Search

Five proteins from human placenta capable of inhibiting pancreatic phospholipase A2 were purified. Two of these proteins were identified as lipocortins I and II. The other three proteins were immunologically distinct from lipocortins I and II and had apparent subunit molecular masses of 32, 33, and 73 kDa as estimated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Amino acid sequence analysis of peptides produced by cyanogen bromide digestion indicated sequence homology of these proteins with lipocortin I and the heavy chain subunit of lipocortin II. Two of these proteins were identified as endonexin II and 67-kDa calelectrin. The third protein appears to be the human form of bovine endonexin I, also characterized as porcine protein II. Sedimentation equilibrium analysis of lipocortin I, endonexin I and II, and the 67-kDa calelectrin suggested monomer-dimer equilibria with dissociation constants in the range of 0.33-1.3 X 10(-3) M and monomer molecular masses of 38,050, 36,400, 36,850, and 73,610 Da, respectively. Self-association of lipocortin II was described by dimerization of a protomer (K12 = 5.3 x 10(-7) M), followed by an indefinite self-association of the dimer (isodesmic dissociation constant, Kiso = 3.6 x 10(-6) M). The protomer molecular mass was 48,800 Da, consistent with a heterodimeric structure composed of one heavy (38,600 Da) and one light (10,944 Da) chain as previously characterized for lipocortin II. Sedimentation equilibrium analysis of mixtures of individual protein inhibitors and purified pancreatic phospholipase A2 indicated weak association between enzyme and inhibitor (Kd greater than or equal to 3 x 10(-5) M), insufficient to account for the observed inhibition of enzyme activity.

The kinetics of insulin binding to isolated rat epididymal fat cells was investigated at 15 degrees C, at which temperature the system was simplified by the absence of lysosomal insulin degradation. The data were fit by maximum likelihood criteria with differential equations describing a number of models for the interaction of insulin and cells. Among those models that yielded a fit, the selection criteria were minimization of the Akaike information criterion and compatibility of the overall equilibrium constant for the system calculated from rate constants with the previously obtained experimental value. The results of the analysis indicated that insulin, I, first reversibly bound to cell surface receptors, R, whereupon this initial insulin-receptor complex, RI, reversibly altered its state or cellular location to R'I, according to the following equation. (Formula: see text) No evidence was found that insulin could either associate or dissociate from R'I directly. The association rate constant was kappa 12 = 1.6 x/divided by 1.4 X 10(5) liter mol-1 s-1, a value shown to be incompatible with diffusion control. The other rate constants were: kappa 21 = 3.4 x/divided by 1.6 X 10(-3) s-1, kappa 23 = 3.2 x/divided by 1.5 X 10(-4) s-1, and kappa 32 = 2.0 x/divided by 1.5 X 10(-4) s-1. From these rate constants, an equilibrium constant of 8.4 x/divided by 1.5 nM was calculated, in excellent agreement with the previously measured value of 8.8 x/divided by 1.3 nM (Lipkin, E. W., Teller, D. C., and de Haën, C. (1986) J. Biol. Chem. 260, 1694-1701). The kinetic analysis also yielded receptor numbers similar to those obtained by equilibrium binding studies. The nature of the R'I state is discussed in terms of an internalized state, in terms of insulin receptor complex in caveolae, in terms of receptor aggregates, and in terms of being a Michaelis complex between insulin bound to the receptor and cell surface-bound insulin protease.

Equilibrium binding of insulin to isolated rat epididymal fat cells was investigated. A temperature of 15 degrees C was chosen for the study to minimize lysosomal degradation of insulin. Indeed, medium insulin lost only 1% of its precipitability in trichloroacetic acid during the 4-h incubation required to approach equilibrium. Binding was measured by a method that did not perturb the equilibrium of the system. A new formalism for analyzing binding data in general was introduced. A correction for trapping of insulin in the interstitial space of cell pellets was both necessary and sufficient to derive specific binding data from raw observations. Thus, so-called “nonspecific binding” was unmasked as a misnomer, and the expression “correction for trapping” was proposed as a substitute. Equations for one and two independent classes of binding sites were fit to the data by the method of maximum likelihood, and the best fit was selected based on Akaike's information criterion, as adapted for a constant fractional error. More than 99.7% of the binding sites were found to be describable by a simple binding isotherm with Kd,app = 8.8 multiplied by over divided by 1.3 nM. Less than 0.3% sites had a higher affinity (Kd approximately equal to 8 multiplied by over divided by 3 pM). There were 99,000 x/divided by 1.6 binding sites/cell. These equilibrium parameters are in agreement with values derived from a kinetic analysis, presented in the subsequent paper (Lipkin, E. W., Teller, D. C., and de Haën, C. (1986) J. Biol. Chem. 260, 1702-1711).

Insulin binding to isolated rat white adipocytes at 15 degrees C, a temperature at which cellular degradation of insulin is negligible, has been found to be described by the Two-step Binding Model: R + I in equilibrium RI in equilibrium R′I (Lipkin, E. W., Teller, D. C., and de Haën, C. (1986) J. Biol. Chem. 261, 1702-1711). RI is the initially formed complex between the receptor, R, and insulin, I, and R′I is the complex in an altered state or cellular location. Here the possibility was examined that R′I results from disulfide exchange between the receptor and insulin, an exchange proposed by Clark and Harrison (Clarke, S., and Harrison, L. C. (1986) J. Biol. Chem. 257, 12239-12244) to occur at 37 degrees C. A number of sulfhydryl reagents representing various chemical reactivities did not affect insulin binding. The exception was 5,5′-dithiobis-(2-nitrobenzoic acid) (DTNB), which enhanced the number of insulin-binding sites up to 2-fold with no effect on the equilibrium constant. The data suggested that this enhancement was due to activation of cryptic binding sites pre-existing on the cell surface, possibly by increasing the valency of the receptor from 1 to 2. Insulin binding was also enhanced by structural congeners of DTNB devoid of sulfhydryl reactivity, the simplest one being benzoic acid. It was concluded that the effects were not related to modification of sulfhydryl groups, that modification of sulfhydryl groups on the receptor either did not take place or was without effect on binding, and finally, that disulfide exchange between insulin and the receptor was an unlikely explanation for the formation of R′I. Also, since it is possible to show insulin action at 15 degrees C, contrary to the proposal by Clark and Harrison (Clark, S., and Harrison, L. C. (1983) J. Biol. Chem. 258, 11434-11437), disulfide exchange does not appear to be necessary for signal transmission by the occupied receptor.