Fetal and adult albumins are indistinguishable by immunological and physicochemical criteria.

The existence of a functionally immature fetal albumin has been postulated to explain the reduced ability of newborn plasma to bind bilirubin and various drugs. In support of this, cord and adult albumin, isolated by a simple salting-out technique, were reported to differ in electrophoretic and chromatographic properties and in their resistance to alkali and proteolytic enzymes. However, the interpretation of these findings has since been questioned. To resolve this controversy, we have purified to homogeneity human serum albumins from pooled umbilical cord and adult donor plasma. The two albumins were compared and found to be indistinguishable by polyacrylamide gel electrophoresis with and without sodium dodecyl sulfate, as well as by immunoelectrophoresis and double immunodiffusion using specific antibodies against both albumins. Furthermore, the amino acid compositions, the aminoterminal sequence (Asp-Ala-His-Lys-Ser-Glu-Val-Ala-), the carboxy terminus (Leu), and the peptide fingerprints were identical in the two albumins. No significant differences were found by circular dichroism in the ultraviolet (200-350 nm). Binding studies with bilirubin showed association constants of 3.7 +/- 0.7 x 10(7) M(-1) for cord and 2.9 +/- 0.3 x 10(7) M(-1) for adult albumin, respectively. The circular dichroic spectra of 1:1 bilirubin.albumin complexes showed considerable variation between the batches but were not significantly different. The only difference was found in the fluorescence spectra of the bilirubin.albumin complexes, in which complexes with adult albumin showed only 75% of the relative fluorescence exhibited with cord albumin. The combined results nevertheless strongly indicate that fetal and adult albumins are very similar, if not identical.

Hereditary fructose intolerance in early childhood: a major diagnostic challenge. Survey of 20 symptomatic cases.

Twenty infants and young children with hereditary fructose intolerance (HFI) were admitted to hospital. None was diagnosed at admission. Referals were for vomiting of unknown aetiology (16X), pyloric stenosis or hiatus hernia (5X), toxic condition (3X), and hepatomegaly of unknown origin (5X). Feeding difficulties (20X), vomiting (18X), and failure to thrive (16X) were leading symptoms. The most frequent clinical findings were hepatomegaly (18X), pallor (14X), haemorrhages (13X). Ascites, oliguria, tachypnoea, fever, splenomegaly and rickets were less frequent. Laboratory findings were indicative of disturbed hepatic and renal tubular function and also of disturbed intermediary metabolism (hypokaliaemia, hypophosphataemia). However, hypoglycaemia was found in only 4 out of 15 patients tested. Differential diagnosis after hospital admission centered on metabolic disorders such as glycogenoses, galactosaemia, tyrosinosis, or Wilson's disease. Hepatitis, toxic hepatosis, liver tumour, intrauterine infection and sepsis were also considered. Eleven children had first ingested fructose within the first 6 weeks of life. The diagnosis was usually established only many weeks or months after first fructose intake and appearance of symptoms. This documents how difficult the diagnosis of this disease can be both in practice and in hospital. The course was severe in 11 children and lethal in 4. In only 5 patients was the course mild. The 16 survivors are doing well under fructose-exclusion diet. Irreversible visual impairment after intraocular haemorrhage occurred once. In each case HFI could have been suspected immediately, had a detailed nutritional history been taken. Practising paediatricians should know the composition of commonly used infant formulae. They should never prescribe sugared condensed milk for intractable vomiting prior to excluding HFI. Solution for intravenous infusion containing fructose and sorbitol are life-threatening for undiagnosed HFI patients.

Affinity labeling of the primary bilirubin binding site of human serum albumin.

A label for the bilirubin binding sites of human serum albumin was synthesized by reacting 2 mol of Woodward's reagent K (N-ethyl-5-phenylisoxazolium-3'-sulfonate) with 1 mol of bilirubin. This yielded a water-soluble derivative in which both carboxyl groups of bilirubin were converted to reactive enol esters. Covalent labeling was achieved by reacting the label with human serum albumin under nitrogen at pH 9.4 and 20 degrees. Under the same conditions, no covalent binding to the monomers of several proteins could be demonstrated. The number of binding sites for bilirubin and the label were found to be the same, and competition experiments with bilirubin showed inhibition of covalent labeling. The absorption, fluorescence and CD spectra of the label in a complex with human serum albumin were similar to those of the bilirubin human serum albumin complex. However, following covalent attachment to the spectral properties were changed, indicating loss of conformational freedom of the chromophore. Labeling ratios were selected to result in the incorporation of less than 1 mol of label/mol of human serum albumin. Under these conditions, labeling is thought to occur primarily at the high affinity binding site.