Binding of 8-methoxypsoralen to human serum proteins and red blood cells

Serum binding of 8-methoxypsoralen (8-MOP) was studied by equilibrium dialysis. In therapeutic concentrations, 8-MOP binding in serum was high, 91.4%, and constant, indicating concentration-independent kinetics. This binding involved the two main proteins, human serum albumin and alpha 1-acid glycoprotein, in a saturable process with one class of binding sites (n) and affinity constants (Ka) of 1.295 X 10(4) mol/l and 2.115 X 10(4) mol/l, respectively. Binding to lipoproteins and gamma globulins was negligible and non-saturable in therapeutic concentrations, with nKa values of 0.35, 0.024, 0.013 and 0.0004 mumol/l for VLDL, LDL, HDL and gamma globulins, respectively. Inhibition of 8-MOP serum binding was observed with salicylic acid and indomethacin, but not with diazepam, warfarin or erythromycin. Over a range of therapeutic concentrations, the ration of 8-MOP concentration in red blood cells (RBCs) and in serum was constant at 20.3% and three times higher than would be expected if a simple diffusion of the 8-MOP plasma free fraction (fu) occurred. According to the measured and calculated parameters, simulations of 8-MOP blood binding in pathological states (hypoalbuminaemia with or without inflammation) showed variations of fu which were partially 'buffered' by RBCs. Simulation of 8-MOP protein binding at cutaneous interstitial fluid level showed that fu is approximately 30% and permitted prediction of a decrease of fu available to the epidermis in case of local or systemic inflammation. This may imply an increase in the minimum phototoxic dose relevant for PUVA and explain some cases of 'poor' responsiveness of psoriatic patients to PUVA therapy.