top bar

along a 2 mm wide vertical band in the middle of the lane. The mobility was taken to be the point that divides the intensity curve into halves of equal areas. In cases where non-binding AP-positive degraded products were significant, the intensity curve for each lane was first corrected by subtracting from it the intensity curve for the degraded product alone, obtained from the intervening lanes. The intensity curve for the degraded product was assumed to follow a Gaussian distribution and determined using the software CurveExpert (Daniel Hyams, Starkville, MS). Non-linear, least-squares curve fitting was used to fit ACE data to the equilibrium equation (Figure 2) using GraphPad Prism (GraphPad Software Inc., San Diego, CA).

 

Results

 

Expression and Purification of the Glypican-AP Fusion Protein in Cell Lines:

The glypican-AP fusion plasmid obtained from transformed E. coli cells (at concentrations of 0.5 to 2 µg/ml, as determined with UV spectrophotometry) was used to transfect three different cell lines: cos-7, 293, and CHO cells. The media of transfected cells, before and after heat treatment, all showed significant levels of AP activity, compared to unobservable AP activity in untransfected cell media. Separating the transfectant media on agarose gels resulted in highly extended bands characteristic of glycanated proteins, indicating the presence of the glypican-AP fusion proteins. This was also confirmed by a Western blot showing a band corresponding to the molecular weight of the fusion protein (done for cos-7 cells only; Figure 3B). The pooled DEAE fractions with high AP activity yielded an extended band (Figure 3A, arrow) that corresponds only to the lower portion of the band from the original medium, indicating that the pooled fractions contain a more highly charged subpopulation of the original glypican-AP sample. Heparitinase, but not chondroitinase ABC, treatment replaced the extended band with a compact band. This latter band corresponds to a de-glycanated protein (Figure 3C); this confirms that the glypican-AP fusion protein contains mostly, if not all, HS chains rather than chondroitin sulfate chains.

 

Affinity Co-electrophoresis Binding Assays of Glypican-AP Fusion Proteins:

The ACE electrophoretograms for the binding of glypican-AP fusion proteins obtained from 293 and CHO cells against three common HS-binding extracellular molecules­Type I collagen, bFGF, and laminin-1­are shown in Figure 4. The ACE gels for CHO-derived glypican-AP also show, in addition to the expected ACE patterns, a dark, non-shifting, horizontal band across all nine lanes near the top (Figure 4, arrow). The low mobility of that band suggests that it represents non-

binding, AP-positive degraded products that probably consist of AP fragments resulting from proteolytic degradation of the fusion protein. To account for the degraded products, the intensity curve for each lane was corrected as described in Materials and Methods (Figure 5). After the necessary corrections, the mobility for each lane was determined and used to calculate R-values, which were plotted against ligand concentrations for each binding assay (Figure 6). A best-fit curve based on the equilibrium equation (Figure 2) was determined, and a Kd value was obtained for each binding assay (Figure 7).

Discussion

 

We have used AP-tagging and ACE to measure the binding affinity of glypican against various extracellular molecules. AP-tagging was used to specifically label glypican as to allow for its detection during ACE. Before interpreting the ACE results, it should be noted that the DEAE-purified glypican fraction also contains other moderately to highly charged proteins, including endogenous PGs that, in theory, can bind the ligands and interfere with the binding assays. However, since the cell lines only express small amounts of endogenous PG, and since the APtag-2 vector selectively amplifies the expression of the glypican-AP fusion protein, the ratio of glypican-AP to endogenous PG is likely large

Line