Alterations in basal and glucose-stimulated voltage-dependent Ca2+ channel activities in pancreatic beta cells of non-insulin-dependent diabetes mellitus GK rats.

In genetically occurring non-insulin-dependent diabetes mellitus (NIDDM) model rats (GK rats), the activities of L- and T-type Ca2+ channels in pancreatic beta cells are found to be augmented, by measuring the Ba2+ currents via these channels using whole-cell patch-clamp technique, while the patterns of the current-voltage curves are indistinguishable. The hyper-responsiveness of insulin secretion to nonglucose depolarizing stimuli observed in NIDDM beta cells could be the result, therefore, of increased voltage-dependent Ca2+ channel activity. Perforated patch-clamp recordings reveal that the augmentation of L-type Ca2+ channel activity by glucose is markedly less pronounced in GK beta cells than in control beta cells, while glucose-induced augmentation of T-type Ca2+ channel activity is observed neither in the control nor in the GK beta cells. This lack of glucose-induced augmentation of L-type Ca2+ channel activity in GK beta cells might be causatively related to the selective impairment of glucose-induced insulin secretion in NIDDM beta cells, in conjunction with an insufficient plasma membrane depolarization due to impaired closure of the ATP-sensitive K+ channels caused by the disturbed intracellular glucose metabolism in NIDDM beta cells.

Detection of patients with cancer by monoclonal antibody directed to lactoneotetraosylceramide (paragloboside).

A hybridoma producing monoclonal antibody (H11) directed to lactoneotetraosylceramide (paragloboside) has been established from spleen cells of a mouse immunized with paragloboside. The monoclonal antibody H11 (immunoglobulin M type) was selected from five clones showing different reactivities with paragloboside. The monoclonal antibody was highly specific to paragloboside and lacked reactivity with other glycolipids including glucosylceramide, lactosylceramide, globotriaosylceramide, globotetraosylceramide, gangliotriaosylceramide, gangliotetraosylceramide, and GalNAc beta 1-4[NeuAc alpha 2-3]Gal beta 1-4Glc beta 1-1Cer. However, the monoclonal antibody (H11) was found to bind to lactosamine-containing glycolipids at their terminals, such as i- and I-type glycolipids as well as paragloboside. A two-step sandwich radioimmunoassay method for paragloboside antigen in serum was established by using the monoclonal antibody. The mean paragloboside antigen concentration in the sera from 20 normal individuals was 25.3 ng/ml. If the cutoff value was set at 80.9 ng/ml [25.3 + 2 x 27.8 (SD)], only 1 of 20 healthy controls had an elevated paragloboside value in the serum, whereas sera from 9 of 12 (75.0%) hepatoma, 4 of 10 (40%) pancreatic cancer, 16 of 40 (40.0%) stomach cancer, and 6 of 10 (60%) lung cancer patients had elevated paragloboside values. Sera from 3 of 8 hepatitis patients and 7 of 10 liver cirrhosis patients were estimated to be positive but sera from 16 patients with benign disease had paragloboside levels lower than the cutoff value. A larger amount of the antigen was found in liver metastases from colorectal carcinoma compared to the normal counterpart. The antigen was also detected in the medium of various human cancer cells and meconium. However, the antigen in the sera, medium, meconium, and cancer tissue seemed to be associated with glycoprotein or lipoprotein, because most of the antigen activity was eluted in the void volume fraction on high-performance liquid chromatography with a gel filtration column.

Accumulation of gangliosides with N-acetylneuraminosyl(alpha 2-6)lactosamine structure in primary human hepatoma.

Gangliosides of hepatomas have been analyzed by using a monoclonal antibody directed to N-acetylneuraminosyl(alpha 2-6)lactoneotetraosylceramide (sialyl(alpha 2-6)paragloboside), which was prepared by injecting the monosialoganglioside fraction of human meconium into BALB/c mice. The monoclonal antibody, named MSG-15, was found to bind sialyl(alpha 2-6)paragloboside, but it failed to react with other gangliosides, including N-acetylneuraminosyl(alpha 2-3)lactoneotetraosylceramide (sialyl (alpha 2-3)paragloboside) and "Ii"-type gangliosides. MSG-15 was found to recognize NeuAc alpha 2-6Gal beta structure of the ganglioside. Gangliosides obtained from human hepatomas were analyzed by immunostaining on high-performance thin-layer chromatography plates using the monoclonal antibody MSG-15. All primary hepatoma samples used in this study (nine samples) were found to contain sialyl(alpha 2-6)paragloboside, which accounted for 13-31% of the monosialoganglioside fractions in the hepatomas. Furthermore, MSG-15 recognized several monosialogangliosides in addition to sialyl(alpha 2-6)paragloboside. These gangliosides apparently also contain a terminal NeuAc alpha 2-6Gal beta structure. Other ganglioside fractions obtained from hepatoma and meconium were immunostained on thin layer chromatography plates with MSG-15. Additionally, another monoclonal antibody (H-11), which recognizes terminal lactosamine structure, was used to immunostain these fractions after sialidase treatment. Bands stained with both monoclonal antibodies showed similar mobilities to each other in the di- and trisialoganglioside fractions as well as monosialoganglioside fraction. In control liver, GM3 ganglioside accounted for 92% of monosialoganglioside fraction, and sialyl(alpha 2-6)paragloboside accounted for less than 1% of the fraction. Immunohistochemical study by using MSG-15 in tissue sections from hepatocellular carcinoma and normal liver tissues demonstrated that only hepatocellular carcinoma cells gave a positive reaction. These results suggest that the biosynthetic pathway of gangliosides containing NeuAc alpha 2-6Gal beta 1-4GlcNAc beta structure is activated in hepatoma cells.

Specific expression of the ret proto-oncogene in human neuroblastoma cell lines.

The expression of the ret proto-oncogene (proto-ret), which possibly encodes two isoforms of a receptor-type tyrosine kinase, was examined in human tumor cell lines. Expression of the proto-ret mRNA was detected in all 11 neuroblastoma cell lines examined. The level of mRNA varied more than 100-fold in these neuroblastoma cell lines and was particularly high in three of them. On the other hand, 19 non-neuroblastoma tumor cell lines derived from solid tumors and a human diploid fibroblast cell line did not express any detectable levels of proto-ret mRNA. No remarkable amplification of the proto-ret or gross structural changes in the coding region were found in these neuroblastoma cell lines. The specific expression of the proto-ret in neuroblastomas suggests that the proto-ret product may have a role in cellular functions specific to neuroblastoma cells.

Human ret proto-oncogene mapped to chromosome 10q11.2.

Using cosmid clones derived from human ret protooncogene as probes, we determined its chromosome localization by fluorescence in situ hybridization. Two overlapping clones, cret-1 and -2, which were cloned using the most 5' part of human ret proto-oncogene cDNA, hybridized to chromosome 10q11.2. Sixty-three and 52% of the grains obtained by cret-1 and -2, respectively, were localized to the same site. No other specific hybridization site was observed. From these data, we assigned the site of ret proto-oncogene to chromosome 10q11.2, where the possible locus responsible for multiple endocrine neoplasia type 2A (MEN2A) was mapped by linkage analysis using interstitial retinol binding protein cDNA and D10S5. These findings suggest that ret proto-oncogene might be a suitable probe for approaching the MEN2A locus.

Some aspects of mechanism of inhibition of cholesterol absorption by beta-sitosterol.

Mixed bile salt micelle solubilized either cholesterol or beta-sitosterol to a comparable extent. When added simultaneously, beta-sitosterol restricted the micellar solubility of cholesterol. beta-Sitosterol also reduced the cholesterol content in the aqueous (micellar) phase of the intestinal contents of rats, the extent of reduction being comparable with that observed in vitro. The intestinal uptake of cholesterol in vivo was equivalent to the micellar incorporation of cholesterol both in vitro and in vivo. beta-Sitosterol had no inhibitory effect on cholesterol absorption from the micellar solution in jejunal loops in situ, whereas the rate of beta-sitosterol uptake was only about one-fifth that of cholesterol. The intestinal uptake of beta-sitosterol intubated into the stomach of rats was about one-fifth that of cholesterol. The intestinal brush-border membrane discriminated these sterols. These results suggest that the restriction of the micellar solubility of cholesterol, rather than the inhibition of uptake from brush-border membrane, is the major determinant for the interference of beta-sitosterol with cholesterol absorption.

Absorption and transport of base moieties of phosphatidylcholine and phosphatidylethanolamine in rats.

The absorption and transport of the base moieties of phosphatidylethanolamine (PE) and phosphatidylcholine (PC) which were fed to rats were compared. The major absorption site of ethanolamine-labeled PE was proximal jejunum while choline-labeled PC was absorbed almost equally throughout the jejunum. Lysophospholipids, glycerophosphoryl bases and constituent bases were the main digested products in intestinal content. This shows that base-labeled phospholipids were hydrolyzed to water-soluble products as well as lysophospholipids before absorption. The radioactivities from both phospholipids existed mainly in their parent phospholipids and water-soluble products in the intestinal mucosa. The amounts of lymphatic transport of the radioactivities from choline-labeled PC and ethanolamine-labeled PE were 17% and 8%, respectively, at 8 h after administration. The liver in lymph-cannulated rats contained 23% and 48% radioactivity from PC and PE, respectively, suggesting that base moieties of phospholipids, especially PE, were transported mainly via a non-lymphatic route, probably the portal vein, to the liver, as water-soluble products. The radioactivity from both base-labeled phospholipids in the liver was distributed in the parent phospholipids and water-soluble fractions. Ethanolamine-labeled PE was also incorporated into PC in the liver. These results indicate that intestinal absorption and transport of the base moiety of dietary PC and PE are similar; however, their intestinal absorption site and the extent of their separation during transport between the lymphatic and portal systems differ markedly.

Age-related changes in lipid metabolism in rats: the consequence of moderate food restriction.

Male Sprague-Dawley rats, 3 weeks of age, received one of the four following dietary manipulations until 9 months of age: group AD, fed ad libitum; group R, restricted to 75% of food intake of group AD; group RAD, restricted until 5 months of age and then fed ad libitum; group ADR, fed ad libitum until 5 months of age and then restricted. The concentration of serum total and HDL cholesterol tended to be higher in group AD than in all groups that had experienced food restriction. Liver cholesterol was higher in groups AD and RAD than in groups R and ADR. Activity of hepatic microsomal hydroxymethylglutaryl-CoA reductase was comparable, whereas sterol synthesis from mevalonate was significantly higher in groups R and ADR than in groups AD and RAD. Cholesterol 7 alpha-hydroxylase activity also tended to be higher in groups R and ADR. It seems likely that hepatic cholesterol homeostasis functions effectively even when moderate food restriction was started after the growing period. In addition, food restriction reduced the ratio of arachidonate to linoleate, suggesting inhibition of the desaturation system.

Molecular cloning of cDNA for the catalytic subunit of rat liver type 2A protein phosphatase, and detection of high levels of expression of the gene in normal and cancer cells.

A cloned cDNA encoding a catalytic subunit of type 2A protein phosphatase from a rat liver cDNA library was obtained by use of a synthetic oligonucleotide corresponding to the tryptic peptide sequence of the purified enzyme. There was only a single amino acid difference between the deduced amino acid sequence of the clone obtained and those of the catalytic subunits, 2A alpha, of the rabbit skeletal muscle, porcine kidney and human liver enzymes, suggesting that this clone was a rat 2A alpha cDNA. On Northern blot analysis using a cDNA fragment as a probe, three mRNA species were detected in rat liver: a major mRNA of 2.0 kb and a minor one of 2.7 kb under high stringency conditions, and also a 1.1 kb mRNA under low stringency conditions. The 2A alpha gene was found to be highly expressed in various tissues of rat, especially the brain. High levels of expression of the gene were also detected in mouse NIH3T3 cells and their transformants, and in human cancer cell lines as well as a human immortalized cell line.

Tea catechins decrease micellar solubility and intestinal absorption of cholesterol in rats.

A(-)-epicatechin (EC) and (-)-epigallocatechin (EGC) mixture and a mixture of their gallates (ECG and EGCG, respectively) markedly lowered lymphatic cholesterol absorption in rats with a cannulated thoracic duct. A mixture of ECG and EGCG was more effective in reducing cholesterol absorption than the EC and EGC mixture. These catechins also tended to decrease lymphatic absorption of triacylglycerols, although not so pronounced as in cholesterol absorption. An in vitro study on micellar solubility of cholesterol showed that these catechin mixtures precipitated cholesterol solubilized in mixed bile salt micelles in a dose-dependent manner. A mixture of ECG and EGCG more effectively precipitated micellar cholesterol than a mixture of EC and EGC. When purified EC, EGC, ECG and EGCG were used, EGCG was more effective in precipitating micellar cholesterol than ECG. The effect of EC and EGC was comparable and weaker than their gallate esters. The bile acid concentration in the micelles was not affected by these catechins. A positive correlation was observed between the amount of coprecipitated EGCG and cholesterol. These results clearly show that tea catechins, in particular their gallate esters, effectively reduce cholesterol absorption from the intestine by reducing solubility of cholesterol in mixed micelles. The observation accounts for the hypocholesterolemic effect of tea catechins.

Digestion and lymphatic transport of eicosapentaenoic and docosahexaenoic acids given in the form of triacylglycerol, free acid and ethyl ester in rats.

Lymphatic transport of eicosapentaenoic (EPA) and docosahexaenoic (DHA) acids given as trieicosapentaenoyl glycerol (TriEPA) and tridocosahexaenoyl glycerol (TriDHA) was compared with that of ethyl ester and free acid in rats cannulated with thoracic duct. Trioleoylglycerol (TO) served as a control. EPA and DHA, compared with oleic acid, were slowly transported in lymph irrespective of fat types administered. Total 24-h recovery of DHA in all fat types and ethyl EPA was significantly lower compared to that of oleic acid. Lymphatic recovery of EPA and DHA in rats given TriEPA and TriDHA was significantly higher at the first 3 h after the administration compared to those given as free acid or ethyl ester. The recovery in rats given free acid at a later stage (9-24 h) was higher than that of the other fat types. As a result, the 24-h recovery was comparable between triacylglycerol (TAG) and free acid, while it was significantly lower in ethyl ester. Although TriEPA and TriDHA were slowly hydrolyzed by pancreatic lipase in vitro compared with TO and TAGs rich in EPA or DHA at the second position, the hydrolysis rate at 60 min incubation was comparable among the TAGs examined. The hydrolysis rate of ethyl esters was extremely low even in 6 h incubation with lipase. These observations show that presence of EPA and DHA at the 1- and 3-positions of TAGs does not result in their lower recovery in lymph. Processes after lipolysis may be responsible for their low recovery in lymph. In a separate study, slower lymphatic recovery of DHA given as free acid than TriDHA was improved by the simultaneous administration of TO, but not by free oleic acid. The observations suggest that the slow recovery of free acid is caused by delayed TAG synthesis in mucosal cells and/or low micellar solubility of fatty acids in the intestinal lumen due to a limited supply of 2-monoacylglycerol (MAG). A large portion of EPA and DHA were recovered in lymph chylomicrons and very low density lipoproteins (VLDL, > 95%) and incorporated into TAG (84-92%) fraction in all fat types examined. Lymphatic recovery rate of simultaneously administered cholesterol was influenced by the fat types given.

Inhibitory effects of vesnarinone in the progression of myocardial damage in experimental autoimmune myocarditis in rats.

OBJECTIVES: Vesnarinone, a positive inotropic and immunomodulatory agent, diminishes nitric oxide (NO) levels by suppressing the induction of inducible NO synthase (iNOS) expressed in cytokine-stimulated macrophages and cardiomyocytes. We examined whether vesnarinone exerts inhibitory effects on the progression of myocardial damage in experimental autoimmune myocarditis in rats through suppression of iNOS. METHODS: Myocarditis was induced in 30 Lewis rats by injection of porcine cardiac myosin and vesnarinone was orally administered to 20 of the 30 rats. On day 21 after immunization (the climax of inflammation), the hemodynamics were examined and the severity of myocarditis was evaluated by determining the area ratio (%) [affected/entire area] of myocardial lesions in histological sections. Levels of serum CK-MB, NOx (NO2(-)+NO3-), TNF-alpha and IL-1 beta, and cyclic GMP, iNOS mRNA, TNF-alpha and IL-1 beta in heart tissues were determined. Expression of iNOS and TNF-alpha protein were examined by immunohistochemical methods. RESULTS: Histopathological examination revealed extensive myocardial destruction and massive infiltration of inflammatory cells in the vesnarinone-untreated rats. The area ratio of the lesions in the treated rats was significantly lower than that in the untreated ones. Levels of CK-MB, NOx, cyclic GMP, cytokines and iNOS mRNA were significantly lower in the vesnarinone-treated rats. Infiltrating macrophages and cardiomyocytes in the untreated rats showed much higher levels of expression of iNOS and TNF-alpha than those in the vesnarinone-treated rats. CONCLUSIONS: Vesnarinone may prove to be useful in the treatment of myocarditis by attenuating NO production through suppression of iNOS induced by cytokines.

Measurement of plasma renin activity by a simple solid phase radioimmunoassay.

A solid phase RIA in which an antibody adsorbed onto polystyrene balls was developed to determine PRA. Complete inhibition of converting enzyme and angiotensinase during enzymatic reaction was achieved by phenyl methyl sulfonyl fluoride and EDTA combination. Pepstatin A was found to be an effective agent to block angiotensin I generation during the RIA, and the sample can be directly incubated at room temperature for RIA without any special treatment to inhibit renin activity. The intra- and interassay coefficients of variation (CV) were 5.5-6.9% and 3.7-8.2%, respectively. The recovery was 91.9-117% of added angiotensin I. The assay is sufficiently sensitive and reliable for routine use and correlates acceptably (r = 0.996) with an established RIA. The antibody-adsorbed balls were compared to the soluble antibody with respect to thermodynamic parameters. It was found that the apparent equilibrium constant of the antibody-adsorbed ball was reduced to approximately 1/2 sec of soluble antibody, which was predominately due to the decrease in unitary entropy change by adsorption.

Two-dimensional crystals of the Kdp-ATPase of Escherichia coli.

A variant form of the Kdp-ATPase of Escherichia coli was overproduced to a level approaching 37% of the protein in the inner membrane of this organism. Membranes from overproducing cells were prepared with an inside-out orientation. Incubation of the membranes on ice for 1-2 weeks in the presence of sodium vanadate resulted in the formation of two-dimensional crystals of the Kdp-ATPase. The calculated projection map of the p1 crystal form showed three prominent density peaks at a resolution of 22 A. This technique is a useful and simple method to obtain low-resolution structures of membrane proteins.

Inhibitory effect of ethanolamine plasmalogen on iron- and copper-dependent lipid peroxidation.

The effect of ethanolamine plasmalogen (EtnPm) on lipid peroxidation was investigated in liposomal suspension of egg yolk phosphatidylcholine. EtnPm inhibited iron- and copper-dependent peroxidation in the presence of preformed hydroperoxides, although it was not effective for radical initiator mediated lipid peroxidation. EtnPm resulted in complete binding of iron to liposomal lipids, suggesting that EtnPm exerts its inhibitory effect on lipid peroxidation through inhibiting preformed peroxide decomposition by trapping transition metal ions.

Comparison of absorption and metabolism of beta-sitosterol and beta-sitostanol in rats.

The fates of [4-14C]beta-sitosterol ([14C]S) and [4-14C]beta-sitostanol ([14C]HS) were compared after after oral or intravenous administration to rats. Excretion into feces of oral [14C]HS was significantly higher than that of [14C]S. More than 97% of [14C]HS and 88% of [14CS]S were recovered in the feces within 7 days. Thus, deposition of [14C]HS was negligible in the tissues that were examined. Turnover in serum of [14C]HS which was injected intravenously appeared to be more rapid than that of [14C]S; [14C]HS was excreted as neutral steroids at a rate more than twice that of [14C]S. The rate of excretion of [3H]cholesterol was slightly greater when HS was administered simultaneously. The liver contained significantly less radioactivity after [14C]HS than after [14C]S administration. More [14C]HS than [14C]S was present in esterified form in serum and liver. The ratio of sterol in very low density lipoprotein to that in high density lipoprotein was less for HS or S than for endogenous cholesterol; this was particularly marked with HS. These results suggest that HS would be a more effective hypocholesterolemic agent than S.

Lipid-lowering activity of phytostanols in rats.

In the present study, the hypocholesterolemic activity of phytosterols and phytostanols was compared. Phytostanols were prepared by hydrogenating a phytosterol mixture from corn oil and were fed at different levels (0.1-1.0%) to male rats for 10 to 14 days with or without cholesterol (1.0%). In an appropriate combination with cholesterol, phytostanols showed significantly greater activity in lowering the plasma and possibly liver cholesterol levels in comparison with the corresponding phytosterols. The stanols further stimulated the fecal recovery of cholesterol. The rate of intestinal absorption of phytostanols appeared obviously lower than that of phytosterols and thus the deposition into plasma and liver lipids was almost negligible.

Lymphatic transport of cholesterol in normocholesterolemic rats treated with pravastatin, an inhibitor of HMG-CoA reductase.

Lymphatic absorption and transport of cholesterol and triacylglycerols were examined in rats treated with pravastatin, an inhibitor of 3-hydroxy-3-methyglutaryl-CoA (HMG-CoA) reductase. Pravastatin-treatment for 1, 7 and 28 days did not affect the recovery of cholesterol and triacylglycerols during 24 h after the lipid administration: the recovery was 52-59% and 82-93% for cholesterol and triacylglycerols, respectively. Rats treated with pravastatin for 28 days had a higher lymphatic recovery of the lipids during 3-6 h after the lipid administration than did control rats. Pravastatin treatment did not affect the ratio of phospholipid to cholesterol in the gut mucosa, the fatty acid composition of the lymph and mucosal lipids. We concluded that an inhibitor of HMG-CoA reductase would exert no adverse effect on absorption of fat-soluble nutrients by gut.

Construction of a specific and sensitive sandwich enzyme immunoassay for 20 kDa human growth hormone.

Twenty kilodalton human growth hormone (20K-hGH) is a naturally occurring isoform lacking amino acid residues 32-46 of 22K-hGH. Due to this structural similarity to 22K-hGH, no one has constructed a specific and sensitive assay system for 20K-hGH, which can be used for measuring physiological concentration of this isoform in the circulation. To construct such a specific assay system, we have generated polyclonal antibodies (pAb) and monoclonal antibodies (mAbs) against recombinant 20K-hGH. Binding characteristics to 20K-, 22K-hGH and monkey GH (mGH) of these five mAbs, designated A23, B13, C02, D05, and D14, were analyzed by enzyme immunoassays (EIAs) and surface plasmon resonance analysis. Only one of them, the mAb D05, does not crossreact with 22K-hGH. It was also observed that mAb B13 does not crossreact with mGH, although the later is 96% homologous to hGH. Using these antibodies we have established several sandwich EIA systems for circulating 20K-hGH. The combination of A23 as a solid-phase antibody and B13 as a labeled antibody permitted both high sensitivity to 20K-hGH (< 0.1 ng/ml) and low cross-reactivities with 22K-hGH (< 2%), mGH (< 0.3%) and rat GH (< 0.1%). The clearances of administered 20K-hGH were determined by this combination in both rats and monkeys. In the assay of physiologically circulating 20K-hGH in humans, the combination of D05 and affinity-purified anti-20K-hGH pAb showed the highest sensitivity to 20K-hGH (< 10 pg/ml) and substantially no cross-reactivity with 22K-hGH (< 0.1%). The plasma 20K-hGH concentration in healthy female subjects was determined by this combination. The assay systems constructed here enables us to directly measure circulating 20K-hGH in physiological condition with no interference of 22K-hGH for the first time.

An unusual regressive germinal center, the 'FDC-only lymphoid follicle,' in lymph nodes of organ transplant recipients.

Follicular lesions include germinal center (GC) hyperplasia, regressive transformation of GCs, and follicle lysis. The present histologic, electron microscopic, and immunohistochemical study of six autopsy cases after organ transplantation accompanied by the administration of immunosuppressive drugs revealed a peculiar regression of lymph node GCs in two cases, which has not been noted previously. The histologic findings of the regressive GCs were classified into three patterns. In pattern A, the GCs had few or no lymphocytes and were surrounded by a poorly developed mantle zone-like structure. Apoptotic cell death of GC lymphocytes was found in a few GCs, but most GCs lacked tingible body macrophages. In pattern B, the GC lymphocytes and tingible body macrophages were absent, showing crowded follicular dendritic cells (FDCs) in a corpuscular shape. In pattern C, the lymphocytic mantle was absent. The GCs were smaller than those in the other patterns, and the shape was irregular because of disintegration of FDCs. The immunostaining for FDC markers revealed dispersed growth of FDCs. On electron microscopy, the lesions were composed of a dense mass of elliptical and oval cells without prominent cytoplasmic processes, a labyrinthlike structure, and emperipolesis of lymphocytes. The distinct desmosomelike adhesive junctions, specific electron microscopic features of FDCs, were evident. We propose to call these follicular lesions "FDC-only lymphoid follicles." It is speculated that this follicle may be evoked after preceding follicular hyperplasia with a complicated mechanism including increased apoptosis of GC lymphocytes and decreased lymphocyte migration to lymph node GCs caused by immunosuppressive drugs.