Dirk K.F. Meijer

Characterization of transport in isolated human hepatocytes: A study with the bile acid taurocholic acid, the uncharged ouabain and the organic cations vecuronium and rocuronium

The uptake and efflux of three categories of substrates were measured in isolated human hepatocytes and compared to those in rat hepatocytes. In addition, the extent to which the in vitro experiments quantitatively reflect liver function in vivo in both species was investigated. The anionic bile acid taurocholic acid was taken up by isolated human hepatocytes at a considerably lower rate than observed in isolated rat hepatocytes. Taurocholic acid uptake both in human hepatocytes and in liver plasma membrane vesicles showed sodium dependency. The uptake rate of taurocholic acid in isolated hepatocytes of both species was quantitatively compatible with the reported liver clearance of the bile acid in vivo. Ouabain uptake rate in isolated human hepatocytes was lower than in rat hepatocytes. This species difference was in accordance with pharmacokinetic studies in vivo on hepatic clearance of ouabain in man and rat. Uptake of vecuronium into human hepatocytes was about a factor of 10 lower than that in rat hepatocytes. Uptake into and efflux from human hepatocytes was comparable for the two short acting muscle relaxants vecuronium and rocuronium. Since distribution to the liver is considered to be a major factor in termination of action of vecuronium and rocuronium these observations were in line with the human pharmacokinetic profiles. In conclusion, the uptake rate of the studied model compounds in human hepatocytes appeared to be lower than that in rat hepatocytes. These observed transport rates reflected the relative hepatic transport rates observed in these species in the intact organism, but the absolute values in both species for some substrates may have been somewhat lower than calculated from in vivo data. It is concluded that transport studies in isolated hepatocytes are suitable for comparative drug transport studies, but are less precise in the prediction of quantitative membrane transport.

The preferential homing of a platelet derived growth factor receptor-recognizing macromolecule to fibroblast-like cells in fibrotic tissue

Platelet derived growth factor (PDGF) is a key factor in the induction and progression of fibrotic diseases with the activated fibroblast as its target cell. Drug targeting to the PDGF-receptor is explored as a new approach to treat this disease. Therefore, we constructed a macromolecule with affinity for the PDGF-beta receptor by modification of albumin with a small peptide that recognises this PDGF-beta receptor. The binding of the peptide-modified albumin (pPB-HSA) to the PDGF-beta receptor was confirmed in competition studies with PDGF-BB using NIH/3T3-fibroblasts and activated hepatic stellate cells. Furthermore, pPB-HSA was able to reduce PDGF-BB-induced fibroblast proliferation in vitro, and proved to be devoid of proliferation-inducing activity itself. We assessed the distribution of pPB-HSA in vivo in two models of fibrosis and related the distribution of pPB-HSA to PDGF-beta receptor density. In rats with liver fibrosis (bile duct ligation model), pPB-HSA quickly accumulated in the liver in contrast to unmodified HSA (P<0.001). The major part of pPB-HSA in the fibrotic liver was localized in hepatic stellate cells. In rats with renal fibrosis (anti-Thy1.1 model), pPB-HSA also homed to the cells that expressed the PDGF-beta receptor, i.e. the mesangial cells in the glomeruli of the kidney. These results indicate that pPB-HSA may be applied as a macromolecular drug-carrier that accumulates specifically in cells expressing the PDGF-beta receptor, thus allowing a selective delivery of anti-fibrotic agents to these cells.

Tumor vasculature targeted therapies - Getting the players organized

Based on their location and central role in solid tumor growth, tumor vascular endothelial cells may present an attractive target for the delivery of therapeutic drugs or cells. The potency of blocking the tumor blood supply in eradicating solid tumors was demonstrated recently in a mouse model of tumor vasculature targeting (Huang et al., Science 275: 547-550, 1997). For clinical application of such strategies, tumor endothelium specific target epitopes need to be identified. Recent studies on angiogenesis have identified angiogenesis-related molecules as potential target epitopes. Among these are vascular endothelial growth factor (VEGF)/VEGF-receptor complex, alpha(v) integrins, and Tie receptor tyrosine kinases. Besides blockade of their signalling cascades leading to inhibition of angiogenesis, these epitopes may also be instrumental in tumor vessel specific delivery of therapeutics. Data on the efficacy of therapeutic modalities aimed at these, mostly heterogeneously distributed tumor endothelial epitopes are scarce, and sophisticated experimentation is required to rationalize the development of new therapeutic strategies. Importantly, only detailed evaluations in cancer patients will provide the blueprint for the development of clinically effective tumor vascular targeted therapies.

A multispecific uptake system for taurocholate, cardiac glycosides and cationic drugs in the liver

To test the hypothesis of multiplicity in carrier-mediated uptake mechanisms for organic cations in the liver and to study the possible relation with bile acid and cardiac glycoside uptake mechanisms, mutual interaction during uptake of various radiolabeled quaternary amines has been studied in isolated rat hepatocytes. Inhibition patterns at low concentrations (1 microM) of the presumed type I monovalent organic cation tri-n-butylmethylammonium were markedly different from those at relatively high concentrations (25 microM). Both the cardiac glycoside K-strophantoside and the bile acid taurocholate clearly reduced the uptake rate of tri-n-butylmethylammonium at 25 microM whereas these agents completely failed to reduce the uptake at low concentrations of the cation. Subsequently, inhibition of uptake of some multivalent amphipathic organic cations (muscle relaxants) for the type II uptake system was investigated. It was found that the uptake of these muscle relaxants both at tracer concentrations (< 1 microM) and at relatively high concentrations (25 microM) was decreased in the presence of low concentrations of the cardiac glycoside K-strophantoside, while taurocholate only inhibited the uptake at the concentration range > 25 microM of the muscle relaxants. Procainamide ethobromide, a typical type I organic cation, did not affect the uptake either at the low or high concentration range of the muscle relaxants. It is concluded that for each of the type I-like compounds and type II-like compounds tested at least two systems are involved in uptake into hepatocytes: tri-n-butylmethylammonium in a concentration range < or = 1 microM is mainly taken up by the type I uptake system and at concentrations > or = 25 microM also by system(s) that can be inhibited by taurocholate and K-strophantoside. Bulky amphipathic organic (type II) cations at concentrations < 1 microM are also transported by an uptake system that is inhibitable by cardiac glycosides but not by bile salts. At concentrations > 25 microM these compounds are predominantly accommodated by an uptake system that possibly mediates uptake of both cardiac glycosides and bile acids. This concept was supported by the observation that both type II organic cations and bile salts can inhibit ouabain uptake, while type II organic cations as well as the cardiac glycosides reduce taurocholate uptake rate. The present data support the idea that the liver seems to be equipped with a multispecific uptake system that transports hydrophobic compounds irrespective of charge, including some type I and type II organic cations at relatively high substrate concentrations.

Differential effects of metabolic inhibitors on cellular and mitochondrial uptake of organic cations in rat liver

The effects of several metabolic inhibitors on the uptake of tri-n-butylmethylammonium (TBuMA) were studied in isolated rat liver mitochondria, isolated rat hepatocytes and isolated perfused rat livers, in order to characterize further the mechanisms for carrier-mediated uptake and cellular accumulation of organic cations in the liver. Treatment of isolated hepatocytes with valinomycin, carbonylcyanide-m-chlorophenyl-hydrazone (CCCP), dinitrophenol, oligomycin or antimycin resulted in a rapid decrease in cellular ATP within 3 min of addition. The initial uptake rate of TBuMA was generally largely affected by these treatments. However, fructose at 10 mM had no effect at all on the uptake rate of the cation whereas cellular ATP was decreased to an extent comparable to that after treatment with the metabolic inhibitors. Consequently it was hypothesized that the metabolic inhibitors affected the initial cellular uptake rate of organic cations due to either altered intracellular sequestration (e.g. mitochondria) or alternatively to direct effects on the plasma membrane rather than by decreasing cellular ATP. Isolated rat mitochondria were shown to take up organic cations very efficiently. Accumulation in this organelle is probably driven by the negative membrane potential as measured by the uptake of the lipophilic cation [3H]tetraphenylphosphonium. Treatment of the isolated mitochondria with various metabolic inhibitors decreased the membrane potential in parallel to the effects on the uptake of TBuMA. Since mitochondria constitute a considerable intracellular volume, they may contribute largely to the storage of the organic cation in the hepatocyte. In isolated perfused livers, preloaded with either TBuMA or tetraphenylphosphonium (TPP+), the addition of valinomycin or CCCP leads to a marked backflux of the cations from the liver into the perfusion medium. This suggests strongly that a large part of the intracellular storage capacity is lost after metabolic inhibitor treatment, probably as the consequence of dissipation of the mitochondrial membrane potential. Since the metabolic inhibitors in contrast to TBuMA uptake did not decrease the initial uptake rate of TPP+ into isolated hepatocytes, it was concluded that mitochondrial uptake (mitochondria are the major storage sites for TPP+) is not an essential determinant of the initial uptake rate in intact hepatocytes.(ABSTRACT TRUNCATED AT 400 WORDS)

Homing of negatively charged albumins to the lymphatic system: general implications for drug targeting to peripheral tissues and viral reservoirs.

The present study shows the lymphatic distribution of the negatively charged anti-HIV-1 agents succinylated or aconytilated human serum albumins (HSAs) in rats. Quantitation of blood and lymphatic concentrations of these proteins was performed through fluorescence detection of the fluorescein isothiocyanate (FITC)-labeled proteins. At several time points after i.v. injection, samples were taken from the cannulated thoracic duct and the carotid artery. Distribution of the negatively charged albumins (NCAs) to lymph was much more rapid than that of albumin itself and was dependent on the total net negative charge added to the protein: the half-life times of lymphatic equilibration were 15, 30, and 120 min for FITC-labeled aconytilated HSA, FITC-labeled succinylated HSA, and FITC-labeled HSA, respectively. Lymph to blood concentration ratios of the studied compounds obtained at steady state approached unity. In addition, the fluorescence in both body fluids was shown to represent unchanged labeled proteins. It was therefore inferred that the NCAs efficiently passed the endothelial barrier from blood to the interstitial compartment. Subsequently, we studied whether a specialized process was involved in the endothelial passage of the NCAs to the lymph. The following observations supported such a mechanism: a) preinjection of the scavenger receptor blockers polyinosinic- and formaldehyde-treated HSA reduced the transport from blood to the lymphatic compartment of FITC-labeled aconytilated HSA by more than 90%; b) the rate of lymphatic distribution was largely reduced when the body temperature of the rat was lowered to 28 degrees; and c) pre-administration of chloroquine resulted in a significant reduction in the lymphatic distribution of the NCAs. These data collectively indicate that a scavenger receptor-mediated process is involved in the transendothelial transport of NCAs. In situ localization in lymph nodes of the rat showed that FITC-labeled aconytilated and succinylated HSA are mainly present in the germinal center and parafollicular zones. The efficient distribution of these anionized proteins to the lymphatic system is of particular interest for HIV therapy, taking into account that replication of HIV mainly takes place in the lymphoid system. The observation that macromolecules, through charge modification, can extravasate through a receptor-mediated transcytotic process is potentially of major importance for the delivery of drugs with macromolecular carriers to cells not directly in contact with the blood.

Delivery of pharmacologically active dexamethasone into activated endothelial cells by dexamethasone-anti-E-selectin immunoconjugate

To deliver selectively anti-inflammatory agents into activated endothelial cells, drug-targeting conjugates were developed. Dexamethasone (Dexa) was covalently linked to a monoclonal antibody specifically recognizing E-selectin, which is strongly upregulated in endothelial cells at inflammatory sites. In the present study, the pharmacological effects of this Dexa-mouse antihuman E-selectin antibody (H18/7) (Ab(hEsel)) conjugate were investigated and compared to the effects obtained by free Dexa in human umbilical vein endothelial cells. Flow cytometry and ELISA were performed to analyze the levels of cell adhesion molecules (ICAM-1 and VCAM-1) and secreted cytokines (IL-6 and IL-8). The studies were extended by analysis of a complex gene expression pattern, using a cDNA expression array containing 268 genes encoding human cytokines/cytokine-receptors. Fifty genes and 28 genes were upregulated (ratio> or =2) upon incubation of human umbilical vein endothelial cells with TNFalpha for 6 and 24hr, respectively. This gene expression profile was markedly altered when cells were activated with TNFalpha in the presence of Dexa (100 nM) or Dexa-Ab(hEsel) conjugate (10 micro g/mL conjugate corresponding to 100 nM Dexa). Relative and competitive RT-PCR analysis verified downregulation of TNFalpha-mediated expression of CD40L and IL-8 by Dexa and Dexa-Ab(hEsel), respectively. These results indicated a successful internalization and processing of Dexa-Ab(hEsel) in activated endothelial cells, allowing the intracellularly delivered Dexa to exert its pleiotropic anti-inflammatory activity.

Inhibition of influenza virus fusion by polyanionic proteins

Anionic charge-modified human serum albumin (HSA) has previously been shown to exert potent in vitro activity against human immunodeficiency virus type 1 (HIV-1). In these studies, introduction of the additional negative charges was performed by derivatizing the epsilon-amino groups of lysine residues with succinic (Suc-HSA) or cis-aconitic anhydride (Aco-HSA), by which primary amino groups are replaced with carboxylic acids. The anti-HIV-1 activity was related to inhibition of gp41-mediated membrane fusion. Here, we investigated the activity of aconitylated and succinylated proteins on influenza virus membrane fusion, which is mediated by the viral membrane glycoprotein hemagglutinin (HA). Aco-HSA and Suc-HSA markedly inhibited the rates and extents of fusion of fluorescently labeled virosomes bearing influenza HA, with target membranes derived from erythrocytes. The inhibitory activity was dependent on the overall negative-charge density; HSA modified with 36 or less extra negative charges failed to inhibit fusion. The inhibition of fusion showed a certain degree of specificity for the protein carrying the negative charges: polyanionic HSA and beta-lactoglobulin A derivatives had fusion-inhibitory activity, whereas succinylated BSA, lactalbumin, lactoferrin, lysozyme, and transferrin were inactive. Aco60-HSA and Aco-beta-lactoglobulin A inhibited influenza virus membrane fusion in a concentration-dependent manner, IC50 values being about 4 and 10 microg/mL, respectively. HA-mediated membrane fusion is pH dependent. Aco60-HSA did not induce a shift in the pH threshold or in the pH optimum. Fusion with liposomes of another low pH-dependent virus, Semliki Forest virus, was not specifically affected by any of the compounds reported here. In view of some structural and functional similarities between influenza HA and the HIV-1 gp120/gp41 complex, it is tempting to postulate that the current results might have some implications for the anti-HIV-1 mechanism of polyanionic proteins.

Mechanism of anti-HIV activity of succinylated human serum albumin

In the present study, we described the interaction of succinylated human serum albumin (Suc-HSA), a negatively charged anti-HIV-1 active protein, with HIV-1 gp120 and in detail with the third variable domain of gp120 (V3 loop). To this end, different assay formats were tested in which gp120- and V3-related peptides were presented in various configurations in order to investigate the effect of the conformational structure of the V3 loop on the interaction with negatively charged albumins. When gp120 presented via a lectin was used, it was observed that Suc-HSA bound to native gp120. The binding site appeared to be located at or near the thrombin digestion site (GPGRAF sequence) in the V3 loop of gp120, since the cleavage of the loop resulted in decreased binding of Suc-HSA. In addition, Suc-HSA was able to protect the V3 region of gp120 from cleaving with thrombin. In contrast, significant binding of Suc-HSA to V3 loop or gp120 peptides was not observed when both were presented in a fluid phase system, suggesting the involvement of a monovalent-low affinity binding of Suc-HSA. Using overlapping peptides delineating the whole V3 loop immobilized to CNBr-Sepharose, we noticed that the interaction of the V3 loop with Suc-HSA was predominantly induced by electrostatic interactions between positively charged linearized peptide fragments and Suc-HSA and was positively influenced by the presence of hydrophobic amino in the V3 loop fragments as well. Moreover, the highest affinity site was located at sites near the GPGRAF sequence. These observations add to the evidence, collected earlier, that Suc-HSA interferes at the level of virus entry, independent of interaction with the CD4 receptor. Since the recently discovered chemokine receptors are negatively charged, we can hypothesize that Suc-HSA is able to prevent the positively charged V3 loop from interacting with these types of receptors, thereby inhibiting virus entry.

Investigations on the hepatic uptake systems for organic cations with a photoaffinity probe of procainamide ethobromide

Azido procainamide methoiodide (APM), a photolabile derivative of the transport model compound procainamide ethobromide (PAEB), shows a close resemblance to PAEB from a physicochemical point of view. Like PAEB it is effectively taken up by the liver and excreted into bile. Kinetics of the uptake of APM in isolated hepatocytes revealed that in addition to a non-saturable process, two saturable uptake systems are involved (Km1 = 3 microM, Vmax1) = 80 pmol/min/10(6) cells, Km2 = 100 microM, Vmax2 = 130 pmol/min x 10(6) cells). The uptake rate of APM was inhibited markedly in the presence of other organic cations. Organic anions and uncharged compounds generally had no inhibitory effect on the APM uptake. These results support the theory that there is a separate hepatic uptake system for organic cations like APM. Photoaffinity labeling of intact hepatocytes as well as plasma membrane sub-fractions enriched with sinusoidal domains disclosed two major binding polypeptides with apparent M(r) of 48,000 and 72,000. Such labeling patterns were not observed in membranes from hepatoma cells that are deficient in organic solute uptake. Differential photoaffinity labeling with other cationic compounds such as tributylmethyl ammonium and d-tubocurarine reduced the incorporation of APM in these polypeptides. The 48- and 72-kDa proteins might be involved in carrier-mediated transport of type I organic cations at the hepatic uptake level.