Structure of R a t Bile Canaliculi as Revealed by Scanning Electron Microscopy PIETRO MOTTA AND GUIDO FUMAGALLI The Department of Anatomy, Medical College, University of Rome, Italy
ABSTRACT Bile canaliculi in the rat liver have been studied by scanning electron microscopy. They appear as half tubules carved out of the hepatocytes' surface. In several cases the bile channel bifurcates to form two or three long branches running on the same face of the cell. Therefore, by SEM it seems obvious that the same side of an hepatocyte may be used for bounding two or more bile canaliculi. Some bile canaliculi display a flexuous course and show lateral sacculations. Some of them are large and apparently similar to the short lateral branches of the bile canaliculus; others are smaller and arise obliquely from the subjacent cortical areas of the hepatocyte cytoplasm. These latter structures are best observed in stereo-views in which they appear as narrow intracellular projections bordered with a few microvilli, actually opening into the lumen of the bile canaliculus. The SEM results suggest that these sacculations probably correspond to short intracellular branches of the bile canaliculi. Zones of minimal distance (0.1 p ) between the space of Disse and the bile channel have been frequently observed. They are actually the zones where the intercellular clefts arising from the subendothelial space of Disse come into closest contact with the bile canaliculus and might serve as sites of simple diffusion of substances.
The architecture of the mammalian biliary system has been extensively studied by transmission electron microscopy (TEM) and it is well known that the bile canaliculi are intercellular spaces limited by two adjacent hepatocytes (Rouiller, '54; Fawcett, '55; Steiner and Carruthers, '61; Cossel, '64; Bruni and Porter, '65; Matter et al., '69). A much debated question of liver cytology concerns the existence in mammals of true intracellular bile canaliculi (Novikoff and Noe, '55). Recently their occurrence has been reported in some mammals (Carruthers and Steiner, '62; Barone and Finocchio, '63) and different vertebrates (David, '61; Yamamoto, '65) but denied by some authors (Rouiller, '56; David, '64; Tanikawa, '68). Further, it is also disputed whether there exists a direct communication between the biliary channels and the spaces of Disse (Fawcett, '55; Rouiller and Jezequel, '63; Cossel, '64; David, '64; Tanikawa, '68; Motta and Porter, '74). The purpose of this paper is to show by scanning electron microscopy (SEM) ANAT. REC., 182: 499-514.
some new topographic details of the surface structure of the bile canaliculus in the rat liver, as these bear on the above questions. MATERIAL AND METHOD
Liver tissue from five healthy, young, adult albino rats was perfused with a Tyrode solution through the left ventricle. The perfusion solution was used at room temperature and contained 0.1% procaine, a dab of heparin and bubbled 0%. Gravity was used to maintain a perfusion flow rate of about 60 mil/min. After about 30 seconds the flow of Tyrode solution was interrupted and 2.5% glutaraldehyde in cacodylate buffer solution (Sabatini, Bensch and Barrnett, '63) (0.18 M, pH 7.3) was perfused for about ten minutes. Small pieces of liver were subsequently excised and immersed in the same fixative. After a period of time between one and five days the tissues were washed for three to four hours in the same buffer and cut Received Sept. 12, '74. Accepted Mar. 18, '75.
499
500
PIETRO MOTTA AND GUIDO FUMAGALLI
into small blocks with a razor blade or carefully fragmented (pulled apart) with jeweler's forceps. Dehydration was carried out rapidly in a graded series of acetone and the specimens were transferred to liquid CO, for critical point drying (Porter et al., '72). The dried samples were mounted on aluminum studs using conductive silver paint, and coated with a layer of carbon and gold in a high vacuum evaporator (DV-502, Denton Vacuum). All the specimens were examined and photographed using a Cambridge Stereoscan Model S4 operated at 10 or 20 kV. OBSERVATIONS
As was shown in a previous publication (Motta and Porter, '74) the liver cells prepared by these techniques seem to separate along their surfaces rather than breaking open. Thus, the hepatocytes readily show the open half of canaliculi running along their sides. As seen with the SEM bile canaliculi form a continuous three dimensional network provided with irregular meshes (fig. 1). Generally, a single canaliculus is present along the central zone of one side of any one hepatocyte (fig. 1 ) . But in a number of cases the bile channel bifurcates in two or three branches that run on the same face of the cell (fig. 2). Some bile canaliculi possess a tortuous course and show frequent lateral sacculations and/or very short branches (fig. 3). The canalicular lumens are about 0.5-0.1 in width (figs. 4, 5). The surfaces limiting the half channel possess numerous short microvilli projecting obliquely or perpendicularly into the lumen (figs. 4, 5). The lateral short branches of the bile canaliculi show the same features as the principal channels (figs. 3, 7). Some of them are larger and apparently more similar to lateral diverticula or sacculations of the bile canaliculus (fig. 3); others are smaller and arise obliquely from the subjacent cortical areas of the cytoplasm (figs. 7, 8, 9 ) . This aspect is easily observed in the stereo pairs (figs. 10, 11) in which apparently an intracellular channel bordered by a few microvilli opens into the lumen of the actual intercellular bile canaliculus. The wall surface of these biliary channels
is not completely smooth; in addition to microvilli, they show small holes and/or round deep invaginations (fig. 8). The nature of these holes and invaginations has not been determined. The smaller ones could represent the openings in coated pits that are common on the surfaces of hepatocytes. Others could well represent vacuoles arising from the subjacent Golgi's zone. On some occasions a granular cloudy material was observed in the lumen of the bile canaliculi and/or on the sides of the cells adjacent to them (figs. 3, 6). On either side of the biliary channels there is an area running along the margins of such canaliculi that likely coincide with the limits of intimate contact between adjacent liver cells (figs. 1, 4, 5, 6, 7). These areas that appear as the smoothest parts of the cell surface vary greatly in width. Some are large and measure about 2.0-5.0 p. The narrowest zones correspond generally to the intercellular spaces between two adjacent hepatocytes. Such areas coincide with the zones where the intercellular space of Disse and its recesses appear in close contact with the wall of the bile canaliculi. We have never, however, found a direct communication between these two spaces (figs. 7, 10). At higher magnification even within these areas the surface of the hepatocytes is not absolutely smooth (figs. 4, 9 ) : they are simply regions where the microvilli are relatively short and scarce (Motta and Porter, '74). Further, a number of small holes and/or pits have been observed on these apparently smooth surfaces (fig. 5); there are also larger holes (deep invaginations) and cellular protrusions (figs. 2, 4, 11).These could correspond to the studlike projections into concavities in the adjacent hepatocytes and vice versa as revealed by transmission micrographs. DISCUSSION
TEM studies indicate that as a rule a bile canaliculus runs between each adjacent pair of cells (Rouiller and Jezequel, '63; David, '64; Tanikawa, '68) but that occasionally three hepatocytes bound the lumen of a single canaliculus (David, '64; Barone et al., '67; Tanikawa, '68). Moreover, a single side of the hepatocyte may frequently be provided with two or three
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI
branching separated bile canaliculi. Therefore, by SEM, it seems obvious that the same side of one hepatocyte facing two adjacent cells may be used for bounding two or eventually more bile canaliculi. It is also evident that bile canaliculi have a rather tortuous course and possess short lateral branches and/or sacculations. Such a branching pattern of bile canaliculi was fully appreciated by Novikoff and Noe (‘55) in sections of liver stained for alkaline phosphatase activity or also in histological preparations in which dyes were injected into the biliary tree (Barone and Finocchio, ’63). Some of these branches, e.g., those depicted in figures 2 and 3, could well represent the intercellular origins of the biliary canalicular tree. In this regard it is also important to note that the scanning micrographs (particularly the stereopairs) show short branches of the bile canaliculi, penetrating obliquely into the cytoplasm. The existence of true “intracellular bile canaliculi” introduced by Pfluger (1869) and von Kupffer (1876) was a much debated question in the past years; but it was widely accepted by a number of light microscopists (see the review of Novikoff and Noe, ’55). With the introduction of the electron microscope, intracellular extensions of bile canaliculi have been extensively described only by Barone and Finocchio (’63) but their occurrence has been repeatedly mentioned in other recent investigations (Carruthers and Steiner, ’62; Tanikawa, ’68; Matter et al., ’69). On the contrary Rouiller (’56) in his original electron microscopic study on the bile canaliculus concluded: “then it is possible that the majority of the structural features considered as intracytoplasmic bile canaliculi represent indeed a part of the Golgi apparatus. In addition their location near the bile canaliculi gives the impression that they are intracytoplasmic branches of bile canaliculi.” This concept seems to be largely accepted since in recent textbooks of Histology it is reported that the descriptions of intracellular branches of bile canaliculi penetrating the cytoplasm . . . .“seem to be a misinterpretation of vacuoles that often occur in the cytoplasm adjacent to the canaliculus and in con-
50 1
junction with the Golgi complex” (Bloom and Fawcett, ’68). The intracellular extensions found by SEM are rather short but always possess a lumen lined by a few microvilli similar to those lining the wall of the intercellular bile canaliculi. Although other techniques such as freeze fracture preparations might provide more evidence of the real existence of these intracellular canalicular projections, the present SEM observations (particularly the stereoviews ) seem to suggest that short branches arising from the bile canaliculi may penetrate the cytoplasm of perfused rat hepatocytes. One important point to be considered concerns the nature of the surfaces of the hepatocytes on either side of the bile canaliculus. In the present study the different types of junctional complexes were not clearly identified. But it is reasonable to suggest that the smoothest areas running along either side of the half bile channel may coincide with the limits of intimate contact between adjacent hepatocytes which in turn represent part of the junction a1 complex. In this regard it is also important to note that the SEM results show the spaces of Disse to be larger than one would gather from TEM pictures. In fact, they frequently extend among adjacent hepatocytes in deep recesses in close contact with the bile canaliculi (Motta and Porter, ’74). Further, a minimal distance of 0.1 was noted between the bile canaliculi and spaces of Disse in a study by TEM of serial sections of rat liver (Matter et al., ’69). Our observations suggest that regions of such minimal distance between the bile canaliculi and the recesses of Disse are not occasional findings but occur more numerously than is generally evident from transmission micrographs. In addition as observed by SEM, they are actually the regions where the intercellular recesses arising from the subendothelial spaces of Disse (Motta and Porter, ’74) come into closest contact with the bile canaliculus. Therefore it may be possible that substances which have the same concentration in the bile as in the blood (i.e., electrolytes and glucose) could be exchanged by simple diffusion (Brauer, ’63) in these narrow are as.
502
PIETRO MOTTA AND GUIDO FUMAGALLI ACKNOWLEDGMENTS
This work was partially supported by a Fulbright grant to Dr. Pietro Motta and undertaken in the Department of Molecular, Cellular and Development Biology, University of Colorado, Boulder, Colorado. The authors are grateful to Dr. Keith R. Porter for his generosity in making laboratory facilities available and for his helpful and most valuable suggestions. Sincere thanks are also due to Dr. A. J. Ladman for his interest and advises in editing the paper. The special support of the Italian Foreign Office was also greatly appreciated. LITERATURE CITED Barone, P., and F. Finocchio 1963 Ricerche morfologiche ed ultrastrutturali sui diverticoli intracellulari dei capillari biliari. Arch. Ital. Anat. Istol. Pat., 36: 366-379. Barone, R., C. Inferrera and G. Carrozza 1967 Aspetti ultrastrutturali dell'epatocita in rapporto alla secrezione biliare ed alle sue alterazioni. Relaz. Atti X Congr. SOC.Ital. Patologia Messina-Taormina, pp. 49-21 1. Bloom, W., and D. W. Fawcett 1968 A textbook of Histology, ninth ed. W. B. Saunders Co. Philadelphia-London-Toronto. Brauer, R. W. 1963 Liver circulation and function. Physiol. Rev. 43: 115-213. Bruni, C., and K .R. Porter 1965 The fine structure of the parenchymal cell of the normal rat liver. I. General observations, Amer. J. Path., 46: 691-755. Carruthers, J. S., and J. W. Steiner 1962 Fine structure of terminal branches of the biliarv tree. Arch. Path., 74: 39-48. Cossel, L. 1964 Die menschliebe Leber in Elektronenmikroskop. Gustav Fischer, Jena. David H. 1961 Zur submikroskopischen Morphologie intrazellularen Gallen-kapillaren. Acta Anat. (Basel), 47: 216-224. Submikroskopische Ortho-und 1964 Pathomorphologie der Leber. Akademie Verlag, Berlin.
Fawcett, D. W. 1955 Observations o n the cytology and electron microscopy of hepatic cells. J. Nat. Cancer. Inst., 15: 1475-1502. Kupffer, C. von 1876 Ueber Sternzellen in der Leber. Arch. f . mikr. Anat., 22: 353-385. Matter, W., L. Orci and C. Rouiller 1969 A study on the permeability barriers between Disse's space and the bile canaliculus. J. Ultrastr. Res., suppl. 11: 1-71. Motta, P., and K. R. Porter 1974 Structure of Rat Liver Sinusoids and Associated Tissue Spaces as revealed by Scanning Electron Microscopy. Cell Tiss. Res., 248: 111-125. Novikoff, A. B., and E. F. Noe 1955 Observations o n fragmented rat liver canaliculi and the problems of intracellular bile canaliculi and Golgi apparatus. J. Morph., 96: 189-211. Pfluger, E. 1869 Pfluger's Arch. Ges. Physiol., 2: 459 (quoted by Rouiller, '56). Porter, K. R., D. Kellev and P. M. Andrews 1972 The preparation of cultured cells and soft tissues for scanning electron microscopy. In: Proceedings of the 5th Annual Stereo-scan Colloquium. Kent Cambridge Scientific Co., Chicago. pp. 1-19. Rouiller, C. 1954 Les canalicules biliares. Etude au microscope electronique. C. R. SOC. Biol., 148: 2008-2011. 1956 Les canalicules biliares. Acta Anat. (Basel), 26: 94-109. Rouiller, C., and A. M. Jezequel 1963 Electron microscopy of the liver. In: The Liver. Vol. I. Ch. Rouiller, ed. Academic Press, New York, pp. 195-264. Sabatini, D. D., K. G. Bensch and R. J. Barrnett 1963 Cytochemistry and electron microscopy. The preservation of cellular ultrastructure and enzymatic activity by aldehyde fixation. J. Cell Biol., 27: 19-58. Steiner, J. W., and J. S. Carruthers 1961 Studies on the fine structure of the terminal branches of the biliarv tree. I. The morphology of normal bile canaliculi and bile productules. Amer. J. Path., 37: 639-650. Tanikawa, K. 1968 Ultrastructural aspects of the liver and its disorders. Springer, BerlinHeidelberg-New York. Yamamoto, T. 1965 Some observations on the fine structure of the intrahepatic biliary passages in goldfish. 2. Zellforsch., 65: 319-330.
PLATE 1 EXPLANATION OF FIGURE
1
Part of a liver lobule showing plates of hepatocytes ( E ) and sinusoids ( S ) . The half bile canaliculi (arrows) r u n along the cell surface forming a continuous network. Kupffer cells (K plus arrows) building into the lumen of the sinusoids are also evident. A few bundles of collagen fibers ( R ) are shown. x 4,132.5.
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 1
503
PLATE 2 EXPLANATION OF FIGURES
504
2
Hepatocyte surface showing bifurcating branches of a bile canaliculus. The lateral surfaces bordering the bile canaliculus possess small plicae ( P plus arrows) and a large circular infolding ( i plus arrow). x 6,250.
3
A bile canaliculus provided with a number of lateral short branches is shown on the surface of an hepatocyte; some granular cloudy material is evident over the borders of the bile channel (arrows). x 7,020.
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 2
505
PLATE 3 EXPLANATION O F FIGURES
506
4
Along the central areas of one side of an hepatocyte is sculptured a bile canaliculus provided with numerous microvilli. The smooth areas running along the margins of the canaliculus probably coincide with the limits of intimate contact between adjacent hepatocytes. The large holes (arrows) correspond to the stud-like projections of adjacent liver cells. x 7,700.
5
Bile canalicular lumen provided with numerous microvilli. Holes and invaginations are present on the wall surface of the biliary channel. A few microvilli (arrows) and small holes ( a plus arrows) are also evident along the margins of the canaliculus. Numerous microvilli populate the surfaces of the hepatocvtes frontinq the space of Disse ( S ) . Possibly reticular fibers ( R) are shown in the subendothelial spaces. x 7,250.
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 3
507
PLATE 4 EXPLANATION O F FIGURES
6
Hepatocytes with branching bile canaliculi (arrows) located over their surfaces. Besides a granular material ( X plus arrows) small and large holes are present o n the relatively smooth margins of the bile canaliculi. S, space of Disse. x 6,890.
7 Bile canaliculus crossing over two hepatocytes. Short diverticula provided with microvilli open i n the lumen of the bile channel (arrows). Pits and microvilli are also present on the surface of the liver cells adjacent to the bile canaliculus. The central area (A) shown in the picture corresponds to a minimal distance zone between the intercellular clefts of Disse ( S ) and the bile canaliculus. x 12,600.
508
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 4
509
PLATE 5 EXPLANATION OF FIGURES
510
8
View of a tortuous bile channel with numerous short microvilli projected obliquely or perpendicularly into the lumen. Small holes (arrows) open into the lumen of the biliary canaliculus. x 14,475.
9
The surface of the interior wall of a bile canaliculus clearly shows a number of large infoldings. They are provided with few microvilli and arise from the intracellular cortical areas of the cytoplasm (arrows). Microvilli, pits and small holes populate the lateral margins of the biliary channel. x 14,490.
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 5
51 1
PLATE 6 EXPLANATION O F FIGURES
512
10
In this stereo pair the lumen of a bile canaliculus is shown in three dimensional view. Numerous microvilli and intracellular extensions (on the left) populate the lumen of the biliary channel. Holes and microvilli are distributed over the surfaces of the hepatocytes ( o n the right). x 4,650.
11
This stereo view shows a bile canaliculus largely branched. An intracellular diverticulum opens into the lumen of the canaliculus (center). Large holes, corresponding to the stud-like infoldings are present along the lateral margins of the bile canaliculus. x 6,800.
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 6
513
ABSTRACT Bile canaliculi in the rat liver have been studied by scanning electron microscopy. They appear as half tubules carved out of the hepatocytes' surface. In several cases the bile channel bifurcates to form two or three long branches running on the same face of the cell. Therefore, by SEM it seems obvious that the same side of an hepatocyte may be used for bounding two or more bile canaliculi. Some bile canaliculi display a flexuous course and show lateral sacculations. Some of them are large and apparently similar to the short lateral branches of the bile canaliculus; others are smaller and arise obliquely from the subjacent cortical areas of the hepatocyte cytoplasm. These latter structures are best observed in stereo-views in which they appear as narrow intracellular projections bordered with a few microvilli, actually opening into the lumen of the bile canaliculus. The SEM results suggest that these sacculations probably correspond to short intracellular branches of the bile canaliculi. Zones of minimal distance (0.1 p ) between the space of Disse and the bile channel have been frequently observed. They are actually the zones where the intercellular clefts arising from the subendothelial space of Disse come into closest contact with the bile canaliculus and might serve as sites of simple diffusion of substances.
The architecture of the mammalian biliary system has been extensively studied by transmission electron microscopy (TEM) and it is well known that the bile canaliculi are intercellular spaces limited by two adjacent hepatocytes (Rouiller, '54; Fawcett, '55; Steiner and Carruthers, '61; Cossel, '64; Bruni and Porter, '65; Matter et al., '69). A much debated question of liver cytology concerns the existence in mammals of true intracellular bile canaliculi (Novikoff and Noe, '55). Recently their occurrence has been reported in some mammals (Carruthers and Steiner, '62; Barone and Finocchio, '63) and different vertebrates (David, '61; Yamamoto, '65) but denied by some authors (Rouiller, '56; David, '64; Tanikawa, '68). Further, it is also disputed whether there exists a direct communication between the biliary channels and the spaces of Disse (Fawcett, '55; Rouiller and Jezequel, '63; Cossel, '64; David, '64; Tanikawa, '68; Motta and Porter, '74). The purpose of this paper is to show by scanning electron microscopy (SEM) ANAT. REC., 182: 499-514.
some new topographic details of the surface structure of the bile canaliculus in the rat liver, as these bear on the above questions. MATERIAL AND METHOD
Liver tissue from five healthy, young, adult albino rats was perfused with a Tyrode solution through the left ventricle. The perfusion solution was used at room temperature and contained 0.1% procaine, a dab of heparin and bubbled 0%. Gravity was used to maintain a perfusion flow rate of about 60 mil/min. After about 30 seconds the flow of Tyrode solution was interrupted and 2.5% glutaraldehyde in cacodylate buffer solution (Sabatini, Bensch and Barrnett, '63) (0.18 M, pH 7.3) was perfused for about ten minutes. Small pieces of liver were subsequently excised and immersed in the same fixative. After a period of time between one and five days the tissues were washed for three to four hours in the same buffer and cut Received Sept. 12, '74. Accepted Mar. 18, '75.
499
500
PIETRO MOTTA AND GUIDO FUMAGALLI
into small blocks with a razor blade or carefully fragmented (pulled apart) with jeweler's forceps. Dehydration was carried out rapidly in a graded series of acetone and the specimens were transferred to liquid CO, for critical point drying (Porter et al., '72). The dried samples were mounted on aluminum studs using conductive silver paint, and coated with a layer of carbon and gold in a high vacuum evaporator (DV-502, Denton Vacuum). All the specimens were examined and photographed using a Cambridge Stereoscan Model S4 operated at 10 or 20 kV. OBSERVATIONS
As was shown in a previous publication (Motta and Porter, '74) the liver cells prepared by these techniques seem to separate along their surfaces rather than breaking open. Thus, the hepatocytes readily show the open half of canaliculi running along their sides. As seen with the SEM bile canaliculi form a continuous three dimensional network provided with irregular meshes (fig. 1). Generally, a single canaliculus is present along the central zone of one side of any one hepatocyte (fig. 1 ) . But in a number of cases the bile channel bifurcates in two or three branches that run on the same face of the cell (fig. 2). Some bile canaliculi possess a tortuous course and show frequent lateral sacculations and/or very short branches (fig. 3). The canalicular lumens are about 0.5-0.1 in width (figs. 4, 5). The surfaces limiting the half channel possess numerous short microvilli projecting obliquely or perpendicularly into the lumen (figs. 4, 5). The lateral short branches of the bile canaliculi show the same features as the principal channels (figs. 3, 7). Some of them are larger and apparently more similar to lateral diverticula or sacculations of the bile canaliculus (fig. 3); others are smaller and arise obliquely from the subjacent cortical areas of the cytoplasm (figs. 7, 8, 9 ) . This aspect is easily observed in the stereo pairs (figs. 10, 11) in which apparently an intracellular channel bordered by a few microvilli opens into the lumen of the actual intercellular bile canaliculus. The wall surface of these biliary channels
is not completely smooth; in addition to microvilli, they show small holes and/or round deep invaginations (fig. 8). The nature of these holes and invaginations has not been determined. The smaller ones could represent the openings in coated pits that are common on the surfaces of hepatocytes. Others could well represent vacuoles arising from the subjacent Golgi's zone. On some occasions a granular cloudy material was observed in the lumen of the bile canaliculi and/or on the sides of the cells adjacent to them (figs. 3, 6). On either side of the biliary channels there is an area running along the margins of such canaliculi that likely coincide with the limits of intimate contact between adjacent liver cells (figs. 1, 4, 5, 6, 7). These areas that appear as the smoothest parts of the cell surface vary greatly in width. Some are large and measure about 2.0-5.0 p. The narrowest zones correspond generally to the intercellular spaces between two adjacent hepatocytes. Such areas coincide with the zones where the intercellular space of Disse and its recesses appear in close contact with the wall of the bile canaliculi. We have never, however, found a direct communication between these two spaces (figs. 7, 10). At higher magnification even within these areas the surface of the hepatocytes is not absolutely smooth (figs. 4, 9 ) : they are simply regions where the microvilli are relatively short and scarce (Motta and Porter, '74). Further, a number of small holes and/or pits have been observed on these apparently smooth surfaces (fig. 5); there are also larger holes (deep invaginations) and cellular protrusions (figs. 2, 4, 11).These could correspond to the studlike projections into concavities in the adjacent hepatocytes and vice versa as revealed by transmission micrographs. DISCUSSION
TEM studies indicate that as a rule a bile canaliculus runs between each adjacent pair of cells (Rouiller and Jezequel, '63; David, '64; Tanikawa, '68) but that occasionally three hepatocytes bound the lumen of a single canaliculus (David, '64; Barone et al., '67; Tanikawa, '68). Moreover, a single side of the hepatocyte may frequently be provided with two or three
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI
branching separated bile canaliculi. Therefore, by SEM, it seems obvious that the same side of one hepatocyte facing two adjacent cells may be used for bounding two or eventually more bile canaliculi. It is also evident that bile canaliculi have a rather tortuous course and possess short lateral branches and/or sacculations. Such a branching pattern of bile canaliculi was fully appreciated by Novikoff and Noe (‘55) in sections of liver stained for alkaline phosphatase activity or also in histological preparations in which dyes were injected into the biliary tree (Barone and Finocchio, ’63). Some of these branches, e.g., those depicted in figures 2 and 3, could well represent the intercellular origins of the biliary canalicular tree. In this regard it is also important to note that the scanning micrographs (particularly the stereopairs) show short branches of the bile canaliculi, penetrating obliquely into the cytoplasm. The existence of true “intracellular bile canaliculi” introduced by Pfluger (1869) and von Kupffer (1876) was a much debated question in the past years; but it was widely accepted by a number of light microscopists (see the review of Novikoff and Noe, ’55). With the introduction of the electron microscope, intracellular extensions of bile canaliculi have been extensively described only by Barone and Finocchio (’63) but their occurrence has been repeatedly mentioned in other recent investigations (Carruthers and Steiner, ’62; Tanikawa, ’68; Matter et al., ’69). On the contrary Rouiller (’56) in his original electron microscopic study on the bile canaliculus concluded: “then it is possible that the majority of the structural features considered as intracytoplasmic bile canaliculi represent indeed a part of the Golgi apparatus. In addition their location near the bile canaliculi gives the impression that they are intracytoplasmic branches of bile canaliculi.” This concept seems to be largely accepted since in recent textbooks of Histology it is reported that the descriptions of intracellular branches of bile canaliculi penetrating the cytoplasm . . . .“seem to be a misinterpretation of vacuoles that often occur in the cytoplasm adjacent to the canaliculus and in con-
50 1
junction with the Golgi complex” (Bloom and Fawcett, ’68). The intracellular extensions found by SEM are rather short but always possess a lumen lined by a few microvilli similar to those lining the wall of the intercellular bile canaliculi. Although other techniques such as freeze fracture preparations might provide more evidence of the real existence of these intracellular canalicular projections, the present SEM observations (particularly the stereoviews ) seem to suggest that short branches arising from the bile canaliculi may penetrate the cytoplasm of perfused rat hepatocytes. One important point to be considered concerns the nature of the surfaces of the hepatocytes on either side of the bile canaliculus. In the present study the different types of junctional complexes were not clearly identified. But it is reasonable to suggest that the smoothest areas running along either side of the half bile channel may coincide with the limits of intimate contact between adjacent hepatocytes which in turn represent part of the junction a1 complex. In this regard it is also important to note that the SEM results show the spaces of Disse to be larger than one would gather from TEM pictures. In fact, they frequently extend among adjacent hepatocytes in deep recesses in close contact with the bile canaliculi (Motta and Porter, ’74). Further, a minimal distance of 0.1 was noted between the bile canaliculi and spaces of Disse in a study by TEM of serial sections of rat liver (Matter et al., ’69). Our observations suggest that regions of such minimal distance between the bile canaliculi and the recesses of Disse are not occasional findings but occur more numerously than is generally evident from transmission micrographs. In addition as observed by SEM, they are actually the regions where the intercellular recesses arising from the subendothelial spaces of Disse (Motta and Porter, ’74) come into closest contact with the bile canaliculus. Therefore it may be possible that substances which have the same concentration in the bile as in the blood (i.e., electrolytes and glucose) could be exchanged by simple diffusion (Brauer, ’63) in these narrow are as.
502
PIETRO MOTTA AND GUIDO FUMAGALLI ACKNOWLEDGMENTS
This work was partially supported by a Fulbright grant to Dr. Pietro Motta and undertaken in the Department of Molecular, Cellular and Development Biology, University of Colorado, Boulder, Colorado. The authors are grateful to Dr. Keith R. Porter for his generosity in making laboratory facilities available and for his helpful and most valuable suggestions. Sincere thanks are also due to Dr. A. J. Ladman for his interest and advises in editing the paper. The special support of the Italian Foreign Office was also greatly appreciated. LITERATURE CITED Barone, P., and F. Finocchio 1963 Ricerche morfologiche ed ultrastrutturali sui diverticoli intracellulari dei capillari biliari. Arch. Ital. Anat. Istol. Pat., 36: 366-379. Barone, R., C. Inferrera and G. Carrozza 1967 Aspetti ultrastrutturali dell'epatocita in rapporto alla secrezione biliare ed alle sue alterazioni. Relaz. Atti X Congr. SOC.Ital. Patologia Messina-Taormina, pp. 49-21 1. Bloom, W., and D. W. Fawcett 1968 A textbook of Histology, ninth ed. W. B. Saunders Co. Philadelphia-London-Toronto. Brauer, R. W. 1963 Liver circulation and function. Physiol. Rev. 43: 115-213. Bruni, C., and K .R. Porter 1965 The fine structure of the parenchymal cell of the normal rat liver. I. General observations, Amer. J. Path., 46: 691-755. Carruthers, J. S., and J. W. Steiner 1962 Fine structure of terminal branches of the biliarv tree. Arch. Path., 74: 39-48. Cossel, L. 1964 Die menschliebe Leber in Elektronenmikroskop. Gustav Fischer, Jena. David H. 1961 Zur submikroskopischen Morphologie intrazellularen Gallen-kapillaren. Acta Anat. (Basel), 47: 216-224. Submikroskopische Ortho-und 1964 Pathomorphologie der Leber. Akademie Verlag, Berlin.
Fawcett, D. W. 1955 Observations o n the cytology and electron microscopy of hepatic cells. J. Nat. Cancer. Inst., 15: 1475-1502. Kupffer, C. von 1876 Ueber Sternzellen in der Leber. Arch. f . mikr. Anat., 22: 353-385. Matter, W., L. Orci and C. Rouiller 1969 A study on the permeability barriers between Disse's space and the bile canaliculus. J. Ultrastr. Res., suppl. 11: 1-71. Motta, P., and K. R. Porter 1974 Structure of Rat Liver Sinusoids and Associated Tissue Spaces as revealed by Scanning Electron Microscopy. Cell Tiss. Res., 248: 111-125. Novikoff, A. B., and E. F. Noe 1955 Observations o n fragmented rat liver canaliculi and the problems of intracellular bile canaliculi and Golgi apparatus. J. Morph., 96: 189-211. Pfluger, E. 1869 Pfluger's Arch. Ges. Physiol., 2: 459 (quoted by Rouiller, '56). Porter, K. R., D. Kellev and P. M. Andrews 1972 The preparation of cultured cells and soft tissues for scanning electron microscopy. In: Proceedings of the 5th Annual Stereo-scan Colloquium. Kent Cambridge Scientific Co., Chicago. pp. 1-19. Rouiller, C. 1954 Les canalicules biliares. Etude au microscope electronique. C. R. SOC. Biol., 148: 2008-2011. 1956 Les canalicules biliares. Acta Anat. (Basel), 26: 94-109. Rouiller, C., and A. M. Jezequel 1963 Electron microscopy of the liver. In: The Liver. Vol. I. Ch. Rouiller, ed. Academic Press, New York, pp. 195-264. Sabatini, D. D., K. G. Bensch and R. J. Barrnett 1963 Cytochemistry and electron microscopy. The preservation of cellular ultrastructure and enzymatic activity by aldehyde fixation. J. Cell Biol., 27: 19-58. Steiner, J. W., and J. S. Carruthers 1961 Studies on the fine structure of the terminal branches of the biliarv tree. I. The morphology of normal bile canaliculi and bile productules. Amer. J. Path., 37: 639-650. Tanikawa, K. 1968 Ultrastructural aspects of the liver and its disorders. Springer, BerlinHeidelberg-New York. Yamamoto, T. 1965 Some observations on the fine structure of the intrahepatic biliary passages in goldfish. 2. Zellforsch., 65: 319-330.
PLATE 1 EXPLANATION OF FIGURE
1
Part of a liver lobule showing plates of hepatocytes ( E ) and sinusoids ( S ) . The half bile canaliculi (arrows) r u n along the cell surface forming a continuous network. Kupffer cells (K plus arrows) building into the lumen of the sinusoids are also evident. A few bundles of collagen fibers ( R ) are shown. x 4,132.5.
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 1
503
PLATE 2 EXPLANATION OF FIGURES
504
2
Hepatocyte surface showing bifurcating branches of a bile canaliculus. The lateral surfaces bordering the bile canaliculus possess small plicae ( P plus arrows) and a large circular infolding ( i plus arrow). x 6,250.
3
A bile canaliculus provided with a number of lateral short branches is shown on the surface of an hepatocyte; some granular cloudy material is evident over the borders of the bile channel (arrows). x 7,020.
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 2
505
PLATE 3 EXPLANATION O F FIGURES
506
4
Along the central areas of one side of an hepatocyte is sculptured a bile canaliculus provided with numerous microvilli. The smooth areas running along the margins of the canaliculus probably coincide with the limits of intimate contact between adjacent hepatocytes. The large holes (arrows) correspond to the stud-like projections of adjacent liver cells. x 7,700.
5
Bile canalicular lumen provided with numerous microvilli. Holes and invaginations are present on the wall surface of the biliary channel. A few microvilli (arrows) and small holes ( a plus arrows) are also evident along the margins of the canaliculus. Numerous microvilli populate the surfaces of the hepatocvtes frontinq the space of Disse ( S ) . Possibly reticular fibers ( R) are shown in the subendothelial spaces. x 7,250.
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 3
507
PLATE 4 EXPLANATION O F FIGURES
6
Hepatocytes with branching bile canaliculi (arrows) located over their surfaces. Besides a granular material ( X plus arrows) small and large holes are present o n the relatively smooth margins of the bile canaliculi. S, space of Disse. x 6,890.
7 Bile canaliculus crossing over two hepatocytes. Short diverticula provided with microvilli open i n the lumen of the bile channel (arrows). Pits and microvilli are also present on the surface of the liver cells adjacent to the bile canaliculus. The central area (A) shown in the picture corresponds to a minimal distance zone between the intercellular clefts of Disse ( S ) and the bile canaliculus. x 12,600.
508
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 4
509
PLATE 5 EXPLANATION OF FIGURES
510
8
View of a tortuous bile channel with numerous short microvilli projected obliquely or perpendicularly into the lumen. Small holes (arrows) open into the lumen of the biliary canaliculus. x 14,475.
9
The surface of the interior wall of a bile canaliculus clearly shows a number of large infoldings. They are provided with few microvilli and arise from the intracellular cortical areas of the cytoplasm (arrows). Microvilli, pits and small holes populate the lateral margins of the biliary channel. x 14,490.
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 5
51 1
PLATE 6 EXPLANATION O F FIGURES
512
10
In this stereo pair the lumen of a bile canaliculus is shown in three dimensional view. Numerous microvilli and intracellular extensions (on the left) populate the lumen of the biliary channel. Holes and microvilli are distributed over the surfaces of the hepatocytes ( o n the right). x 4,650.
11
This stereo view shows a bile canaliculus largely branched. An intracellular diverticulum opens into the lumen of the canaliculus (center). Large holes, corresponding to the stud-like infoldings are present along the lateral margins of the bile canaliculus. x 6,800.
SCANNING ELECTRON MICROSCOPY OF BILE CANALICULI Pietro Motta and Guido Fumagalli
PLATE 6
513
