Michiyasu Awai

CAPILLARY GROWTH FROM REVERSED RAT AORTIC SEGMENTS CULTURED IN COLLAGEN GEL

The process of angiogenesis from aortic segments turned inside out and embedded in collagen gel was studied. Two to three days after inoculation, fibroblastic cells migrated from both ends of the segments. Later, capillary sprouts also appeared from both ends of the segments but not from the outer surface, even though there was a covering of endothelial cells. If the outer surface was injured, capillaries sometimes appeared at the damaged site. This may suggest that endothelial cells have more affinity for basement membrane than collagen gel and that they migrate only from an injured site. Immuno‐histochemical staining demonstrated factor VHI‐related antigen in the capillary structures but not in the fibroblastic cells. Electron microscopically, capillary lumina were lined with several endothelial cells, and fibroblastic cells had the characteristics of smooth muscle cells. Since these fibroblastic cells have been known to appear under angiogenetic conditions in vivo, they may play an important role in angiogenesis, and the present culture technique may be a useful model for studying this process. ACTA PATHOL JPN 38: 1503∼1512, 1988.

Transferrin Receptor Expression in Normal, Iron‐deficient and Iron‐overloaded Rats

The distribution of transferrin receptor (TfR) positive cells and their staining intensity were examined in the liver, duodenum, pancreas, spleen, kidney and brain of iron deficient, iron‐overloaded and normal Wistar rats to elucidate the regulatory effects of iron on TfR expression in vivo. Iron deficiency was produced by an iron deficient food and water regimen, and iron overload by repeated intraperitoneal injections of ferric nitrilotriacetate (Fe3‐NTA) for 12 weeks. In iron deficient rats, levels of hemoglobin (Hb= 5.9 ± 0.7) and serum iron (Sl = 29.9 ± 4.4) were lower, and total iron binding capacity (TIBC = 624.4 ± 72.7) was higher than in normal rats (Hb = 15.6 ± 0.9, Sl = 206.5 # 20.5, TIBC = 416.0 ± 56.0), and Wee versa for SI (217.7 ± 15.5) and TIBC (307.1 ± M45.4) in iron‐overloaded rats. In normal rats, TfR positive granules were observed in hepatocytes and Kupffer cells of the liver, absorptive epithelium of the duodenum, acinar and Langerhans islet cells of the pancreas, macrophages and red pulp erythro‐blasts of the spleen, and distal convoluted tubular epithelium of the kidney. Although the tissue distribution pattern of TfR positive cells was similar in normal, iron deficient and iron overloaded rats, the staining intensity and number of TfR positive cells were obviously higher in iron deficient, and lower in iron overloaded rats. We conclude that TfR expression is negatively regulated by the tissue concentration of iron. Acta Pathol Jpn 39: 759‐764, 1989.

Stimulation of glutathione synthesis in iron-loaded mice

We have previously reported that the iron-loading of mice, by feeding them carbonyl iron, caused an elevation of hepatic glutathione concentration and an increase in glutathione excretion from the liver (Kawabata, T., Ogino, T. and Awai, M. (1989) Biochim. Biophys. Acta 1004, 89-94). To elucidate the mechanism of glutathione elevation, hepatic cysteine concentration and gamma-glutamylcysteine synthetase (L-glutamate: L-cysteine gamma-ligase (ADP-forming), EC 6.3.2.2) activity were measured and possible changes in cysteine metabolism were also compared between iron-loaded and control mice. Hepatic cysteine concentration was higher in iron-loaded mice (185 +/- 12 nmol/g wet wt.) than in the controls (164 +/- 8 nmol/g wet wt.), and gamma-glutamylcysteine synthetase activity was also elevated in iron-loaded mice (34.3 +/- 3.2 nmol/mg protein per min) compared with the controls (28.6 +/- 3.8 nmol/mg protein per min). A comparison of the metabolic pathways with intravenously injected [35S]cysteine showed that organ distribution of the isotope was not significantly different, and also the rate of [35S]cysteine uptake into the hepatic glutathione fraction exhibited no difference between the two groups of mice. This shows that hepatic cysteine turnover may not be different between the two groups of mice. Since hepatic cysteine concentration was higher in iron-loaded mice, the apparently equal turnover of hepatic cysteine suggests that GSH synthesis may be elevated in iron-loaded mice. The high gamma-glutamylcysteine synthetase activity is suggested to stimulate GSH synthesis in iron-loaded mice.

Capillary growth of rat aortic segments cultured in collagen gel without serum.

The process of angiogenesis was studied under serum‐free conditions using rat aortic segments in three‐dimensional collagen gel. In serum‐free and growth factor free conditions, the capillaries formed networks and tube‐like structures, and the endothelial cells produced von Willebrand factor, laminin and type IV collagen, but the number of capillaries was lower and their growth was slower than in medium containing 20% fetal calf serum (FCS). Incorporation of bromodeoxyuridine (BrdU) and inhibition of growth by hydroxyurea suggested that capillary growth depended mainly on cell proliferation and not on migration. Capillary growth in PRMI 1640 or DM EM was similar and more efficient than in MEM. Only slight growth was seen in Medium 199 and HAM‐F12. The addition of serum to the medium accelerated capillary growth in proportion to the amount added. In serum‐free conditions, ITS(+) and fibroblast growth factor (FGF) promoted capillary growth, but not to a significant extent. There ware no differences in capillary growth among the gel matrices used (type I collagen, type I+ 11 collagen, type I IV collagen, fibrin and plasma clot).

Protective effects of glutathione against lipid peroxidation in chronically iron-loaded mice

To elucidate the protective effects of glutathione against iron-induced peroxidative injury, changes in the hepatic glutathione metabolism were studied in chronically iron-loaded mice. When the diets of the mice were supplemented with carbonyl iron, iron deposition occurred primarily in the parenchymal cells of the liver. In addition, expiratory ethane production was elevated, suggesting an enhancement in lipid peroxidation. In iron-loaded mice, the total hepatic glutathione contents were higher (6.21 +/- 0.53 mumol/g wet wt.) than in control mice (4.61 +/- 0.31 mumol/g wet wt.), primarily due to an increase in the reduced glutathione contents. The value of oxidized glutathione was also higher (98.5 +/- 8.1 nmol/g wet wt.) than in the controls (60.8 +/- 9.5 nmol/g wet wt.), and the ratio of oxidized glutathione to total glutathione increased. The excretion rate of glutathione from the hepatocytes in iron-loaded mice also increased. These observations suggest that chronic iron-loading of mice stimulates lipid peroxidation and oxidation of glutathione and that peroxidized molecules may be catabolized using reduced glutathione.

DISSEMINATED INTRAVASGULAR COAGULATION (DIC)

Renal tissues of 208 autopsied cases were examined. Malignant neoplasm with hematological malignancies often accompanied DIG. Tissue sections were stained with hematoxylin and eosin and Mallorys phosphotungstic acid hematoxylin (PTAH), and were applied for immunoperoxidase method (IP), using antisera against human fibrinogen, FDP‐D and FDP‐E. Histologically in 80 cases (38%) fibrin or fibrinogen related materials (FRMs) were observed in the glomerular capillary or the intratubular area or in both. FRMs were PTAH or IP positive or both in 23 of the 26 cases (88%) clinically diagnosed as DIC. In the remaining three cases anticoagulants probably Interfered with FRMs observation. This study showed the PTAH stain was nonspecific and insensitive to FRMs, and that IP was necessary for a pathological diagnosis of DIC. The presence of FRMs in the renal tubuli is an important finding in confirming DIG. DIC may be present histologically in the absence of clinical DIC symptoms.

Lipid peroxidation and cytotoxicity of Ehrlich ascites tumor cells by ferric nitrilotriacetate

Ferric nitrilotriacetate, which causes in vivo organ injury, induced lipid peroxidation and cell death in Ehrlich ascites tumor cells in vitro. The process was inhibited by butylated hydroxyanisole and enhanced by vitamin C and linolenic acid, indicating a close relationship between cytotoxicity and the lipid peroxidizing ability of Fe3+ NTA. The cytotoxicity was suppressed by glucose and a temperature below 20 degrees C. Lipid peroxidation of Fe3+ NTA-treated cells was greater at 0 degree C than at 37 degrees C, contrary to results with Fe3+ NTA-treated plasma membranes of Ehrlich ascites tumor cell. These results suggested that metabolism and membrane fluidity are important factors in the expression of the Fe3+ NTA-induced cytotoxicity. H2O2 showed a lower cytotoxicity than did Fe3+ NTA but a greater lipid peroxidizing ability. H2O2 appeared to damage the cells less, and was quenched rapidly by cellular metabolism unlike Fe3+ NTA. In transferrin-free medium, Ehrlich ascites tumor cell readily incorporated Fe3+ NTA, and iron uptake was greater than NTA-uptake in Fe3+ NTA-treated cells, suggesting that Ehrlich ascites tumor cell incorporated iron from Fe3+NTA and metabolized it into an inert form such as ferritin.

Transferrin Receptors and Selective Iron Deposition in Pancreatic B Cells of Iron‐overloaded Rats

Iron overload was produced in Wistar rats by repeated intraperitoneal injections of ferric nitrilotriacetate (Fe3+‐NTA) for one to six months. Pancreatic tissues from these iron‐overloaded rats and untreated controls were examined for insulin (for B cells), glucagon (for A cells), transferrin receptor (TfR), transferrin (Tf) and ferritin (Ft) using immunohistochemical methods, and for iron by histochemical Berlin blue staining. In the islets of iron‐overloaded rats, increased Ft staining appeared prior to deposition of Berlin blue‐stainable iron, and the staining intensity of Ft and iron was stronger in B cells than in A cells. In the islets of untreated control rats, the staining intensity of TfR was stronger in B cells than in A cells. TfR staining of the islets was weaker in iron‐overloaded rats than in the controls. These findings suggest that 1) iron uptake by islet cells in vivo is regulated and mediated by TfR, 2) intracytoplasmic Ft transforms into stainable iron in iron‐overloaded rats, and 3) predominance of TfR expression in B cells may result in selective deposition of iron and predispose B cells to damage and diabetes mellitus in iron‐overloaded rats. Acta Pathol Jpn 41: 647–652, 1991.

Rapid in vitro transformation system for liver epithelial cells by iron chelate, Fe-NTA

We have previously found toxic effects of iron chelate, Fe-NTA on cultured normal rat liver epithelial cells (RL34). In the present study, when RL34 cells were exposed to 50 μg/ml iron of Fe-NTA for 15 days, besides the expected cytolytic effects in most cells, the appearance of resistant cells was observed. The resistant cells showed drastic morphological transformation, grew in soft agar, and induced hepatocellular carcinomas when transplanted into syngeneic newborn rats in a short period of time. Since DNA instability in the transformed cells was ascertained by differential AO staining, it is suggested that DNA damage by Fe-NTA is of a critical importance for extremely rapid neoplastic transformation of normal epithelial cells.

Glial Fibrillary Acidic Protein Immunoreactivity of Chondrocytes in Immature and Mature Teratomas

The immunoreactivity of chondrocytes for glial fibrillary acidic protein (GFAP), other intermediate filament proteins and S‐100 protein was studied in formalin‐fixed paraffin‐embedded sections. A total of 95 cartilage specimens were examined from five immature teratomas, 12 mature teratomas, and a teratocarcinoma. GFAP‐immunoreactive chondrocytes were abundant in immature cartilages, and as the cartilages maturated, these chondrocytes decreased and became distributed peripherally. Elastic cartilage had more GFAP‐immunoreactive chondrocytes than non‐elastic cartilage. GFAP‐immunoreactive cartilage was often located close to central nervous tissue. lmmunostaining for vimentin and S‐100 protein revealed extensive distribution of immunoreactive chondrocytes in immature and mature cartilages, but in mature cartilage, chondrocytes at the center had less vimentin immunoreactivity. GFAP‐immunoreactive chondrocytes also showed apparent immunostaining for vimentin. There was no difference in immunohistochemical staining for the α and α subunits of S‐100 protein. The immunoreactivities of teratoma cartilage specimens were quite similar to those of respiratory tract cartilage.