THE ANATOMICAL RECORD 232:126-132 (1992) Pathways of Lymph Flow Through Intestinal Lymph Nodes in the Horse STEPHEN A. NIKLES AND TREVOR J. HEATH Department of Anatomy, University of Queensland, St. Lucia, 4067, Australia ABSTRACT In the horse, several thousand lymph nodes receive lymph from the intestine, part of which is very large and contains microorganisms that enable the animal to utilize refractory dietary constituents such as cellulose. The aim of this study was t o describe the pathways by which lymph is delivered into, traverses, and is drained from these lymph nodes. These pathways were studied with either Microfill or methacrylate casting materials and with light and electron microscopy. The afferent lymphatic vessel delivering lymph into one of the nodes divides over the capsular surface and within trabeculae into terminal branches, and these are continuous with the subcapsular and trabecular sinuses through rounded holes up to 30 pm across. Lymph is conveyed from the subcapsular and trabecular sinuses through the cortex by four types of sinuses: trabecular sinuses, cortical tubular sinuses, tubulelike sinuses with a network of stellate cell processes, and sinuses between cortical cords. It is conveyed through the medulla by sinuses both within and between medullary cords. Lymph is drained from these sinuses by initial efferent lymphatics of three types: those between medullary cords, those within the subcapsular sinus overlying medullary or cortical cords, and those within trabeculae. All three types are continuous with surrounding sinuses through holes 5-30 pm across. These three alternative routes for lymph drainage may ensure adequate lymph flow during different intranodal conditions that may exist when the node is responding to microorganisms or other foreign materials. Major differences exist in the pathways by which lymph traverses the lymph nodes of pigs, sheep, and rodents (Heath and Brandon, 1983; Heath et al., 1986; Heath and Spalding, 1987; Spalding and Heath, 1987). Some information is available from the horse, but only from the legs, on pathways by which lymph enters and leaves lymph nodes (Heath and Perkins, 1989; Perkins and Heath, 19901, but no information is available on the pathways by which lymph flows through the node parenchyma. This information could help provide a better understanding of how antigens, such as those derived from intestinal parasites and microorganisms and carried in the intestinal lymph, interact with cells of the immune system. This is of particular interest for the horse intestine because of the vast population of microorganisms in its huge cecum and colon, and because of the very large number, several thousand, of lymph nodes that are associated with the intestine (Schummer et al., 1981). Our aim was to investigate the pathways by which lymph is delivered into these lymph nodes by afferent lymphatic vessels, traverses the node within lymph sinuses, and drains from these sinuses into efferent lymphatic vessels. This was undertaken using Microfil and methacrylate casts and by light and electron microsCOPY. 0 1992 WILEY-LISS, INC MATERIALS AND METHODS Seven female and four male horses, 6-48-monthsold, were killed by shooting at a local knackery, and their intestines were removed. These horses were also used by Nikles and Heath (1990). An intestinal lymphatic vessel was cannulated and the lymph node on which it terminated was perfused with either yellow or white Microfil Rubber Injection Compound (Canton Biomedical Products, Boulder, CO), or Batson’s #17 Anatomical Corrosion Compound (Paul Valley Industrial Park, Warrington, PA) using methods described by Spalding and Heath (1987). Fixatives were perfused into the node through a cannula in a n afferent lymphatic. First, 0.5-1 ml of saline was injected under manual pressure over 15-20 minutes, followed by 0.25-0.5 ml of fixative containing 2% glutaraldehyde and 2% paraformaldehyde at pH 7.2, and then 0.25-0.5 ml of a second fixative containing 4% glutaraldehyde, over a total of 50-60 minutes. This Received September 5, 1990; accepted May 28, 1991. Address reprint requests to Professor T.J. Heath, Department of Anatomy, University of Queensland, St. Lucia, 4067, Australia. LYMPH PATHWAYS IN HORSE LYMPH NODES was followed by 0.01 ml of Evans Blue dye (0.05%) to mark the perfused area of the node. The slow perfusion rate was essential to avoid the distention of the lymph sinuses. All measurements were made on nodes fixed in this way. Immediately after the perfusion of fixative, the tissue was cut into slices 1-2 mm thick, then immersed in the second fixative for a further 24 hours. Tissues for electron microscopy were prepared a s described by Heath et al. (1986) and Spalding and Heath (1987) and examined in a Phillips 505 scanning electron microscope. Sections for light microscopy included some 7 pm thick stained with hematoxylin and eosin, and some 1 pm stained with toluidine blue. A total of 282 lymph nodes were examined. These came from the jejunum and ileum (26 nodes), cecum (64), right ventral colon (5), left ventral colon (381,pelvic flexure (4),left dorsal colon (12), right dorsal colon (16), descending colon (9), and near the root of the mesentery in the region of the cranial mesenteric artery (108). The latter region is referred to by Schummer et al. (1981) as the cranial mesenteric root. This term will be used here. RESULTS Studies with light and scanning electron microscopy and with Microfil casts revealed that each lymph node was composed of parenchyma surrounded by a capsule from which trabeculae extended inwards at about a right angle, and then branched extensively. The cortical parenchyma consisted of a relatively continuous peripheral layer which was continuous with a series of cortical cords. These cords, which contained fewer nodules, varied considerably in width. For example, they were 10-350 pm (87 16; n = 34) wide in one node from the cranial mesenteric root. It was not possible to identify clearly a separate paracortical region. The medullary parenchyma was arranged in cords which were generally smaller than the cortical cords; most were 20-50 pm across. Both cortical and medullary parenchyma was penetrated by a labyrinth of lymph sinuses. Lymph was delivered into these sinuses by afferent lymphatics and drained from them by efferent lymphatics (Fig. 1). * Afferent Lymph Pathways Afferent lymphatic vessels generally divided into 24 primary branches and less commonly into as many as seven (Fig. 1).Valves were present in the afferent vessels and in their branches that were more than 40 pm in diameter. Most branches remained on the node surface dividing progressively into terminal vessels, and these penetrated obliquely through the capsule. Sometimes the afferent vessels on the node surface were closely associated with efferent vessels. Other afferent lymphatics, up to 600 pm in diameter, penetrated deeply into the node within trabeculae, and then divided to yield terminal vessels some of which were 10 p m or less across (Figs. 1,2). Terminal afferent lymphatics were continuous with the surrounding sinuses through many rounded holes, and these varied from about 10-30 pm across (Fig. 2). 127 Cortical Lymph Sinuses The subcapsular sinus extended over the whole of the cortical parenchyma (Fig. 1)but varied in width from less than 10 pm to more than 200 pm. It was traversed by a complex network of stellate cell processes, which were generally 0.5-8 p.m across but were occasionally as wide as 13 pm. Sinuses of four other types conveyed lymph through the cortex. These were: Trabecular sinuses These surrounded trabeculae and penetrated into the cortical parenchyma, branching extensively and becoming progressively narrower. In one node, for example, they ranged in width from 275 pm at their origin from the subcapsular sinus to 15 pm in terminal regions. A network of processes, similar to that within the subcapsular sinus, pervaded the lumen (Fig. 2). Cortical tubular sinuses These sinuses were generally circular in cross section and about 10-30 pm in diameter. In one node from the cranial mesenteric root, 20 measurements revealed a mean width of 17 i 1.3 pm. The lumen contained only occasional stellate cell processes (Fig. 3), and the wall was punctuated by rounded holes, generally 6-10 pm across their major axis (Figs. 3, 4).Processes 1-4 pm across radiated outward from the sinus wall (Fig. 4). Cortical tubular sinuses were mainly distributed through the more-continuous parenchyma and were only occasionally present within the cortical cords (Fig. 1). They were often the predominant sinus in the vicinity of nodules. Tubule-like sinuses with a luminal network These sinuses differed from cortical tubular sinuses in having a complex network of luminal stellate cell processes, and in being larger. For example, in one node from the cranial mesenteric root they were 25-85 pm (43 i 3.6 pm n = 25) in diameter (Fig. 5).They were the predominant sinus within the inner regions of the cortex and appeared to be extensions of the sinuses between cortical cords. Sinuses between cortical cords Although occurring mainly between cortical cords some of these sinuses penetrated deeply into the cortical parenchyma (Fig. 1).They contained a complex network of luminal stellate cell process, and occasional trabeculae, up to 30 pm across. These sinuses varied considerably in size; in one node from the left ventral colon they were found to be 25-225 pm across. The lining of these sinuses was variable: in some places it appeared continuous, while in other there were many rounded holes, 1-9 pm or more across. Medullary Lymph Sinuses Lymph was conveyed through the medullary parenchyma by sinuses both within and between medullary cords. 128 S.A. NIKLES AND T.J. HEATH 1 CORTICAL TUBULAR SINUS I TRABECULAR SINUS PENETRATING AFFERENT SINUS WI MEDULLARY EFFERENT LYMPHATICS 2. FROM MEDULLARY SINUS 3. WITHIN TRABECULUM Fig. 1. Diagrammatic representation of the lymph pathways through an intestinal lymph node of the horse. This is based on light and electron micrographs, and Microfil casts. An afferent lymphatic vessel approaches the dorsal surface of the node and branches over a large portion of the surface. Some of these branches remain on the node surface, dividing into terminal vessels which deliver lymph into the subcapsular sinus. Others penetrate into the node within trabeculae and deliver lymph into trabecular sinuses. Lymph flows from the subcapsular and trabecular sinuses to the sinuses between cortical cords and medullary cords. It is conveyed there either directly or through cortical tubular sinuses or tubule-like sinuses with networks of stellate cell processes. Initial efferent lymphatics arise from sinuses between medullary cords, within trabeculae in sinuses between cortical or medullary cords, and from the subcapsular sinus overlying cortical and medullary cords. Only one of each type of initial efferent lymphatic is shown. Sinuses within medullary cords These were rounded in cross section and less than 10 pm across (8 f 2.2 pm, n = 11 in a node from the cranial mesenteric root) (Fig. 6). Occasional processes 0.5-3 pm across, extended across the lumen, and the sinus lining appeared continuous (Fig. 6B). great as 12 pm in diameter, extended across their lumen. The processes of surrounding sinuses were attached to the outer wall of these initial efferent lymphatics. Rounded holes, 5-25 pm across (16 2.0 pm n = lo), were present in the walls of these vessels. These holes were generally smaller than those in terminal afferent lymphatics, and they were present in greater numbers and distributed more extensively along the length of the vessel. As the efferent vessels increased in size, the numbers of holes in the wall and of luminal processes decreased, but valves appeared. The smallest vessel observed with valves was 90 pm in diameter. Sinuses between medullary cords These sinuses contained a network of luminal stellate cell processes, similar to that within the sinuses between cortical cords, with which these sinuses were continuous (Fig. 6A). The sinus width varied considerably: it was 70-200 pm in one node from the cranial mesenteric root. Initial and larger efferent lymphatics, and trabeculae, passed through some of these sinuses. Efferent Lymph Pathways Lymph was drained from lymph sinuses by initial efferent lymphatic vessels of three types. These were: Vessels between medullary cords These vessels were similar to those described by Heath and Perkins (1989). They were 40-100 pm across with processes, mostly 0.5-5 pm, but some a s * Vessels within trabeculae Some of the initial efferent lymphatics within trabeculae were as narrow a s 9 km across. They were continuous with the surrounding sinuses through rounded holes in the surface of the trabeculae; these were generally 5-30 pm (11 1.2 pm, n = 22) but occasionally as large as 85 km across their major axis. Luminal processes were present, but their frequency varied considerably between vessels. The initial efferents often formed complex labyrinths * LYMPH PATHWAYS IN HORSE LYMPH NODES 129 medullary and cortical cords. They were continuous with the sinus through holes in their wall, and these were similar to those in the initial efferent vessels described above (Fig. 8). After penetrating the capsule, the vessels joined with other efferent vessels and formed an interlacing network on the surface. DISCUSSION Afferent Lymph Pathways The pathways by which lymph from the intestine of the horse is delivered into lymph nodes by afferent lymphatics, then conveyed through the parenchyma in lymph sinuses and drained through efferent lymphatics, have some similarities to those in the pig and the sheep, but they are much more complex than the traditional concepts derived from work on rodents. In the horse, each intestinal lymph node, like each superficial node (Perkins and Heath, 1990), is generally supplied with lymph from a single afferent vessel. By contrast, the superficial nodes of the sheep and pig, the only ones that have been described in any detail, receive lymph from a number of different vessels: 6-12 in the sheep, and 1-4 in the pig (Heath and Brandon, 1983; Spalding and Heath, 1987). The terminal afferent lymphatics in the horse are continuous with the subcapsular and trabecular sinuses through rounded holes up to 30 pm across. In this respect the situation in the horse is similar to that in sheep (Heath and Brandon, 1983). However in the horse nodes, both more and larger vessels penetrate into the substance of the nodes. Furthermore, they penetrate deeply into the node in a manner reminiscent of that in the pig (Spalding and Heath, 1987).These penetrating afferent lymphatics in the horse have the potential to deliver a large proportion of the lymph deep within the node, and in the general vicinity of some initial efferent lymphatics. The effect of this on the phagocytic and immune capacity of the node can only be speculated upon. Lymph Sinuses Fig. 2. Scanning electron micrograph of a lymph node from the cranial mesenteric root showing a n afferent lymphatic (a) within a trabeculum (asterisks). A shows a terminal branch of the afferent lymphatic that has a hole about 10 pm across in its wall. Detail of this hole is shown in B. This hole allows communication between the vessel and the surrounding trabecular sinus (s). cp, cortical parenchyma. A, X 400; B, x 2,400. within the trabeculae before coalescing to form larger vessels (Fig. 7). Direct connections between the lumen of these larger vessels, up to 130 pm across, and the sinuses were observed. Valves were present in vessels as small as 40 pm in diameter. Luminal processes became less common as the vessels size increased up to 100 pm and were generally absent in still larger vessels (Fig. 7). These vessels emerged from the node over a large area of its surface, and occasionally from a deep groove in the node. Vessels within subcapsular sinus Initial efferent lymphatics, 12-25 pm across, drained lymph from the subcapsular sinus overlying The lymph flows through a labyrinth of sinuses of far greater complexity than in nodes that have been described from pigs, sheep, and rodents. Lymph from afferent lymphatics that has entered either the subcapsular sinus or trabecular sinus flows through the cortex via an extensive network of trabecular sinuses. It may then flow directly into the sinuses between cords of parenchymal tissue, or be further distributed throughout the cortical parenchyma within either the cortical tubular sinuses, or the tubule-like sinuses. Some lymph may also flow around the periphery of the node within the subcapsular sinus, and from there into sinuses between either cortical or medullary cords. The possible alternative pathways for lymph flow in rodents are more restricted (Sainte-Marie et al., 1982). In the rat, for example, Sainte-Marie and Sin (1968) reported that the subcapsular sinus occasionally yields narrow extensions that run across the medulla and transform into the wider and irregular medullary sinuses. In the superficial nodes of sheep, the subcapsular and trabecular sinuses, and the sinuses between medullary cords, are comparable to those described in S.A. NIKLES AND T.J. HEATH 130 Fig. 6. Scanning electron micrographs showing a region of medullary parenchyma in a lymph node from the left ventral colon. A shows sinuses (s) between medullary cords (m), which are traversed by a complex network of luminal stellate cell processes. A sinus within the medullary cord is indicated by the arrow, and B shows luminal detail of one of these sinuses. The sinus contains few luminal processes (p) and the sinus lining appears continuous. A, x 400; B, x 2,400. Figs. 3,4. Scanning electron micrographs of cortical tubular sinuses within the cortical parenchyma (cp) from a lymph node from the left ventral colon. Figure 3 shows a longitudinal section and Figure 4 a transverse section, with processes (p) radiating outward from the sinus wall. Arrows indicate holes up to 30 km across in the sinus wall. Figure 3, x 800; Figure 4, x 1,200. Figs. 3-5. Fig. 5. Scanning electron micrograph of a lymph node from the left ventral colon showing two tubule-like sinuses which are traversed by a network of stellate cell processes (asterisks), and a cortical tubular sinus which is smaller (arrow). This sinus does not contain luminal stellate cell processes. cp, cortical parenchyma. x 320. LYMPH PATHWAYS I N HORSE LYMPH NODES 131 Kurokawa and Ogata, 1980; Heath and Spalding, 1987). The tubular cortical sinuses of horses, though smaller, resemble those of sheep (Heath et al., 1987). They are also similar to the sinuses located within the medullary cords in the horse-sinuses which do not appear to have a counterpart in the other species that have been studied. The complex labyrinth of sinuses, coupled with the lack of structural compartmentalisation, should allow lymph that enters through the single afferent lymphatic to be distributed extensively throughout the parenchyma of the horse intestinal lymph node. This would have important implications for the ability of the node to remove and respond to foreign material carried from the intestine in the lymph. The real significance of this, and the relationship of this feature to the vast microbial population of the horse intestine, is not known. It appears that these features are not common to all species. In some, for example, the cortical parenchyma may be divided into functionally separate units. These include the “deep cortical units” or “physiological units” of the rat, each of which is related to the opening of the afferent lymphatic, the “lobular units” of the rabbit, and, on a somewhat broader scale, the node anlagen of the pig (Kelly, 1975; Sainte-Marie et al., 1982; Spalding and Heath, 1987). Functional studies using the pig and the rat have shown that lymph from each afferent lymphatic terminating on a node is regionally distributed within the parenchyma (Sainte-Marie et al., 1982; Spalding and Heath, 1987). It would appear that in these animals, in contrast to the horse, the lymph does not normally interact with all the parenchyma of the node, but only that in the region to which it is delivered by afferent lymphatics. Efferent Lymph Pathways Lymph is drained from lymph sinuses in the horse by initial efferent lymphatic vessels of three types: those Fig. 7. Scanning electron micrograph of an efferent lymphatic (e) within sinuses between medullary cords, those within from a lymph node from the right ventral colon, showing points of the subcapsular sinus overlying cortical or medullary entry of smaller efferent lymphatics (asterisks). c, capsule; s, sinuses cords, and those within trabeculae. These lymphatics between medullary cords; v, valve. x 200. are continuous with the surrounding sinuses through Fig. 8. Scanning electron micrograph of a lymph node from the rounded holes in their walls. Most of these holes are 5-30 pm across, and are smaller than those described cranial mesenteric root showing the initial efferent lymphatic vessel (e) within the subcapsular sinus (s).The vessel is continuous with the in the sheep (20-50 pm) and pig (80 pm or more) sinus through holes in its wall. c, capsule. x 760. (Heath and Spalding, 1987; Spalding and Heath, 1989). The initial efferent lymphatics within sinuses have thin walls but are supported by processes extending this paper (Heath et al., 1986; Heath and Spalding, from their outer walls into the network of the surrounding sinuses (Heath and Perkins, 1989). However, 1987). However, in contrast to the sheep, the cortex of the the initial lymphatics within trabeculae appear to be horse intestinal nodes has two types of tubular sinuses: well buttressed by the connective tissue of the trabeccortical tubular sinuses, which have a circular cross- ulum. These mechanisms should help keep the lumen section and few luminal processes, and tubule-like si- of the vessel open if the intranodal pressure rises as nuses, which have a complex network of stellate pro- may occur when the node responds t o foreign material carried t o it in the lymph. cesses in their lumen. These latter sinuses are 25-85 Initial efferent lymphatics arising within the node pm across, and so are larger than the cortical tubular sinuses; these vary between 10-30 pm across, and they have been reported from sheep (Heath and Spalding, can be readily distinguished with the scanning electron 19871,but not rats. In this species, as Sainte-Marie and Sin (1968) reported, the medullary sinuses join tomicroscope. In distribution, both types of sinus resemble “para- gether with the end-portion of the subcapsular sinus at cortical” and “tubular cortical” sinuses in rabbits, the hilus to form the efferent lymphatic vessels. The “mudstreams” of rodents, and “tubular” sinuses of pig, which differs from most other animals in the censheep (Sonderstrom and Stenstrom, 1969; Kelly, 1975; tral location of its nodular tissue, has some initial lym- S.A. NIKLES AND T.J. HEATH 132 phatics which are similar to those within trabeculae in the horse (Spalding and Heath, 1987). It is thought that in the pig these vessels were once on the node surface, but after fusion of node anlagen come to lie within remnants of the capsule in the enlarged, adult node (Spalding and Heath, 1989). There is no evidence for a similar origin for the initial efferent lymphatics within trabeculae in the horse. Furthermore, those within trabeculae in the horse are present in far greater numbers and form more complex networks than those in the pig. In the horse, the efferent lymphatics generally emerge over a large area of the node surface. This arrangement is comparable to that seen in the pig and the sheep, but it contrasts to that in the rat, where a single efferent lymphatic leaves the node at a clearly defined hilus (Sainte-Marie and Sin, 1968; Heath and Brandon, 1983; Spalding and Heath, 1987). ACKNOWLEDGMENTS We are grateful to Gary Godbold for help with the horses and the Director and staff of the UQ Electron Microscope Centre for their assistance. We thank the Queensland Equine Research Foundation for financial support. LITERATURE CITED Heath, T., and R. Brandon 1983 Lymphatic and blood vessels of the popliteal node in sheep. Anat. Rec., 207;461-472. Heath, T.J., R.L. Kerlin, and H.J. Spalding 1986 Afferent pathways of lymph flow within the popliteal node in sheep. J. Anat., 149: 65-75. Heath, T.J., and H.J. Spalding 1987 Pathways of lymph flow to and from the medulla of lymph nodes in sheep. J . Anat., 155:177-188. Heath, T.J., and N.R. Perkins (1989) Pathways between lymph vessels and sinuses in lymph nodes: A study in horses. Anat. Rec., 223,420-424, Kelly, R.H. 1975 Functional anatomy of lymph nodes. I. The paracortical cords. Inter. Arch. Allergy Appl. Immunol., 48:836-849. Kurokawa, T., and T. Ogata 1980 A scanning electron microscope study on the lymphatic microcirculation of the rabbit mesenteric lymph node: A corrosion cast study. Acta Anat., 107:439-466. Nikles, S.A., and T.J. Heath (1990) Pathways of lymph flow from the intestine of the horse. Anat. Rec., 229521-524. Perkins, N.R., and T.J. Heath 1990 Pathways of lymph flow from the superficial tissues in the legs of horses. Res. Vet. Sci., 48:119123. Sainte-Marie, G., F.S. Peng, and C. Belisle 1982 Overall architecture and patterns of lymph flow in the rat lymph node. Am. J. Anat., 164:275-309. Sainte-Marie, G., and Y.M. Sin 1968 Structures ofthe lymph node and their possible function during the immune response. Rev. Can. Biol., 27:191-207. Schummer, A., H. Wilkens, B. Vollmerhaus, and K.-H. Habermehl 1981 The Circulatory System, the Skin, and the Cutaneous Organs of the Domestic Animals. Verlag Paul Parey, Berlin, pp. 328-330,432-434. Sonderstrom, N., and A. Stenstrom 1969 Outflow paths of cells from the lymph node parenchyma to the efferent lymphatics. Scand. J. Haematol., 6t186-196. Spalding, H.J., and T. Heath 1987 Pathways of lymph flow through superficial inguinal lymph nodes in the pig. Anat. Rec., 217:188195. Spalding, H.J., and T.J. Heath 1989 The fine structure of lymph pathways in nodes from the superficial inguinal lymph centre in the pig. J . Anat., 166:43-54.