Promiscuous tumor targeting phage proteins

Abstract

Cancer cell-specific targeting ligands against numerous cancer cell lines have been selected previously and used as ligands for cell-specific delivery of chemotherapies and various nanomedicines. However, tumor heterogeneity is one recognized problem hampering clinical translation of targeted anti-cancer medicines. Therefore, a novel class of targeting ligands is required that recognize receptors expressed between a variety of cancer phenotypes, identified here as 'promiscuous' ligands. In this work, promiscuous phage fusion proteins were first identified by a novel selection scheme to enrich for pan-cancer cell binding abilities, as indicated by conserved structural motifs identified previously in other cancer types. Additionally, peptide sequences containing a combination of motifs were identified to modulate binding. A panel of phage fusion proteins was studied for their specificity and selectivity for lung and pancreatic cancer cells. Phage displaying the fusion peptides GSLEEVSTL or GEFDELMTM, the two predominate clones with greatest binding ability, were used to modify preformed, doxorubicin-loaded, liposomes. These modified liposomes increased cytotoxicity up to 8.1-fold in several cancer cell lines when compared with unmodified liposomal doxorubicin. Taken together, these data indicate that promiscuous phage proteins, selected against different cancer cell lines, can be used as targeting ligands for treatment of heterogeneous tumor populations.

Keywords: lung cancer; multi-target selection; pancreatic cancer; phage display; targeted nanomedicines.

Fig. 1

Promiscuous phage display selection scheme. (A) Selection strategy utilizing both lung and pancreatic cancer cells to generate the eluate–eluate, eluate–lysate, lysate–eluate and lysate–lysate fractions after each round of selection. The output fractions generated from each round of selection is indicated by an arrow pointing from the input fraction. The target cells and experimental conditions for each round of selection are shown at the right. (B) Total phage yield after each round of selection as calculated for both the eluate (white bars) and lysate (black bars) fractions for comparison between each round of selection. (C) Histogram of the information content of the phage sublibrary enriched for promiscuous cancer cell-binding (black line) compared to the parent f8/9 library (gray line).

Fig. 2

Selectivity and specificity assays of phage with target cells. (A) Phage capture assay of representative phage clones interacting with either pancreatic cancer cells (black bars) or serum (white bars). (B) Phage capture assay of representative selected phage clones interacting with pancreatic cancer cells (black bars), lung cancer cells (striped bars), normal lung cells (gray bars) and serum (white bars). (C) Phage capture assay of representative phage clones containing a DGR and/or VDG motifs interacting with pancreatic cancer cells (black bars) or serum (white bars). (D) Phage capture assay comparing binding of phages containing VDG and DGR motives in pancreatic cancer cells (black bars), lung cancer cells (striped bars), normal lung cells (gray bars) and serum (white bars). For all plots, the phage clone VPEGAFSSD was included as a negative control. Data are presented as the mean percent yield ± standard deviation (SD). * indicates a P-value < 0.05, when comparing paired output yields. (1) EVNVEEINL; (2) EGMNYGIDQ; (3) EVSVEEINL; (4) DPNWEATVG; (5) EDARTAAMA; (6) ATYNESVNE; (7) EPWSPTMGD; (8) GSLEEVSTL; (9) GPYVGDLDS; (10) DGRADLSYD; (11) DGRPDTVDG; (12) GGEDEESTV; (13) DHVWAEGDS; (14) ADTAEVSTL; (15) DPRVESMSG; (16) DYGEEAINV; (17) DNGREVGND; (18) DGRMGSEVS; (19) GEFDELMTM; (20) AEYGESVNA; (21) DGRTIGDND; (22) DGRFSDMPT; (23) DGRDHSGQD; (24) DGRHLDQVD; (25) DGRFGVDGS; (26) VDGRMGDMG; (27) negative control.

Fig. 3

Intracellular accumulation and fate of cancer cell-specific phage. (A) Intracellular fate of isolated phage clones in PANC-1 and Calu-3 cells. Cancer cells were incubated with 10⁷ virions of each corresponding phage for 4 h and visualized using confocal immunofluorescence microscopy. Phage were visualized with an Alexa Fluor^®-488-conjugated anti-fd IgG (green), membranes visualized with a WGA-Alexa Flour^®-555 conjugate (red), and nuclei visualized with TO-PRO-3 (blue). All scale bars are 10 µm. (B) Intracellular accumulation of phage clone GSLEEVSTL. Orthogonal slices of PANC-1 or Calu-3 cells after 4 h incubation with GSLEEVSTL phage. Phage were visualized with an Alexa Fluor^®-488-conjugated anti-fd IgG (green), membranes visualized with a WGA-Alexa Flour^®-555 conjugate (red), and nuclei visualized with TO-PRO-3 (blue).

Fig. 4

Physicochemical characterization of phage protein-modified lipodox. Presence of phage protein was assayed by western blot with an anti-fd IgG after separation by SDS–PAGE. Samples: (1) 10–250 kDa marker (Bio-Rad), (2) VNGRAEAP protein, (3) EPSQSWSM protein, (4) GSLEEVSTL protein, (5) Mock sample control, (6) VNGRAEAP-modified Lipodox, (7) EPSQSWSM-modified Lipodox, (8) GSLEEVSTL-modified Lipodox, (9) Mock Lipodox sample control and (10) Unmodified Lipodox.

Fig. 5

Cytotoxicity of phage protein-modified lipodox. Cell viability as determined by MTT assay of GSLEEVSTL-modified Lipodox (A, black bars) or GEFDELMTM-modified Lipodox (B, black bars) compared with unmodified Lipodox (white bars) in several relevant cell lines. Percent viability was calculated by comparison of treated cells to untreated control cells, which were determined as 100% viable. * denotes a P-value < 0.05.