Search

A novel low mol wt inositol phosphoglycan antagonist of insulin action of oxidative glucose metabolism in isolated rat adipocytes was partially purified from normal human plasma and shown to be increased in type II diabetic plasma. It was characterized chemically as a myo-inositol phosphoglycan containing a cyclic 1,2-phosphate. This antagonist, termed fraction V3, is now shown to inhibit the action of an inositol glycan insulin pH 2.0 mediator that stimulates pyruvate dehydrogenase phosphatase in a similar manner to insulin. In addition, fraction V3 inhibits stimulation of the pyruvate dehydrogenase (PDH) phosphatase by Mg2+, the enzyme's required metal, and by spermine, a polyamine. Fraction V3 does not inhibit active PDH itself. The inhibitory effect is dose dependent and apparently noncompetitive or nonsurmountable for the insulin inositol glycan pH 2.0 mediator, thus comparing kinetically with its insulin antagonistic action on intact adipocytes. Its inhibitory action on PDH phosphatase is dose dependent and competitive for Mg2+ stimulation of the phosphatase. Additionally, fraction V3 is shown to inhibit stimulation by Mg2+ of cloned recombinant PDH phosphatase catalytic subunit. Inhibition by fraction V3 of Mg(2+)-stimulated PDH phosphatase and its cloned catalytic subunit helps explain its mechanism of action to inhibit insulin-stimulated oxidative glucose metabolism in adipocytes and its potential clinical significance in insulin resistance.

A novel low mol wt inositol phosphoglycan inhibitor (M tau 1200-1500) of insulin action in rat adipocytes has been partially purified from normal human plasma. This inhibitor, termed fraction V after the first purification step and fraction V3 after the second, is different from other reported serum insulin antagonists. It contains myoinositol, galactosamine, and mannose in approximate molar ratios of 1:1:3.3. The myoinositol has a 1,2-cyclic phosphate substituent, which is essential for the inhibitory activity. Its inhibitory activity is significantly elevated (161%, P < 0.05 for fraction V; 278%, P < 0.05 for fraction V3) in plasma of humans with noninsulin-dependent diabetes mellitus as compared with plasma of nondiabetic controls. These findings represent the first report of a naturally occurring mammalian inositol 1,2-cyclic phosphate containing phosphoglycan related to insulin action.

Synalbumin and albumin to which an insulin B-chain has been attached have been tested in vivo for insulin antagonistic activity by an intraperitoneal injection method. Both inhibit the effect of insulin on the rat diaphragm, but not on the epididymal fat pad. These findings are similar to those reported in vitro. The results are compatible with the hypothesis that the synalbumin insulin antagonist is the same as the B-chain of the insulin molecule.

Patients with essential diabetes including prediabetes have more synalbumin antagonism to insulin than normal subjects or patients with “pancreatic” diabetes, e.g. after pancreatectomy. Thus excessive synalbumin antagonism (synalbumin positive) can be regarded as a biochemical marker to ascertain whether a given person is constituted as an essential diabetic-without reference to carbohydrate intolerance.-On this premise, the genetic transmission of essential diabetes has been examined by studying the relatives of essential diabetics. The results strongly suggest that excessive synalbumin antagonism is inherited as an autosomal “dominant” character.-Ninety-seven members of nine families were studied, 39 were synalbumin negative whereas 58 were synalbumin positive, but only 16 of these latter have overt carbohydrate intolerance. A further three have spontaneous hypoglycaemia whereas the remainder are quite asymptomatic. These observations suggest that overt carbohydrate intolerance is relatively uncommon or will be a late event in many people constituted as essential diabetics.

Studies on ultrafiltrates of the plasma of normal and diabetic subjects have shown that a polypeptide of similar characteristics to the growth-hormone-derived polypeptide In-G is present in higher concentrations in diabetic subjects, thus indicating a possible role in the pathogenesis of diabetes mellitus. The polypeptide is absent from the plasma of hypophysectomized diabetic patients.

Continued interest in the possible pathogenetic role of the SIA in diabetes mellitus has been perpetuated primarily by three observations: (1) that plasma and sera from many diabetic subjects and their close relatives contain higher levels of the SIA than most nondiabetics. (2) that the SIA appears to affect glucose metabolism of muscle but not adipose tissue, possibly accounting for the increased prevalence of obesity in adultonset diabetics; and (3) that the reduced B-chain of insulin when bound to albumin reproduces most of the in vitro and in vivo metabolic and physiochemical characteristics of the SIA. On the other hand, a critical appraisal of the present state of knowledge reveals serious obstacles to acceptance of a role for synalbumin in diabetes mellitus. These are: (1) the potential presence of a dialysis casing artifactual insulin antagonist; (2) the spectrum of antagonistic, neutral and stimulatory activity in albumin fractions prepared by methods not utilizing dialysis indicating that preparative technique can significantly affect the degree of measured insulin antagonism; (3) the failure of the SIA to demonstrate an unequivocal in-vivo effect; (4) a mechanism of action inconsistent with the metabolic alterations observed in diabetes mellitus; (5) the use of the relatively insensitive and imprecise rat hemidiaphragm bioassay to assess qualitative and quantitative differences in insulin antagonism; and (6) the lack of specificity since albumins of other species as well as egg albumin and gelatin are indistinguishable from the SIA in the assays currently in use. Although these difficulties do not absolutely deny a physiological role for the SIA, they do urge a very cautious interpretation of the available data until a much more selective means of assessment becomes possible.

Cellulose acetate, starch gel, and immunoelectrophoresis of human serum “albumin” extracts were performed. TCA/ethanol extraction altered the albumin so that polymers were present in addition to normal monomer albumin. Alpha1globulin was also present in the extracts, proving them to be heterogeneous in nature. Nevertheless, extracts from nondiabetic and diabetic sera demonstrated insulin antagonism when measured on rat hemidiaphragms in physiologic serum concentration (4.0 per cent). Reduced concentration (1.25 per cent) of diabetic extracts still demonstrated insulin antagonism; normal extracts did not. These results, consistently found, support the idea that a difference exists between nondiabetic and diabetic humans in quantity of insulin antagonist associated with circulating protein. Extracts of the same blood sample, or of blood from the same subject on different days, gave reproducible results from one observation to another. Despite constancy of insulin concentration in hemi-diaphragm preparations, biologic effect of insulin (glucose uptake increase) varied. At all levels of insulin effect diabetic extracts significantly antagonized insulin compared to nondiabetic extracts. To show differences in effect among individual extracts, comparison was made between per cent change from basal uptake produced by extract-plus-insulin and that produced by insulin alone. In twenty-seven of thirty-three nondiabetic subjects (82 per cent) change in glucose uptake produced by insulin-plus-extract was over 60 per cent of that produced by insulin alone; in six (18 per cent) the change lay between 50 per cent and 60 per cent. In all thirty-nine diabetic subjects change from basal uptake was less than 60 per cent of that produced by insulin alone.