Effect of the conformation of concanavalin A on its affinity for manganous ion.

The stoichiometry of Mn2+ binding to concanavalin A at pH 6.4-7 which had been established in two independent studies [J.A. Sophianopoulos, A.J. Sophianopoulos, and W.C. MacMahon (1983) Arch. Biochem. Biophys. 223, 350-359; D.J. Christie, G.R. Munske, and J.A. Magnuson (1979) Biochemistry 18, 4638-4644] was challenged [C.F. Brewer, R.D. Brown, III, and S.H. Koenig (1983) Biochemistry 22, 3691-3702] on grounds of possible experimental errors. Additional evidence is presented in this study in support of the previous finding that at pH 6.4 only one Mn2+ binds per concanavalin A monomer of Mr 25,550. Also, evidence is presented showing that the results of Sophianopoulos et al. could not have been due to contamination by Ca2+. A comparison is made of the results in the three studies cited above which indicates that the concanavalin A used by Brewer et al. had decreased affinity for Mn2+ and it contained an appreciable fraction of concanavalin A incompetent of binding saccharides.

Manganese, calcium, and saccharide binding to concanavalin A, as studied by ultrafiltration.

The binding of the ligands Mn2+, Ca2+, and methyl alpha-D-glucopyranoside to concanavalin A, purified as described (A.J. Sophianopoulos and J.A. Sophianopoulos (1981) Prep. Biochem. 11, 413-435), was studied by ultrafiltration in 0.2 M NaCl, pH 5.2 and pH 6.5 to 7, and at 23 to 25 degrees C. The association constant (Ka) of methyl alpha-D-glucopyranoside to concanavalin A was (2 +/- 0.2) X 10(3) M-1, both at pH 5.2 and 7. At pH 5.2 and in the absence of Ca2+, the Ka of Mn2+ to concanavalin A was (5 +/- 1) X 10(3) M-1, and in the presence of 1 mM Ca2+, the Ka was (9.1 +/- 2.1) X 10(5) M-1. At pH 6.5 Mn2+ bound to concanavalin A with a Ka of (7.3 +/- 1.8) X 10(5) M-1, and the binding affinity was virtually independent of the presence of Ca2+. Experiments of binding of 4-methylumbelliferyl alpha-D-mannopyranoside to concanavalin A indicated that at pH 5.2, binding of a single Mn2+ per concanavalin A monomer was sufficient to induce a fully active saccharide binding site. Ca2+ is not necessary for such activation, but rather it increases the affinity of concanavalin A for binding Mn2+.

Preparation of homogeneous concanavalin A.

Concanavalin A (Con A), obtained either commercially or by affinity chromatography, was further purified by incubating at 6-8 hr at pH 3.0-3.2 in 1 M NaCl, 0.08 M glycine and 3 mM each Ca2+ and Mn+2, heat treating at 45 degrees C for 2 hr and centrifuging. The supernatant was neutralized to pH 5 and stored in the cold. Te overall yield was 70-80%. Some of the properties of Con A at pH 5 are: The absorption coefficient of a l g/dl solution is 13.7 at 280 nm; the mean residue ellipticity at 224.5 nm is -9,300 degrees to -9,800 degrees; by sedimentation equilibrium, its molecular weight is 53,000 between pH 3.0 and pH 5.2. Con A solutions standing at room temperature at pH 7 for ten days lose through precipitation only 5-8% of the protein in 0.2 M NaCl and 15% of the protein in 0.1 M NaCl. In the solution conditions of SDS and urea-SDS gels, Con A not only unfolds slowly and incompletely, but it also forms high molecular weight aggregates. Thus, electrophoresis of Con a in such gels is unsuitable for tests of homogeneity. However, as judged by sedimentation equilibrium in 6.5 M quanidine at pH 8.1, purified Con A was monodisperse.

Isolation of a novel glycoprotein (EDC1) from the urine of a patient with acute myelocytic leukemia.

Gel filtration of urine from Patient ED with acute myelocytic leukemia showed a prominent protein peak with elution position corresponding to molecular weights of 20,000 to 35,000. The protein (EDC1) was isolated in pure form by sequential gel filtration and ion-exchange chromatography. Molecular weight of purified EDC1 was 27,000; it contained 27% carbohydrate and was rich in half-cystine (5% of residues). EDC1 was antigenically and chemically distinct from the recognized glycoproteins of normal plasma. With a specific rabbit antiserum and 125l-labeled EDC1, a radioimmunoassay for the glycoprotein was developed. Both noncancer and cancer plasmas contained immunoreactive material. In noncancer plasma, all the immunoreactivity was eluted from Sephadex G-75 and G-200 in position corresponding to molecular weights of 60,000 to 100,000 (Peak 1). In cancer plasma, an additional peak of immunoreactivity was eluted in the position corresponding to EDC1 (M.W., 20,000 to 30,000; Peak 2). Eighty-six % of urines from patients without clinical cancer were nonreactive in radioimmunoassay (less than 0.1 microgram immunoreactive EDC1 per ml); 11 and 3%, respectively, contained immunoreactivity equivalent to 0.1 to 0.9 and 1 to 9 microgram EDC1 per mi, entirely of Peak 1 type. Ninety-one % of urines from patients with disseminated cancer contained immunoreactivity equivalent to 10 to 9,999 microgram EDC1 per ml, primarily of Peak 2 type.