Fully synthetic macromolecular prodrug chemotherapeutics with EGFR targeting and controlled camptothecin release kinetics

Abstract

Herein, we developed a fully polymerizable, peptide-targeted, camptothecin polymeric prodrug system. Two prodrug monomers were synthesized via esterification of campothecin (20Cam) and 10-hydroxycamptothecin (10Cam) with mono-2-(methacryloyloxy)ethyl succinate (SMA) resulting in polymerizable forms of the aliphatic ester- and aromatic ester-linked drugs respectively. These monomers were then incorporated into zwitterionic polymers via RAFT copolymerization of the prodrug monomers with a tert-butyl ester protected carboxy betaine monomer. Subsequent deprotection of the tert-butyl residues with TFA yielded carboxy betaine methacrylate (CBM) scaffolds with controlled prodrug incorporation. Reverse phase HPLC was then employed to establish drug release kinetics in human serum at 37 ^oC for the resultant polymeric prodrugs. Copolymers containing 10Cam residues linked via aromatic esters showed faster hydrolysis rates with 59 % drug released at 7 days, while copolymers with Cam residues linked via aliphatic esters showed only 28 % drug release over the same time period. These differences in drug release kinetics were then shown to correlate with large differences in cytotoxic activity in SKOV3 ovarian cancer cell cultures. At 72 hours, the IC50s of aromatic- and aliphatic- ester linked prodrugs were 56 nM and 4776 nM, respectively. An EGFR-targeting peptide sequence, GE11, was then directly incorporated into the polymeric prodrugs via RAFT copolymerization of the polymeric prodrugs with a peptide macronomer. The GE11-targeted polymeric prodrugs showed enhanced targeting and cytotoxic activity in SKOV3 cell cultures relative to untargeted polymers containing the negative control sequence HW12. Following pulse-chase treatment (15 min, 37 °C), the 72 hour IC50 of GE11 targeted prodrug was determined to be 1597 nM, in contrast to 3399 nM for the non-targeted control.

Graphical Abstract

The development of fully polymerizable, peptide-targeted, camptothecin polymeric prodrugs are reported.

A. Introduction

Cancer remains a leading cause of morbidity and mortality worldwide. The American Cancer Society projects that in 2018 approximately 609,640 Americans will die of cancer.¹ While chemotherapeutic agents such as camptothecin are highly cytotoxic, their therapeutic efficacy is often limited by off-target side effects in healthy tissues; low therapeutic windows prevent the administration of drug doses that achieve therapeutic concentrations in cancer tissues.²

A promising strategy for addressing this problem is the development of polymeric prodrugs in which the therapeutic agent is covalently linked to a hydrophilic, macromolecular scaffold via a hydrolytic or enzymatically degradable linkage.^3–6 Advantages offered by this strategy include increased drug solubility and stability, increased circulation half-lives, reduced immunogenicity, and passive targeting to cancer tissues as a result of the enhanced permeability and retention (EPR) phenomenon.^7–9 Furthermore, cancer-targeting ligands such as antibodies, peptides, carbohydrates, and vitamins can be incorporated into this drug delivery system to further enhance tumor specificity.^10–12 In particular, antibody-drug conjugates, in which a cytotoxic agent is coupled to a cancer-specific antibody, have already demonstrated clinical success in some lymphomas¹³ and breast cancers.^14,15 However, peptide-based targeting ligands promise a number of advantages over antibodies in drug delivery. Peptides typically show lower immunogenicity relative to proteins, are inexpensive to produce at multi-ton scales, and can be incorporated at defined conjugation sites into a diverse range of drug delivery systems.¹⁶ Additionally, sequences targeting a wide range of receptors can be readily developed using phage display¹⁷, and peptides are approximately two orders of magnitude smaller than antibodies allowing them to more readily penetrate into the high-pressure interstitial tumor environment.¹⁸ However, peptides also typically display significantly lower binding affinities and the naturally occurring l-enantiomers usually exhibit low stability in biological systems. By polymerizing them into multivalent displays in a zwitterionic matrix, these issues may be minimized, and it is also possible to create equivalent non-natural peptide sequences that maintain affinity and stability.^19–21

One target of great interest in cancer is the epidermal growth factor receptor (EGFR), which is a tyrosine kinase receptor that is known to promote cell proliferation, differentiation, migration, and inhibit apoptosis.²² EGFR is overexpressed in a wide array of epithelial cancers making it a promising target for preferential tumor delivery.²³ The EGFR receptor has also been shown to constitutively internalize via receptor-mediated endocytosis making it an attractive target for peptide-mediated drug delivery.²⁴ Despite these advantages, native EGFR has strong mitogenic and neoangiogenic activity that would be counterproductive for an anticancer therapeutic.²² In order to circumvent these limitations, Lie et. al. employed in vitro phage display to identify a novel peptide, GE11, that binds to EGFR with high affinity (K_d ~22 nM).²⁵ This sequence is internalized by EGFR-overexpressing cell lines and displays negligible mitogenic activity. Conjugation of GE11 to both poly(PEI) and liposomal drug carriers has been shown to enhance therapeutic delivery in EGFR-overexpressing tumor cells both in vitro and in tumor xenograft models.^25,26 Taken together, these findings suggest GE11 as an inexpensive alternative for targeting cytotoxic drugs to EGFR positive cancers.

An attractive route for the synthesis of peptide-targeted polymeric prodrugs is the direct reversible deactivation radical polymerization (RDRP) of therapeutic agents that have been reversibly modified with suitable vinyl functionality.^27–32 This strategy allows for one or more drug classes to be incorporated into the final polymer at predetermined ratios without the need for additional conjugation and purification steps. Recently, we employed this approach to synthesize methacrylate-based polymeric prodrugs derived from the antibiotic ciprofloxacin.^29,33 Drug release studies conducted in human serum showed that phenyl ester-linked antibiotics were cleaved from the polymer scaffold at significantly higher rates relative to aliphatic ester linked ciprofloxacin. These differences in the relative antibiotic release rates were found to strongly influence the antimicrobial activity of the polymeric prodrug with ciprofloxacin linked via phenyl esters showing significantly lower minimum inhibitory concentrations than the aliphatic ester linkage.

Based on these findings we decided to investigate the relative drug release rates and chemotherapeutic activity of polymeric prodrugs based on camptothecin (20Cam) and 10-hydroxy camptothecin (10Cam). Previously we have observed that polymeric prodrugs prepared from the copolymerization of aliphatic ester linked 20Cam with the hydrophilic comonomer poly(ethylene glycol) methyl ether methacrylate (FW_avg ~950 Da) (O950) showed slow drug release rates with a t_1/2 of ~ 20 days.³⁴ In contrast esterification of 10Cam at position 10 on the pyrrolo[3,4-β]-quinoline ring yields an aromatic ester that could be expected to show more rapid drug release from the polymer scaffold. Indeed polymeric prodrugs based on campothecin linked at the 10-position have recently been shown to provide prolonged long-acting anticancer activity both in vitro and in vivo.^35–37 For example, Zhu et. al. recently showed that polymeric micelles derived from poly(2-methylacryloyloxyethyl phosphorylcholine)-b-poly(10-hydroxy-camptothecin methacrylate) (pMPC-b-pHCPT) displayed long circulation times and high levels of tumor growth inhibition in HeLa tumor-bearing nude mice.³⁵

Building on this work, herein we detail the development of a fully polymerizable, GE11-targeted, camptothecin polymeric prodrug system. Cancer-specific peptide targeting was incorporated into the polymeric prodrugs using peptide macro-monomer technology previously developed for the intracellular delivery of biologic drugs. ^38,39 First, two camptothecin polymeric prodrugs were synthesized with either an aliphatic ester- or aromatic ester-linkage and their drug release kinetics and corresponding cytotoxic activities were compared. Next, a GE11-targeting peptide monomer was directly incorporated into the polymeric prodrug via block copolymerization to enhance drug uptake and cytotoxic activity in SKOV3 ovarian cancer cells.

B. Results and Discussion

Synthesis of zwitterionic polymeric prodrugs containing aliphatic ester-linked (20Cam) or aromatic ester-linked (10Cam) camptothecin.

Polymerizable prodrug monomers provide a facile route by which therapeutic agents may be integrated into a hydrophilic polymer scaffold without the need for costly and often ill-defined post polymerization conjugation steps. Shown in Scheme 1, is the synthetic strategy for the preparation of zwitterionic carboxybetaine (CBM)-based copolymers with Cam linked to the polymer backbone via aliphatic or aromatic ester groups. Both Cam prodrug monomers were first synthesized by conjugating the carboxylic acid of mono-2-(methacryloyloxy)ethyl succinate (SMA) to hydroxyl groups present in the drug. The use of SMA instead of a smaller carboxylic acid monomer such as methacrylic acid yields prodrug monomers where the sterically bulky drug moieties are separated from the polymer backbone by a small spacer, which may improve copolymerization behavior with other methacrylate-based monomers.

Scheme 1.

Synthetic scheme for the preparation of peptide-targeted polymeric prodrugs containing Cam linked to the polymer scaffold via esters at position 10 or 20.

Synthesis of the aliphatic ester linked 20Cam-SMA monomer was conducted as described previously via carbodiimide coupling of SMA to Cam, which contains a single hydroxyl residue.³¹ In order to synthesize a monofunctional prodrug monomers with a single aromatic ester linkage from 10Cam it was necessary to develop reaction conditions that preferentially react with 10Cam at the phenolic 10-position. This was accomplished by first converting SMA into the acid chloride via treatment with thionyl chloride.⁴⁰ SMA acid chloride was then reacted with 10Cam in tetrahydrofuran in the presence of triethylamine to trap the by-product HCl, as depicted in Scheme 2. Analysis of the ¹H NMR for 10Cam and the resultant 10Cam-SMA monomer (Fig. 1a, b) shows that conjugation primarily occurred at the desired phenolic 10-hydroxyl group over the aliphatic 20-hydroxyl group. Here selective conjugation was confirmed by the complete disappearance of the phenolic 10-hydroxyl group signal at 10.30 ppm, while the 20-hydroxyl group signal at 6.46 ppm remained constant throughout the course of the esterification reaction. This specificity has been shown to arise from the stronger hydrogen bonding interaction between 20-hydroxyl group and adjacent carbonyl group on the lactone ring in polar tetrahydrofuran solvent.⁴¹ Synthesis of a mono-functionalized species was further confirmed by mass spectroscopy (Fig. 1c) showing [M+1] value as 577.7.

Scheme 2.

Synthetic scheme for the preparation of 10-Hydroxycamptothecin-SMA (10Cam-SMA) via selective esterification of the phenolic group at position 10 with 2-methacryloyloxyethylsuccinoyl chloride.

Fig. 1.

¹H NMR of (a) 10-hydroxy camptothecin (10Cam) and (b) 10-Hydroxycamptothecin-SMA (10CamSMA) confirming selective esterification of phenolic group at position 10 over the aliphatic hydroxyl at position 20. Positive mode mass spectroscopy spectra for (c) 10CamSMA, (d) GE11MA and (e) HW12MA confirming the synthesis of the desired mono-functionalized prodrug and peptide macromonomers respectively.

The resulting CamSMA monomers were then polymerized with tertiary- butyl ester-protected CBM (tQuat), after which the resulting copolymers were deprotected with trifluoroacetic acid (TFA) (Scheme 1). Polymerizations were conducted in the presence of the RAFT agent CTP with ABCVA as the primary radical source with an initial monomer to CTA to initiator ratio ([M]_o:[CTA]_o:[I]_o) of 50:1:0.1. An initial molar feed ratio of 16 mol % was used for both chemotherapeutic monomers. These feed compositions correspond to approximately 34 wt % prodrug monomer residues (21 wt. % drug) in the final deprotected copolymer. This composition was selected in order to yield copolymers with high solubility in aqueous buffers. Shown in Fig. 2d,e are aqueous SEC chromatograms for both poly(tQuat-co-10CamSMA) and poly(tQuat-co-20CamSMA) copolymers synthesized under these conditions. In both cases the molecular weight distributions are symmetric and narrow with molecular weights and molar mass dispersity values of 18 500:1.10 and 17 200:1.20 g/mol respectively. Shown in Fig. 2a, is the ¹H NMR spectrum for poly(tQuat-co-10CamSMA) in d₆ DMSO where resonances associated with both monomer residues can be clearly visualized. Here the tQuat tertiary butyl ester residues are observed as an intense resonance at 1.48 ppm along with overlapping ester resonances associated with both monomers and 10Cam between 3.9 and 5.1 ppm. Quantitative removal of the tertiary butyl ester protecting groups upon treatment of the copolymers with TFA was confirmed via the absence of the intense tertiary butyl ester resonances (Fig. 2b, c). Also apparent in Fig. 2b and Fig. 2c are resonance between 7 and 9 ppm that are associated with distinct aromatic protons on both 10Cam-SMA and 20Cam-SMA residues. The presence of these resonances in combination with complete removal of the tert-butyl ester resonance suggests that the copolymers were successfully deprotected without cleaving the labile phenyl ester groups connecting the prodrug residues to the polymer scaffold. Dynamic light scattering measurements conducted in PBS (150 mM NaCl, 20 mM phosphate buffer) at pH 7.4 suggest that both poly(CBM-co-20CamSMA) and poly(CBM-co-10CamSMA) have hydrodynamic diameters that are consistent with molecular dissolved unimeric species (d_h < 10 nm).

Fig. 2.

¹H NMR spectrums of (a) poly(tQuat-co-10CamSMA), (b) poly(CBM-co-10CamSMA), and (c) poly(CBM-co-20CamSMA) in DMSO-d6. Quantitative removal of the tert-butyl ester protecting groups was confirmed via the disappearance of the intense resonance at 1.5 ppm. SEC chromatograms (refractive index channel) of (d) poly(tQuat-co-10CamSMA), and (e) poly(tQuat-co-20CamSMA).

Determination of drug release kinetics in human serum and in vitro cytotoxicity for poly(CBM-co-20CamSMA) and poly(CBM-co-10CamSMA).

Drug release kinetics for the two polymeric prodrugs were evaluated in 100% human serum at 37 °C in order to establish differences in the rate at which the aliphatic (20CamSMA) and aromatic (10CamSMA) ester linked drugs were released from the hydrophilic polymer scaffold (Fig. 3). Drug release as a function of time was then quantified via reverse phase HPLC in comparison to free drug standards. Drug release from poly(CBM-co-10CamSMA), where the drug is linked at the position 10 of the pyrrolo[3,4-β]-quinoline ring, was found to be significantly faster than poly(CBM-co-20CamSMA) containing the aliphatic ester linked drug. For example, 37 % of the aromatic ester linked 10Cam was released after 4 days while only 12 % release for the aliphalic ester-linked 20Cam was observed over the same time period. These results suggest that the position of the ester linkage on Cam significantly affects the rate at which the hydrophobic prodrug residues are cleaved from the polymer scaffold and provides a method by which binary release kinetics could be introduced into a single copolymer.

Fig. 3.

Drug release kinetics measured by high-performance liquid chromatography (HPLC) as a function of time for poly(CBM-co-10CamSMA) and poly(CBM-co-20CamSMA) showing faster ester cleavage for the phenolic ester linked 10Cam residues relative to aliphatic ester linked 20Cam residues. All drug release studies were conducted in 100 % human serum at 37 °C. Error bars correspond to the standard deviation for n =4.

The cytotoxic activity of free 10- and 20-Cam as well as the polymeric prodrugs were next evaluated in SKOV3 ovarian cancer cell cultures. Shown in Fig. 4 is the dose response curve for free 10- and 20-Cam as well as poly(CBM-co-20CamSMA) and poly(CBM-co-10CamSMA). Here cells were treated with free drug or polymer for 72 hours, after which cell viability was measured using the MTS assay. As can be seen in Fig. 4, both free 10- and 20-Cam were extremely cytotoxic to SKOV3 cells with an IC50 value of 14 and 21 nM respectively. In comparison, cells treated with poly(CBM-co-20CamSMA) under these conditions yielded an IC50 value of 4746 nM, which is approximately 227 times less active than the free drug controls. In strong contrast, the aromatic ester linked poly(CBM-co-10CamSMA) system was substantially more cytotoxic to SKOV3 cancer cells with an IC50 of 49.7 nM which is approximately 96 times more active than poly(CBM-co-20CamSMA) and only 2.4 times less active than the free drug. Based on drug release studies (Fig. 3), only 30 % of the covalently linked 10Cam is released by 72 h, which corresponds to a free 10CAM concentration of 17 nM at an initial dose of 50 nM. It also should be noted that cells incubated with free drug receive the maximum drug dose over the entire treatment period while cells exposed to the conjugates experienced a tapered dose as the ester linkages hydrolyze. However, in vivo this feature is expected to increase drug circulation times, exposure in diseased tissues, and therapeutic efficacy. In contrast clinically administered camptothecin derivatives, such as topotecan, exhibit short in vivo circulation times with terminal half-lives of 2 to 3 hours.⁴² To illustrate the potential utility of this feature, we recently showed that similar polymeric prodrugs containing the antibiotic ciprofloxacin linked to a poly(oligo ethylene glycol methacrylate) scaffold via a phenyl ester provided high cure efficiencies in a completely lethal F.t. novacida pulmonary mouse challenge models, while mice treated with free ciprofloxacin were all dead within 6 days of exposure to the bacteria.⁴³ No notable cytotoxic activity was observed for the drug free poly(CBM) control polymer at any of the concentrations evaluated.

Fig. 4.

Cell viability as a function of drug concentration for SKOV3 ovarian cancer cells treated with poly(CBM-co-10CamSMA),poly(CBM-co-20CamSMA), and free 10- and 20-Cam for 72 hours. Cytotoxic activity was quantified using MTS assay relative to untreated controls. Error bars correspond to the standard deviation for n =5.

Synthesis of polymeric prodrugs targeting epidermal growth factor receptor (EGFR).

The GE11 peptide, identified via phage display, has been shown to bind strongly and specifically to the epidermal growth factor receptor overexpressed in a wide array of epithelial malignancies, making it a promising agent for enhancing the specificity of polymeric prodrug.²⁵ Prior to employing GE11 for targeting a polymeric drug delivery system in SKOV3 ovarian cancer cells, its ability to bind to EGFR overexpressed in SKOV3s was validated using GE11 and HW12 (control) peptides labelled on their N-termini with rhodamine B. Flow cytometric analysis showed that GE11 bound to SKOV3s at significantly higher levels. Results are included in supplemental section (Fig. S1). Based on these studies, a polymerizable form of the GE11 sequence was synthesized in order to facilitate facile incorporation into chemotherapeutic polymers as shown in Scheme 1. This monomer (GE11-MA) consisted of the GE11 sequence covalently linked to a polymerizable methacrylamide group at the peptide N-terminus with a short hexyl spacer (Ahx) separating the two. A non-targeting peptide macromonomer (HW12-MA) was also synthesized to serve as a negative control.

Next, the ability of GE11-MA to target a polymeric delivery platform was evaluated. Employing RAFT polymerization, both the GE11 peptide monomer and HW12 control monomer were copolymerized with tQuat. The molar compositions of the reaction feeds were 4% peptide monomer and 96% CBM. In addition, to fluorescently label the polymers for cell targeting studies, a trace amount (< 0.1 mol %) of rhodamine B monomer was added to each reaction. To assess the ability of GE11 to enhance polymer targeting, SKOV3 cells were pulsed with 1 μM polymer for 15 min at 37 °C, after which the binding/uptake of the rhodamine-B labeled polymers was assessed by flow cytometry. GE11 lead to a significant 18-fold increase in polymer binding and uptake (Fig. 5), confirming GE11’s utility as a targeting agent.

Fig. 5.

(a) Flow cytometric analysis of SKOV3 ovarian cancer cell binding by peptide-CBM copolymers containing either the GE11 sequence targeting EGFR receptors or the HW12 negative control sequence. Cells were treated with 1 μM of rhodamine B labeled polymer for 15 minutes at 37 C. (b) Average fluorescence intensity for GE11-targeted polymer is significantly larger than untargeted and untreated controls. Error bars correspond to the standard deviation for n =5.

The targeting capabilities of GE11 were then employed to enhance the delivery and cytotoxic activity of the 10Cam- based polymeric prodrug. This prodrug monomer was selected for further study based on its high cytotoxic activity. In these studies, both the GE11 peptide monomer and 10CAM prodrug monomer were incorporated into a single carboxybetaine-based polymeric scaffold. Here a poly(10CamSMA-co-tQuat) macro chain transfer agent (macro-CTA) was first synthesized from which peptide macromonomers were then copolymerized with tQuat and then subsequently deprotected with TFA to yield the desired poly[(CBM-co-10Cam)-b-(CBM-co-GE11MA)] and poly[(CBM-co-10Cam)-b-(CBM-co-HW12MA)] (Scheme 1). A small amount of Rhodamine B methacrylate (REMA) was also introduced into the block copolymer in order facilitate direct analysis of SKOV3-binding via flow cytometry. The molar feed ratio of peptide to CBM in the second block was 4:96, which corresponds to approximately 2 peptides monomer residues per polymer chain at a target DP of 50. Aqueous SEC analysis of the GE11 and HW12 containing diblock copolymers yielded M_n values of 29 600 and 32 100 respectively. The impact of peptide targeting on polymeric prodrug cytotoxic activity was next tested in SKOV3 cells. To more closely mimic in vivo exposure, cells were pulsed with polymer (100–10,000 nM) for 15 minutes at 37 °C and then washed and given fresh media. The viability was then measured following a 72-hour incubation period. As can be seen in Fig. 6, neither of the non-drug control polymers exhibited cytotoxicity at doses up to 10 μM. In contrast, both of the polymers containing 10CAM killed SKOV3 cells in a dose dependent manner. Here the poly[(CBM-co-10Cam-SMA)-b-(CBM-co-GE11MA)] was observed to significantly enhance the cytotoxic activity of the polymeric prodrug. Under the pulse treatment conditions tested, the IC50s of EGFR targeting poly[(CBM-co-10CAM)-b-(CBM-co-GE11MA)] and untargeted poly[(CBM-co-10CAM)-b-(CBM-co-HW12MA)] control were determined to be 1597 nM and 3399 nM respectively. The targeting effect of GE11 in the context of the prodrug diblock copolymers was also evaluated by incubating SKOV3 cells with the polymers and measuring binding by flow cytometry (Fig. S2). In comparison to the HW12 control polymer, GE11 was found to enhance polymer binding by approximately 2-fold, which is consistent with the approximately 2-fold increase in cytotoxic activity. This targeting effect is less pronounced than that observed for the peptide-CBM copolymers, possibly as a result of the larger size and/or lower peptide content of the polymers. Further increases in both specificity and activity could potentially be observed by incorporating high molar feed percentage into the block copolymer and/or through the use of a nanoparticle morphology that could yield multivalent ligands.

Fig. 6.

GE11 targeting peptides enhance the cytotoxic activity of Cam containing polymeric prodrugs. SKOV3 cells were pulse-treated with poly[(CBM-co-10CamSMA)-b-(CBM-co-GE11MA)] (targeting sequence) or poly[(CBM-co-10CamSMA)-b-(CBM-co-HW12MA)] (control sequence) for 15 minutes at 37 C and cell viability was measured at 72 hours by an MTS assay. Control polymers containing either the GE11 or HW12 sequences but lacking prodrug residues showed negligible cytotoxicity over the entire concentration range evaluated. Error bars correspond to the standard deviation for n =5.

Conclusions

RAFT polymerization was employed to synthesize zwitterionic polymeric prodrugs based on the chemotherapeutic agents 20Cam and 10Cam. Two prodrug monomers were synthesized via esterification of SMA with Cam at either the 10 or 20 positions to yield aromatic and aliphatic esters respectively. Copolymers containing Cam linked via aromatic esters showed significantly faster hydrolysis rates relative to the aliphatic linked 20CamSMA. In vitro cytotoxicity measurements in SKOV3 cells showed that differences in Cam release rates resulted in considerable differences in chemotherapeutic activity with the 10Cam linked polymeric prodrugs being shown to be nearly two orders of magnitude more active than the aliphatic linked 20Cam. A peptide macromonomer targeting EGFR was synthesized and incorporated into polymeric prodrugs. Flow cytometry showed a two-fold increase in binding for the targeted polymers relative to untargeted polymer controls. Pulse chase experiments showed a similar level of enhancement in cytotoxic activity in SKOV3 cells. These results taken together suggest that the RAFT copolymerization of prodrug monomers with peptide macromonomers can be used to prepare chemotherapeutic polymers with cancer cell-specific targeting and cytotoxic activity. Furthermore, the controlled drug release kinetics and cell-specific uptake demonstrated by this system promise to enhance drug efficacy and minimize off-target side effects in vivo.

C. Experimental

Materials

Chemicals and all materials were supplied by Sigma-Aldrich unless otherwise specified. Rhodamine B-Methacrylate, tert-Butoxycarbonylmethyl-dimethyl-[2-(2-methyl-acryloyloxy)-ethyl]-ammonium; bromide (tQuat) and Succinic acid 4-ethyl-3,13-dioxo-3,4,12,13-tetrahydro-1H-2-oxa-6,12a-diaza-dibenzo[b,h]fluoren-4-yl ester 2-(2-methyl-acryloyloxy)-ethyl ester (20Cam-SMA) were synthesized as described previously.^31,44,45

Cell Culture.

SKOV3 human ovarian cancer cells (ATCC) were maintained in RPMI 1640 Medium (+L-glutamine, + HEPES) supplemented with 10% FBS (GIBCO) and 1% penicillin/streptomycin (GIBCO). Cells were maintained in log-phase growth at 37 °C and 5% CO₂.

Cell Viability Assay.

For peptide targeting studies, SKOV3 cells were plated in 96-well plates at a density of 3,000 cells per well and pulsed with polymer solutions in culture medium (0–10,000 nM) for 15 min. Cells were then washed twice with media and cell viability was measured after 72 hours using an MTS Cell Proliferation Colorimetric Assay (Biovision). For studying the camptothecin prodrug polymers, SKOV3 cells were plated in 96-well plates at 5,000 cells per well and treated continuously with polymer or free drug for 72 hours, after which cell viability was measured.

Flow Cytometry.

For measuring polymer binding and uptake, polymers were synthesized containing trace quantities of a rhodamine B fluorescent monomer. SKOV3 cells were plated in 6-well plates at a density of 100,000 cells per well, allowed to adhere overnight, and treated with 1 μm polymer in media for 15 min at 37 °C. Cells were then trypsinized, washed with PBS, and analyzed on a BD LSRII flow cytometer. For measuring the binding of rhodamine B-labeled peptides, SKOV3 cells were treated with either 0.1 or 1 µm peptide for 30 minutes at 4 °C.

Drug Release Kinetics.

To evaluate the release of Cam from the prodrugs, the CBM-prodrug copolymers (6 mg/mL) were incubated in 100% human serum at 37 °C. At various time points, reaction aliquots were taken and diluted 1:1 with methanol/water (75/25, v/v) and 1:1 with acetonitrile to precipitate serum proteins. Samples were centrifuged (15 min, 12,000 g) and supernatants were collected, filtered (0.45 μm low protein binding), and analyzed by RP-HPLC (abs 370 nm). RP- HPLC analysis was conducted at ambient temperature using a Zorbax RX-C18 (4.6 × 150 mm; 5 μm) analytical column (Agilent Technologies, CA) on an Agilent 1260 system. The amount of released drug was quantified in comparison to free drug standards. The total drug content of the polymers was determined by polymer dissolution in 10% aq. H₂SO₄ (72 h, 25 °C), and percent drug release was calculated according to the following equation: % drug released = [Peak(t_x)- Peak(t_o)]/[Peak(H₂SO₄)], where t_x and to are the peaks at time x and zero, and Peak(H₂SO₄) is for total drug.

Size Exclusion Chromatography.

Absolute molecular weights and molar mass dispersity indices for the copolymers were determined using PolySep-SEC GFC-P 3000, LC Column 300 × 7.8 mm (Phenomenex Inc.) connected to an Agilent 1200 Series Liquid Chromatography System (Santa Clara, CA) and Wyatt Technology miniDAWN TREOS, 3 angle MALS light scattering instrument and Optilab TrEX, refractive index detector (Santa Barbara, CA). HPLC-grade water containing 10% aq. acetic acid was used as the mobile phase with PolySep-GFC-P 3000 SEC column (Phenomenex) at a flow rate of 0.5 mL min⁻¹.

Synthesis of 2-Methacryloyloxyethylsuccinoyl chloride.

2-Methacryloyloxyethylsuccinoyl chloride was synthesized following the reported procedure¹: Briefly, mono-2-(methacryloyloxy)ethyl succinate (SMA) 11.5g (50 mmol) was added to neat thionyl chloride 23 mL (320 mmol). The reaction mixture was stirred at room temperature for 30 min and at 50 °C for 1h. After evaporation of the excess thionyl chloride and other volatiles, the resulting 2-methacryloyloxyethylsuccinoyl chloride was used for the next step without further purification.

Synthesis of Succinic acid 4-ethyl-4-hydroxy-3,13-dioxo-3,4,12,13-tetrahydro-1H-2-oxa-6,12a-diaza-dibenzo[b,h]fluoren-9-yl ester 2-(2-methacryloyloxy)-ethyl ester (10Cam-SMA).

2-Methacryloyloxyethylsuccinoyl chloride 1.84 g (7.42 mmol) was added slowly to a solution of 10-hydroxylcamptothecin 900 mg (2.47 mmol) and triethylamine 6.98 mL (50 mmol) in 250 mL tetrahydrofuran at 4 °C. The reaction mixture was stirred at 4 °C for 15 min and at room temperature for 1h. Triethylammonium chloride formed was filtered off and the filtrate was concentrated under reduced pressure. The crude ester was purified by column chromatography with 30% tetrahydrofuran in chloroform. Column purified product was redissolved in 10 mL tetrahydrofuran and precipitated in 90 mL of 80 % ether/hexane. Yield = 976 mg (68.7%). ¹H NMR (300 MHz, CDCl₃, ppm): δ 8.32 (s, 1H), 8.22 (d, J = 9.2 Hz, 1H), 7.68 (d, J = 2.5 Hz, 1H), 7.66 (s, 1H), 7.57 (dd, J₁ = 9.2 Hz, J₂ = 2.5 Hz, 1H), 6.12 (s, 1H), 5.74 (d, J = 16.3 Hz,1H), 5.57 (s, 1H), 5.30 (d, J = 16.3 Hz,1H), 5.28 (s, 2H), 4.33 – 4.48 (m, 4H), 3.85 (s, 1H), 2.98 (t, J = 6.0 Hz, 2H), 2.83 (t, J = 6.0 Hz, 2H), 1.80 – 1.98 (1m and 1s merged, 5H), 1.04 (t, J = 7.4 Hz, 3H); ESI-MS (C₃₀H₂₈N₂O₁₀): Calculated = 577.6 [M +1]⁺; Found m/z = 577.7 [M +1]⁺ and 599.4 [M +Na]⁺.

Synthesis and Characterization of GE11MA and HW12MA peptide macromonomers.

Peptide macromonomers were synthesized using an automated PS3 peptide synthesizer (Protein Technologies). Peptides were synthesized on a solid support (rink amide MBHA resin (100–200 mesh), EMD Millipore) from FMOC protected (L) amino acids (EMD Millipore) and FMOC protected 6-aminohexanoic acid (Ahx) spacer (AnaSpec). For synthesis of peptide monomers, the amino termini of the peptides were coupled to N-succinimidyl methacrylate (TCI America) prior to cleavage from the resin. The peptides macromonomers were deprotected and cleaved from the resin by treatment with trifluoroacetic acid/triisopropylsilane/H₂O (9.5:2.5:2.5, v/v/v) for 4 hours and precipitated in cold ether. Crude peptides/monomers were purified by reverse phase high performance liquid chromatography (RP-HPLC) on a Jupiter 5 μm C18 300Å column (Phenomenex) with an Agilent 1260 HPLC. Ion trap mass spectrometry with electrospray (Bruker Esquire) was used to confirm the molecular weights of the purified peptides. Peptide macromonomers were synthesized with an Ahx spacer between the targeting sequence and monomer. Sequences for GE11MA and HW12MA are Methacrylamido (MA)-Ahx-YHWYGYTPQNVI and MA-Ahx-HYPYAHPTHPSW respectively.

Synthesis of poly(GE11MA-co-CBM) and poly(HW12MA-co-CBM).

Zwitterionic copolymers containing the GE11 targeting sequence as well as the HW12 negative sequence were prepared as follows. To a 5 mL round bottom flask was added: CTP (8.4 mg, 30 μmol), ABCVA (1.12 mg, 3 μmol), peptide monomer (60 μmol), tQuat (0.432 g, 1.44 mmol), REMA (4 mg, 6.75 μmol) and DMSO 2.2 mL ([M]o:[CTA]o:[I]o = 50:1:0.1). The polymerization solutions were then transferred to a septa-sealed vial and purged with nitrogen for 30 minutes. After this time the polymerizations were transferred into a preheated oil bath at 70 °C and allowed to polymerize for 16 hours. The polymers were then isolated via precipitation into diethyl ether followed by dialysis against deionized water at 5 °C and subsequent lyophilization. The dry polymers were then dissolved in neat TFA at a concentration of 20 mg/mL and allowed to react at room temperature for 8 hours. After this time the polymers were precipitated into a 25 times excess of cold diethyl ether and isolated via centrifugation. The polymers were then neutralized with 50 mL of ice cold phosphate buffer (0.2 M pH 7.4) and then dialyzed at 5 °C against deionized water. The final deprotected copolymers were then isolated via lyophilization.

Synthesis of rhodamine-labeled poly(10Cam-co-CBM), poly(20Cam-co-CBM), and poly(CBM).

The synthesis of zwitterionic polymeric prodrugs and negative control polymers were synthesized as follows: To a 10 mL round bottom flask was added CTP (37 mg, 132 μmol) ABCVA (3.7 mg, 13.2 μmol), 10CAM (0.312 g, 0.541 mmol) or CAM (0.303 g, 0.541 mmol), tQuat (0.936 g, 2.77 mmol), REMA (7.82 mg, 13.2 μmol) and DMSO 5.0 mL ([M]_o:[CTA]_o:[I]_o = 25:1:0.1). For synthesis of a non-drug CBM control polymer, CTP (66 mg, 237 μmol) ABCVA (6.6 mg, 23.7 μmol), tQuat (4.00 g, 11.8 mmol), REMA (14.0 mg, 23.7 μmol) and DMSO 16 mL ([M]o:[CTA]o:[I]o = 50:1:0.1) were added to a 10 mL round bottom flask. The polymerization solutions were then transferred to a septa-sealed vial and purged with nitrogen for 30 minutes. After this time the polymerizations were transferred into a preheated oil bath at 70 °C and allowed to polymerize for 16 hours. The polymers were then isolated via precipitation into diethyl ether followed by dialysis against deionized water at 5 °C and subsequent lyophilization. The dry polymers were then dissolved in neat TFA at a concentration of 20 mg/mL and allowed to react at room temperature for 8 hours. After this time the polymers were precipitated into a 25 times excess of cold diethyl ether and isolated via centrifugation. The polymers were then neutralized with 50 mL of ice cold phosphate buffer (0.2 M pH 7.4) and then dialyzed at 5 °C against deionized water. The final deprotected copolymers were then isolated via lyophilization. Copolymer composition was determined to be 14 mol % 10CAM-SMA / 86 mol % tQuat (21.7 mass % 10Cam-SMA residue and 78.3 mass % tQuat) which corresponds to 18.7 % 10-hydroxy camptothecin in the final copolymer.

Synthesis of poly[(10Cam-co-CBM)-block-(GE11MA-co-CBM) and poly[(10Cam-co-CBM)-block-(HW12MA-co-CBM).

Peptide-targeted polymeric prodrug was synthesized via the copolymerization of the appropriate peptide macromonomer with tQuat from a poly(10Cam-co-tQuat) macro-CTA. To a macro-CTA solution (21 μmol) in DMSO (1.77 mL) were added tQuat (0.344 g, 1.02 mmol), GE11MA or HW12MA (42.44 μmol), and ABCVA (0.59 mg, 2.12 μmol) ([M]o:[CTA]o:[I]o = 50:1:0.1). The polymerization solutions were then transferred to a septa-sealed vial and purged with nitrogen for 30 minutes. After this time the polymerization vial was transferred into a preheated oil bath at 70 °C and allowed to polymerize for 16 hours. The polymers were then isolated via precipitation into diethyl ether followed by dialysis against deionized water at 5 °C and subsequent lyophilization. The dry polymers were then dissolved in neat TFA at a concentration of 20 mg/mL and allowed to react at room temperature for 8 hours. After this time the polymers were precipitated into a 25 times excess of cold diethyl ether and isolated via centrifugation. The polymer was then neutralized with 50 mL of ice-cold phosphate buffer (0.2 M pH 7.4) and then dialyzed at 5 °C against deionized water. The final deprotected copolymer was then isolated via lyophilization.