Amphiphilic hyperbranched polymers with a biodegradable hyperbranched poly(ε-caprolactone) core prepared from homologous AB₂ macromonomer

Abstract

Amphiphilic hyperbranched polymers with a biodegradable hyperbranched poly(ε-caprolactone) core were synthesized by one-pot polyesterification of AB₂ macromonomer (including two carboxyl groups and one hydroxyl group) and poly(ethylene glycol) methyl ether, as a very promising biodegradable and biocompatible biomaterial for drug controlled release.

---

Amphiphilic polymers are able to form self-assembled nanostructures in a selected solvent.¹ The nanostructures include micelles,² vesicles,³ nanoparticles,⁴ nanospheres,⁵ and multicompartmental microcapsules.⁶ They possess some interesting shapes, such as spherical,⁷ cylindrical,⁸ rod,⁹ hexagonal,¹⁰ onion-like,¹¹ and worm-like.¹² Because of their special shapes in a selected solvent and their potential applications in the fields of biomedicine and nanotechnology, amphiphilic polymers have received much attention and extensive research.¹³ Among them, polymeric micelles with small particle size (generally less than 100 nm) are one of the most valuable drug carriers developed in recent years and have some significant advantages, such as stable structure, excellent tissue penetration, wide drug-loading application, long in vivo detention time, and targeted delivery of the drug.¹⁴

Polymeric micelles are typically prepared by self-assembly of block polymers.¹⁵ As the most common type of amphiphilic polymers, block polymers contain diblock polymers, triblock polymers, and multiblock polymers. In addition, amphiphilic polymers also exhibit the different structures, such as linear,¹⁶ comb-like,¹⁷ toothbrush-like,¹⁸ miktoarm,¹⁹ dendritic,²⁰ Y-shaped,²¹ H-shaped,²² and tadpole-shaped.²³ The structures of amphiphilic polymers have a great effect on the performance of polymeric micelles.²⁴ Compared to linear block polymers, diversified structures improve the stability of polymeric micelles and enhance the drug release property of polymeric micelles.²⁵ In the previous studies, we found that the micelles of miktoarm,²⁶ block–graft,^27,28 and comb-like²⁹ polymers have more stable and sustained in vitro drug release than that of linear block polymers. However, amphiphilic hyperbranched polymers are rarely reported so far to form polymeric micelles. It is probably that amphiphilic hyperbranched polymers are difficult to be synthesized³⁰ or the conventional hyperbranched polymers are non-biodegradable polymers such as polystyrene derivatives and polyether derivatives,³¹ which greatly limits in vivo applications of polymeric micelles.

In the past decades, only a few literature examples report on the synthesis of biodegradable hyperbranched polymers.³² For example, 4-(2-hydroxyethyl)-ε-caprolactone,³³ 6-hydroxymethyl-1,4-dioxan-2-one,³⁴ 5-hydroxymethyl-1,4-dioxan-2-one,³⁵ 5-{3-[(2-hydroxyethyl)thio]propoxy}-1,3-dioxan-2-one,³⁶ and 5-(4-hydroxybutyl)-1,3-dioxan-2-one³⁷ were synthesized and applied for the preparation of biodegradable hyperbranched polymers by self-condensing ring-opening polymerization (SCROP). However, the synthesis of these monomers requires a lot of steps and the monomers needs to be purified to high purity in order to be used for SCROP. Inspired by the smart works of Hedrick et al. and others for the preparation of hyperbranched poly(ε-caprolactone) derived from intrinsically branched AB₂ macromonomers,^38–41 here we design homologous AB₂ macromonomer (carboxyl-terminated poly(ε-caprolactone), CPCL) with two carboxyl groups and one hydroxyl group, prepared by ring-opening polymerization of ε-caprolactone with dibenzyl malate⁴² as an initiator and then removing benzyl groups. Then, homologous AB₂ macromonomer CPCL was used to synthesize amphiphilic hyperbranched polymers mPEG–HPCL with biodegradable hyperbranched poly(ε-caprolactone) core by one-pot polyesterification of CPCL and poly(ethylene glycol)methyl ether (mPEG) (Scheme 1). Amphiphilic hyperbranched polymers mPEG–HPCL were used to prepare polymeric micelles. The size, morphology, drug loading property, in vitro drug release behavior, and in vitro cytotoxicity of the micelles were investigated.

Scheme 1 Synthesis of amphiphilic hyperbranched polymers mPEG–HPCL.

In order to be able to calculate molar ratio and molecular weight of mPEG–HPCL from the ¹H NMR spectrum, a small amount of benzyl-protected carboxyl-terminated poly(ε-caprolactone) (BPCL) is added and the integration of the peak areas for the benzyl groups is used as the base. We synthesized two different molecular weights of BPCL by adjusting the feed ratios of dibenzyl malate (Fig. 1a) and ε-CL in the polymerization. The hydrogenolytic deprotection of BPCL was carried out to synthesize two different molecular weights of CPCL. The synthetic results are displayed in Table S1.† The yields of BPCL and CPCL are both high. The degree of polymerization (DP) of BPCL is calculated by comparing the integrations of the signals at 5.17–5.10 ppm (PhCH₂–) for the initiator group and 2.43–2.22 ppm (–CH₂CH₂CH₂CH₂CH₂OCO–) for PCL segments (Fig. 1b). The molar ratios of Bn, and ε-CL in BPCL calculated from the ¹H NMR spectra are close to the ratios in feed. The molecular weights were also determined by GPC. The values are close to those calculated from the ¹H NMR spectra. The results indicate that BPCL are prepared in a high yield through adjusting the feed ratio of initiator and monomer. Compared to the ¹H NMR spectrum of BPCL, the characteristic peaks (7.35 ppm, 5.15 ppm, and 5.11 ppm) of benzyl in the ¹H NMR spectrum of CPCL in Fig. 1c disappear completely, indicating that the protecting groups of BPCL are removed completely and the debenzylation ratio is nearly 100%. The characteristic signals of terminated –CH₂OH are observed from the ¹H NMR spectra of BPCL and CPCL. We synthesized four amphiphilic hyperbranched polymers by adjusting the feed ratios of BPCL, mPEG, and CPCL in the polyesterification. The synthetic results are shown in Table 1. The ¹H NMR spectrum of mPEG_2.7–(HPCL₁₀)_5.2 (where the numbers on the lower right corner of mPEG, HPCL, and brackets stand for the grafted number of mPEG chain, the DP of ε-caprolactone in BPCL, and the conjugated number of CPCL, respectively) is shown in Fig. 1d. The characteristic peaks of benzyl group at 7.35, 5.15, 5.11, and 2.97 ppm, mPEG at 3.65 and 3.38 ppm, and PCL segments at 2.43–2.22 ppm are observed. The DPs are calculated by comparing the integration of these peaks. The M_n data calculated from ¹H NMR spectra are in good agreement with the predicted M_n.

Fig. 1 ¹H NMR spectra (A) of dibenzyl malate (a), BPCL₁₀ (b), CPCL₁₀ (c), and mPEG_2.7–(HPCL₁₀)_5.2 (d) (300 MHz, CDCl₃). The characteristic signals of dibenzyl malate, mEPG, and PCL are labeled using Arabic numerals, “+”, and “×”, respectively. ¹H NMR spectra in the boxes (B–D) are the partial enlarged drawings of ¹H NMR spectra of BPCL₁₀ (b), CPCL₁₀ (c), and mPEG_2.7–(HPCL₁₀)_5.2 (d). The residual solvent signal (CHCl₃) is labeled using “*”. The characteristic signals of terminated benzyl groups and terminated –CH₂OH are labeled using “#” and “&”.

Table 1 Synthetic results and micelle properties of mPEG-b-PCL and mPEG–HPCL

Polymers	Feed ratio^c	^1H NMR		GPC		WmPEG^e	Drug-free micelle^f		Drug-loaded micelle^g
		Ratio^d	Mn	Mn	PDI		Size (nm)	PDI	Size (nm)	PDI	DLC^h (%)	EE^i (%)
a Polymerization conditions: [monomer]/[Sn(Oct)2] = 1000, 130 °C, 24 h, in bulk mPEG^1 with molecular weight of 5000 Da.b DCC, DMAP, THF, room temperature, 24 h, mPEG with molecular weight of 2000 Da.c [BPCL] : [mPEG] : [CPCL] in feed.d [mPEG] : [CL] or [BnO–] : [mPEG] : [CPCL] calculated from ^1H NMR spectra.e Weight fraction of mPEG chain in the polymers.f Measured at a concentration of 0.5 mg mL^−1.g Measured with 10 mg of polymers and 1.0 mg of methotrexate as a model drug.h Drug loading content (DLC).i Entrapment efficiency (EE).
mPEG^1-b-PCL48^a	—	1 : 48	10 500	10 900	1.41	0.48	58	0.135	74	0.145	2.11	21.6
mPEG^1-b-PCL105^a	—	1 : 105	17 000	18 300	1.52	0.29	83	0.157	103	0.162	2.96	30.5
mPEG2.7–(HPCL10)5.2^b	1 : 3 : 5	2 : 2.7 : 5.2	11 300	12 400	1.58	0.48	67	0.144	79	0.154	3.03	31.3
mPEG2.6–(HPCL10)8.4^b	1 : 3 : 9	2 : 2.6 : 8.4	14 800	15 300	1.69	0.35	76	0.136	85	0.143	3.49	36.2
mPEG2.6–(HPCL22)4.8^b	1 : 3 : 5	2 : 2.6 : 4.8	17 700	18 600	1.64	0.29	81	0.164	97	0.156	3.24	33.5
mPEG5.3–(HPCL22)8.6^b	1 : 6 : 9	2 : 5.3 : 8.6	33 000	35 400	1.71	0.32	88	0.132	113	0.152	3.31	34.2

---

The molecular weight and polydispersity index (PDI) determined by GPC further confirm the successful synthesis of amphiphilic hyperbranched polymers. The results are listed in Table 1. The M_n data determined by GPC are consistent with those calculated on the basis of ¹H NMR spectra. Normalized GPC curves of BPCL₁₀, CPCL₁₀, and mPEG_2.7–(HPCL₁₀)_5.2 are shown in Fig. S1.† BPCL₁₀ and CPCL₁₀ are low molecular weights with unimodal molecular weight distributions. mPEG_2.7–(HPCL₁₀)_5.2 is high molecular weight with unimodal molecular weight distribution. The result proves the successful synthesis of mPEG–HPCL.

Polymeric micelles of amphiphilic polymers were prepared by dialysis of a THF solution of the polymer against deionized water. Although amphiphilic hyperbranched polymers mPEG–HPCL contain the mixed hydrophobic (PCL) and hydrophilic (mPEG) segments, interior hyperbranched PCL and exterior linear mPEG can be considered as the separate segments on a molecular level. Therefore, amphiphilic hyperbranched polymers mPEG–HPCL will form core–shell polymeric micelles by self-assembly of many molecules in aqueous solution with HPCL as the core and mPEG as the shell. The micelle size and size distribution were determined by dynamic light scattering (DLS). The micelle sizes increase with the increasing of the molecular weight of hydrophobic HPCL core when the amount of hydrophilic mPEG is kept constant (Table 1). Furthermore, the increasing of the amount of hydrophilic mPEG also increases the micelle sizes. For comparison, amphiphilic linear block polymers mPEG¹-b-PCL²⁶ (where mPEG¹ has a different molecular weight from mPEG in mPEG–HPCL, which is noted in Table 1) are used as a control in investigating micelle property, in vitro drug release, and in vitro cytotoxicity. The DLS histograms and the TEM images of the micelles prepared by mPEG¹-b-PCL₄₈ (where the number on the lower right corner of PCL stands for degree of polymerization of ε-caprolactone) and mPEG_2.7–(HPCL₁₀)_5.2 are shown in Fig. 2. The measured values of DLS are 58 and 67 nm for mPEG¹-b-PCL₄₈ micelles and mPEG_2.7–(HPCL₁₀)_5.2 micelles, respectively. Nanosized spherical particles are observed in the TEM images. The mean diameters of mPEG¹-b-PCL₄₈ micelles and mPEG_2.7–(HPCL₁₀)_5.2 micelles based on the TEM images are about 40 and 50 nm, respectively.

Fig. 2 Size distribution profile determined by DLS for mPEG¹-b-PCL₄₈ micelles (A), mPEG_2.7–(HPCL₁₀)_5.2 micelles (B), and TEM images of mPEG¹-b-PCL₄₈ micelles (C), mPEG_2.7–(HPCL₁₀)_5.2 micelles (D). The micelles were produced at a concentration of approximately 0.5 mg mL⁻¹ for DLS measurement. The micelles were negatively stained by phosphotungstic acid for TEM observation.

Methotrexate was used as a model drug to test the properties of polymeric micelles. The drug-loaded micelles were prepared from the solutions of 10 : 1 (by weight) polymer/drug mixtures by a procedure similar to that for drug-free micelles. To determine the drug loading content (DLC) and entrapment efficiency (EE), the drug-loaded micelle solution was dried by rota-evaporation and the residue was dissolved in DMF to determine the drug concentration through measuring the UV absorbance at 303 nm. The results are summarized in Table 1. The drug-loaded micelles possess a bigger diameter than the corresponding drug-free micelles. When the amount of hydrophilic mPEG is similar, the drug loading capacity is dependent to both the amount and the length of hydrophobic PCL. Increasing the hydrophobic PCL amount or length is beneficial for improving the drug loading capacity. However, the longer hydrophobic PCL is disadvantageous to load the drug for the same molecular weight of hydrophobic HPCL core. The longer hydrophobic PCL has the fewer branching points and the decreasing of the branching point causes hyperbranched polymers to tend to block polymers. Hyperbranched polymers mPEG–HPCL possess better drug loading capacity than block polymers mPEG¹-b-PCL. For example, mPEG_2.7–(HPCL₁₀)_5.2 has an EE of 31.3%, which is approximately 45% higher than the 21.6% of mPEG¹-b-PCL₄₈. The hyperbranched structure of mPEG–HPCL increases the disorder and space between PCL chains, thus enhance the interaction of polymer chains with drug molecules.⁴³

The drug release behavior was investigated in PBS (pH 7.4) by monitoring the drug amounts released from the drug-loaded micelle solution that was placed in a dialysis bag. The in vitro drug release profiles of methotrexate from mPEG¹-b-PCL₄₈ micelles and mPEG_2.7–(HPCL₁₀)_5.2 micelles are shown in Fig. S2.† The drug release from mPEG¹-b-PCL₄₈ micelles is fast with a burst release. The drug release from mPEG_2.7–(HPCL₁₀)_5.2 micelles is sustained for about a week. Approximately 78% and 51% of drug enwrapped in the micelles is release from mPEG¹-b-PCL₄₈ micelles and mPEG_2.7–(HPCL₁₀)_5.2 micelles after 2 days, respectively. These results indicate hyperbranched polymers mPEG–HPCL are a superior biomaterial for drug controlled release.

The in vitro cytotoxicity of block polymers mPEG¹-b-PCL and hyperbranched polymers mPEG–HPCL was investigated by MTT assay. The cell viability data at different concentrations of mPEG¹-b-PCL₄₈ and mPEG_2.7–(HPCL₁₀)_5.2 are shown in Fig. S3.† In a wide range of polymer concentrations, the cell viability is higher than 85%, indicating block polymers mPEG¹-b-PCL and hyperbranched polymers mPEG–HPCL have good biocompatibility.

Conclusions

In summary, we describe a facile strategy for synthesizing amphiphilic hyperbranched polymers mPEG–HPCL with biodegradable hyperbranched poly(ε-caprolactone) core from homologous AB₂ macromonomer. mPEG–HPCL self-assemble to form polymeric micelles. The diameters of polymeric micelles determined by DLS are tens of nanometers and nanosized spherical particles are observed from the TEM images. Hyperbranched polymers mPEG–HPCL have better drug-loading property and more sustained drug release than the corresponding linear block polymers mPEG¹-b-PCL. Therefore, mPEG–HPCL will be a very promising biodegradable and biocompatible biomaterial for drug controlled release.

Acknowledgements

This work was financially supported by the National Basic Research Program of China (2011CB606202) and the National Natural Science Foundation of China (51403241).

Notes and references

1. N. Wiradharma, Y. Zhang, S. Venkataraman, J. L. Hedrick and Y. Y. Yang, Nano Today, 2009, 4, 302–317.
2. M. Licciardi, C. Scialabba, C. Sardo, G. Cavallaro and G. Giammona, J. Mater. Chem. B, 2014, 2, 4262–4271.
3. A. Palanisamy and Q. P. Guo, RSC Adv., 2014, 4, 54752–54759.
4. N. Engelhardt, A. Ernst, A. L. Kampmann and R. Weberskirch, Macromol. Chem. Phys., 2013, 214, 2783–2791.
5. B. E. McKenzie, F. Nudelman, P. H. H. Bomans, S. J. Holder and N. Sommerdijk, J. Am. Chem. Soc., 2010, 132, 10256–10259.
6. I. Choi, S. T. Malak, W. N. Xu, W. T. Heller, C. Tsitsilianis and V. V. Tsukruk, Macromolecules, 2013, 46, 1425–1436.
7. S. Venkataraman, Z. A. Chowdhury, A. L. Lee, Y. W. Tong, I. Akiba and Y. Y. Yang, Macromol. Rapid Commun., 2013, 34, 652–658.
8. J. R. Finnegan, D. J. Lunn, O. E. C. Gould, Z. M. Hudson, G. R. Whittell, M. A. Winnik and I. Manners, J. Am. Chem. Soc., 2014, 136, 13835–13844.
9. G. Rizis, T. G. M. van de Ven and A. Eisenberg, Angew. Chem., Int. Ed., 2014, 53, 9000–9003.
10. H. C. Tsai, C. H. Chang, Y. C. Chiu, S. Y. Lin, C. P. Lin and G. H. Hsiue, Macromol. Rapid Commun., 2011, 32, 1442–1446.
11. G. Q. Cai, H. W. Zhang, P. Liu, L. Q. Wang and H. L. Jiang, Acta Biomater., 2011, 7, 3729–3737.
12. N. Ghasdian, D. M. A. Buzza, P. D. I. Fletcher and T. K. Georgiou, Macromol. Rapid Commun., 2015, 36, 528–532.
13. A. Blanazs, S. P. Armes and A. J. Ryan, Macromol. Rapid Commun., 2009, 30, 267–277.
14. Z. L. Tyrrell, Y. Q. Shen and M. Radosz, Prog. Polym. Sci., 2010, 35, 1128–1143.
15. H. Q. Hu, M. Gopinadhan and C. O. Osuji, Soft Matter, 2014, 10, 3867–3889.
16. T. Rudolph, M. D. Luhe, M. Hartlieb, S. Norsic, U. S. Schubert, C. Boisson, F. D'Agosto and F. H. Schacher, ACS Nano, 2015, 9, 10085–10098.
17. K. H. Kim, J. C. Lee and J. Lee, Macromol. Biosci., 2008, 8, 339–346.
18. W. L. Zhang, Y. L. Li, L. X. Liu, Q. Q. Sun, X. T. Shuai, W. Zhu and Y. M. Chen, Biomacromolecules, 2010, 11, 1331–1338.
19. K. Yoon, H. C. Kang, L. Li, H. Cho, M. K. Park, E. Lee, Y. H. Bae and K. M. Huh, Polym. Chem., 2015, 6, 531–542.
20. X. S. Fan, Z. G. Hu and G. W. Wang, RSC Adv., 2015, 5, 100816–100823.
21. W. W. Zhang, D. L. Zhang, X. S. Fan, G. Y. Bai, Y. M. Guo and Z. G. Hu, RSC Adv., 2016, 6, 20761–20771.
22. P. Tirino, C. Conte, M. Ordegno, R. Palumbo, F. Ungaro, F. Quaglia and G. Maglio, Macromol. Chem. Phys., 2014, 215, 1218–1229.
23. X. J. Wan, T. Liu and S. Y. Liu, Biomacromolecules, 2011, 12, 1146–1154.
24. L. T. Yan and X. M. Xie, Prog. Polym. Sci., 2013, 38, 369–405.
25. Y. Wang and S. M. Grayson, Adv. Drug Delivery Rev., 2012, 64, 852–865.
26. X. J. Zhang, J. A. Cheng, Q. R. Wang, Z. L. Zhong and R. X. Zhuo, Macromolecules, 2010, 43, 6671–6677.
27. X. J. Zhang, F. J. Chen, Z. L. Zhong and R. X. Zhuo, Macromol. Rapid Commun., 2010, 31, 2155–2159.
28. X. J. Zhang, Z. G. Zhang, Z. L. Zhong and R. X. Zhuo, J. Polym. Sci., Part A: Polym. Chem., 2012, 50, 2687–2696.
29. X. J. Zhang, Z. G. Zhang, X. Su, H. Dong, Z. L. Zhong and R. X. Zhuo, Macromol. Chem. Phys., 2015, 216, 1712–1717.
30. M. Hu, M. S. Chen, G. L. Li, Y. Pang, D. L. Wang, J. L. Wu, F. Qiu, X. Y. Zhu and J. Sun, Biomacromolecules, 2012, 13, 3552–3561.
31. D. Ernenwein, A. M. Vartanian and S. C. Zimmerman, Macromol. Chem. Phys., 2015, 216, 1729–1736.
32. Y. Huang, D. L. Wang, X. Y. Zhu, D. Y. Yan and R. J. Chen, Polym. Chem., 2015, 6, 2794–2812.
33. M. J. Liu, N. Vladimirov and J. M. J. Frechet, Macromolecules, 1999, 32, 6881–6884.
34. X. H. Yu, J. Feng and R. X. Zhuo, Macromolecules, 2005, 38, 6244–6247.
35. P. G. Parzuchowski, M. Grabowska, M. Tryznowski and G. Rokicki, Macromolecules, 2006, 39, 7181–7186.
36. P. G. Parzuchowski, M. Jaroch, M. Tryznowski and G. Rokicki, Macromolecules, 2008, 41, 3859–3865.
37. M. Tryznowski, K. Tomczyk, Z. Fras, J. Gregorowicz, G. Rokicki, E. Wawrzynska and P. G. Parzuchowski, Macromolecules, 2012, 45, 6819–6829.
38. M. Trollsas and J. L. Hedrick, Macromolecules, 1998, 31, 4390–4395.
39. M. Trollsas, B. Atthoff, H. Claesson and J. L. Hedrick, Macromolecules, 1998, 31, 3439–3445.
40. J. Choi and S. Y. Kwak, Macromolecules, 2003, 36, 8630–8637.
41. C. Gottschalk and H. Frey, Macromolecules, 2006, 39, 1719–1723.
42. R. H. Lin, J. Castells and H. Rapoport, J. Org. Chem., 1998, 63, 4069–4078.
43. A. Kowalczuk, R. Trzcinska, B. Trzebicka, A. H. E. Muller, A. Dworak and C. B. Tsvetanov, Prog. Polym. Sci., 2014, 39, 43–86.

---

Footnote

† Electronic supplementary information (ESI) available: General experimental details. See DOI: 10.1039/c6ra08531h

---