A thermodynamic molecular switch in biological systems: ribonuclease S' fragment complementation reactions.

It is well known that essentially all biological systems function over a very narrow temperature range. Most typical macromolecular interactions show DeltaH degrees (T) positive (unfavorable) and a positive DeltaS degrees (T) (favorable) at low temperature, because of a positive (DeltaCp degrees /T). Because DeltaG degrees (T) for biological systems shows a complicated behavior, wherein DeltaG degrees (T) changes from positive to negative, then reaches a negative value of maximum magnitude (favorable), and finally becomes positive as temperature increases, it is clear that a deeper-lying thermodynamic explanation is required. This communication demonstrates that the critical factor is a temperature-dependent DeltaCp degrees (T) (heat capacity change) of reaction that is positive at low temperature but switches to a negative value at a temperature well below the ambient range. Thus the thermodynamic molecular switch determines the behavior patterns of the Gibbs free energy change and hence a change in the equilibrium constant, K(eq), and/or spontaneity. The subsequent, mathematically predictable changes in DeltaH degrees (T), DeltaS degrees (T), DeltaW degrees (T), and DeltaG degrees (T) give rise to the classically observed behavior patterns in biological reactivity, as may be seen in ribonuclease S' fragment complementation reactions.

Thermodynamic molecular switch in macromolecular interactions.

It is known that most living systems can live and operate optimally only at a sharply defined temperature, or over a limited temperature range, at best, which implies that many basic biochemical interactions exhibit a well-defined Gibbs free energy minimum as a function of temperature. The Gibbs free energy change, deltaG(o) (T), for biological systems shows a complicated behavior, in which deltaG(o)(T) changes from positive to negative, then reaches a negative value of maximum magnitude (favorable), and finally becomes positive as temperature increases. The critical factor in this complicated thermodynamic behavior is a temperature-dependent heat capacity change (deltaCp(o)(T) of reaction, which is positive at low temperature, but switches to a negative value at a temperature well below the ambient range. Thus, the thermodynamic molecular switch determines the behavior patterns of the Gibbs free energy change, and hence a change in the equilibrium constant, Keq, and/or spontaneity. The subsequent, mathematically predictable changes in deltaH(o)(T), deltaS(o)(T), deltaW(o)(T), and deltaG(o)(T) give rise to the classically observed behavior patterns in biological reactivity, as demonstrated in three interacting protein systems: the acid dimerization reaction of alpha-chymotrypsin at low pH, interaction of chromogranin A with the intraluminal loop peptide of the inositol 1,4,5-triphosphate receptor at pH 5.5, and the binding of L-arabinose and D-galactose to the L-arabinose binding protein of Escherichia coli. In cases of protein unfolding of four mutants of phage T4 lysozyme, no thermodynamic molecular switch is observed.

New thermodynamic studies on ribonuclease A at low pH.

The Planck-Benzinger thermal work function, delta W zero(T), represents the heat-flux term which is responsible for breaking or forming the noncovalent bonds in macro-molecular interactions, while the temperature-invariant chemical bond energy, delta H zero(T zero), gives molecules the cohesiveness to form macromolecular structures. A method is described for evaluating delta H zero(T zero) at different temperatures for ribonuclease A at low pH in the standard state, in order to determine the effect of pH on the thermodynamic stability of this protein. Ribonuclease A at pH 1.13 has a delta H zero(T zero) value of 58 kcal mol-1; at pH 2.50, delta H zero(T zero) is 58, at pH 2.77, 59, and at pH 3.15, 60 kcal mol-1. In the conformational thermal transition of ribonuclease A at pH 2.8, the compensatory temperature ranges from 50 to 320 K, and delta H zero(T zero) is approximately 5-6 kcal mol-1. This widening temperature range is typical of the unfolding process. Such differences in the magnitude of the temperature-invariant chemical bond energy can be attributed to specific changes in the solvent ordering in the immediate domain of ribonuclease A.

Molecular conformation of ubiquitinated structures and the implications for regulatory function.

The molecular conformation of ubiquitinated structures and the validity of the N-end rule were examined by simulating the molecular mechanics to ascertain the global energy-minimized structure. We examined the chemical linkage involved in attaching the ubiquitin carboxyl terminus to the N-terminus of three different x-hexapeptides, where x is the amino group of the acceptor peptide--either valine, arginine or glutamic acid--(x-K linkage) and to the epsilon-amino group of lysine of the acceptor hexapeptide x-glu1-his2-lys3-gly4-lys5-val6 (K-K linkage) through the formation of an isopeptide bond. Changes in conformation and molecular stability of the multi-ubiquitinated structures were determined by energy-minimization procedures using the SYBYL program developed by Tripos Associates. In the x-K linkage, the ubiquitin molecule is stretched in the beta-pleated sheets and beta-turns while the alpha-helices expand, as the molecule continues to unfold linearly. In the K-K linkage, the ubiquitin molecules have turned into a u-shaped, semi-circular alignment, contracting into a compact, folded structure.

The effect of bile salts on carbonic anhydrase.

Bile salts are potent inhibitors of bovine carbonic anhydrase and human carbonic anhydrase I and human carbonic anhydrase II. To further characterize the binding of bile salts to carbonic anhydrase, rate constants for the CO2 hydration reaction in the presence of deoxycholate, cholate, glycocholate and taurocholate were determined using stop-flow experiments. Values for the Michaelis-Menton dissociation constant for bovine carbonic anhydrase, human carbonic anhydrase I and human carbonic anhydrase II were found to be 5.2, 9.2 and 13.2 mmol/L, respectively. The inhibition constant values for the various bile salts tested ranged from 0.1 to 1 mmol/L for bovine carbonic anhydrase, 1.6 to 2.4 mmol/L for human carbonic anhydrase I and 0.09 to 0.7 mmol/L for human carbonic anhydrase II. Our results suggest a mechanism of noncompetitive carbonic anhydrase inhibition for bile salts. Bile-salt binding to carbonic anhydrases as measured by scanning molecular sieve chromatography resulted in an increase in partition radius, molecular volume and surface area. The partition radius increased from 24 A to 28 A in the presence of 2.5 mmol/L sodium deoxycholate at critical micelle concentration. As determined by sedimentation equilibrium measurements, approximately 1 gm of carbonic anhydrase will bind 0.03 gm of deoxycholate, suggesting three to six binding sites for bile salt on the carbonic anhydrase molecule. The conformational changes and inhibition of carbonic anhydrases resulting from bile-salt binding may be important to the regulation of enzymatic activity in tissues along the enterohepatic circulation; by limiting bicarbonate availability this interaction may also contribute to the metabolic derangements seen in patients with cholestatic liver disease.

Molecular mechanics of the formation of cholic acid micelles.

The molecular mechanics of cholic acid micelle formation were simulated using the Sybyl energy minimization program (MAXIMIN2), developed by Tripos Associates, interfaced with micro-Vax. Before energy minimization, the molecular dimensions of the cholic acid dodecamer C24H40O6, in terms of the unit cell axes a, b, and c in the cubic crystal class, had values of 13, 18, and 6.7 A, respectively. After energy minimization, at 9370 kcals/dodecamer, these values had increased to 21.6, 42.8 and 20.9 A. At an energy minimization level of 21,626 kcals/dodecamer, the micelle structure is stabilized by hydrophobic interaction, forming distinct horizontal channels along the b-axis, directing the carboxyl and hydroxyl groups toward the surface. These structural changes remain relatively constant as the process of energy minimization continues, down to the lowest energy level we considered, 9370 kcals/dodecamer. The cholic acid layers are highly dissimilar, forming channels of irregular size and shape in a somewhat helical structure. The carboxyl groups and phenanthrene rings are in a puckered orientation, which permits compact packing of the sandwiched multilayers. From the dimension of the channels, it is apparent that guest molecules, such as phospholipid, cholesterol, or inorganic calcium, can be incorporated into the micelle through more than one channel, forming inclusion complexes, such as gallstones.

Quantitative reappraisal of general expressions for multivalent protein binding in subunit-exchange chromatography.

Quantitative expressions have been derived for bivalent equilibria with immobilized ligand systems and for the equilibria for an immobilized protein whose self-association is modified by binding with a soluble ligand, as analyzed by affinity chromatography. These general expressions have been applied in a reexamination of multivalency in the affinity chromatography of antibodies, as reported by Eilat and Chaiken (Biochemistry 18 (1979) 790) and also to studies of neurophysin-peptide hormone interactions using glass matrices reported by Swaisgood and Chaiken (Biochemistry 25 (1986) 4148).

Purification, secretion and immunocytochemical localization of the uterine milk proteins, major progesterone-induced proteins in uterine secretions of the sheep.

Restriction of the conceptus to one uterine horn of the pregnant ewe results in the accumulation of fluid called uterine milk (UTM) in the contralateral horn. Two basic polypeptides, called the uterine milk proteins (UTM-proteins; Mr = 55,000 and 57,000 as determined by polyacrylamide-gel electrophoresis using sodium dodecyl sulfate), accounted for the majority of the protein in uterine milk. The two UTM-proteins were glycoproteins and were readily purified from uterine fluids by cation-exchange chromatography on carboxymethyl (CM)-cellulose followed by Sephacryl S-200 gel-filtration. The purified UIM-proteins had a weight-average molecular weight of 50,700 +/- 4,200, as determined by equilibrium sedimentation analysis. Endometrial explants from pregnant ewes were cultured in the presence of radioactive amino acids and released UTM-proteins into the medium as their major secretory products. The UTM-proteins were secreted into the uterine lumen of nonpregnant, ovariectomized ewes given daily injections of progesterone. Estrone alone was ineffective in inducing UTM-protein production. Immunocytochemical studies indicated that synthesis of the UTM-proteins was confined to the surface and glandular epithelium of the uterus.

Interaction of human low density lipoprotein and apolipoprotein B with ternary lipid microemulsion. Physical and functional properties.

Based on data from sedimentation velocity experiments, electrophoresis, electron microscopy, cellular uptake studies, scanning molecular sieve chromatography using a quasi-three-dimensional data display and flow performance liquid chromatography (FPLC), models for the interaction of human serum low density lipoprotein (LDL) and of apolipoprotein B (apo B) with a ternary lipid microemulsion (ME) are proposed. The initial step in the interaction of LDL (Stokes radius 110 A) with the ternary microemulsion (Stokes radius 270 A) appears to be attachment of the LDL to emulsion particles. This attachment is followed by a very slow fusion into particles having a radius of approx. 280 A. Sonication of this mixture yields large aggregates. Electron micrographs of deoxycholate-solubilized apo B indicate an arrangement of apo B resembling strings of beads. During incubation, these particles also attach to the ternary microemulsion particles and, upon sonication, spherical particles result which resemble native LDL particles in size. Scanning chromatography corroborates the electron microscopy results. By appropriate choice of display angles in a quasi-three-dimensional display of the scanning data (corrected for gel apparent absorbance) taken at equal time intervals during passage of a sample through the column, changes in molecular radius of less than 10 A can be detected visually. Such a display gives a quantitative estimate of 101 +/- 2 A for these particles (compared to 110 A for native LDL). The LDL-ME particles and apo B-ME particles compete efficiently with native LDL for cellular binding and uptake. Cellular association studies indicate that both LDL- and apo B-ME particles are effective vehicles for lipid delivery into cells.

Scanning molecular sieve chromatography of interacting protein systems: nonlinear least-squares analysis of small zone experiments.

A simple routine for nonlinear least-squares analysis is applied to small zone scanning data, where calibration of the gel column requires the use of fully characterized markers of known molecular size. Application of nonlinear least-squares analysis eliminates the difficulty encountered because of scarcity of calibrating markers for gels whose porosities span certain size ranges by providing a sensitive measure of changes in molecular size of the interacting protein system, in order to determine the interaction parameters, size heterogeneity, and the centroid position for time difference chromatographic experiments.

Low density lipoprotein receptor regulation. Kinetic models.

The macromolecular species distribution in a receptor-mediated endocytotic pathway was computer simulated based on kinetic data reported in the literature. In the proposed model, the rapidity with which the recycled receptor is shuttled to the cell surface is indicated by the magnitude of k-3, the shuttling constant. The magnitude of k-3 will vary with the experimental conditions, but when this value is large, the internalized receptor is shuttled back to the cell surface with a traverse time of 14 min. Under steady-state conditions, after the cells have been incubated in the presence of LDL for 5 h (M.S. Brown and J.L. Goldstein, Cell 9 (1976) 663), the time required for a receptor to traverse the entire endocytotic pathway is 52 min. Our simulation suggests that normal LDL binding in such a short-term experiment may be independent of receptor synthesis. Thus, the degradation of LDL and resultant build-up of cholesterol would have no apparent inhibitory effect on the down-regulation of receptor synthesis.

Structural studies on apolipoprotein B: controllable heterogeneity of the complex formed with the surfactant, Triton X-100.

Apolipoprotein B complexed with Triton X-100 (T-ApoB) has been isolated from human low density lipoprotein (LDL). Preparations are heterogeneous when analyzed by sedimentation velocity, with a major 12 S species and minor 17 S species present. The 12 S T-ApoB complex possesses a molecular weight of 880,000 containing 400,000 daltons of protein. Hydrodynamic measurements on this complex are consistent with a prolate ellipsoid model having an axial ratio of 13:1 and 0.22 g/g of bound water. Heterogeneity results from the irreversible aggregation of 12 S complexes into discrete 17 S and faster sedimenting components. A significant finding is that three determinants of this T-apoB heterogeneity could be elucidated and controlled. First, the initial state of aggregation is mainly influenced by the technique by which Triton and LDL are mixed. Second, once isolated, T-ApoB complexes slowly but spontaneously undergo further aggregation at 4 degrees C; the rate and extent of aggregation is enhanced remarkably with increasing temperature. Finally, reagents that unfold and expose protein structure (perchlorate, thiocyanate, and reducing reagents) lead to increased aggregation. The ability to control heterogeneity carries important implications for other studies concerning interactions of apoB with surfactants and lipids.

Up- and down-regulation of insulin receptors. Kinetic models.

A theoretical model for insulin receptor synthesis and degradation in differentiating 3T3-L1 adipocytes is described. This three-step irreversible ordered sequence model explains the up- and down-regulation of receptors in terms of the level of insulin concentration. Kinetic expressions were derived for the model. Numerical solutions for these equations, based on data reported by Reed and Lane (Reed, B.C., and Lane, M. D. (1980) Proc. Natl. Acad. Sci. U.S.A. 77, 285-289) were used for computer-generated curves illustrating insulin-dependent receptor synthesis and degradation. Results show that this model provides the best fit to the reported data and lend support to the suggestion that the free recycled receptor may differ from the newly synthesized receptor. A possible role for the recycled receptor in signal modulation is suggested.

Interaction of human serum apolipoprotein B with sodium deoxycholate.

Preparation of apolipoprotein B (Apo B)-deoxycholate (DOC) complexes by gel filtration chromatography in the presence of 20 mM DOC, pH 8.5, gave two populations of particles with 5% (peak I) and 13% (peak II) lipid remaining bound. These complexes were initially shown to be very large and elongated, with partition radii of approx. 131 +/- 0.5 A, weight average molecular weights of approx. 164 000 +/- 1 000, and an intrinsic viscosity of 80.19 +/- 2.21 ml/g. Additionally, they appeared very similar to native low-density lipoprotein on sodium dodecyl sulfate-polyacrylamide gels, giving one major band. Incubation of these samples for 10 days under nitrogen at 4 degrees C in the presence of antibiotics and protease inhibitor resulted in dissociation to many smaller subunits. Results of scanning molecular sieve chromatography and analytical ultracentrifugation showed that dissociation of these complexes was relatively slow and indicated the presence of at least two classes of components in fresh samples: one a very elongated complex with a radius directly correlated to the DOC/Apo B ratio and inversely correlated to sample aging; and another of much smaller radius which was independent of DOC/Apo B ratio but directly correlated to sample aging; indicating that these dissociated subunits interact with each other to an appreciable extent. Furthermore, these complexes were found to undergo a preferential hydration upon interaction with DOC, which may contribute to large changes in their effective specific volumes, as well as to dissociation of subunits.

Human serum low density lipoprotein-sodium deoxycholate interaction.

The binding of sodium deoxycholate to low density lipoprotein (LDL) is a relatively fast process, as compared to the lipid displacement reaction and solubilization. The initial reaction has a bimolecular rate constant of approximately 539 M-1 S-1. In the presence of 1 mM sodium deoxycholate, below critical micelle concentration, 0.04 g of sodium deoxycholate was bound per g of LDL, resulting in an increase in particle radius from 102 to 128 +/- 0.3 A and an axial ratio of 5.6 for a prolate ellipsoid. Such increases were also observed using higher sodium deoxycholate concentrations and are apparently due to sodium deoxycholate-induced elongation or distortion of certain portions of LDL.

The effect of the propagation coefficient on size distribution in micellar systems.

In this study the effect of the propagation coefficient on the molar distribution function in a modified shell model for micellar systems was examined. The sharpness of the micelle size distribution boundary was found to depend less on the degree of polymerization, n, than on the propagation coefficient, P. Although Kegeles (J. Phys. Chem. 83 (1979) 1728) has reported a marked sharpening of the distribution boundary when P = 2.0. we found the boundary to be fairly broad at this point. However, as values of the propagation coefficient were increased from 3 to 10, the micelle distribution boundary became increasingly sharp. The possibility of such a change in the reaction boundary arising from a structural transition, accompanied by a change in the rate of dissociation of monomer from the shell, is also discussed.

Iron content and molecular weight of uteroferrin and a comparison of its iron and copper forms.

The molecular weight of the purple, iron-containing glycoprotein uteroferrin has been determined to be 35140 +/- 1610 by sedimentation equilibrium centrifugation. This is consistent with measurements carried out by other procedures. The dry weight of a solution of uteroferrin with an absorption of 1.00 cm -1 at 280 nm was 0.87 mg/ml. Highly purified uteroferrin has a ratio of absorbance at 280 nm to 545 nm of about 13.2 and a revised extinction coefficient at 545 nm of 3.1 . 10(3) M -1 is presented. The iron content of uteroferrin has been determined both colorimetrically and by atomic absorption spectroscopy. One iron atom is present per polypeptide chain. Reduction with dithionite leads to loss of iron, allowing the apoprotein to be prepared. Enzymatic activity can be restored to the apoprotein with Fe(III) or with Cu(II). The copper enzyme has no visible color and has a higher pH optimum than the Fe-uteroferrin. Whereas the phosphatase activity of the latter is increased several-fold by beta-mercaptoethanol, reduction inactivates Cu-uteroferrin. Both forms of uteroferrin have similar Km values for p--nitrophenyl phosphate and are inhibited by molybdate but not by tartrate. Excess Cu(II) and Fe(III) strongly inhibit uteroferrin phosphatase activity and these results may explain the failure of other to restore activity to the apoprotein using Cu(II) and Fe(III) salts.

Density and composition models for lipoproteins.

Two theoretical models for lipoprotein particle structure are proposed, based on density and composition and utilizing the Stokes' coefficient as evaluated from the Maude-Whitemore expression, using analytical ultracentrifugation. These density and composition models are compared with the Corey-Pauling-Koltun space-filling model proposed by Verdery and Nichols, and that proposed by Shen et al., which has a sharply defined boundary between the hydrophobic core and the amphipathic layer surrounding it. Results of our calculation of the percentage of total apolar lipid accommodated in the outer hydrophobic core -28, 64 and 94% for LDL, HDL2 and HDL3, respectively - suggest that there may be differences among these three lipoproteins classes in both the packing of cholesteryl ester and triglyceride as well as the conformation of apolipoproteins at the particle's surface.

The dispersion of gram-negative lipopolysaccharide by deoxycholate. Subunit molecular weight.

The lipopolysaccharide (LPS) was isolated from three strains of Salmonella typhimurium, a "smooth" strain, STM 7, the Ra mutant, TV 119, and the Re mutant, SL 1102. The effect of depletion of divalent cations on structure and the effect of deoxycholate on hydrodynamic behavior were studied. The results confirmed previous work by others that divalent cations and hydrophobic forces are important factors influencing LPS size and morphology. The binding of deoxycholate to LPS was measured. When the weight average molecular weights of the deoxycholate-dissociated LPS were determined by sedimentation equilibrium and corrected for bound deoxycholate, the values 5,555, 10,607, and 15,592, respectively, for Re, Ra, and "smooth" LPS were in good agreement with calculated formula weights. Although others have suggested that the basic LPS subunit is a trimer, our results suggest that it exists in the dimeric form.