Unconjugated bilirubin (UCB) causes encephalopathy in severely jaundiced neonates by damaging astrocytes and neurons. Astrocytes, which help defend the brain against cytotoxic insults, express the ATP-dependent transporter, multidrug resistance-associated protein 1 (Mrp1), which mediates export of organic anions, probably including UCB. We therefore studied whether exposure to UCB affects the expression and intracellular localization of Mrp1 in cultured mouse astroglial cells (>95% astrocytes). Mrp1 was localized and quantitated by confocal laser scanning microscopy and double immunofluorescence labeling by using specific antibodies against Mrp1 and the astrocyte marker glial fibrillary acidic protein, plus the Golgi marker wheat germ agglutinin (WGA). In unexposed astrocytes, Mrp1 colocalized with WGA in the Golgi apparatus. Exposure to UCB at a low unbound concentration (B_f)of 40 nM caused rapid redistribution of Mrp1 from the Golgi throughout the cytoplasm to the plasma membrane, with a peak 5-fold increase in Mrp1 immunofluorescence intensity from 30 to 120 min.B_fabove aqueous saturation produced a similar but aborted response. Exposure to this higherB_ffor 16 h markedly decreased Trypan blue exclusion and methylthiazoletetrazoilum activity and increased apoptosis 5-fold by terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling assay. These toxic effects were modestly increased by inhibition of Mrp1 activity with 3-([3-(2-[7-chloro-2-quinolinyl]ethenyl)phenyl-(3-dimethylamino-3-oxopropyl)-thio-methyl]thio)propanoic acid (MK571). By contrast,B_f= 40 nM caused injury only if Mrp1 activity was inhibited by MK571, which also blocked translocation of Mrp1. Our conclusion is that in astrocytes, UCB up-regulates expression of Mrp1 and promotes its trafficking from the Golgi to the plasma membrane, thus moderating cytotoxicity from UCB, presumably by limiting its intracellular accumulation.

Revised concepts of bilirubin encephalopathy have been revealed by studies of bilirubin toxicity in cultured CNS cells and in congenitally jaundiced Gunn rats. Bilirubin neurotoxicity is related to the unbound (free) fraction of unconjugated bilirubin (B_f), of which the dominant species at physiological pH is the protonated diacid, which can passively diffuse across cell membranes. As the binding affinity of plasma albumin for bilirubin decreases strikingly as albumin concentration increases, previously reported B_fvalues were underestimated.Newer diagnostic tests can detect reversible neurotoxicity before permanent damage occurs from precipitation of bilirubin (kernicterus). Early toxicity can occur at B_fonly modestly above aqueous saturation and affects astrocytes and neurons, causing mitochondrial damage, resulting in impaired energy metabolism and apoptosis, plus cell‐membrane perturbation, which causes enzyme leakage and hampers transport of neurotransmitters.The concentrations of unbound bilirubin in the cerebro‐spinal fluid and CNS cells are probably limited mainly by active export of bilirubin back into plasma, mediated by ABC transporters present in the brain capillary endothelium and choroid plexus epithelium. Intracellular bilirubin levels may be diminished also by oxidation, conjugation and binding to cytosolic proteins.These new concepts may explain the varied susceptibility of neonates to develop encephalopathy at any given plasma bilirubin level and the selective distribution of CNS lesions in bilirubin encephalopathy. They also can suggest better strategies for predicting, preventing and treating this syndrome.

Albumin binding is a crucial determinant of bilirubin clearance in health and bilirubin toxicity in certain disease states. However, prior attempts to measure the affinity of albumin for bilirubin have yielded highly variable results, reflecting both differing conditions and the confounding influence of impurities. We therefore have devised a method based on serial ultrafiltration that successively removes impurities in [(14)C]bilirubin until a stable binding affinity is achieved, and then we used it to assess the effect of albumin concentration and buffer composition on binding. The apparent binding affinity of human serum albumin for [(14)C]bilirubin was strongly dependent on assay conditions, falling from (5.09 +/- 0.24) x 10(7) liters/mol at lower albumin concentrations (15 microm) to (0.54 +/- 0.05) x 10(7) liters/mol at higher albumin concentrations (300 microm). To determine whether radioactive impurities were responsible for this change, we estimated impurities in the stock bilirubin using a novel modeling approach and found them to be 0.11-0.13%. Formation of new impurities during the study and their affinity for albumin were also estimated. After correction for impurities, the binding affinity remained heavily dependent on the albumin concentration (range (5.37 +/- 0.26) x 10(7) liters/mol to (0.65 +/- 0.03) x 10(7) liters/mol). Affinities decreased by about half in the presence of chloride (50 mm). Thus, the affinity of human albumin for bilirubin is not constant, but varies with both albumin concentration and buffer composition. Binding may be considerably less avid at physiological albumin concentrations than previously believed.