Relationship of hyperthermia-induced hemolysis of human erythrocytes to the thermal denaturation of membrane proteins

Abstract

Hemolysis of human erythrocytes as a function of time of exposure to 47.4–54.5°C was measured and correlated to thermal transitions in the membranes of intact erythrocytes as determined by differential scanning calorimetry (DSC). Curves of hemoglobin leakage (a measure of hemolysis) as a function of time have a shoulder region exhibiting no leakage, indicative of the ability to accumulate sublethal damage (i.e., damage not sufficient to cause lysis), followed by a region of leakage approximating pseudo-first-order kinetics. Inverse leakage rates ( D o) of 330-21 min were obtained from 47.4–54.5°C, respectively. A relatively high activation energy of 304 ± 22 kJ/mol was obtained for leakage, eliminating the involvement of metabolic processes but implicating a transition as the rate-limiting step. Membrane protein involvement was suggested by the very low rate (10 −2 of the rate from erythrocytes) and low activation energy (50 ± 49 kJ/mol) of hemoglobin leakage from liposomes containing no membrane protein. A model was developed that predicts a transition temperature ( T m) for the critical target (rate-limiting step) of 60°C when measured at a scan rate of 1 K/min. DSC scans were obtained from intact erythrocytes and a procedure developed to fit and remove the transition for hemoglobin denaturation which dominated the scan. Three remained (Transitions A, B, and C) with T m values of 50.0, 56.8, and 63.8°C, respectively. These correspond to, but occur at slightly different temperatures than, the A, B, and C transitions of isolated erythrocyte membranes in the same salt solution ( T m = 49.5, 53–58, and 65.5°C, respectively). In addition, the relative enthalpies of the three transitions differ between isolated membranes and erythrocytes, suggestive of membrane alterations occurring during isolation. Thus, all analyses were conducted on DSC scans of intact erythrocytes. The B transition is very broad and probably consists of several transitions. An inflection, which is seen as a distinct peak (transition B3) in fourth-derivative curves, occurs at 60.8°C and correlates well with the predicted T m of the critical target. Ethanol (2.2%) lowers the T m of B3 by 4.0–4.5 K, close to the shift of 3.3 K predicted from its effect on hemolysis. Glycerol (10%) has very little effect on both hemolysis and the T m of B3, but it stabilizes spectrin ( ΔT m = 1.5 K) against thermal denaturation. Thus, we propose a model for hyperthermic hemolysis in which the major denaturation of spectrin ( T m = 50°C), while possibly necessary for lysis, is not sufficient, and that the rate-limiting step is transition at 60°C, possibly due to the denaturation of an additional membrane protein.