Nine cases of gastric carcinoma with excessive production of α‐fetoprotein (AFP) were analyzed morphologically, histochemically and biochemically. Consequently, it was proposed that AFP‐producing gastric carcinomas should be divided into three subtypes: (i) hepatoid type; (ii) yolk sac tumor‐like type; and (iii) fetal gastrointestinal type. The data from the study suggested that the hepatoid type and the yolk sac tumor‐like type are derived from liver cell metaplasia and yolk sac cell metaplasia of common poorly differentiated medullary adenocarcinoma, respectively. The fetal gastrointestinal type seemed to be a result of the imitation of fetal gastrointestinal epithelium by common tubular adenocarcinoma. The hepatoid type was also the most common in the file of AFP‐producing gastric carcinoma. Unfortunately, most of the hepatoid types seemed to be highly malignant.

Several in vitro studies have suggested that high mobility group (HMG) protein 1 has a role in gene regulation as a trans activator or quasi-transcription factor. However, data on the molecular functions of HMG1 protein in these reactions are contradictory or obscure. In order to assess whether HMG1 protein does, in fact, have transcriptional activation potential, two assay systems in cultured cells were employed. HMG1 protein introduced into COS-1 cells as a complex with a reporter plasmid carrying the lacZ gene enhanced the level of the gene expression. Cotransfection of an expression plasmid carrying HMG1 cDNA into the cells with the reporter plasmid enhanced the activity of beta-galactosidase 2-3-fold in comparison with that of the control effector plasmid. The enhancement was proved to be dependent not on the replication but on the transcription of the reporter plasmid. In the cotransfection experiments, an expression plasmid the HMG1 molecule lacking the acidic carboxyl terminus repressed the expression of the reporter gene. The binding of an HMG1 protein variant lacking the acidic carboxyl terminus to DNA gave an extremely large shift of gel retardation in comparison with the complete HMG1 molecule. Together, these results indicate that HMG1 protein can enhance expression in cells in culture at the step of gene transcription and that the DNA binding domains comprising two-thirds of the HMG1 protein molecule are responsible for the inhibition property. Also, the acidic terminus of the HMG1 molecule is essential for the enhancement of gene expression in addition to elimination of the repression caused by the DNA binding. (ABSTRACT TRUNCATED AT 250 WORDS)

High mobility group (HMG) 2 is a sequence-nonspecific DNA-binding protein consisting of a repeat of DNA-binding domains called HMG1/2 boxes A and B and an acidic C-terminal. To understand the mode of HMG2 interaction with DNA, we expressed various HMG2 peptides containing HMG1/2 box(es) in Escherichia coli cells and purified them. Gel retardation and DNA supercoiling assay indicated that the region essential for the preferential binding of HMG2 with negatively supercoiled DNA and DNA unwinding activity is located in box B, but not sufficient alone. The flanking C-terminal basic region or box A linked by a linker region is necessary to express activities. The SPR measurements certified that the intrinsic DNA binding affinity of box B is weaker (Kd = 170 microM), and these adjoining regions largely strengthen the affinity (Kd

High mobility group (HMG) proteins 1 and 2 contain two similar but non-identical repeats of DNA-binding domains and an acidic C-terminal. The proposed functions of HMG proteins 1 and 2 imply a probable difference in their DNA-binding abilities. The primary studies by gel retardation assay showed that HMG2 has higher affinity than HMG1 for supercoiled and linear DNA. The DNA-binding of HMG2 appeared strong enough to allow exchange with HMG1 molecule already bound to DNA, while the DNA-binding region of HMG1 showed higher affinity than that of HMG2. In order to compare more quantitatively the affinities, surface plasmon resonance (SPR) measurements using a BIAcore instrument were conducted. The kinetic data indicated that the Kd for the complex of HMG2 with DNA is smaller than that of HMG1, in contrast to the situation for the DNA-binding region of these proteins. The sequence between the second DNA-binding domain and the acidic C-terminal of HMG proteins is required for tight DNA-binding. Also, the acidic C-terminal strongly modulates the DNA-binding ability of each protein. The usefulness of SPR measurement for quantitative analysis of affinity and regions involved in DNA-binding under conditions nearly identical to those in solution is discussed.