Ion-exchange equilibria of lysozyme, myoglobin and bovine serum albumin. Effective valence and exchanger capacity.

A series of ion-exchange equilibrium experiments were completed and analyzed by a mass action model. The isotherms at various counter-ion concentrations are regressed to determine the average number of binding sites, Z, the mass action constant, K, and the effective capacity of the exchanger, Pm. The estimates for Pm are 1-10 mmol of protein adsorbed per liter of packed Sepharose exchanger, in agreement with values determined from fitting the Langmuir isotherms. A geometric model, developed to relate the charge distribution of the exchanger and the size of the protein to the protein capacity, also confirms the estimates of Pm. The Z values estimated (and confirmed with an independent slope method) range from 0.8 to 3.6, indicating that most of the ionized sites on the protein do not undergo ion exchange. Geometric model estimates of exchanger sites covered by a protein at maximum surface coverage show that only 1-32% of the sites actually undergo ion exchange. This result indicates that adsorbed protein molecules "block" many exchanger sites, preventing other molecules from ion exchanging at those sites. This high degree of blocking by proteins also indicates that there is an excess of exchanger sites, as evidenced by Z showing no dependence on charge separation distance over the range 6-10 A. Scatchard pots of the equilibrium data indicate that there is competition between at least two different binding forms of each protein studied. For the high-affinity systems, cooperative binding was observed at low exchanger loading; the character then changed to heterogeneous competitive binding as the exchanger loading increased. The effect on an isotherm plot is a subtle but systematic deviation from the Langmuir and mass action models. More important, the competition means that Z represents an average for competing binding forms and, as such, Z can be fractional.