SynopsisA method is presented which allows one to calculate a distribution of sedimentation coefficients from the boundaries of sedimentation-velocity experiments with mono-or paucidisperse solutes. With two solutes differences in S value as small as 20% can be resolved. In absence of heterogeneity and concentration-dependent effects, the analysis also provides values for the diffusion coefficient within an accuracy of -10 to +5%. Tests with both simulated data and ultracentrifugation experiments on short DNA fragments show the value and the limitations of the method.

Thermal denaturation of very homogeneous preparations of core particles from chicken erythrocyte chromatin is studied by several techniques. The change in absorbance, which is very closely paralleled by changes in heat capacity, is a biphasic process with inflexions at 600C and 740C. In contrast, isolated DNA of the same length denatures in a single transition around 440C. Monitoring the circular dichroism of the cores during thermal denaturation reveals biphasic changes in the secondary structure of the DNA, preceding the base unstacking by 100C in the first and 30C in the second phase. However, measurable alterations in the secondary structure of the histones are confined to the second phase with a melting temperature at 710C. Increase in the ionic strength of the buffer from 1 mM to 10 mM leads to almost monophasic melting curves as measured by absorbance and CD, while not causing any measurable conformational changes at room temperature. The melting of core particles is interpreted as a denaturation of about 40 base pairs in the first phase, followed by a massive breakdown of the native structure of a tight histone-DNA complex, which frees the remaining 100 base pairs for unstacking.

We have removed histone H1 specifically from calf thymus nuclei by low pH treatment, and studied the digestion of such nuclei in comparison with undepleted nuclei. By a number of criteria the nuclei do not appear damaged. The DNA repeat-length in nuclear chromatin is found to be the same (192 +/- 4 bp) in the presence or absence of H1. These experiments demonstrate that the core histone complex of H2A, H2B, H3, and H4 can itself protect DNA sequences as long as 168 bp from nuclease. Our interpretation is that this represents an important structural element in chromatin, carrying two full turns of superhelical DNA. Depending on conditions of digestion this 168 bp fragment may be metastable and is normally rapidly converted by exonucleolytic trimming to the well-known "core-particle" containing 145 bp. Larger stable DNA fragments observed indigestion of H-1 depleted nuclei appear to arise from oligomers assembled from 168 bp cores in close contact exhibiting trimming of 0-20 bp at the ends. Electrophorograms of undepleted nuclear digests reveal oligomer bands in several size classes, each corresponding to one or more combinations of 168 bp particles, H1-protected spacers of about 20 bp length, and particles with ends trimmed to varying degrees.

Binding to the alpha subunit of tryptophan synthase induces extrinsic Cotton effects in the substrates indole (IND), indoleglycerol phosphate (IGP), and D-glyceraldehyde-3-P (D-GAP) and in the inhibitor indolepropanol phosphate (IPP). These effects disappear when the enzyme is denatured in guanidinium chloride. The induced circular dichroism (CD) was used to determine the dissociation constant and the number of binding sites for IPP. The dissociation constant so determined is equal to 48 muM and is in good agreement with the value of 48 muM obtained by equilibrium dialysis. From the temperature dependence of the dissociation constant, a value of -2.8 kcal/mol for the binding enthalpy was obtained. The determination of dissociation constants by means of extrinsic Cotton effects is shown to be quite feasible. CD competition experiments with glycerol phosphate (GP) suggest that IPP binds bifunctionally to the enzyme: via its indole part and its phosphate group. Indolepropanol, which lacks the phosphate group, does not show an extrinsic Cotton effect. Since the induced CD is strongly dependent on the binding geometry, the close similarity between the induced spectra in IPP and IGP is additional evidence that IPP is a good substrate analog. Binding to the enzyme results in a blue shift of the IPP fluorescence emission maximum. The dissociation constant determined by fluorescence titration equals 46 muM and agrees well with the values determined by the other two methods. Previous biochemical and fast kinetic studies suggested the existence of multiple conformational states for the enzyme and of ligand-induced conformational changes. No evidence was found in the far-uv CD spectra for conformational changes upon binding of IND and D-GAP. For IPP a very small effect was observed.

The binding of indole and indolepropanol phosphate, an analogue of the substrate indoleglycerol phosphate, to the individual a and P,-subunits and to the a2 8,-cornplex of tryptophan synthase was studied by equilibrium dialysis. The use of ['4C]indole and indolepropanol [32P]phosphate permitted simultaneous binding studies to be carried out. Competition between indole and indolepropanol phosphate in binding to a particular site was taken as evidence for that site being part of the active site of the a-subunit.The binding of indole to the active site of the a-subunit is weak (Kd = 18 mM). A second distinct site binds indole more strongly (Kd = 1.5 mM) and interacts with the active site indirectly. It is therefore designated an effector site. Furthermore, the binding of indole and/or indolepropanol phosphate appears to stabilize different conformations of the a-subunit. The B,-subunit binds indole only weakly (Kd = 12 mM) to many (n = 10) sites per polypeptide chain. The a,P,-cornplex retains one or two sites per ap-equivalent of relatively high affinity (Kd = 1.2 mM). The active sites of the component a and 8-subunits probably belong to the second class of many (n = 40) sites of low (Kd = 30 mM) affinity for indole. These findings support conclusions from the literature that both bi-substrate reactions involving indole catalyzed by tryptophan synthase and its subunits must follow strictly ordered addition mechanisms with the respective other substrate adding first. +L-tryptophan + H20. (reaction 3 )The a-subunit is capable of catalyzing reaction (2) and the P,-subunit reaction (3) albeit only 1 -2 % as efficiently as the a,P,-complex. Reactions (2) (3) can formally be considered to be partial reactions of the physiologically relevant reaction (I), which can be catalyzed only by the native a,P,-complex [I]. This hypothesis implies that indole is a free intermediate. However, no indole could be detected in various trapping experiments designed to prove its transient accumulation in the course of the reaction (1) [ 2 , 3 ] . These findings have been interpreted in terms of a channelling effect, whereby the particular organization of a hypothetical composite active site [4] precludes the equilibration of indole with the bulk solvent. However, this is not a unique explanation and more direct evidence pertaining to the relationship between the partial reactions (2) and ( 3 ) and the overall reaction (1) appears desirable. In particular, it is pertinent to ask whether the indole subsites in the active sites of the a and the P2-subunits are retained and/or modified in the a,~,-complex. Such studies would also provide ancillary evidence for interpreting steady-state kinetic data.