In the study, we first synthesized a Bi(PEG-PLA)-Pt(IV) polymer-cisplatin prodrug conjugate following the protocol illustrated in . Briefly, carboxyl-functionalized poly(ethylene glycol) (COOH-PEG-OH, Mn = 3500 gmol−1) was used as macroinitiator for the ring opening polymerization of l-lactide in the presence of stannous octoate as a catalyst. The resultant poly(ethylene glycol)-b-poly(l-lactide) (PEG-PLA) was characterized by gel permeation chromatography (GPC) with a molecular weight of 10000 gmol−1 and a polydispersity index of 1.12. Then the PEG-PLA was activated by 4-nitrophenyl chloroformate and derivatized to its hydrazine derivative (PEG-PLA-NH-NH2). Subsequently, this hydrazine terminated polymer was reacted with pre-synthesized cis,trans,cis-PtCl2(OCOCH2CH2COCH3)2 (NH3)2 cisplatin analogue prodrug, in which levulinic acid was used as a spacer, resulting in a Bi(PEG-PLA)-Pt(IV).
Schematic description of the synthesis of Bi(PEG-PLA)-Pt(IV) polymer-prodrug conjugate.
The entire reaction processes were monitored by 1
H-NMR spectroscopy and GPC. The 1
H NMR spectra of all intermediate products were provided in the Supporting Information (Figure S1–S4)
. As indexed in , the 1
H-NMR spectrum of the final polymer-cisplatin prodrug conjugate included all characteristic resonance peaks of the PEG-PLA polymer, the levulinic acid spacer, and the Pt(IV) prodrug, for example, the characteristic peaks of −NH-N= at δ 8.28 ppm, −CH2
of levulinic acid at δ 2.1 and 2.95 ppm, and −NH3
of cisplatin at δ 8.0 ppm. A considerable 1
H NMR resonance shift was observed for the −CH2
of levulinic acid and the −NH3
of cisplatin in the final product as compared to those in PtCl2
). This is likely due to the presence of two polyester chains that sandwich the Pt(IV) metal from the axial position in the Bi(PEG-PLA)-Pt(IV) conjugate. Such conformation creates local magnetic field inhomogeneity accounting for the resonance shift. In addition, the quadrupolar effect of the 14
N nucleus may also contribute to the resonance shift because it makes protons resonating in a broad spectrum.14
The formation of Bi(PEG-PLA)-Pt(IV) conjugate was further confirmed by the GPC measurements. As shown in , the characteristic peak of PEG-PLA at 6.6 min disappeared in the chromatogram of the polymer-cisplatin prodrug conjugate. Instead, a dominant peak appeared at the retention time of 5.5 min, corresponding to the molecular weight of about 20000 gmol−1
. This clearly indicates the formation of Bi(PEG-PLA)-Pt(IV) conjugate, which contains two PEG-PLA polymer chains.
Characterization of Bi(PEG-PLA)-Pt(IV) polymer-prodrug conjugate. (A) 1H-NMR spectrum of the synthesized Bi(PEG-PLA)-Pt(IV) conjugate. (B) GPC chromatogram of PEG-PLA polymer and Bi(PEG-PLA)-Pt (IV) conjugate.
Next we measured the critical micellar concentration (CMC) of the synthesized Bi(PEG-PLA)-Pt(IV) conjugate to evaluate their feasibility of forming NPs. CMC was determined using pyrene as a hydrophobic probe that has been widely used for this purpose because of its characteristic fluorescence spectra sensitive to environmental polarity.22
The fluorescence emission of pyrene was fixed at 390 nm while its excitation spectra were monitored at various concentrations of Bi(PEG-PLA)-Pt(IV) conjugate. The intensity ratio at 332 nm and 329 nm was plotted against polymer-cisplatin prodrug concentration in a semi-log graph. As shown in , the CMC of the Bi(PEG-PLA)-Pt(IV) conjugate was 3.6±0.25 mg/L. This very low CMC value indicates that the conjugate is prone to form NPs via precipitation method. Indeed, dynamic light scattering (DLS) measurement showed that the self-assembled Bi(PEG-PLA)-Pt(IV) conjugate NPs had an unimodal size distribution with an averaged diameter of 86±2.0 nm ( inset), which was consistent with the findings from SEM images (). The surface zeta potential of the NPs was about −33±1.2 mV ( inset). We further found that the size and surface zeta potential of the polymer-prodrug conjugate NPs were similar to those of the corresponding PEG-PLA polymeric NPs, 83±2.0 nm and −36±2.0 mV, respectively. This suggests that conjugation of cisplatin prodrug to the PEG-PLA polymer chain has negligible effect on formation of the polymeric NPs.
Figure 4 (A) Determination of the critical micelle concentration (CMC) of Bi(PEG-PLA)-Pt(IV) conjugate using pyrene probe. Inset: NP size, surface zeta potential of Bi(PEG-PLA)-Pt(IV) NPs and PEG-PLA NPs measured by DLS. (B) Representative SEM image of Bi(PEG-PLA)-Pt(IV) (more ...)
After having prepared Bi(PEG-PLA)-Pt(IV) NPs, we then quantified cisplatin loading yield of the NPs and cisplatin release kinetics from the NPs at different pH values using inductively coupled plasma optical emission spectrometry (ICP-OES). Here the drug loading yield is defined as the weight ratio of the cisplatin payload to the NPs including both polymer excipients and cisplatin. The drug release kinetics represents how fast the drugs leak out of the NPs, plotting as the weight ratio of the accumulative released cisplatin to the total cisplatin payload against time. As shown in , when more Pt(IV) prodrugs were used to react with PEG-PLA during the polymer-prodrug conjugation process, a higher cisplatin drug loading yield was achieved. For example, the initial Pt(IV)/PEG-PLA reaction molar ratio of 2:1, 4:1, and 6:1 resulted in a final cisplatin drug loading yield of 0.35±0.01 wt%, 0.89±0.02 wt%, and 1.05±0.03 wt% (mean±SD, n=3), respectively. However, when the Pt(IV)/PEG-PLA molar ratio was higher than 6:1 (e.g., 8:1), no considerable drug loading yield increase occurred. This is likely due to the saturation of polymer chains, in which all PEG-PLA polymers have been conjugated with Pt(IV) prodrugs. showed the cisplatin release kinetics from the Bi(PEG-PLA)-Pt(IV) NPs at three distinct pH values, pH=5.0, 6.0, and 7.4, respectively. The cisplatin release rate from the NPs at pH=5.0 and 6.0 was significantly faster than at pH=7.4. When the cisplatin loading yield was 1.05 wt% (), it took the Bi(PEG-PLA)-Pt(IV) NPs around 4 hrs and 6 hrs to release 50% of total cisplatin payload at pH=5.0 and 6.0, respectively, versus
22 hrs at pH=7.4. The contrast of the cisplatin release rate was even more sharply within the first a few hrs. For example, during the first 2 hrs period, 17% and 15% of the cisplatin payload was released at pH=5.0 and 6.0, respectively, while only 2% was released at pH=7.4. These results suggest that cisplatin release kinetics from the Bi(PEG-PLA)-Pt(IV) NPs is pH-dependent. This is mainly because the cisplatin analogue Pt(IV) prodrugs were covalently conjugated to the polymer chains through hydrazone bond, which is an acid-labile bond. At pH=5~6, hydrazone bond can be easily cleaved within a few minutes to free the drugs which will diffuse out of the NPs. In contrast, this bond is relatively stable at pH=7.4.23
The observed sustained cisplatin release at pH=7.4 may be due to the degradation of the PLA polymers, to which the cisplatin analogue prodrugs were covalently linked. As a biodegradable polymer, PLA ester can be hydrolyzed to small segments or monomers at both neutral pH and acidic pH.24
Here we incubated the PEG-PLA NPs in pH=5.0, 6.0, and 7.4 PBS solutions at 37 °C, respectively. At each time point, an aliquot of the PEG-PLA NPs were collected to measure the polymer molecular weight (Mw
) using GPC. As shown in inset, after 50 hrs incubation, the polymer Mw
decreased by a factor of 20%, 18%, and 8 % at pH=5.0, pH=6.0, and pH=7.4, respectively. This data reasonably explains the sustained drug release kinetics at neutral pH as shown in but also raises a concern that the observed rapid drug release at pH=5.0 and 6.0 might be because of fast PLA degradation at acidic pH rather than the cleavage of hydrazone bond. However, negligible difference of Mw
loss was observed within the first 24 hrs of incubation at these three pH values. This confirms that the drug burst at pH=5.0 and 6.0 during the first a few hrs is due to the cleavage of the hydrazone bond but not polymer degradation.
Figure 5 (A) Cisplatin loading yield of Bi(PEG-PLA)-Pt(IV) NPs at various initial Pt(IV)/PEG-PLA reaction molar ratios. (B) Cisplatin release profile from Bi(PEG-PLA)-Pt(IV) NPs at pH=5.0 (open circles), pH=6.0 (open triangles), and pH=7.4 (solid circles) PBS (more ...)
Lastly, we examined the in vitro
cellular cytotoxicity of the synthesized acid-responsive Bi(PEG-PLA)-Pt(IV) NPs. To this end, we chose A2780 human ovarian carcinoma cell line as a model cancer cell because of the well-known toxicity of cisplatin against ovarian cancer. Following a well established cellular cytotoxicity measurement protocol,25, 26
the A2780 cells were incubated with Bi(PEG-PLA)-Pt(IV) NPs for 4 hrs. After the incubation, the excess NPs were removed and the cells were washed three times with fresh buffer followed by the addition of fresh culture media. Subsequently the cells were incubated for 72 hrs before being assessed by MTT assay. Cell culture media and PEG-PLA NPs (without cisplatin analogue prodrugs) were used as negative controls. Free cisplatin drug at different concentrations (10 μM, 50 μM, and 100 μM respectively) served as positive controls. As shown in , the cell viability of the Bi(PEG-PLA)-Pt(IV) NPs decreased to about 65% after 4 hrs incubation. In contrast, PEG-PLA NPs had negligible cytotoxicity against ovarian cancer cells, similar as the cell media. The cell viability of free cisplatin at 10 μM, 50 μM, and 100 was 99%, 62%, and 35%, respectively. Based on the cisplatin loading yield of 1.05 wt% measured in , we calculated that the amount of cisplatin loaded in the Bi(PEG-PLA)-Pt(IV) NPs for this cytotoxicity study was equivalent to 7 μM free cisplatin. Surprisingly, the NPs with an equivalent 7 μM free cisplatin had cellular cytotixicity against ovarian cancer cells as high as 50 μM free cisplatin. To ensure that the measured cytotoxicity was due to the internalized NPs but not the free drugs in the media released from the NPs, the culture media were filtered through a membrane with a molecular weight cut-off of 10 KDa after the 4 hrs incubation with Bi(PEG-PLA)-Pt(IV) NPs. The filtrate was collected to quantify Pt content using ICP-OES. Negligible amount of free Pt drug (0.05 μM) was observed in the media, which was consistent with the slow drug release profile of the NPs at pH=7.4. This approximate 7 fold cytotoxicity increase of Bi(PEG-PLA)-Pt(IV) NPs might be attributed to the burst drug release in the acidic intracellular environment. Upon internalization, the acid-responsive NPs caused a surge in intracellular drug concentration that possibly overwhelmed some chemoresistance mechanisms of tumor cells, such as the P-glycoprotein (P-gp) membrane proteins mediated drug efflux mechanism.27
Our results are consistent with an earlier study by Xu et al., who have examined the activity of cisplatin encapsulated in different NP systems and have observed enhanced cytotoxicity of those NPs with fast cisplatin release profile. 28
Figure 6 MTT assay to measure the cytotoxicity of Bi(PEG-PLA)-Pt(IV) NPs against A2780 human ovarian carcinoma cell line in comparison with cell culture media, PEG-PLA NPs and free cisplatin (10 μM, 50 μM and 100 μM). The amount of cisplatin (more ...)