Intralobular heterogeneity of perisinusoidal stellate cells in porcine liver

The aim of the present investigation was to elucidate the intralobular heterogeneity of the perisinusoidal stellate cells (fat-storing cells, lipocytes) in the porcine liver. Their three-dimensional structure, desmin immunoreactivity and vitamin-A storage were studied by use of the Golgi silver, immunocytochemical and gold chloride methods. In order to locate the stellate cells, the hepatic lobules were divided into 10 zones. The stellate cells were readily identified in Golgi preparations by their striking dendritic appearance with branching processes encompassing the sinusoids. The stellate cells in the centrolobular zones were conspicuously dendritic with longer processes in conspicuously dendritic with longer processes in comparison to those emitted by periportal elements. Such arborizations were studded with numerous thorn-like microprojections. Desmin immunoreaction in the periportal zones was stronger than that in the centrolobular zones. Vitamin-A storage in the stellate cells was well developed in zones 2–4, but reduced gradually toward the central region. The perisinusoidal etellate cells display marked heterogeneity in morphology and function based on their zonal location in the hepatic lobule.     

Perisinusoidal stellate cells of the liver: important roles in retinol metabolism and fibrosis

In mammals, liver perisinusoidal stellate cells play an important role as a main store of body retinol (vitamin A). This fat-soluble vitamin is essential for vision, and regulates differentiation and growth of many cell types during embryonal development as well as in adult tissues. Thus, many cell types require a continuous supply of retinol. The storage of retinol (as retinyl esters) in stellate cells ascertains ample access of retinol to such cells also during periods with a low dietary intake. In lower vertebrates such as fish, vitamin A-storing stellate cells are found not only in the hepatic lobule, but also in the connective tissues of organs like intestine, kidney, ovaries, testes, and gills. Extrahepatic vitamin A-storing stellate cells are found in higher vertebrates when excessive doses of vitamin A are administered. It is not clear at present whether these cells also play a role in retinol metabolism under normal conditions. Stellate cells proliferate in a fibrotic liver, and they have been found to synthesize connective tissue compounds such as collagen. It was recently demonstrated that stellate cells are the principal cellular source of collagen and other extracellular substances in normal as well as fibrotic livers. Therefore, stellate cells, which seem to be a specialized type of pericyte, have a central role in the pathological changes observed during the development of liver fibrosis.     

Expression of cell-adhesion molecules in embryonic induction. I. Morphogenesis of nestling feathers

The potential relationship of cell adhesion to embryonic induction during feather formation was examined by immunohistochemical analysis of the spatiotemporal distribution of three cell-adhesion molecules (CAMs), neural CAM (N-CAM), liver CAM (L-CAM), and neuron-glia CAM (Ng-CAM), and of substrate molecules (laminin and fibronectin) in embryonic chicken skin. The N-CAM found at sites of embryonic induction in the feather was found to be similar to brain N-CAM as judged by immuno-cross-reactivity, migratory position in PAGE, and the presence of embryonic to adult conversion. In contrast to the N-CAM found in the brain, however, only one polypeptide of Mr 140,000 was seen. N-CAM-positive dermal condensations were distributed periodically under L-CAM-positive feather placodes at those sites where basement membranes are known to be disrupted. After initiation of induction, L-CAM-positive placode cells became transiently N-CAM-positive. N-CAM was asymmetrically concentrated in the dorsal region of the feather bud, while fibronectin was concentrated in the ventral region. During feather follicle formation, N-CAM was expressed in the dermal papilla and was closely apposed to the L-CAM-positive papillar ectoderm, while the dermal papilla showed no evidence of laminin or fibronectin. The collar epithelium was both N-CAM- and L-CAM-positive. During the formation of the feather filament, N-CAM appeared periodically and asymmetrically on basilar cells located in the valleys between adjacent barb ridges. In contrast to the two primary CAMs, Ng-CAM was found only on nerves supplying the feather and the skin. These studies indicate that at each site of induction during feather morphogenesis, a general pattern is repeated in which an epithelial structure linked by L-CAM is confronted with periodically propagating condensations of cells linked by N-CAM.     

Postnatal changes in sublobular distribution of NADPH-cytochrome P-450 reductase in rat liver

Summary Immunohistochemical distribution of NADPH-cytochrome P-450 reductase (NADPH-ferrihaemoprotein reductase; EC 1.6.2.4.) in the liver lobule was examined during development of the rat. From the 19th day of gestation to 4 days after birth, the enzyme was distributed uniformly throughout the lobule. The immunostaining for the enzyme was weak before birth, and became slightly stronger after birth. A slightly uneven distribution of immunoreactivity, stronger in perivenular zones, appeared at 5 days after birth. Then, the staining intensity in perivenular zones became progressively stronger with age, except for a slight increase between 10 and 20 days of age. The intensity in periportal zones also increased gradually, although it remained weaker than that in perivenular zones. Around 30 days of age, the distribution of the immunostaining, stronger in perivenular than in periportal zones, was similar to that seen in the lobules of adult animals. thus, heterogeneity among hepatocytes with respect to the enzyme content is not present in fetal and newborn rats but develops gradually during postnatal development; the postnatal growth of the liver is accompanied by a change in the pattern of the distribution of this enzyme within the lobule.     

Immunoelectron microscopic localization of neural cell adhesion molecules (L1, N-CAM, and myelin-associated glycoprotein) in regenerating adult mouse sciatic nerve

The localization of the neural cell adhesion molecules L1, N-CAM, and the myelin-associated glycoprotein was studied by pre- and postembedding staining procedures at the light and electron microscopic levels in transected and crushed adult mouse sciatic nerve. During the first 2-6 d after transection, myelinated and nonmyelinated axons degenerated in the distal part of the proximal stump close to the transection site and over the entire length of the distal part of the transected nerve. During this time, regrowing axons were seen only in the proximal, but not in the distal nerve stump. In most cases L1 and N-CAM remained detectable at cell contacts between nonmyelinating Schwann cells and degenerating axons as long as these were still morphologically intact. Similarly, myelin-associated glycoprotein remained detectable in the periaxonal area of the degenerating myelinated axons. During and after degeneration of axons, nonmyelinating Schwann cells formed slender processes which were L1 and N-CAM positive. They resembled small-diameter axons but could be unequivocally identified as Schwann cells by chronical denervation. Unlike the nonmyelinating Schwann cells, only few myelinating ones expressed L1 and N-CAM. At the cut ends of the nerve stumps a cap developed (more at the proximal than at the distal stump) that contained S-100-negative and fibronectin-positive fibroblast-like cells. Most of these cells were N-CAM positive but always L1 negative. Growth cones and regrowing axons expressed N-CAM and L1 at contact sites with these cells. Regrowing axons of small diameter were L1 and N-CAM positive where they made contact with each other or with Schwann cells, while large-diameter axons were only poorly antigen positive or completely negative. 14 d after transection, when regrowing axons were seen in the distal part of the transected nerve, regrowing axons made L1- and N-CAM-positive contacts with Schwann cells. When contacting basement membrane, axons were rarely found to express L1 and N-CAM. Most, if not all, Schwann cells associated with degenerating myelin expressed L1 and N-CAM. In crushed nerves, the immunostaining pattern was essentially the same as in the cut nerve. During formation of myelin, the sequence of adhesion molecule expression was the same as during development: L1 disappeared and N-CAM was reduced on myelinating Schwann cells and axons after the Schwann cell process had turned approximately 1.5 loops around the axon. Myelin-associated glycoprotein then appeared both periaxonally and on the turning loops of Schwann cells in the uncompacted myelin.(ABSTRACT TRUNCATED AT 400 WORDS)     

Differential distribution of B16F10 melanoma cells in the liver lobule

The distribution pattern and the number of tumor cells arrested in the liver were studied in mouse livers. Mice were perfused intravascularly with a suspension of B16F10 melanoma cells. The animals were sacrificed at 0, 1, 5, and 20 min after tumor cell perfusion. The pattern of tumor cell distribution was studied by morphological methods, and by a combined method of fluorescent-tumor cell labelling and histochemical succinate dehydrogenase activity on frozen sections, in order to define the localization of tumor cells arrested in the liver lobule. The results show that the tumor cells have an exclusive distribution in the periportal regions of the liver lobule (identified as the high succinate dehydrogenase activity areas), and that the cells are not arrested in the pericentral regions (identified as the low succinate dehydrogenase activity areas). In addition, indomethacin treatment (2 mg/kg/day) induced an increase in the number of melanoma cells arrested in the liver, but a different distribution with respect to controls was not observed. These results show that periportal regions of the liver lobule constitute a particular domain in which the B16F10 melanoma cells present a special retention ability that can be modulated by indomethacin treatment.     

A new method for the isolation of fresh hepatocytes from periportal and pericentral regions of the liver lobule

A simple method which avoids the use of perfusion with calcium free buffer, hydrolytic enzymes and detergents has been developed to obtain fresh hepatocytes from periportal and pericentral regions of the liver lobule. Cylindrical plugs (200 x 500 microns) of periportal and pericentral areas of the rat liver lobule weighing about 1 mg were collected with a micropunch from fresh or perfused liver. Ninety percent of cells were intact as assessed from trypan blue staining. Glutamine synthetase activity was detected predominantly (ca. 85%) in plugs isolated from pericentral regions indicating that this method allows selective harvesting of pure sublobular zones of the liver lobule. Rates of oxygen uptake measured at 25 degrees C by plugs from livers perfused in the anterograde direction were 56 +/- 5 and 33 +/- 7 mumol/g/h by periportal and pericentral plugs, respectively, values similar to data obtained from the intact organ. This method provides new opportunities to study the regulation of basic metabolic processes in cells from sublobular areas under nearly physiological conditions.     

Intercellular communication among liver cells in the perisinusoidal space of the injured liver: Pathophysiology and therapeutic directions

Hepatic stellate cells (HSCs) in the perisinusoidal space are surrounded by hepatocytes, liver sinusoidal endothelial cells, Kupffer cells, and other resident immune cells. In the normal liver, HSCs communicate with these cells to maintain normal liver functions. However, after chronic liver injury, injured hepatocytes release several proinflammatory mediators, reactive oxygen species, and damage-associated molecular patterns into the perisinusoidal space. Consequently, such alteration activates quiescent HSCs to acquire a myofibroblast-like phenotype and express high amounts of transforming growth factor-β1, angiopoietins, vascular endothelial growth factors, interleukins 6 and 8, fibril forming collagens, laminin, and E-cadherin. These phenotypic and functional transdifferentiation lead to hepatic fibrosis with a typical abnormal extracellular matrix synthesis and disorganization of the perisinusoidal space of the injured liver. Those changes provide a favorable environment that regulates tumor cell proliferation, migration, adhesion, and survival in the perisinusoidal space. Such tumor cells by releasing transforming growth factor-β1 and other cytokines, will, in turn, activate and deeply interact with HSCs via a bidirectional loop. Furthermore, hepatocellular carcinoma-derived mediators convert HSCs and macrophages into protumorigenic cell populations. Thus, the perisinusoidal space serves as a critical hub for activating HSCs and their interactions with other cell types, which cause a variety of liver diseases such as hepatic inflammation, fibrosis, cirrhosis, and their complications, such as portal hypertension and hepatocellular carcinoma. Therefore, targeting the crosstalk between activated HSCs and tumor cells/immune cells in the tumor microenvironment may also support a promising therapeutic strategy.     

Cells with morphological and immunohistochemical features of hepatic stellate cells (Ito cells) form an extralittoral (extrasinusoidal) compartment in the cirrhotic rat liver.

Systematic studies on hepatic stellate cells and myofibroblasts have so far mainly focused on cells located in the perisinusoidal space of Disse, the so-called littoral compartment. Here, these cells play a key role for intralobular fibrogenesis and sinusoidal capillarization. However, advanced hepatic fibrosis and cirrhosis are characterized by portal tract fibrosis and septal fibrosis, thus involving cells outside the perisinusoidal space. To study the question as to whether hepatic stellate cells occur and are expanded in an extralittoral (extrasinusoidal) compartment in cirrhogenesis, we systematically analyzed the distribution and density of desminreactive stellate cells in a rat model of hepatic fibrosis. Fibrosis and remodeling of the liver were induced by bile duct ligation, and stellate cells were identified by single and double immunohistochemistry. We can show that desmin-reactive cells are reproducibly detectable in extralittoral compartments of the normal and fibrotic rat liver. Periductular extralittoral stellate cells are significantly more frequent in cirrhosis, indicating that extralittoral stellate cells expand in concert with proliferating ductules. The findings suggest that ductular proliferation thought to represent a pacemaker of hepatic remodeling is accompanied by a population of cells exhibiting the same phenotype as perisinusoidal stellate cells.     

Phenotypic changes and loss of N-CAM-mediated adhesion in transformed embryonic chicken retinal cells

Transformation of 6-d-old embryonic chicken retinal cells by Rous sarcoma virus (RSV) was found to cause significant changes in several cellular properties including adhesiveness, motility, and state of differentiation. The alterations in cell adhesivity were analyzed by means of specific antibodies to the calcium-independent neural cell adhesion molecule, N-CAM. In the RSV-transformed cells the amount of N-CAM present at the cell surface was significantly decreased relative to normal cells, as assessed by immunofluorescent staining, specific immunoprecipitation, and immunoblotting experiments. This decrease was reflected in a marked reduction in N-CAM-mediated adhesiveness measured in vitro. A different, calcium-dependent, adhesive system also present on neurons was not detectably altered by RSV transformation and, in contrast with previous studies on normal neurons, this adhesive system was detected without treatment by proteases. In culture, the transformed cells formed fewer and less compact colonies than the normal retinal cells. Observation of the RSV-transformed retinal cells by time-lapse cinematography confirmed the reduction in adhesiveness and also revealed that the transformed cells were more highly motile than their normal counterparts. In addition, RSV transformation appeared to alter the differentiation of the cultured retinal cells. Immunofluorescent staining studies indicated that in contrast to mature neurons, transformed neural retinal cells expressed the 34,000-mol-wt tyrosine kinase substrate and reduced amounts of a neuron-specific ganglioside recognized by monoclonal antibody A2B5. These characteristics are shared by untransformed glial cells. In double immunofluorescent staining experiments, many cells expressed both N-CAM and pp60src shortly after viral infection, which implies that the N-CAM-positive neuroepithelial cells were transformed by RSV. In addition, a highly purified population of N-CAM-positive neural retinal cells, selected using a fluorescence-activated cell sorter, was rapidly and extensively transformed by RSV at rates comparable to those of the unfractionated population. These results established that the transformed cells were largely derived from RSV-infected neuroepithelial cells rather than from a small population of retinal glial cells present in the primary culture. The findings suggest reconsideration of the possible origin of tumors classified by morphological criteria as derived from glia and raise the possibility that the normal homologue of pp60src may play a role in the commitment of neuroepithelial cells to neuronal or glial differentiation pathways.     

Hepatic retinol metabolism. Distribution of retinoids, enzymes, and binding proteins in isolated rat liver cells

The main retinoids and some binding proteins and enzymes involved in retinol metabolism have been quantified in different types of rat liver cells. Hepatic perisinusoidal stellate cells contained 28-34 nmol of retinoids/10(6) cells, and parenchymal liver cells contained 0.5-0.8 nmol of retinoids/10(6) cells, suggesting that as much as 80% of more of total liver retinoids might be stored in stellate cells with the rest stored in parenchymal cells. Isolated endothelial cells and Kupffer cells contained very low levels of retinoids. More than 98% of the retinoids recovered in stellate cells were retinyl esters. Isolated parenchymal and stellate cell preparations both contained considerable retinyl palmitate hydrolase and acyl-CoA:retinol acyltransferase activities. Parenchymal cells accounted for about 75-80% of the total hepatic content of these two enzyme activities, with the rest located in stellate cells. On a cell protein basis, the concentrations of both of these activities were much greater in stellate cells than in parenchymal cells. In contrast, cholesteryl oleate and triolein hydrolase activities were fairly evenly distributed in all types of liver cells. Large amounts of cellular retinol binding proteins were also found in parenchymal and stellate cells. Although parenchymal cells accounted for more than 90% of hepatic cellular retinol binding protein, the concentration of the protein in stellate cells (per unit protein) was 22 X greater than that in parenchymal cells. Stellate cells were also enriched in cellular retinoic acid binding protein. Thus, both parenchymal and stellate cells contain substantial amounts of retinoids and of the enzymes and intracellular binding proteins involved in retinol metabolism. Stellate cells are particularly enriched in these several components.     

Liver cell heterogeneity. The distribution of pyruvate kinase and phosphoenolpyruvate carboxykinase (GTP) in the liver lobule of fed and starved rats.

Pyruvate kinase and phosphoenolpyruvate carboxykinase activities were determined in microdissected freeze-dried liver cells from the periportal and pericentral area of the liver lobule. Pyruvate kinase activity was measured by a microfluorimetric procedure adapted to 20-200 ng tissue dry weight. In livers from fed rats, its activity was twice as high in the central zone as in the periportal cells; starvation reduced this gradient by decreasing central activities. Phosphoenolpyruvate carboxykinase activity was measured by a microradiochemical technique in 100-300 ng tissue dry weight. In livers from fed rats, this enzyme was nearly 3 times more active in the periportal cells than in the central area. Starvation increased this enzyme in both zones with a more pronounced change in the central cells. The results indicate a heterogeneous distribution of enzymes of carbohydrate metabolism in the liver lobule. Gluconegenesis seems to be localized preferentially in periportal hepatocytes, whereas the glycolytic enzyme was found to be more active in cells surrounding the pericentral liver cells.     

Immunoelectron microscopic localization of the neural cell adhesion molecules L1 and N-CAM during postnatal development of the mouse cerebellum

The cellular and subcellular localization of the neural cell adhesion molecules L1 and N-CAM was studied by pre- and postembedding immunoelectron microscopic labeling procedures in the developing mouse cerebellar cortex. The salient features of the study are: L1 displays a previously unrecognized restricted expression by particular neuronal cell types (i.e., it is expressed by granule cells but not by stellate and basket cells) and by particular subcellular compartments (i.e., it is expressed on axons but not on dendrites or cell bodies of Purkinje cells). L1 is always expressed on fasciculating axons and on postmitotic, premigratory, and migrating granule cells at sites of neuron-neuron contact, but never at contact sites between neuron and glia, thus strengthening the view that L1 is not involved in granule cell migration as a neuron-glia adhesion molecule. While N-CAM antibodies reacting with the three major components of N-CAM (180, 140, and 120 kD) show a rather uniform labeling of all cell types, antibodies to the 180-kD component (N-CAM180) stain only the postmigratory granule cell bodies supporting the notion that N-CAM180, the N-CAM component with the longest cytoplasmic domain, is not expressed before stable cell contacts are formed. Furthermore, N-CAM180 is only transiently expressed on Purkinje cell dendrites. N-CAM is present in synapses on both pre- and post-synaptic membranes. L1 is expressed only preterminally and not in the subsynaptic membranes. These observations indicate an exquisite degree of fine tuning in adhesion molecule expression during neural development and suggest a rich combinatorial repertoire in the specification of cell surface contacts.     

Gluconeogenesis predominates in periportal regions of the liver lobule

Rates of gluconeogenesis from lactate were calculated in periportal and pericentral regions of the liver lobule in perfused rat livers from increases in O2 uptake due to lactate. When lactate (0.1-2.0 mM) was infused into livers from fasted rats perfused in either anterograde or the retrograde direction, a good correlation (r = 0.97) between rates of glucose production and extra O2 uptake by the liver was observed as expected. Rates of oxygen uptake were determined subsequently in periportal and pericentral regions of the liver lobule by placing miniature oxygen electrodes on the liver surface and measuring the local change in oxygen concentration when the flow was stopped. Basal rates of oxygen uptake of 142 +/- 11 and 60 +/- 4 mumol X g-1 X h-1 were calculated for periportal and pericentral regions, respectively. Infusion of 2 mM lactate increased oxygen uptake by 71 mumol X g-1 X h-1 in periportal regions and by 29 mumol X g-1 X h-1 in pericentral areas of the liver lobule. Since the stoichiometry between glucose production and extra oxygen uptake is well-established, rates of glucose production in periportal and pericentral regions of the liver lobule were calculated from local changes in rates of oxygen uptake for the first time. Maximal rates of glucose production from lactate (2 mM) were 60 +/- 7 and 25 +/- 4 mumol X g-1 X h-1 in periportal and pericentral zones of the liver lobule, respectively. The lactate concentrations required for half-maximal glucose synthesis were similar (0.4-0.5 mM) in both regions of the liver lobule in the presence or absence of epinephrine (0.1 microM). In the presence of epinephrine, maximal rates of glucose production from lactate were 79 +/- 5 and 59 +/- 3 mumol X g-1 X h-1 in periportal and pericentral regions, respectively. Thus, gluconeogenesis from lactate predominates in periportal areas of the liver lobule during perfusion in the anterograde direction; however, the stimulation by added epinephrine was greatest in pericentral areas. Differences in local rates of glucose synthesis may be due to ATP availability, as a good correlation between basal rates of O2 uptake and rates of gluconeogenesis were observed in both regions of the liver lobule in the presence and absence of epinephrine. In marked contrast, when livers were perfused in the retrograde direction, glucose production was 28 +/- 5 mumol X g-1 X h-1 in periportal areas and 74 +/- 6 mumol X g-1 X h-1 in pericentral regions.(ABSTRACT TRUNCATED AT 400 WORDS)     

Serum components and activated Ha-ras antagonize expression of perivenous marker genes stimulated by beta-catenin signaling in mouse hepatocytes

Hepatocytes of the periportal and perivenous zones of the liver lobule show marked differences in the contents and activities of many enzymes and other proteins. Previous studies from our and other groups have pointed towards an important role of beta-catenin-dependent signaling in the regulation of expression of genes encoding proteins with preferential perivenous localization, whereas, in contrast, signaling through Ras-dependent pathway(s) may induce a 'periportal' phenotype. We have now conducted a series of experiments to further investigate this hypothesis. In transgenic mice with scattered expression of an activated Ha-ras (Ha-ras(G12V)) mutant in liver, expression of the perivenous markers glutamine synthetase and two cytochrome P450 isoforms was completely abolished in those hepatocytes demonstrating constitutively activated extracellular signal-regulated kinase activity, even though they were located directly adjacent to central veins. Similarly, incubation of primary hepatocytes or hepatoma cells with increasing amounts of serum caused a concentration-dependent attenuation of expression of perivenous marker mRNAs, whereas the expression of periportal markers was increased. The inhibitory effect of high amounts of serum on the expression of perivenous markers was also observed if their expression was stimulated by activation of beta-catenin signaling, and comparable inhibitory effects were seen in cells stably transfected with a T-cell factor/lymphoid-enhancing factor-driven luciferase reporter. Epidermal growth factor could partly mimic serum effects in hepatoma cells, and its effect could be blocked by an inhibitor of extracellular signal-regulated kinase activity. These data suggest that activation of the Ras/mitogen-activated protein kinase (extracellular signal-regulated kinase) pathway favors periportal gene expression while simultaneously antagonizing a perivenous phenotype of hepatocytes.     

Study of the function and regulation of liver N-CAM in Xenopus laevis.

The liver of Xenopus laevis is a unique exception in terms of the cell adhesion molecules (CAM) which it expresses. In most species, hepatocytes are characterized by the expression of the epithelial Ca(2+)-dependent CAM E-cadherin or of closely related variants of this molecule (e.g., L-CAM); in Xenopus liver, however, the levels of expression of epithelial cadherins is very low while a thyroxine-inducible isoform of N-CAM is expressed in postmetamorphic hepatocytes. Since Xenopus liver N-CAM is localized in regions of contact between hepatocytes, it has been proposed that it might be involved in mediating hepatocyte adhesion in this species. In this study, we demonstrate that N-CAM can indeed act as a functional adhesion molecule in the liver of Xenopus and that its expression is correlated with a number of profound morphological changes of this organ. After thyroxine treatment, hepatocytes are no longer organized in long loose cords but in compact lobules of cells. Furthermore, at the ultrastructural level, plasma membranes are in much closer proximity with the appearance of electron-dense material in areas of closer contact. We have established two novel culture systems for premetamorphic Xenopus hepatocytes as adherent and non-adherent cells, and we describe the induction of expression of N-CAM in these cells. Given the difference in the profile of adhesion molecules present in the liver of Xenopus and of other species, our results are discussed in view of the importance of the expression of a specific set of cell adhesion molecules in defining the development of homologous organs in different species.     

Expression of cell adhesion molecules in the olfactory system of the adult mouse: presence of the embryonic form of N-CAM

The expression of the neural cell adhesion molecules N-CAM and L1 was investigated in the olfactory system of the mouse using immunocytochemical and immunochemical techniques. In the olfactory epithelium, globose basal cells and olfactory neurons were stained by the polyclonal N-CAM antibody reacting with all three components of N-CAM (N-CAM total) in their adult and embryonic states. Dark basal cells and supporting cells were not found positive for N-CAM total. The embryonic form of N-CAM (E-N-CAM) was only observed on the majority of globose basal cells, the precursor cells of olfactory neurons, and some neuronal elements, probably immature neurons, since they were localized adjacent to the basal cell layer. Differentiated neurons in the olfactory epithelium did not express E-N-CAM. In contrast to N-CAM total, the 180-kDa component of N-CAM (N-CAM180) and E-N-CAM, L1 was not detectable on cell bodies in the olfactory epithelium. L1 and N-CAM180 were strongly expressed on axons leaving the olfactory epithelium. Olfactory axons were also labeled by antibodies to N-CAM180 and L1 in the lamina propria and the nerve fiber and glomerular layers of the olfactory bulb, but only some axons showed a positive immunoreaction for E-N-CAM. Ensheathing cells in the olfactory nerve were observed to bear some labeling for N-CAM total, L1, and N-CAM180, but not E-N-CAM. In the olfactory bulb, L1 was not present on glial cells. In contrast, N-CAM180 was detectable on some glia and N-CAM total on virtually all glia. Glia in the nerve fiber layer were labeled by E-N-CAM antibody only at the external glial limiting membrane. In the glomerular layer, E-N-CAM expression was particularly pronounced at contacts between olfactory axons and target cells. The presence of E-N-CAM in the adult olfactory epithelium and bulb was confirmed by Western blot analysis. The continued presence of E-N-CAM in adulthood on neuronal precursor cells, a subpopulation of olfactory axons, glial cells at the glia limitans, and contacts between olfactory axons and their target cells indicates the retention of embryonic features in the mammalian olfactory system, which may underlie its remarkable regenerative capacity.     

Abnormal HNK-1 expression in the cerebellum of an N-CAM null mouse

Human natural killer antigen-1 (HNK-1) is a carbohydrate epitope associated with sulfoglucuronylglycolipids and glycoproteins. Biochemical analyses have demonstrated associations between the HNK-1 epitope and isoforms of the neural cell adhesion molecule (N-CAM) family. In the cerebellum, HNK-1 is prominently expressed in Purkinje cell dendrites and Golgi cells. Purkinje cell expression of HNK-1 reveals an array of parasagittal stripes and transverse zones. Interestingly, the parasagittal expression pattern of HNK-1 is different from those reported with several other markers such as zebrin II/aldolase C and the small heat shock protein HSP25. N-CAM null knockout mice were used to explore the possible role of the HNK-1/N-CAM interaction during the topographical organization of the cerebellar cortex. N-CAM null mice have no N-CAM immunoreactivity but otherwise the cerebellum appears morphologically normal. Further, in the N-CAM null HNK-1 immunoreactivity is abolished from Purkinje cell dendrites but is retained on Golgi cells and neurons of the cerebellar nuclei. Despite the absence of N-CAM/HNK-1, parasagittal stripes and transverse zones in the cerebellum as revealed by using zebrin II immunocytochemistry appear normal.     

Direct mobilization of retinol from hepatic perisinusoidal stellate cells to plasma.

We have studied the mechanism for mobilization of retinol from stellate cells. Our data show that perisinusoidal stellate cells isolated from liver contained retinol-binding protein (RBP) mRNA. By Western blot analysis we found that cultivated liver stellate cells secreted RBP into the medium. Cultivated stellate cells loaded in vitro with [3H]retinyl ester mobilized radioactive retinol as a complex with RBP. Furthermore, exogenous RBP added to the medium of cultured stellate cells increased the secretion of retinol to the medium. These data suggest that liver stellate cells in vivo mobilize retinol directly to the blood and that a transfer to parenchymal cells for secretion as holo-RBP is not required. The direct mobilization of retinol from liver stellate cells as retinol-RBP to blood is indirectly supported by the demonstration of RBP mRNA production and RBP secretion by lung stellate cells. The data suggest that the same mechanism for retinol mobilization may exist in hepatic and extrahepatic stellate cells. This is, vitamin A-storing stellate cells in liver, lungs, and probably also in other organs may synthesize their own RBP (or alternatively use exogenous RBP) and mobilize holo-RBP directly to the blood.