Combinatorial synthesis and screening of cancer cell-specific nanomedicines targeted via phage fusion proteins

Abstract

Active tumor targeting of nanomedicines has recently shown significant improvements in the therapeutic activity of currently existing drug delivery systems, such as liposomal doxorubicin (Doxil/Caelyx/Lipodox). Previously, we have shown that isolated pVIII major coat proteins of the fd-tet filamentous phage vector, containing cancer cell-specific peptide fusions at their N-terminus, can be used as active targeting ligands in a liposomal doxorubicin delivery system in vitro and in vivo. Here, we show a novel major coat protein isolation procedure in 2-propanol that allows spontaneous incorporation of the hydrophobic protein core into preformed liposomal doxorubicin with minimal damage or drug loss while still retaining the targeting ligand exposed for cell-specific targeting. Using a panel of 12 structurally unique ligands with specificity toward breast, lung, and/or pancreatic cancer, we showed the feasibility of pVIII major coat proteins to significantly increase the throughput of targeting ligand screening in a common nanomedicine core. Phage protein-modified Lipodox samples showed an average doxorubicin recovery of 82.8% across all samples with 100% of protein incorporation in the correct orientation (N-terminus exposed). Following cytotoxicity screening in a doxorubicin-sensitive breast cancer line (MCF-7), three major groups of ligands were identified. Ligands showing the most improved cytotoxicity included: DMPGTVLP, ANGRPSMT, VNGRAEAP, and ANDVYLD showing a 25-fold improvement (p < 0.05) in toxicity. Similarly DGQYLGSQ, ETYNQPYL, and GSSEQLYL ligands with specificity toward a doxorubicin-insensitive pancreatic cancer line (PANC-1) showed significant increases in toxicity (2-fold; p < 0.05). Thus, we demonstrated proof-of-concept that pVIII major coat proteins can be screened in significantly higher throughput to identify novel ligands displaying improved therapeutic activity in a desired cancer phenotype.

Keywords: breast cancer; doxorubicin; liposomal drug delivery; pancreatic cancer; phage display; targeted nanomedicines.

Figure 1

Phage fusion protein-targeted nanoparticles. Target-specific phage fusion protein selected from phage libraries can be introduced into drug-loaded PEGylated liposomes exploiting their intrinsic properties to spontaneously integrate into lipid bilayers.

Figure 2

(A) UV/Vis spectrum (1 nm scan) of pVIII major coat protein solubilized in 75% v/v 2-propanol/1X TBS, pH 7.4 blanked with solution of 75% v/v 2-propanol/1X TBS, pH 7.4. (B) SDS-PAGE of pVIII major coat protein solubilized in 75% 2-propanol/1X TBS, pH 7.4 separated on a 4–20% polyacrylamide gel. Bands were identified by staining with a colloidal blue staining kit (Invitrogen) to reveal a single, low molecular weight band of ~7.5 kDa (as noted by the arrow) in samples 1 and 2. M (Marker)—1:10 dilution of Precision Plus Protein WesternC Standards (Bio-Rad), Sample 1–1.36 μg DMPGTVLP phage protein, Sample 2–0.27 μg DMPGTVLP phage protein.

Figure 3

Purification and characterization of DMPGTVLP-modified Lipodox. Size exclusion chromatography of modified Lipodox on a Superose 6 column (30 × 1 cm), 2.5 mL/fractions 4–11 of the elution profile are shown. Western blot of recovered fractions probed with a polyclonal anti-fd rabbit primary IgG, followed by a biotinylated goat anti-rabbit secondary IgG and detected with NeutrAvadin-HRP and Pico West luminol substrate. Physicochemical characterization of recovered fractions for size distribution (d.nm), zeta potential (mV), doxorubicin recovery (μg), and percent doxorubicin recovery (%).

Figure 4

(A) Schematic of phage protein orientation assay, where it is expected that the N-terminus of the protein is exposed to proteinase K degradation while the C-terminus of the protein is protected from degradation by the lipid bilayer. (B) N-terminal orientation assay in DMPGTVLP-modified Lipodox. SDS-PAGE of concentrated DMPGTVLP-modified Lipodox followed by assay by Western Blot. M (Marker)—1:10 dilution of Precision Plus WesternC Standards, Samples 1 and 2–500 ng DMPGTVLP isolated protein, Samples 3 and 4—DMPGTVLP-modified Lipodox (~500 ng DMPGTVLP protein). Samples 1 and 3—untreated controls, Samples 2 and 4—Proteinase K (PK) treated samples. (C) Phage protein orientation assay by dot blot analysis with data quantified by densitometry and presented as the mean ± standard deviation of output signal normalized to 1X untreated protein (~375 ng protein). Untreated samples (dark bars) are compared to proteinase K treated samples (light bars) N = 3.

Figure 5

Flow cytometry analysis of (A) unmodified Lipodox and (B) 488-labeled-DMPGTVLP-modified Lipodox. Doxorubicin quantified on 670LP emission channel and 488-labeled phage protein quantified on 533/30 emission channel. Samples were excited using a 488 nm laser.

Figure 6

MTT viability assay of MCF-7 cells treated with dilutions of Lipodox (dark bars) or DMPGTVLP-modified Lipodox (light bars) after 24 h of incubation. Data are presented as the mean ± sample standard deviation of the percent viable fraction compared to untreated control cells, which were taken as 100% viable. N = 3; *P < 0.05, paired, two-tailed Student's t-test vs. unmodified Lipodox.

Figure 7

Doxorubicin uptake assay in MCF-7 cells treated with 2 μg of Lipodox (dark bars) or 2 μg DMPGTVLP-modified Lipodox (light bars) over 24 h. Data are presented as the mean ± sample standard deviation of the relative fluorescence of doxorubicin at an excitation wavelength of 470 nm and emission wavelength of 590 nm. N = 3; *P < 0.05, paired, two-tailed Student's t-test vs. unmodified Lipodox.

Figure 8

MTT viability assay of (A) MCF-7 or (B) PANC-1 cells treated with dilutions of phage protein-modified Lipodox after 24 h of drug incubation followed by 48 h of drug washout. Modified samples are identified by their displayed 8- or 9-mer fusion peptide sequence in the legend above. Data are presented as the mean ± sample standard deviation of the percent viable fraction compared to untreated control cells, which were taken as 100% viable. N = 3; *P < 0.001, paired, two-tailed Student's t-test vs. unmodified Lipodox; ^# P < 0.05, paired, two-tailed Student's t-test vs. unmodified Lipodox.

Figure 9

Spontaneous insertion of the major coat protein into lipid membranes. (A) In the first step, the isolated coat protein binds to the lipid membrane using electrostatic interactions of the C-terminus with negatively charged phosphate headgroups of lipids, followed by the insertion of the hydrophobic region of the protein into the hydrocarbon core of the lipid bilayer. (B) In the second step, the hydrophilic tail is released into the trans side of lipid bilayer through a process similar to cell-penetrating peptides. From Petrenko and Jayanna (2014).