Cell biology of cytochrome P-450 in the liver.

Cytochromes P-450 (P-450) are members of a multigene superfamily of hemoproteins consisting the microsomal monooxygenase system with NADPH P-450 reductase (reductase) and/or reducing equivalents. Expression of many P-450 isoforms in hepatocytes is shown to be regulated at the level of transcription through interaction between cis-acting elements in the genes and DNA-binding (transacting) factors. Some isoforms of the CYP1A, 2B, 2E, and 3A subfamilies are regulated at the posttranscriptional level. For the topology of P-450 and reductase molecules in ER membrane of hepatocytes, models from stopped flow analysis and electron spin resonance are proposed. The densities of total P-450 and reductase molecules are revealed to be high enough to support the cluster model, suggesting that about ten P-450 molecules form an aggregate and surround one reductase molecule, and therefore the two enzymes form large micelles. ER proliferation after PB administration, which had been correlated with increase in P-450 level, is shown to be probably independent of the increase in P-450 level. There are considerable discrepancies among results reported on sublobular expression of various P-450 isoforms. Causes of the discrepancies are likely to be differences in experimental conditions of histochemical detection carried out and/or in species, strain, and/or sex.

Endoplasmic reticulum proliferates without an increase in cytochrome P-450 in hepatocytes of mice treated with phenobarbital and cobalt chloride.

To determine whether endoplasmic reticulum (ER) proliferation in hepatocytes after phenobarbital (PB) administration relates closely to cytochrome P-450 (P-450) increase, we have measured the amount of total P-450 per unit cytoplasmic volume (P-450 content) by microphotometry and estimated the area of ER per unit cytoplasmic volume (ER area) by morphometry in periportal, midzonal, and perivenular hepatocytes of mice injected daily with PB (100 mg/kg), or with PB (100 mg/kg) plus cobalt chloride (50 mg/kg) for three days. After injection of PB, the P-450 content and ER area increased in hepatocytes of the three sublobular zones. In mice treated with PB plus cobalt chloride, however, the ER area increased, but the P-450 content decreased or remained unchanged in hepatocytes of the three zones. We conclude that cobalt chloride inhibits the increase in total P-450 but has no effect on the proliferation of ER of hepatocytes in mice treated with PB, indicating a dissociation of ER proliferation and P-450 increase after administration of PB.

Quantitative analysis of endoplasmic reticulum and cytochrome P-450 in hepatocytes from rats injected with methylcholanthrene.

To examine whether the smooth endoplasmic reticulum (SER) proliferates in hepatocytes from animals treated with methylcholanthrene (MC) frequently used as an inducer for the enzymes of the microsomal mono-oxygenase system, we estimated the area of (smooth and rough) ER per unit cytoplasmic volume by morphometry in periportal, midzonal and perivenular hepatocytes from rats injected with 25 mg/kg MC once a day for 3 days. In addition, immunostaining intensity of major MC-inducible cytochrome P-450 (P-450) forms (1A1/1A2) and total P-450 content in the cytoplasm of hepatocytes in the three zones were measured by microphotometry to ascertain whether P-450 is sufficiently induced in each sublobular zone by the administration. In spite of significant increase in the staining intensity of P-450 1A1/1A2 and amount of total P-450, the proliferation of SER (and RER) did not occur in the three-zone hepatocytes from rats injected with MC. In perivenular hepatocytes, constitutive forms of P-450 other than 1A1/1A2 decreased (to 10%) instead of marked increase in P-450 1A1/1A2 (about 20 times), while the constitutive forms decreased to 50% in midzonal hepatocytes and remained unchanged in periportal hepatocytes after MC administration. In addition, the present results show divergence between biochemical and immumohistochemical results previously reported on MC-inducible P-450 after MC administration to be due primarily to a curvilinear relationship between content and intensity.

Culture of stromal cells derived from medullary cavity of human long bone in the presence of 1,25-dihydroxyvitamin D3, recombinant human bone morphogenetic protein-2, or ipriflavone.

We previously showed that stromal cells derived from bone marrow specimens formed at the fracture site of human long bone differentiated during culture to polygonal cells and spindle cells, and polygonal cells, but not spindle cells, produced calcified matrix. To clarify the origin of polygonal and/or spindle cells, and factors necessary for differentiation of marrow stromal cells to osteogenic cells, we cultured stromal cells derived from the normal (unfractured) medullary cavity (SCN) as well as stromal cells from the medullary cavity distant from the fracture site (SCF). After 3 weeks of primary culture and 2 days of secondary culture, the cells were cultured in medium containing 1,25-dihydroxyvitamin D3 (VD), recombinant human bone morphogenetic protein-2 (BMP), or ipriflavone (IF) for 3 weeks. For biochemical analysis, cells reaching confluence after 3 weeks of secondary culture were cultured with one of the factors for 3 days. Some of SCF cultured with VD or IF were transformed to polygonal cells, and showed high alkaline phosphatase (ALPase) activity and high osteocalcin and insoluble calcium production. Cloned polygonal cells from the SCF formed nodules and aggregates consisting of calcium. Other SCF cultured with VD or IF and SCF cultured with BMP were spindle shaped. Some spindle-shaped cells from SCF cultured with BMP or IF revealed high ALPase activity and high osteocalcin production, comparable with the spindle cells from the fracture site. However, spindle-shaped cells from SCF cultured with VD and other spindle-shaped cells from SCF cultured with BMP or IF showed low ALPase activity and low osteocalcin production. The results show that SCF probably contain at least three subpopulations: (a) cells that differentiate to polygonal cells by the influence of VD or IF; (b) cells that differentiate to the spindle cells by the influence of BMP or IF; and (c) cells that are not transformed by the influence of VD, BMP, or IF.

Culture of marrow stromal cells derived from bone marrow specimens formed at fracture site of human long bone.

To study the process of the internal callus formation, we cultivated marrow stromal cells derived from bone marrow specimens formed at fracture sites of human long bone in alpha-modified Eagle's medium containing 10% fetal calf serum. After 2 weeks of culture, the cells formed two types of colonies; one consisted of spindle cells, and the other comprised of polygonal cells. The two types of colonies were separated and cultured further. The spindle and polygonal cells proliferated to confluence within 3 weeks and after 4 weeks, respectively, after the separation. Both the spindle and polygonal cells showed on the plasma membrane moderate intensity of staining reaction of alkaline phosphatase (ALPase) activity before the period of confluence and strong intensity during the period of confluence. Then, the spindle cells did not produce calcified matrix, but detached from the dish after 6-8 weeks of culture. In the colonies of polygonal cells, however, dense nodules were formed after 9 weeks of culture, which became visible to the naked eye as white aggregates after 11-12 weeks. Electron microscopic studies on the polygonal cells demonstrated matrix vesicles in the intercellular ground substance after 6 weeks of culture, and electron-dense needle-like crystals on the matrix vesicles after 8-10 weeks of culture. On the basis of infrared spectroscopic analysis, the aggregates were composed of hydroxyapatite. Thus, stromal cells derived from bone marrow specimens formed at fracture site of human long bone differentiated to spindle and polygonal cells containing high ALPase activity (a marker for osteogenic capacity) during culture, and the polygonal cells but not spindle cells produced the calcified matrix.

An improved microphotometry system for measurement of cytochrome P-450 in hepatocyte cytoplasm.

To measure cytochrome P-450 (P-450) content in hepatocyte cytoplasm, we developed a dual monochromator-equipped microphotometry system (KWSP-1). Simultaneous measurements of absorbance at 450 and 490 nm with narrow band width (0.5 nm) and small spot size (2 microns) were accomplished by this system. Corresponding fields in serial sections could be easily and rapidly identified under the Nomarski imaging mode of KWSP-1. Photometric accuracy and repeatability of wavelength setting of KWSP-1 were also satisfactory for measurement of P-450. With this system, it is thus possible to measure the extinction of P-450 from many small measuring areas and to precisely determine P-450 content in the cytoplasm of rat hepatocytes. A microphotometric method was developed using cuvette slides and two serial 10-microns thick sections (mapping method). The intracellular distribution of P-450 in individual hepatocytes could be visualized by the mapping method with KWSP-1. However, this method was not applicable to tissue sections containing hemoglobin larger than 4 microM.

Measurement of NADPH-ferrihemoprotein reductase content in sections of liver.

We developed a method for measuring the content of NADPH-ferrihemoprotein reductase in sections of liver. First, reductase in sections of rat liver was detected with the indirect immunoperoxidase reaction. Subsequently, specific absorbances were measured in the stained sections by microphotometry. Then, the resulting specific absorbances were converted into the reductase content in the sections using an apparent extinction coefficient obtained from a nitrocellulose binding assay. The average of the reductase content in hepatocytes in periportal, intermediate, and perivenous zones thus measured was consistent with the value in liver homogenates estimated by enzyme-linked immunosorbent assay. Therefore, the present method gave accurate measurement of the reductase content in the sections. Perivenous hepatocytes contained 1.5 times as much reductase (1.15 nmol/g liver, mean for five animals) as that in periportal hepatocytes (0.74 nmol/g liver). The reductase content in hepatocytes in the intermediate zone (0.93 nmol/g liver) was intermediate between values of the periportal and perivenous hepatocytes.

Glucagon receptors in endothelial and Kupffer cells of mouse liver.

To determine whether hepatic sinusoidal cells contain glucagon receptors and, if so, to study the significance of the receptors in the cells, binding of [125I]-glucagon to nonparenchymal cells (mainly endothelial cells and Kupffer cells) isolated from mouse liver was examined by quantitative autoradiography and biochemical methods. Furthermore, the pathway of intracellular transport of colloidal gold-labeled glucagon (AuG) was examined in vivo. Autoradiographic and biochemical results demonstrated many glucagon receptors in both endothelial cells and Kupffer cells, and more receptors being present in endothelial cells than in Kupffer cells. In vivo, endothelial cells internalized AuG particles into coated vesicles via coated pits and transported the particles to endosomes, lysosomes, and abluminal plasma membrane. Therefore, receptor-mediated transcytosis of AuG occurs in endothelial cells. The number of particles present on the abluminal plasma membrane was constant if the amount of injected AuG increased. Therefore, the magnitude of receptor-mediated transcytosis of AuG appears to be regulated by endothelial cells. Kupffer cells internalized the ligand into cytoplasmic tubular structures via plasma membrane invaginations and transported the ligand exclusively to endosomes and lysosomes, suggesting that the ligand is degraded by Kupffer cells.

Postnatal development and sublobular distribution of cytochrome P-450 in rat liver: a microphotometric study.

To study the process of expression of cytochrome P-450 (P-450) in hepatocytes during development, we measured microphotometrically the P-450 content in periportal and perivenular hepatocytes of male rats during peri- and postnatal growth. From Day 19 of gestation to Day 5 after birth, P-450 content in both periportal and perivenular hepatocytes increased markedly (periportal 1046%; perivenular 819%). The content in periportal hepatocytes remained unchanged from 5 to 20 days of age, and increased slightly (24%) from 20 to 45 days of age. However, the content in perivenular hepatocytes increased progressively (105%) between 5 and 45 days of age. The difference in P-450 content became apparent between periportal and perivenular hepatocytes after 7 days of age. The content in periportal or perivenular hepatocytes reached the adult level at 45 days of age. Therefore, the perinatal period is the time at which a marked increase in P-450 occurs in hepatocytes throughout the liver lobule. The subsequent period before weaning is the time at which the sublobular heterogeneous distribution of P-450 appears. The period after weaning is the time at which a slight increase in P-450 content in periportal hepatocytes and a marked increase in the enzyme in perivenular hepatocytes takes place.

Relationship between immunostaining intensity and antigen content in sections.

We studied the relationship between staining intensity of immunohistochemical reaction and antigen content in sections. Alpha-fetoprotein (AFP) and albumin in sections cut from livers of newborn, 5-, 10-, 20-, and 60-day-old rats were examined as examples. First, we compared average immunostaining intensity (sum of specific absorbance in pixel/number of pixels) measured by image processing (IP), with antigen content measured by immunochemical assay to determine whether the intensity is proportional to antigen content. The intensity of AFP was proportional to the antigen content, whereas that of albumin was not. Subsequently, the antigen preservation test was carried out to determine whether the intensity was decreased by fixation and, if so, which type of decrease (proportional or disproportionate) occurred. Thereafter, antigen content in the same portion in the same immunostained section was measured by the microphotometric (MP) method followed by the IP method, because the MP method gives a low average antigen content when a decrease in antibody binding occurs in sections, whereas the average antigen content measured by the IP method is unchanged. The intensity of AFP decreased primarily by a proportional decrease in antigenicity during fixation. However, the intensity of albumin decreased not only by a proportional decrease during fixation but also by a disproportionate reduction in antibody binding during immunostaining or before fixation. The results indicate that AFP content in sections is measurable by quantitative immunohistochemical methods, whereas albumin content is not.

Effect of 3-methylcholanthrene administration on expression of cytochrome P-450 isoforms induced by phenobarbital in rat hepatocytes.

The effects of an inducer on expression of cytochrome P-450 (P-450) isoforms induced antecedently by another inducer are unknown. Thus, we examined the amount of phenobarbital (PB)-inducible P-450 isoforms (P-450 2B1/2B2) in hepatocytes from rats injected first with PB and then with 3-methylcholanthrene (MC) (PB+MC-treated animals) by quantitative immunohistochemistry. In addition, expression of P-450 2B2 mRNA was examined by in situ hybridization. In PB-treated animals, P-450 2B1/2B2 content increased in perivenular and midzonal hepatocytes. In PB+MC-treated animals, however, the PB-induced increase in 2B1/2B2 content was suppressed in perivenular hepatocytes but promoted in midzonal hepatocytes. The hybridization signal for P-450 2B2 mRNA appeared almost exclusively in perivenular hepatocytes after 24 hr of PB injection and disappeared after 48 hr of injection. In PB+MC-treated animals, however, strong hybridization signal was observed in midzonal and perivenular hepatocytes after 48 hr of PB injection. The promotion of the increase in P-450 2B1/2B2 content in midzonal hepatocytes in PB+MC-treated animals probably corresponds to the strong hybridization signal, whereas there appeared to be a divergence between the intensity of the signal and the content in perivenular hepatocytes. The results indicate that MC administration drastically influences the pattern of expression of P-450 isoforms induced by PB in perivenular and midzonal hepatocytes.

Localization of xenobiotic-responsive element binding protein in rat hepatocyte nuclei after methylcholanthrene administration as revealed by in situ Southwestern hybridization.

Xenobiotic-responsive element binding protein (XRE-BP), a heterodimer of aryl hydrocarbon receptor (AhR) and its nuclear translocator (Arnt), regulates the transcription of cytochrome P-450 1A1 gene (CYP1A1) through XRE in response to xenobiotic inducers. For a better understanding of the regulatory mechanism of CYP1A1 through XRE, localization of XRE-BP was examined in liver sections or isolated hepatocyte nuclei from control and 3-methylcholanthrene (MC)-treated rats by in situ Southwestern hybridization, using synthetic XRE as a probe, and was observed by confocal laser scanning microscopy and electron microscopy. Gel mobility shift assay and competitive binding assay showed specificity of the synthetic XRE probe. XRE-BP was exclusively localized in hepatocyte nuclei in liver sections from animals 3 hr after MC injection, whereas the protein was absent in hepatocyte cytoplasm in MC-treated animals and in hepatocyte nuclei and cytoplasm in control animals. In isolated hepatocyte nuclei, XRE-BP began to accumulate in the central region between 0.5 and 3 hr, showed a peak between 3 and 6 hr, decreased gradually between 6 and 72 hr, and disappeared at 72 hr after MC injection. The protein was scarce in peripheral and nucleolar regions of the nucleus. Therefore, XRE-BP is formed in the nuclei of hepatocytes after MC stimulation. In addition, XRE-BP was found in isolated hepatocyte nuclei from control animals after preincubation with cytoplasmic lysate from MC-treated animals, although the protein was absent in the nuclei before the preincubation. These findings strongly suggest that AhR translocates from hepatocyte cytoplasm to the nucleus and forms XRE-BP in the nucleus after MC stimulation.

Microphotometric analysis of cytochrome P-450 in periportal, midzonal, and perivenular hepatocytes of mice treated with phenobarbital.

To obtain detailed information on the increase of cytochrome P-450 (P-450) content in periportal, midzonal, and perivenular hepatocytes after phenobarbital (PB) administration, and to study the mechanism of increased P-450 in the endoplasmic reticulum (ER), we estimated microphotometrically the P-450 content and morphometrically the area of ER in hepatocytes of three zones from mice injected with 35, 50, 100, or 150 mg/kg of PB for 3 days. The amount of P-450 per unit cytoplasmic volume and the number of P-450 molecules per unit ER area (P-450 number) were increased by injection of 50, 100, or 150 mg/kg, and the ER area per unit cytoplasmic volume was increased by injection of 100 or 150 mg/kg, in hepatocytes from all three zones. Thus, the amount of P-450 in hepatocytes appeared in general to increase multiplicatively by simultaneous increases in both the P-450 number and the ER area. Furthermore, we could recognize two general types of relationship in the P-450 number and ER area between the patterns of change and the increasing doses: (a) increase in the P-450 number without ER proliferation (active type) in periportal and perivenular hepatocytes after injection of low doses; and (b) increase in ER proliferation without increase in the P-450 number (passive type) in hepatocytes of all three zones after injection of high doses.

A new microphotometric method for measurement of cytochrome P-450 in sections of liver.

We developed a new microphotometric method for measuring the amounts of cytochrome P-450 (P-450) in fresh frozen sections of liver. Four serial frozen sections cut from the liver were separately incubated in 50 mM Tris-HCl buffer (pH 8.0) alone, in buffer containing sodium dithionite, in buffer saturated with carbon monoxide (CO), and in buffer saturated with CO and containing sodium dithionite. The difference between absorbance at 450 nm and that at 490 nm was measured in these sections with a simple microphotometer system. This method yielded precise amounts of P-450 in sections by measuring the true extinction of P-450 and by minimizing the effect of contaminating hemoproteins. Livers of adult rats contained large amounts of P-450, which was greater in perivenular hepatocytes than in periportal hepatocytes. In livers of newborn rats, however, small amounts of the enzyme were distributed evenly throughout the lobule.

High and testosterone-dependent glucose 6-phosphatase activity in epithelium of mouse seminal vesicle.

For study of the mechanism of seminal fructogenesis, glucose 6-phosphatase activity was examined cytochemically (a method modified from that of Wachstein and Meisel) and biochemically (the method of Leskes et al.) in seminal vesicles from normal, castrated, and castrated and testosterone-treated mice. The reaction product for the activity was localized in the endoplasmic reticulum and nuclear envelope of all cell types composing the seminal vesicle. In normal seminal vesicle, the reaction product was apparently more abundant in columnar and basal cells than in other cell types. Ten, 20, and 30 days after castration, the abundant amount of reaction product in columnar and basal cells decreased to the level in other cell types. In animals treated with testosterone after castration, however, the reaction product in columnar and basal cells remained abundant. If fructose 6-phosphate was added to the reaction medium in place of glucose 6-phosphate, the amount and pattern of deposition of the reaction product did not change. Changes in biochemical activity in castrated or castrated and testosterone-treated animals paralleled the cytochemical results. The results show that the high activity in columnar and basal cells is under the control of testosterone, and the role of this enzyme is probably to release fructose into the seminal fluid.

Measurement of NADPH-cytochrome P-450 reductase content in rat liver sections by quantitative immunohistochemistry with a video image processor.

We have a quantitative light microscopic immunohistochemical method using video image processing. First, an antigen (NADPH-cytochrome P-450 reductase) content in homogenates of livers of rats was measured by enzyme immunoassay. Then frozen sections from rat livers were incubated with the anti-NADPH-cytochrome P-450 reductase antibody under saturation conditions by the indirect immunoperoxidase method. Subsequently, relative staining intensities in small portions and those in wide areas in the sections were measured with a video image processor. Finally, the resulting relative values obtained from the small portions were converted into absolute NADPH-cytochrome P-450 reductase contents using the results of enzyme immunoassay and the average relative staining intensity obtained from the wide areas in the sections. The reductase content in sections from rat livers measured by the image processing method coincided with the content measured by the microphotometric method using a nitrocellulose model system. The present image processing method is applicable to measurement of contents of antigens that can not be immobilized in model systems.

Significance of high glucose-6-phosphatase activity in rat oviduct epithelium.

To study the origin of glucose in the oviduct fluid, we cytochemically examined glucose-6-phosphatase (G6Pase) activity in rat oviduct. The activity in the whole oviduct was also assayed biochemically. During proestrous, estrous, and metestrous phases, staining reaction for the activity was moderate in the epithelium of the caudal isthmus (CaI) and uterotubal junction (UJ), whereas it was weak in that of the ampulla (A) and cephalic isthmus (CeI). In the diestrous phase, staining reaction in the epithelium of CaI and UJ became strong although it remained weak in that of A and CeI. Reaction product for the activity was localized in the endoplasmic reticulum and nuclear envelope of all cell types in the epithelium. The amount of reaction product in secretory cells was small to moderate in CaI and UJ, and small in A and CeI during proestrus, estrus, and metestrus. In diestrous the amount became abundant in CaI and UJ and moderate in A and CeI. However, the amount in ciliated cells remained small in the four segments during the four phases. The biochemical activity in diestrous was greater than that in proestrus, estrus, or metestrus. This shows that the activity is high in secretory cells in the epithelium of CaI and UJ in the diestrous phase and suggests that the role of the high activity is to release glucose into the oviduct fluid for use by the embryo passing down the CaI and UJ to the uterus.

Relation between cytochrome P-450 increase and endoplasmic reticulum proliferation in hepatocytes of mice treated with phenobarbital: a microphotometric and morphometric study.

To obtain detailed information on phenobarbital (PB)-induced cytochrome P-450 (P-450) increase and endoplasmic reticulum (ER) proliferation in hepatocytes, we estimated microphotometrically the amount of P-450 per unit cytoplasmic volume and morphometrically the area of ER per unit cytoplasmic volume in hepatocytes adjacent to the portal area or central venule (1 periportal or 1 perivenular cells) and in the second and third layers from the portal area or central venule (2, 3 periportal or 2, 3 perivenular cells) from mice injected with 35, 50, 100, or 150 mg/kg PB once a day for 3 days. By dividing the P-450 amount by the ER area, the number of P-450 molecules per unit ER area was also calculated. In 1 and 2, 3 perivenular cells, except for 2, 3 perivenular cells after injection of 150 mg/kg PB, the amount of P-450 increased with ER proliferation and the number of P-450 molecules in ER remained unchanged after injection of 50, 100, or 150 mg/kg PB. In 2, 3 periportal cells, however, the P-450 amount and the number of P-450 molecules in ER increased markedly without or with some ER proliferation after injection of 50, 100, or 150 mg/kg PB; the P-450 increase appears to be generally independent of ER proliferation. The 1 periportal cells are probably exceptional hepatocytes that usually did not respond to PB stimulation.

Densities of NADPH-ferrihemoprotein reductase and cytochrome P-450 molecules in the endoplasmic reticulum membrane of rat hepatocytes.

In hepatocytes, NADPH-ferrihemoprotein reductase (reductase) has been hypothesized to exist as aggregates or micelles in endoplasmic reticulum (ER) membrane. However, if the number of reductase molecules per unit area of ER is low, this hypothesis cannot explain how a few reductase molecules efficiently reduce many P-450 molecules. To test this hypothesis, we estimated the numbers of reductase and P-450 molecules per unit ER area (reductase and P-450 densities) by microphotometry of the two enzymes in conjunction with morphometry of ER in periportal, midzonal, and perivenular rat hepatocytes. The reductase density in periportal, midzonal, and perivenular hepatocytes (107-179 molecules/microns 2 of ER) was high enough to efficiently reduce all P-450 molecules in the ER, although the value in perivenular hepatocytes was lowest owing to the relatively greater amount of ER in this region. The pattern of sublobular gradient in the reductase density was similar to that in the P-450 density. Consequently, the molar ratio of P-450 to reductase in ER was similar (about 40:1) in hepatocytes regardless of their positions within the liver lobule.

High glucose-6-phosphatase activity in non-pigmented epithelial cells of rabbit ciliary body.

For study of the origin of glucose in the aqueous humor, glucose-6-phosphatase (G6Pase) and hexokinase activities, and glycogen, were cytochemically examined in the ciliary body (CB) of rabbit. G6Pase activity was also assayed biochemically. The staining reaction for G6Pase activity was strong in the non-pigmented epithelium (NPE) in the pars plana and tips of ciliary processes in the region containing large ciliary pockets within the pars plicata. NPE cells contained abundant reaction product for G6Pase activity in the endoplasmic reticulum (ER) and nuclear envelope. However, NPE in other regions of the CB and pigmented epithelium (PE) of CB, and other areas surrounding the anterior and (PE) of CB, and other areas surrounding the anterior and posterior chambers, showed weak or no G6Pase staining reaction. Biochemical G6Pase activity in the whole ciliary body was relatively high. Both NPE and PE in the pars plana and the tips showed strong staining reaction for hexokinase activity but no staining for glycogen. Furthermore, NPE cells in the tips bore large aggregates of smooth ER and many Golgi apparati. These suggest that the high G6Pase activity in NPE cells in the pars plana and the tips is related to glucose release into the aqueous humor.