71

CELL STRUCTURE AND FUNCTION 26: 71–77 (2001)

© 2001 by Japan Society for Cell Biology

Nuclear Deviation in Hepatic Parenchymal Cells on Sinusoidal Surfaces in ArcticAnimals

Mitsuru Sato1�, Mitsutaka Miura1, Naosuke Kojima1, Nobuyo Higashi1, Katsuyuki Imai1, Takeya Sato1,Heidi L. Wold2, Jan Øivind Moskaug2, Rune Blomhoff2, Kenjiro Wake3, Nobert Roos4, Trond Berg5,Kaare R. Norum2, and Haruki Senoo1

1Department of Anatomy, Akita University School of Medicine, Akita 010-8543, Japan, 2Institute for NutritionResearch, Faculty of Medicine, University of Oslo, Oslo, Norway, 3Liver Research Unit, Minophagen Phar-maceutical Co. Ltd., Tokyo 160-0004, Japan, 4Electronmicroscopical Unit for Biological Sciences, Faculty ofMathematics and Natural Sciences, University of Oslo, Oslo, Norway, and 5Division of Molecular Cell Biol-ogy, Institute of Biology, University of Oslo, Oslo, Norway

ABSTRACT. In normal rat and human, most of the nuclei of hepatic parenchymal cells are centrally located inthe cytoplasm. However, it is reported that the nuclei of hepatic parenchymal cells are situated at a deviatedposition on sinusoidal surfaces under pathological situations such as chronic hepatitis, hepatocellular carcinoma,adenomatous hyperplasia, or regeneration. During a study on the mechanism of extreme vitamin A-accumulationin hepatic stellate cells of arctic animals including polar bears, arctic foxes, bearded seals, and glaucous gulls, wenoticed that these arctic animals displayed the nuclear deviation in hepatic parenchymal cells on sinusoidalsurfaces. In this study, we assessed the frequency of hepatic parenchymal cells showing the nuclear deviation onthe sinusoidal surfaces in arctic animals. A significantly higher frequency of the nuclear deviation in hepaticparenchymal cells was seen in polar bears (89.8±3.4%), arctic foxes (68.6±10.5%), bearded seals (63.6±8.4%), andglaucous gulls (24.2±5.8%), as compared to that of control rat liver (9.8±3.5%). However, no pathologicalabnormality such as fibrosis or necrosis was observed in hepatic parenchymal and nonparenchymal cells of arcticanimals, and there were no differences in the intralobular distribution of parenchymal cells displaying the nucleardeviation in the livers from either arctic animals and control rats. The hepatic sinusoidal littoral cells such asstellate cells or extracellular matrix components in the perisinusoidal spaces may influence the nuclear positioningand hence the polarity and intrinsic physiological function of parenchymal cells.

Key words: nuclear positioning/hepatic parenchymal cell/hepatic sinusoid/polar bear/arctic fox

Nuclear positioning seems to be a microtubule- or actin-de-pendent process (Kopecká and Gabriel, 1998; Morris, 2000;Reinsch and Karsenti, 1997; Reinsch and Gönczy, 1998;Wang et al., 1979). Various types of polarized epithelialcells have apical, lateral, and basal surfaces, and their nucleiare usually displaced to the basal surfaces. However, hepat-ic parenchymal cells have a unique polarity when comparedto other epithelial cell types, as their nuclei are centrally lo-cated in the cytoplasm (Hubbard et al., 1994). The polarityof the hepatic parenchymal cells, as other epithelial cells, is

an important modulator for proper physiological liver func-tion. In the liver tissue parenchymal cells have three distinctregions of plasma membrane: sinusoidal, lateral, and bilecanalicular (Hubbard et al., 1994). The sinusoidal or basalplasma membrane faces the perisinusoidal space of Disse.The lateral membranes are in contact with each other be-tween neighboring parenchymal cells and allow cells toshare small molecules through gap junctions. The canalicu-lar, or apical membrane is crucial for the exocrine function(bile excretion). Each of the distinct plasma membrane do-mains contains the specific integral membrane proteins es-sential for proper function. The polarity of hepatic paren-chymal cells is likely to be regulated by cell-cell and/orcell-extracellular matrix interactions accompanying thereorganization of cytoskeletons and organelles including

* To whom correspondence should be addressed: Department of Anato-my, Akita University School of Medicine, Akita 010-8543, Japan.

Tel: +81–18–884–6057 , Fax: +81–18–834–7808E-mail: mitsuru@med.akita-u.ac.jpAbbreviations: PB, phosphate buffer; HE, hematoxylin and eosin.

72

M. Sato et al.

nuclei, Golgi complexes, or mitochondria.In hepatic parenchymal cells of normal human and rat,

the nuclei are ordinarily localized to the center incytoplasm. However, it has been reported that the nuclei ofhepatic parenchymal cells exhibit the deviating position onsinusoidal surfaces in human liver accompanying chronichepatitis, hepatocellular carcinoma, adenomatous hyperpla-sia, or regeneration (Popper, 1977; Tanaka et al., 1998).We have found that arctic animals, including polar bears,arctic foxes, and bearded seals, which are situated in the up-per part of the food chain in the Arctic, display an accumu-lation of a large quantity of vitamin A as retinyl esters in he-patic stellate cells located between parenchymal cells andsinusoidal endothelial cells (Senoo et al., 1999; Senoo et al.,in press). Recently we have also noticed a high frequency ofnuclear deviation in hepatic parenchymal cells of thesearctic animals. In this study, we assessed the frequency ofparenchymal cells displaying the nuclear deviation onsinusoidal surfaces in hepatic lobules of the arctic animals.

Materials and Methods

Arctic animals

Under permission to hunt the animals by the District Governor ofSvalbard, we obtained 8 arctic foxes (Alopex lagopus), 14 beardedseals (Erignathus barbatus), and 13 glaucous gulls (Larus hyper-boreus) near Longyearbyen (78� N, 15� E) in the Svalbard archi-pelago in the period from August of 1996 to April of 1998, and an-alyzed the liver tissues from them. Three polar bears (Ursusmaritimus) were shot in self-defense at Svalbard in February andAugust of 1998 in Ny Ålesund and Hornsund. A brown bear(Ursus arctos) and Asiatic black bears (Selenarctos thibetanus)were obtained in Hokkaido and Akita, respectively, Japan.

Light microscopy and image analysis

Human liver tissues used in this study as controls were obtained byautopsy under permission. Tissue blocks (3�3�0.5 cm) of the liverfrom arctic animals, as well as normal human and rat livers as con-trols, were fixed in 3.7% (v/v) formaldehyde in 0.1 M sodiumphosphate (PB), pH 7.4, dehydrated in a graded ethanol series, par-affin-embedded, and sectioned in 3 µm thickness. The sectionswere stained with hematoxylin and eosin (HE), photographed us-ing an Olympus OMD-200 microscope (Olympus, Tokyo, Japan),and analyzed using a Macintosh computer with Adobe Photoshop5.5 and NIH Image 1.6. Nuclear deviated cells were arbitrarily de-fined when the ratio of the distance between the nuclear membraneand the close sinusoidal surface to the distance between the nuclearmembrane and the opposite sinusoidal surface was less than 0.5.The frequency of hepatic parenchymal cells showing the nucleardeviation on sinusoidal surfaces was evaluated at the specified areain the hepatic lobules, which were divided into the central, inter-mediate, and peripheral regions from the central vein according tothe classical concept of hepatic lobule structure (Fawcett, 1997)

and into zones I, II, and III according to Rappaport (Rappaport,1973). Student’s t-test was used for statistical analysis using Mi-crosoft Excel 5.0. For pathological analysis by staining of connec-tive tissue components, liver blocks were fixed in 3.7% formalde-hyde in PB, sectioned using a freezing microtome in 10 �mthickness, and then visualized by Ishii-Ishii’s silver impregnation(Ishii and Ishii, 1965), or by azan staining.

Transmission electron microscopy

Tissue blocks (3�3�2 cm) of the liver from arctic animals wereperfused with 1.5% (v/v) glutaraldehyde in 0.06 M cacodylatebuffer, pH 7.4, containing 1% (w/v) sucrose for 1 to 2 min byinjection through blood vessels that appeared on the cut surface ofthe block. The pieces (2�2�1 mm) of the fixed liver blocks werepostfixed with 2% (w/v) osmium tetroxide in PB for 2 hr, dehy-drated in a graded ethanol series, embedded in Epon 812 resin(TAAB, Berkshire, UK), and then sectioned in 70 nm thicknessusing an Ultramicrotome 2088-V (LKB, Uppsala, Sweden). Theultrathin sections were stained with 2% (w/v) uranyl acetate and0.05% (w/v) lead citrate, and then examined under a transmissionelectron microscope JEM-1200EX (JEOL, Tokyo, Japan) at an ac-celeration voltage of 80 kV.

Results

Nuclear deviation in hepatic parenchymal cells onsinusoidal surfaces in arctic animals

In normal rat liver as a control, most of the nuclei were cen-trally located in parenchymal cells (Fig. 1A), as well as innormal human liver (data not shown), whereas most of thehepatic parenchymal cells of polar bears (Fig. 1B), arcticfoxes (Fig. 1C), and bearded seals (Fig. 1D) showed thedeviated nuclei on sinusoidal surfaces. Other arctic animalsincluding glaucous gulls (Fig. 1E) displayed a modestbut higher frequency of nuclear deviation of hepaticparenchymal cells, as compared to that of control rat (Fig.1A).

No pathological abnormality such as fibrosis or necrosiswas observed after HE-staining (Fig. 1), Ishii-Ishii’s silverimpregnation (Fig. 2), or azan staining (data not shown) inhepatic parenchymal and nonparenchymal cells of arctic an-imals, although a marked increase in lipid droplets contain-ing vitamin A was obvious in hepatic stellate cells of polarbears, arctic foxes, or bearded seals, as previously described(Senoo et al., 1999; Senoo et al., in press). It has been re-ported that the nuclear deviation is associated with chronicliver diseases toward hepatic fibrosis (Tanaka et al., 1998).However, there was no increase in fibrous components ofconnective tissue after Ishii-Ishii’s silver impregnation (Fig.2) in the liver tissue from arctic animals including polarbears (Fig. 2C and D), arctic foxes (Fig. 2E and F), beardedseals (Fig. 2G and H), and glaucous gulls (Fig. 2I and J), ascompared to the rat liver (Fig. 2A and B).

Nuclear Deviation in Hepatic Parenchymal Cell

73

As the representatives of hepatic parenchymal cells withor without nuclear deviation, transmission electron micro-graphs of the hepatic parenchymal cells from normal con-trol rats and arctic foxes were shown in Fig. 3A and B, re-spectively. The results shown in Fig. 1 and 3 indicate theincrease of nuclear-deviated parenchymal cells in hepaticlobules from arctic animals examined. However, there wereno differences in the number and distribution of mitochon-dria, Golgi complexes, or peroxisomes in hepatic parenchy-mal and nonparenchymal cells of arctic animals, as com-pared to those of normal rats (Fig. 3) and human (data notshown). It was unlikely that the high frequency of nucleardeviation seen in hepatic parenchymal cells from arctic ani-mals was derived from artifact during sampling or histolog-ical staining because control rat liver displayed much lowerfrequency of nuclear deviation in hepatic parenchymal cells,whereas a brown bear or black bears obtained in Hokkaidoor Akita, Japan, respectively, showed higher frequency ofnuclear-deviated parenchymal cells than normal rats butlower frequency than polar bears and arctic foxes (data notshown); there were also substantial differences in the fre-quency of nuclear-deviated parenchymal cells among arcticanimals, for example polar bears or arctic foxes versusglaucous gulls, in spite of virtually the same processing forsampling and histology.

Estimation of parenchymal cell population displayingnuclear deviation on sinusoidal surfaces in arctic ani-mal liver

For statistical analysis of the frequency of parenchymalcells showing nuclear deviation on sinusoidal surfaces, par-affin-embedded sections were stained with HE (Fig. 4A andC), and the parenchymal cells with or without nuclear devi-ation in hepatic lobules were separately marked on photo-graphs (Fig. 4B and D) obtained under a light microscopeand then counted by using NIH Image 1.6 and MicrosoftExcel 5.0 on Macintosh computer. To estimate the frequen-cy of hepatic parenchymal cells having deviated nucleus, atleast 100 cells were surveyed at random in each lobule, andthe results were summarized and statistically analyzed (Fig.5). Extremely high frequency of nuclear deviation was seenin hepatic parenchymal cells of polar bears, arctic foxes,and bearded seals; other arctic animals such as glaucousgulls showed only modest but significantly high frequencyof nuclear-deviated, hepatic parenchymal cells (Fig. 5).Thus, arctic animals situated at upper position in food webappeared to tend to have a higher frequency of nuclear devi-ation in parenchymal cells on hepatic sinusoids.

Intralobular distribution of hepatic parenchymal cellsdisplaying nuclear deviation in arctic animals

No significant differences were observed in intralobular dis-tribution of the nuclear-deviated parenchymal cells in theclassical hepatic lobule (Fig. 4B and D) and Rappaport’s

Fig. 1. Nuclear localization in hepatic parenchymal cells on hepaticsinusoids in rats and arctic animals. Paraffin-embedded sections of livertissue from rats (A), polar bears (B), arctic foxes (C), and bearded seals (D)most of the nuclei in hepatic parenchymal cells are deviated on sinusoidalsurfaces, and only a few cells, as indicated by black arrows, display thecentrally positioned nuclei. The glaucous gull liver also show the increasein nuclear-deviated parenchymal cells (E). Scale bars, 50 �m.

74

M. Sato et al.

Fig. 2. Fibrous connective tissue components visualized by Ishii-Ishii’s silver impregnation in liver tissues from rats and arctic animals. Ishii-Ishii’s silverimpregnation was used to visualize connective tissue components in the liver tissues from normal control rats (A and B) and arctic animals including polarbears (C and D), arctic foxes (E and F), bearded seals (G and H), and glaucous gulls (I and J). The right panel (B, D, F, H, and J; scale bars, 50 �m) indicatesa higher magnification of the left panel (A, C, E, G, and I; scale bars, 100 �m, respectively).

Nuclear Deviation in Hepatic Parenchymal Cell

75

Fig. 3. Transmission electron micrograph indicating nuclear deviation in hepatic parenchymal cells on sinusoidal surface in arctic foxes. In the rat liver(A) the nucleus is centrally located, whereas in the arctic fox liver (B) the nucleus is almost in contact with the sinusoidal surface. There were no differencesin the number and distribution of mitochondria, peroxisomes, lysosomes, or Golgi complexes. N: nucleus, S: sinusoid. Scale bars, 5 �m.

Fig. 4. Intralobular distribution of parenchymal cells with or without deviated nucleus in the liver from polar bears. The parenchymal cells with or withoutdeviated nucleus on sinusoidal surfaces are marked by ‘ ’ or ‘ ’, respectively (B and D), on photographs under a light microscope of liver sections stainedwith HE (A and C) from polar bears, as well as other arctic animals and control rats. The hepatic lobules (A and C) were divided into the central, intermedi-ate, and peripheral regions, which were bordered by solid lines (B and D), from the central vein. The lower panels (C and D; scale bar, 50 �m) showed high-er magnifications of the regions indicated by rectangles on the upper panels (A and B; scale bar, 50 �m, respectively).

76

M. Sato et al.

lobule (data not shown) of either arctic animals such aspolar bears (Fig. 6) and control rats (data not shown).

DiscussionIn the liver from control rat and human, most of the nucleiwere centrally localized in parenchymal cells, whereasamong arctic animals a significantly higher frequency ofnuclear deviation in hepatic parenchymal cells on sinusoidalsurfaces was seen in polar bears, arctic foxes, bearded seals,and glaucous gulls. In particular, the high frequency of nu-clear deviation in hepatic parenchymal cells was found inthe arctic animals such as polar bears, arctic foxes, or beard-ed seals, which are situated at the upper position in the foodweb in the Arctic. It has been reported that the appearanceof hepatic parenchymal cells with deviated nuclei, whichare mainly located in zone II (intermediate region) accord-ing to the Rappaport’s classification (Rappaport, 1973), isassociated with pathological changes such as chronic hepa-titis, hepatocellular carcinoma, adenomatous hyperplasia, orregeneration (Popper, 1977; Tanaka et al., 1998). However,there were no differences in the intralobular distribution ofparenchymal cells showing nuclear deviation in the liverfrom arctic animals. No pathological signs were detectableafter staining of connective tissue components by Ishii-Ishii’s silver impregnation, as well as HE-staining and azan

staining. Analysis by transmission electron microscopy alsoindicated neither degeneration, nor apoptosis, nor necrosisin either hepatic parenchymal cells and nonparenchymalcells. There were no differences in the number and distribu-tion of organelles such as endoplasmic reticulum, mitochon-dria, Golgi complexes, or peroxisomes in either parenchy-mal and nonparenchymal cells, except for the extremelyhigh accumulation of lipid droplets containing vitamin A inhepatic stellate cells, as previously described (Senoo et al.,1999; Senoo et al., in press). These results suggest that theincrease in nuclear-deviated parenchymal cells in arctic ani-mal liver is independent of pathological situations such asliver fibrosis.

We have found that the liver of arctic animals includingpolar bears and arctic foxes accumulates 10–20 times morevitamin A in hepatic stellate cells rather than parenchymalcells, as compared to that of human liver or experimental ratliver (Senoo et al., 1999; Senoo et al., in press). It is also tobe noted that these arctic animals are highly contaminatedwith organic pollutants (Henriksen et al., 2000; Skaare etal., 2000), although no pathological signs were observed in

Fig. 5. Increase in frequency of parenchymal cells displaying deviatednuclei on sinusoidal surfaces in arctic animals. The frequency (%) of nucle-ar-deviated parenchymal cells per total parenchymal cells surveyed was es-timated by analyzing the photographs shown in Fig. 4 using Macintoshcomputer with Adobe Photoshop 5.5 and NIH Image 1.6 for hepatic lob-ules from control rats (R) and arctic animals including polar bears (PB),arctic foxes (AF), bearded seals (BS), and glaucous gulls (GG). Statistical-ly significant difference (P<0.05) indicated by ‘ ’ from the value of con-trol rats was assessed by Student’s t-test using Microsoft Excel 5.0.

Fig. 6. Intralobular distribution of parenchymal cells displaying nucleardeviation on sinusoidal surfaces in polar bears. The frequency (%) of nu-clear-deviated parenchymal cells on sinusoidal surfaces per total parenchy-mal cells surveyed was compared among the central (C), intermediate (I),and peripheral (P) regions from the central vein in hepatic lobules of polarbears, as shown in Fig. 4 as a representative. There were no significant dif-ferences in the frequency of nuclear-deviated parenchymal cells among in-tralobular regions in hepatic lobules of polar bears, as well as other arcticanimals examined and control rats (data not shown).

Nuclear Deviation in Hepatic Parenchymal Cell

77

the liver from arctic animals, as described above. Thus, it isstill not clear whether or not the highly accumulated vitaminA or organic pollutants affect the nuclear localization ofhepatic parenchymal cells; that is, it is unclear whether ornot the nuclear deviation is a biomarker for accumulation ofexcess vitamin A or organic pollutants in the liver.

Alternatively, nuclear positioning is largely dependent onmicrotubule or F-actin organization (Kopecká and Gabriel,1998; Morris, 2000; Reinsch and Karsenti, 1997; Reinschand Gönczy, 1998; Wang et al., 1979). The polarity of thehepatic parenchymal cells is important for the proper physi-ological functioning of the liver (Hubbard et al., 1994), andappears to be regulated by cell-cell and cell-extracellularmatrix interactions in association with reorganization of cy-toskeletons and organelles including nuclei, Golgi complex-es, or mitochondria. Hence, the sinusoidal littoral cells suchas stellate cells or extracellular matrices such as basementmembrane components in the perisinusoidal spaces may in-fluence the nuclear positioning in hepatic parenchymalcells. Liver tissue has the unique feature that both hepaticparenchymal cells and sinusoidal endothelial cells lack anintegrated basement membrane structure (Hubbard et al.,1994; Wack et al., 2001) as seen in other epithelial tissues,although the basement membrane structure has beenobserved to transiently form during pathological conditioninvolving regeneration and thereafter disappear (Wack etal., 2001). Immunohistochemically we used commerciallyavailable antibodies to clarify whether or not the cytoskele-ton organization or the deposition of basement membranecomponents such as laminin affects the nuclear positioningin hepatic parenchymal cells. There was neither alteration incytoskeleton organization nor deposition of laminin in thearea displaying the presence of hepatic parenchymal cellswith nuclear deviation (data not shown), although we couldnot exclude the possibility of involvement of cytoskeletalproteins and/or basement membrane components in the nu-clear deviation before obtaining antibody recognizable sucha protein component of polar animal species. Thus, the pre-cise mechanism by which the nucleus of parenchymal cellsis deviated on hepatic sinusoids in arctic animals, as well asthe effects of nuclear deviation on cell function, is still enig-matic and under investigation in this laboratory.

This is the first report showing the nuclear deviation onsinusoidal surfaces in hepatic parenchymal cells in wildanimals.

Acknowledgments. The authors are grateful to Jørn Eldar Fortun andAnders Friberg for their excellent hunting skills. This work was supportedin part by Grants-in-Aid for International Scientific Research (Joint Re-search) (08044237 and 09044251) and Grant-in-Aid for Scientific Re-search (B) (2) (11694238) from the Ministry of Education, Science, Sports,and Culture, Japan. We also thank the Norwegian Polar Institute, the Dis-

trict Governor at Svalbard, and the University Courses at Svalbard fortheir generous assistance in allowing us to use their laboratory facilities inLongyearbyen.

References

Fawcett, D.W. 1997. Liver and gallbladder. In Bloom & Fawcett: ConciseHistology (D.W. Fawcett and R.P. Jensh, eds.), Chapman & Hall, NewYork, pp. 211–220.

Henriksen, E.O., Gabrielsen, G.W., Trudeau, S., Wolkers, J., Sagerup, K.,and Skaare, J.U. 2000. Organochlorines and possible biochemical ef-fects in glaucous gulls (Larus hyperboreus) from Bjørnøya, the Barentssea. Arch. Environ. Contam. Toxicol., 38: 234–243

Hubbard, A.L., Barr, V.A., and Scott, L.I. 1994. Hepatocyte surface polar-ity. In The Liver: Biology and Pathobiology, (I.M., Arias, J.L. Boyer, N.Fausto, W.B. Jakoby, D.A. Schachter, and D.A. Shafritz, eds.), RavenPress, New York, pp.189–213.

Ishii, T. and Ishii, T. 1965. Zur Darstellung der argyrophiren Fasern. (EineModifikation der Silber-impregnationmethode von Bielschowsky-Maresch). Mikroskopie, 20: 1–11.

Kopecká, M. and Gabriel, M. 1998. The aberrant positioning of nuclei andthe microtubular cytoskeleton in Saccharomyces cerevisiae due to im-proper actin function. Microbiology, 144: 1783–1797.

Morris, N.R. 2000. Nuclear migration: from fungi to the mammalian brain.J. Cell Biol., 148: 1097–1101.

Popper, H. 1977. Pathologic aspects of cirrhosis. Am. J. Pathol., 87: 228–264.

Rappaport, A.M. 1973. The microcirculatory hepatic unit. Microvasc.Res., 119: 11–34.

Reinsch, S. and Gönczy, P. 1998. Mechanism of nuclear positioning. J.Cell Sci., 111: 2283–2295.

Reinsch, S. and Karsenti, E. 1997. Movement of nuclei along microtubulesin Xenopus egg extracts. Curr. Biol., 7: 211–214.

Senoo, H., Imai, K., Higashi, N., Wake, K., Kojima, N., Miura, M., Wold,H.L., Moskaug, J.Ø., Sato, T., Sato, M., Roos, N., Berg, T., Norum,K.R., and Blomhoff, R. 2001. Transport and hepatic storage of vitaminA in arctic animals. Cells Hepatic Sinusoid, in press.

Senoo, H., Wake, K., Imai, K., Kojima, N., Matano, Y., Miura, M., Sato,M., Roos, N., Berg, T., Norum, K.R., and Blomhoff, R. 1999. VitaminA-storing system in mammals and birds in Arctic area – A study in theSvalbard archipelago –. Cells Hepatic Sinusoid, 7: 34–35.

Skaare, J.U., Bernhoft, A., Derocher, A., Gabrielsen, G.W., Goksøyr, A.,Henriksen, E., Larsen, H.J., Lie, E., and Wiig, Ø. 2000. Organochlo-rines in top predators at Svalbard – occurrence, levels and effects. Toxi-col. Lett., 112–113: 103–109.

Tanaka, Y., Ichida, T., Nomoto, M., Matsuda, Y., and Asakura, H. 1998.Areas of sinusoidal surface hepatocyte nuclear predominance in type Cchronic hepatitis. Liver, 18: 383–390.

Wack, K.E., Ross, M.A., Zegarra, V., Sysko, L.R., Watkins, S.C., andStolz, D.B. 2001. Sinusoidal ultrastructure evaluated during the revas-cularization of regenerating rat liver. Hepatology, 33: 363–378.

Wang, E., Cross, R.K., and Choppin, P.W. 1979. Involvement of microtu-bules and 10-nm filaments in the movement and positioning of nuclei insyncytia. J. Cell Biol., 83: 320–327.

(Received for publication, January 31, 2001and in revised form, March 1, 2001)