Chinese Chemical Letters  2020, Vol. 31 Issue (7): 1822-1826   PDF    
Hyperbranched polymer micelles with triple-stimuli backbone-breakable iminoboronate ester linkages
Xuan Zhanga, Jushan Gaob, Xiaoye Zhaob, Zhaotie Liua, Zhongwen Liua, Ke Wangb,*, Guo Lia,*, Jinqiang Jianga,*     
a Key Laboratory of Syngas Conversion of Shaanxi Province, Key Laboratory of Applied Surface and Colloid Chemistry, Ministry of Education, School of Chemistry and Chemical Engineering, Shaanxi Normal University, Xi'an 710062, China;
b School of Pharmacy, Health Science Center, Xi'an Jiaotong University, Xi'an 710061, China
Abstract: Hyperbranched polymers have attracted increasing interests due to their unique structure-related advantages and have been utilized as drug carriers for controlled delivery. However, it is still challenging to prepare multi stimuli-responsive hyperbranched polymers to sense and response the complex yet delicate changes in physiological environment. Herein, we propose a triple-stimuli backbone-breakable hyperbranched polymer (HBP(OEG-IB)), which is prepared via the convenient iminoboronate multicomponent reaction of α, w-di(1, 2-diol)s oligo(ethylene glycol), tris(3-aminopropyl)amine and 2-formylphenylboronic acid. Upon the stimulation of CO2, lactic acid and glutathione, micelles formed by HBP(OEG-IB) could be disrupted via the dissociation of iminoboronate ester bond to subsequent release incorporated camptothecin (CPT). Cell experiments show that the HBP(OEG-IB) is non-toxic but can enhance the therapeutic effect of CPT thanks to the effect of the protonated tertiary amino groups. The demonstration made in this work can enrich the design of responsive HBPs and can be readily applied to other systems with tunable responsive properties and functions.
Keywords: Hyperbranched polymer    Multi-stimuli responsiveness    Iminoboronate esters    CPT    Drug delivery    

The highly branched and globularly architected topology of hyperbranched polymers endows them with the advantages including low viscosity, good solubility, and high functionality, compared to linear polymers [1,2]. Recently, backbone-responsive amphiphilic hyperbranched polymers, which combine the advantages of hyperbranched polymers and backbone-responsive polymers, have attracted considerable attention due to their potential in the fields such as drug delivery, tissue engineering, and biomedical application [3-8]. These backbone-responsive hyperbranched polymers can be prepared via recently-developed facile methods like proton transfer polymerization [9,10] or random copolymerization [4,11,12], and the responsiveness of prepared hyperbranched polymers can be conveniently tuned by changing the type of monomers and their compositions. However, the physiological environment is changeable and rather complex [5,13], requiring the drug carriers to simultaneously sense different environmental signals and respond more flexibly and dynamically, but so far the design and synthesis of hyperbranched polymers responsive to more than three kinds of stimuli are still challenging, and the relevant works are rare.

On another front, the covalent yet reversible nature of boronate esters inspires the development of dynamic polymers toward the applications such as blue-emissive materials, self-healing materials, covalent organic frameworks and RNA mimics [3,14-18]. Boronate esters can be formed by boronic acid and various types of 1, 2- and 1, 3-diols, making it an ideal motif to bind with versatile hydroxylated targets [19-22]. Boronic ester-based block copolymers have also been synthesized to exhibit their potentials as drug-releasing carriers, as they response fast and drastic to pH changes and competing diols [14,23]. However, multi-stimuli (more than two) responsive hyperbranched polymers based on the reversibility of boron esters, to the authors' knowledge, have been seldom reported. In this work, we aim to demonstrate the multistimuli responsive drug delivery behavior of micelles formed by hyperbranched oligo(ethylene glycol) with iminophenylboronate ester linkages (HPB(OEG-IB)). We prepared the target polymer via a one-pot multicomponent reaction, which amino, phenylboronate acid and 1, 2-diol groups are condensed to form boronate ester bonds stabilized by a dative bond formed with adjacent imine nitrogen. The formed iminoboronate ester bonds are located in the backbone of HPB(OEG-IB), exhibiting a faster and more sensitive responsive behavior to the stimulation of acidic or reductive signals. The obtained polymer can form micelles in aqueous solution and can be degraded upon the stimulation of bubbling CO2, adding lactic acid (LA) or glutathione (GSH), which are found to have higher enrichment levels inside tumor cells compared to that of normal cells [13,24].

The synthesis of the designed HBP(OEG-IB) and its stimuliresponsive dissociation behavior for drug delivery are schematically shown in Fig. 1. The convenient iminoboronate reaction of α, w-di(1, 2-diol)s oligo(ethylene glycol) (Diols-OEG), tris(3-aminopropyl)amine and 2-formylphenylboronic acid (FPBA) results in a backbone-breakable iminoboronate linkage. In low pH environments, the tertiary amines are protonated and the iminoboronate esters are hydrolyzed, leading to the breakage of the amphiphilic balance and subsequent dissociation of micelles. The successful synthesis of the product is evidenced by 1H NMR results (Fig. S2 in Supporting information), as the intensity of the aldehyde peak located at 9.92 ppm gradually decreased and a new peak at 8.57 ppm emerged and the intensity increased within the reaction time (from 0 to 24 h). The aldehyde peak totally disappeared after 24 h, indicating the completeness of the iminoboronate reaction to prepare the designed HBP(OEG-IB).

Download:
Fig. 1. The self-assembly of the synthesized hyperbranched polymer (HBP(OEG-IB)) and its multi-stimuli triggered dissociation and drug release behavior upon the stimulation of CO2, LA and GSH.

The triggered dissociation behavior of the formed micelles under the stimulation of CO2, LA and GSH are then investigated, respectively. Fig. 2a shows the absorption spectra of polymer solution before and after bubbling CO2. It can be seen that polymeric solution before CO2 bubbling exhibits an characteristic peak at 252 nm with a shoulder peak at 285 nm, corresponding to the phenyl ring in aldehyde phenylboronic acid monomers and iminoboronate ester groups in the HBP(OEG-IB). After bubbling CO2, the intensity of the peak at 252 nm increased while that at 285 nm weakened, indicating the hydrolysis of iminoboronate ester bonds and dissociation of the micelles formed by HBP(OEGIB) in a weak acidic environment after CO2 incorporation. Moreover, the curves of fluorescence microscopy (Fig. S8b in Supporting information) show a rapid increasing trend at 422 nm under the excitation of 285 nm after bubbling CO2. 1H NMR results also show that the disappearance of the peaks of imine at 8.57 ppm and the peak of -CH2-N < adjacent to iminoboronate ester at 6.07 ppm, accompanied by the emergence of the peak of aldehyde group at 9.92 ppm (Fig. S5 in Supporting information). These results prove that the dissociation of iminoboronate ester groups and disruption of the formed micelles triggered by CO2. Similarly, the presence of LA or GSH can also induce the breakage of the HBP (OEG-IB) and disruption of the micelles. These are evidenced by the same characterization protocols. After adding LA or GSH into the micellar solution, the curves of absorption spectra shows an increase of intensity at 252 nm and simultaneous decrease at 285 nm, the fluorescence emission band exhibits a decrease of peak intensity at 422 nm. 1H NMR analyses also demonstrate the disappearance of peak at 6.07 nm and 8.57 nm along with the emergence at 9.92 nm for aldehyde group. These results clearly indicate the effectiveness of LA and GSH acting as triggering cues to induce dissociation of polymeric micelles.

Download:
Fig. 2. (a) Absorption spectra of PM (0.10 mg/mL ×1.5 mL) upon CO2, where the inset shows the plot of the normalized absorption intensity at 252 nm (252). (b) TEM images of HBP(OEG-IB) polymer micelles.

TEM and DLS measurements were conducted to investigate changes of the micellar morphologies before and after stimulation. Fig. 2b shows the spherical-shaped micelles of HBP(OEG-IB) with an average diameter of 23.9 ± 5.3 nm. As for DLS, the determined average hydrodynamic diameter of the micelles is 78.8 ± 19.1 nm in aqueous medium, which is almost unchanged at different concentrations, indicating a good stability of the micellar structure (Fig. S4a in Supporting information). After the stimulation of CO2, LA or GSH, the micellar aggregates can no longer be observed, with only some precipitants left on the copper mesh, implying the dissociation of micellar structure (Figs. S11a-c in Supporting information).

We then studied the application potential of this triple-stimuli responsive HBP(OEG-IB) as nanocarriers for controlled drug delivery. Here camptothecin (CPT) is selected as a model drug, which has a good therapeutic effect for cancer treatment [25]. However, its application is limited by its high toxicity, poor watersolubility and structure instability [25]. After encapsulated into micelles, its premature release can be minimized in neutral environment. Once reaching the acidic endosomal/lysosomal locations, the CPT-loaded hyperbranched polymer micelles can be taken up by the target cancer cells via the endocytosis of cell membranes [26], which significantly improve the therapeutic effect of CPT to tumor cells and simultaneously minimizing undesirable side effects. Under physiologically reductive environment, CPT-loaded micelles could be triggered to dissociate via the cleavage iminoboronate ester bonds by the above investigated signals that are enriched in tumor cells.

Afterwards, the controlled drug release behaviors were investigated. As shown in Fig. 3a, the absorption band of CPT at 366 nm declined drastically after bubbling 20.0 mL CO2, indicating the release of CPT from the micelles and subsequent precipitate after acid-induced disruption. Similar results are obtained after adding LA or GSH, the normalized intensity (366) experienced a notable drop (Figs. S14a and S15a in Supporting information), suggesting the efficient disruption of the micellar structure and CPT release. However, the extent of decrease of sample after addition of GSH is less prominent compared to that obtained with the CO2 and LA stimulated samples, ascribing to the addition of thiol group in GSH to the electrophilic imine carbon center [27,28], thus resulting in a lower extent of micelle dissociation and CPT release.

Download:
Fig. 3. Absorption changes and release profiles of CPT-PM (0.20 mg/mL ×1.5 mL) upon CO2 (a, b), LA (c) and GSH (d). The inset showing the normalized absorption changes at 366 nm (366).

To investigate the kinetics of drug releasing behavior upon stimulation, the micellar solution was sealed in a dialysis cap with the dialysis membrane immersed in the cuvette water. Afterwards, the CPT content in the cuvette water was detected by absorption spectrometry. Fig. 3b indicated that the release of CPT from CPTPM is much enhanced upon CO2. For the micelle solution without CO2, a continuous release of CPT is observed, and about 29.7% of the loaded drugs are released after 10 h. In contrast, the ratio of drug delivery is accelerated by the stimulation of CO2, as after 10 h about 48.0% ± 1.82%, 57.3% ± 0.86%, 67.7% ± 1.66% and 77.7% ± 1.47% of CPT molecules are released from the micelles with the added amount of CO2 are 0.5 mL, 1.5 mL, 3.0 mL and 12.0 mL, respectively. The relative slow leakage of CPT from micelles without stimulation is advantageous in their circulation in blood, and much enhanced releasing ratio is desired to exhibit its therapeutic effect at tumor sites. Similar results are obtained by the solutions responsive to LA and GSH. There can be seen that the drug-releasing ratio is increased to different extents after adding certain amounts of LA or GSH, and a complete release of CPT can be obtained after adding 2.5 mmol/L LA or 8.0 mmol/L GSH, respectively (Figs. 3c and d).

TEM images of the micelles before and after drug release were also obtained (Figs. S19a-d in Supporting information). After loaded with CPT, the average diameter of the micelles increased to 113.5 ± 28.1 nm, which is much larger compared to that of native micelles without drug loading, indicating the effective encapsulation of CPT into the hydrophobic core of the formed micelles. After stimulation, spherical structures cannot be observed in the TEM images, with only some precipitants left on the substrate, again proves the disruption of the micelles and the release of the loaded CPT.

It is known that the microenvironment conditions near the tumor tissue are at a lower pH and a higher reduction potential compared to normal tissue, thus the micellar dissociation and drug release of encapsulated CPT can be effectively triggered. To verify this, in vitro cytotoxicity of the micelles are investigated. Fig. 4a showed that the cell viability of HeLa cells co-cultured with blank micelles for 24 h is still higher than 85.6%, clearly proving the good biocompatibility of the designed polymer. In sharp contrast (Fig. 4b), the cell viability of HeLa cells with CPT-loaded micelles (0.01–10.0 μg/mL) decreased to 88.1%, 71.7%, 54.8%, 54.3%, 52.4%, 50.0%, 45.1% and 38.0%, respectively, lower than that of the batches with directly co-culturing CPT. The possible mechanism is that the tertiary amines in HBP(OEG-IB) can be protonated and interact with the charges on the surfaces of cell membranes [29,30], leading to enhanced uptake of the drug-loaded micelles by cells, and effective release of CPT from the micelles inside the cells with an acidic pH and high GSH level.

Download:
Fig. 4. In vitro cell viability of HeLa cells co-cultured with HBP(OEG-IB) (a), free CPT and CPT-PM (b) for 48 h.

Confocal laser scanning microscopy (CLSM) was used to track the cellular internalization of free CPT (Fig. 5a) and CPT-loaded micelles (Fig. 5b). The CLSM images of free CPT and CPT loaded micelles were labeled by the fluorescence of PI (red) for nuclear sites, and CPT (blue) and lysotracker (green) for lysosomes, respectively. As shown in Figs. 5a and b, after incubation, the blue fluorescence signals for free CPT molecules and CPT-loaded micelles show a time-dependent enhancement within 12 h, which can be visually noticed in the nucleus of Hela cells, indicating the CPT molecules entered cells in the way of passively cytoplasm [26], and the micelles can also be protonated by the amino groups and interact with the charges on the surfaces of cell membranes, indicating an enhanced uptake of the drug-loaded micelles by cells. Moreover, after 12 h cultivation, the red fluorescence enabled by PI is significantly increased in the nucleus, indicating the apoptosis of a large number of HeLa cells after co-culturing of the cells with free CPT molecules or CPT-loaded micelles, as PI molecules can only penetrate the membrane of dead cells. Moreover, the blue fluorescence of CPT molecules loaded within micelles has a larger overlapped area with the red fluorescence of PI compared to that of free CPT molecules, which suggests that more CPT molecules can enter the cells by the transportation of micelles based on the effect of the protonated amino groups. This phenomenon could also prove that CPT-PM first entered lysosome and then permeated into the nucleus of tumor cells [31].

Download:
Fig. 5. CLSM images of HeLa cells co-cultured with free CPT (a) and CPT-PM (b) at 1 μg/mL CPT for 12 h, respecctively. Fleuroscent PI, lyso tracker green and CPT were used to label the nuleus and lysosomes of HeLa Cells. The scale bars represent 50 μm.

An Annexin V-FITC/PI dual-staining assay was further utilized to investigate the death process in the way of apoptosis of HeLa cells by flow cytometry after incubation with free CPT and CPT-PM (Fig. 6). It can be seen that 68.5% of the cells are apoptotic in the early stage for the group administrated by free CPT molecules, and this value increased to 90.7% after the later stage. For the cell group regulated by CPT-loaded micelles, although only 0.3% apoptotic cells were observed in the early stage, the amount of apoptotic cells can be even higher than that administrated by free CPT molecules. In contrast, the control cell group without drug administration shows an apoptosis ratio of about only 5.5%. The above results indicate the effectiveness of drug administration by loading the CPT into HBP(OEG-IB)-formed micelles to exhibit an even improved therapeutic effect.

Download:
Fig. 6. Flow-cytometric analysis of annexinV-FITC/PI dual-staining of HeLa cells co-cultured with free CPT and CPT-PM at CPT concentration of 5 μg/mL for 24 h.

In this work, an amphiphilic hyperbranched polymer (HBP (OEG-IB)) bearing triple stimuli-sensitive dissociable iminoboronate linkages was successfully synthesized by a one-pot multi-component reaction process. The obtained polymer can selfassemble into micellar aggregates in water, which featured in narrow particle size distribution, high drug loading capacity and high micellar stability. The iminoboronate linkages are liable in acidic and reductive environments, thus CO2, lactic acid and GSH are utilized to realize triggered dissociation of the formed micelles and subsequent release of CPT, as these chemicals are enriched near the tumor sites. DLS, TEM, 1H NMR, absorption spectroscopy and fluorescence spectroscopy were utilized to characterize the dissociation and release behaviors, and the results clearly prove the effectiveness of stimuli-triggered cleavage of iminoboronate ester bonds under these stimuli. The cytotoxicity and cell experiments revealed that the micellar carriers are non-toxic but can enhance the therapeutic effect of loaded CPT for cell apoptosis by the tertiary amines protonated and interacted with the charges on the surfaces of cell membranes. The design demonstrated in this study renders a direct, facile and general way for synthesis of multi-stimuli responsive hyperbranched polymers with more flexible and tunable responsive behaviors toward applications in the complex physiological environment.

Declaration of competing interest

There is no declaration of interest in this investigation. The authors declare no competing financial interest.

Acknowledgments

This work was financially supported by the Nature Science Foundation of China (Nos. 51803115, 81773686 and 21636006), the Nature Science Foundation of Shaanxi Province (No. 2019JQ-528), the Fundamental Research Funds for the Central Universities (Nos. 2017TS024, GK201801003, GK201802009 and GK201901001), the China Postdoctoral Science Foundation Funded Project (No. 2017M623106), and Young Talent Fund of University Association for Science and Technology in Shaanxi, China (No. 20180602), the Program of Introducing Talents of Discipline to Universities (No. B14041).

Appendix A. Supplementary data

Supplementary material related to this article can befound, in the online version, at doi:https://doi.org/10.1016/j.cclet.2020.03.018.

References
[1]
D.L. Wang, Y. Jin, X.Y. Zhu, et al., Prog. Polym. Sci. 64 (2017) 114-153. DOI:10.1016/j.progpolymsci.2016.09.005
[2]
T. Nishikubo, H. Kudo, K. Maruyama, et al., Polym. J. 38 (2006) 447-456. DOI:10.1295/polymj.38.447
[3]
Z. Wei, J.H. Yang, Z.Q. Liu, et al., Adv. Funct. Mater. 25 (2015) 1352-1359. DOI:10.1002/adfm.201401502
[4]
B. Yu, X.S. Jiang, R. Wang, et al., Macromolecules 43 (2010) 10457-10465. DOI:10.1021/ma1023632
[5]
Y.Y. Zhuang, H.P. Deng, Y. Su, et al., Biomacromolecules 17 (2016) 2050-2062. DOI:10.1021/acs.biomac.6b00262
[6]
L. Yang, H. Sun, Y. Liu, et al., Angew. Chem. Int. Ed. 57 (2018) 17048-17052. DOI:10.1002/anie.201809753
[7]
H. Li, W.Z. Chen, F.F. Xu, et al., Macromol. Rapid Comm 39 (2018) 1800053. DOI:10.1002/marc.201800053
[8]
Y. Huang, F. Qiu, L.Y. Shen, et al., ACS Nano 10 (2016) 10489-10499. DOI:10.1021/acsnano.6b06450
[9]
Z.F. Jia, G.L. Li, Q. Zhu, et al., Chem. -Eur. J. 15 (2009) 7593-7600. DOI:10.1002/chem.200900345
[10]
H. Chen, Z.F. Jia, D.Y. Yan, et al., Macromol. Chem. Phys. 208 (2007) 1637-1645. DOI:10.1002/macp.200700049
[11]
J.Y. Liu, Y. Pang, W. Huang, et al., Macromolecules 43 (2010) 8416-8423. DOI:10.1021/ma1015819
[12]
G.X. Cao, G. Li, Q. Yang, et al., Macromol. Rapid Commun 39 (2018) 1700684. DOI:10.1002/marc.201700684
[13]
W.X. Gao, Y.L. Hu, L. Xu, et al., Chin. Chem. Lett. 29 (2018) 1795-1798. DOI:10.1016/j.cclet.2018.05.022
[14]
Y.P. Li, W.W. Xiao, K. Xiao, et al., Angew. Chem. Int. Ed. 51 (2012) 2864-2869. DOI:10.1002/anie.201107144
[15]
O.R. Cromwell, J. Chung, Z.B. Guan, J. Am. Chem. Soc. 137 (2015) 6492-6495. DOI:10.1021/jacs.5b03551
[16]
A.P. Bapat, D. Roy, J.G. Ray, et al., J. Am. Chem. Soc. 133 (2011) 19832-19838. DOI:10.1021/ja207005z
[17]
S. Delpierre, B. Willocq, G. Manini, et al., Chem. Mater. 31 (2019) 3736-3744. DOI:10.1021/acs.chemmater.9b00750
[18]
P.M.S.D. Cal, J.B. Vicente, E. Pires, et al., J. Am. Chem. Soc. 134 (2012) 10299-10305. DOI:10.1021/ja303436y
[19]
S.R. Chaudhari, N. Suryaprakash, J. Org. Chem. 77 (2012) 648-651. DOI:10.1021/jo202334d
[20]
R.J. Ma, C. Zhang, Y. Liu, et al., RSC Adv. 7 (2017) 21328-21335. DOI:10.1039/C7RA01742A
[21]
A. Pettignano, S. Grijalvo, M. Haring, et al., Chem. Commun. 53 (2017) 3350-3353. DOI:10.1039/C7CC00765E
[22]
Y. Xu, Y.Y. Lu, L. Wang, et al., Colloid. Surface. B 141 (2016) 318-326. DOI:10.1016/j.colsurfb.2016.01.044
[23]
X.Y. Xu, H. Shang, T.Y. Zhang, et al., Int. J. Pharm. 548 (2018) 649-658. DOI:10.1016/j.ijpharm.2018.07.020
[24]
Y. Ding, C. Du, J.W. Qian, et al., Polym. Chem. 9 (2018) 3488-3498. DOI:10.1039/C8PY00579F
[25]
L. Qiu, L.F. Zhao, C.F. Xing, et al., Chin. Chem. Lett. 29 (2018) 301-304. DOI:10.1016/j.cclet.2017.09.048
[26]
G.C. Yu, X.L. Zhao, J. Zhou, et al., J. Am. Chem. Soc. 140 (2018) 8005-8019. DOI:10.1021/jacs.8b04400
[27]
F.M.F. Santos, A.I. Matos, A.E. Ventura, et al., Angew. Chem. 129 (2017) 9474-9478. DOI:10.1002/ange.201703492
[28]
R.M.R.M. Lopes, A.E. Ventura, L.C. Silva, et al., Chem. -Eur. J. 24 (2018) 12495-12499. DOI:10.1002/chem.201802515
[29]
M. Masui, H. Takata, T. Kominami, Electrophoresis 23 (2002) 2087-2095. DOI:10.1002/1522-2683(200207)23:13<2087::AID-ELPS2087>3.0.CO;2-R
[30]
V.A. Sethuraman, Y.H. Bae, J. Control. Release 118 (2007) 216-224. DOI:10.1016/j.jconrel.2006.12.008
[31]
X.X. Shi, M.L. Hou, X.Q. Ma, et al., Biomacromolecules 20 (2019) 1190-1202. DOI:10.1021/acs.biomac.8b01566