Characterization of the H₂N-PEEP-b-PBYP-hyd-DOX

In this study, a polymeric prodrug based on polyphosphoesters was prepared by the following steps and the synthesis routes were shown in Scheme 2. Firstly, the polyphosphoester containing EABoc fragment at one end was prepared via one-pot sequential feeding of EOP and BYP monomers by ROP, using 2-(tert-butoxycarbonylamino)-1-ethanol (EABoc) as initiator. Secondly, an amino-terminated polyphosphoester H₂N-PEEP-b-PBYP was obtained by hydrolysis. Then, DOX-hyd-N₃ was prepared and conjugated to the backbone via "click" chemistry between the alkynyl group of PBYP segment and azide group of DOX-hyd-N₃. Finally, we got the polymeric prodrug H₂N-PEEP-b-PBYP-hyd-DOX. The detailed experimental process is described in the electronic supplementary information (ESI).

2Full-screen View

Synthesis routes to polymeric prodrug H₂N-PEEP-b-PBYP-hyd-DOX.

In order to confirm the chemical structures and the molecular weights of the diblock copolymers, we have made the following characterizations: ³¹P-NMR, GPC, ¹H-NMR, FTIR, UV-Vis and HPLC analysis, respectively. The derivative DOX-hyd-N₃, monomers EOP and BYP were separately prepared. Figs. S1, S2 and S3 in ESI prove that the required DOX derivative and monomers have been synthesized. Fig. S4(A) (in ESI) is the ³¹P-NMR spectrum of the BocNH-PEEP product after 60 min of EOP polymerization, indicating that all EOP monomers have been polymerized completely. In Fig. S4(B) (in ESI), the appearance of the new peak assigned to the PBYP repeat units indicates the successful synthesis of BocNH-PEEP-b-PBYP. From the GPC traces (Fig. 1), we can further confirm the successful synthesis of the diblock copolymer due to the increase of molecular weight.

1Full-screen View

GPC traces of BocNH-PEEP₆₄ ($\overline M _{\rm{n}}$=1.25×10⁴ g·mol⁻¹, Ð=1.05) and BocNH-PEEP₆₄-b-PBYP₃₆ ($\overline M _{\rm{n}}$=1.69×10⁴ g·mol⁻¹, Ð=1.09), the number of repeating units is calculated by ¹H-NMR spectra.

In addition, the results of ¹H-NMR spectra in Fig. 2 indicate that we have obtained the precursors BocNH-PEEP, BocNH-PEEP-b-PBYP, H₂N-PEEP-b-PBYP and the prodrug H₂N-PEEP-b-PBYP-hyd-DOX, respectively. In Fig. 2(A), the chemical shift peaks at δ=1.43 ppm (peak a), δ=1.34 ppm (peak f), δ= 4.33−4.22 ppm (peak h) and δ=4.22−4.12 ppm (peak e) show that the polymer BocNH-PEEP was synthesized. In Fig. 2(B), the new characteristic peaks at δ=2.04 ppm (peak g) and δ=2.61 ppm (peak i) indicate that the diblock copolymer BocNH-PEEP-b-PBYP was obtained. In Fig. 2(C), the chemical shift at δ=8.59 ppm (peak k) indicates the appearance of the amino group at one end of the copolymer chain after hydrolysis. After click reaction between H₂N-PEEP-b-PBYP and DOX-hyd-N₃, the characteristic signal (peak q) that appears at δ=7.55 ppm in Fig. 2(D) prove that the prodrug H₂N-PEEP-b-PBYP-hyd-DOX have been prepared successfully. On the basis of ¹H-NMR spectrum in Fig. 2(B), the degree of polymerization of both PEEP (x) and PBYP (y) can be calculated by the following Eqs. (1) and (2):

2Full-screen View

¹H-NMR spectra of (A) BocNH-PEEP₆₄, (B) BocNH-PEEP₆₄-b-PBYP₃₆, (C) H₂N-PEEP₆₄-b-PBYP₃₆ and (D) H₂N-PEEP₆₄-b-PBYP₃₆-hyd-DOX.

1 $x{\rm{ = }}\frac{{3A_{\rm{f}}}}{{A_{\rm{a}}}}$

2 $y{\rm{ = }}\frac{{9A_{\rm{g}}}}{{A_{\rm{a}}}}$

where A_a is the integral area of the protons at δ=1.43 ppm in BocNH― (peak a, ―CH₃), A_f is the integral area of the protons at δ=1.34 ppm in PEEP chain segment (peak f, ―CH₃) and A_g is the integral area of the protons at δ=2.04 ppm in PBYP chain segment (peak g, ―CH₂C≡CH).

Furthermore, the molecular weights of BocNH-PEEP and BocNH-PEEP-b-PBYP could be calculated by Eqs. (3) and (4), in which 152.01 and 176.11 represent the molecular weights of one repeating unit of PEEP and PBYP, respectively, 144 is the molecular weight of EABoc. The data of molecular weights ($\overline M _{\rm{n}}$) and molecular weight distributions (Ð) of copolymers are listed in Table 1. Considering the experimental effect of assembly, we chose BocNH-PEEP₆₄-b-PBYP₃₆ for follow-up experiments.

1Full-screen View

Molecular weights and molecular weight distributions of BocNH-PEEP and BocNH-PEEP-b-PBYP.

Sample	${\overline M _{\rm{n}}}^{\rm{\;a}}$ (g·mol−1)	${\overline M _{\rm{n}}}^{\rm{\;b}}$ (g·mol−1)	${\overline M _{\rm{w}}}^{\rm{\;b}}$ (g·mol−1)	Ð b
a Calculated from 1H-NMR spectra. (solvent: CDCl3); b Measured by GPC. (eluent: DMF; standard: polystyrene).
BocNH-PEEP64	9700	12500	13300	1.05
BocNH-PEEP64-b-PBYP36	16000	16900	18400	1.09
BocNH-PEEP30-b-PBYP23	8600	7500	10600	1.41
BocNH-PEEP30-b-PBYP30	8500	9600	10800	1.13
BocNH-PEEP45-b-PBYP30	9100	9900	10900	1.10

Show more

3 $\overline M _{\rm{n (BocNH-PEEP)}} = 152.01x{\rm{ + 144}}$

4 $\overline M _{{\rm{n (BocNH-PEEP-}}b{\rm{-PBYP) }}}= \overline M _{\rm{n (BocNH-PEEP)}} + 176.11y$

FTIR, UV-Vis and HPLC were also measured to prove the successful preparation of the prodrug H₂N-PEEP-b-PBYP-hyd-DOX. The characteristic absorption at 2105 cm⁻¹ (ν_−C≡CH) in Fig. S5(B) (in ESI) indicate the successful synthesis of the copolymer BocNH-PEEP-b-PBYP. The characteristic absorption at 3280 cm⁻¹ ($ν_{-{\rm{NH}}_2}$) and 1673 cm⁻¹ ($ν_{-{\rm{NH}}_2}$) in Fig. S5(C) (in ESI) confirm the amino group at the end of the copolymer was obtained. The disappearance of the characteristic absorption at 2105 cm⁻¹ (ν_−C≡CH) shows that the prodrug H₂N-PEEP-b-PBYP-hyd-DOX has been synthesized. What's more, as shown in the Fig. S6 (in ESI), there is a significant red shift between the H₂N-PEEP-b-PBYP-hyd-DOX and DOX in the UV-Vis spectrum. Therefore, it can illustrate that DOX has been bound to the side groups of the copolymer. DOX, DOX-hyd-N₃ and H₂N-PEEP-b-PBYP-hyd-DOX were measured by HPLC and showed different retention time (Fig. S7 in ESI), thus proving that the required prodrugs have been synthesized.

Self-Assembly Behavior of H₂N-PEEP-b-PBYP-hyd-DOX

The prodrug H₂N-PEEP-b-PBYP-hyd-DOX is an amphiphilic copolymer, which can self-assemble into nanoparticles (abbreviated as PDNPs) with the PBYP chain segment and the hydrophobic DOX as the core, and the PEEP chain segment as the shell in aqueous solution. The critical aggregation concentration (CAC) can be used to evaluate the self-assembly capability of amphiphilic polymers.^[45] In general, the low CAC value means that the copolymer can form nanoparticles at low concentration. We used the pyrene fluorescence probe method for measuring the fluorescence intensity of a series of concentrations of polymer solutions. Fig. S8 (in ESI) shows the relationship between the logarithm of the concentration of the copolymer solution (logC) and the ratio of the fluorescence intensity of the third peak to the first peak (I₃/I₁), and the CAC value (0.096 mg·mL⁻¹) of H₂N-PEEP-b-PBYP-hyd-DOX is obtained by intersecting the two straight lines.

The average particle size ($\overline D _{\rm{Z}}$) and size polydispersity index (size PDI) are two key parameters for the delivery of polymer prodrug. It has been reported that the nanoparticles in the range of 50−200 nm can be transported to the tumor tissue effectively.^[46] In this study, we characterized the particle sizes and the morphologies by DLS and TEM measurements. From Fig. 3(a) we can get the result that the prodrug H₂N-PEEP-b-PBYP-hyd-DOX can self-assemble into nanoparticles with uniform particle size in aqueous solution. The average particle size of PDNPs is about 70 nm. Fig. 3(b) shows the particle size distribution histogram of H₂N-PEEP-b-PBYP-hyd-DOX (0.5 mg·mL⁻¹) with particle size of 93 nm and size PDI of 0.092. The average particle size measured by DLS is larger than that observed by TEM, which is mainly because the average particle size of nanoparticles with hydrophilic chains is different in the aqueous solution and the dry environment. After the surface of polymer prodrug nanoparticles was combined with monoclonal antibody, namely mAb-PDNPs, the particle morphology and particle size distribution histogram are shown in Figs. 3(c) and 3(d). The slight increase in particle size is due to the binding of CD147 monoclonal antibodies to the surface of PDNPs, forming mAb-PDNPs. When the PDNPs were exposed to the environment of PB 5.0 buffer for 48 h, the original self-assembled nanoparticles almost dissociated, and no spherical particles were observed, as shown in Fig. 3(e). And the corresponding particle size distribution becomes wider (Fig. 3f), which means that the polymeric prodrug has acid-sensitiveness.

3Full-screen View

TEM images of the nanoparticles: (a) PDNPs in PB 7.4 after 48 h; (c) mAb-PDNPs in PB 7.4 and (e) PDNPs in PB 5.0 after 48 h. (b, d and f) Particle-size distribution histograms of the samples in panels (a), (c) and (e), respectively. All the micellar concentrations were kept at 0.5 mg·mL⁻¹; the scale bars of images are 200 nm.

Characterization of CD147-PEEP-b-PBYP-hyd-DOX

To confirm whether the CD147 mAb was successfully bound to PDNPs, XPS was measured for elemental analysis of the two samples, PDNPs and mAb-PDNPs, as shown in Fig. S9 (in ESI). Since PDNPs do not contain sulfur, and antibody contains disulfide bond, the sulfur peak in the XPS spectrum (Fig. S9B in ESI) indirectly proves that CD147 mAb has been attached to the surface of prodrug nanoparticles PDNPs.

In addition, in the FTIR spectum of Fig. S10 (in ESI), the migration of the carbonyl absorption peak from 1665 cm⁻¹ to 1638 cm⁻¹ can also account for the coupling of CD147 mAb to PDNPs. To further test the modification rate of CD147 mAb, different batches of samples were tested by BCA protein kit to determine the modification rate of CD147 mAb and the results are shown in Table 2. By measuring the zeta potential of the particle, it can also be proved that the antibody binds to the prodrug nanoparticle. Due to the amino groups on the surface of the prodrug nanoparticles, the zeta potential of PDNPs was +7.3 mV. When the monoclonal antibody CD147 mAb was linked to PDNPs, the amino group on the surface of the PDNPs will be consumed, the zeta potential would gradually decrease. Considering the experimental effect and cost, the sample with a mass ratio of CD147 mAb to PDNPs of 4 (CD147 mAb : PDNPs, μg/mg) was chosen for the subsequent experiments.

Enzymatic Degradation

In the presence of phosphodiesterase I (PDE I) and PB 7.4 buffer solution, a degradation experiment on the diblock copolymer H₂N-PEEP₆₄-b-PBYP₃₆ was performed. The process of degradation was monitored by ¹H-NMR spectroscopy. As shown in Fig. 4, we could get the results that the original peaks assigned to protons of the PEEP and PBYP blocks gradually weakened or disappeared along with the incubating time from 0 h to 72 h, and several new signals appeared at δ=0.83 ppm, δ=1.25 ppm, and δ=1.33 ppm, respectively. Therefore, the polyphosphoesters as the main chain have good biodegradability.

4Full-screen View

¹H-NMR spectra of (A) H₂N-PEEP₆₄-b-PBYP₃₆ and its degradation products at predetermined time intervals for (B) 24 h, (C) 48 h and (D) 72 h, respectively. (Solvent: CDCl₃).

Particle Stability and pH Sensitivity

The nanoparticles PDNPs formed by self-assembly of amphiphilic polymer prodrug in water should remain stable in the normal in vivo environment. When entering the tumor cell microenvironment, due to the internal weak acidity, the hydrazone bond is broken, the nanoparticles would dissociated and DOX be released.^[38] These processes can be confirmed by observing the changes of particle size ($\overline D {{z}}$) and size PDI of the polymer prodrug nanoparticles under specific conditions. As shown in Figs. 5(a) and 5(b), there were inconspicuous changes of the particle size over 48 h, that can illustrate the favorable stability of PDNPs under neutral conditions. However, under the condition of PB 6.0 and PB 5.0 buffer, the particle size and size PDI of PDNPs gradually increased with time, which indicate that PDNPs will disintegrate through hydrazone bond in weakly acidic medium and effectively release drug. At the same time, combined with the results of the enzymatic degradation of the polyphosphoester backbond, we can believe that the prepared polymeric prodrug nanoparticles have good pH-responsiveness and drug release performance in the tumor microenvironment.

5Full-screen View

Particle size distribution histograms of PDNPs (0.5 mg·mL⁻¹) in different pH media: (a) PB 7.4, (b) PB 6.8, (c) PB 6.0 and (d) PB 5.0, as determined by DLS.

In Vitro Release of DOX

In recent years, pH-responsive polymers are playing an increasingly important role in drug delivery system.^[47−49] In this study, as a performance acid-sensitive bond, the hydrazone bond will ensure efficient release of the DOX under acidic conditions while maintaining equilibrium under physiological conditions. The cumulative DOX release experiments in vitro were performed in different conditions: (A) PB 5.0, (B) PB 6.0, (C) PB 7.4 with PDE I, (E) PB 7.4 for PNDPs, and (D) PB 7.4 for mAb-PDNPs, respectively, to investigate the release of DOX, as shown in Fig. 6. We can see that after 50 h of incubation in thermostatic oscillator, the cumulative release of DOX was up to 65% for PDNPs under the condition of PB 5.0 buffer (Curve A), while 10% of DOX release for both PDNPs and mAb-PDNPs after 50 h of incubation in PB 7.4 buffer, indicating that mAb-PDNPs have similar stability to PDNPs in normal physiological environment. By comparing Curve D with Curve E, we can know that the modification of CD147 mAb on the surface of prodrug nanoparticles has little effect on the release of DOX. Furthermore, in the presence of PDE I, PDNPs would dissociate and release DOX, as shown in Curve C. These results indicate that the prodrug nanoparticles PDNPs can release DOX by dissociation under acidic conditions and the polyphosphoesters have good biodegradability.

6Full-screen View

In vitro DOX release curves for (D) mAb-PDNPs in PB 7.4 buffer solution, and PDNPs under different conditions: (A) PB 5.0, (B) PB 6.0, (C) PB 7.4 buffer solution with PDE I, and (E) PB 7.4, respectively. All the sample concentrations were kept at 0.5 mg·mL⁻¹ and the temperature at 37 °C.

Biocompatibility

The biocompatibility of materials is one of the important bases to evaluate whether they can be used as drug carriers in drug delivery system. In this study, methyl thiazolyl tetrazolium (MTT) assays were used to study the cytotoxicity of the polymer H₂N-PEEP-b-PBYP against human umbilical vein endothelial cells (HUVEC cells), mouse fibroblast cells (L929 cells) and human hepatocellular carcinomas cells (HepG2 cells), respectively. These cells were incubated with different concentrations of polymer solution for 48 h. Fig. 7 illustrates that the viability of the cells are all over 82% no matter the normal cells or the hepatoma carcinoma cells. Therefore, the copolymer H₂N-PEEP-b-PBYP has low toxicity to cells and has good biocompatibility. The above results show that polyphosphoester has promising application for drug carriers.

7Full-screen View

Cell viability of HUVEC cells, HepG2 cells and L929 cells with H₂N-PEEP₆₄-b-PBYP₃₆ of different concentrations for 48 h incubation.

In Vitro Cytotoxicity

MTT assays were also used to study the cytotoxicity of PDNPs and mAb-PDNPs against HepG2 cells, human cervical cancer cells (HeLa cells) and HUVEC cells, respectively, in which the two later cells have no specific binding sites and can be used as the controls. As shown in Fig. 8(a), after culture HUVEC cells with mAb-PDNPs and PDNPs for 48 h, the cell survival rate was higher, indicating that both mAb-PDNPs and PDNPs hardly released drugs in human healthy cells. In Figs. 8(b) and 8(c), the concentration dependence of the three groups of nanoparticles was observed, that is, the cell viability decreased gradually with the increase of the concentration of nanoparticles. The values of the half maximal inhibitory concentration (IC₅₀) are listed in Table 3, which can be used to evaluate the anti-tumor effect of the nanoparticles. In Fig. 8(b), the IC₅₀ values of free DOX, mAb-PDNPs and PDNPs are 0.19, 1.12 and 2.67 mg·L⁻¹, respectively. The mAb-PDNPs show a lower IC₅₀ value compared with PDNPs. This is due to the fact that mAb-PDNPs can specifically bind to the antigens on the surface of HepG2 cells and promote the enrichment of nanoparticles. On the other hand, compared with the other two groups, free DOX has the lowest IC₅₀ value, indicating the highest toxicity. This may be due to the use of polyphosphoester as a drug carrier to reduce the toxicity of the active drug. What is more, since DOX was chemically bound to the polymer, the structure was stable and the bond need to be broken to release DOX into the tumor microenvironment, which takes more time. In Fig. 8(c), the IC₅₀ values of PDNPs and mAb-PDNPs are 14.25 and 13.06 mg·L⁻¹, which are similar, because there is no antigen on the surface of HeLa cells that can be specifically recognized with CD147 monoclonal antibody. And the IC₅₀ value of free DOX is the lowest of 4.75 mg·L⁻¹, which is similar to that of the case of HepG2 cells in Fig. 8(b). The above results can indicate mAb-PDNPs can realize targeted drug delivery and polyphosphoester as drug carrier can reduce the toxicity of active drug.

8Full-screen View

Cell viabilities of (a) HUVEC cells, (b) HepG2 cells and (c) HeLa cells with mAb-PDNPs, PDNPs and free DOX of different concentrations for 48 h incubation.

Cellular Uptake

It is also important to consider whether the drug-loaded nanoparticles can recognize tumor cells and be internalized by the cells to effectively release the drugs in the tumor cells.^[50] Therefore, HepG2 cells and HeLa cells were used to evaluate the cellular uptake for PDNPs and mAb-PDNPs, respectively. A live cell imaging system was used to monitor the uptake process in real time, as shown in Fig. 9. Comparing Fig. 9(a) with Fig. 9(c), DOX fluorescence intensity of mAb-PDNPs is stronger than that of PDNPs in the same period, this is because mAb-PDNPs act on HepG2 cells surface by active targeting, while PDNPs enter the cells by cellular uptake. In addition, as shown in Fig. 9(b), the fluorescence intensity of free DOX is much smaller, because DOX enters the cells through diffusion by virtue of the concentration difference between inside and outside the cells. Besides, by comparing Fig. 9(c) with Fig. 9(f), the fluorescence intensity of HepG2 cells is stronger than that of HeLa cells, because there are antigenic sites on the surface of HepG2 cells that can be targeted and recognized by CD147 monoclonal antibodies. These results show that both nanoparticles could effectively release drug in tumor cells, and mAb-PDNPs possess good targeting ability to HepG2 cells.

9Full-screen View

Live cell imaging system images: HepG2 cells incubated with (a) PDNPs, (b) free DOX and (c) mAb-PDNPs, respectively; HeLa cells incubated with (d) PDNPs, (e) free DOX and (f) mAb-PDNPs, respectively, at different time. For all test, DOX dosage was maintained at 25 mg·L⁻¹. For each group, images from left to right show the cell nuclei stained with H 33342 (blue), the DOX fluorescence in cells (red), and overlays of the two images. All the scale bars of images are 20 μm.