MICROSCOPY RESEARCH AND TECHNIQUE 37:497–508 (1997) Blood Flow Patterns in the Rat Pancreas: A Simulative Demonstration by Injection Replication and Scanning Electron Microscopy TAKURO MURAKAMI,1* TSUYOSHI MIYAKE,2 MARI TSUBOUCHI,1 YUTAKA TSUBOUCHI,1 AIJI OHTSUKA,1 AND TSUNEO FUJITA3 1Department 2Department 3Department of Anatomy, Okayama University School of Medicine, Okayama, 700 Japan of Internal Medicine, Okayama University School of Medicine, Okayama, 700 Japan of Anatomy, Niigata University School of Medicine, Niigata, 951 Japan KEY WORDS pancreas; microvascular architecture; microcirculation; corrosion casting; scanning electron microscopy ABSTRACT Scanning electron microscopy of vascular casts prepared by arterial injections of intentionally reduced amounts of resin showed that in the rat pancreas, the casting medium fills blood capillaries in the endocrine islets more promptly than those in the exocrine lobules and secretory ducts. Furthermore, the exocrine lobules containing endocrine islets allowed a more rapid resin flow through the insulo-acinar portal route than those lobules lacking an islet. The capillaries of secretory ducts were the last portions to be filled with resin. Since the resin used in this study was as viscous as blood and injected under a physiological pressure, the microcirculatory patterns demonstrated by the present method reflect the physiological flow pattern of blood in the pancreas. Microsc. Res. Tech. 37:497–508, 1997 r 1997 Wiley-Liss, Inc. INTRODUCTION Complete or sufficient injection of low viscosity casting media into arteries reproduces the whole extent of blood vascular beds, including thick venous portions (Murakami, 1971; Murakami et al., 1973). Incomplete or insufficient injection of such media into arteries partially reproduces the vascular beds; the arteriolecapillary, capillary-venular or venous systems remain unfilled (Murakami et al., 1983). Similar injection into veins can reproduce the venous or veno-capillary system (Murakami et al., 1983). Thorough injection into veins is not suited for casting a whole blood vascular bed, since it causes rupture on the capillary level (Murakami et al., 1983). A combination of adequate casting methods, both complete and partial, thus, allowed a precise scanning electron microscopic analysis of the fine vascular arrangement or microcirculatory patterns in the kidney, cerebral hypophysis and other organs (Murakami, 1971, 1972; Murakami et al., 1983, 1987, 1993a), including the pancreas (Fujita and Murakami, 1973; Murakami and Fujita, 1992; Murakami et al., 1992, 1993b, 1994). This paper describes some findings in the pancreas, especially those obtained by complete and incomplete arterial injections (Miyake et al., 1992). The target organ of the present study is the rat pancreas, since its microcirculatory pattern has raised much controversy and discussion (Bonner-Weir and Orci, 1982; Ohtani et al., 1986), including a complicated insulo-acinar portal r 1997 WILEY-LISS, INC. system which additionally has separate or collateral venous drainage (Miyake et al., 1992; Murakami and Fujita, 1992). MATERIALS AND METHODS Fourteen adult male Wistar rats weighing 350–360 g were anesthetized with ethyl ether, and their thoracic aorta was cannulated at the level of the seventh thoracic vertebra. The animal’s circulatory system was then rinsed by perfusion with Ringer solution (15.0–20.0 ml) and injected with a semi-polymerized and diluted low viscosity methacrylate casting medium (Murakami, 1971; Murakami et al., 1973) at an injection pressure of 120–130 mm Hg. As a substitute for this laboratory-prepared medium, a commercially available casting medium (Mercox 2R or 2B, Oken shoji, Tokyo, Japan) was sometimes used. This medium was also diluted with 20–30% monomeric methyl methacrylate (Katayama Kagaku, Osaka, Japan) prior to the injection. The casting media used in the present study were thus as viscous as blood (or had a viscosity of about 4.0 centipoise at room temperature). *Correspondence to: Prof. Takuro Murakami, Section of Human Morphology, Department of Anatomy, Okayama University School of Medicine, 2–5–1 Shikatacho, Okayama 700 Japan. Received 20 September 1994; revised 6 April 1995; accepted 31 May 1995. 498 T. MURAKAMI ET AL. Fig. 1. Low power scanning electron micrograph of the thoroughly replicated blood vascular bed of the adult rat pancreas (20 ml resin injection through the thoracic aorta). Complete injection of low viscosity casting medium through the thoracic aorta reproduces the whole extent of the pancreatic blood vascular bed. BD, common bile duct; EL, exocrine lobule. Bar, 1.0 mm. MICROCIRCULATORY PATTERNS IN THE RAT PANCREAS Fig. 2. Thoroughly replicated blood vascular bed of the adult rat pancreas. The pancreatic blood vascular bed mainly consists of the capillary networks of the exocrine lobules (EL), endocrine islets (EI and II) and secretory ducts (SD). The endocrine islets are either 499 embedded as the intralobular islets (II) in the exocrine lobules, or as the interlobular (extralobular) islets (EI) in the interlobular tissue spaces or along the secretory ducts. A part of this figure was shown elsewhere (Miyake et al., 1992). Bar, 500 µm. 500 T. MURAKAMI ET AL. Fig. 3. Two exocrine lobules (EL1 and EL2) isolated from a thoroughly replicated specimen. The EL1 lobule contains an intralobular islet (II). This islet issues no insulo-venous efferent vessel, but emits some marked insulo-acinar portal vessels (arrowheads) directly draining into the lobular capillaries. The EL1 lobule has thus a dual blood supply: by the portal vessels and by the acinar branches (la) of the lobular artery. The EL2 lobule contains no islet. IA, interlobular artery; IV, interlobular vein; SD, secretory duct; ia, islet afferent artery (insular branch of the lobular artery). Bar, 100 µm. The amount of the casting medium used for injection was changed as follows: 17.0–20.0 ml of the casting medium was perfused in four rats; 8.0–11.0 ml in four rats; 3.0–6.0 ml in four rats; 0.5–1.0 ml in the remaining two rats. In some animals of the latter three groups, the hepatic portal vein was cut off closely below the portal spiral valve (Booz, 1964) prior to the perfusion with Ringer solution. The resin-injected animals were placed for 1 hour in a hot water bath (60°C), and then the pancreas was isolated ‘‘en bloc’’ together with the liver, stomach, duodenum, spleen and kidneys. The isolated organs were immersed in a hot 10% NaOH solution (60°C) overnight, and washed for 8 hours in running tap water. This NaOH treatment followed by washing in tap water was repeated several times. The vascular casts of the pancreas and other abdominal organs were frozen in distilled water, freezedried, and dissected with forceps or needles. The pan- creatic casts were then microdissected with sharpened forceps and needles under a binocular light microscope, stained with osmium vapor (Murakami et al., 1973), coated with gold in a vacuum evaporator, and observed with a scanning electron microscope (S-2300, Hitachi) using an acceleration voltage of 5 KV. RESULTS The injection of 17.0–20.0 ml resin produced complete vascular casts of the kidney, stomach, duodenum, spleen, pancreas and liver. Incomplete resin injection allowed only partial replication of the blood vascular beds of these organs. In general, injection of 0.5–1.0 ml resin reproduced only the arterial systems in the duodenum and spleen as well as arterial and arterio-capillary systems of the gastric mucosa. MICROCIRCULATORY PATTERNS IN THE RAT PANCREAS 501 Fig. 4. A lobule (EL) containing an intralobular islet (II). This lobule is supplied both by the insulo-acinar portal vessels (arrowheads) and by the acinar branch (la) of the lobular artery. In addition to the portal vessels, the islet issues a collateral venous (insulo- venous) efferent vessel (iv) draining into the lobular vein. IA, interlobular artery; IV, interlobular vein; SD, secretory duct; ia, islet afferent artery. A part of this figure was shown elsewhere (Miyake et al., 1992). Bar, 150 µm. In the kidney, the glomeruli were reproduced almost completely by the 1.0 ml resin injection. The 3.0–6.0 ml resin injection sufficiently reproduced the arterial systems of the duodenum and spleen, and the capillary plexus of the gastric mucosa. The kidney glomeruli, including their efferent vessels, were completely reproduced by the 3.0–6.0 ml resin injection. When 8.0–11.0 ml of resin was used for injection, the vascular plexuses of the stomach and kidneys were thoroughly reproduced, though those of the duodenum and spleen were insufficiently replicated. 502 T. MURAKAMI ET AL. Fig. 5. An isolated interlobular islet (EI). The interlobular islet has a fine (capsular) capillary meshwork (arrowheads), which is homologous with the lobular capillary network. IA, interlobular artery; IV, interlobular vein; SD, secretory duct; ia, islet afferent artery; iv, islet efferent vein (insulo-venous efferent vein). Bar, 100 µm. In the thoroughly replicated specimens (i.e. when injected with 17–20 ml of resin), it was clearly confirmed that the rat pancreas contains a rich blood vascular bed which mainly consists of a lobular capillary network supplying the exocrine lobules, insular capillary networks supplying the endocrine islets of Langerhans, and ductal capillary networks supplying the pancreatic ducts (Figs. 1, 2). The capillary networks of the endocrine islets were more or less conglomerated. The islets thus identifiable in the cast specimens were observed as intralobular islets embedded in the lobular capillary network (Figs. 2–4), or as interlobular (extralobular) islets located among the lobular capillary networks or along a ductal capillary plexus (Figs. 2, 5). The intralobular islets issued insulo-acinar portal vessels draining into the lobular capillary networks (Figs. 3, 4). These intralobular islets were distributed in relatively small or thin lobules, so that they usually issued some collateral venous (insulo-venous) efferent Fig. 6. Incompletely replicated adult rat pancreas (1.0 ml resin injection through the thoracic aorta). The capillary networks of the islets (arrowheads) are reproduced, whereas those of the lobules and secretory ducts are not replicated. IA, interlobular artery. Bar, 300 µm. MICROCIRCULATORY PATTERNS IN THE RAT PANCREAS 503 Fig. 7. An intralobular islet (II) isolated from the cast prepared by 1.0 ml resin injection. In the islet, resin filling begins from the superficial aspects. The islet efferent vein (iv) arises from the deep (thick arrow) and superficial (thin arrow) efferent rootlets in the islet. Arrowhead indicates an insulo-acinar portal vessel. IA, interlobular artery; ia, islet afferent artery. Bar, 100 µm. vessels directly draining into the lobular veins (Fig. 4). The interlobular islets issued only venous (insulovenous) efferent vessels which drained into the ductal or interlobular veins (Fig. 5). In this manner, many exocrine lobules contained one or more well-distinguishable islets, and received dual blood supply from the islets via the insulo-acinar portal vessels and from the acinar arteries. With the 0.5–1.0 ml resin injection, the insular capillary networks were almost sufficiently reproduced, whereas the capillary networks of exocrine lobules and ducts remained unfilled (Fig. 6). The insulo-venous efferent and insulo-acinar portal vessels of the islets were insufficiently replicated with the 0.5–1.0 ml resin injection. With this injection, however, it was clearly noted that the arterially injected resin reached the islets from their superficial aspects, and that the resin after reaching the islets moved into the insulo-acinar or insulo-venous efferent vessels from their superficial and deep aspects (Fig. 7). With the 3.0–6.0 ml resin injection, the insular capillary networks were completely reproduced, whereas Fig. 8. Incompletely replicated rat pancreas (5.0 ml resin injection). The islets (arrowheads) and their venous efferent (insulovenous) systems are promptly filled with resin, so that the related systemic venous vessels (V), including the interlobular veins (IV), are also replicated. IA, interlobular artery. Bar, 500 µm. the capillary networks of exocrine lobules and ducts still remained unfilled (Figs. 8–10). The insulo-venous efferent and insulo-acinar portal vessels of the islets were sufficiently replicated with this injection. The venous branches continuous with the insulo-venous efferent vessels as well as the lobular capillaries supplied by the insulo-acinar portal vessels were also partially replicated (Fig. 9). 504 T. MURAKAMI ET AL. Fig. 9. An intralobular islet (II) and its efferent vessels replicated by 5.0 ml resin injection (closer view of a part of Figure 8). The insulo-venous efferent vessels (iv) are sufficiently reproduced together with their connecting lobular or interlobular veins (V). Replication of the insulo-acinar portal system is insufficient; the lobular capillaries continuous with the portal vessel (arrowhead) are not reproduced. Bar, 100 µm. With the 8.0–11.0 ml resin injection, the islets and their insulo-venous efferent and insulo-acinar portal vessels were sufficiently reproduced together with the lobular capillaries directly supplied by the insulo-acinar portal vessels (Figs. 11, 12). With this injection, replication of the lobular capillaries supplied by the acinar arteries was still insufficient; the same was true for the ductal capillary plexus. The venous efferent vessels of the islets were usually filled with larger amount of resin than insulo-acinar portal vessels, being sufficiently reproduced together with their connecting intralobular or interlobular veins. In these veins, some other peripheral branches were usually reproduced by the retrograde flow of resin (Fig. 11). This back flow of resin into the peripheral veins typically occurred in specimens in which the hepatic portal vein was neither cut nor opened. Replication of the ductal capillary networks was insufficient with the 8.0–11.0 ml resin injection. Suffi- cient casting of these ductal networks was noted in the 17.0–20.0 ml resin injections (Figs. 3–5). DISCUSSION The present results confirm that thoroughly replicated specimens are useful for a precise morphological analysis of blood vascular bed under the scanning electron microscope. Moreover, they indicate that incompletely replicated specimens (specimens prepared by injections with intentionally reduced amounts of resin) are useful to study the inflow modes of the casting medium into the blood vascular bed as well as its outflow modes from the vasculature. Our casting medium was as viscous as blood and perfused under a physiological pressure. Our previous experiments by intravital microscopy using the mesenterium of ethyl ether-anesthetized adult rats have shown that our medium flows through mesenterial capillaries in a fashion similar to that of blood MICROCIRCULATORY PATTERNS IN THE RAT PANCREAS Fig. 10. An interlobular or periductal islet (EI) replicated by 6.0 ml resin injection. The islet is sufficiently reproduced, whereas the secretory duct (SD) is replicated partially. IA, interlobular artery; IV, interlobular vein; ia, islet afferent artery; iv, islet efferent vein. Bar, 200 µm. (Miyake et al., 1992). These data indicate that our vascular casts prepared by injections of intentionally reduced amounts of resin through arteries properly reflect the actual blood flow patterns, and provide important, additional information concerning the dynamic flow of blood. Recent experiments by us with incomplete and complete arterial injections of low viscosity resin have confirmed that in the rat cerebral hypophysis, the resin perfused into the median eminence flows into the anterior lobe (Murakami et al., 1993a). It is noteworthy that the stomach, especially its mucosa, is perfused with the casting medium as promptly as the kidney. Also noteworthy is that the pancreatic islets are replicated as promptly as the kidney glomeruli. These findings indicate that the stomach and the pancreatic islets are the organs which are most rapidly and richly supplied with blood. 505 This paper supplements our previous study (Miyake et al., 1992), and demonstrates that in the rat pancreas, the vascular plexuses of the islets are filled with the casting medium more promptly than those of the lobules and ducts, and that in the lobules containing the islets, the medium flows more rapidly through the insulo-acinar portal routes than the lobular routes via the acinar arteries. It also shows that even in the intralobular islets with well-developed insulo-venous efferent vessels, the casting medium flows in the first instance into the portal vessels. These facts prove that the arterial blood in the rat pancreas preferentially flows through the insulo-acinar portal routes. Moreover, we show that neither the insulo-venous routes nor the lobular and ductal routes interfere with the insuloacinar portal routes, which convey high concentration of insular hormones to the exocrine acini (Fujita, 1973; Fujita and Murakami, 1973). Our recent study of thoroughly and partially replicated specimens has shown that even in the human pancreas in which most of the islets are intralobular and have few collateral venous (insulo-venous) drainage (Murakami et al., 1992), the arterially injected resin preferentially flows through the insulo-acinar portal routes (Murakami et al., 1994). Prompt filling with resin of the cortical capillaries of the islets and later filling of their deep capillaries support our previous findings that in the rat islets, the blood mainly flows from the A–D cell mantle to the B cell core (Fujita, 1973; Ohtani et al., 1986), and strengthen our idea that in the rat islets, the B cells’ release of insulin is regulated by glucagon and somatostatin from A and D cells more effectively than otherwise by this intra-insular microcirculatory design from the superficial to deep layers (Fujita and Murakami, 1973). Late filling of the ductal capillaries with resin suggests that the ductal plexuses receive a small amount of blood to reabsorb water from the pancreatic juice (Murakami and Fujita, 1992). Microcirculation patterns of blood have been studied by intravital light microscopy of living tissues or organs which are sometimes injected with India ink, fluorescent dyes, microspheres or other substances (Bunnag et al., 1963; Frazor and Henderson, 1980; Lifson et al., 1980; McCuskey and Chapman, 1969; Ohtani, 1983; Ohtani et al., 1986). Although these methods are useful for direct observation of blood flow, their use is limited; deep and wide areas are hardly observed by the intravital microscopy. Furthermore, the images from the intravital methods are not always clear because of the limited resolution and shallow focus of the light microscope (Ohtani, 1983; Ohtani et al., 1986). This shortcoming is supplemented by the incomplete casting/scanning technique introduced in the present study. Retrograde flow of the injected resin in the veins seems to be a major problem of this technique. To overcome this problem, complete opening of the efferent vessels of the organs to be examined is useful. The incompletely injected or cast specimens are easily deformable so that they should be freeze-dried. 506 T. MURAKAMI ET AL. Fig. 11. Incompletely replicated rat pancreas (8.0 ml resin injection). The islets (arrowheads) and their insulo-venous and insuloacinar portal systems are reproduced together with their connecting interlobular veins (IV), whereas the lobular capillaries directly sup- plied by the acinar branches of the lobular artery are not replicated. Arrows indicate the retrograde flow in the veins. IA, interlobular artery. Bar, 500 µm. MICROCIRCULATORY PATTERNS IN THE RAT PANCREAS 507 Fig. 12. An insulo-lobular complex replicated by 8.0 ml resin injection (closer view of a part of Figure 11). The lobular capillary plexus (EL) is sufficiently reproduced together with the insulo-acinar portal vessels (arrowheads) which arise from the islet (II). The collateral insulo-venous efferent vessels (iv) are also sufficiently reproduced. The capillary plexus of the secretory duct is partially replicated (SD). IA, interlobular artery; IV, interlobular vein. Bar, 100 µm. REFERENCES Miyake, T., Murakami, T., and Ohtsuka, A. (1992) Incomplete vascular casting for a scanning electron microscope study of the microcirculatory patterns in the rat pancreas. Arch. Histol. Cytol., 55:397–406. Murakami, T. (1971) Application of the scanning electron microscope to the study of the fine distribution of the blood vessels. Arch. Histol. Jpn., 32:445–454. Murakami, T. (1972) Vascular arrangement of the rat renal glomerulus. A scanning electron microscope study of corrosion casts. Arch. Histol. Jpn., 34:87–107. Murakami, T. and Fujita, T. (1992) Microcirculation of the rat pancreas, with special reference to the insulo-acinar portal and insulovenous drainage systems: A further scanning electron microscope study of corrosion casts. Arch. Histol. Cytol., 55:453–476. Murakami, T., Unehira, M., Kawakami, H., and Kubotsu, A. (1973) Osmium impregnation of methyl methacrylate vascular casts for scanning electron microscopy. Arch. Histol. Jpn., 36:119–124. Murakami, T., Ohtani, O., Ohtsuka, A., and Kikuta, A. (1983) Injection replication and scanning electron microscopy of blood vessels. In: Biomedical Research Applications of Scanning Electron Microscopy, vol. 3. G.M. Hodges and K.E. Carr, eds. Academic, London, pp. 1–30. Murakami, T., Kikuta, A., Taguchi, T., Ohtsuka, A., and Ohtani, O. Bonner-Weir, A. and Orci, L. (1982) New perspective on the microvasculature of the islets of Langerhans in the rat. Diabetes, 31:883–889. Booz, K.H. (1964) Zur Morphologie und funktionellen Bedeutung einer Spiralklappe in der V. portae der Nagetiere. Anat. Anz., 115:141–147. Bunnag, S.C., Bunnag, S., and Warner, N.E. (1963) Microcirculation in the islets of Langerhans of the mouse. Anat. Rec., 146:117–123. Frazor, P.A. and Henderson, J.R. (1980) The arrangement of endocrine and exocrine pancreatic microcirculation observed in the living rabbit. Q. J. Exp. Physiol., 65:151–158. Fujita, T. (1973) Insulo-acinar portal system in the horse pancreas. Arch. Histol. Jpn., 35:161–171. Fujita, T. and Murakami, T. (1973) Microcirculation of monkey pancreas with special reference to the insulo-acinar portal system. A scanning electron microscope study of vascular casts. Arch. Histol. Jpn., 35:255–263. Lifson, N., Kramlinger, K.G., Mayrand, R.R., and Lender, E.J. (1980) Blood flow to the rabbit pancreas with special reference to the islets of Langerhans. Gastroenterology, 79:466–473. McCuskey, R.S. and Chapman, T.M. (1969) Microscopy of the living pancreas in situ. Am. J. Anat., 126:395–408. 508 T. MURAKAMI ET AL. (1987) Blood vascular architecture of the rat cerebral hypophysis and hypothalamus. A dissection/scanning electron microscopy of vascular casts. Arch. Histol. Cytol., 50:133–176. Murakami, T., Fujita, T., Taguchi, T., Nonaka, Y., and Orita, K. (1992) The blood vascular bed of the human pancreas, with special reference to the insulo-acinar portal system. Scanning electron microscopy of vascular casts. Arch. Histol. Cytol., 55:381–395. Murakami, T., Miyake, T., Ohtsuka, A., Kikuta, A., and Taguchi, T. (1993a) Microcirculatory patterns in adult rat cerebral hypophysis: A scanning electron microscope study of replicated specimens. Arch. Histol. Cytol., 56:243–260. Murakami, T., Fujita, T., Miyake, T., Ohtsuka, A., Taguchi, T., and Kikuta, A. (1993b) The insulo-acinar and insulo-venous drainage systems in the pancreas of the mouse, dog, monkey and certain other animals: A scanning electron microscopic study of corrosion casts. Arch. Histol. Cytol., 56:127–147. Murakami, T., Fujita, T., Tanaka, T., Tsubouchi, M., Tsubouchi, Y., Taguchi, T., Ohtsuka, A., and Kikuta, A. (1994) Microcirculatory patterns in human pancreas: Supplementary observations of vascular casts by scanning electron microscopy. Arch. Histol. Cytol., 57:9–16. Ohtani, O. (1983) Microcirculation of the pancreas: A correlative study of intravital microscopy with scanning electron microscopy of vascular corrosion casts. Arch. Histol. Jpn., 46:315–325. Ohtani, O., Ushiki, T., Kanazawa, H., and Fujita, T. (1986) Microcirculation of the pancreas in the rat and rabbit with special reference to the insulo-acinar portal system and emissary vein of the islet. Arch. Histol. Jpn., 49:45–60.