On the interaction of alpha-lactalbumin and galactosyltransferase during lactose synthesis.

The regulatory effect of alpha-lactalbumin in the lactose synthase system has been ascribed to its reversible association with a complex of galactosyltransferase with Mn2+ and UDP-galactose, prior to the binding of monosaccharides; the resulting complex has a higher affinity for various monosaccharides. Two steps in the postulated catalytic cycle have been investigated; UDP-galactose binding to enzyme-Mn2+ by equilibrium dialysis and alpha-lactalbumin binding to enzyme-Mn2+-UDP-galactose by sedimentation velocity and kinetics. There is a single binding site for UDP-galactose on the enzyme-Mn2+ complex, and the dissociation constant for UDP-galactose from enzyme-Mn2+-UDP-galactose was found to be 72 muM at 37 degrees. The formation of a complex between galactosyltransferase and alpha-lactalbumin in the presence of Mn2+ and UDP-galactose was observed as an increase in sedimentation coefficient of enzyme activity So20,w from 3.25 +/- 0.03 in the absence of alpha-lactalbumin to 4.22 +/- 0.03 at saturating concentrations of alpha-lactalbumin, a value closely similar to that of a cross-linked 1:1 complex of the proteins under the same conditions (4.35 +/- 0.03). No interaction was observed in the absence of substrates or with UDP-galactose and EDTA. From the ultracentrifuge data and steady state kinetics, dissociation constants for alpha-lactalbumin from the enzyme-Mn2+-UDP-galactose-alpha-lactalbumin complex were determined at several temperatures and salt concentrations. These showed good internal agreement. The free energy change delta G degrees for the association of the two proteins is calculated, and the results are discussed in relation to the nature of the interaction.

The regulatory effect of a-lactalbumin in the lactose synthase system has been ascribed to its reversible association with a complex of galactosyltransferase with Mn2+ and UDP-galactose, prior to the binding of monosaccharides; the resulting complex has a higher affinity for various monosaccharides. Two steps in the postulated catalytic cycle have been investigated; UDP-galactose binding to enzyme.Mn'+ by equilibrium dialysis and cY-lactalbumin binding to enzyme .Mn2-.UDP-galactose by sedimentation velocity and kinetics. There is a single binding site for UDP-galactose on the enzyme .Mn*+ complex, and the dissociation constant for UDP-galactose from enzyme .Mn2+ .UDP-galactose was found to be 72 FM at 37". The formation of a complex between galactosyltransferase and cu-lactalbumin in the presence of Mn2+ and UDP-galactose was observed as an increase in sedimentation coefficient of enzyme activity sOZO.w from 3.25 f 0.03 in the absence of cu-lactalbumin to 4.22 f 0.03 at saturating concentrations of a-lactalbumin, a value closely similar to that of a cross-linked 1:l complex of the proteins under the same conditions (4.35 f 0.03). No interaction was observed in the absence of substrates or with UDP-galactose and EDTA. From the ultracentrifuge data and steady state kinetics, dissociation constants for ol-lactalbumin from the enzyme .Mn2+ .UDP-galactose'cu-lactalbumin complex were determined at several temperatures and salt concentrations.
These showed good internal agreement. The free energy change AG" for the association of the two proteins is calculated, and the results are discussed in relation to the nature of the interaction.

Interaction
of the two components of lactose synthase (EC 2.4.1.22), cu-lactalbumin (1) and galactosyltransferase (2), is required for significant catalysis of lactose synthesis at physiological concentrations of glucose. In isolation, the galactosyltransferase catalyzes the transfer of galactose from UDP-galactose to free or protein-bound N-acetylglucosamine (2). Transfer to free GlcNAc' is inhibited by ol-lactalbumin at monosaccharide concentrations above 4 mM (2). The mechanism of the reactions catalyzed by galactosyltransferase in the absence of a-lactalbumin, appears to involve an ordered sequential addition of the substrates Mn2+, UDP-galactose, and monosaccharide, and an ordered release of products: disaccharide and MnUDP (3,4 According to previous kinetic studies of galactosyltransferase, Mn*+ is an obligatory first binding substrate to the enzyme, whose attachment directly precedes that of UDPgalactose (4,12). For the homogenous galactosyltransferase from human milk or bovine colostrum, with which these studies were performed, the role of Mn2+ in the reaction appears from steady state kinetics not to be as simple as was initially suggested (10,13), and there is a possibility that two manganous ions may attach to the enzyme during the catalytic cycle. While studies are currently in progress to clarify this, we have chosen for the purposes of the current study to work at a concentration of Mn2+ sufficient to saturate most of the enzyme.
The dissociation constant for the Mn2+ complex of bovine colostrum galactosyltransferase is 0.95 mM (12), and the concentration of Mn2+ used (10 mM) is sufficient to give more than 90% saturation.
The interpretation of the results must take this into consideration, as well as the fact that the dissociation constant for MnUDP-galactose is 7.5 mM (4)

Preparation of Cross-Linked
Lactose Synthuse Complex-Covalent cross-linking of colostrum galactosyltransferase and u-lactalbumin (bovine) was performed essentially as previously described (11) except that dimethysuberimidate was substituted for the pimelimidate as the bifunctional reagent, and 0.1 M borate, pH 8.1, replaced triethanolamine as the buffer. Cross-linking was allowed to proceed for 2 hours at 22" in the presence of 3 mM MnCl,, 0.3 mM UDP-galactose, and 105 MM a-lactalbumin.
The reaction was terminated by the addition of 0.2 volume of 0.5 M glycine, and the cross-linked lactose synthase was isolated by gel filtration on a column of Bio-Gel PI50 (1.5 x 87 cm) followed by affinity chromatography with a-lactalbumin-Sepharose (11). ously obtained with bovine colostrum galactosyltransferase at 37" was 25.5 FM (12). Similar initial velocity experiments were performed at three lower temperatures (27", 17.5", and 10.7"), and as the same characteristic kinetic patterns were obtained (i.e. the presence of slope but not intersect effects in double reciprocal plots of initial velocity and glucose concentration at a series of fixed concentrations of oc-lactalbumin) they were interpreted in the same way, to give the corresponding dissociation constants at the lower temperatures (Table II). As the binding of UDP-galactose is tighter at these lower temperatures, it is valid to assume that the dissociation constant for cu-lactalbumin need not be corrected for UDP-galactose concentration.
That such a correction is unnecessary at 37" has been previously demonstrated (4   b Obtained by extrapolation of the In K, versus I/T plot (Fig. 3).  Table I. 93 3.  constant K, (kinetic parameter K,, (4, 12)) against I/T (Fig. 3)  The plot of lactose synthase activity against n-lactalbumin concentration (Fig. 4) shows that these concentration variations will give errors of less than 1% in the determination of lactose synthase activity. The sedimentation coefficient of free galactosyltransferase S pl may be determined in the absence of cu-lactalbumin. velocity ultracentrifugation, using an ultraviolet scanner, has been used previously to demonstrate complex formation between galactosyltransferase and oc-lactalbumin, in the presence of either UDP-galactose or GlcNAc (8). This previous study utilized a multicomponent galactosyltransferase from bovine milk, and while a progressive increase in sedimentation coefficient of enzyme with increasing concentration of a-lactalbumin was observed in the presence of GlcNAc, no quantitative data with respect to dissociation constants were presented.
Our previous ultracentrifuge studies, performed in the presence of 5 ITIM GlcNAc, gave a value of 3.3 for the sedimentation coefficient (s2,,J of b ovine colostrum galactosyltransferase. The sedimentation coefficient determined in the presence of 10 mM MnCl, and 0.3 mM UDP-galactose in the present work is very similar sZO+ 3.25. The sedimentation coefficient of the enzyme, determined as lactose synthase activity, increases hyperbolically with cu-lactalbumin concentration to reach a plateau at 400 pg/ml of sZO+ 4.22. The sZO+ was not changed significantly at 1600 Kg/ml. The saturation curve (Fig. 5) was determined at a constant temperature of 16". Sedimentation coefficients for the free enzyme (zero concentration of cY-lactalbumin) and enzyme oc-lactalbumin complex (800 /*g of a-lactalbumin/ml) were also determined as sZO,w at a series of temperatures (23.3", 20", 12", and So), and were in all cases 3.25 + 0.03 and 4.22 * 0.03, respectively.
As saturation with UDP-galactose as discussed previously, using values for K,* derived from the Van't Hoff plot shown in Fig. 1. The results are summarized in Table II. A direct comparison of the K, values for a-lactalbumin derived from kinetics and ultracentrifugation is precluded by the presence of 0.1 M NaCl in the latter measurements and its absence in the former. However, from the Van't Hoff plots of these data in Fig. 3 (16). Small but distinct changes observed in sedimentation coefficient of galactosyltransferase in the presence of GlcNAc, 3.3 S (12), and in the presence of Mn*+, UDP-galactose, 3.25 S, may reflect conformational changes upon the binding of those substrates.
A more pronounced change is observed with crosslinked galactosyltransferase-cu-lactalbumin in the presence of Mn2+ and UDP-galactose, 4.35 S and absence of substrates, 4.51 s.
Two further aspects of the attachment of substrates during the lactose synthase catalytic cycle remain to be investigated; the first, the nature of manganese binding, being difficult to investigate.
The second is the nature of the attachment of monosaccharides, and how this is modified by cu-lactalbumin. The present work provides a basis for the study of this second aspect which is currently in progress.