Scanning Electron Microscopy

Volume 1986 Number 3 Part III Article 21

7-22-1986

The Three-Dimensional Microstructure of the A Review by Scanning Electron Microscopy

G. Macchiarelli University of Rome

P. M. Motta University of Rome

Follow this and additional works at: https://digitalcommons.usu.edu/electron

Part of the Life Sciences Commons

Recommended Citation Macchiarelli, G. and Motta, P. M. (1986) "The Three-Dimensional Microstructure of the Liver A Review by Scanning Electron Microscopy," Scanning Electron Microscopy: Vol. 1986 : No. 3 , Article 21. Available at: https://digitalcommons.usu.edu/electron/vol1986/iss3/21

This Article is brought to you for free and open access by the Western Dairy Center at DigitalCommons@USU. It has been accepted for inclusion in Scanning Electron Microscopy by an authorized administrator of DigitalCommons@USU. For more information, please contact [email protected]. SCANNING ELECTRON MICROSCOPY /1986/111 (Pages 1019-1038) 0586-5581/86$1.00+05 SEM Inc., AMF O'Hare (Chicago), IL 60666-0507 USA

THETHREE-DIMENSIONAL MICROSTRUCTURE OF THE LIVER A Review by Scanning Electron Microscopy G. Macchiarelli and P.M. Motta*

Department of Anatomy, University of Rome Italy

(Received for publication March 13, 1986, and in revised form July 22, 1986)

Abstract Introduction

The improvement in scanning electron micros­ The structure of the liver reflects its ex­ copy (SEM) techniques has permitted us to des­ tremely complex function. In fact, the liver can cribe the microstructure of the liver. By SEM,the be considered as both exocrine and, to some ex­ liver peritoneal surface is composed of flat me­ tent, endocrine in nature (81). In addition, it sothelial cells possessing microvilli and cilia. metabolizes, detoxifies and inactivates many dif­ Hepatic sinusoids connect the portal vessels with ferent substances of both endogenous and exo­ the terminal branches of the hepatic (cen­ genous origin ( 46). It further plays an i mpor­ tral ).Endothelial cells of the portal space tant role in the hemodynamic regulation of the arteries are elongated and arranged longitudinal­ splanchnic system. ly, while those of the central and portal veins When the first microscopists began studying are polygonal and flattened,possessing microvil­ the histological organization of the liver, it li. The sinusoidal endothelial cells show both immediately appeared clear that a visualization small fenestrations (sieve plates),up to 200 nm of its three-dimensional structure was funda­ in diameter, and large ones,up to l ~m. Within mental to unravelling its complex functional mys­ the sinusoids are seen bridging structures,cover­ teries. Several ingenious attempts to draw the ed by fenestrated ,seeming to have a liver's three-dimensional pattern have been made fibrillar core. Kupffer cells resemble macropha­ (7, 15-17). Only in the second half of this cen­ ges,showing microvilli, blebs, lamellipodia and tury, however, through careful studies using ste­ filopodia. Within the Space of Disse are seen the reological methods and statistical analysis of fat-storing cells,having laminar dendritic pro­ transmission electron microscopic (TEM) pictures jections. The polyhedral liver cell faces the (18, 126, 145), has it been possible to confront Space of Disse (vascular pole) or faces an adja­ the heterogeneous complexity of the hepatic cent (biliary pole). Vascular facets gland. The limits of these techniques have been are evenly covered by microvilli. Biliary facets recently reviewed (73) and, at the present time, show a central longitudinal depression,bordered scanning electron microscopy (SEM) seems to be by microvilli (bile hemicanaliculi). Canaliculo­ the technique of choice for three-dimensional ductular junction and epithelia show studies. Thus, after more than 15 years since the blebs, microvilli and cilia. Up to now,fetal li­ first SEM reports of mammalian liver lobule (4, ver and liver pathology have been scarcely inves­ 23), so many SEMliver images have been obtained, tigated by SEM: in the future, they can be suc­ that we feel confident in our description of the cessfully approached by three-dimensional studies. "microstructure of the liver" (73). In fact, SEM has been extensively used to obtain data from the KEY WORDS:Liver, Sinusoid, Sinusoidal Cells, of many mammalian species including hu­ Hepatocyte, Biliary System, Pathology, Fetal, mans, and other animals, such as birds and fish Scanning Electron Microscopy. (see Table 1). Most of this literature will be reviewed in this paper. Further, new SEMimages of sinusoids *Address for correspondence: from an unexplored rodent (Praomys Mastomys Nata­ P. M. Motta, Department of Anatomy, University lensis) will be presented. of Rome, via A. Borelli 50, 00161 Rome, Italy. Phone No. (39-6) 49.26.23.

1019 G. Macchiarelli and P.M. Motta

such as endothelial fenestrae (156, 157). Table 1. Primary fixation is usually performed with solutions of 1-3% glutaraldehyde in cacodylate or SOURCEOF LIVERTISSUE FOR SEMINVESTIGATIONS. phosphate buffer. Paraformaldehyde has also been (Bibliography update 1985) used, alone (4% in phosphate buffer), or with glutaraldehyde (1 .5 - 4% in phosphate buffer). No side effects have been demonstrated when the al­ Source Bibliography dehyde phase of fixation has been prolonged. This suggests that the samples may also be left in the CAT 62; 69-72; 81. primary fixative for a long period in order to CHICK 113. obtain optimal tissue fixation and hardening (5). DOG 69; 70. In a blood-rich organ such as liver, it is FISH 97. necessary to accurately wash out all blood from FROG 111. the vascular bed prior to fixation (81,116). In HUMANS 27; 36-40; 44; 62; 76; 81; 89; 91; this way, precipitation of plasma proteins which 134; 136-138; 153; 160; 161. obscure vessel surfaces, can be avoided (27). PIG 62; 70; 71. Several washing solutions have been used but no PRIMATES data can be referred as to which is the solution Baboon 57. of choice. Monkey 87; 108; 133; 135; 139. Both the addition of anticoagulants (hepa­ RODENTS rin) or vasodilatants (procaine), to infusates, Guinea-Pig 35; 62; 69-71. and the injection of either during anesthesia im­ Hamster 59; 160. proves the quality of perfusion, which can be Mouse 27; 69-71; 92; 121; 130; 141. monitored by noting the change in liver colour Rabbit 27; 105; 111; 125; 158. from a dark-red to a sandy-beige (116). Rat 8; 9; 11; 14; 21; 22; 26; 27; 30; Some authors suggest the maintenance of in­ 34; 41; 43; 45; 51; 54; 55; 62; 65; fusates at body temperature in order to mimic the 67-72; 74-79; 81; 86; 88; 90; 93; physiological environment and to enhance the 94; 99; 101; 112; 114; 115; 123; speed of fixative penetration (5). On the other 130-132; 142; 144; 149; 151-154; hand, the use of lower temperatures may reduce 158; 160; 162. anoxic damage. SHEEP 157. Attention must be paid to the infusion pres­ FETAL sure (153). In fact, the use of vasodilatant animals 71; 72; 76; 113; 157. drugs and aldehyde fixation through the arterial humans 58; 71; 76. circulation, has the effect of converting the blood vessels into a rigid pipe system, causing loss of the arteriolar "barrage" (22). Thus, the Preparation of Liver Samples for SEM infusates may reach the postarteriolar vessels (sinusoids) with a pressure higher than physiolo­ The improvement in electron microscopic gical values, which range from 6 to 10 mmHgin techniques during the past 15 years has permitted the mammalian portal system. In addition, there us to obtain optimal liver specimens for SEM is evidence that a high infusion pressure of in­ investigation with relatively simple procedures. fusates, whether injected via the portal vein or In order to avoid artifacts and to yield a clean via the arterial system, is associated with an surface for observation, however, it seems increase in the number of larger gaps in the si­ imperative to respect certain crucial points dur­ nusoidal endothelium (21). Considering certain ing the preparation of samples (27, 81). properties of endothelium, such as filtering and In order to obtain good preservation of hep­ sieving, these larger gaps have been also des­ atic subcellular structures, fixation must be cribed as artifacts (155) or pathological featu­ performed by vascular perfusion (148). In fact, res ( 21). immersion primary fixation alone does not ade­ In animal experimentation, perfusion pre­ quately avoid cellular anoxic damage, does not ferably is performed via the portal vein, but may preserve endothelium (150) and does not display a also be done via the aorta, or directly via the homogeneous fine structure in the whole sample left ventricle. Biopsy samples also have been (119).Furthermore, immersion fixation may create perfused, through the introduction of a small spatial artifacts due to the non-homogeneous needle into the parenchyma (puncture-perfusion) hardening of the liver tissue and it may also (38,40,86,134,137) or through cannulation of the cause the disappearance of delicate structures cut vessels (58,76,81).

1020 SEMof the Liver Osmication is not always required for SEM. useful to obtain stereo pictures during SEM In fact, it doesn't seem to offer any advantage observation, to allow a better understanding of during observation of liver samples (133). How­ the three-dimensional relationships of liver ever, the metal-impregnation technique with tan­ cells (73). nin-osmium may enhance the conductivity under the electron beam of non-coated specimens (84,85). The Peritoneal Liver Surface Dehydration can be carried out gradually, either with ethyl alcohol or acetone (81), fol­ The external surface of the liver is almost lowed by critical-point drying (see Howard and completely covered by a monolayer of serosal Postek, 1979 for further references) (28). cells that rest with the basal lamina on a vari­ Freeze-drying also has been performed, and is ad­ able amount of connective-tissue fibres, forming vantageous in exposing more of the surface area the fibrous capsule (73). (8), but it seems inferior to critical-point dry­ The fine granules and filamentous struc­ ing in high-magnification SEM(98). tures, intermingled with fibroblasts and colla­ Samples are usually coated with gold or genous fibrils of the fibrous capsule, can be gold-palladium; in our experience, carbon-gold easily recognized by means of SEM in chemically also gives good results (74). dissociated tissue (31) or through breaks of the The surface tissue exposure is of particular peritoneal layer. Further, in vascular corrosion importance. Several techniques have been sug­ casts, Glisson's capsule, which follows the hepa­ gested. Methods such as mechanical tissue dis­ tic vessels into the liver parenchyma, has a poor sociation by digital pressure (performed in dried network supplied by arterioles from the samples) (55,99,157), tissue dissociation with branches of the hepatic arteries and emptying in­ jewelry forceps after cutting with a sharp razor to the sinusoids 109). blade (both in wet or dried samples) (81), It is demonstrated by SEMthat the perito­ freeze-fracture (in alcohol-infiltrated or crit­ neal sheath which is absent only within the area ical-point dried samples) with freon or liquid of the liver-diaphragmatic attachment (46), is nitrogen (8,29,30,81,128), all have valuable ap­ flat, and is seen as such in samples fixed by the plications in delineating different structures. perfusion technique (Fig. 1). But this sheath ap­ Thus, internal hepatocytic organelles can be well pears corrugated when the samples are directly visualized in freeze-fractured dried samples (81) placed in the fixative (Fig. 2). This is probably and bile hemicanaliculi are exposed after simple due to a reactive contraction of the underlying cutting or fragmentation of wet tissues (75). cells, such as smooth muscle cells. Isolated (from mechanical-dissociated Serosal cells, which, when examined by con­ dried samples) can be collected on a double-face ventional electron microscopy, appear coupled to­ tape placed on a stub (81). Hepatocytes can also gether by junctional complexes (73) and covered be studied after isolation by elutriation and by a highly positive ruthenium-red coat (glycoca­ culture (114,142). In addition, intracellular lyx) (56,68), are flattened and polygonally hepatocytic components have been exposed by shaped, often populated by abundant microvilli of cracking of resin-infiltrated samples (125), by 1-2 ~min length when examined by SEM(Fig. l) fracturing of tannin-osmium treated samples (36, (81). Observed at high magnification, microvilli 39), by chemical dissociation with boric acid show granules and filamentous structures (Fig. 3) (130), or by the osmium-DMSO-distilled water me­ which are thought to correspond to material de­ thod (31). The hepatocyte cytoskeleton has been rived from glycocalyx or serosal exudate (68). recently studied after treatment with Triton-X The pits and the blebs that are often detected on l 00 ( 112). the serosal surface may be considered as signs of Liver vasculature can also be studied well the surf ace activity of the serosa l ce 11s. In by vascular corrosion cast methods (for details fact, mesothelial cells are involved in the pro­ about this technique see Ohtani et al, 1983 and cesses of secretion, absorption and exchange of Lametschwandtner et al., 1984) (53,111) that vis­ fluids which may act as surface lubricants (1,2, ualize an exact reproduction of vessel disposi­ 73). tion in space (3,83,160). In addition, the outer Furthermore, the serosal cells covering the sinusoidal wall and adjacent intercellular spaces liver, like those seen in other organs, possess a may be investigated in resin-cast non-corroded single cilium of variable length (Fig. 4). The tissue specimens (12). cilium might represent some rudimentary struc­ Always, all of these methods should be used ture, or it might have some chemoreceptorial or in preparing liver samples, because they comple­ motile function (80,129). In any case, the role ment each other. of these cilia is still to be clarified. Finally, we believe that it is extremely

1021 G. Macchiarelli and P.M. Motta Liver Architecture and Vasculature The branches of the hepatic artery show a thick wall formed by several layers of smooth The Lobules and Acini. muscle cells . Arterial endothelium is arranged Modern SEMtechniques permit us to fully ap­ in longitudinal columns running along the major preciate the complex microarchitecture of the li­ axis of the vessels. Endothelial cells are elon­ ver gland (73). The liver is supplied by an ex­ gated, with a slight central (nuclear) pro­ traordinary number of nutritive and functional trusion. Their surface is smooth, but, occasion­ vessels (hepatic arteries and portal veins). Such ally, microvillous projections and blebs can be vessels contribute to create a three-dimensional detected. Endothelial cell margins easily can be labyrinthic structure (72), basically character­ recognized, and sometimes there is overlapping of ized by the morpho-functional tissue unit (117) the borders of adjacent cells (Fig. 7). called the lobule or "acinus", as Wepfer (146) The portal veins, as a rule, are larger than and Malpighi (61) first supposed. As recently re­ arterial vessels, but their walls are thinner, viewed (73), the three tissue units of the hepa­ presenting one or two layers of smooth muscle tic architecture represent different ways of cells. Venous endothelial cells are polygonally looking at the same structure (46): the classic shaped or elongated. On their flat surface micro­ lobule of Kiernan (50), the portal lobule of Mall villous projections and pits are often observed. (60) and the liver acinus (118) emphasize pecu­ Occasionally, outlets of smaller vessels and si­ liar aspects revealed in different mammalian nusoids can be seen (Fig. 8). species according to the morphophysiological By means of vascular corrosion cast methods, condition considered. The smallest hepatic func­ SEMhas revealed a peribiliary capillary plexus tional unit seems to be the "acinus hepaticus" supplied by the branches of the hepatic arteries (117), and it offers explanations for many histo­ of the portal space (87,108). These vessels ana­ pathological liver changes (46). The classic stomose sinusoidal channels or portal veins (96, lobule that can be called, in some instances, an 109). These two kinds of efferent branches of the endocrine lobule (if the direction of the blood plexus have been described in human and rabbits flow is considered) and the portal lobule that with the same frequency. On the other hand, the can be called an exocrine lobule (when one consi­ peribiliary plexus is drained mostly by portal ders the direction of the bile flow) (81) are veins in the rat, and by sinusoids in the monkey more easily recognized in tissue preparations by (87,109). This type of "portal" vascular system, light or electron microscopy (72). considering its close connection with the bile When fractured liver samples from properly ducts, could play a role in the intralobular perfused animal livers are examined by SEM, hepa­ feedback control of bile production (81,87,105, tic lobules can be identified as irregular poly­ 110). hedral structures (Fig. 5) with a centrally lo­ Terminal Hepatic Veins. cated terminal hepatic vein or portal triad, Terminal hepatic veins characteristically which are surrounded by the hepatic laminae (26, may be observed, in cross sections, at the center 45,49,81,92, 139). of converging hepatic laminae of the classic Different aspects can be noted in the ar­ lobule (central veins) (Fig. 9). They present a rangement of the hepatic laminae. This seems to wall mostly composed of the intimal lining, rest­ be correlated with the changes in the blood­ ing on a very thin muscular layer. The vascular pressure gradient along the ramifications of the lumen shows numerous sinusoidal openings often hepatic veins producing a plastic deformation of crossed by bridging structures (151). These brid­ the liver laminae which may temporarily alter ges, sometimes seen trapping blood cells (Fig. their location in space. Thus, the observer may 10), are formed by a collagen core covered by have the impression that the hepatic laminae ra­ fenestrated endothelium (151). The fenestrated diate from the portal triad (portal lobule or endothelium even may be recognized around the si­ acinus) or that they converge toward a central nusoidal outlets for a small area (Fig. 10). En­ vein (classic lobule) (81). dothelial cells have a polygonal shape. The Vessels within the Portal Space. rounded nucleus is clearly seen to bulge in the Arterial and venous vessels within the por­ central area of the cells, whose borders are· ea­ tal space present remarkable differences in their sily detectable (Fig. ll). Leukocytes and Kupffer wall structure and endothelial surfaces (73). cells are frequently adherent to venous endothe­ In cross sections vascular components appear lium (Fig. 11). The different hemodynamic and close to the lymphatic vessels and the bile duct, structural patterns of these vessels (high pres­ surrounded by connective tissue fibres that also sure and thick wall in arteries; low pressure and delimit spaces for the interstitial fluids (Fig. thin wall in veins) may be responsible for 6) ( 26,81). the above-mentioned ultrastructural endothelial

1022 SEMof the Liver

~ The serosal cells covering the surface of ~ An isolated cilium (arrow) arising from the liver are populated by numerous microvilli(m) the serosa l surface of the liver. (m, microvi 11i) Arrows delineate cell borders (Bar= 10 Mm). Rat. ( Bar = l ).Jm). Cat. ~ Liver surface appears corrugated in sam­ ~ Mammalianportal lobule.(P, portal space; ples fixed by immersion (Bar = 10).Jm). Humming­ V, terminal hepatic vein; HL, hepatic laminae) bird. (Bar = l 00 ,um). Rat . ~ High magnification reveals granules {g) ~ Portal area. (V, branch of portal vein; adhering to microvilli (m) of mesothelial cells a, artery; b, bile duct; ly, lymphatics) (Bar = (Bar= 1 µm). Mouse. 10 µm). Rat.

1023 G. Macchiarelli and P.M. Motta differences. The deforming transmural pressure Kupffer cells, whose long cellular projections indeed may stimulate vessels with a highly reac­ can also be seen penetrating and enlarging the tive musculature (such as arteries) that respond endothelial fenestrations (70,81). Endothelial with a functional contraction, creating the lon­ dynamism also can be proven by our recent SEMob­ gitudinal folding of the arterial endothelium. servations on liver of the rodent, Praomys Masto­ Sinusoids, Sinusoidal Cells and Space of Disse. mys Natalensis. In this animal, intrasinusoidal As seen by SEM, in fractured samples, as bridges arising from endothelium are often noted well as in corroded vascular casts, the liver (Fig. 13). By SEM, these bridges (up to 4 µm in "sinusoidal" network anastomoses the portal ves­ length) appear to be covered by fenestrated endo­ sels with the terminal branches of hepatic veins thelium, and they probably possess a collagen (26,48,68,73,106, 107). core (Fig. 14). Considering their intraluminal A flat fenestrated endothelium, forming the location, their possible role may be related to sinusoidal wall, separates the capillary lumen intrasinusoidal regulation of the blood flow, a from the perivascular interstitial spaces (peri­ role similar to that which the long cytoplasmic sinusoidal space of Disse) (Fig. 12) (72,88, projections of the Kupffer cells are thought to 148,). Sinusoidal cells are the cellular ele­ p 1ay ( 73). ments populating the sinusoids and the Space of The three-dimensional aspect and the loca­ Disse (6,73,103,104). tion of the Kupffer cells are best revealed in Three principal sinusoidal cell-types have stereo SEM pictures (Fig. 15) (70,81). These been morphologically identified by SEM-TEMstud­ cells have an irregular cell body, which, depend­ ies: the endothelial cell, the and ing upon the activation of the cell, appears to the perisinusoidal stellate cell. The latter is be provided with a variable number of blebs, mi­ also called "Ito cell" or "fat-storing cell" (32, crovilli, holes, lamellipodia and filopodia (26, 140). Others types of perisinusoidal cells have 35,51,69,70,79,89) that are characteristic fea­ been reported, such as "pit cells" (6,47, 149) and tures of macrophages (70,121). Kupffer cells, "pericytes" (97), but through employment of SEM, through their cell bodies and/or cytoplasmic pro­ it is difficult to adequately detect their cesses, are in contact with endothelial cells surface features. (Fig. 16) (42,69,124), with which they may form In perfusion-fixed samples, endothelial junctional complexes (33). Blood cells are also cells are flattened. Their only protrusion is frequently seen associated with the Kupffer cells seen over the nuclear region. Sinusoidal endothe­ (Fig. 15), possibly as an expression of the immu­ lial cells are of an extremely variable size, and nological role of liver macrophages . Further, they possess several cytoplasmic openings evenly Kupffer cell projections may, in some instances, distributed over their extensive cytoplasmic create a kind of intrasinusoidal micro-labyrinth processes (Fig. 12) (71,78,79). These openings, that may play a role in the regulation of blood or "fenestrae", permit a communication between fl ow ( 70). Disse's space and the sinusoidal lumen. Ito cells, fibroblasts and fibrocytes are Dimensions and distribution of fenestrae have the elements populating the space of Disse. These been extensively studied (26,27,35,66,79,89,127, cells, which easily can be differentiated in TEM 132,152-155,157,158). Two principal types of sections, are intermingled within a delicate net­ openings are usually described (35,73): small work of fibrils and collagen fibers supporting fenestrae (100-200 nm), often clustered, forming the endothelial lining (73). The fat-storing the so-called "sieve plate" (148); and large fe­ cells (perisinusoidal stellate cells or Ito nestrae (up to 1 µm) (Fig. 12).Larger gaps gener­ cells) may be considered as vitamin A-storing li­ ally are considered as artifacts or as a result pocytes (140). In SEM they can be identified of some toxic agent (21,22,57,81,94,150). This through the larger gaps of sinusoidal endothe­ is supported by the increased number of large fe­ lium, possessing numerous laminar dendritic pro­ nestrations under particular experimental and/or cesses, and rarely, microvillous projections pathological conditions (21,22,72,79). On the (Fig. 17) (34,44,81,136). Occasionally, a single other hand, endothelium should not be considered cilium floating in Disse's space, or emerging as a rigid structure, but as a dynamic one (67, through a sinusoidal gap into the vascular lumen, 81,139). Its morphology changes from organ to can be recognized (81,104). The function of organ (120) and even within the same organ, the fat-storing cells appears related to the adapting its ultrastructural features to several metabolism and storage of vitamin A, as document­ different pathological and physiological condi­ ed by the large increase of these elements in Vi­ tions. According to such considerations, larger tamin A-treated animals (30, 52,140), and also by fenestrae may be an expression of temporary in­ the features they have in commonwith similar juries that may involve both endothelial and stellate vitamin A-storing cells found in other

7024 SEMof the Liver

~ Endothelium of the hepatic artery in the endothelial fenestrations (F) can be recognized. portal space. (E, endothelial cell; arrows, cell (*,bridge; B, red blood cell). (Bar= 10 µm).Rat. limits) (Bar= 10 ,l(m). Rat. Fig.11 Endothelium of the terminal hepatic ~ Endothelium of the portal vein of the vein.(N, nucleus; E, endothelial cell; L, lympho­ portal space. (E, endothelial cell; S, sinusoidal cyte; arrow, cell limits; S, small sinusoidal opening) (Bar= 10 µm). Rat. opening). (Bar = 10 ,(Jm). Rat. ~ Central vein (V) surrounded by hepatic ~ Fenestrated endothelium.Sieve plates (SV) laminae (HL) (Bar = 10 jjm). Rat. and large fenestrae (G) of sinusoidal endothelium Fig. 10 A sinusoidal opening in the central vein, (*,microvilli of the hepatocyte).(Bar= l ,um).Rat.

1025 G. Macchiarelli and P.M. Motta organs (73,140}. These data demonstrate that si­ microvilli (Fig.20). Bile canaliculi may possess nusoidal cells should be considered morpholo­ branches and lateral sacculations (75). In ste­ gically as three distinct cell types whose re­ reo views, the canalicular bed appears to be pro­ spective roles seem to be somewhat integrated. In vided with numerous microvilli, large holes and fact, sinusoidal cells have an intimate spatial small diverticulae that may be the intracellular relationship, and this seems to further represent sacculations observed in TEMsections (73,75). coordination of certain activities carried out Both surfaces at the canalicular sides are within the sinusoidal spaces, including filtra­ relatively smooth. Such smooth bands (0.1-0.4 ~m tion and discernment of blood substances. Thus, width) are the areas of adjacent hepatocytes' the mechanical sieving activity of the en­ junctional attachments; the latter represent the dothelial sieve plates (148) may be enhanced by anatomical barrier between vascular and biliary the anatomical barrier created within the sinu­ compartments. It is believed that where these soids by the Kupffer cell prolongations. In this barriers are very narrow, bile regurgitation may way the Kupffer cells contact blood elements and occur under experimental or pathological condi­ substances, exerting a phagocytic, and/or immuno­ tions (73,90). Protrusions and holes also can be logical role. Such functional complementarity of distinguished on these bands. The former to­ sinusoidal cells also may be suggested by expe­ gether may serve as an additional system of at­ rimental studies involving treatment of choles­ tachment for adjacent hepatocytes, as well as for terol remnants that reach hepatic sinusoids. In­ sealing bile canaliculi (73,81). deed, endothelial cells and Kupffer cells may Bile canaliculi become wider (1-2.5 ,um) near work together in recycling or removing choles­ the portal zones, where they may be seen emptying terol remnants that, when too large, cannot pass into the intralobular ductules of Hering (62,38). through the endothelial fenestrae (14,20,21,158) Two kinds of connections of bile canaliculi with or are phagocytized by Kupffer cells (147). bile ductules (canalicular-ductular junctions) are described (41). Bile canaliculi usually emp­ Liver Parenchymal Cells and lntrahepatic Biliary ty into bile ductules, forming an ampulla just Tree. before the junction. Sometimes, bile canaliculi may also be connected directly with the bile duc­ The hepatic parenchymal cell is able to car­ tules without any sacculation or dilatation. The ry on its various functional activities simul­ ductule's wall is made of a few epithelial cells taneously due to the close spatial relationship joined to converging hepatocytes (37,62,81). that this cell shares with both the vascular com­ Numerous microvilli and, rarely, long cilia ari­ partment and biliary system. Such morphological sing from epithelial cells, are also described relations can be successfully studied by SEM (27,74,). (73). In fact, SEMclearly demonstrated that the Ductules empty bile directly into the bili­ physiological polarity of the hepatocytes cor­ ary ducts located in the portal space (Figs. 5- responds to an anatomical polarity (46,81). By 6). The epithelial cells forming the bile duct SEM, the nepatocyte appears as a polyhedral cell wall possess microvilli and cilia (37). Micro­ with 6 or more facets, possessing vascular poles villi are characteristically disposed in longitu­ facing the perivascular, interstitial spaces and dinal rows (62,81) that, in the larger ducts, biliary poles delimiting the bile canaliculi seem to delimit hexagonal profiles corresponding (Fig. 18) (75,82). Hepatocytic facets may have to the cell borders (Fig. 21). Almost all epithe­ various morphological appearances in relation to lial cells of the larger ducts show a centrally their different functional polarity (19). The located cilium on their surface (Fig. 21). Lumi­ vascular facets, observable in isolated hepato­ nal holes (0.1 µm diameter) are present in the cytes or through the larger sinusoidal gaps, are smallest bile ducts. These openings represent richly covered by short microvilli (Fig. 19) (77) the inlets of bile canaliculi that occasionally which considerably increase the membrane surface may empty directly into biliary ducts without area destined for the absorption and treatment of ductular passage (62). the interstitial fluids (73). Biliary facets, Intracellular components of the liver paren­ having a different functional polarity, have dif­ chymal cells are poorly studied by SEM(36,81, ferent features. In properly fractured samples, 112,125). In fractured hepatocytes, the nucleus biliary facets present a centrally located longi­ appears spherical and centrally located (81) tudinal depression (width: 0.5 µmin the central (Fig. 22). Nuclear pores have been demonstrated areas of the lobule; up to 2.5 µm, near the peri­ in chemically dissociated tissues (36). When ex­ phereal area} (73) that runs along the entire amined three-dimensionally, other cytoplasmic or­ cell surface.(75,135). These channels (bile hemi­ ganelles and inclusions appear evenly dispersed canaliculi} are bordered by short, thick throughout a network of fibrous tubular

1026 SEMof the Liver

Fig. 13 An endothelial bridge (B) covered by fenestrated endothelium,within a sinusoid.(*,space 0f Dis­ se; H,hepatocyte; bc,. Fig. 14 A bridge where endothelium (El has been removed.Note the fibrillar composition of this structure.(S,sinusoid; H,hepatocyte). (Bars= l ~m). Praomys.

Fig.15 A stereo pair show­ ing the intimate rela­ tionship between a Kupffer cell and a lymphocyte. (Bar= l µm). Rat.

Fig. 16 Within a sinusoid is present a Kupffer cell (K) having a long cytoplasmic extension (arrow). (be, bile canaliculus). Fig. 17 Through an artifactual gap of the sinusoidal endothelium it is possible to see the laminar dendritic processes of a fat-storing cell (Fl located in the space of Disse. (SV, sieve plate; m, microvilli of the hepatocyte). (Bars= l µm). Rat.

10Z7 G. Macchiarelli and P.M. Motta components (in cracked resin-infiltrated samples) and the space of Disse have been revealed by SEM (125) or filamentous structures (as recently evi­ (75). These areas, which correspond to the nar­ denced in liver perfused with Triton-XlOO) (112). rowest smooth bands present at the canalicular In any case, identification and classification of sides, may be considered "loci minoris resisten­ intracellular structures and organelles, along tiae" for their supposed tendency to rupture and with their three-dimensional relationships, will leak bile in response to increases in intrabil­ be better clarified in SEM-TEM(freeze-etching) iary pressure (73,90,122). However, SEMdata on correlated studies. the occurrence of free communications between perivascular spaces and bile canaliculi are Future Applications of SEMin Hepatology contradictory (11,63,90), so further studies must be done to clarify this problem. SEMalso has The actual three-dimensional arrangement of been successfully employed in studies of hepatic the liver's cells, as well as of the cells of ma­ vascular diseases (138), including vascular tu­ ny other organs, has been revealed by SEM(24,49, mors (59,137,159). In addition, parenchymal di­ 80,81). Normal liver structure easily can be sorders, such as hepatic necrosis, cirrhosis and studied by comparing SEM data with data col­ tumors, both in animals (43,81,93,99-102,115,l41, lected through employment of other morphological 143,144,) and in humans (39,91,137,161 ), have techniques, including morphometrical analysis been investigated three-dimensionally. All of (73). The achievement of this primary step will these reports clearly have shown that SEMis a be extremely useful for the development of other helpful technique for the elucidation of the pa­ biological fields. Unfortunately, hepatic physio­ thogenesis of hepatic alterations (73), as well pathology and organogenesis are important areas as for the collection of data for differential of biological research that are still scarcely diagnosis in human pathology (95,137), especially investigated by means of SEM. when compared with other morphological and immu­ Morphological studies, whether in experimen­ no-histochemical studies (10,64). tal animal models or in human pathology, are one The few SEMpapers concerning liver develop­ of the interesting topics in biological investi­ ment that are presently available, lend support gations, today. SEM may offer valuable informa­ to the assertion that the study of the dynamic tion toward a better understanding of many evolution of cell and tissue during hepatogenesis physio-pathological processes occurring in all may be approached through a three-dimensional e­ organs (10), especially in liver. Up to the pre­ valuation (58,76,113). Nevertheless, various de­ sent, however, the liver has not been extensively velopmental stages of the liver, and hemopoiesis, studied by SEM pathologists, hence, very few pa­ have been extensively investigated by TEM(25, pers describing human pathological samples are 124,163). The architecture of the fetal liver is available. characterized by the presence of extensive vascu­ Hepatobiliary lesions produced by bile duct lar areas and by a greater number of sinusoids ligation (9, 11,13,81,90,131) or treatment with than is found in the adult liver (58,76), allow­ cholestatic agents (54,55,65,162) in animals, as ing increased blood filtration and absorption well as those biopsied in humans (137), have been (113). Sinusoidal endothelial cells possess nu­ studied. In intra-extra-hepatic cholestasis, SEM merous gaps, and they seem morphologically dif­ easily revealed diffuse dilatation of bile cana­ ferent from Kupffer cells, even in early develop­ liculi that had lost their bordering microvilli mental stages (71). Hepatoblasts, which may be and emitted numerous ramifications (81,90). In classified best through correlated SEM-TEMstud­ addition, increased microvillosity of newly ies, show surfaces extensively covered by short formed side-branches of the bile canaliculi has microvilli (71). These microvilli, which increase been observed in humans (137). Diffuse prolifer­ as the liver matures (113), are sometimes seen to ation of the intrahepatic biliary tree, enlarged be arranged in longitudinal parallel rows. They canaliculo-ductular junctions and numerous endo­ are probably related to the formation of the bile exocytotic formations of the hepatocytic surfaces canaliculi (73). also have been reported (11). These changes may As demonstrated in this review, the "world help to explain the mechanism of bile regurgita­ of the liver" offers many other suitable topics tion into the vascular compartment, i.e., jaun­ for future three-dimensional studies. The exact dice formation. This would seem to suggest that function of endothelial fenestrae and their chan­ bile regurgitation is related to increased duc­ ges during pathological processes, the morpho­ tular reabsorption of bile and trans-hepatocytic functional relationships of endothelium with transport, leading to release of bile into the other sinusoidal cells, the origin and the trans­ space of Disse (11). On the other hand, areas formation of hepatic cells during the fetal with minimal distance between the bile canaliculi period or during regeneration and, finally, the

1028 SEMof the Liver three-dimensional microstructure of liver patho­ logical processes, a subject which is virtually unexplored and unresolved, are all fascinating problems. Indeed, we have demonstrated that many of the liver cell 's components can be easily well identified only by means of SEMalone. We therefore conclude that SEMinvestiga­ tions will surely help to elucidate many of the problems regarding physiological and pathological processes occurring in the liver (27,73). It is our hope, then, that scanning electron micros­ copists will continue in their efforts to unravel the mysteries entwined within this complex organ, especially so that clinicians and surgeons might find new approaches to treatment of the several hepatic disorders still known to be incurable.

Fig. 18 Hepatic laminae. (H, hepatocytes; S, si­ arrows show holes which are part of the junctio­ nusoids; be, bile canaliculi).(Bar = 10 µm). Rat. nal complex (*, Space of Disse; S, sinusoid). Fig. 19 The vascular pole of a hepatocyte toward (Bar= l µm). Guinea Pig. the space of Disse is seen through a large arte­ Fig. 21 Epithelial surface of a bile duct of the factual endothelial rupture. (H, hepatocyte; m, portal area. The arrows evidence cilia. (m, mi­ microvilli; E, endothelium). (Bar= l µm). Rat. crovilli). (Bar= 1 µm). Rat. Fig. 20 Bile facet of a hepatocyte showing the Fig. 22 Fractured hepatocytes. (N, nucleus; C, bile groove (be) bordered by microvilli (m). The cytoplasm; S, sinusoid). (Bar = l µm). Rat.

1029 G. Macchiarelli and P.M. Motta References 13. CompagnoJ,Grisham JW.(1974). Scanning elec­ tron microscopy of extrahepatic biliary obstruc­ 1. Andrews PM, Porter KR.(1973).The ultrastruc­ tion. Arch Pathol 97:348-351. tural morphology and possible functional signi­ ficance of mesothelial microvilli. Anat Rec 177: 14. De Zanger RB, Wisse E.(1982). The filtration 409-426. effect of rat liver fenestrated sinusoidal endo­ thelium on the passage of (remnant) chylomicrons 2. Barberini F, Carpino F, Renda T, Motta PM. to the space of Disse,in: Sinusoidal Liver Cells, (1977). Etude au microscope electronique a bala­ DL Knook,E Wisse (eds),Elsevier/Amsterdam, 69-76. yage du peritoine du rat. Anat Anz 142:486-496. 15. Disse J. (1890). Ueber die lymphbahnen der 3. Bernatzky G, Lametschwandtner A. (1979). The saugethierleber. Arch Mikrosk Anat 36:203-224. angioarchitecture of some regions of the liver of the Atlantic hagfish, Myxine Glutinosa and the 16. Elias H. (1949). A re-examination of the toad, Bufo Bufo. A SEMstudy of vascular cor­ structure of the mammalian liver. I. Parenchymal rosion cast. Ber Nat Ver Med Salzburg 3/4:45-49. architecture. AmJ Anat 84:311-334.

4. Bierring F, Skaaring P.(1973).Scanning elec­ 17. Elias H. (1949). A re-examination of the tron microscopy of liver. JEOL News lle:16-17. structure of the mammalian liver. II. The hepatic lobule and its relation to the vascular and bi­ 5. Boyde A.(1972). Biological specimen prepara­ liary system. AmJ Anat ~: 379-465. tion for the scanning electron microscope. An overview. Scanning Electron Microsc.1972: 257-264. 18. Elias H, Sherrick JC. (1969). Morphology of the Liver, Academic Press, New York, 12-22. 6. Bradfield JWB.(1984). Liver sinusoidal cel­ ls. J Pathol 142:5-6. 19. Evans WH.(1980). A biochemical dissection of the functional polarity of the plasma membrane 7. Braus E, Vierling H. (1930). In: Atlas der of the hepatocyte. Biochimica et Biophysica Acta Normalen Mikroskopischen Anatomie des Menschen, 604:27-64. M Clara, K Herschel, H Ferner H, (eds), JA Barth, Leipzig, (1974), 130. 20. Fraser R, Bosanquet AG, Day WA. (1978). Fil­ tration of chylomicrons by the liver may influen­ 8. Brooks SEH, Haggis GH.(1973).Scanning elec­ ce cholesterol metabolism and atherosclerosis. tron microscopy of rat's liver. Applicaton of Atherosclerosis 29:113-123. freeze-fracture and freeze-drying techniques. Lab Invest 29:60-64. 21. Fraser R, Bowler LM, Day WA, Dobbs B, John­ son HD, Lee D.(1980). High perfusion pressure da­ 9. Brooks SEH, Reynolds P, Audretsch JJ Haggis mages the sieving ability of sinusoidal endothe­ G. (1975). Scanning electron microscopy of proli­ lium in rat livers. Br J exp Path ~:222-228. ferating bile ductules. Lab Invest 33:311-315. 22. Frenzel H, Kremer B, Hucker H.(1977).The li­ 10. Buss H, Hellweg G. (1980). Application of ver sinusoids under various pathological condi­ scanning electron microscopy to diagnostic patho­ tions. A TEM and SEM study of rat liver after logy. A critical review. Scanning Electron Mi­ respiratory hypoxia, telecobalt-irradiation and crosc. 1980;111:139-153. endotoxin application,in: Kupffer and Other Liver Sinusoidal Cells, E Wisse, DL Knock (eds), 1977 11. Carpino F, Gaudio E, Marinozzi G, Melis M, Elsevier/North-Holland Biomedical Press, Amster­ Motta PM.(1981).A scanning and transmission elec­ dam, 213-222. tron microscopic study of experimental extra­ hepatic cholestasis in the rat. J Submicrosc 23. Fujita T, Tokunaga J, Inoue H. (1971). Atlas Cytol _ll:581-598. of Scanning Electron Microscopy in Medicine, Iga­ ku-Shoin Ltd., Tokyo, 20-21. 12. Castenholz A.(1983). The outer surface mor­ phology of blood vessels as revealed in scanning 24. Fujita T, Tanaka T, Tokunaga J. (1981).Scan­ electron microscopy in resin cast, non-corroded ning Electron Microscopic Atlas of Cells and Tis­ tissue specimens. Scanning Electron Microsc. sues, Igaku-Shoin Ltd., Tokyo, 144-157. 1983;IV:1955-1962.

1030 SEMof the Liver

25. Fukuda T.(1974).Fetal hemopoiesis. II. Elec­ electron microscopy. Arch histol jap lZ_:15-24. tron microscopic studies on human hepatic hemopo­ iesis. Virchows Arch B (Cell Pathol) 16:249-270. 36. Itoshima T, Shimada Y, Hayashi N, Murakami T (1976). A Scanning electron microscopy of subcel­ 26. Grisham JW, Nopanitaya W, CompagnoJ, Nagel lular structures of the human hepatic cell. Arch AEH.(1975).Scanning electron microscopy of normal histol jap ~:15-21. rat liver: the surface structure of its cells and tissue components. AmJ Anat ~:295-322. 37. Itoshima T, Yoshino K, YamamotoK, Ohata W, Kubota M, Ukida M, Ito T, Hirakawa H, MunetomoF, 27. Grisham JW, Nopanitaya W, CompagnoJ.(1976). Shimada Y.(1977). Scanning electron microscopy of Scanning electron microscopy of the liver: a re­ the bile ductule.Gastroenterologia Jpn }l:476-482. view of methods and results, in: Progress in Li­ ver Diseases, H Popper, F Schaffner (eds), 1976, 38. Itoshima T, Yoshino K, YamamotoK, Munetomo Grune & Stratton Inc, NewYork, V:l-23. F, Murakami T. (1978). Scanning electron micros­ copy of puncture-perfused human liver biopsy sam­ 28. Howard KS, Postek MT.(1979). Dehydration of ples. Observation of bile ductules and bile cana­ scanning electron microscopy specimens. A biblio­ liculi. Scanning ElectronMicrosc.1978;II:169-174. graphy (1974-1978). Scanning Electron Microsc. 1979;11:892-902. 39. Itoshima T, Yoshino K, YamamotoK, Munetomo F, Shimada Y, Nagashima H. (1978). Alterations of 29. Humphreys WJ, Spurlock BO, Johnson JS. liver cell nuclei in hepatitis as revealed by (1974). Critical point drying of ethanol­ scanning electron microscopy. Scanning Electron infiltrated, cryofractured biological specimens Microsc. 1978;II:203-208. for scanning electron microscopy. Scanning Elec­ tron Microsc. 1974:276-282. 40. Itoshima T, Yoshino K, YamamotoK, Munetomo F, Murakami T. (1978). Cleaning of scanning elec­ 30. Ikejiri N, Tanikawa K.(1977).Effects of vi­ tron microscope specimens by puncture perfusion. tamin A and estrogen on the sinusoidal cells in Microsc Acta 80:207-210. the rat liver, in: Kupffer Cells and Other Liver Sinusoidal Cells, E Wisse, DL Knook (eds), 1977 41. Itoshima T, Kiyotoshi S, Kawaguchi K, Yoshi­ Elsevier/North-Holland Biomedical Press, Amster­ no K, MunetomoF, Ohta W, Shimada Y, Nagashima H. dam, 83-93. (1980). Scanning electron microscopy of rat bile canalicular-ductular junction. Scanning Electron 31. Inoue T.(1983). Use of distilled water as a Microsc. 1980;III:373-378. rinsing solution for intracellular observation by scanning electron microscopy. Scanning Electron 42. Itoshima T, Kiyotoshi S, Kawaguchi K, Ogawa Microsc. 1983;1:227-233. H, Ohta W, Ito T, Hirakawa H, Shimada Y, Nagashi­ ma H. (1981). Kupffer cell hyperplasia in liver 32. Ito T.(1951).Cytological studies on stellate diseases. Demonstration by scanning electron cells of Kupffer and fat-storing cells in the ca­ microscopy of biopsy samples. Gastroenterol Jpn pillary wall of the human liver. Acta Anat Nippon 16: 223-231. 26:42-74. 43. Itoshima T, Kiyotoshi S, Kawaguchi E, Ito T, 33. Ito T, Tanuma Y, Shibasaki S. (1980). Junc­ Ogawa H, Nagashima H, YamamotoK, Kobayashi T, tions between Kupffer cells and hepatic sinu­ Murakami T. (1981). Scanning electron microscopy soidal endothelium. A review. Okajimas Folia Anat of rat degenerated hepatocytes after carbon Jpn 57:145-158. tetrachloride intoxication. Scanning Electron Microsc. 198l;III:131-136. 34. Ito T, Itoshima T, Ukida M, Tobe K, Kiyoto­ shi S, Kawaguchi K, Ogawa H, Yamamoto K, Hattori 44. Itoshima T, Ito T, Tobe K, Kiyotoshi S, Ka­ S, Kitadai M, Mizutani S, Tuchiya T, Kita K, Ta­ waguchi K, Ogawa H, Hattori S, Kitadai M, Mizuta­ naka R, Nagashima H. (1984). Scanning electron ni S, Tsuchiya T, Ukida M, Nagashima H. (1982). microscopy of Ito's fat-storing cells in the rat Scanning electron microscopy of fat-storing cells liver. Acta Med Okayama 38:1-9. in liver biopsy specimens of liver diseases, in: Sinusoidal Liver Cells, DL Knook, E Wisse (eds), 35. Itoshima T, Kobayashi T, Shimada Y, Murakami Elsevier Biomedical Press, Amsterdam, 21-28. T.(1974).Fenestrated endothelium of the liver si­ nusoids of the guinea-pig as revealed by scanning

1031 G. Macchiarelli and P.M. Motta

45. Jones AL, Schmucker DL. (1977). Current con­ of baboons fed alcohol: a scanning electron cepts of liver structure as related to function. microscopic study. Hepatology 4:386-391. Gastroenterol 73:833-851. 58. Makabe S, Motta PM.(1980) Scanning electron 46. Jones AL, Spring-Mills E.(1983). The liver microscopy of foetal liver sinusoids in humans. and , in: Histology. Cell and Tissue Folia Morphol Praha 28:85-87. Biology, L Weiss (ed),Elsevier Science Publishing Co Inc, NewYork, 708-748. 59. Malick L, Toth B. (1977). Injection replica­ tion of the vasculature of angiosarcomas in liver 47. Kaneda K, Wake K. (1983). Distribution and of Syrian Golden Hamster. Scanning Electron morphological characteristics of the pit cells in Microsc. l977;II:533-540. the liver of the rat. Cell Tiss Res 233:485-505. 60. Mall FP. (1906). A study of the structural 48. Kardon RH, Kessel RG. (1980). Three-dimen­ unit of the liver. AmJ Anat 5:227-308. sional organization of the hepatic microcircula­ tion in the rodent as observed by scanning elec­ 61. Malpighi M. (1666). De Hepate, Bologna. tron microscopy of corrosion cast. Gastroenterol 79: 72-81 . 62. Marinozzi G, Muto M, Correr S, Motta P. (1977).Scanning electron microscope observations 49. Kessel RG, Kardon RH.(1979). Tissues and Or­ of intrahepatic biliary tree. I, canaliculo­ gans. A Text Atlas of Scanning Electron Micros­ ductular junction. Bile ductules and ducts. J copy, WHFreemann, San Francisco CA, 112-121. Submicr Cytol ~:127-143.

50. Kiernan F.(1833). The anatomy and physiology 63. Metz J, Aoki A, Merlo M, Forssmann WG.(1977) of the liver. Philos Trans Roy Soc Lond 123: 711- Morphological alterations and functional changes 770. of inter-hepatocellular junctions induced by bile duct ligation. Cell Tiss Res 182:299-310. 51. Kiyotoshi S. (1982). Kupffer cell hyperpla­ sia in rats intoxicated by carbon tetrachloride 64. Miyai K, Wagner RM, Richardson AL. (1974). as demonstrated by scanning electron microscopy. Preparation of liver for combined SEMand TEM Gastroenterologia Jpn .!1.:422-429. study. Scanning Electron Microsc. 1974:283-290.

52. Kusumoto Y, Fujita T.(1977).Vitamin A uptake 65. Miyai K, Richardson AL, Mayr W, Javitt NB. cells distributed in the liver and other organs (1977).Subcellular pathology of rat liver in cho­ of the rat. Arch histol jap 40:121-135. lestatis and choleresis induced by bile salts. I. Effects of lithocolic , 3 beta-hydroxy-5-chole­ 53. Lametschwandtner A, Lametschwandtner U, Wei­ noic, cholic and dehydrocholic acids. Lab Invest ger T.(1984).Scanning electron microscopy of vas­ 36:249-258. cular corrosion casts. Technique and applica­ tions. Scanning Electron Microsc.1984;II:663-695. 66. Montesano R, Nicolescu P. (1978). Fenestra­ tions in endothelium of rat liver sinusoids re­ 54. Layden T, Schwarz J, Boyer JL. ( 1974). Scan­ visted by freeze-fracture. Anat Rec 190:861-870. ning electron microscopic (SEM) evaluation of mo­ dels of cholestasis-lithocholate produces a dis­ 67. Motta PM.(1975). A scanning electron micros­ tinct injury to bile canalicular membranes. copic study of the rat liver sinusoid: endotheli­ Gastroenterol 67:806. al and Kupffer cells. Cell Tiss Res 164:371-385.

55. Layden T, Schwarz J, Boyer JL. (1975). Scan­ 68. Motta PM. (1977). The three-dimensional fine ning electron microscopy of the rat liver. Stu­ structure of the liver as revealed by scanning dies of the effect of taurolithocholate and other electron microscopy.Int Rev Cytol suppl _§_:347-397. models of cholestatis. Gastroenterol 69:724-738. 69. Motta P.(1977). Kupffer cells as revealed by 56. Luft JH.(1971).Ruthenium red and violet. II. scanning electron microscopy, in: Kupffer Cells Fine structural localization in animal tissues. and Other Liver Sinusoidal Cells, E Wisse, DL Anat Rec 171:369-416. Knook (eds), Elsevier/North Holland Biomedical Press, Amsterdam, 93-102. 57. Mak KM, Lieber CS. (1984). Alterations in endothelial fenestrations in liver sinusoids 70. Motta PM. (1979) . The location of Kupffer

1032 SEMof the Liver cells in the liver of different mammals. A three Shain, Tokyo/New York, Pp.: Preparation of Liver dimensional analysis by scanning electron micros­ Samples for SEM, 1-4; The Peritoneal Liver Sur­ copy. Verh Anat Ges 73:827-830. face, 5-13; Liver Architecture and Vasculature, 14-29, 83-143; Liver Parenchymal Cells and Intra­ 71. Motta PM.(1981). Three-dimensional architec­ hepatic Biliary Tree, 30-81; Future Applications ture of the mammalian liver. A scanning electron of SEMin Hepatology, V, 145-171. microscopy review, in: Three Dimensional Micro­ anatomy of Cell and Tissue Surfaces, LJA DiDio, 82. Motta PM, Fujita T, Nishi M.(1982). Scanning PMMotta, DJ Allen (eds), Elsevier/North Holland electron microscopy of the mammalian liver, in: Inc, 33-50. Basic and Clinical Hepatology, PM Motta, LJA DiDio (eds), Martinus Nijhoff, The Hague/Boston/ 72. Motta PM.(1982).Scanning electron microscopy London, 31-50. of the liver, in: Progress in Liver Diseases, H Popper, F Schaffner (eds), Grune & Stratton Inc, 83. Murakami T. (1971 ). Application of the scan­ New York, VII:l-16. ning electron microscope to the study of the fine distribution of blood vessels. Arch histol jap 73. Motta PM.(1984).The three-dimensional micro­ 32:445-454. anatomy of the liver. Arch histol jap 47:1-30. 84. Murakami T.(1973).A metal impregnation met­ 74. Motta P, Fumagalli G. (1974). Scanning elec­ hod of biological specimens for scanning electron tron microscopy demonstration of cilia in rat in­ microscopy. Arch histol jap ~:323-326. trahepatic bile ducts. Z Anat Entwickl-Gesch 145: 223-226. 85. Murakami T.(1974). A revised tannin-osmium method for non-coated scanning electron micros­ 75. Motta P, Fumagalli G.(1975).Structure of rat cope specimens. Arch histol jap ~:189-193. bile canaliculi as revealed by scanning electron microscopy. Anat Rec ~:499-514. 86. Murakami T. (1976). Puncture perfusion of small tissue pieces for scanning electron micros­ 76. Motta PM, Makabe S.(1980). Foetal and adult copy. Arch histol jap 39:99-103. liver sinusoids and Kupffer cells as revealed by scanning electron microscopy, in: The Reticuloen­ 87. Murakami T, Itoshima T, Shimada Y. (1974). dothelial System and the Pathogenesis of Liver Peribiliary portal system in the monkey liver as Disease, H Liehr, M Grun (eds), Elsevier/North evidenced by the injection replica scanning elec­ Holland Biomedical Press, ll-16. tron microscope method.Arch histol jap lZ_:245-260.

77. Motta P, Porter KR. (1974). Rat liver cells 88. Muto M. (1975). A scanning electron micros­ and associated spaces of Disse as revealed by copic study on endothelial cells and Kupffer scanning electron microscopy. Proceedings 8th In­ cells in rat liver sinusoids. Arch histol jap lZ_: ternational Congress of Electron Microscopy. GW 369-386. Saunders, DJ Goodchild (eds), The Australian Aca­ demy of Science, Canberra, Australia, II:410-411. 89. Muto M, Nishi M, Fujita T. (1977). Scanning electron microscopy of human liver sinusoids. 78. Motta P, Porter KR. (1974). Etude au micros­ Arch histol jap 40:137-151. cope electronique a balayage des espaces de Disse et des sinusoides dans le foie du rat. Bull Assn. 90. Nishi M.(1978). Scanning electron microscope des Anatomistes 58:162. study on the mechanism of obstructive jaundice in rats. Arch histol jap i!_:411-426. 79. Motta P, Porter KR. (1974). Structure of rat liver sinusoids and associated tissue spaces as 91. Nishi M, KawamuraT, KamimuraT, Murayama H, revealed by scanning electron microscopy. Cell Sasaki H, lchida F.(1976). Scanning and transmis­ Tiss Res 148:111-125. sion electron microscopic observation of primary biliary cirrhosis. J Clin Electron Microsc 2_:5-6. 80. Motta PM, Andrews PM, Porter KR. (1977). Mi­ croanatomy of Cell and Tissue Surfaces. An atlas 92. Nopanitaya W, Grisham JW. (1975). Scanning of SEM, Lea & Febiger, Philadelphia PA, 105-115. electron microscopy of mouse intrahepatic struc­ tures. Exp Mol Pathol Q:441-458. 81. Motta P, Muto M, Fujita T.(1978). The Liver. An Atlas of Scanning Electron Microscopy, Igaku 93. Nopanitaya W, Grisham JW, Carson JL, Dotson

1033 G. Macchiarelli and P.M. Motta MM.(1976). Surface features of cirrhotic liver. remarks on the fine structure of sinusoidal Virchows Arch A Path Anat and Histol 372:97-108. cells. Okajimas folia anat jpn 58:325-368.

94. Nopanitaya W,LambJC, Grisham JW, Carson JL. 104. Ohata M, Tanuma Y, Ito T.(1984). A transmis­ (1976). Effect of hepatic venous outflow obstruc­ sion electron microscopic study on sinusoidal tion on pores and fenestrations in sinusoidal cells of Guinea Pig liver, with special reference endothelium. Br J exp Path 2.Z_:604-609. to the occurrence of a canalicular system and "pored domes" in the endothelium. Arch histol jap 95. Nopanitaya ?I, Grisham JW, Lesene HR. (1977). 47:359-376. A preliminary assessment of the use of scanning electron microscopy for diagnostic evaluation of 105. Ohtani 0. ( 1979). The peri bi 1i ary porta 1 liver biopsies. Scanning Electron Microsc. system in the rabbit liver. Arch histol jap 1977; I I: 159-169. 42:153-167.

96. Nopanitaya W, Grisham JW, Aghajanian JG, 106. Ohtani 0.(1981). Microcirculation studies by Carson JL. (1978). Intrahepatic microcirculation: the injection-replica method with special refe­ SEM study of the terminal distribution of the rence to the portal circulations, in: Three Di­ hepatic artery. Scanning Electron Microsc. mensional Microanatomy of Cells and Tissue Sur­ 1978;II:837-842. faces, DJ Allen, PM Motta, LJA DiDio (eds), Elsevier/North Holland, Amsterdam, 51-70. 97. Nopanitaya W, Aghajanian J, Grisham JW, Carson JL. (1979). An ultrastructural study on a 107. Ohtani 0.(1984).Review of scanning and light new type of hepatic perisinusoidal cell in fish. microscopic methods in microcirculation research Cell Tiss Res 198:35-42. and their application in pancreatic studies. Scanning Electron Microsc. 1984;II:653-661. 98. Nordestgaard BG, Rostgaard J. (1985). Criti­ cal-point drying versus freeze drying for scan­ 108. Ohtani 0, Murakami T.(1978).Peribiliary por­ ning electron microscopy: a quantitative and qua­ tal system in the rat liver as studied by the in­ litative study on isolated hepatocytes. J Microsc jection-replica scanning electron microscope met­ 137:189-207. hod. Scanning Electron Microsc. 1978;II:24l-245.

99. Ogawa H.(1982). Scanning electron microscopy 109. Ohtani 0, Murakami T.(1985).The blood supply of rat liver hyperplastic nodules induced by of the liver and the : scanning electron diethylnitrosamine. Scanning Electron Microsc. microscopy of vascular cast and intravital mi­ 1982;IV:1793-1798. croscopy of living tissues. XII International Congress of Anatomy, London, Abstract book, Book 100. Ogawa H, Itoshima T, Ukida M, Ito T, Kiyoto­ Production Consultants, Cambridge, 522. shi S, Kitadai M, Hattori S, Mizutani S, Kita K, Tanaka R, Tobe K, Nagashima H, Kobayashi T, Ha­ 110. Ohtani 0, Murakami T, Jones AL.(1982).Micro­ maya K. (1984). Scanning electron microscopy of circulation of the liver, with special reference experimentally induced hepatocellular carcinoma. to the peribiliary portal system, in: Basic and Scanning Electron Microsc. 1984;I:359-368. Clinical Hepatology, PMMotta, LJA DiDio (eds), Martinus Nijhoff, The Hague, II:85-96. 101. Ogawa H, Itoshima T, Ukida M, Ito T, Kiyoto­ shi S, Kitadai M, Hattori S, Mizutani S, Kita K, 111. Ohtani 0, Kikuta A, Ohtsuka A, Taguchi T, Tanaka R, Tobe K, Nagashima H, Kobayashi T, Ham­ Murakami T.(1983). Microvasculature as studied by aya K. (1984). Scanning electron microscopy of the microvascular corrosion casting. Scanning experimentally induced sequential alterations of electron microscope method. I. Endocrine and di­ rat liver hyperplastic nodules. Scanning Electron gestive system. Arch histol jap 46:1-42. Microsc. 1984;I:369-374. 112. Okanoue T, Ohta M, Fushiki S, Ou 0, Kachi K, 102. Ogawa H, Itoshima T, Nagashima H. (1985). Okuno T, Takino T, French SW. (1985). Scanning Cavernous structures composed of hepatocytes in electron microscopy of the liver cell cytoske­ experimentally induced hepatocellular carcinoma. leton. Hepatology ~:1-6. Acta Pathol Jpn ~:1151-1162. 113. Overton J, Meyer R. (1984). Aspects of liver 103. Ohata M, Tanuma Y, Ito T. (1982). Electron and gut development in the chick. Scanning Elec­ microscopic study on avian livers with special tron Microsc. l984;II:737-746.

1034 SEMof the Liver 114. Penasse W, Bernaert D, Mosselmans R, Wanson sinusoidal endothelial cells in liver of normal JC, Drochmans P.(1979). Scanning electron micros­ and experimental rabbits.Anat Embryol ~:125-142. copy of adult rat hepatocytes in situ, after iso­ lation of pure fractions by elutriation and after 125. Tanaka K, Iino A. (1973). Demonstration of culture. Biol Cellulaire 34:175-186. fibrous components in hepatic interphase nuclei by high resolution scanning electron microscopy. 115. Perrissoud D, Auderset G, Reymond0, Maignan Exptl Cell Res ~:40-46. MF. (1981). The effect of carbon tetrachloride on isolated rat hepatocytes. Early morphological al­ 126. Tanikawa K. (1979). Ultrastructural Aspects terations of the plasma membrane. Virchows Arch of the Liver and Its Disorders. Igaku-Shoin, (Cell Pathol) 35:83-91. Tokyo, l -98.

116. Postek MT, Howard KS, Johnson AH, McMichael 127. Tanuma Y, Ohata M, Ito T. (1983). Electron KL. (1980). Scanning Electron Microscopy. A Stu­ microscopic studies on the sinusoidal cells in dent's Handbook, Ladd Researching Industries Inc, the monkey liver. Arch histol jap 46:401-426. Burlington VT, 115-238. 128. Tokunaga J, Edanaga M, Fujita T, Adachi K. 117. Rappaport AM.(1973).The microcirculatory he­ (1974). Freeze cracking of scanning electron patic unit. Microvasc Res ~:212-228. microscope specimens. A study of the kidney and spleen. Arch Histol jap lZ_:165-182. 118. Rappaport AM, Borowy ZJ, Lougheed WM,Lotto WN. (1954). Subdivision of hexagonal liver 129. Van Blerkom J, Motta PM. (1979). The Cel­ lobules into a structural and functional unit. lular Basis of Mammalian Reproduction, Urban & Role in hepatic physiology and pathology. Anat Schwarzenberg, Munich, 5-51. Rec 119: ll-34. 130. Vial JD, Porter KR. (1975). Scanning micros­ 119. Reith A, Kraemer M, Vassy J. (1984). The in­ copy of dissociated tissue cells. J Cell Biol fluence of mode of fixation, type of fixative and 67:345-360. vehicles on the same rat liver: a morphometric/ stereologic study by light and electron micros­ 131. Vial JD, Simon FR, Mackinnon AM. (1976). Ef­ copy. Scanning Electron Microsc.l984;II:645-65l. fect of bile duct ligation on the ultrastructural morphology of hepatocytes. Gastroenterol 70:85- 120. Renkin EM, Curry FE. (1982). Endothelial 92. permeability: pathways and modulations. Ann NY Acad Sci 401:248-259. 132. Vidal-Vanaclocha F,Barbera-Guillem E. (1985) Fenestration patterns in endothelial cells of rat 121. Satodate R, Sasou S, Oikawa K, Hatakeyama N, liver sinusoids. J Ultrastruct Res 90:115-123. Katsura S. (1977). Scanning electron microscopi­ cal studies on the Kupffer cell in phagocytic 133. VonnahmeFJ. (1977). A scanning electron mi­ activity, in: Kupffer Cells and Other Liver croscopic study of Kupffer cells in the monkey Sinusoidal Cells. E Wisse, DL Knook (eds), Else­ liver,in: Kupffer Cells and Other Liver Sinu­ vier/North Holland Biomedical Press, Amsterdam, soidal Cells. Wisse E, Knook DL (eds), Elsevier/ 121-129. North Holland Biomedical Press,Amsterdam,103-108.

122. Scharschmidt BF. (1982). Bile formation and 134. VonnahmeFJ. (1980). An improved method for cholestasis, metabolism and enterohepatic circu­ transparenchymal fixation of human liver biopsies lation of bile acids, and gallstone formation, for scanning electron microscopy. Scanning Elec­ in: Hepatology. A Text-Book of Liver Disease, tron Microsc. l980;III:l77-l80. D Zakim, TD Boyer (eds), WBSaunders Co., Phila­ delphia PA, 297-351. 135. VonnahmeFJ. (1981). A scanning electron mi­ croscopic study of the liver of the monkey Macaca 123. Skaaring P, Bierring F. (1976). On the in­ Speciosa. II. Intra- and extrahepatic biliary trinsic innervation of normal rat liver. Histo­ system. Cell Tiss Res 215:207-214. chemical and scanning electron microscopical studies. Cell Tiss Res 171:141-155. 136. VonnahmeFJ. (1982). The structure and fun­ ction of subendothelial processes of per1s1nu­ 124. Tamaru T, Fujita H. (1978). Electron-mi­ soidal cells in human liver disease; in: Sinu­ croscopic studies on Kupffer's stellate cells and soidal Liver Cells , DL Knook, E Wisse (eds) ,

1035 G. Macchiarelli and P.M. Motta Elsevier Biomedical Press, Amsterdam, 13-20. 149. Wisse E. (1977). Ultrastructure and function of Kupffer cells and other sinusoidal cells in 137. Vonnahme FJ. (1984). The scanning electron the liver; in: Kupffer Cells and Other Liver Si­ microscope as a diagnostic tool in liver patho­ nusoidal cells. E Wisse, DL Knook (eds), Elsevier logy. Scanning Electron Microsc. 1984;1:173-182. North Holland,Biomedical Press, Amsterdam, 33-66.

138. Vonnahme FJ, Brolsch C. (1981) . Ultras­ 150. Wisse E, Knook DL. (1979). The investigation tructural observations of the hepatic vascular of sinusoidal cells: a new approach to the study system in Budd-Chiari 's syndrome. Biblthca Anat of liver function, in: Progress in Liver Di­ 20:98-101. seases, H Popper, F Shaffner (eds),Grune & Strat­ ton Inc, New York, Vol VI, 153-171. 139. Vonnahme FJ, Muller 0. (1981). A scanning electron microscopic study of the liver of the 151. Wisse E, De Zanger RB, Jacobs R, McCuskey monkey Macaca Speciosa. I. Vascular system of the RS. (1983). Scanning electron microscope obser­ hepatic lobule. Cell Tiss Res ~:193-205. vations on the structure of portal veins, sinu­ soids and central veins in rat liver. Scanning 140. Wake K. (1980) . Perisinusoidal stellate Electron Microsc. 1983;III:l441-l452. cells (fat-storing cells,interstitial cells, li­ pocytes), their related structure in and around 152. Wisse E, De Zanger R , Roland J. (1983). the liver sinusoids, and vitamin A-storing cells Scanning EMobservations on rat liver sinusoids in extrahepatic organs. Int Rev Cytol 66:303-353. relevant to microcirculation and transport processes. J Clin Electron Microsc 16:427-430. 141. Walker RM, Racz WJ, Mc Elligott TF. (1983). Scanning electron microscopic examination of 153. Wisse E, De Wilde A, De Zanger R. (1983). acetaminophen-induced hepatotoxicity and con­ Perfusion fixation of human and rat liver tissue gestion in mice. AmJ Pathol 113:321-330. for light and electron microscopy: a review and assessment of existing methods with special em­ 142. Wanson JC, Bernaert D, May C. (1979). Mor­ phasis on sinusoidal cell and microcirculation, phology and functional properties of isolated and in: The Science of Biological Specimen Prepara­ cultured hepatocytes; in: Progress in Liver Di­ tions for Microscopy and Microanalysis, JP Revel, seases, H Popper, F Schaffner (eds), Grune & T Barnard, GHHaggis (eds), SEMInc, AMFO'Hare, Stratton Inc, New York, Vol VI, l-22. Chicago IL, 31-38.

143. Wanson JC, Penasse W, Bernaert D, May C. 154. Wisse E, Charles K, Vandersmissen P, De Zan­ (1979). Cell surface properties of control and ger RB.(1984). Comparative morphometry on rat li­ carcinogen-treated hepatocytes, isolated in sub­ ver sinusoids,endothelial fenestrae,and red blood populations by elutriation. Scanning Electron cells fixed by constant low pressure perfusion Microsc. l979;III:l6l-l68. fixation,in: Proceedings of the 42nd Annual Meet­ ing of Electron Microscopy Society of America 144. Wanson JC, Penasse W , Bernaert D, May C. Bailey GW(ed), San Francisco Press Inc, 208-209. (1980). SEMstudy of adult rat hepatocytes from preneoplastic and hyperplastic foci induced in 155. Wisse E, De Zanger RB, Charles K, Van Der the liver by diethylnitrosamine. Scanning Elec­ Smissen P,McCuskey RS.(1985).The liver sieve: tron Microsc. l980;III:29-34. considerations concerning the structure and fun­ ction of endothelial fenestrae,the sinusoidal 145. Weibel ER. (1974). Selection of the best me­ wall and the space of Disse.Hepatology f:683 692. thod on stereology. J Microsc .!..Q_Q_:261-269. 156. Wright PL, Clemett JA, Smith KF, Day WA, 146. Wepfer F . (1664) . Quoted by Jones and Fraser R. (1983). Hepatic sinusoidal endothelium Spring-Mills, 1983. (See ref. 46 above). in goats. Aust J Exp Biol Med Sci 61 :739-741.

147. Werb Z. (1983). Howthe macrophage regulates 157. Wright PL, Smith KF, Day WA,Fraser R. its extracellular environment. AmJ Anat 166:237- (1983). Hepatic sinusoidal endothelium in sheep: 256. an ultrastructural reinvestigation. Anat Rec 206:385-390. 148. Wisse E.(1970).An electron microscopic study of the fenestrated endothelial lining of rat li­ 158. Wright PL, Smith KF, Day WA, Fraser R. ver sinusoids. J Ultrastruct Res 31 :125-150. (1983). Small liver fenestrae may explain the

1036 SEMof the Liver susceptibility of rabbits to atherosclerosis. tissue preparation. In a review like this, we Arteriosclerosis 3:344-348. might expect a point of view or a conclusion with regard to this matter. Would you please formulate 159. YamamotoK, Itoshima T, Ito T, Ukida M, Oga­ your opinion on how we should judge the quality wa H, Kitadai M, Hattori S, Mizutani S, Nagashima and correctness of a good SEMliver preparation? H. (1983). Scanning electron microscopy of a li­ Authors: It is not easy to define in a few words ver cavernous hemangioma. Gastroenterolgia Jpn what the standard of quality is for a good SEM 18: 15-20. liver preparation. Actually, as demonstrated in this paper, one singular mode of preparation of 160. YamamotoK, Sherman I, Phillips MJ, Fisher liver samples does not exist. In fact, due to the MM.(1985). Three-dimensional observations of the complicated structure of the liver, different hepatic arterial terminations in rat, hamster and techniques are needed to evidence different human liver by scanning electron microscopy of structures: most of the time, it is not possible microvascular casts. Hepatology ~:452-456. to obtain a good conservation of sinusoids and of bile canaliculi in the same sample. As to how we 161. Yoshino K. (1979). Scanning electron micros­ should judge the SEMliver preparations,our meter copy of the liver of primary biliary cirrhosis. is represented by the pictures offered in this Scanning Electron Microsc. l979;III:697-704. paper as well as those presented in other publications of our group (text ref. 73,81,82). 162. Yoshino K. (1980). Scanning electron micros­ copy on the rat liver with alpha-naphthyliso­ E. Wisse: You state that the morphology of endo­ thiocyanate-induced cholestasis. Gastroente­ thelial cells differs from organ to organ, even rolgia Jpn ]_i:550-563. within one organ. Do you think there only exists one type of endothelial cell with different mor­ 163. Zamboni L. (1965). Electron microscopic stu­ phological expression (adapted to local circum­ dies of blood embryogenesis in humans. I. The stances), instead of a multitude of differently ultrastructure of the fetal liver. J Ultrastruct programmed endothelial cell types? Res 12:509-524. Authors: Both hypotheses are likely. According to morphological data we cannot express any further Discussion with Reviewers comment. However "local circumstances" easily may induce changes of the endothelium. E. Wisse: In your introduction you mention that new SEMimages on the liver sinusoids of the ani­ E. Wisse: Have you seen junctional complexes mal Praomys wi 11 be shown. Maywe know your reason between endothelial and Kupffer cells? for looking at the sinusoids of these animals? Authors: No, personally, we have not seen them, What were the results? but they have been reported (text reference: 33). Authors: Praomys Mastomys Natalensis (PMN) is a rodent native to South Africa which easily deve­ T. Itoshima: The bridge in Fig.14 seems likely lops neoplastic processes in certain organs. Fur­ but broken.We would like to see an unbroken ther, it has a female prostate that often deve­ bridge in the whole. lops cancer. Up to now, its ultrastructural mi­ Authors: See Fig. 23. croanatomy has not been extensively studied. Our group, in collaboration with Dr. DiDio's group at the Medical College of Ohio, is reviewing the mi­ croanatomy of the liver as well as that of other organs (ovary, uterus, heart), in order to eluci­ date the basis of the pathological processes. As far as the results are concerned, PMNliver, in healthy animals, appeared similar to that of other rodents, except for the sinusoidal bridging structures described in this paper. We reported these structures, in the present paper, because it is a new finding and it is of interest in un­ derstanding the hemodynamics of the liver micro­ circulation.

E. Wisse: You are summing up quite a number of data and considerations concerning fixation and Fig. 23 Praomys liver. An entire intrasinusoidal bridge (arrow). (S = sinusoid). Bar is 10 µm.

1037 G. Macchiarelli and P.M. Motta T. Itoshima: The arrangement of the hepatic lami­ but correlated SEM-TEMstudies are needed to nae: you describe that changes in the blood pres­ better elucidate this problem. sure gradient produce a deformation of the liver laminae, which may temporarily alter their loca­ S.G. McClugage: Endothelial bridges are described tion in space. I cannot understand what you say. in figures 10, 13 and 14. Have these structures Do you mean to say that hepatic laminae change been described in other parts of the vascular their interrelationships dynamically? As you know system such as the cavernous sinus in the cranial hepatic labyrinth is interconnected three-dimen­ vault? Does their incidence vary from one end of sionally and their relationships are rigidly re­ the sinusoid (peri-portal) to the other (cen­ stricted by hepatocyte continuity. tral)? Since such endothelial-covered structures Authors: We agree that the interrelations of the would increase the potential surface area for hepatic labyrinth are rigidly restricted by hepa­ thrombosis could they not affect nutritional tocytic continuity. Actually, when we state that blood flow through the lobule under certain pa­ deformation of the liver laminae may occur, we thologic conditions? mean to say that the aspect of the liver archi­ Authors: We have no data concerning the presence tecture is dependent upon the filling state of of similar structures in organs other than liver. the sinusoids with blood. In fact, as you know, - The endothelial bridges were found uniformly not all the liver sinusoids are filled with blood throughout the entire sinusoid. - Any structure at the same time, but distribution of blood flow that changes the regular laminar flow through a changes according to the momentary functional vessel creates turbulence. This happens at the state of the liver. Thus, we can find both empty level of any vascular branch or at the site of an and filled sinusoids. This causes variations in endothelial lesion (disendothelization). When the sinusoidal caliber. It is precisely due to turbulent blood flow occurs, platelets may be ac­ the close relationship existing between sinusoids tivated, with consequent thrombus formation at and hepatic laminae, that the latter changes the level of an endothelial lesion. Partial or their location in the space. The same phenomena total vascular occlusion occurs as a consequence happen in all soft, highly vascularized tissues, of thrombotic process. (Spaet TH, Gaynor E, Ste­ including the liver. mermann MB; Thrombosis, atherosclerosis and endo­ thelium. AmHeart J ~: 661-667, 1974 - Ross R, R.D. Soloway: What is known about the SEMstruc­ Faggiotto A, Bowen-Pope D, Raines E; The role of ture of the sinusoidal valves that control the endothelial injury and platelet and macrophage microcirculation within the lobule? It is known interactions in atherosclerosis. Circulation that kinetic studies of the hepatic microvascula­ 2.Q_(suppl III):77-82, 1984). Thus, structures such ture show that blood flows intermittently through as the intrasinusoidal bridges may actually crea­ many of the sinusoids. te a risk factor for vascular occlusion, but only Authors: Up to now, we have no SEMdata to con­ when disendothelization occurs. firm the existence of sinusoidal valves. We be­ lieve that the regulation of the "intermittent" F.J. Vonnahme: Is there any evidence that the blood flow through the sinusoids is controlled pits which you describe on the serosal surface of mostly by the smooth muscle cell contraction of the liver may play a role in the formation of the intrahepatic vessels, and partially by struc­ asci tes? tures present inside the sinusoids themselves. Authors: No. But it is likely that they may. These latter structures may include the intrasi­ nusoidal bridges or the prolongations of the F.J. Vonnahme: You did not mention the functional Kupffer cells. role of "fat-storing cells" in fibrogenesis. Would you please comment? F. Low: Regarding the question of the basement Authors: Often, fat-storing cells are seen close­ membrane in the liver could you make some com­ ly associated with collagen fibrils only in those ments about what is revealed by SEM along the areas fronting the hepatocytes (text reference lines of Burkel and Low, Am J Anat 118:769-784 73). Therefore, it has been suggested that they (1966)? may have a role in secreting collagen and rein­ Authors: Unfortunately, SEMgives very little in­ forcing the reticular network in the Space of formation on those structures that are not ex­ Disse. This may support the hypothesis that these posed during preparation. Particular techniques cells have some commonor1g1n with fibroblasts; are needed in order to expose basement membrane but further evidence is needed (text references surface, and it is poorly studied by SEMmethods. 73, 126, 140). Up to now, fat-storing cells are However, SEMseems to confirm the fact that the surely recognized as lipocytes,storing vitamin A. basement membrane is not continuous in the liver,

1038