NH2-terminal signals in ATP7B Cu-ATPase mediate its Cu-dependent anterograde traffic in polarized hepatic cells.

Cu is an essential cofactor of cellular proteins but is toxic in its free state. The hepatic Cu-ATPase ATP7B has two functions in Cu homeostasis: it loads Cu+ onto newly synthesized apoceruloplasmin in the secretory pathway, thereby activating the plasma protein; and it participates in the excretion of excess Cu+ into the bile. To carry out these two functions, the membrane protein responds to changes in intracellular Cu levels by cycling between the Golgi and apical region. We used polarized hepatic WIF-B cells and high-resolution confocal microscopy to map the itinerary of endogenous and exogenous ATP7B under different Cu conditions. In Cu-depleted cells, ATP7B resided in a post-trans-Golgi network compartment that also contained syntaxin 6, whereas in Cu-loaded cells, the protein relocated to unique vesicles very near to the apical plasma membrane as well as the membrane itself. To determine the role of ATP7B's cytoplasmic NH2 terminus in regulating its intracellular movements, we generated seven mutations/deletions in this large [approximately 650 amino acid (AA)] domain and analyzed the Cu-dependent behavior of the mutant ATP7B proteins in WIF-B cells. Truncation of the ATP7B NH2 terminus up to the fifth copper-binding domain (CBD5) yielded an active ATPase that was insensitive to cellular Cu levels and constitutively trafficked to the opposite (basolateral) plasma membrane domain. Fusion of the NH2-terminal 63 AA of ATP7B to the truncated protein restored both its Cu responsiveness and correct intracellular targeting. These results indicate that important targeting information is contained in this relatively short sequence, which is absent from the related CuATPase, ATP7A.

Absence of direct delivery for single transmembrane apical proteins or their "Secretory" forms in polarized hepatic cells.

The absence of a direct route to the apical plasma membrane (PM) for single transmembrane domain (TMD) proteins in polarized hepatic cells has been inferred but never directly demonstrated. The genes encoding three pairs of apical PM proteins, whose extracellular domains are targeted exclusively to the apical milieu in Madin-Darby canine kidney cells, were packaged into recombinant adenovirus and delivered to WIF-B cells in vitro and liver hepatocytes in vivo. By immunofluorescence and pulse-chase metabolic labeling, we found that the soluble constructs were overwhelmingly secreted into the basolateral milieu, which in vivo is the blood and in vitro is the culture medium. The full-length proteins were first delivered to the basolateral surface but then concentrated in the apical PM. Our results imply that hepatic cells lack trans-Golgi network (TGN)-based machinery for directly sorting single transmembrane domain apical proteins and raise interesting questions about current models of PM protein sorting in polarized and nonpolarized cells.

Disruption of a yeast very-long-chain acyl-CoA synthetase gene simulates the cellular phenotype of X-linked adrenoleukodystrophy.

X-linked adrenoleukodystrophy (X-ALD) is characterized biochemically by elevated levels of saturated very long-chain fatty acids (VLCFAs) in plasma and tissues. In X-ALD, peroxisomal very-long-chain acyl-CoA synthetase (VLCS) fails to activate VLCFAs, preventing their degradation via beta-oxidation. However, the product of the defective XALD gene (ALDP) is not a VLCS, but rather a peroxisomal membrane protein (PMP). Disruption of either or both of two yeast PMP genes related to the XALD gene did not produce a biochemical phenotype resembling that found in X-ALD fibroblasts. The authors identified a candidate yeast VLCS gene (the FAT1 locus) by its homology to rat liver VLCS. Disruption of this gene decreased VLCS activity, but had no effect on long-chain acyl-CoA synthetase activity. In FAT1-disruption strains, VLCS activity was reduced to 30-40% of wild-type in both a microsome-rich 27,000 g supernatant fraction and a peroxisome- and mitochondria-rich pellet fraction of yeast spheroplast homogenates. Separation of the latter organelles by density gradient centrifugation revealed that VLCS activity was peroxisomal and not mitochondrial. VLCS gene-disruption strains had increased cellular VLCFA levels, compared to wild-type yeast. The extent of both the decrease in peroxisomal VLCS activity and the VLCFA accumulation in this yeast model resembles that observed in cells from X-ALD patients. Characterization of the gene(s) responsible for the residual peroxisomal VLCS activity may suggest new therapeutic approaches in X-ALD.

Role of very-long-chain acyl-coenzyme A synthetase in X-linked adrenoleukodystrophy.

X-linked adrenoleukodystrophy (X-ALD) is characterized biochemically by decreased ability of cells to activate (via very-long-chain acyl-coenzyme A synthetase [VLCS]) and subsequently degrade very-long-chain fatty acids in peroxisomes. It is noteworthy that the gene defective in X-ALD encodes ALDP, a peroxisomal membrane protein unrelated to VLCS. We cloned human VLCS (hVLCS) and found that peroxisomes from X-ALD fibroblasts contained immunoreactive hVLCS, refuting the earlier hypothesis that ALDP is required to anchor VLCS to the peroxisomal membrane. Furthermore, hVLCS was topographically oriented facing the peroxisomal matrix in both control and X-ALD fibroblasts, contradicting the alternative hypothesis that ALDP is required to translocate VLCS into peroxisomes. However, overexpression of both hVLCS and ALDP in X-ALD fibroblasts synergistically increased very-long-chain fatty acid beta-oxidation, indicating that these proteins interact functionally.

X-linked adrenoleukodystrophy: genes, mutations, and phenotypes.

X-linked adrenoleukodystrophy (X-ALD) is a complex and perplexing neurodegenerative disorder. The metabolic abnormality, elevated levels of very long-chain fatty acids in tissues and plasma, and the biochemical defect, reduced peroxisomal very long-chain acyl-CoA synthetase (VLCS) activity, are ubiquitous features of the disease. However, clinical manifestations are highly variable with regard to time of onset, site of initial pathology and rate of progression. In addition, the abnormal gene in X-ALD is not the gene for VLCS. Rather, it encodes a peroxisomal membrane protein with homology to the ATP-binding cassette (ABC) transmembrane transporter superfamily of proteins. The X-ALD protein (ALDP) is closely related to three other peroxisomal membrane ABC proteins. In this report we summarize all known X-ALD mutations and establish the lack of an X-ALD genotype/phenotype correlation. We compare the evolutionary relationships among peroxisomal ABC proteins, demonstrate that ALDP forms homodimers with itself and heterodimers with other peroxisomal ABC proteins and present cDNA complementation studies suggesting that the peroxisomal ABC proteins have overlapping functions. We also establish that there are at least two peroxisomal VLCS activities, one that is ALDP dependent and one that is ALDP independent. Finally, we discuss variable expression of the peroxisomal ABC proteins and ALDP independent VLCS in relation to the variable clinical presentations of X-ALD.

Peroxisomal very long chain fatty acid beta-oxidation activity is determined by the level of adrenodeukodystrophy protein (ALDP) expression.

Impaired peroxisomal beta-oxidation of saturated very long chain fatty acids (VLCFA, >/=C22:0) results in increased VLCFA levels in the tissues and body fluids of patients with disorders of peroxisomal biogenesis (i.e., Zellweger syndrome and neonatal adrenoleukodystrophy) and single peroxisomal protein defects (i.e., X-linked adrenoleukodystrophy (X-ALD) and acyl-CoA oxidase deficiency). We show that SV40T transformation also results in impaired peroxisomal beta-oxidation and VLCFA accumulation despite the presence of abundant peroxisomes. To explore the mechanism responsible for this observation, we have examined expression of key components of peroxisomal VLCFA beta-oxidation. We found that expression of both acyl-CoA oxidase, the rate limiting enzyme of peroxisomal VLCFA beta-oxidation and the adrenoleukodystrophy protein (ALDP), the defective gene product in X-ALD, are reduced after SV40T transformation. Surprisingly, ALDP overexpression by itself restores peroxisomal VLCFA beta-oxidation in SV40T-transformed control and X-ALD cells. These results demonstrate that ALDP is a fundamental component in VLCFA peroxisomal beta-oxidation and may serve as a "gatekeeper" for VLCFA homeostasis.

Gene redundancy and pharmacological gene therapy: implications for X-linked adrenoleukodystrophy.

As more functional redundancy in mammalian cells is discovered, enhanced expression of genes involved in alternative pathways may become an effective form of gene therapy. X-linked adrenoleukodystrophy (X-ALD) is a peroxisomal disorder with impaired very-long-chain fatty acid metabolism. The X-ALD gene encodes a peroxisomal membrane protein (ALDP) that is part of a small family of related peroxisomal membrane proteins. We show that 4-phenylbutyrate treatment of cells from both X-ALD patients and X-ALD knockout mice results in decreased levels of and increased beta-oxidation of very-long-chain fatty acids; increased expression of the peroxisomal protein ALDRP; and induction of peroxisome proliferation. We also demonstrate that ALDP and ALDRP are functionally related, by ALDRP cDNA complementation of X-ALD fibroblasts. Finally, we demonstrate the in vivo efficacy of dietary 4-phenylbutyrate treatment through its production of a substantial reduction of very-long-chain fatty acid levels in the brain and adrenal glands of X-ALD mice.

Disruption of the Saccharomyces cerevisiae FAT1 gene decreases very long-chain fatty acyl-CoA synthetase activity and elevates intracellular very long-chain fatty acid concentrations.

Activation of fatty acids to their coenzyme A derivatives is necessary for subsequent metabolism. Very long-chain fatty acids, which accumulate in tissues of patients with X-linked adrenoleukodystrophy, are activated by very long-chain acyl-CoA synthetase (VLCS) normally found in peroxisomes and microsomes. We identified a candidate yeast VLCS gene (FAT1), previously identified as encoding a fatty acid transport protein, by its homology to rat liver peroxisomal VLCS. Disruption of this gene decreased, but did not abolish, cellular VLCS activity. Fractionation studies showed that VLCS activity, but not long-chain acyl-CoA synthetase activity, was reduced to about 40% of wild-type level in both 27,000 x g supernatant and pellet fractions. Separation of organelles in the pellet fraction by density gradient centrifugation revealed that VLCS activity was associated with peroxisomes and microsomes but not mitochondria. FAT1 deletion strains exhibited decreased growth on medium containing dextrose, oleic acid, and cerulenin, an inhibitor of fatty acid synthesis. FAT1 deletion strains grown on either dextrose or oleic acid medium accumulated very long-chain fatty acids. Compared with wild-type yeast, C22:0, C24:0, and C26:0 levels were increased approximately 20-, 18-, and 3-fold in deletion strains grown on dextrose, and 2-, 7-, and 5-fold in deletion strains grown on oleate. Long-chain fatty acid levels in wild-type and deletion strains were not significantly different. All biochemical defects in FAT1 deletion strains were restored to normal after functional complementation with the FAT1 gene. The level of VLCS activity measured in both wild-type and deletion yeast strains transformed with FAT1 cDNA paralleled the level of expression of the transgene. The extent of both the decrease in peroxisomal VLCS activity and the very long-chain fatty acid accumulation in the yeast FAT1 deletion model resembles that observed in cells from X-linked adrenoleukodystrophy patients. These studies suggest that the FAT1 gene product has VLCS activity that is essential for normal cellular very long-chain fatty acid homeostasis.

Suppression of peroxisomal membrane protein defects by peroxisomal ATP binding cassette (ABC) proteins.

X-Linked adrenoleukodystrophy (X-ALD) is a neurodegenerative disorder characterized by reduced peroxisomal very long chain fatty acid (VLCFA) beta-oxidation. The X - ALD gene product (ALDP) is a peroxisomal transmembrane protein with an ATP binding cassette (ABC). ALDP and three other ABC proteins (PMP70, ALDR, P70R) localize to the peroxisomal membrane. The function of this family of peroxisomal membrane proteins is unknown. We used complementation studies to begin analysis of their role in VLCFA beta-oxidation and on the peroxisomal membrane. Expression of either ALDP or PMP70 restores VLCFA beta-oxidation in X-ALD fibroblasts, indicating overlapping functions. Their expression also restores peroxisome biogenesis in cells that are deficient in the peroxisomal membrane protein Pex2p. Thus it is likely that complex protein interactions are involved in the function and biogenesis of peroxisomal membranes that may contribute to disease heterogeneity.

Altered expression of ALDP in X-linked adrenoleukodystrophy.

X-linked adrenoleukodystrophy (ALD) is a neurodegenerative disorder with variable phenotypic expression that is characterized by elevated plasma and tissue levels of very long-chain fatty acids. However, the product of the gene defective in ALD (ALDP) is a membrane transporter of the ATP-binding cassette family of proteins and is not related to enzymes known to activate or oxidize fatty acids. We generated an antibody that specifically recognizes the C-terminal 18 amino acids of ALDP and can detect ALDP by indirect immunofluorescence. To better understand the mechanism by which mutations in ALDP lead to disease, we used this antibody to examine the subcellular distribution and relative abundance of ALDP in skin fibroblasts from normal individuals and ALD patients. Punctate immunoreactive material typical of fibroblast peroxisomes was observed in cells from seven normal controls and eight non-ALD patients. Of 35 ALD patients tested, 17 had the childhood-onset cerebral form of the disease, 13 had the milder adult phenotype adrenomyeloneuropathy, 3 had adrenal insufficiency only, and 2 were affected fetuses. More than two-thirds (69%) of all patients studied showed no punctate immunoreactive material. There was no correlation between the immunofluorescence pattern and clinical phenotype. We determined the mutation in the ALD gene in 15 of these patients. Patients with either a deletion or frameshift mutation lacked ALDP immunoreactivity, as expected. Four of 11 patients with missense mutations were also immunonegative, indicating that these mutations affected the stability or localization of ALDP. In the seven immunopositive patients with missense mutations, correlation of the location and nature of the amino acid substitution may provide new insights into the function of this peroxisomal membrane protein. Furthermore, the study of female relatives of immunonegative ALD probands may aid in the assessment of heterozygote status.

In vitro integration of retrotransposon Ty1: a direct physical assay.

Retrotransposon Ty1 of Saccharomyces cerevisiae inserts a double-stranded Ty1 cDNA into the yeast genome by a reaction analogous to the integration mechanism used by retroviruses. A quantitative in vitro integration assay that directly detects integrative recombination products was developed for Ty1. Blunt-ended artificial radioactive substrates bearing Ty1 termini integrate into circular or linear target DNAs. The reaction is specific for native integrase isolated in the form of virus-like particles; virus-like particles prepared from integrase mutants were completely inactive in this assay. The products are radioactive, allowing direct detection after gel electrophoresis by autoradiography. Using this simple and amenable system, we characterized the biochemical requirements of the system and the structures of the major integration products. Two classes of products were detected: those that were the result of bona fide complete integration events (concerted reactions) and single-end joinings of substrate to target (half-reactions). Additionally, we used a genetic selection scheme to identify and characterize target sites of complete integration events into a circular target plasmid; a 5-bp target site duplication flanking the inserted DNA resembling the duplication characteristic of in vivo integration was observed.

Ty1 in vitro integration: effects of mutations in cis and in trans.

Mutations within the TYB gene of Ty1 encoding integrase (IN) as well as alterations in its substrate, a linear DNA molecule, were examined for their effects on in vitro IN activity, using a recently developed physical assay. Five different codon-insertion mutations, two frameshift mutations, and one missense mutation, previously identified as transposition-deficient mutations, were tested. Virus-like particles, the source of IN, from two different protease mutants and a reverse transcriptase mutant exhibited near-normal to normal IN activity. Two frameshift mutations mapping within the phylogenetically variable C-terminal domain of IN resulted in significant in vitro IN activity. In contrast, three mutations within the amino-terminal conserved domain of IN completely abolished IN activity. When the substrate termini were mutated, we found that substrates with as few as 4 bp of Ty1 termini were capable of efficiently generating integration products. Surprisingly, certain substrates that lacked obvious similarity to Ty1 termini were also readily integrated into both linear and circular targets, whereas others were not used as substrates at all. Termini rich in adenosine residues were among the more active substrates; however, certain substrates lacking terminal adenosine residues can form small quantities of integration products, including complete integration reactions.

In-frame linker insertion mutagenesis of yeast transposon Ty1: phenotypic analysis.

A plasmid bearing a transpositionally functional GAL1::Ty1 fusion was mutagenized by insertion of four or five codons semirandomly throughout the plasmid. This collection of mutant plasmids was introduced into yeast cells and studied with regard to the properties of the mutant Ty1-encoded proteins and the transposition phenotypes observed. All of the transposition-inactivating mutations were previously found to be recessive with the exception of a single mutation in TYA. In this mutant, TYA protein of normal abundance is produced, but the virus-like particles containing this protein are unstable and have aberrant behavior. The effects of mutations in noncoding regions, as well as the capsid protein coding region TYA, and the regions encoding the protease, integrase and reverse transcriptase proteins are described. Effects on gene expression, types of proteins produced, proteolysis of precursor proteins, virus-like particle structure, and biochemical activities of the encoded proteins are summarized. In addition, we show that one of the mutations in the 3' LTR represents a new nonessential site into which foreign marker DNA can be inserted without compromising transposition.

In-frame linker insertion mutagenesis of yeast transposon Ty1: mutations, transposition and dominance.

A plasmid bearing a GAL1::Ty1 fusion that is competent to transpose was mutagenized by insertion of oligodeoxyribonucleotides that precisely introduce four or five codons semirandomly throughout the plasmid. Approximately one quarter of these resulted in inactivation of transposition; these include inactivating insertions in both the TYA and TYB genes, corresponding to retroviral gag and pol genes. Examples of transposition-inactivating mutations map within each of the known or proposed functional domains of TYB, suggesting that these are all required for retrotransposition. All of the transposition-inactivating mutations were found to be recessive with the exception of a single mutation in TYA. The remaining mutations have slightly deleterious to no effect on Ty1 transposition.

Hotspots for unselected Ty1 transposition events on yeast chromosome III are near tRNA genes and LTR sequences.

A collection of yeast strains bearing single marked Ty1 insertions on chromosome III was generated. Over 100 such insertions were physically mapped by pulsed-field gel electrophoresis. These insertions are very nonrandomly distributed. Thirty-two such insertions were cloned by the inverted PCR technique, and the flanking DNA sequences were determined. The sequenced insertions all fell within a few very limited regions of chromosome III. Most of these regions contained tRNA coding regions and/or LTRs of preexisting transposable elements. Open reading frames were disrupted at a far lower frequency than expected for random transposition. The results suggest that the Ty1 integration machinery can detect regions of the genome that may represent "safe havens" for insertion. These regions of the genome do not contain any special DNA sequences, nor do they behave as particularly good targets for Ty1 integration in vitro, suggesting that the targeted regions have special properties allowing specific recognition in vivo.

Sequence, tissue distribution, and differential expression of mRNA for a putative insulin-responsive glucose transporter in mouse 3T3-L1 adipocytes.

The cDNAs for two putative glucose transporters from mouse 3T3-L1 adipocytes were isolated and sequenced. One of these cDNAs encodes the murine homolog of the human hepG2/erythrocyte glucose transporter, termed GT1. GT1 mRNA is most abundant in mouse brain and is expressed in both 3T3-L1 preadipocytes and adipocytes. The other cDNA encodes a glucose transporter-like protein, termed GT2, that has a unique amino acid sequence and tissue distribution. GT2 cDNA encodes a protein with 63% amino acid sequence identity and a similar structural organization to GT1. GT2 mRNA is found at high levels in mouse skeletal muscle, heart, and adipose tissue, all of which exhibit insulin-stimulated glucose uptake. GT2 mRNA is absent from 3T3-L1 preadipocytes but is induced dramatically during differentiation into adipocytes. This increase in mRNA content correlates closely with the acquisition of insulin-stimulated glucose uptake. We propose that GT2 is an insulin-regulated glucose transporter.

Cellular location of the cleavage event of the polymeric immunoglobulin receptor and fate of its anchoring domain in the rat hepatocyte.

Transcytosis of polymeric immunoglobulin (pIg) across glandular and mucosal epithelia is mediated by a member of the immunoglobulin supergene family, the pIg receptor. During transcellular routing, the receptor is cleaved and its ectoplasmic domain, known as secretory component (SC), is released into secretions bound to pIg. Using receptor-domain-specific antibodies, we have combined cell fractionation and immunoblotting of rat liver to examine the cellular routing of the receptor, the cellular location of the cleavage event and the fate of the anchor domain. Cleavage is a late event in receptor processing. It appears to occur at the canalicular plasma membrane, since intact receptor is present in this membrane domain and no SC is detected in whole liver homogenate or in cell fractions. The membrane anchor remaining after cleavage can be recovered in bile, as well as in a low-density fraction obtained after equilibrium centrifugation of liver (microsomal fractions) on sucrose density gradients. These data suggest that the membrane-anchor domain may be internalized as well as secreted together with SC into bile.

The major subunit of the rat asialoglycoprotein receptor can function alone as a receptor.

Mammalian hepatic asialoglycoprotein receptors (ASGP-R) are composed of two unique, but closely related polypeptides, which in the rat are designated rat hepatic lectins 1 and 2/3 (RHL 1, RHL 2/3). Despite numerous studies, the composition of a functional ASGP-R has remained unclear. We examined this question in rat hepatoma tissue culture (HTC) cells (which lack endogenous ASGP-R) that were co-transfected with cDNAs for both RHL 1 and RHL 2/3. The original population was cloned, but derivatives were unstable. We therefore used fluorescence-activated cell sorting to separate a subpopulation of cells (positive) that specifically endocytosed fluoresceinated asialoorosomucoid (ASOR) from one that did not (negative). We then used indirect immunofluorescence with polypeptide-specific ASGP-R antibodies, immunoanalysis, and binding and uptake studies with two Gal ligands (ASOR and NAc-galactosylated poly-L-lysine (Gal-Lys] to further define the ASGP-R status in these two populations. As reported by others, we found that expression of both RHL 1 and RHL 2/3 in the positive cells resulted in binding, uptake and degradation of ASOR, the most commonly used ASGP-R ligand. The negative cells expressed only RHL 1 and neither bound nor processed ASOR. However, the presence of RHL 1 was sufficient for specific high affinity binding and processing of the synthetic ligand, Gal-Lys, by negative cells. These results show that RHL 1 can function as an ASGP-R, given a highly galactosylated ligand, and that RHL 2/3 must play an important role in the organization of native ASGP-R in the membrane.

Biochemical characterization of domain-specific glycoproteins of the rat hepatocyte plasma membrane.

Seven integral proteins (CE 9, HA 21, HA 116, HA 16, HA 4, HA 201, and HA 301) were isolated from rat hepatocyte plasma membranes by immunoaffinity chromatography on monoclonal antibody-Sepharose. Six of the proteins (all but HA 16) exhibit domain-specific localizations (either bile canalicular or sinusoidal/lateral) about the hepatocyte surface. We identified three of these protein antigens as leucine aminopeptidase (HA 201), dipeptidyl peptidase IV (HA 301), and the asialoglycoprotein receptor (HA 116). We also developed 125I-lectin blotting procedures that, when used in conjunction with chemical and glycosidase treatments, permitted a comparison of the types of oligosaccharides present on the seven proteins. All seven are sialoglycoproteins, based upon the effects of prior neuraminidase and periodate-aniline-cyanoborohydride treatments of blots on labeling by 125I-wheat germ agglutinin. 125I-labeled Ricinus communis agglutinin I and 125I-peanut agglutinin blotting of the desialylated proteins revealed few if any conventional O-linked oligosaccharides, suggesting that the sialyl residues represent termini of N-linked complex-type oligosaccharides. Depending upon the protein, we estimated the presence of 2-26 N-linked oligosaccharides/polypeptide chain from the Mr reductions accompanying chemical or enzymatic deglycosylation. Three of these mature plasma membrane proteins (HA 21, HA 116, and HA 4) have both high mannose-type and complex-type oligosaccharides on every copy of their polypeptide chains. The labeling of these three proteins by 125I-concanavalin A was sensitive to treatment with endoglycosidase H, and each exhibited a quantitative reduction in Mr after the treatment, as assessed independently by 125I-wheat germ agglutinin blotting. At this level of analysis, we were unable to discern differences in the types of oligosaccharides present on these seven glycoproteins that correlate with their patterns of expression within the plasma membrane domains of this polarized epithelial cell.

Identification of rat hepatocyte plasma membrane proteins using monoclonal antibodies.

We have localized and identified five rat hepatocyte plasma membrane proteins using hybridoma technology in combination with morphological and biochemical methods. Three different membrane preparations were used as immunogens: isolated hepatocytes, a preparation of plasma membrane sheets that contained all three recognizable surface domains of the intact hepatocyte (sinusoidal, lateral, and bile canalicular), and a glycoprotein subfraction of that plasma membrane preparation. We selected monoclonal IgGs that were hepatocyte specific and localized them using both immunofluorescence on 0.5-micron sections of frozen liver and immunoperoxidase at the ultrastructural level. One antigen (HA 4) was localized predominantly to the bile canalicular surface, whereas three (CE 9, HA 21, and HA 116) were localized predominantly to the lateral and sinusoidal surfaces. One antigen (HA 16) was present in all three domains. Only one antigen (HA 116) could be detected in intracellular structures both in the periphery of the cell and in the Golgi region. The antigens were all integral membrane proteins as judged by their stability to alkaline extraction and solubility in detergents. The apparent molecular weights of the antigens were established by immunoprecipitation and/or immunoblotting. In a related study (Bartles, J.R., L.T. Braiterman, and A.L. Hubbard, 1985, J. Cell. Biol., 100:1126-1138), we present biochemical confirmation of the domain-specific localizations for two of the antigens, HA 4 and CE 9, and demonstrate their suitability as endogenous domain markers for monitoring the separation of bile canalicular and sinusoidal lateral membrane on sucrose density gradients.