Resolving multiple overlapping calorimetric transitions by use of a microcomputer: studies on erythrocyte membranes.

Buffer changes and certain drugs cause temperature shifts, amplitude changes, and transition broadening in the differential scanning calorimetric (DSC) analysis of erythrocyte membranes. However, it has been difficult to interpret and quantitate these shifts and changes because the scans are composed of multiple overlapping transitions and because more than one transition may be simultaneously affected. An empirical approach has been developed by using Gaussian modeling to resolve these calorimetric transitions. Data analysis was carried out on a microcomputer using a nonlinear regression program (PCNONLIN) to fit the data scans. These results show that changes in the calorimetric scans of erythrocyte membranes due to alterations in the buffer environment, such as pH and osmolarity, can be resolved by fitting the data scans with the proposed mathematical model and optimizing the resolution parameters with PCNONLIN. In addition, resolution uncovered hidden characteristics that may not have been readily evident. Under certain conditions, for example, apparent transition shifts were shown to actually be amplitude changes and transition broadening. Determination of the limitations and validity of this method was accomplished with simulation studies. This technique offers a simple means for fitting overlapping DSC transitions by use of a commercially available nonlinear regression program that can be run on a microcomputer.