Purchase this article with an account.
Thomas Barnes, Joseph Kovalchin, Allyson Masci, Michael Schmidt, Pamela Pegman, Patricia Lowden, Christian Dombrowski; Use of galectin-3 fusions to extend the surface residence time of proteins topically applied to the eye. Invest. Ophthalmol. Vis. Sci. 2013;54(15):1090. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
One of the key limiting parameters in the penetration and effectiveness of topically applied ocular therapeutics is their mean residence time. Most of an instilled solution clears through drainage, with first-order clearance about 4 times that of bulk tear flow (~5 ul/min). For protein drugs, the problem becomes more acute due to the additional burden of reduced penetrance through the cornea and conjunctiva. Residence time might be increased however by fusing a protein of interest to a mucosal surface binding protein, such as galectin-3 (gal-3). Gal-3 is a small pentameric ocular surface resident protein that cross-links O-type mucins.
We used Gaussia luciferase (luc) as our test protein, fused at its N-terminus to one or 2 copies of the carbohydrate binding domain of gal-3 (gal). Fusions were expressed in HEK cells and purified by IMAC chromatography, and evaluated mucin binding by ELISA and by their ocular surface half-lives on mouse eyes and rabbit corneas, using luciferase activity as the readout.
We compared luc, luc-gal and luc-gal-gal in a variety of assays. Addition of gal-3 moieties to luciferase did not affect either the specific or relative activities of luciferase. On ELISA plates coated with MUC1 and asialofetuin, one gal-3 copy increased the bound luciferase by 10-20-fold, and two copies increased it by 200-250-fold. Binding to isolated rabbit corneal punches after a 30-min wash was increased by 6 and 12-fold with 1 or 2 gal copies, respectively. 30 mins after topical instillation in mouse eyes, luc and luc-gal had near background luciferase activity in enucleated eyes, while luc-gal-gal was 95-fold higher. In a time-course study, luc was lost from the surface by 15 min, luc-gal by 30 min, while luc-gal-gal required 180 min or more to reach background levels. The t1/2 of luc-gal-gal was at least 6-fold higher than that of luc. We also showed that gal3 can be readily fused to therapeutic moieties, including our novel 17kD chimeric IL-1 blocker which is the active substance in our clinical development product EBI-005.
We have demonstrated that fusion of a protein to the small sugar-binding domain of gal-3 increases the half-life of its ocular surface residence by at least 6-fold. This might afford a means to increase the absorption of biologics from the ocular surface, reducing either the dose required, the frequency of administration, or both.
This PDF is available to Subscribers Only