Rotaviruses contain a double-stranded ribonucleic acid genome consisting of 11 segments. Gel electrophoresis separates genome segments and allows identification of strain differences. This electrophoretic typing technique was applied to rotavirus specimens from 116 children and 72 newborn babies. Between 1973 and 1979, 17 different electropherotypes of rotavirus were observed in children with acute gastroenteritis. These electropherotypes showed a sequential pattern of appearance, with a limited number of electropherotypes present at any given time. By contrast, only two electropherotypes were identified from isolates from newborn babies in seven hospitals during 1975 to 1979. These two electropherotypes were very similar and were never identified in children with acute gastroenteritis. One of the neonatal electropherotypes was found in the nurseries of five different hospitals and persisted in one hospital for 4 years. Electrophoretic typing techniques can be applied routinely and reproducibly to small samples of feces and could prove to be of value in epidemiological studies of rotavirus infection.

Levels of antirotaviral secretory immunoglobulin A were measured by enzymelinked immunosorbent assay in colostrum and milk samples collected daily for the first 5 days postpartum from 49 mothers breast-feeding their infants. The trypsin-inhibitory capacity of these lacteal secretion samples was assessed by their ability to inhibit the hydrolysis of aN -benzoyl-DL-arginine-p-nitroaniide by trypsin. Stools passed by these breast-fed infants and by an additional 43 bottlefed infants were pooled by individual and examined by electron microscopy for rotavirus. Stool trypsin levels were estimated with the gelatin hydrolysis test. Breast-fed infants were significantly less likely to become infected with rotavirus and showed significantly lower stool tryptic activity than did bottle-fed infants. Breast-fed infants who did not excrete rotavirus over the 5-day period received milk of significantly higher antirotaviral secretory immunoglobulin A or trypsininhibitory capacity or both than breast-fed infants who were infected with rotavirus. A case of probable maternal rotavirus infection during pregnancy, producing greatly elevated lacteal antirotaviral secretory immunoglobulin A levels lasting for 2 years, was detected. Results of this study suggest that both antibodies and trypsin inhibitors in human milk can be associated with the protection of neonates against rotavirus infection in the first 5 days of life.

A solid-phase, enzyme-linked immunospecific assay for measurement of different immunoglobulin classes of human rotavirus antibodies is described. The antigen, which was adsorbed directly to polyvinyl microtiter plates, consisted of a clarified cell culture stock of the simian rotavirus SA 11. The assay was sensitive and reproducible and could readily be calibrated to determine concentrations of each class of antibody. The assay was applied to measurements of rotavirus antibodies in serum, colostrum, milk, and fecal samples. It particularly facilitates investigations of the role of immunoglobulin A antibodies in immunity to rotavirus infections.

Levels of rotavirus-specific immunoglobulin G (IgG), IgA, IgM, and secretory immunoglobulin in maternal and cord serum, colostrum and milk, and infants' stools were measured by enzyme-linked immunosorbent assay in 92 mothers and their infants. Although antirotaviral IgG, IgA, and secretory immunoglobulin were present in most maternal sera, only IgG crossed the placenta. All samples of colostrum and milk tested contained antirotaviral secretory immunoglobulin and IgA except those of two women in whom IgA deficiency was subsequently described. Specific IgM and IgG were also detected in many colostral samples. Antirotaviral IgA was detected in many colostral samples. Antirotaviral IgA was detected in stools of breast-fed but not bottle-fed neonates. Apparently the human infant receives rotaviral antibodies both transplacentally and via maternal colostrum and milk.