Printing buffer composition: Peptides were first dissolved in DMSO, followed by addition of 50 mM EDTA (pH 8.0) to a final DMSO content of 20% (DMSO was omitted for peptide EBP-0). Glycerol was then added to a final concentration of 7.3% (v/v). Slide wash buffer consisted of phosphate buffered saline (PBS) with 0.05% Tween 20 (v/v). Cell capture, slide blocking, and wash buffers were formulated as follows: PBS with 0.5% bovine serum albumin (BSA) (w/v) and 2 mM EDTA. ICC blocking buffer consisted of PBS with 2% BSA and 10% goat serum. T24 cells were cultured in McCoy's 5a Modified Medium, and J82 cells were cultured in Eagle's Minimum Essential Medium. Cultures were incubated at 37°C, 5% CO2 and supplemented with 10% fetal bovine serum (FBS) and penicillin/streptomycin.Phage display
Cell-sur face biopanning of phage-display libraries was done on human urothelial cells lines using magnetically activated cell sorting as previously described (18). In order to identify urothelial cell binding peptides that do not bind WBCs, phage display was carried out in the presence of urothelial cells (T24 or J82) and WBCs (1:9 ratio). Urothelial cells were biotinylated and conjugated to streptavidin magnetic microbeads, and phage libraries were incubated with the mixture of cells. The cell-phage mixture was then added to an LS Column and placed in a magnetic field. After multiple washes to remove the WBC population and loosely bound phage, the column was removed from the magnetic field, and urothelial cells were eluted. Urothelial cell binding phage were amplified overnight in E. coli (XL1-Blue), followed by three additional rounds of biopanning. Using a fluorescently labeled anti-M13 antibody, phage pools from each round were analyzed by fluorescently activated cell sorting (FACS) and enzyme-linked immunosorbent assay (ELISA) to identify enriched cell binding pools. Selected pools were titrated on lawns of E. coli, and individual phage were amplified and screened by FACS. Phage that bound T24 and J82 cells without binding WBCs were selected for DNA sequencing by rolling circle amplification (Sequetech, Mountain View, CA), and peptide sequences were deduced from the resulting DNA sequences.Peptide synthesis
Peptides (Table 1) were synthesized on a PTI Symphony synthesizer using standard Fmoc chemistry (19). To facilitate unidirectional attachment to NHS-derivatized glass slides, the peptide amino termini and any internal lysine residues were acetylated, and a flexible PEG linker and lysine residue were added to the C terminus. For measurements of peptide density on the slides, peptides were also synthesized with a trypsin-cleavable site near the C terminus. Following synthesis, simultaneous cleavage and side chain deprotection was achieved by treatment with a trifluoroacetic acid (TFA) cocktail. Crude peptide was precipitated with cold diethyl ether and purified by reverse-phase HPLC using a Vydac C18 silica column (10 µm, 120 Å, 250 × 22 mm) with a linear gradient of water/acetonitrile (0.1% TFA). Homogeneity of the purified peptides was evaluated by analytical RP-HPLC, and molecular mass was confirmed by MALDI-TOF-MS.Table 1.
Covalent attachment of peptides to glass slides
Peptides were attached to glass slides using N-hydroxysuccinimide (NHS) chemistry. To create discrete peptide-conjugated regions, the NHS-derivatized glass slides were mounted in a ProPlate tray with 4, 8, or 16 distinct wells (2.67 cm2, 1.13 cm2, or 0.28 cm2 per well, respectively), and peptides were dissolved in printing buffer (0 or 200 µM) and pipetted into each well (946, 396, or 100 µl, respectively). Slides were incubated for 2 h (30°C, 75% relative humidity) then washed 3 times (5 min each) with washing buffer. Unreacted NHS groups were hydrolyzed using deactivation solution.Peptide density on slides
Peptide density on the slides was quantified by cleaving the peptide with trypsin (5 µg) in ammonium bicarbonate buffer (1.7 mL, 100 mM, pH 8.5). Slides were covered with the trypsin solution, sealed, and incubated for 5 h (37°C with shaking), after which the released peptide fragment was collected and dried in a SpeedVac. The pellet was suspended in water, and peptide was quantified by HPLC using a standard curve prepared from the trypsinized peptide fragment.Selectivity of peptide-conjugated slides
WBCs (labeled with CellTracker Green) and urothelial cells (J82; labeled with CellTracker Red) were mixed (7:1, 8:1, or 72:1 ratio) and applied to peptide-conjugated slides. Slides without peptide were used as a control. After 15 min, unbound cells were removed by tipping the slide to allow the liquid to drain. Slides were then washed 3 times by immersing in wash buffer, then fixed in 4% paraformaldehyde. The number of each cell type bound was counted manually or by Image-Pro Plus 7.0 software (Media Cybernetics, Rockville, MD) using fluorescence microscopy (QIClick camera, QImaging, Surrey, BC, Canada) with a 10× objective (N = 5 fields per well; N = 2 wells). After determining the final ratio of WBC:J82 on the slides, WBC depletion was calculated as 1 - (final ratio/start ratio).