Purification of Wnt3a liposomes
Mouse Wnt3a protein was purified as described 
, without the heparin purification step. The product, containing approximately equal amounts of Wnt3a and bovine serum albumin, was concentrated further to 250 ng/µl in PBS with 1% CHAPS.
Generating Wnt3a liposomes
Many different lipid compositions were attempted. In all cases, 14 µmol of lipid were added; when multiple lipids were used, they were added in a 90
4 molar ratio as indicated. 1,2-Dipalmitoyl-sn
-Glycero-3-Phosphocholine (DPPC)(cat#: 850355C), 1-Myristoyl-2-Palmitoyl-sn
-Glycero-3-Phosphocholine (MPPC)(cat#: 850445C), 1,2-Distearoyl-sn
-Glycero-3-Phosphocholine (DMPC)(cat#:850345C), and 1-Palmitoyl-2-Oleoyl-sn
-Glycero-3-Phosphocholine (POPC)(cat#:850457C) were obtained from Avanti Polar Lipids, Inc. (Alabaster, AL).
Unless otherwise indicated, 14 µmol of DMPC in chloroform were dried to a thin film in a 10 ml round bottom flask using nitrogen gas and were further evaporated in a vacuum overnight. Purified Wnt3a in 1% CHAPS in 1× PBS was then diluted in 1× PBS to a total concentration of 1–1.3 µg/ml. This solution was then added to the 10 ml flask and vortexed vigorously until the solution was cloudy and there was no lipid visible on the bottom of the flask. The lipid solution was then extruded 40 times through a 100–200 nm polycarbonate membrane in a thermo-barrel extruder held at 30–32°C (Avanti Polar Lipids, Inc).
To separate liposome-associated Wnt3a from free Wnt3a, the extruded solution was spun in an Optima TLX Ultracentrifuge (Beckman Coulter, Fullerton, CA) at 28,000 rpm for a minimum of 30 minutes at 4°C. The supernatant was removed and the lipid pellet re-suspended in 1× DMEM (Mediatech, Inc., Herndon, VA). The liposomes, if not used immediately, were stored at 4°C.
Cell culture and activity assays for Wnt3a liposomes
Mouse LSL cells were grown at 37°C and 5% CO2 in 1× DMEM, 10% FBS, and 1% Penicillin/Streptomycin (Mediatech, Inc., Herndon, VA). The cells were plated in 96-well plates with an initial density of 25,000 cells/well and allowed to recover overnight. The cells were then treated as described and incubated for an additional 17 hours. As the LSL cells constitutively express β-galactosidase and express luciferase in response to Tcf/Lef binding, activity was assessed via the Dual-Light® Combined Reporter Gene Assay System (Applied Biosystems). Relative luciferase units were measured and normalized against β-galactosidase activity. Error bars indicate standard deviation. All assays were done in triplicate. Student's T test was employed to determine statistical significance.
Determination of effective Wnt3a concentration in Wnt liposomes
Increasing amounts of purified Wnt3a protein were added to LSL cells, grown in 96-well plates, in order to generate an activity gradient. In a parallel set of experiments, LSL cells were exposed to different volumes of Wnt3a liposomes. All experiments were performed in triplicate. Activity of the purified protein was then determined by luciferase activity (as described above) and liposomal Wnt3a activity was plotted on the same graph. From these data we interpolated the concentration of active Wnt3a in a given volume of liposomes.
Trypsin digestion of Wnt3a liposomes
Exo-liposomal Wnt3a protein was removed by subjecting liposomes to trypsin digestion. Briefly, 75 µl of Wnt3a liposomes were added to 405 µl of 1× DMEM and 20 µl 1× Trypsin solution (Mediatech, Inc., Herndon, VA) for a final mixture containing 4.0% trypsin. Liposomes were then incubated at 37°C for 0, 5, 10, 15, or 20 minutes. The proteolytic reaction was quenched by the addition of 1 ml of 1× DMEM containing 10% FBS. The samples were ultra-centrifuged at 28,000 rpm for 30 min at 4°C; the supernatant was then removed and the lipid pellet re-suspended in 1× DMEM. The samples were then added to LSL cells in triplicate, or prepared for standard Western blot analysis.
Isolation and characterization of Wnt antagonists
We undertook a small molecule Wnt antagonist high-throughput screen using a mammalian cell–based reporter assay. The reporter plasmid (Super8xTOPflash) was stably introduced into the mouse C3H10T1/2 cell line, and the best 10T1/2 Super8xTOPflash clone gave ~800 fold up-regulation of luciferase activity after 24 hour incubation with Wnt3a conditioned media.
Under these assay conditions we screened ~205,000 small molecule compounds at a concentration of 10 µM and identified several potential Wnt antagonists (hereafter referred to as Ant). These antagonists were further evaluated using a luciferase reporter driven by the SV40 large T antigen promoter (for detecting general non-specificity) and a minimal promoter with multimerized Gli binding sites (for detecting activity on the Hedgehog pathway). The primary screen hit for the compound series described in this paper was effective at inhibiting Wnt signaling (IC50~1–2 µM) but did not function as a Hedgehog inhibitor (IC50>20 µM) and showed no inhibitory effect on SV40 promoter driven luciferase reporter at concentrations up to 30 µM. In the derivative Ant 1.4, either a bromide or a chlorine atom was substituted in the ortho position to enhance activity over the unsubstituted molecule.
Evaluation of Wnt antagonist activity in autocrine and paracrine cultures
C3H10T1/2 cells were transfected with a Super8xTOPflash luciferase reporter plasmid (Upstate Cell Signaling Solutions). PA-1 cells (ATCC) were transfected with a TOPflash (Upstate Cell Signaling Solutions) luciferase reporter plasmid. Both cell lines were maintained using standard tissue culture protocols.
PA-1 cells exhibit autocrine Wnt signaling 
; to monitor the ability of antagonists to block autocrine Wnt activity, cells were plated in 96-well plates at 20,000 cells/well in growth medium; 24 hrs later, the cells were changed to fresh growth medium and the antagonists were added in the presence of 20% Wnt3a condition medium, or 0.1 µg/ml purified Wnt3a protein. After 24 hrs, plates were assayed for luciferase activity with the LucLite kit (Packard).
To test the ability of the antagonists to block an exogenous Wnt signal and autocrine Wnt signal simultaneously, PA-1 cells transfected with Super8xTOPflash reporter were plated as described above; 24 hrs later, the cells were changed to fresh growth medium and the antagonists were added in the presence of 20% Wnt3a condition medium, or 0.1 µg/ml purified Wnt3a protein. After 24 hrs, plates were assayed for luciferase activity with the LucLite kit (Packard).
To test the ability of the antagonists to block paracrine Wnt signaling, we co-cultured 10T1/2 Super8xTOPflash reporter cells with L cells stably expressing Wnt 3a protein. In all assays, hFzd8 CRD, an antagonist of Wnt signaling 
, served as a positive control for Wnt inhibition.
Wnt antagonist activity: liposomal preparation
Wnt antagonist (Ant) 1.4Cl was diluted from 20 mM in DMSO to final concentrations of 2 µM or 20 µM in 1× DMEM. LSL cells were treated with 15 µL of Wnt3a liposomes or 15 µL of purified Wnt3a protein and varying concentrations of Ant 1.4Cl ranging from 0.01 µM to 10 µM and allowed to incubate overnight. The activity assay was conducted as previously mentioned and was conducted in triplicates. Percent inhibition values are relative to non-antagonist treated Wnt3a liposome or free Wnt3a protein activity as appropriate.
All experiments were performed in accordance with Stanford University Animal Care and Use Committee guidelines. Six three-week old, male CD-1 mice were purchased from Charles River Laboratories, Inc. (Wilmington, MA). Animals were housed in a light- and temperature-controlled environment and given food and water ad libitum.
Before subcutaneous injection of Wnt3a/PBS liposomes, the back of the mice was shaved and cleansed with antiseptic Betadine. Two regions, each 1 cm in diameter, were marked approximately 5 mm lateral from the midline. Mice received subcutaneous injections of either 20 µl Wnt3a liposomes or PBS liposomes every other day. Tissues were harvested after 14 days (7 injections) and fixed in 4% Paraformaldehyde overnight. Samples were processed for paraffin embedding, sectioned at 8 µm thickness and stained with H&E for histological evaluation.