Identification of a forskolin-like molecule in human renal cysts.

Renal cyst enlargement is increased by adenosine cAMP, which is produced within mural epithelial cells. In a search for modulators of cAMP synthesis cyst fluids from 18 patients with autosomal dominant or recessive polycystic kidney disease (PKD) were analyzed, and in 15 of them, a stable lipophilic molecule that increased cAMP levels, stimulated transepithelial chloride and fluid secretion, and promoted the proliferation of human cyst epithelial cells was characterized. With the use of HPLC-mass spectrometry, a bioactive lipid with the same mass spectral fingerprint, the same chromatographic retention time, and the same biologic properties as forskolin, a widely known, potent adenylyl cyclase agonist, has been isolated and identified within the cyst fluid. Forskolin is synthesized by the plant Coleus forskohlii, but its appearance or compounds like it have not been reported in animals. The origin of forskolin in patients with PKD was not revealed by this study. Synthesis by mural cyst epithelial cells or an exogenous source are the most likely possibilities. Forskolin is sold for weight management and as a cardiovascular tonic in health stores and through the Worldwide Web. It is concluded that forskolin may have a role in promoting the enlargement of cysts in autosomal dominant PKD and recommended that patients avoid oral and parenteral preparations that contain this compound.

Structure of PITPbeta in complex with phosphatidylcholine: comparison of structure and lipid transfer to other PITP isoforms.

Phosphatidylinositol transfer protein (PITP) is a ubiquitous eukaryotic protein that preferentially binds either phosphatidylinositol or phosphatidylcholine and catalyzes the exchange of these lipids between membranes. Mammalian cytosolic PITPs include the ubiquitously expressed PITPalpha and PITPbeta isoforms (269-270 residues). The crystal structure of rat PITPbeta complexed to dioleoylphosphatidylcholine was determined to 2.18 A resolution with molecular replacement using rat PITPalpha (77% sequence identify) as the phasing model. A structure comparison of the alpha and beta isoforms reveals minimal differences in protein conformation, differences in acyl conformation in the two isoforms, and remarkable conservation of solvent structure around the bound lipid. A comparison of transfer activity by human and rat PITPs, using small unilamellar vesicles with carefully controlled phospholipid composition, indicates that the beta isoforms have minimal differences in transfer preference between PtdIns and PtdCho when donor vesicles contain predominantly PtdCho. When PtdCho and PtdIns are present in equivalent concentrations in donor vesicles, PtdIns transfer occurs at approximately 3-fold the rate of PtdCho. The rat PITPbeta isoform clearly has the most diminished transfer rate of the four proteins studied. With the two rat isoforms, site-directed mutations of two locations within the lipid binding cavity that possess differing biochemical properties were characterized: I84alpha/F83beta and F225alpha/L224beta. The 225/224 locus is more critical in determining substrate specificity. Following the mutation of this locus to the other amino acid, the PtdCho transfer specific activity became PITPalpha (F225L) approximately PITPbeta and PITPbeta (L224F) approximately PITPalpha. The 225alpha/224beta locus plays a modest role in the specificity of both isoforms toward CerPCho.

Cyclic AMP promotes growth and secretion in human polycystic kidney epithelial cells.

BACKGROUND: Progressive cyst enlargement, the hallmark of autosomal-dominant polycystic kidney disease (ADPKD) and autosomal-recessive (ARPKD) polycystic kidney disease, precedes the eventual decline of function in these conditions. The expansion of individual cysts in ADPKD is determined to a major extent by mural epithelial cell proliferation and transepithelial fluid secretion. This study determined if common receptor-mediated agonists and an anonymous lipid stimulate the production of 3' 5'-cyclic monophosphate (cAMP) in mural epithelial cells from the two major types of human cystic diseases. METHODS: cAMP responses to maximally effective concentrations of renal agonists were determined together with measurements of transepithelial anion current and cellular proliferation and extracellular signal-related kinase (ERK 1/2) expression in primary cultures of epithelial cells from human ADPKD and ARPKD cysts. RESULTS: The rank orders of responses to ligands for ADPKD and ARPKD cells were identical: epinephrine > desmopressin (DDAVP) approximately arginine vasopressin (AVP) > adenosine > prostaglandin E(2) (PGE(2)) > parathyroid hormone (PTH). cAMP concentrations elevated by epinephrine, DDAVP, adenosine, and PGE(2) were diminished by receptor-specific inhibitors. Pools of cyst fluid collected individually from 16 of 19 ADPKD kidneys increased, to varying degrees, cAMP levels in ADPKD and ARPKD cells. PGE(2), beta-adrenergic and AVP antagonists partially inhibited cAMP accumulation in response to fluids from three kidneys, but a large portion of the endogenous activity was attributed to yet-to-be identified bioactive lipid, designated cyst activating factor (CAF). CAF stimulated cAMP production in ADPKD and ARPKD cells, activated ERK(1/2), and increased cellular proliferation in ADPKD cells. CAF increased positive short circuit current (I(SC)) in polarized ADPKD and T-84 monolayers, indicating stimulation of net anion secretion. CONCLUSION: Endogenous adenylyl cyclase agonists promote cell proliferation and electrolyte secretion of human ADPKD and ARPKD cells in vitro. We suggest that increased levels of cAMP may accelerate cyst growth and overall renal enlargement in patients with PKD.

Exposure of phosphatidylinositol transfer proteins to sphingomyelin-cholesterol membranes suggests transient but productive interactions with raft-like, liquid-ordered domains.

Both isoforms of rat phosphatidylinositol transfer protein (PITP) mediate the intermembrane transfer of sphingomyelin (CerPCho). In the plasma membrane, CerPCho often segregates with cholesterol into microdomains such as lipid rafts and caveolae. To test the hypothesis that PITP exhibits a preference for CerPCho- and cholesterol-rich membranes, we prepared unilamellar vesicles containing variable amounts of these two lipids. We also used CerPCho species with different acyl composition and treated vesicles with agents known to sequester and remove cholesterol. We observed that the beta isoform of rat PITP was more sensitive to membrane cholesterol than was the alpha isoform, as shown by increases in specific activities of lipid transfer of 2-6-fold. A relatively high membrane content of cholesterol (mole fraction > 0.4) was required to elicit such enhancements. Treatment of cholesterol-rich membranes with a series of beta cyclodextrins demonstrated that, upon depletion of cholesterol from participating membranes, the PITPbeta activity changes were fully reversible. We finally noted that the mechanism by which cholesterol enhances the activity of PITPbeta appeared to involve a decreased affinity of the protein for the membrane surface, in a manner that was independent of vesicle size and membrane microviscosity. We conclude that PITPbeta interacts transiently but productively with the liquid-ordered phase formed by CerPCho and cholesterol and discuss the possibility of PITP interactions in vivo with sphingolipid- and cholesterol-rich membrane microdomains.

Sterol carrier protein-2 functions in phosphatidylinositol transfer and signaling.

Over 20 years ago, it was reported that liver cytosol contains at least two distinct proteins that transfer phosphatidylinositol in vitro, phosphatidylinositol transfer protein (PITP) and a pH 5.1 supernatant fraction containing sterol carrier protein-2 (SCP-2). In contrast to PITP, there has been minimal progress on the structural and functional significance of SCP-2 in phosphatidylinositol transport. As shown herein, highly purified, recombinant SCP-2 stimulated up to 13-fold the rapid (s) transfer of radiolabeled phosphatidylinositol (PI) from microsomal donor membranes to highly curved acceptor membranes. SCP-2 bound to microsomes in vitro and overexpression of SCP-2 in transfected L-cells resulted in the following: (i) redistribution of phosphatidylinositols from intracellular membranes (mitochondria and microsomes) to the plasma membrane; (ii) enhancement of insulin-mediated inositol-triphosphate production; and (iii) 5.5-fold down regulation of PITP. Like PITP, SCP-2 binds two ligands required for vesicle budding from the Golgi, PI, and fatty acyl CoA. Double immunolabeling confocal microscopy showed SCP-2 significantly colocalized with caveolin-1 in the cytoplasm (punctate) and plasma membrane of SCP-2 overexpressing hepatoma cells (72%), HT-29 cells (58%), and SCP-2 overexpressing L-cells (37%). Taken together, these data show for the first time that SCP-2 plays a hitherto unrecognized role in intracellular phosphatidylinositol transfer, distribution, and signaling.

Both isoforms of mammalian phosphatidylinositol transfer protein are capable of binding and transporting sphingomyelin.

The structurally related mammalian alpha and beta isoforms of phosphatidylinositol (PtdIns) transfer protein (PITP) bind reversibly a single phospholipid molecule, preferably PtdIns or phosphatidylcholine (PtdCho), and transport that lipid between membrane surfaces. PITPbeta, but not PITPalpha, is reported extensively in the scientific literature to exhibit the additional capacity to bind and transport sphingomyelin (CerPCho). We undertook a detailed investigation of the lipid binding and transfer specificity of the soluble mammalian PITP isoforms. We employed a variety of donor and acceptor membrane lipid compositions to determine the sensitivity of recombinant rat PITPalpha and PITPbeta isoforms toward PtdIns, PtdCho, CerPCho, and phosphatidate (PtdOH). Results indicated often striking differences in protein-phospholipid and protein-membrane interactions. We demonstrated unequivocally that both isoforms were capable of binding and transferring CerPCho; we confirmed that the beta isoform was the more active. The order of transfer specific activity was similar for both isoforms: PtdIns>PtdCho>CerPCho>>PtdOH. Independently, we verified the binding of CerPCho to both isoforms by showing an increase in holoprotein isoelectric point following the exchange of protein-bound phosphatidylglycerol for membrane-associated CerPCho. We conclude that PITPalpha and PITPbeta are able to bind and transport glycero- and sphingophospholipids.