Dexamethasone modulates rat renal brush border membrane phosphate transporter mRNA and protein abundance and glycosphingolipid composition.

Glucocorticoids are important regulators of renal phosphate transport. This study investigates the role of alterations in renal brush border membrane (BBM) sodium gradient-dependent phosphate transport (Na-Pi cotransporter) mRNA and protein abundance in the dexamethasone induced inhibition of Na-Pi cotransport in the rat. Dexamethasone administration for 4 d caused a 1.5-fold increase in the Vmax of Na-Pi cotransport (1785 +/- 119 vs. 2759 +/- 375 pmol/5 s per mg BBM protein in control, P < 0.01), which was paralleled by a 2.5-fold decrease in the abundance of Na-Pi mRNA and Na-Pi protein. There was also a 1.7-fold increase in BBM glucosylceramide content (528 +/- 63 vs. 312 +/- 41 ng/mg BBM protein in control, P < 0.02). To determine whether the alteration in glucosylceramide content per se played a functional role in the decrease in Na-Pi cotransport, control rats were treated with the glucosylceramide synthase inhibitor, D-threo-1-phenyl-2-decanoyl-amino-3-morpholino-1-propanol (PDMP). The resultant 1.5-fold decrease in BBM glucosylceramide content (199 +/- 19 vs. 312 +/- 41 ng/mg BBM protein in control, P < 0.02) was associated with a 1.4-fold increase in Na-Pi cotransport activity (1422 +/- 73 vs. 1048 +/- 85 pmol/5 s per mg BBM protein in control, P < 0.01), and a 1.5-fold increase in BBM Na-Pi protein abundance. Thus, dexamethasone-induced inhibition of Na-Pi cotransport is associated with a decrease in BBM Na-Pi cotransporter abundance, and an increase in glucosylceramide. Since primary alteration in BBM glucosylceramide content per se directly and selectively modulates BBM Na-Pi cotransport activity and Na-Pi protein abundance, we propose that the increase in BBM glucosylceramide content plays an important role in mediating the inhibitory effect of dexamethasone on Na-Pi cotransport activity.

Expression of glycosphingolipids in serum-free primary cultures of mouse kidney cells: male-female differences and androgen sensitivity.

The expression of neutral glycosphingolipids was examined in primary kidney cell cultures derived from adult male and female beige mutant mice (C57BL/6J;bgj/bgj) with enrichment for proximal tubule cells during preparation of explants and using defined serum-free medium for the culture conditions. Cells proliferated for 7 days in vitro to provide confluent or nearly confluent monolayers of epithelial-type growth indicative of proximal tubule cells. The male vs female differences in neutral glycosphingolipids seen in the kidney in vivo were retained in these 7 day cultures. Cultures derived from males contained galacto- and digalactosylceramides whereas those from females did not express these types of glycolipids. Also, male cells had higher ratios of sphingosine: phytosphingosine containing species in Nfa (non-hydroxy fatty acid) globotriaosylceramide and in glucosylceramide than females. The shift in sphingosine: phytosphingosine to male ratios in Nfa globotriaosylceramide and in glucosylceramide could be stimulated in female kidney cells by treatment with 10(-5) M testosterone or 5 alpha-dihydrotestosterone. The male-specific expression of neutral glycosphingolipids, then, appears to be stable character of male-type differentiation in mouse kidney that is passed on during proliferation in culture. Female kidney cells retain an ability to respond to androgens with specific changes in neutral glycosphingolipid expression during 7 days of growth in vitro in serum-free conditions, but do not respond with the induction of the male-specific glycolipids galacto- and digalactosylceramides as seen in vivo.

Pathogenesis of Shigella diarrhea: XVII. A mammalian cell membrane glycolipid, Gb3, is required but not sufficient to confer sensitivity to Shiga toxin.

Shiga toxin recognizes a galactose-alpha 1-->4-galactose terminal glycolipid, globotriaosylceramide (Gb3), in sensitive mammalian cells and is translocated by endocytosis to the cytoplasm, where it blocks protein synthesis. To determine if Gb3 is both required and sufficient for toxicity, Gb3 content in cells was altered by blocking key biosynthetic or degradative path enzymes with specific inhibitors. The resulting decrease or increase in cellular Gb3 was associated with a decrease or increase in binding of and response to Shiga toxin. Toxin-resistant Gb3-deficient variants of sensitive cells fused with liposomes containing Gb3 but not globotetraosylceramide (Gb4) became susceptible, whereas fusion of Gb3 liposomes to naturally resistant Gb3-deficient CHO cells increased toxin binding but not cytotoxicity. These data demonstrate that Gb3 is required, but not sufficient, for the action of Shiga toxin and suggest the existence of a toxin translocation mechanism linked to surface glycolipids that is not expressed in CHO cells.

The accumulation and metabolism of glycosphingolipids in primary kidney cell cultures from beige mice.

In the normal C57BL/6J male mouse a specific subset of the kidney glycosphingolipids which is associated with multilamellar bodies of lysosomal origin and represents about 10% of the total kidney glycolipids, is excreted into the urine each day. This excretion is blocked and glycosphingolipids accumulate in the kidney of bgJ/bgJ mutants of this strain. To examine this process in vitro, glycosphingolipid metabolism and excretion were studied in beige mouse kidney cell cultures. Primary kidney cell cultures from male C57BL/6J control and bgJ/bgJ beige mutants were grown in D-valine medium and glycosphingolipids labeled with [3H]palmitate. As we have shown previously, the giant lysosomes of altered morphology were maintained in cultures of the beige kidney cells. Beige-J and control cells synthesized the same types of glycosphingolipids, but the mutant cells had quantitatively higher levels of these compounds than control cells, as determined by high performance liquid chromatography. Beige-J cells incorporated more [3H]palmitate into glycosphingholipids than control cells on a cpm/mg protein basis and the specific activity (cpm/pmole glycosphingolipid) was lower in beige cells. Medium from beige-J cells accumulated more glycosphingolipids than that from control cells in a 24 h period. The glycosphingolipids released into the medium as determined by HPLC were primarily non-lysosomal types and both control and mutant cells retained the glycosphingolipids associated with lysosomal multilamellar bodies excreted in vivo. The elevated levels of lysosomal glycosphingolipids and the dysmorphic lysosomes in primary cultures of beige cells, then, are not caused by a mutant block in secretion of lysosomes.

Lipid composition of lysosomal multilamellar bodies of male mouse urine.

Previous studies from our laboratory have shown that male C57BL/6J mice excrete into the urine multilamellar lysosomal bodies that contain specific neutral glycosphingolipids. These mice excrete approximately 20-30% of their kidney glycolipids each day. The significance and function of this secretion of multilamellar lysosomal organelles is unknown. To characterize these excreted bodies further, we report here their neutral lipid and phospholipid composition. The bodies were collected by differential centrifugation, extracted with chloroform-methanol, and lipids were fractionated into neutral lipids, glycolipids, and phospholipids. The neutral lipids consisted primarily of cholesterol, dolichol, and ubiquinone. The phospholipid fraction consisted primarily of a single molecular species of phosphatidylcholine. This lipid which comprises more than 90% of the total phospholipids was found to contain 16:0 ether and C22:6 n-3 fatty acid as determined by gas-liquid chromatography-mass spectrometry. The glycosphingolipids as reported previously consisted primarily of galabiosylceramides and globotriaosylceramides. This membrane lipid composition is different from any previously reported cellular organelle.

Quantitation of the rabbit intestinal glycolipid receptor for Shiga toxin. Further evidence for the developmental regulation of globotriaosylceramide in microvillus membranes.

Shiga toxin, produced by Shigella dysenteriae 1, causes enterotoxic, cytotoxic, and neurotoxic effects, which may be mediated by a glycolipid receptor, globotriaosylceramide, Gb3. To study the relationship of this receptor and toxin effects, globotriaosylceramide was quantitated and further characterized in rabbit small intestinal microvillus membranes at various ages. Glycolipids were extracted from rabbit microvillus membranes, purified on Unisil columns, and quantitated by high-performance liquid chromatography. The major glycolipid peaks were hydroxylated fatty acid-containing glucosylceramide, lactosylceramide, and globotriaosylceramide. There was a marked increase of globotriaosylceramide levels with age, ranging from 0.02 to 16.2 pmol/micrograms microvillus membrane protein in neonates and adults, respectively. The globotriaosylceramide peak was susceptible to alpha-galactosidase treatment, which produced an elevation in the lactosylceramide peak, but markedly reduced globotriaosylceramide content in 34-day-old rabbits. Binding of iodinated Shiga toxin to globotriaosylceramide was documented on high-performance thin-layer chromatography plates by autoradiography. The glycolipid receptor for Shiga toxin in rabbit microvillus membranes is thus a hydroxylated fatty acid-containing globotriaosylceramide. This moiety is virtually absent in neonates and gradually increases with age. Quantitative differences in globotriaosylceramide may be the underlying basis for the age-specific differences in functional responsiveness of rabbit intestinal tissue to Shiga toxin.

Glycolipid stage-specific embryonic antigens (SSEA-1) in kidneys of male and female C57BL/6J and beige adult mice.

SSEA-1 glycolipids from the kidneys of normal male and female as well as beige mutant mice were isolated and their structures were examined by component analysis, mass spectrometry, immunoblotting, and permethylation studies. These antigens were shown to be extended globoside derivatives as reported by Sekine et al. (1987. J. Biochem. 101:553-562). Quantitative high performance liquid chromatography analyses revealed that the concentration of SSEA-1 glycolipids were four-to fivefold greater in male than female mice. Essentially no SSEA-1 glycolipids were excreted in the urine of normal male mice and thus are not components of the multilamellar lysosomes normally excreted. Testosterone is known to induce the hypertrophy of proximal tubule cells that involves the formation of multilamellar lysosomes and results in the accumulation and excretion of these bodies and associated lysosomal enzymes and specific glycolipids. The present results indicate that in male mice there is also an increase in subcellular structures that contain SSEA-1 glycolipids. The amount of SSEA-1 glycolipids in male beige mice were greater than in normal mice on a per kidney basis. Thus, the increase is in proportion to the kidney hypertrophy seen in beige mouse kidneys. Beige mutant mice appear to have a primary defect in the excretion of multilamellar lysosomes which produces a secondary hypertrophy with an accompanying increase in SSEA-1 glycolipids.

Biochemical and morphological characterization of primary kidney cell cultures from beige mutant mice.

Primary kidney cultures from adult beige-J (bgJ/bgJ) mice were selected for epithelial cell growth using D-valine medium. After 2 weeks of attachment and proliferation in vitro, the cells form a confluent or nearly confluent monolayer that retains several phenotypic characteristics of the beige-J mutant. These include large, multilamellar inclusion bodies that are apparently dysmorphic lysosomes, and higher concentrations of neutral glycosphingolipids and dolichols than control cells. beta-Glucuronidase activity, used as a lysosomal enzyme marker, is not elevated in beige-J-cultured kidney cells compared with controls, as it is in the intact kidney. The high levels of beta-glucuronidase activity in both control and mutant cells may mask expression of this difference in vitro. The action of the beige-J mutation in kidney cells is thought to be due to a block in exocytosis that results in the accumulation of abnormal lysosomes and their components. The maintenance of the beige phenotype in vitro indicates that the mutation is not suppressed in primary kidney cell cultures. The expression of the beige phenotype in vitro should be useful for studies concerning the primary lesion of this mutation.

Glycosphingolipid patterns in primary mouse kidney cultures.

Primary kidney cultures from C57BL/6J mice, 6 weeks of age or older, were produced using D-valine medium to select for epithelial cell growth. After allowing the cells to attach and proliferate for 1 week following plating, medium was changed once per week. Cells formed nearly confluent monolayers during the second week of culture. The cultured cells contained all of the glycosphingolipids seen in the adult kidney, analyzed by high performance liquid chromatography as their perbenzoyl derivatives. Glucosylceramide, however, was highly predominant in the cultured cells, whereas dihexosyl- and trihexosylceramides predominate in the intact kidney. Sex differences in glycolipid contents found in the intact kidney were also apparent in these cultured cells: The concentration of neutral glycolipids, in general, was higher in male cells than in those derived from females, and the male-specific glycolipid nonhydroxy fatty acid digalactosylceramide was high in male cells but very low in female cells. Neutral glycosphingolipids were labeled in 2-week-old cultures using [3H]palmitate. The [3H]palmitate was incorporated into all of the glycolipids within 2 hr of labeling. Hence, adult mouse kidney cells in D-valine medium retain their differentiated characteristics for a sufficient period of time to allow investigation of glycolipid syntheses in monolayer cultures of epithelial cells derived from this organ.

Glucosylceramide in the androgen-responsive kidney of the lizard Anolis carolinensis.

During the spring breeding season of the American chameleon, Anolis carolinensis, elevated levels of glucosylceramides which contain hydroxy fatty acids are produced in the kidneys of males but not females. Hyperproduction of this glycolipid is also induced by testosterone. The testosterone-induced hypertrophy of epithelial cells in the proximal tubules of the mouse kidney seems an analogous phenomenon and an elevated concentration of specific glycolipids in the male mouse kidney has been previously demonstrated. Thus the formation of renal glycolipids in response to testosterone may be a widespread feature in vertebrates.

Biosynthesis of neutral glycosphingolipids in kidney slices from male and female mice.

Previous reports from our laboratory (1981. J. Biol. Chem. 256: 13112-13120 and 1983. Endocrinology. 113: 251-258) showed the absence of Nfa-GalCer and Nfa-GaOse2Cer in kidneys of several strains of female mice. These lipids are always present in male kidneys and several other glycolipids are also elevated in males. To test whether this phenomenon is due to lowered biosynthesis in females, glycosphingolipid formation was assessed in kidney slices with [3H]galactose as precursor. The glycolipids were extracted after various incubation periods (from 30 min to 90 min) and individual glycolipids were separated and quantitated by high performance liquid chromatography and radioactivity was determined. The rate of formation of hydroxy fatty acid-containing galactosylceramide was the same in both sexes. The glycolipids which were low or not detectable in female kidney, Nfa-GalCer, Nfa-GaOse2Cer and Hfa-GaOse2Cer were rapidly labeled in the male kidney slices. These results suggest that nonhydroxy fatty acid-containing ceramide:UDP-Gal galactosyltransferase and hydroxy fatty acid-containing galactosylceramide:UDP-Gal galactosyltransferase have elevated activities in males. While the glucosylceramides are labeled at the same rates in both sexes, lactosylceramide appears to be labeled at higher rates in the male tissue. This suggests that glucosylceramide:UDP-Gal galactosyltransferase also has elevated activity in males. In addition, these data show that monohexosylceramides with different ceramide compositions are labeled at different rates.

Altered secretion and accumulation of kidney glycosphingolipids by mouse pigmentation mutants with lysosomal dysfunctions.

The kidney and urine glycosphingolipids of five pigmentation mutants which are known to have altered secretion of kidney lysosomal enzymes were examined. Among 34 pigmentation mutants which have been studied (Novak, E. K., Wieland, F., Jahreis, G. P., and Swank, R. T. (1980) Biochem. Genet. 18, 549-561) eight are known to have a 1.5- to 2.5-fold increase in kidney beta-glucuronidase in testosterone-treated females. These mutants appear to have defects in lysosomal processing, and because the mutations are at separate loci, each mutant probably affects different steps in assembly and/or exocytosis of lysosomes and related subcellular organelles. To test whether the neutral glycosphingolipids, galabiglycosylceramides, and globotriglycosylceramides thought to be associated with kidney lysosomes (McCluer, R. H., Williams, M. A., Gross, S. K., and Meisler, M. H. (1981) J. Biol. Chem. 256, 13112-13120) also exhibit abnormal secretion in the mutants with lysosomal enzyme abnormalities, the mutants beige-J, pale ear, light ear, pallid, and ruby eye-2-J were studied. The kidney and urine neutral glycosphingolipids from males of each mutant and C57BL/6J control mice were analyzed by high performance liquid chromatography. Beige-J, light ear, and pale ear showed marked increases in total kidney glycolipids; globotriglycosylceramides accounted for the bulk of the increase. Ruby eye-2-J showed less marked but significantly increased quantities of one galabiglycosylceramide and the globotriglycosylceramides in kidney. Pallid showed no significant increase in total kidney glycolipids but the globotriglycosylceramides appeared slightly elevated. In terms of the decrease in total urinary glycosphingolipids, the mutants fell into 2 categories. Beige-J, light ear, and pale ear were severely affected, whereas ruby eye-2-J and pallid were affected to a much lesser extent. Within the most severely affected group the excretion of the globotriglycosylceramides was more severely affected than that of the galabiglycosylceramides. The galabiglycosylceramides and globotriglycosylceramides appear to be specific markers of lysosomal membranes, but the independent behavior of these two classes of lipids during testosterone induction in normal mice and the differential effects on their secretion by different mutants indicate that they do not always exist in a characteristic ratio in a single type of subcellular organelle. All of the mutants accumulate organelles in their kidney proximal tubules which have distinct morphological characteristics as seen by electron microscopy.

Testosterone-responsive mouse kidney glycosphingolipids: developmental and inbred strain effects.

Mouse kidney and urinary glycosphingolipids from developing C57BL/6J and adults of several other inbred strains were analyzed by high performance liquid chromatographic techniques. Glycosphingolipids from male and female C57BL/6J kidneys were similar until the fifth week of age. Galactosylceramide containing nonhydroxy fatty acids and galabiglycosylceramide containing nonhydroxy fatty acids first appeared in male kidneys, followed by an increase in galabiglycosylceramide containing hydroxy fatty acids. Galabiglycosylceramide was observed in male urine from the earliest collection period (26 days of age). At 5 weeks, globotriglycosylceramides were present in male urine, and by 6 weeks, they became the major glycolipid species. Analysis of the glycosphingolipids from adult male and female DBA/2J, CBA/J, C3H/HeJ, and AKR/J kidneys revealed that galactosylceramides and galabiglycosylceramides which contain nonhydroxy fatty acids were absent in all females and present in all males. The globotriglycosylceramides were elevated in male kidneys of all strains. Galabiglycosylceramides and globotriglycosylceramides were present in male urine of all strains. Each strain exhibited a characteristic pattern of urinary glycosphingolipids which varied not only in the different levels of di- and triglycosylceramides but also in the ratio of components that are distinguished by their fatty acid and long chain base composition. These data provide evidence that in several inbred strains of mice, testosterone induces the production of specific di- and triglycosylceramides, which are components of lysosomal organelles that are normally excreted in the urine.

Testosterone effects on the induction and urinary excretion of mouse kidney glycosphingolipids associated with lysosomes.

Male C57BL/6J mice were shown to excrete in their urine approximately 50 times as much glycolipid as do female mice. These lipids were primarily galabiglycosylceramides and globotriglycosylceramides which contained nonhydroxy and hydroxy fatty acids, sphingosine, and phytosphingosine as components of their ceramide moieties. The lipids were shown to be associated with multilamellar bodies. Neutral glycosphingolipids from kidney and urine were separated and analyzed by high performance liquid chromatographic techniques. Male kidneys were shown to contain about twice as much neutral glycosphingolipid as female kidneys. The female kidneys were virtually devoid of galactosylceramides and galabiglycosylceramides. Implantation of testosterone pellets in female mice resulted in a concomitant increase in urinary glycosphingolipids (greater than 100-fold) and in urinary lysosomal enzymes (about 10-fold). In mutant light ear mice, which are defective in excretion of lysosomal enzymes, the urinary excretion of glycolipids is also blocked. Taken together these data suggest that testosterone induces the formation of multilamellar kidney lysosomes which contain a complement of glycosphingolipids and which are normally excreted in the urine.

Quantitative analysis of monosialogangliosides by high-performance liquid chromatography of their perbenzoyl derivatives.

A quantitative high-performance liquid chromatographic method for the analysis of monosialogangliosides as their perbenzoyl derivatives has been devised. Samples containing as little as 3 nmol were converted to their perbenzoyl derivatives by reaction with 0.1 ml of 10% benzoyl chloride in pyridine at 60 degrees C for 1 hr. The products were purified by silicic acid chromatography and analyzed by high-performance liquid chromatography (HPLC). The HPLC analysis was performed with a 50 cm X 2.1 mm LiChrosphere SI 4000 column and a linear gradient of 7-23% dioxane in hexane in 18 min. Detection was at 230 nm. The detector response was found to be proportional to the amount of monosialoganglioside analyzed. As little as 50 pmol of injected material could be conveniently quantitated. The overall yield from derivatization and chromatography, as determined with radiolabeled GM1, was found to be 86%. To take advantage of the high sensitivity of the HPLC, a small-scale isolation method for gangliosides was devised. This method coupled with HPLC isotope dilution analysis was used to analyze the GM3 content of 1 ml of human plasma.

Glycolipid synthesis in normal and virus-transformed hamster cell lines.

Studies have been continued on the synthesis of glycolipids by the NIL 2 line of hamster cells. Several clones were isolated from this line. These clones vary in morphology, saturation density, and glycolipid composition. Contrary to expectation there was no correlation between saturation density and complexity of the glycolipid pattern. In fact, the clone with the highest saturation density was the only one to show the complete set of glycolipids found previously in NIL 2 cells. All untransformed NIL clones show an increase in the level of "higher" (more than two saccharides/ceramide) glycolipids as the cells approach confluence. In the line containing all three "higher" neutral glycolipids, all three increased as cells approached saturation density. In the line containing only hematoside, this compound increased at confluence. Cells transformed by hamster sarcoma virus or polyoma virus showed no density-dependent glycolipid synthesis.