Joel G. Parkes

The composition of phospholipids in outer and inner mitochondrial membranes from guinea-pig liver

Abstract 1. 1. Mitochondria, outer and inner mitochondrial membranes and microsomes were prepared from guinea-pig liver. 2. 2. The inner mitochondrial membrane exclusively contained succinate:cytochrome c reductase while outer membrane was characterized by a high level of monoamine oxidase. Low levels of microsomal contamination of the outer membrane were indicated by assays of glucose-6-phosphatase and NADPH:cytochrome c reductase. 3. 3. The phospholipids were separated by thin-layer chromatography and the fatty acids analysed by gas-liquid chromatography. In all the membranes lecithin had a relatively high content of stearate and linoleate. Phosphatidylethanolamine contained less linoleate than lecithin but had a significantly ligher level of arachidonate. Phosphatidylinositol was the most highly saturated lipid, containing large amounts of stearate, while cardiolipin was over 90% unsaturated, due mostly to linoleic acid. It is concluded that in guinea-pig liver each phospholipid has a rather distinctive fatty acid pattern which is approximately the same in all membranes. 4. 4. Some implications of the results are discussed.

Modulation by iron loading and chelation of the uptake of non-transferrin-bound iron by human liver cells

Hepatic non-transferrin-bound Fe (NTBI) flux and its regulation were characterized by measuring the uptake of Fe from [59Fe]/nitrilotriacetate (NTA) complexes in control and Fe-loaded cultures of human hepatocellular carcinoma cells (HepG2). Exposure to ferric ammonium citrate (FAC) for 1 to 7 days resulted in a time- and dose-dependent increase in the rate of NTBI uptake. In contrast to previous studies showing a dependence of the rate of Fe uptake on extracellular Fe, this was positively correlated with total cellular Fe content. The Fe3+ chelating agents deferoxamine (DFO), 1,2-dimethyl-3-hydroxypyrid-4-one (CP 020) and 1,2-diethyl-3-hydroxypyrid-4-one (CP 094) prevented or diminished the increase in NTBI transport when present during Fe loading and reversed the stimulation in pre-loaded cells in relation to their abilities to decrease intracellular iron. Although saturation of the Fe uptake process was not achieved in control cells, kinetic modelling to include linear diffusion-controlled processes yielded estimated parameters of Km = 4.3 microM and Vmax = 2.6 fmol/micrograms protein/min for the underlying process. There was a significant increase in the apparent Vmax (31.2 fmol/micrograms protein per min) for NTBI uptake in Fe-loaded cells, suggesting that Fe loading increases the number of a rate-limiting carrier site for Fe. Km also increased to 15.2 microM, comparable to values reported when whole liver is perfused with FeSO4. We conclude that HepG2 cells possess a transferrin-independent mechanism of Fe accumulation that responds reversibly to a regulatory intracellular Fe pool.

Speciation of tissue and cellular iron with on-line detection by inductively coupled plasma-mass spectrometry

Iron accumulating to excess in tissues of humans and animal models occurs mainly as complexes with transferrin, ferritin, other hemoproteins, and insoluble hemosiderin particles. To determine the distribution of Fe amongst these molecular species, we have used inductively coupled plasma-mass spectrometry as a means of on-line, isotope-specific detection for their liquid chromatographic separation. The stable isotope 57Fe is a suitable isotope for monitoring the Fe content of each fraction, and its availability at high isotopic enrichment makes it an attractive choice for tracer studies when the use of a radioisotope is undesirable, e.g., in human subjects. The detection system offers the advantages of high sensitivity (detection limits in the parts per billion range), a wide dynamic range (linearity of the calibration curve over several orders of magnitude), and on-line analysis facilitating real-time evaluation of the chromatographic separation, in addition to isotope-specific information. The Fe distributions in healthy rat livers, liver and heart tissue from Fe-loaded human subjects, and human hepatocyte cultures are reported. The ferritin:hemosiderin ratio in these samples is shown to be an indicator of the degree of Fe loading and correlates well with that determined by Zeeman-corrected electrothermal atomic absorption as an alternative means of detection.

Characterization of Fe2+ and Fe3+ transport by iron-loaded cardiac myocytes

Plasma iron overload causes cardiac iron accumulation leading to toxicity and organ failure. In order to understand the basis of iron acquisition, we examined mechanisms of Fe3+ and Fe2+ uptake in control and iron-loaded cardiomyocyte cultures. Iron loading increased rates of Fe3+ and Fe2+ uptake, primarily by increasing Vmax. Inhibition of Fe3+ transport by impermeable Fe2+ chelators and the presence of a cell surface ferricyanide reductase activity are consistent with a role for redox cycling in Fe3+ uptake. However, flavoproteins and copper-dependent oxidases known to be required for redox-active iron transport in yeast do not appear to be involved in iron uptake by cardiac myocytes, nor do the abundant cardiac L-type Ca2+ channels. The data suggest that both redox states of iron contribute to cardiac iron accumulation in iron overload.

Differential accumulation of non-transferrin-bound iron by cardiac myocytes and fibroblasts.

Cardiac myocytes accumulate iron preferentially over fibroblast-like non-myocytes, both in clinical iron overload and when the cells are grown together in culture. In order to determine whether this reflects the tissue context or is an inherent property of the cells, we studied iron transporters, transport kinetics, and iron efflux in homogeneous cultures of rat cardiac myocytes and fibroblasts. In both cells, the rate of uptake of 59Fe from transferrin was insignificant, compared to the rate of uptake from non-transferrin-bound iron (NTBI). Expression of transferrin receptor mRNA and protein, and divalent metal transporter 1 (DMT1) mRNA, could not account for any difference in iron accumulation, and proportional efflux after iron loading was similar in both cells. Nevertheless, iron accumulation from NTBI over 72 h was greater in myocytes as determined by histological staining and quantitative iron measurement. NTBI uptake was greater for Fe2+ than Fe3+ in both cells, was increased by iron loading in both cells to a similar extent, and was characterized bysimilar Michaelis constants (K(m)) in all cases (redox state and presence or absence of iron loading). However, V(max) values were about 10-fold higher in myocytes. We conclude that preferential iron accumulation in cardiac myocytes, compared to fibroblasts, is due to a higher capacity of the NTBI-transporter system, and reflects an inherent difference in NTBI acquisition by the individual cell types.

Modulation of stellate cell proliferation and gene expression by rat hepatocytes: effect of toxic iron overload.

Mechanisms by which hepatocytes and transdifferentiated hepatic stellate cells (HSC) initiate liver fibrosis in chronic iron toxicity are unknown. This study was to determine if factors in media from control and iron-loaded rat hepatocyte cultures modulate HSC gene expression and proliferation. Conditioned medium (CM) from both control and iron-loaded hepatocytes increased serum-stimulated DNA synthesis by HSC to 140% of control values (P<0.05). Heating CM (15 min, 80 degrees C) caused a suppression of DNA synthesis that was partially reversed by a TGF-beta-neutralizing antibody. Addition of TGF-beta1 reproduced the suppression. Levels in HSC of mRNA for collagen type I, collagen type IV, TGF-beta, and plasminogen activator inhibitor-1 were unaffected by exposure to CM but increased significantly when CM from iron-loaded hepatocytes was heat-treated. In HepG2 cell cultures, iron loading increased total (but not activated) TGF-beta secretion into the medium approximately 2-fold. We conclude that increased secretion of latent TGF-beta by hepatocytes injured by iron is a potential factor influencing fibrogenic behavior of HSC.

Microsomal induction, alcohol, and lipoprotein metabolism: Is there a three-way relationship?

The role of ethanol as a microsomal enzyme-inducing agent and as a modulator of lipid metabolism is reviewed. In an attempt to ascertain the mechanisms underlying the latter effects we examined the changes in hepatic triglyceride lipase (HTGL), hepatic high density lipoprotein (HDL) binding, and apolipoprotein secretion mediated by ethanol in a variety of experimental situations. Chronic administration of ethanol to rats decreased the ability of the liver to secrete HTGL, but primary liver cultures prepared from both ethanol and sucrose-fed rats secreted more HTGL when acutely exposed to ethanol over a 3-day period than when grown in a control medium. Hep G2 cells when grown in ethanol-containing medium for 14-28 days manifested increased HDL-binding capacity; apolipoprotein-A1 secretion was increased by ethanol but apolipoprotein B secretion was not affected. These findings suggest that increased plasma HDL concentrations which follow chronic ethanol ingestion may be due, at least in part, to increased hepatic secretion and reduced intravascular conversion of the lipoprotein despite enhanced reuptake by the liver; they are not consistent with an ethanol-mediated alteration in very low density lipoprotein secretion by the liver.

Synthesis In vivo of phospholipids op liver mitochondria and endoplasmic reticulum from glycerol and fatty acids

Abstract The kinetics of labelling of phospholipids of mitochondria and endoplasmic reticulum in guinea-pig liver in vivo from both [2-3H]glycerol and I-14C-labelled fatty acids are consistent with differential rates of transfer (phosphatidylinositol > phosphatidylcholine > phosphatidylethanolamine) from the endoplasmic reticulum to mitochondria. Double-label experiments suggest that the acyl ester bonds of constituent phospholipids are established largely in the endoplasmic reticulum and retained in mitochondria and that any independent acylating activity in mitochondria must be small in comparison.

Lipolytic enzymes as markers of induction and differentiation.

Factors leading to microsomal enzyme induction are associated with hypertriglyceridemia in man. Phenobarbital (PB) increases hepatic synthesis of triglyceride but lowers its serum concentration in rats due to increased postheparin plasma activities of lipoprotein lipase (LPL) and hepatic triglyceride lipase (HTGL); these changes are accompanied by increased activity of these lipolytic enzymes in adipose tissue and liver. The present work explores the cellular mechanisms whereby PB increases the tissue content of these enzymes, using primary cultures of rat liver hepatocytes and a continuous cell line of mouse fibroblasts (preadipocytes) that undergo differentiation into mature fat cells. Secretion and synthesis of HTGL in primary rat hepatocytes increased 50% with insulin; when PB was added with insulin, activity was enhanced an additional 50%. By contrast, insulin inhibited HTGL secretion from the well differentiated rat hepatoma cell line, FU-5-5, C8, and this inhibition was partly overcome by PB. These results suggest that different control mechanisms govern the synthesis and secretion of HTGL in normal rat liver cells and hepatoma. In cultured pre-adipocytes (3T3-F442A) insulin promoted differentiation when added to confluent cultures. PB (0.5 mM) resulted in marked enhancement of conversion of adipocytes characterized by a two- to threefold increase in extracellular LPL and a 10-fold increase in intracellular enzyme. These results suggest that PB promotes conversion of uncommitted cells into pre-adipocytes at an early stage in the differentiation of adipose tissue.