These results do not provide evidence for a causal relationship between RAP and H. pylori.

The systemic humoral immune response to Helicobacter pylori antigens was investigated in 36 children with recurrent abdominal pain (RAP). H. pylori was cultured and Helicobacter‐like organisms (HLO) were seen in six children, three of whom had active and two inactive chronic gastritis. None of these children had endoscopic abnormalities. All sex children had increased IgG antibodies to heat‐stable H. pylori antigens which were of the IgG1 and IgG3 subclasses. Using six other IgG tests, four of which were commercially available, two to five H. pylori‐positive children were found seropositive. Five of six H. pylori‐negative children with inactive chronic gastritis and no endoscopic abnormalities had increased IgM antibody levels in addition to increased or borderline increased IgG antibody levels to H. pylori, indicating activity in a chronic H. pylori infection. Five children without H. pylori and with no morphological changes, but with gastritis or duodenitis by endoscopy, had significantly lower IgG and IgA antibody levels compared to other groups. Six of nineteen children without H. pylori, and with no morphological or endoscopic changes had increased IgG and IgM antibody levels to H. pylori. All H. pylori‐negative children were seronegative by the four commercial kits. Overall, 12 (33%) of 36 children with RAP were either H. pylori positive by culture and microscopy or had increased IgG antibody levels to H. pylori, which is significantly different from the 10–14% seropositive rate of asymptomatic children. H. pylori may therefore be a cause of RAP in one quarter to one third of the children with RAP in whom other etiologies of RAP are excluded. Further studies on a large number of children are needed for an extended evaluation of the humoral immune response to H. pylori and for further examination of commercial kits which seem to give a high number of false‐negative results.

In the present study, we report a more detailed biochemical analysis of the B16 melanoma, metastasis-associated, Met-72 antigen. Specifically, we have examined (1) the molecular forms of Met-72 isolated during synthesis, surface expression and 'shedding' and (2) the cell-surface expression of Met-72 during the cell cycle. These experiments show that the 72 kD species originally described has an isoelectric point of between 6.3 and 6.9, but is the desialylated derivative of an 83 kD native molecule whose isoelectric point ranges between pH 4.9 and 5.6. In addition, a 90 kD glycoprotein doublet was immunoprecipitated from biosynthetically labelled B16 melanoma cells, but does not appear to be a precursor of the 83 kD or 72 kD molecule. These findings have led us to interchangeably use the terminology Met-72 and Met 72/83. The latter terminology more accurately describes the physical forms which can be identified by different labelling procedures. When culture supernatants from 3H-leucine labelled cells were subjected to anti-Met-72 immunoprecipitation, a 35 kD species was identified as a possible 'shed' product of these cells. Met-72/83 expression during the cell cycle was analyzed by flow cytometry and found not to be restricted to any particular stage. In addition, experiments were performed to determine whether low levels of Met-72 expression on poorly metastatic B16 melanomal clones was a direct result of low levels of synthesis, or if other control mechanisms regulated intracellular pools of Met-72 prior to cell-surface expression.