RGD conjugated, Cy5.5 labeled polyamidoamine dendrimers for targeted near-infrared fluorescence imaging of esophageal squamous cell carcinoma

Abstract

The early detection of esophageal squamous cell carcinoma (ESCC), one of the most common human neoplasms, is of great importance in improving prognosis. However, there is a lack of screening methods with high sensitivity for early diagnosis of ESCC. In this study, we developed a αvβ3 integrin targeted near-infrared (NIR) fluorescence nanoprobe by conjugation of NIR dye Cy5.5 and tumor vasculature targeted cyclic RGDfK peptide onto a polyamidoamine (PAMAM) dendrimer (RGD–PEG–PAMAM–Cy5.5). The applicability of the as-prepared targeted nanoprobe for NIR fluorescence imaging of 4-NQO induced large and tiny esophageal neoplasms in mice was examined. It was demonstrated that the NIR fluorescence imaging with our targeted nanoprobe enables the identification of ESCC at an early stage.

---

Introduction

Esophageal squamous cell carcinoma (ESCC) is one of the most common human neoplasms and ranks the sixth leading cause of cancer related deaths worldwide.^1,2 The high mortality rate of ESCC is primarily associated with late diagnosis because of the limited sensitivity of currently used white light (WL) upper endoscopy.^3–5 Therefore, development of new imaging technology with high sensitivity is highly desirable for early detection of esophageal neoplasms.

Fluorescence imaging, which provides real-time information about tumor margins and extent of tumor spread, is emerging as a promising modality for the detection of tumors with improved accuracy.^6–9 Generally, near-infrared (NIR) fluorescence (650–900 nm) is preferred owning to its low absorption and relatively low auto-fluorescence, which thereby greatly improves the sensitivity.^10,11 With the advances in nanotechnology, a variety of nanoparticles-based NIR fluorescent probes have been developed for in vivo imaging applications.^12–15 However, one of the major concerns for in vivo NIR fluorescence imaging is efficient delivery of the nanoprobes in the tissue of interest. To improve the targeting ability, a variety of targeting moieties such as small-molecule, peptide, and antibody-based ligands have been conjugated to NIR nanoprobes.^16–20 Targeted fluorescence nanoprobe has been reported for detection of esophageal adenocarcinoma.²¹ However, the development of a targeted nanoprobe for in situ NIR imaging of ESCC, especially, at its early stage, remains unexplored so far.

Herein, we designed and fabricated a novel targeted NIR nanoprobe by conjugation of Cy5.5 and cyclic RGDfK peptide (denoted as RGD thereafter) onto polyamidoamine (PAMAM) dendrimer (Scheme 1). The highly branched architecture of PAMAM dendrimers not only provide a nanoplatform to simultaneously load drugs, targeting molecules, and imaging agents, but also promote multivalent binding of target ligands to target cellular receptors.^22–25 Meanwhile, RGD was selected as the targeting moiety because of its high affinity towards αvβ3 integrin, which is over-expressed on the endothelial cells of esophageal neoplasms. In addition, the linear PEG was used to link the RGD and PAMAM. Moreover, the incorporated PEG improves the biocompatibility and prolongs the blood circulation time of nanoprobe. The potential of as-prepared targeted nanoprobe (RGD–PEG–PAMAM–Cy5.5) for NIR fluorescence imaging of early ESCC in mice was exploited.

Scheme 1 Synthesis of targeted and non-targeted NIR nanoprobes.

Experimental

Materials

G5.0 PAMAM dendrimer (5 wt% in methanol), 4′,6-diamidino-2-phenylindole (DAPI), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), dimethyl sulfoxide (DMSO), and carcinogen 4-nitroquinoline 1-oxide (4-NQO) were purchased from Sigma-Aldrich (St. Louis, USA). Cy5.5 was obtained from Thermofisher Scientific (USA). c(RGDfK) was supplied by Gl Biochem Ltd (Shanghai, China). Malemide–PEG2000–NHS and malemide–PEG2000–SMC were received JenKem Technology (Beijing, China). NHS, EDC and other from JenKem Technology (Beijing, China). NHS, EDC and other chemicals were supplied by Sinopharm Chemical Reagent (Beijing, China). Fetal bovine serum, RPMI medium 1640, trypsin were purchased from Hyclone (Logan, USA). Alexa Fluo 488 labeled goat anti-rabbit secondary antibodies were received from Abcam (Cambridge, USA). Rabbit anti-mouse/human β3-integrin primary antibodies were obtained from Epitomics (Burlingame, USA).

Synthesis of NIR nanoprobes

Cy5.5 labeled PAMAM dendrimers (PAMAM–Cy5.5) was first synthesized. Briefly, 0.58 mg of EDC (3 μmol) and 0.86 mg of NHS (7.5 μmol) were added into 2.54 mg Cy5.5 (2 μmol) in 0.5 ml of NaAc–HAc buffer (0.1 M, pH 6.0). After stirring at room temperature for 30 min, the resulting solution was added dropwise into 0.72 ml of PAMAM (1 μmol) and the pH was adjusted to 8.0 by HCl, the mixture was kept stirring at room temperature for 24 h to produce PAMAM–Cy5.5. Meanwhile, PEG–RGD was prepared by reaction of 7.78 mg of RGD (11 μmol, dissolved in 0.5 ml of 0.1 M, pH 6.0 NaAc–HAc buffer) with 20 mg of malemide–PEG2000–NHS (10 μmol). To produce RGD conjugated PAMAM–Cy5.5, PAMAM–Cy5.5 was added into PEG–RGD (PAMAM : PEG : RGD = 1 : 20 : 22, molar ratio). After stirring at room temperature overnight, the mixture was transferred into an ultrafiltration tube (Millipore, USA, MWCO 5000) and centrifuged at 6000 rpm for 10 min for 5–6 times. The products were retrieved by freeze drying. The yield of RGD–PEG–PAMAM–Cy5.5 was 71.3% (percentage of the total amount of raw materials, PAMAM, PEG, and RGD, w/w). For comparison purpose, the non-targeting nanoprobe PEG–PAMAM–Cy5.5 was also prepared as well with the same procedure described above except the PAMAM–Cy5.5 was added into malemide–PEG2000–SMC in 0.5 ml of 0.1 M NaAc–HAc buffer (pH 6.0) with the molar ratio of PAMAM : PEG of 1 : 14.

Characterization

¹H nuclear magnetic resonance (NMR) spectra were acquired on an AS500 NMR spectrometer (Varian, USA) at 500 MHz. Fourier transform infrared (FTIR) spectra were collected on a Nicolet Avatar 300 FT-IR spectrometer (Thermo Nicolet Instrument Corp., USA). Sample powders were mixed with KBr and then pressed into a pellet with minimum pressure. A total of 50 scans were accumulated with a resolution of 4 cm⁻¹ for each spectrum. The fluorescence emission spectrum and ultraviolet absorbance spectrum were recorded on an F-2500 fluorescence spectrophotometer (Hitachi, Japan) and a UV-3010 ultraviolet spectrophotometer (Hitachi, Japan), respectively.

Cell culture

Human esophageal squamous carcinoma TE-1 cell lines were purchased from Riken Cell Bank (Tsukuba, Japan) and cultured in Roswell Park Memorial Institute (RPMI) 1640 supplemented with 10% fetal bovine serum (FBS) at 37 °C in a humidified 5% CO₂ atmosphere.

Cytotoxicity

To evaluate the cytotoxicity of NIR nanoprobes, the MTT cell proliferation assay was conducted. In brief, suspensions of TE-1 cell with a cell density of 5 × 10³ cells per well in 100 μl cell culture medium were added to 96-well plates by serial dilutions. Eight replicates were performed under the same conditions. After 24 h for cell attachment, the cells were treated with NIR nanoprobes (targeted or non-targeted) or unmodified G5.0 PAMAM dendrimer with final concentrations in a range of 0.01–4 μM and incubated for 48 h at 37 °C in 5% CO₂. Then, the medium was removed and MTT (100 μl, 0.5 mg ml⁻¹) was added to each well. After incubation for another 4 h, the medium was removed and DMSO was added. The optical density (OD) value was then measured at wavelength of 490 nm (Multiskan Spectrum microplate spectrophotometer, USA) and the cell viability was calculated by the ratio of optical density values between the experimental and the control groups. Data were presented as the mean value with standard deviations from three independent experiments.

Flow cytometry analysis

TE-1 cells were seeded in 6-well plates with a density of 1 × 10⁵ cells per well. After 24 h of incubation, TE-1 cells were treated with nanoprobes at different concentrations and incubated for another 24 h. The cells were then washed, centrifuged, fixed, and analyzed by Guava EasyCyte Mini flow cytometry system (Millipore, USA) equipped with a 630 nm Ar laser.

Esophageal neoplasms model

All animal experiments were in accordance with the National Institute of Health Guide for Care and Use of Laborartory Animals (publication no. 85-23, revised 1985). Moreover, the experimental procedures were previously approved by the ethics committee of Capital Medical University, Beijing, China. The carcinogen 4-NQO was induced in mice esophagus by the drinking water method reported elsewhere.²⁶ In brief, six-week-old female C57Bl/6 mice, obtained from Charles River Laboratories (Beijing, China), were subjected to drinking water contained 100 μg ml⁻¹ 4-NQO continuously for 16 weeks. Then, the mice were kept for another 12 weeks with free access to normal drinking water for further use.

NIR fluorescence imaging

NIR fluorescence images of esophagus bearing large or tiny neoplasms were acquired on a NightOWL LB 983 in vivo imaging system (Berthold Technologies, Germany) equipped with a 630 nm excitation filter and a 680 nm emission band-pass filter set and the exposure time was 0.5 s. The fluorescence intensities on acquired NIR fluorescence images were quantified by ImageJ software (NIH, USA).

Histological and immunohistological staining

After NIR fluorescence imaging, the mice esophagus bearing neoplasms were fixed, dehydrated, and sectioned with a thickness of 6 μm. The esophagus sections from each mouse were divided into two groups. Group one was stained with hematoxylin and eosin (H&E), while group two was stained with β3-integrin primary antibody followed the fluorophore-labeled secondary antibody and DAPI. CLSM images were collected on a LEICA TCS SP5 confocal microscope (Leica Microsystems, Germany) using a 40× oil-immersion lens.

Statistical analysis

Data were expressed as mean ± SD when the sample number is above 3. Statistical difference was evaluated with two-tailed corrected Student's t test (SPSS, IBM) and p < 0.05 was considered as significant.

Results and discussion

Characterization

The successful synthesis of targeted nanoprobe was supported by ¹H NMR spectra (Fig. 1). Fig. 1A provides the ¹H-NMR spectrum of PAMAM G5.0 dendrimer and the assignment of observed peaks was as follows. δ 2.25–2.40 (m, –CH₂CH₂CONH–); δ 2.45–2.58 (m, –CH₂CH₂N[double bond splayed right]); δ 2.60–2.72 (m, –NCH₂CH₂CO–); δ 2.75–2.85 (m, –CH₂CH₂NH₂); δ 3.10–3.20 (m, –CONHCH₂CH₂–).^27,28 In addition to these characteristic peaks, the ¹H-NMR spectra of both non-targeted and targeted NIR nanoprobes exhibited newborn multiple peaks around 3.6 ppm due to the –CH₂CH₂O– repeating units of PEG (Fig. 1B and C), confirming the conjugation of PEG to PAMAM. Moreover, the ¹H-NMR spectrum of targeted NIR nanoprobe presented couples of weak peaks between 7.1 and 7.3 ppm, which were originated from phenyl protons of RGD. These peaks were absent in the ¹H-NMR spectrum of non-targeted NIR nanoprobes (Fig. 1C), validating the presence of RGD in targeted NIR nanoprobes. Furthermore, the number of PEG chains and RGD per nanoprobe was estimated from the nanoprobe was estimated from the integrated areas of ¹H-NMR peaks. It was found that the molar ratio of RGD : PEG : PAMAM in the targeted NIR nanoprobes was 7 : 15 : 1, while the molar ratio of PEG : PAMAM in the non-targeted nanoprobes was 14 : 1, implying an equivalent PEG moiety.

Fig. 1 ¹H NMR spectra of PAMAM (A), non-targeted (B), and targeted (C) nanoprobes acquired in D₂O at 500 mHz.

The labeled NIR dye Cy5.5 renders the targeted nanoprobe a characteristic absorption peak at 675 nm in the UV-vis spectrum (Fig. 2, solid lines) and a fluorescence emission peak at 700 nm in the fluorescent emission spectrum (Fig. 2, broken lines, excited at 630 nm). Moreover, it is worthwhile to note that the targeted nanoprobe could be well dispersed in water at concentration up to 100 μM and could remain stable for at least 1 month (inset, Fig. 2).

Fig. 2 UV-vis (solid lines) and fluorescence (broken lines) spectra of targeted nanoprobe. Inset, digital photo of the targeted nanoprobe dispersed in water.

Cytotoxicity

Prior to NIR fluorescent imaging, the cytotoxicity of both targeted and non-targeted nanoprobes was evaluated via MTT cell viability assay. For comparison, the cytotoxicity of PAMAM was also evaluated. As depicted in Fig. 3A, the cell viabilities were greater than 80% at all tested concentrations of both nanoprobes, indicating a low toxicity of nanoprobes towards TE-1 cells.

Fig. 3 (A) Cell viabilities of TE-1 cells 48 h after treatment with G5.0 PAMAM dendrimer, non-targeted, and targeted nanoprobes, respectively. (B) MFI of TE-1 cells incubated with non-targeted and targeted nanoprobes at 0.05 and 0.1 μM for 2 and 24 h, respectively. The data represent mean ± SD (n = 4). (*) p < 0.05, (**) p < 0.01, (***) p < 0.001. (C) CLSM images of the TE-1 cells treated with PBS, non-targeted, and targeted nanoprobes for 24 h, respectively. Cy5.5 is displayed in red and the nuclei stained with DAPI are shown in blue. Scale bar: 25 μm.

CLSM imaging

Next, the uptake of nanoprobes by TE-1 cells was examined by CLSM images. Obviously, the TE-1 cells treated with targeted nanoprobes emitted a marked higher fluorescence than that of TE-1 cells incubated with non-targeted nanoprobes, suggesting a greater cellular uptake (Fig. 3C). Moreover, it was noted that the uptake of nanoprobes by TE-1 cells exhibited a dose and time-dependent pattern. This was further supported by the mean fluorescence intensity (MFI) of labeled Cy5.5 in TE-1 cells measured by flow cytometry. As shown in Fig. 3B, no significant difference between the targeted and non-targeted nanoprobes during the first 2 h. However, with the lapsed time and increased concentration, the cellular uptake of targeted NIR nanoprobes became significantly greater than that of non-targeted ones. Notably, the MFI of targeted nanoprobes in TE-1 cells was 2.19 times higher than that of non-targeted nanoprobes at 24 h.

NIR fluorescence imaging of esophageal neoplasms

Mouse esophageal neoplasm model induced by 4-NQO was chosen for NIR imaging study because it is similar to the tumorigensis of human. By setting the threshold of 2 mm in size, the esophageal neoplasms subject to NIR fluorescence imaging were divided into large and tiny groups. The large esophageal neoplasm represents the existence of neoangiogenesis while the tiny one enables us to examine the applicability of targeted nanoprobes for the early detection of ESCC. To ensure the imaging was performed in the conditions simulating the living ones as much as possible, the mucosa of esophagus was imaged immediately after exposure. To identify anatomic localization of esophageal neoplasms, the white light (WL) images were first obtained. Nevertheless, it can be seen that WL images failed to provide adequate visual contrast for delineating the neoplasms (Fig. 4) due to the presence of mucosal erosion or inflammation nearby the neoplasms. In contrast, esophageal neoplasms with a clear margin could be easily identified on the NIR fluorescence images. In particular, the NIR fluorescence signal in the large neoplasm was more intensive when treated with the targeted nanoprobe, signifying the high specificity of RGD modified nanoprobes towards endothelial cells of esophageal neoplasm. Although this phenomenal became less distinctive in the tiny ones, it was adequate to detect the neoplasms with a clear margin, suggesting that the targeted NIR nanoprobe possess great potential not only for early detection of ESCC but also for ensuring a clear surgical margin for minimally invasive endoscopic surgery.

Fig. 4 Representative WL and NIR fluorescence images of esophagus bearing tiny (A) and large (B) esophageal neoplasms at 24 h after injection of non-targeted and targeted nanoprobes, respectively. (C) TBR values obtained on WL and NIR fluorescence images of esophageal neoplasms treated with non-targeted and targeted nanoprobes, respectively. The values represent mean ± SD (n = 4). (*) p < 0.05.

As the effectiveness of in vivo imaging is invariably defined by the ability of the probe to provide a high target to background ratio (TBR), we conducted the TBR analysis on both NIR and WL images. The TBR values on NIR images of huge neoplasm treated with non-targeted and targeted nanoprobes were 3.09 ± 0.18 and 5.09 ± 0.22, respectively. Both TBR values were pronounced higher than that obtained on corresponding WL images, which were 1.59 ± 0.06 and 1.69 ± 0.08, respectively. Similarly, the TBR values on NIR image of tiny neoplasms treated with non-targeted and targeted nanoprobes (2.53 ± 0.15 and 4.31 ± 0.31, respectively) were higher than that of the corresponding WL images (1.70 ± 0.08 and 1.03 ± 0.08, respectively) as well. More importantly, it was noted that the difference between non-targeted and targeted nanoprobes was statistical significance (p < 0.05). The enhanced TBR thus validates the effectiveness of the targeted nanoprobe for better detection of esophageal neoplasms.

Pathology

The esophageal neoplasms identified by NIR fluorescence imaging were excised and examined by pathology and H&E staining. It was confirmed that the pathological type of the tiny neoplasms is papilloma (Fig. 5A), while the pathological type of the tiny neoplasms is invasive squamous cell carcinoma with tumor cells invading submucosal of the esophagus (Fig. 5B). These two types of esophageal neoplasms well represented the precancerous lesion and early stage of ESCC, both are vital in esophagus screening for the detection of ESCC.

Fig. 5 Pathological evidence of carcinogenesis in mouse esophagus after 4-NQO treatment. (A) Papilloma found in the tiny esophageal neoplasms, and (B) invasive squamous cell carcinoma found in the large esophageal neoplasms. Scale bar: 50 μm.

CLSM and immunofluorescence imaging

Furthermore, the ex vivo CLSM images of esophageal sections bearing 4-NQO induced neoplasms were acquired 24 h post-injection of nanoprobes (Fig. 6A). According to the fluorescence intensity, it is obvious that the intratumoral cumulation of targeted nanoprobe was much higher than that of non-targeted nanoprobe, for either large or tiny esophageal neoplasms. Further quantitative analysis confirmed this observation. As shown in Fig. 6C, the fluorescence intensities in both tiny and large esophageal neoplasms treated with the targeted nanoprobe were 2.27 and 1.58 times higher than that treated with the non-targeted nanoprobe, respectively.

Fig. 6 (A) Representative CLSM images of 4-NQO induced tiny and large esophageal neoplasms sections at 24 h post-injection of nanoprobe (5.0 nmol per mouse). Cy5.5 is in red and the nuclei stained with DAPI appear blue. Scale bar: 100 μm. (B) Representative CLSM images of 4-NQO induced esophageal neoplasms sections at 24 h post-injection of targeted nanoprobe (5.0 nmol per mouse). The β3 integrin antibody immunofluorescence is in green. Scale bar: 100 μm. (C) Cellular fluorescence intensity at 24 h post-injection of the nanoprobe. The data were quantified by normalizing the Cy5.5 fluorescence with the number of nuclei in the indicated areas.

Moreover, the targeting specificity of the targeted nanoprobe to αvβ3 integrin was attested by immunofluorescence imaging (Fig. 6B). It is well known that the β3 integrin is predominately located in the tumor core area with high vasculature density. The red fluorescence representing the targeted nanoprobe co-localized well with the β3 integrin immunofluorescence (green) with a co-localization coefficient value as high as 0.77, signifying the specific targeting efficiency of RGD conjugated nanoprobes towards esophageal neoplasms.

Conclusion

In sum, we developed a targeted NIR nanoprobe by conjugation of Cy5.5 and RGD onto G5.0 PAMAM dendrimers. The successful conjugation of Cy5.5 and RGD onto PAMAM dendrimers was confirmed by ¹H NMR, FT-IR and UV-vis. The as-prepared nanoprobe emitted NIR fluorescence at 630 nm upon excitation at 680 nm and exhibited a low cytotoxicity towards TE-1 cells. Compared with the non-targeted counterpart, the targeted nanoprobe led to an improved targeting efficiency and consequently, an enhanced TBR of tiny esophageal neoplasms in mice, which was histologically confirmed as the early ESCC. Therefore, our targeted nanoprobe has the translatable potential for early detection of ESCC as well as for intra-operative optical-image-guided resection in combination with upper endoscopy.

Acknowledgements

The authors gratefully acknowledge the financial supports from National Natural Science Foundation of China (81271639, 81570507), Beijing Municipal Science and Technology Commission (Z151100003915097), The Capital Health Research and Development of special (2014-1-2021) and Capital Medical University Basic-Clinical Research Cooperation Fund (16JL46).

References

1. A. Pennathur, M. K. Gibson, B. A. Jobe and J. D. Luketich, Lancet, 2013, 381, 400–412.
2. J. Ferlay, I. Soerjomataram, R. Dikshit, S. Eser, C. Mathers, M. Rebelo, D. M. Parkin, D. Forman and F. Bray, Int. J. Cancer, 2015, 136, E359–E386.
3. P. C. Enzinger and R. J. Mayer, N. Engl. J. Med., 2003, 349, 2241–2252.
4. A. Pennathur, A. Farkas, A. M. Krasinskas, P. F. Ferson, W. E. Gooding, K. G. Michael, M. J. Schuchert, R. J. Landreneau and J. D. Luketich, Ann. Thorac. Surg., 2009, 87, 1048–1055.
5. P. Lao-Sirieix and R. C. Fitzgerald, Nat. Rev. Clin. Oncol., 2012, 9, 278–287.
6. J. Rao, A. Dragulescu-Andrasi and H. Yao, Curr. Opin. Biotechnol., 2007, 18, 17–25.
7. S. AnderssonEngels, C. afKlinteberg, K. Svanberg and S. Svanberg, Phys. Med. Biol., 1997, 42, 815–824.
8. G. Choy, P. Choyke and S. K. Libutti, Mol. Imaging, 2003, 2, 303–312.
9. R. Rossin, P. R. Verkerk, S. M. Bosch, R. C. M. Vulders, I. Verel, J. Lub and M. S. Robillard, Angew. Chem., Int. Ed., 2010, 49, 3375–3378.
10. S. A. Hilderbrand and R. Weissleder, Curr. Opin. Chem. Biol., 2010, 14, 71–79.
11. G. M. van Dam, G. Themelis, L. M. A. Crane, N. J. Harlaar, R. G. Pleijhuis, W. Kelder, A. Sarantopoulos, J. S. de Jong, H. J. G. Arts, A. G. J. van der Zee, J. Bart, P. S. Low and V. Ntziachristos, Nat. Med., 2011, 17, 1315–1319.
12. P. Sharma, S. Brown, G. Walter, S. Santra and B. Moudgil, Adv. Colloid Interface Sci., 2006, 123, 471–485.
13. X. He, K. Wang and Z. Cheng, Wiley Interdiscip. Rev.: Nanomed. Nanobiotechnol., 2010, 2, 349–366.
14. J. Yan, M. C. Estevez, J. E. Smith, K. Wang, X. He, L. Wang and W. Tan, Nano Today, 2007, 2, 44–50.
15. S. Gao, D. Chen, Q. Li, J. Ye, H. Jiang, C. Amatore and X. Wang, Sci. Rep., 2014, 4, 4384.
16. J. Liu, Z. Li, X. Yang, W. Liu, B. Wang, Y. Zhu, K. Mu and W. Zhu, Chem. Commun., 2015, 51, 13369–13372.
17. H. Wang, H. Wu, H. Shen, S. Geng, B. Wang, Y. Wang, X. Ma, G. Li and M. Tan, J. Mater. Chem. B, 2015, 3, 8832–8841.
18. Y. Yuan, J. Zhang, L. An, Q. Cao, Y. Deng and G. Liang, Biomaterials, 2014, 35, 7881–7886.
19. R. Tang, J. Xue, B. Xu, D. Shen, G. P. Sudlow and S. Achilefu, ACS Nano, 2015, 9, 220–230.
20. C. Shao, S. Li, W. Gu, N. Gong, J. Zhang, N. Chen, X. Shi and L. Ye, Anal. Chem., 2015, 87, 6251–6257.
21. P. Habibollahi, J. L. Figueiredo, P. Heidari, A. M. Dulak, Y. Imamura, A. J. Bass, S. Ogino, A. T. Chan and U. Mahmood, Theranostics, 2012, 2, 227–234.
22. P. Kesharwani and A. K. Lyer, Drug Discovery Today, 2015, 20, 536–547.
23. Z. Qiao and X. Shi, Prog. Polym. Sci., 2015, 44, 1–27.
24. P. Kesharwani, K. Jain and N. K. Jain, Prog. Polym. Sci., 2014, 39, 268–307.
25. D. Shcharbin, A. Shakhbazau and M. Bryszewska, Expert Opin. Drug Delivery, 2013, 10, 1687–1698.
26. X. H. Tang, B. Knudsen, D. Bemis, S. Tickoo and L. J. Gudas, Clin. Cancer Res., 2004, 10, 301–313.
27. L. Han, J. Li, S. Huang, R. Huang, S. Liu, X. Hu, P. Yi, D. Shan, X. Wang, H. Lei and C. Jiang, Biomaterials, 2011, 32, 2989–2998.
28. X. Wu, B. Ding, J. Gao, H. Wang, W. Fan, X. Wang, W. Zhang, X. Wang, L. Ye, M. Zhang, X. Ding, J. Liu, Q. Zhu and S. Gao, Int. J. Nanomed., 2011, 6, 1747–1756.

---

Footnote

† These authors contributed equally.

---