Plectin-targeted liposomes enhance the therapeutic efficacy of a PARP inhibitor in the treatment of ovarian cancer

Abstract

Advances in genomics and proteomics drive precision medicine by providing actionable genetic alterations and molecularly targeted therapies, respectively. While genomic analysis and medicinal chemistry have advanced patient stratification with treatments tailored to the genetic profile of a patient's tumor, proteomic targeting has the potential to enhance the therapeutic index of drugs like poly(ADP-ribose) polymerase (PARP) inhibitors. PARP inhibitors in breast and ovarian cancer patients with BRCA1/2 mutations have shown promise. About 10% of the patients who received Olaparib (PARP inhibitor) showed adverse side effects including neutropenia, thrombocytopenia and in some cases resulted in myelodysplastic syndrome, indicating that off-target effects were substantial in these patients. Through proteomic analysis, our lab previously identified plectin, a cytolinker protein that mislocalized onto the cell surface during malignant transformation of healthy ovarian tissue. This cancer specific phenotype allowed us to image pancreatic cancer successfully using plectin targeted peptide (PTP) conjugated to nanoparticles or displayed on capsid protein of adeno-associated virus (AAV) particles. Objective: The goal of this study was to integrate the available pharmacogenomics and proteomic data to develop effective anti-tumor therapies using a targeted drug delivery approach. Methods: Plectin expression and localization in human ovarian tumor specimens were analyzed followed by in vitro confirmation of cell surface plectin localization in healthy and ovarian cancer cell lines. PTP-conjugated liposomes were prepared and their specificity for plectin+ cells was determined in vitro and in vivo. A remote loading method was employed to encapsulate a PARP inhibitor (AZ7379) into liposomes. An ideal buffer exchange method and remote loading conditions were determined based on the amount of lipid and drug recovered at the end of a remote loading process. Finally, in vivo tumor growth studies were performed to determine the efficacy of PTP liposomes in preventing PARP activity in mice bearing OVCAR8 (high grade epithelial ovarian cancer (EOC)) tumors. Results: PTP liposomal AZ7379 delivery not only enhanced PARP inhibition but also resulted in decelerated tumor growth in mice bearing subcutaneous and intraperitoneal OVCAR8 tumors. In mice bearing subcutaneous or intraperitoneal tumors, treatment with PTP liposomes resulted in a 3- and 1.7-fold decrease in tumor volume, respectively, compared to systemic drug treatment. Conclusion: Targeted drug delivery assisted by genomic and proteomic data provides an adaptable model system that can be extended to effectively treat other cancers and diseases.

Keywords: epithelial ovarian cancer; genomics; pharmacodynamics; pharmacokinetics; plectin; poly(ADP-ribose) polymerase; proteomics; targeted drug delivery.

Figure 1

Representative images of Plectin-1 immunohistochemistry of Ovarian TMAs. (A) Representative images of serous, clear cell, poorly differentiated, mucinous, and endometrioid carcinomas and benign cystadenoma tumors. In serous, clear cell and poorly differentiated carcinomas, plectin was highly expressed and localized mostly in the cell membrane. In endometrioid and mucinoius carcinomas, plectin was mostly localized in the cytosol. In benign samples (serous and mucinous cystadenoma), plectin expression was low and localized in the cytosol. Insets in the images are magnified regions from the corresponding black squares. (Scale bar: 20 μm) (B) Western blot for plectin and cell membrane protein, alpha 1 sodium potassium ATPase (ATP1A1) from surface biotinylated fraction of proteins from FT132, SKOV3 and OVCAR8 cells. (C) Densitometric ratios of plectin to ATP1A1 were plotted for FT132, SKOV3 and OVCAR8 to determine plectin expression in relation to ATP1A1. (* represents p < 0.05).

Figure 2

Plectin-targeted peptide (PTP) liposomes bind to Plectin. (A) PTP peptide was conjugated to DSPE-PEG-maleimide via the thiol present on the C-terminus of the peptide. (B) Binding of PTP liposomes to the C-terminus of Plectin was measured using biolayer interferometry (BLI). The anti-His antibody coated sensors were used to bind his-tagged plectin fragment followed by association with no peptide, NCP and PTP liposomes. Association and dissociation response was seen by PTP liposomes but not with No peptide and NCP liposomes. (B, baseline).

Figure 3

PTP liposomes bind to SKOV3 and OVCAR8 cells. Cell binding assay (ELISA) was carried out to assess PTP liposome association to cell surface plectin on SKOV3 and OVCAR8 cells. Liposomes (No Pep, NCP and PTP) at varying concentrations were incubated with FT132, SKOV3 and OVCAR8 cells for 1 h at 37 °C followed by quantification of DiR using fluorescent plate reader. The binding of No Pep liposomes was subtracted from NCP to determine the non-specific binding. NCP binding was subtracted from PTP liposomes to determine specific binding of PTP liposomes to (A) FT132, (B) SKOV3 and (C) OVCAR8 cells.

Figure 4

PTP liposomes accumulate in OVCAR8 tumors to a greater extent than SKOV3. (A) In vivo images of mice bearing SKOV3 and OVCAR8 tumors injected with No Peptide (No Pep), Negative Control Peptide (NCP) and PTP liposomes (t=4 h). (B) The same animals were imaged at 24 h post-injection. (C-D) Graphs showing pharmacokinetics (percent injected dose of DiR) of three liposomes obtained using FMT imaging followed by image analysis using the tumor as the region-of-interest (ROI; white dotted circles). FMT imaging of the liposomal preparations was performed in live mice (n=6). Liposomes displaying a NCP were included to compare random vs. targeted liposome kinetics. In this instance, targeted liposome accumulation was largely consistent with the density of available cellular targets (* represents p < 0.05). (E) Biodistribution of liposomes 24 h post-injection from different organs including the tumors (* represents p < 0.05). (F) Following 24 h of in vivo imaging, tumor sections were prepared and stained for plectin followed by confocal microscopy to look at plectin and DiR colocalization. Confocal images of the tumor sections revealed DiR-labeled liposomes (red) accumulated in the plectin+ (green) SKOV3 and OVCAR8 cells (DAPI, blue; scale bar, 20 μm).

Figure 5

PARP inhibition in OVCAR8 tumors following AZ7379 delivery. (A) Experimental design for measuring PD following AZ7379 delivery. There were 7 groups - 1. Untreated (n=5), 2. Systemic AZ7379-0.5 (n=5), 3. Systemic AZ7379-1 (n=5), 4. No Peptide liposomes AZ7379 (No Pep AZ7379) (n=6), 5. NCP liposomes AZ7379 (n=6), 6. PTP liposomes AZ7379-0.5 (n=5), 7. PTP liposomes AZ7379-1 (n=6). Systemic AZ7379 was delivered three times a week by oral gavage and liposomal groups were delivered twice a week via tail vein injections. (B) Tumor volume (mm³) was measured every week following treatment and plotted over time to determine the efficacy of each treatment group. (C) Final tumor volume from all the 7 groups indicates significantly lower tumor volumes with PTP liposomes-0.5,-1 (* represents p < 0.05, n.s- non-significant).

Figure 6

PTP liposomes delay tumor growth of OVCAR8 cells expressing iRFP. (A) Experimental design for measuring PD following AZ7379 delivery. There were 4 groups: 1) untreated (n=6), 2) systemic AZ7379 (n=6), 3) No Pep AZ7379 (n=6), 4) PTP liposomes AZ7379 (n=6). Systemic AZ7379 was delivered three times a week by oral gavage and liposomal groups 3 and 4 were delivered twice a week via tail vein injections. (B) FMT images of mice from different treatment groups at days 2, 12, and 22 after treatment. (C) Tumor volume was measured every week following treatment by imaging the mice via FMT and plotted over time to determine the efficacy of each treatment (p< 0.05). (D) Ex vivo FMT images of organs after 3 weeks of treatment. iRFP fluorescence was normalized to the weight of the organs and represented as pmol of iRFP per g tissue. Tumor burden included tumor isolated from the omentum and diaphragm, which were the majors regions where tumor mass was observed. (E) PAR and actin were quantified from tumor lysates by Western blot. The Western blot images represent lysates from 3 animals of each group chosen randomly from this study. (F) PAR expression was normalized to actin expression in the same lysates/ blot (* represents p< 0.05).