Would you expect a carboxylic acid group within the water-free interior of a protein to have a higher or lower Ka than it would have if it occured on the protein's surface, where it is hydrated. I think it is higher due to the oxygen H-bonding with the water.