We describe an algorithm for finding the most statistically significant non-overlapping subtrees of a hierarchical clustering of gene expression data with respect to a set of secondary data labels on genes. The method is implemented as a Java plug-in for a commercial gene expression analysis program (GeneSpring).

Analogs of human choriogonadotropin (hCG) lacking N-glycosyl chains at alpha52Asn and alpha78Asn were purified from the culture media of insect cells by immunoaffinity chromatography using a monoclonal antibody column. As previously reported, while analogs lacking carbohydrate at alpha52Asn and alpha78Asn had similar receptor binding activities compared with the wild type recombinant hCG (hCGwt), they differed in their signal transduction properties. The mutant lacking carbohydrate at alpha78Asn had 20% less cAMP-stimulating activity than hCGwt, but the absence of glycosylation at alpha52Asn resulted in the reduction of cAMP accumulation by 90-95%. A similar effect of the mutations was observed on the stimulation of steroidogenesis. Circular dichroism spectra of the two mutants showed significant differences. The mutant lacking carbohydrate at alpha52Asn had a much higher negative mean residue ellipticity (MRE) at 200 nm and a lower negative MRE at 220 nm than that lacking carbohydrate at alpha78Asn and hCGwt. The dissociation rates of the alpha52Asn and alpha78Asn carbohydrate deficient mutants at pH 3 and room temperature, measured by using 1-anilino-8-naphthalenesulfonate, were 9.4 x 10(-5) and 3.8 x 10(-5) s-1, respectively, as compared with 1.5 x 10(-5) s-1 for hCGwt. The results of both CD measurements and dissociation studies strongly suggest that the absence of carbohydrate at alpha52Asn results in conformational changes in the mutant which might explain the loss in its signal transduction function. This is further supported by indirect evidence from two other lines of experimentation. Unlike the mutant lacking carbohydrate at alpha78Asn, the one lacking carbohydrate at alpha52Asn cross-reacted with the two subunit specific monoclonal antibodies, anti-hCGalpha and anti-hCGbeta, which normally did not cross-react with the native or the hCGwt. Also, polyclonal anti-hCGbeta but not anti-hCGalpha was able to restore the cAMP-producing activity of the alpha52Asn carbohydrate deficient mutant. From all the data taken together, it appears that the loss of second messenger-producing activity of hCG with the absence of the glycosyl chain at alpha52Asn was probably due to a conformational change in the heterodimer rather than due to the loss of the alpha52Asn-carbohydrate-receptor interaction.

No abstract

No abstract

No abstract