The Liver: Biology and Pathobiology

By James L. Boyer and Nelson Fausto

In its Fifth Edition, this classic book retains its traditional strength of relating molecular physiology to understanding disease pathology and treatment as it explores the current state and future direction of hepatology.

Painstakingly revised, this edition includes 60 new chapters. As in previous editions, a section called Horizons summarizes advances of extraordinary nature in areas expected to have a substantial impact on hepatology. The Fifth Edition’s Horizons section includes emerging topics such as tissue engineering of the liver, liver-directed gene therapy, decoding the liver cancer genome, and imaging cellular proteins and structure.

To preserve essential background information which has not changed while making room for the panoply of major new contributions to understanding of liver disease, 14 chapters from the previous edition are freely available online at gastrohep.com.   To view these chapters visit - http://www.gastrohep.com/theliver/

• Language: English
• Publisher: Wiley
• Release date: Aug 24, 2011
• ISBN: 9781119964223

Book preview

1 Organizational Principles of the Liver

Joe W. Grisham

Department of Pathology and Laboratory Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

PRINCIPLES OF LIVER ORGANIZATION

The liver provides functions required to maintain homeostasis in the organism. To accomplish this the liver synthesizes numerous essential molecules of diverse sort; extracts and metabolizes a plethora of nutrients and xenobiotics brought into the body through the alimentary tract (and substances entering by other routes), as well as worn-out molecules and cells; stores, exports and/or excretes the metabolic products; and neutralizes numerous foreign antigens and microbes from the gut. These varied functions take place in a structurally complex, multicellular tissue with a unique angioarchitecture that has slowly evolved to its present form.

Major features of liver structure are a functional tissue (parenchyma) composed of at least seven distinct types of cell—hepatocytes, cholangiocytes, sinusoidal endothelial cells, macrophages, lymphocytes of several different phenotypes, dendritic cells, and stellate cells—that conjointly possess the capacities to synthesize, metabolize and eliminate a wide range of complex molecules and to carry out immune fnctions, all arranged in a matrix that facilitates their cooperative interaction. This cellular matrix is perfused with blood at low pressure and flow rate through uniquely structured capillary-size blood vessels, which are supplied by two sources of blood: (i) venous blood that has already circulated through the gut, pancreas, and spleen, is reduced in oxygen and pressure, and is enriched in nutrients and toxins absorbed from the gut and in viscerally generated hormones and growth factors; and (ii) arterial blood at systemic levels of oxygen, pressure, and composition.

The most fundamental feature of liver organization is a unique vascular pattern in which afferent (supplying) and efferent (draining) blood vessels of all sizes interdigitate uniformly, always maximally separated by parenchymal tissue and connected almost exclusively by the smallest capillary-size vessels (the sinusoids). Afferent blood vessels branch to form up to 8–10 orders of diminishing size from their entrance at the liver hilum; terminal portal veins, which supply blood to sinusoids, arise from the smallest two or three orders of preterminal portal veins. Sinusoids are interposed between afferent terminal portal veins and small efferent hepatic (central) veins, which collect sinusoidal blood and merge to form larger hepatic veins. This vascular pattern provides a large volume of blood at a high flow rate through large vessels with high compliance and capacity to supply the sinusoids at a low flow rate and pressure. Total liver blood flow is large only because there are myriad sinusoids.

Hemodynamic conditions resulting from this vascular pattern create watersheds or flow currents that segment the continuous mass of liver parenchyma into separate afferent and efferent vascular units at both macroscopic and microscopic levels. The liver parenchyma is not divided by connective tissue into identical capsule-delimited lobules, as is characteristic of all other glands, but is segmented only by hemodynamic patterns of afferent and efferent blood flow (see below). The smallest hemodynamic segments of parenchyma perfused with blood by sinusoids supplied by terminal afferent vessels and drained by small efferent vessels (the afferent and efferent microvascular segments often termed lobules, acini, etc.) are the parenchymal units in which the metabolic activity of the liver parenchyma is focused (see below). The human liver contains millions of virtually identical microvascular segments. Large afferent and efferent vessels hemodynamically divide the parenchyma into at least eight macrovascular segments, which enable the surgical resection of large portions (segments) of the blood-filled, sponge-like liver through hemodynamic fissures between afferent and efferent blood flows (see below).

PHYLOGENESIS AND EVOLUTION OF THE LIVER

The multiple functions of the mammalian liver are carried out by a combination of exocrine, endocrine, and paracrine/juxtacrine mechanisms involving the several types of cell listed above, as well as stromal cells (fibroblasts), and cells forming nerves and large blood vessels. The liver is the largest visceral organ in humans and other vertebrates, comprising generally 2–5% of the body weight in most species. It is a comparatively late evolutionary development, emerging more or less in concert with the vertebral column. Although the functions carried out by the various cellular components of the liver parenchyma are not limited to vertebrates, the aggregation of these different cells into the tissue of a single organ is unique. In invertebrates the liver’s functions are distributed among various cells and tissues dispersed throughout the body; glandular appendages to the gut (midgut or digestive glands) provide a few (but not all) of the metabolic functions of the vertebrate liver and pancreas [1]. A more liver-like gut appendage, associated with a vestigial portal circulation and some hepatocyte-like functions, occurs in the cephalochordate, amphioxus or lancelet (Branchiostoma sp.) [2, 3], a primitive chordate which occupies a proximate position in the evolution of vertebrates [4]. Development of a portal circulation may be the crucial event that enabled the formation of the liver in vertebrates [1]. The livers of mammals, birds, amphibians, and fish pass through similar developmental sequences, but vary structurally in adults [5]. In some cartilaginous fish, for example, the pancreatic and hepatic elements derived from the primitive foregut are incompletely separated; in primitive cartilaginous fish the biliary ducts regress in adults [6, 7]; and in bony fish there is a simpler ramification of blood vessels (and, consequently, vascular segments) and a different arrangement of hepatocytes and terminal bile ducts [8, 9].

EMBRYOLOGY OF THE LIVER

The evolutionary steps that eventuated in the emergence of the mammalian liver containing multiple types of functioning cell are obscure. To the extent that the ontogenesis of the liver mirrors its phylogenesis, the embryonic development of the mammalian liver suggests the way in which the aggregation of multiple types of cell into the hepatic parenchyma may have evolved (Figure 1.1).

Figure 1.1 Schematic depiction of some of the major events and cell–tissue interactions during the embryogenesis of the liver. Under influences (growth factors, etc.) from the developing heart (cardiac area) and from endogenous endothelial cells of the septum traversum, endordermal cells (the liver bud) migrate from the ventral foregut into the septum traversum, forming the primordial hepatoblasts. Hematopoietic tissue also migrates to the septum traversum from sites in the fetal yolk sac and the aortic-gonadal-mesonephros (AGM) region of the embryo. Hematopoietic tissue and hepatoblasts interact to further the development of both. Hepatoblasts that contact the mesenchyme around the emerging portal veins progressively change to form the ductal plates and, subsequently, cholangiocytes. A few hepatoblasts become liver stem cells. Late in the development of the fetal liver, in conjunction with the maturation of hepatocytes, hematopoietic tissue migrates from the liver to bone, but a few representatives of several hematopoietic cells—sinusoid endothelial cells, macrophages, immunocytes, and stellate cells—remain in the liver. Development of the liver is completed postnatally

c01_image001.jpg

Liver embryogenesis in mammals (see Chapter 2) begins in a few endodermal cells of the ventral foregut (the liver bud) located adjacent to the mesoderm of the septum transversum, which lies between the developing heart and the yolk sac and contains several types of mesenchymal cells and small blood vessels [10]. Interactions between epithelial cells of the foregut endoderm and mesenchymal cells of adjacent mesoderm direct the early development of the liver through the mediation of growth factors and cytokines [10, 11]. Hepatic commitment of endodermal cells of the ventral foregut is first signaled by the neoexpression of a group of genes for transcription factors and for liver-enriched or specific proteins, including α-fetoprotein and albumin [10]. Committed endodermal cells progressively invade the septum transversum, directed by competent endothelial cells in the latter [12], to form the first vestiges of the plate-like structure of the adult liver parenchyma, in which endodermal cells (hepatoblasts) that eventually become hepatocytes are separated by primordial sinusoids.

Bipotent hepatoblasts associated with endothelial cells in the septum transversum differentiate into both hepatocytes and cholangiocytes of the adult liver [10, 11, 13]. Hepatoblasts of humans express cytokeratins (CK) 8, 18, and 19 [14], and most of them amplify expression of CK 8 and 18 and mature to form hepatocytes in conjunction with cessation of CK 19 expression [14]. The relatively few hepatoblasts that touch the mesenchyme of the primitive portal tracts form a double-layer of cells termed the ductal plate, acquire the expression of CK 7 and amplify expression of CK 19, while continuing to express CK 8 and 18 [15], gradually gaining the form and function of cholangiocytes. Early cholangiocytes in ductal plates remodel to form separate ducts, while intervening ductal plate cells regress. Larger ducts are incorporated into the developing portal tracts and the smallest ducts connect to bile canaliculi in hepatic plates. Development of cholangiocytes and bile ducts from hepatoblasts is regulated by cytokines produced by portal mesenchyme [10, 13].

With the early migration of primitive hematopoietic tissue from the yolk sac and the embryonic aortagonad- mesonephros, the developing liver becomes the main hematopoietic organ of the mammalian embryo [16]. Cytokines produced by hematopoietic tissue promote the further development of the liver parenchyma and vice versa [10, 17]. Hematopoiesis migrates from the mammalian liver to the bone marrow during later stages of embryogenesis, but several cells derived from hematopoietic tissue—including macrophages, immunocytes, some sinusoidal endothelial cells, and possibly stellate cells—become permanent residents of the liver [18–20]. Small vessels in the septum transversum are the source of endothelial cells required for emergence of endodermal cells from the foregut [10, 12], but some endothelial cells originate from angioblasts in hematopoietic tissue [21]. Stellate cells have been posited to originate from such widely disparate sources as gut endoderm, mesoderm of the septum transversum, the neural crest, and/or hematopoietic tissue [22], but recent studies exclude a neural crest origin [23] and support an origin from hematopoietic tissue [24–26]. In fact, after transplantation of either bone marrow ([27], references S3, S8, S64 therein) or liver ([27], references S67–S69 therein), all of the liver’s non-epithelial cells are partially replenished from hematopoietic tissue in mammals, but they can all also proliferate locally (see below).

The pattern of emergence of the liver in fish, birds, and mammals is similar [4, 28, 29], but endothelial cells do not appear to direct the emergence of endodermal cells from the gut during development of the liver in zebrafish [29]. Furthermore, endothelial cells with scavenger activity are located in the gills and kidneys of cartilaginous and bony fish, and not in the liver as in all terrestrial animals [30]; the location of scavenger endothelial cells in the liver reflects a late step in the evolution of the mammalian liver. Nevertheless, the general pattern of expression of transcription factors and genes involved in liver development is conserved in all of these species [10], suggesting a common transcriptional strategy for assembling the liver. When this strategy first emerged awaits further genetic analysis of gut appendages in chordate ancestors of vertebrates.

ANATOMY OF THE LIVER

The liver of adult humans weighs from 1300 to 1700 g, depending on sex and body size. It is a continuous sponge-like parenchymal mass penetrated by tunnels (lacunae) that contain the interdigitating networks of afferent and efferent vessels [31].

The blood vessels and their investments of connective tissue provide the soft, spongy liver with its major structural support, or skeleton. Larger afferent vessels, portal veins, and hepatic arteries, are contained together in connective tissue—the portal tracts—which are continuous with the mesenchymal components of the liver’s mesothelium-covered surface capsule (Glisson’s capsule). Portal tract connective tissue is progressively diminished as the portal veins decrease in size. In humans and many other mammals, the smallest (terminal) portal vein is almost bare of connective tissue. (In contrast, connective tissue accompanies the terminal portal veins of adult swine and a few other animals, giving the false appearance of capsule-encased lobules.) Portal tracts also contain bile ducts, lymphatic vessels, nerves, and varying populations of other types of cells, such as macrophages, immunocytes, stellate cells, and possibly hematopoietic stem cells ([32] and references therein). The collagenous investment of the efferent vessels is less robust and lacks large numbers of adventitious cells.

The hepatic artery is distributed to the tissues of portal tracts, the liver capsule, and the walls of large vessels [33–36]. In portal tracts arterial branches form a capillary network (the peribiliary plexus) arborized around bile ducts [36, 37]. Efferent twigs from the peribiliary plexus empty into adjacent portal veins and sinusoids, forming an intrahepatic portal system [36, 37]. The portal vein supplies blood to the parenchymal mass only through its terminal branches [36–38]. Portal and arterial blood appear to be well mixed before entering sinusoids [38], but the direct supply of arterial blood to sinusoids by small branches of the hepatic artery is uncertain [33–37].

Glisson’s capsule, including portal tracts, contains extraparenchymal cells and tissues that make important contributions to liver function, including bile metabolism (cholangiocytes in bile ducts), vascular regulation (sympathetic nerves), pain perception (parasympathetic nerves), immune function (immunocytes), and lymph formation (lymph vessels). A large volume of lymph (up to half of all lymph) is produced in the liver, mostly in the ramifications of Glisson’s capsule [39]. Lymph vessels are not found in the parenchyma [40] and no satisfactory mechanism has been proposed that could separate the countercurrent flow of lymph in spaces of Disse and flow of blood in sinusoids with large fenestrations, although tracer studies suggest that lymph originates within the parenchyma [41].

In most mammalian species the liver is multilobed, the individual lobes reflecting the distribution of the major branches of afferent and efferent blood vessels. In contrast, the human liver parenchyma is fused into a continuous parenchymal mass with two major lobes, right and left, delineated only by being supplied and drained by separate first- and second-order branches of the portal and hepatic veins. Right and left lobes are topographically separated by the remnants of the embryonic umbilical vein (the falciform ligament), but this landmark does not locate the true anatomic division. Anatomically, the medial segment of the left lobe is located to the right side of the falciform ligament, centered on the anterior branches of the left portal vein. Interdigitation of first- and second-order branches of the portal and hepatic veins produces eight or more macrovascular parenchymal segments centered on large portal veins and separated by large hepatic veins [42]. Hemodynamic watersheds or fissures separating afferent and efferent macrovascular segments permit the surgical resection of individual or adjacent segments.

LIVER HEMODYNAMICS

The hepatic vasculature is characterized by high capacity, high compliance, and low resistance [43]. Blood vessels encompass about 22% of the liver’s mass/volume [44] and the liver contains about 12% of the total blood volume under physiological conditions [43], a sizeable fraction of which can be expelled by contraction of larger vessels by sympathetic nerve stimulation: the liver is a blood reservoir. The pressure of portal venous blood is reduced as the major afferent vessels dichotomize through the parenchyma, from about 130mm of water in the extrahepatic portal vein to about 60mm of water in the preterminal portal veins of the exteriorized liver of the anesthetized rat, amounting to about 60% of the total transhepatic pressure gradient [44]. A similar portal pressure gradient has been found in humans [43]. Blood flow through the liver amounts to about 1500–2000 ml minute−1 in adult humans, about 25% of the resting cardiac output [43]. About 25% of the total liver blood flow is derived from the hepatic artery at prevailing arterial pressure and oxygenation. The portal venous blood (about 75% of total liver blood flow) arrives at the liver partially depleted of oxygen and at a reduced pressure as a consequence of having already perfused the splanchnic viscera.

In aggregate, sinusoids comprise about 60% of the liver’s vascular volume, or about 13% of the total liver mass/volume [44]. A significant decrement in blood pressure occurs in sinusoids (about 40% of the transhepatic pressure gradient), the pressure declining to about 25mm of water in terminal hepatic veins of exteriorized liver of anesthetized rats [45]; the pressure gradient in the short sinusoids is especially steep. Blood pressure in the inferior vena cava approximates that in the terminal hepatic vein [43]. Consequently, although flow of blood through sinusoids faces little resistance, it is slow and somewhat intermittent and is assisted by negative pressure produced by respiratory expiration [43]. Possible mechanisms of regulating blood flow within sinusoids are controversial. Sinusoids appear to have limited contractile ability, possibly produced by contraction of encircling stellate cells (pericytes) [46, 47]. Studies of the exteriorized liver of rodents suggest that sinusoidal flow may be regulated at both inlet and outlet levels [46], but other similar studies have not detected sphincters at either point [45]. However, sinusoidal flow is strongly affected by postsinusoidal resistance [43].

Unlike capillaries elsewhere, liver sinusoids are composed of endothelial cells that are penetrated by holes (fenestrae) and lack a basal membrane [48], features that allow free egress of the fluid components and solutes of the perfusing blood. For example, tagged albumin has access to a space in the liver that is about 48% larger than the sinusoidal volume, in contrast to other tissues in which capillary space and albumin space are nearly the same [44].

HEMODYNAMIC MICROSEGMENTATION OF THE LIVER PARENCHYMA

Profiles of portal tracts and hepatic veins of various sizes are a prominent feature of liver histology [33–35]. Profiles of smaller branches of the afferent and efferent vessels (together with their stromal components) predominate in tissue sections taken from peripheral, subcapsular locations, whereas tissue sections taken from more proximal areas nearer to the hilum also contain larger vascular structures [35]. These vascular/stromal elements are contained in tunnels (lacunae) that penetrate the parenchymal mass [33]. While hepatic plates have been analogized as brick-like walls (muralia) of hepatocytes one cell (one brick) thick, this description somewhat oversimplifies the situation. Hepatic plates merge and branch frequently [41], and at branch points they are focally more than one cell thick [33, 49]. Furthermore, in livers undergoing growth or repair, hepatic plates are focally several cells thick at sites of hepatocyte proliferation before they are remodeled to a thinner structure [50, 51].

In histological sections of mammalian liver afferent and efferent vessels interdigitate regularly in an approximate ratio of 2–3 portal tract profiles for each profile of a hepatic vein, to form a pattern of cross-sections of portal tracts and hepatic veins separated by parenchyma [34, 35]. Most of the cross-sectioned portal tracts contain preterminal (penultimate) portal veins that represent the 7th–10th-order branches from the hilar portal vein in large mammals, such as humans. These small portal tracts and hepatic (central) veins penetrate the parenchyma in nearly parallel orientations about 0.5–1.0mm apart. The terminal portal veins (septal or inlet venules) branch from preterminal portal veins at points on the circumference of the latter separated by about 120 radial degrees (triradial branching) and penetrate the parenchyma approximately perpendicular to and midway between two adjacent terminal hepatic veins [34, 35]. (Afferent liver blood vessels branch one order more than do efferent vessels.) The disposition of terminal portal veins in two dimensions yields a roughly hexagonal pattern, forming the edges of the more-or-less hexagonal Kiernan lobule (see below); the portal tracts that contain the parental preterminal portal veins are positioned at alternating corners of the hexagons. This description of the distribution pattern of preterminal and terminal portal veins and of terminal hepatic veins is, again, somewhat idealized; the distribution of terminal portal veins in two dimensions actually forms patterns not limited to regular hexagons since not every terminal hepatic vein is bordered by exactly three portal tracts containing preterminal portal veins and not every terminal portal vein is of the same length.

During their course through the parenchyma terminal portal (inlet) veins break up completely into sinusoids, which are oriented more or less perpendicularly to the veins (Figure 1.2). Because they are hardly larger than sinusoids, terminal portal veins are not conspicuous in humans and other mammals that lack a connective sheath around them. However, in adult swine their course through the parenchyma is clearly marked by connective tissue.

Figure 1.2 Schematic diagram of the distribution of the terminal branches of the portal and hepatic veins and of the flow of blood through sinusoids that connect them. Terminal hepatic veins and the portal veins of several orders from which terminal portal veins arise penetrate the parenchyma more or less in parallel. Terminal portal veins branch from the parental vessels at points around the circumference about 120° apart and at somewhat different points along the length of the latter, to form a quasi-hexagonal pattern of distribution. Terminal portal veins branch dichotomously along their entire circumferences and lengths to form sinusoids that merge with terminal hepatic veins, which, in turn, connect with larger (sublobular) hepatic veins. Not depicted here, terminal twigs of the hepatic artery and intrahepatic bile duct (the latter as Canals of Hering) accompany the terminal portal veins without a connective tissue sheath in most mammals, including humans. Peripheral sinusoids near terminal portal veins are tortuous, becoming straighter in the remainder of their course toward terminal hepatic veins. Sinusoids branch and merge freely throughout their lengths to form a vascular sponge. Plates of hepatocytes (not shown) fill the spaces between sinusoids. Arrows depict the prevailing directions of blood flow from terminal portal veins through sinusoids into terminal hepatic veins. The relatively high blood pressure near terminal portal veins forms a hemodynamic barrier that directs sinusoidal flow into the nearest terminal hepatic vein where the pressure is at its lowest (see text). The sinusoidal network between adjacent terminal portal veins is the locus of the hemodynamic barrier that delimits microsegments of parenchyma. The segmentation of the parenchymal mass into microscopic units is a hemodynamic phenomenon produced by the in-flow sources located at the junctions of terminal portal veins and sinusoids, and sinks at the junctions of sinusoids and terminal hepatic veins

c01_image002.jpg

Capillary-size sinusoids occupy the smallest and most numerous tunnels (lacunae) in the parenchymal mass [33]. In favorably oriented histological sections more or less parallel, longitudinal profiles of sinusoids alternate with hepatic plates [49]. In other orientations, sinusoids may appear in histological sections as circular cross-sections located in a slightly larger circular cross-section of a small tunnel in the parenchymal mass. A narrow cleft, termed the space of Disse, separates sinusoids from hepatocytes located in adjacent hepatic plates [48, 52]. At their proximal (portal venous) ends, sinusoids are narrow and somewhat tortuous, whereas their middle and distal (hepatic venous) portions are larger and straighter [37, 53, 54]. Sinusoids and hepatic plates are disposed radially around the draining hepatic veins and extend more or less directly to the supplying terminal portal veins [54]. Microvascular segments (often incorrectly termed lobules) of the liver parenchyma are delineated by this pattern of distribution of microvessels and by the resulting hemodynamic properties of blood flow through them (alternating sources and sinks), which separate the flow of blood to individual microsegments. In humans and other mammals that lack connective tissue along the terminal portal veins, the parenchyma is segmented only by the hemodynamic barrier (vascular septum or watershed) formed by currents of blood flowing from the terminal portal veins into the sinusoids. These currents effectively separate the flow of blood in adjacent microvascular segments.

Concepts of the smallest structural and functional units of the liver have generated controversy for more than 400 years [55]. To a large extent this controversy has centered on efforts to determine the true lobular structure of the liver, when, in fact, the liver is a continuous parenchymal mass that lacks lobules, even in swine! Whether the correct equivalent of the liver lobule should be centered on the draining efferent vessels (as in the Kiernan lobule) [56] or on the supplying afferent vessels (as in the Mall lobule) [57] has consumed much argument [55]. The need for this controversy is eliminated when one realizes that the liver lacks true lobules; this focuses concepts and terminology on hemodynamically determined parenchymal microsegments.

In hemodynamic terms the classic hexagonal Kiernan lobule [56] is the efferent microvascular segment, being the smallest unit of parenchyma that is drained of blood by a single efferent (terminal hepatic or central) vein. The nature of the afferent microvascular segment, the smallest unit of parenchyma supplied with blood by an individual afferent (terminal portal) vein, has continued to be controversial, but recent studies have greatly calmed the controversy. The afferent microvascular segment, which is supplied with blood by a single terminal portal vein, was defined by Rappaport, who called this unit of parenchyma the liver acinus [34, 58]. More recent studies by Matsumoto and Kawakami [35, 59] and by Ekataksin and Wake [60] have refined the structure of the afferent microvascular segment of liver parenchyma that is supplied with blood by a single terminal portal vein. In their terminologies, the approximate unit of parenchyma encompassed by the Rappaport acinus is called the primary lobule [38, 59] or the hepatic microcirculatory subunit [60]. Matsumoto et al. defined the basis of the vascular septum, caused by the flow of blood from adjacent terminal portal veins into intervening sinusoids [35], which produces a hemodynamic barrier that separates the flow of blood in adjacent microvascular segments. An afferent microvascular segment (acinus, primary lobule, microcirculatory subunit) overlaps into the two neighboring efferent microvascular segments (hexagonal lobules) that center on adjacent terminal hepatic veins; and each efferent microvascular segment is composed of cone-shaped bits of parenchyma from six (on average) afferent microvascular segments. Ekataksin and Wake showed that the afferent microvascular segment is also the smallest unit of parenchyma drained of bile by terminal bile ducts [60], demonstrating that this hemodynamic segment is also the smallest excretory unit of parenchyma. Terminal bile ducts accompany terminal portal veins through the parenchyma, but the biliary epithelial cells in these tiny ducts are difficult to see in histological sections unless stained with an antibody to CK 19 [61]. Terminal bile ducts accompanying the terminal portal veins merge with hepatic plates to form the Canals of Hering, tiny tubules composed of both biliary epithelial cells and small hepatocytes [61].

THE CELLS OF THE LIVER PARENCHYMA

Hepatocytes are commonly denoted as parenchymal cells and the other cells of the liver tissue matrix as non-parenchymal cells. This convention is somewhat artificial since hepatocytes alone are not competent to perform all essential hepatic functions, and the several types of cell in the liver tissue matrix function as an integrated community to carry out conjointly the multiplicity of hepatic functions. Functional integration of this cellular community is accomplished by several communication mechanisms, including signaling networks involving numerous cytokines and chemokines, and by direct transfer of small molecules through gap junctions [62].

Hepatocytes, responsible for most of the synthetic and many of the metabolic functions of the liver (see Chapters 3–25), are large polygonal cells (averaging about 25–30μm in cross section and 5000–6000μm³ in volume [63]). They are the most numerous cells in the liver parenchyma; the adult human liver probably contains about 10−11 hepatocytes, representing about 60% of all cells in the parenchymal matrix and composing about 80% of its mass/volume [52]. Hepatocytes are shaped as complex rhomboids with several distinct surfaces [34] that comprise functionally distinct domains (see Chapter 6). About 35% of the total hepatocyte surface faces sinusoids, and the area of this surface is greatly amplified by the folding of the plasma membrane to form innumerable microvilli that extend into the space of Disse [63]. About 50% of the total hepatocyte surface faces adjacent hepatocytes [63]. The plasma membrane of these intercellular surfaces is mostly flat except where it is infolded to form bile canaliculi, which comprise about 13% of the total hepatocyte surface [63], also amplified by microvilli. The flat intercellular surface membranes contain intercellular adhesion complexes (tight junctions, intermediate junctions, and desmosomes) that pin together the adjacent hepatocytes and form a permeability barrier between the perisinusoidal space of Disse and bile canaliculi. Bile canaliculi form a belt-like extracellular space (about 1μm in diameter) that is continuous along the lengths of hepatic plates, connecting at the portal ends with bile ducts. The flat intercellular hepatocyte surface also contains gap junctions that allow communication between adjacent hepatocytes by transfer of small molecules.

Hepatocytes are polarized by molecular specializations of their various surface membranes in the forms of receptors, pumps, transport channels, and carrier proteins (see Chapters 6, 14, 21, 23 and 24). The canalicular membrane is modified for bile excretion, whereas the sinusoidal surface is equipped for extraction of a great variety of molecules from the blood and for the simultaneous secretion into the blood of other molecules that have been modified or synthesized by hepatocytes. As befits their numerous metabolic functions, hepatocytes contain a complex array of mitochondria (~1700 per cell on average), peroxisomes (~370 per cell), lysosomes (~250 per cell), Golgi complexes (~50 per cell), aggregates of rough and smooth endoplasmic reticulum (~15% of cell volume), and numerous microtubules/microfilaments [63].

Cholangiocytes comprise much less than 1% of the total number of cells in the liver parenchyma, since most are located in bile ducts in portal tracts [64]. Only the smallest bile ducts penetrate the parenchymal mass in the company of terminal portal veins, where they connect with bile canaliculi in hepatic plates. The points of connection of ducts with hepatic plates are defined by tubular structures, the Canals of Hering, composed of both small cholangiocytes and small hepatocytes [61], which are the location of liver epithelial stem cells that can differentiate into both hepatocytes and cholangiocytes [65, 66]. Larger bile ducts contain cholangiocytes that rest on a basal membrane and vary in number and size in proportion to duct size [64]. Their luminal surface membranes are expanded by microvilli and contain stereocilia, which signal from bile flow/content to cell cytoplasm [67]. Although they contain fewer mitochondria and sparser endoplasmic reticulum than do hepatocytes, cholangiocytes in intrahepatic bile ducts, together with the network of capillaries that surrounds them (the peribiliary plexus), form a metabolic unit that modifies the composition of canalicular bile by altering the content of water and solutes [68].

Endothelial cells of sinusoids comprise about 3% of the parenchymal mass/volume [52] and probably number about 3 × 10¹⁰ in an adult human liver. The thin cytoplasm of these flattened cells is penetrated by holes (fenestrations), each about 150–170 nm in diameter, that form groups termed sieve plates [48, 69]. Unlike other capillaries, sinusoids lack a basal membrane, but they are surrounded by a complex mixture of molecules, including collagens I, III, IV, V, and VI, laminin, heparin sulfate and dermatan sulfate proteoglycans, fibronectin, and chondroitin sulfate [70]. The unique structure of liver endothelium and sinusoids enables the free escape of fluid components of blood, but the matrix molecules located in the space of Disse can bind some molecules ([71] and references therein) and may retard the access of some solutes to hepatocyte surfaces. Sinusoidal endothelial cells are actively pinocytic and avidly clear effete proteins and colloids from the perfusing blood through several scavenger receptors, providing the main pathway for clearance of effete molecules from the circulation [69, 72]. (see Chapters 26–28).

Macrophages (Kupffer cells) comprise about 2% of the parenchymal volume/mass [51] and perhaps number about 2 × 10¹⁰/adult human liver. They are located within the lumens of sinusoids, most numerous in the portal regions, and are held in place by loose attachments to the sinusoidal endothelium [69]. Liver macrophages are avidly phagocytic through C3 and Fc receptors, clearing the sinusoidal blood of relatively large particulate materials including bacteria, effete cells (worn-out erythrocytes, dead or damaged hepatocytes, etc.) [69, 72]. Together with sinusoidal endothelial cells they form the organism’s major system for removing worn-out cells and proteins from perfusing blood. Activated macrophages produce a large number of chemokines and cytokines that have a fundamental role in the implementation of the liver’s acute phase reaction, coordinating the responses to injury of all of the parenchymal cells [73].

Immunocytes of the liver—T, NK, and NKT (and a few B) lymphocytes and dendritic cells—are components of a liver-centered immune system, largely segregated from the rest of the body’s immune system [19, 73]. The human liver is estimated to contain about 10¹⁰ lymphocytes of different phenotypes, located along sinusoids and in portal tracts [73]. The liver-centered immune system includes a major fraction of the body’s innate (native) immune capacity, as well as a small component of its acquired (adaptive) immune capacity [19, 73]. In addition to lymphocytes and dendritic cells, liver macrophages (Kupffer cells) and sinusoidal endothelial cells are also essential components of the liver-centered immune system [73]. Both liver macrophages and sinusoidal endothelial cells function as antigen presenters and both types of cell when activated secrete chemokines/cytokines that help stimulate the acute phase reaction and thymus-independent maturation of antigen-specific clones of T lymphocytes [19, 69, 73].

Liver immunity is involved in the removal/neutralization of numerous foreign antigens that reach the liver from the gut, including bacteria, particularly by the mechanisms of innate (native) immunity. In addition to the clearance of foreign antigens, the liver’s native immune system has a major regulatory role in the repair of the liver after cell injury and loss (see below). The small population of liver T lymphocytes, part of the body’s acquired (adaptive) immune capacity, are involved in virus elimination, clearance of activated T lymphocytes, and development of antigen tolerance [19, 73].

Stellate cells, located in the space of Disse outside of and partly encircling sinusoids (pericytes), comprise about 1.5% of the parenchymal volume/mass [52]. Stellate cells are multifunctional (see Chapters 28–29); together with hepatocytes they participate in the metabolism of vitamin A and store this fat-soluble vitamin (as retinyl esters) in lipid inclusions; this gave origin to previous designations as lipocytes, fat-storing, and/or vitamin A-storing cells [74]. Stellate cells synthesize, secrete, and degrade components of the perisinusoidal extracellular matrix [74]. They respond to several cytokines by becoming actively migratory and by acquiring a myofibroblastic phenotype, signaled by their expression of desmin, α-smooth muscle actin, and several neuroendocrine proteins [74]. As myofibroblasts, stellate cells have a major role in the fibrotic responses of the liver to injury of various sorts.

FUNCTIONAL AND STRUCTURAL HETEROGENEITY ALONG HEPATIC PLATES AND SINUSOIDS

Hepatic plates and adjacent sinusoids form associations that are structurally similar in all parts of the liver. Various liver cells show numerical, structural, and functional heterogeneities related to their location along the afferent–efferent axis of hepatic plates and sinusoids. Among the structural differences are ploidy variations in hepatocytes; in adult mammals hepatocytes located at the portal ends of hepatic plates are diploid, while cells of higher ploidy are located further downstream [75]. Gap junctions containing connexin 26 are more numerous on portal hepatocytes, whereas junctions containing connexin 32 are distributed on hepatocytes in all parts of hepatic plates [76]. In the portal regions, hepatic plates merge and branch frequently, accompanied by narrow, tortuous sinusoids [53, 54]. A larger fraction of stellate cells of the portal regions express desmin [74] and macrophages are larger and more numerous in this part of the parenchyma [73], whereas endothelial cell fenestrations are smaller and less numerous [48, 69]. Distribution of the molecular components of the perisinusoidal matrix also varies along the afferent–efferent axis of the space of Disse [50].

These variations in structure and cellular composition are associated with functional differences among hepatocytes located at different points along the afferent– efferent axis of plates–sinusoids. Rappaport divided the portal-hepatic (afferent–efferent) lengths of hepatic plates into three arbitrary zones (termed I, II, and III) and cited published studies documenting that hepatocytes located in these zones differ in their functional capabilities and susceptibilities to pathological damage [58]. Metabolic zonation of hepatocytes is strikingly exemplified by regional differences in carbohydrate metabolism (gluconeogenesis and glycogen storage by periportal hepatocytes; glycolysis by perihepatic vein hepatocytes), the enzymes of ammonia metabolism (carbamoyl phosphate synthetase is concentrated in periportal vein hepatocytes; glutamine synthetase is confined to a single ring of hepatocytes bordering terminal hepatic veins). Recent research has shown that many liver functions are dispersed heterogeneously, with dispersed functions often acting as integrated parts of coordinate metabolic systems [77].

Zonation of liver functions is related to sinusoidal hemodynamics, which produces gradients in blood-borne substances available to hepatocytes and other cells of the parenchymal matrix [44]. Hepatocytes and other cells located at afferent and efferent ends of hepatic plates are subjected to different microenvironmental conditions. Certain molecules are largely extracted by the first hepatocytes that encounter the perfusing blood, lowering their concentration downstream. For example, oxygen levels in the blood at afferent and efferent ends of sinusoids differ greatly because oxygen is efficiently extracted by hepatocytes located at the afferent ends of hepatic plates, exposing downstream hepatocytes to relatively hypoxic conditions; the oxygen gradient alone can explain much of the heterogeneity of hepatocyte function related to position in plates [78]. Other molecules modified or produced by upstream hepatocytes are excreted into the sinusoidal blood and may be removed by hepatocytes located further downstream. The complex interplay of metabolite concentration in the perfusing blood, coupled with extraction, modification, secretion, re-extraction, and further modification, influence the metabolic events that occur in individual cells and define unequal parenchymal territories that produce zonal variations in different physiological capabilities and pathological susceptibilities [44]. Differential concentration of certain growth factors (termed morphogens) by hepatocytes in various parts of the hepatic plates may be especially important in producing functional heterogeneity [79].

A countervailing hypothesis holds that functional heterogeneity of hepatocytes results from age-dependent differentiation, producing cells with hard-wired differences in metabolic capabilities. This hypothesis is based on the concept of the streaming liver [80], which posits that hepatocytes are born at the afferent ends of hepatic plates and mature functionally as they migrate proximodistally along the length of the plates, becoming functionally variant as they mature, losing the capacity to proliferate and senescing/dying at the efferent ends of the plates [75, 81], analogous to the lifespan sequence of enterocytes migrating along the length of the small intestinal cryptvillus. Although this hypothesis is superficially attractive, many studies of the kinetics of hepatocyte proliferation provide overwhelming evidence that hepatic plates do not constitute tracks along which hepatocytes regularly migrate during their life cycles [82]. All hepatocytes (including polyploid cells) can proliferate, obviating the possibility that functional heterogeneity results from their age-dependent differentiation and ultimate senescence.

TURNOVER OF LIVER CELLS

All liver parenchymal cells have a finite lifespan which can be shortened by physiological or pathological conditions that increase cell death and cell birth. Physiological turnover of hepatocytes occurs slowly with a lifespan of about 400 days in an adult steady-state hepatocyte population, about 0.025% of which typically will be engaged in DNA synthesis [83]. Cholangiocytes undergo turnover at similar rates.

Analysis of the specific lifespans of the other cells of the liver parenchyma is complicated by the fact that their number can be augmented both by local proliferation of cells resident in the liver populations [84] and by repletion from the bone marrow. Repletion of these hematopoietically-derived cells from bone marrow is evident in recipients of bone marrow transplants, in which they are replaced by cells of the new bone marrow genotype [27], and in recipients of liver transplants, in which these types of liver cell are replaced with cells of the host genotype [27]. In contrast, hepatocytes are not generated from bone marrow cells in significant number under either circumstance [27].

LIVER PARENCHYMAL REPAIR

Three distinct processes have evolved to produce the new hepatocytes needed to meet physiologically increased functional demand or to replace hepatocytes that are lost pathologically to trauma and toxicity. These processes center either on the temporary reactivation of cell cycle transit in fully differentiated, mitotically quiescent hepatocytes, with or without the coordinate proliferation and integration of other types of cells into a parenchymal matrix, and/or on the generation of entirely new hepatocyte lineages from adult liver stem cells (see Chapters 36, 38).

The most direct and rapid of these parenchymal augmentation/replacement processes involves the upregulation of hepatocyte birth in the absence of a preceding increase in hepatocyte death, often associated with increased hepatic functional demand due to physiological need [85]. Hyperplasia of hepatocytes by this mechanism enlarges the parenchymal mass and increases hepatocyte functional capacity. This process is regulated by the binding of ligands to hepatocyte nuclear receptors, of which nearly 50 have been identified [86]. Nuclear receptors are transcription factors that, when bound to ligands, directly up-regulate the combination of genes required to drive hepatocytes through the cell cycle [85, 86]. Several ligands for nuclear receptors (termed primary hepatocyte mitogens), including adrenal corticoids, bile acids, sex steroids, thyroid hormone, peroxisome proliferators, and 9-cis-retinoic acid directly stimulate the proliferation of hepatocytes and increase liver mass after binding to nuclear receptors [85]. Although it would seem that new endothelial cells would be needed to support the additional hepatocytes, no documentation of coordinate endothelial cell proliferation has been presented; it is possible that needed endothelial cells are derived from bone marrow.

Next in process complexity and in rapidity of response is the replacement of cellularly diverse liver parenchyma by the sequential proliferation of all of the component cells (hepatocytes, cholangiocytes, endothelial cells, macrophages, stellate cells, and immunocytes), and the melding of the new cells into tissue that closely reiterates the functional units of the undamaged liver [87]. This process, which is capable of replacing up to 70% of the parenchymal mass in mammals, is often called liver regeneration—a misnomer since in mammals the part of the liver removed surgically does not regenerate in the same way as body parts in certain lower animal species. Instead, the liver remaining after resection is enlarged by the formation of new tissue microunits that accurately correct deficient liver functions. (In contrast to the process in mammals, liver repair after partial hepatectomy in fish most intensively involves cells at the resection margin [88, 89] and may culminate in the regrowth (regeneration) of the resected tissue [88].)

The cell proliferation phase of this reparative process in mammals has been subjected to intensive kinetic and regulatory analyses ([87] and references therein). After tissue loss, residual hepatocytes are activated to proliferate within a few hours; hepatocyte proliferation begins at the portal ends of plates [83], and successive waves of hepatocyte proliferation ultimately involve virtually all residual hepatocytes [83, 90]. Hepatocyte replacement is followed sequentially by proliferation of sinusoidal endothelial cells and macrophages [83, 84], and the other cells of the parenchymal matrix. To the extent that it has been elucidated (see Chapters 36, 38), regulation of hepatocyte proliferation is effected by a complex mixture of cytokines and growth factors [87]. Most of the regulatory molecules are produced by various liver cells or released from storage sites within the liver [87], and many are components of the acute-phase reaction [91] and other elements of the liver’s native immune system [92, 93]. The less completely analyzed remodeling phase centrally involves endothelial cells and likely the other cells of the liver parenchyma. For example, proliferating hepatocytes initially form focal multicellular clumps [50, 51], which are cleaved into one-hepatocyte-wide plates by signaling from and separation by endothelial and stellate cells [50, 51]. The mechanism of liver tissue reformation at higher organizational levels (development of new afferent and efferent microvessels and establishment of new parenchymal microunits) is yet to be elucidated.

Although known regulatory mechanisms drive the reparative process, the mechanism that triggers the onset of repair after loss of liver tissue is still obscure. Since the liver vasculature must accept the entire portal blood volume, it has long been suspected that the trigger may be the massive increase in portal blood flow per unit of residual mass that follows loss of liver tissue [94]. Increased portal blood flow and pressure cause shear stress in sinusoids [94], which produces a burst of nitric oxide and prostaglandin production by sinusoidal endothelial cells, possibly providing the molecular trigger [95, 96]. Alternatively (or in concert), early activation of the nuclear receptor mechanism of hepatocyte proliferation may function as a trigger [97], and it is possible that multiple alterations in the physiological status of the liver remaining after tissue loss may converge to comprise a mass action trigger.

The mechanistically most complex (and slowest) process of liver parenchymal repair involves the generation of new hepatocyte and biliary epithelial cell lineages from adult liver stem cells (located in and around the Canals of Hering) [61, 65, 66], and the subsequent neoformation of liver functional units in a process that resembles aspects of liver embryogenesis. In this process, progeny of liver stem cells proliferate to form populations of poorly differentiated cells (termed oval cells in rodents) that have developmental analogies to embryonic hepatoblasts. In rodents, activation of hepatocyte formation through the oval cell reaction appears to occur only when liver repair from mature hepatocytes is blocked [65]. However, parenchymal repair after surgical resection of liver in teleost fish includes both hepatocyte proliferation and the generation of new hepatocyte lineages from stem cells [89]. The oval cell reaction in rodents is also regulated by elements of hepatic immunomodulation centered on the acute-phase reaction [98].

Although the subject of intensive scrutiny recently, there is no substantial evidence that hematopoietic stem cells are a significant source for the generation of hepatocytes or biliary epithelial cells in either humans or experimental animals [27]. This situation contrasts with the replenishment from hematopoietic sources of other cells of the liver parenchyma [27].

REFERENCES

1. Cornelius, C.E. (1985) Hepatic ontogenesis. Hepatology, 5, 1213–21.

2. Liang, Y.J. and Zhang, S.C. (2006) Demonstration of plasminogen-like protein in amphioxus with implications for the origin of vertebrate liver. Acta Zool, 87, 141–45.

3. Tian, J.X., Zhang, S.C., Liu, Z.H. et al. (2007) Characterization and tissue-specific expression of phosphatidylcholine transfer protein gene from amphioxus Branchiostoma belcheri. Cell Tissue Res, 330, 53–61.

4. Holland, L.Z., Albalet, R., Azumi, K. et al. (2008) The amphioxus genome illustrates vertebrate origins and cephalochordate biology. Genome Res, 18, 1100–11.

5. Elias, H. (1955) Origin and early development of the liver in various vertebrates. Acta Hepatol, 3, 1–567.

6. Sidon, E.W., Peek, W.D., Youson, J.H. and Fisher, M.M. (1980) Fine structure of the liver in the larval lamprey, Petromyzon marinus L; bile ducts and gall bladder. J Anat, 131, 501–19.

7. De Vos, R., De Wolf-Peters, C. and Desmet, V. (1973) A morphologic and histochemical study of biliary atresia in lamprey liver. Z Zellforsch, 136, 85–96.

8. Hinton, D.E. and Couch, J.A. (1998) Architectural pattern, tissue and cellular morphology in livers of fishes: relationship to experimentally-induced neoplastic processes, in Fish Ecotoxicology (eds T. Braubeck, D.E. Hinton and B. Streit), Birkh¨auser Verlag, Basel, pp. 142–64.

9. Hardman, R.C., Volz, D.C., Kullman, S.W. and Hinton, D.E. (2007) An in vivo look at vertebrate liver architecture: three-dimensional reconstructions from medaka (Oryzias latipes). Anat Rec, 290, 770–82.

10. Zhao, R. and Duncan, S.A. (2005) Embryonic development of the liver. Hepatology, 41, 956–67.

11. Zaret, K.S. (2007) Regulatory phases of early liver development: paradigms of organogenesis. Nat Rev Genet, 3, 499–512.

12. Matsumoto, K., Yoshitomi, H., Rossant, J. and Zaret, K.S. (2001) Liver organogenesis promoted by endothelial cells prior vascular function. Science, 294, 559–63.

13. Lemaigre, F. and Zaret, K.S. (2004) Liver development update: new embryo models, cell lineage control, and morphogenesis. Curr Opin Genet Dev, 14, 582–90.

14. Desmet, V.J., Van Eyken, P. and Sciot, R. (1990) Cytokeratins for probing cell lineage relationships in developing liver. Hepatology, 12, 1249–51.

15. Van Eyken, P., Sciot, R., Callea, F. et al. (1988) The development of the intrahepatic bile ducts in man: a keratin-immunohistochemical study. Hepatology, 8, 1586–95.

16. Durand, C. and Dzierzak, E. (2005) Embryonic beginnings of adult hematopoietic stem cells. Haematologica, 90, 100–8.

17. Kinoshita, T. and Miyajima, A. (2002) Cytokine regulation of liver development. Biochim Biophys Acta, 1592, 303–12.

18. Takezawa, R., Watanabe, Y. and Akaike, T. (1995) Direct evidence of macrophage differentiation from bone marrow cells in the liver: a possible origin of Kupffer cells. J Biochem, 118, 1175–93.

19. Doherty, D.G. and O’Farrelly, C. (2000) Innate and adaptive lymphoid cells in the human liver. Immunol Rev, 174, 5–20.

20. Murasawa, S. and Asahara, T. (2005) Endothelial progenitor cells for vasculogenesis. Physiology, 20, 36–42.

21. Zhang, N., Mustin, D. Reardon, W. et al. (2006) Blood-borne stem cells differentiate into vascular and cardiac lineages during normal development. Stem Cells Dev, 15, 17–28.

22. Geerts, A. (2001) History, heterogeneity, developmental biology, and functions of quiescent hepatic stellate cells. Semin Liver Dis, 21, 311–35.

23. Cassiman, D., Barlow, A. Borght, S.V. et al. (2006) Hepatic stellate cells do not derive from the neural crest. J Hepatol, 44, 1098–104.

24. Baba, S., Fujii, H., Hirose, T. et al. (2004) Commitment of bone marrow cells to hepatic stellate cells in mouse. J Hepatol, 40, 255–60.

25. Miyata, E., Masuya, M., Yoshida, S. et al. (2008) Hematopoietic origin of hepatic stellate cells in the adult liver. Blood, 111, 2427–35.

26. Forbes, S.J., Russo, F.P., Rey, V. et al. (2004) A significant proportion of myofibroblasts are of bone marrow origin in human liver diseases. Gastroenterology, 126, 955–63.

27. Thorgeirsson, S.S. and Grisham, J.W. (2006) Hematopoietic cells as hepatocyte stem cells: a critical review of the evidence. Hepatology, 43, 2–8.

28. Le Douarin, N.M. (1975) An experimental analysis of liver development. Med Biol , 53, 427–55.

29. Field, H.A., Ober, E.A., Roeser, T. and Stainier, D.Y.R. (2003) Formation of the digestive system in zebrafish. I. Liver morphogenesis. Dev Biol , 253, 279–90.

30. Seternes, T., Sørensen, K. and Smedsrød, B. (2002) Scavenger endothelial cells of vertebrates: a nonperipheral leukocyte system for high capacity elimination of waste macromolecules. Proc Natl Acad Sci USA, 99, 7594–97.

31. Elias, H. (1949) A re-examination of the mammalian liver, I. Parenchymal architecture. Am J Anat, 84, 311–33.

32. Kotton, D.W., Fabian, A.J. and Mulligan, R.C. (2005) A novel cell population in adult liver with potent hematopoietic reconstitution activity. Blood, 106, 1574–80.

33. Elias, H. (1949) A re-examination of the mammalian liver, II. The hepatic lobule and its relation to the vascular and biliary systems. Am J Anat, 84, 379–456.

34. Rappaport, A.M., Borowy, Z.J., Lougheed, W.M. et al. (1954) Subdivision of hexagonal liver lobules into a structural and functional unit. Role in hepatic physiology and pathology. Anat Rec, 119, 11–34.

35. Matsumoto, T., Komori, R., Magara, M. et al. (1979) A study of the normal structure of the human liver, with special reference to its angioarchitecture. Jikeikai Med J , 26, 1–40.

36. Mitra, S.K. (1966) The terminal distribution of the hepatic artery with special reference to arterio-portal anastomosis. J Anat, 100, 651–63.

37. Kardon, R.H. and Kessel, R.G. (1980) Threedimensional organization of the hepatic microcirculation in the rodent liver as observed in the scanning electron microscopy of corrosion casts. Gastroenterology, 79, 72–81.

38. Watanabe, Y., P¨uschel, G.P., Gardemann, A. and Jungermann, K. (1994) Presinusoidal and proximal intrasinusoidal confluence of hepatic artery and portal vein in rat liver: functional evidence by orthograde and retrograde bivascular perfusion. Hepatology, 19, 1198–207.

39. Trutmann, M. and Sasse, D. (1994) Lymphatics of the liver. Anat Embryol, 190, 201–9.

40. Yamamoto, K. and Phillips, M.J. (1986) Threedimensional observation of the intrahepatic lymphatics by scanning electron microscopy of corrosion cast. Anat Rec, 214, 67–70.

41. Ohtani, Y., Wang, B., Poonkum, R. and Ohtani, O. (2003) Pathways for movement of fluid and cells from hepatic sinusoids to the portal lymphatic vessels and subcapsular region in rat livers. Arch Histol Cytol , 66, 239–52.

42. Couinand, C. (1999) Liver anatomy: portal (and suprahepatic) or biliary segmentation. Dig Surg, 16, 459–67.

43. Lautt, W.W. and Greenway, C.V. (1987) Conceptual view of the hepatic vascular bed. Hepatology, 7, 952–63.

44. Goresky, C.A. (1963) A linear method for determining the liver sinusoidal and extravascular volumes. J Physiol , 204, 626–40.

45. Nakata, K., Leong, G.F. and Brauer, R.W. (1960) Direct measurement of blood pressure in the minute vessels of the liver. Am J Physiol , 199, 1181–8.

46. McCuskey, R.S. (2000) Morphological mechanisms for regulating blood flow through sinusoids. Liver, 20, 3–7.

47. Reynaert, H., Thompson, M.G., Thomas, T. and Geerts, A. (2002) Hepatic stellate cells: role in microcirculation and pathophysiology of portal hypertension. Gut, 50, 571–81.

48. Wisse, E., DeZanger, R.B., Charels, K. et al. (1985) The liver sieve: considerations concerning the structure and function of endothelial fenestrae, the sinusoidal wall and the space of Disse. Hepatology, 5, 683–92.

49. Grisham, J.W., Nopanitaya, W., Compagno, J. and Nagel, A.E.H. (1975) Scanning electron microscopy of normal rat liver: the surface structure of its cells and tissue components. Am J Anat, 144, 295–322.

50. Martinez-Hernandez, A. and Amenta, P. (1995) The extracellular matrix in hepatic regeneration. FASEB J , 9, 1401–10.

51. Wack, K.E., Ross, M.A., Zegara, V. et al. (2001) Sinusoidal ultrastructure during revascularization of regenerating liver. Hepatology, 33, 363–78.

52. Blouin, A., Bolender, R.P. and Weibel, E.R. (1977) Distribution of organelles and membranes between hepatocytes and nonhepatocytes in the rat liver parenchyma. A stereological study. J Cell Biol, 72, 441–55.

53. Miller, D.L., Zanolli, C.S. and Gumucio, J.J. (1979) Quantitative morphology of the sinusoids of the hepatic acinus. Gastroenterology, 76, 965–69.

54. Grisham, J.W. and Nopanitaya, W. (1981) Scanning electron microscopy of casts of hepatic microvessels, in Hepatic Circulation in Health and Disease (ed. W. Laut), Raven Press, New York, pp. 87–107.

55. Desmet V.J. (1999) Organizational principles, in The Liver: Biology and Pathobiology, 4th edn (eds I.M. Arias, J.M. Boyer, F.V. Chisari et al.), Lippincott Williams & Wilkins, Philadelphia, pp. 3–15.

56. Kiernan, F. (1833) The anatomy and physiology of the liver. Trans R Soc Lond, 123, 711–70.

57. Mall, F.P. (1906) A study of the structural unit of the liver. Am J Anat, 5, 227–308.

58. Rappaport, A.M. (1973) The microcirculatory hepatic unit. Microvasc Res, 6, 212–28.

59. Matsumoto, R. and Kawakami, M. (1982) The unitconcept of hepatic parenchyma—a re-examination based on angioarchitectural studies. Acta Pathol Jpn, 32 (Suppl 2), 285–314.

60. Ekataksin, W. and Wake, K. (1991) Liver units in three dimensions: I. Organization of argyrophil connective tissue skeleton in porcine liver with particular reference to the compound hepatic lobule. Am J Anat, 191 (2), 113–53.

61. Saxena, R. and Theise, N. (2004) Canals of Hering: recent insights and current knowledge. Semin Liver Dis, 24, 43–48.

62. Kmeic, Z. (2001) Cooperation of liver cells in health and disease. Adv Anat Embryol Cell Biol , 161, 1–151.

63. Weibel, E.R., St¨aubli, W., Gn¨agi, H.R. and Hess, F.A. (1969) Correlated morphometric and biochemical studies on the liver cell. 1. Morphometric model, stereologic methods, and normal morphometric data for rat liver. J Cell Biol, 42, 68–90.

64. Benedetti, A., Bassotti, C., Rapno, K. et al. (1996) A morphometric study of the epithelium lining the rat intrahepatic biliary tree. J Hepatol, 24, 335–42.

65. Thorgeirsson, S.S. and Grisham, J.W. (1997) Liver stem cells, in Stem Cells (ed. C.S. Potten), Academic Press, London, pp. 233–85.

66. Kuwahara, R., Kofman, A.V., Landis, C.S. et al. (2008) The hepatic stem cell niche: identification fy label-retaining assay. Hepatology, 47, 1994–2002.

67. Bing, Q., Masyuk, T.V., Muff, M.A. et al. (2006) Isolation and characterization of cholangiocyte primary cilia. Am J Physiol Gastrointest Liver Physiol , 291, G500–9.

68. Alpini, G., Ulrich, C., Roberts, S. et al. (1997) Molecular and functional heterogeneity of cholangiocytes from rat liver after bile duct ligation. Am J Physiol Gastrointest Liver Physiol , 272, G289–97.

69. Smedsrød, B., DeBleser, P.J., Braet, F. et al. (1994) Cell biology of endothelial and Kupffer cells. Gut, 35, 1509–16.

70. Martinez-Hernandez, A. and Amenta, P.S. (1993) The hepatic extracellular matrix. I. Components and distribution in normal liver. Virchows Arch A, 423, 1–11.

71. Schuppan, D., Schmid, M., Somasundaram, R. et al. (1998) Collagens in the liver extracellular matrix bind hepatocyte growth factor. Gastroenterology, 114, 139–52.

72. Smedsrød, B. (2004) Clearance function of scavenger endothelial cells. Comparative Hepatol, 3 (Suppl 1), S22.

73. Parker, G.A. and Picut, C.A. (2005) Liver immunobiology. Toxicol Pathol, 33, 52–62.

74. Friedman, Sl. (2008) Hepatic stellate cells: protean, multifunctional, and enigmatic cells of the liver. Physiol Rev, 88, 125–72.

75. Gupta, S. (2000) Hepatic polyploidy and liver growth control. Semin Cancer Biol , 10, 161–71.

76. Traub, O., Look, J., Dermietzel, R. et al. (1989) Comparative characterization of 21-kD and 26-kD junction proteins in murine liver and cultured hepatocytes. J Cell Biol , 108, 1039–51.

77. Jungermann, K. and Katz, N. (1989) Functional specialization of different hepatocyte populations. Physiol Rev, 69, 708–64.

78. Jungermann, K. and Kietzmann, T. (2000) Oxygen: modulator of metabolic zonation and disease of the liver. Hepatology, 31, 255–60.

79. Gebhardt, R., Baldysiak-Figiel, A., Kr¨ugel, V. et al. (2007) Hepatocellular expression of glutamatesynthetase: an indicator of morphogen actions as master regulators of zonation in adult liver. Prog Histochem Cytochem, 41, 201–66.

80. Zajicek, G., Oren, R. and Weinreb, M. Jr. (1985) The streaming liver. Liver, 5, 293–300.

81. Sigel, S.H., Brill, S., Fioreno, A.S. and Reid, L.M. (1992) The liver as a stem cell and lineage system. Am J Physiol , 263, G139–48.

82. Grisham, J.W. (1994) Migration of hepatocytes along hepatic plates and stem cell-fed hepatocyte lineages. Am J Pathol , 144, 849–55.

83. Grisham, J.W. (1962) A morphologic study of deoxyribonucleic acid synthesis and cell proliferation in regenerating rat liver; autoradiography with thymidine H3. Cancer Res, 22, 842–49.

84. Widmann, J.J. and Fahimi, H.D. (1975) Proliferation of mononuclear phagocytes (Kupffer cells) and endothelial cells in regenerating rat liver. Am J Pathol , 80, 349–66.

85. Columbano, A. and Ledda-Columbano, G.M. (2003) Mitogenesis by ligands of nuclear receptors: an attractive model for the study of molecular mechanisms implicated in liver growth. Cell Death Differ, 10, S19–21.

86. Nagy, L. and Schwabe, J.W. (2004) Mechanism of the nuclear receptor molecular switch. Trends Biochem Sci , 29, 317–24.

87. Michalopoulos, G.K. (2007) Liver regeneration. J Cell Physiol , 213, 286–300.

88. Sadler, K.L., Krahn, K.N., Gaur, N.A. and Ukomadu, C. (2007) Liver growth in the embryo and during liver regeneration in zebrafish requires the cell cycle regulator, uhrf1. Proc Natl Acad Sci USA, 104, 1570–75.

89. Okihiro, M.S. and Hinton, D.E. (2000) Partial hepatectomy and bile duct ligation in rainbow trout (Oncorhynchus mykiss): histologic, immunohistochemical and enzyme histochemical characterization of hepatic regeneration and hyperplasia. Toxicol Pathol, 28, 342–35.

90. Rabes, H.M., Wirsching, R., Tuczek, H.V. and Iseler, G. (1976) Analysis of cell cycle compartments of hepatocytes after partial hepatectomy. Cell Tissue Kinet , 9, 517–32.

91. Fulop, A.K., Pocsik, E., Brozik, E. et al. (2001) Hepatic regeneration induces transient acute phase reaction: systemic elevation of acute phase reactants and soluble cytokine receptors. Cell Biol Int , 25, 585–92.

92. Fausto, N. (2006) Involvement of the innate immune system in liver regeneration and injury. J Hepatol, 45, 347–49.

93. Dong, Z., Wei, H., Sun, R. and Tian, Z. (2007) The roles of innate immune cells in liver injury and regeneration. Cell Mol Immunol, 4, 241–52.

94. Sato, Y., Kayama, S., Tsukada, K. and Hatakeyama, K. (1997) Acute portal hypertension reflecting shear stress as a trigger to liver regeneration following partial hepatectomy. Surg Today (Jpn J Surg), 27, 518–26.

95. Schoen, J.M., Wang, H.H., Minuk, G.Y. and Lautt, W.W. (2001) Shear stress-induced nitric oxide release triggers the liver regeneration cascade. Nitric Oxide, 5, 453–64.

96. Schoen-Smith, J.M. and Lautt, W.W. (2005) The role of prostaglandins in triggering the liver regeneration cascade. Nitric Oxide, 13, 111–17.

97. Huang, W., Ma, K., Zhang, J. et al. (2006) Nuclear receptor-dependent bile acid signaling is required for normal liver regeneration. Science, 312, 233–36.

98. Thorgeirsson, S.S., Factor, V.M. and Grisham, J.W. (2004) Early activation and expansion of hepatic stem cells, in Handbook of Stem Cells, Introduction to Adult and Fetal Stem Cells, vol. 3 (eds R. Lanza, H. Blau, D. Melton et al.), Academic Press, San Diego, pp. 497–512.

2 Embryonic Development of the Liver

Roque Bort¹ and Kenneth S. Zaret²

¹Unidad de Hepatología Experimental, CIBEREHD, Centro de Investigación, Hospital Universitario La Fe, Valencia, Spain

²Cell and Developmental Biology Program, Fox Chase Cancer Center, Philadelphia, PA, USA

INTRODUCTION

The liver is one of the first organs to develop in the embryo and it rapidly becomes one of the largest organs in the fetus. The most essential function of the mammalian fetal liver is to provide a site for hematopoiesis. The early dependence of the fetus on its