D. A. Steeber and S. Gong, J. Mater. Chem., 2009, 19, 7879.
62. P. J. Shiny, A. Mukherjee and N. Chandrasekaran, Bioprocess Biosyst. Eng., 2014, 37, 991–997.

---