Three large dissectable rat glomerular models reconstructed from wide-field electron micrographs.код для вставкиСкачать
THE ANATOMICAL RECORD 196:431-440 1980) Three Large Dissectable Rat Glomerular Models Reconstructed from Wide-Field Electron Micrographs GRACE C. H. YANG AND ASHTON B. MORRISON Department of Pathology, Rutgers Medical School, College of Medicine and Dentistry of New Jersey, Piscataway, New Jersey 08854 ABSTRACT Three very large Styrofoam models of glomerular capillary tufts have been built, based on serial ultrathin sections which cut through three complete rat glomeruli isolated from all regions of renal cortex. The images of glomeruli are enlarged on 2 0 x 24" wide-field electron micrographs, from which each capillary is traced on Styrofoam discs and then individually connected. These models can be disassembled to measure the width and length of each capillary segment and to map the complicated glomerular vascular arrangement. Each of the three randomly selected glomeruli has three arteriolar openings passing through the Bowman's capsulw.g., an afferent opening and two closely situated efferent openings. Each of the three glomeruli consists of three major lobules, e.g., two larger lateral lobules and a smaller central lobule. Assuming the model is positioned so that the vascular pole is superior and the afferent arteriole is closer to the observer than the two efferent vessels, the lobule occupying the frontal right portion originates from the right major branching of the afferent arteriole and reconverges a t the right efferent opening. The lobule occupying the frontal left portion originates from the left major branching of the afferent arteriole and reconverges a t the left efferent opening. The lobule occupying the central portion originates from the central major branching and connects both lateral lobules with a few interlobular anastomoses. The diameter and the length of interlobular anastomoses, and the volume of major lobules are listed. The technique of the isolation and reembedment of glomeruli and the construction of models is described in detail. In seeking to study the arrangement of vascular arrangement are as follows: 1) the glomerular capillary tufts, a few models have glomerulus consists of a n anastomosing plexus been built; those include 1) a plastic model re- of capillaries, the lobulation is a technical arconstructed from serial light microscopic sec- tifact (Boyer, '56);2) the glomerular capillaries tions of 2 pm thickness (Boyer, '56).This model form a complicated network divided into adequately shows the outside of the vascular lobules which communicate with each other by arrangement, but it cannot be taken apart to a few interbridging capillaries (Murakami et reveal the internal structure because of the ri- al., '71). The objective of the present work is to take gidity of the plastic. 2) An ironwire model from serial light microscopic sections of 0.5 pm advantage of the newly developed technique of thickness (Aeikens et al., '77). This model wide-field electron microscopy (Yang and Morshows a network of wires without portraying rison, '75) to build glomerular models which are large enough to keep track of the smallest capilthe actual shapes of the capillary. Murakami et al. ('71) have shown scanning lary, accessible for the measurement and mapelectron micrographs of stretched corrosion ping of the complicated capillary network, and casts of r a t glomeruli. Excellent as these finally to observe the vascular arrangement of stretched corrosion casts are in showing the the glomerulus. lobular arrangement, they are too small to be measured or mapped. Received June 5, 1979, accepted September 18, 1979. Other early studies on the glomerular vascular arrangements include Johnston, 1899; Address Reprint Requests to Grace C.H. Yang, M.S. Dept of AnatVimtrup, '28; Wilmer, '41; Trabucco and Mar- omy, CMDNJ-Rutgers Medical School, Piscataway, N.Y. 08854. A.B. Morrison, M.D., Ph.D., is Dean and Vice President for Academic quez, '52; Hall, '55;Zlabek, '57; and Lewis, '58. Affairs, Eastern Virginia Medical School, P.O. Box 1980. Norfolk, Va. 23501. The more recent views on the glomerular 000-3276X/80/1964-0431$02.000 1980 ALAN R. LISS. INC. 43 1 432 GRACE C.H. YANG AND ASHTON B. MORRISON MATERIALS AND METHODS A kidney of a normal male Sprague-Dawley rat, weighing 191g was perfused via abdominal aorta with 1%glutaraldehyde and 1.5%tannic acid in 0.1 M phosphate buffer (pH 7.2). The cortex was cut into 1.5 mm thick slices, which were then postosmicated and then stained en bloc with WOuranyl acetate in maleate buffer (pH 6) a t 60°C overnight. After rinsing in maleate buffer (pH 5.4) twice, the tissue was dehydrated in graded ethanols and embedded in Epon-Araldite at the flat end of a BEEM capsule. The isolation and reembedment of glomeruli After polymerization, the cortical tissue was cut half way with a razor blade. The tissue was then fractured at the cut. Consequently, the glomeruli encapsulated in Bowman's capsules were exposed (Fig. la). One end of a finepointed tweezers was used to pick out the exposed glomeruli which were then separated from the rest of the tissue a t their vascular poles without the macula densa. Several glomeruli were thus isolated from all regions of the cortex and collected in a BEEM capsule (Fig. lb). A sharpened applicator stick, dipped in fresh Epon-Araldite mixture, was used to draw a thin line on the flat cap of another BEEM capsule, where four randomly picked glomeruli were then aligned. The fresh Epon in the BEEM capsule was polymerized before it was filled with more resin mixture to avoid disturbing the linear arrangement of the glomeruli. Finally, a block containing four glomeruli in a straight line was obtained. It was then trimmed into a thin block face and was ready to be sectioned (Fig. Id). The above procedures were done under a dissecting microscope. The preparation of serial sections Every effort was made to reduce the tremendous workload to a more manageable level. Two steps were taken to avoid the staining ofmasses of grids and to eliminate the danger of section breakage during staining. First, kidney was stained en bloc with uranyl acetate prior to embedment. Secondly, ultrathin sections thicker than usual were used. They are of dark gold interference color a t a n estimated thickness of 100 nm (Yang and Shea, '75). Therefore, once the sections were mounted on slot grids coated with plain 0.5% Formvar, they could be photographed immediately without additional treatment. The serial sections were made with a diamond knife mounted on a LKB Ultratome 111. The entire process of sectioning took five days, during which time it was necessary to use extreme caution to avoid damaging either the diamond knife or the block face. The entire apparatus was left overnight on the ultramicrotome to minimize handling and to avoid shifting cutting planes which would result in the discontinuity of glomerular images. The block face was trimmed several times while still mounted on the ultramicrotome, thus reducing the number of grids carrying the serial sections. Consequently, the necessity for changing grids in and out of the electron microscopic column was reduced to a minimum. The whole glomerular images were recorded by a Philips 300 electron microscope at x 420 using half mask, so that each negative could record two whole glomerular images. The selection of negatives Though serial sections were cut, it was found to be unnecessary to record every section. Each negative was compared to the adjacent negative under a dissecting microscope which was placed on top of a narrow illuminated viewer. Only negatives showing the changes in the capillary lumina were chosen, so that it was possible t o delete unnecessary films and yet still follow the pathways of every capillary. Fewer negatives near the top or bottom portions of the glomeruli were selected, and more negatives were used from the central portion where small capillaries appeared more frequently and were more tortuous in their courses. Nearly 150 films were chosen for each glomerulus. Hundreds of 2 0 x 2 4 wide-field electron micrographs were then obtained (Yang and Morrison, '75). The final magnification was x 3,909. The height of the model was compensated for the unselected sections as follows: first, the theoretical height of the glomerulus was obtained by multiplying the section thickness by the total number of sections occupying the entire glomerulus; then the theoretical height was divided by the number of selected sections; in this manner, the use of Styrofoam discs of 3 mm thickness was determined and custom ordered from the manufacturer. The construction of the models Each print was given a section number in the order of its sequence. The outlines of all the capillary lumina were traced on 2 0 x 24" transparent gel sheets (photographic supply store) with a marking pen. A micrograph with LARGE DISSECTABLE GLOMERULAR MODELS 433 Fig. 1. The isolation, reembedment, and sectioning of glomeruli. a) A piece of osmicated tissue fractured from an Epon block, exposing a glomerulus (arrow). b) Several osmicated and plasticized glomeruli isolated from tissue. c) Four glomeruli aligned in a row. d) The trimmed block face ready to be sectioned. e) Serial sections showing 32 images of glomeruli. 0 Close-up view of the serial sections showing glomeruli encapsulated in Bowman’s capsule. the most complicated capillary lumina from the center part of the glomerulus was chosen as the standard for orientation. First, it was traced on a gel sheet, then given a center axis on which subsequent tracings were aligned. This was necessary because the glomerular images on the micrographs were not printed at the exact same angles. A ball pen was used to press down the outlines of all the capillary lumina from the gel sheet onto styrofoam sheets of 3 mm thick- ness (Tekniplex, Inc. Somerville, N.J.). Styrofoam discs of various shapes were then cut using a n X-act0 knife, and labeled with the section number and a n arrow showing the orientation. The discs were attached to the matching outlines on the gel sheets by pieces of double-stick tape. Using the centering arrow on the gel sheet as a guide, the discs were removed from the gel sheet, pasted in the correct position with Elmer’s glue, and reinforced by small pins. 434 GRACE C.H. YANG AND ASHTON B. MORRISON The micrographs were checked frequently to ensure the accuracy of various connections. The interlobular anastomoses were determined by gently pulling the lobules away from each other to locate the connecting points. These were then separated and fastened with Velcro tapes (fabric store). This enabled the models to be pulled apart to reveal the internal details and reconnected to show the entire capillary tuft. Each of the major lobules was weighed, and the volume was calculated from the density of the Styrofoam. Only three glomerular models were built, because one glomerulus was damaged at its vascular pole. Though the pieces of the model were glued together, the models could be easily dissected because they were made of Styrofoam discs, which did not absorb the glue as thoroughly as discs made of wood or other material. The thicker capillary segments, e.g., the primary afferent branchings, were dissected by a jeweler’s saw. Other segments were dissected by simply inserting a knife between the discs and moving it up and down until the discs separated. The dried glue could be easily peeled away and the pins could be removed by forceps. RESULTS Each of the three glomerular models has three arteriolar openings passing through the Bowman’scapsule verified on the original electron micrographs (Fig. 5)-e.g., an afferent opening (identified by the J G granules on its arteriolar wall), and two closely situated efferent openings (Figs. 2a, 3a, 4a). Each of the three models can be easily separated into three lobules-two larger lateral lobules and a smaller central lobule (Figs. 2b, 3b, 4b). The lobule occupying the frontal right portion originates from the right major afferent branching and reconverges a t the right efferent opening. The lobule occupying the frontal left portion originates from the left major afferent branching and reconverges a t the left efferent opening. The lobule occupying the central portion originates from the central major branching as a separate unit, but later connects both lateral lobules with a few interlobular anastomoses. No direct connection is observed between two lateral lobules, although they are next to each other in the frontal plane with their own efferent arteriolar openings situated close together as a pair. Each lobule consists of a very complicated anastomosing network of capillaries, and no “blind-end” was found. The classically described simple capillary “loops”were not found in any of the three glomerular models. The diameter of the capillaries ranges from about 2 pm to 22 pm; most of them measure between 5 pm to 10 pm. The course of the pathways is quite tortuous. A capillary turns a t sharp angles in short distances and isjoined by other capillaries at various points. It was found that the diameter of a capillary may vary a great deal within a segment. Therefore, it could be said that a single red blood cell must travel extensively before it passes through the maze of the capillary network. The diameter of three openings, the diameters of the major branchings of the three openings, the length and diameter of each interlobular anastomosis, and the volume of the capillary lumina of the major lobules are illustrated in diagrams showing simplified versions of the glomerular capillary tufts (Figs. 5,6). Model #2 has been completely dissected to measure and map the complex capillary network. The data have been mathematically analyzed and their physiological significance have been discussed (Shea, ’79). DISCUSSION In all three models, the glomerulus divides into three major lobules, with the center smaller lobule acting as the bridge between the two larger lateral lobules. Our findings are somewhat similar to Murakami et al. (’71). However, all three models show double efferent openings extending out of Bowman’s capsule, whereas only 11out of 1,200rat glomeruli have true double efferent arterioles (Murakami et al. ’71). This discrepancy may be due to a difference in technique. In their corrosion cast method, the Bowman’s capsule is corroded away and the location of its borderline cannot be determined. With our isolation technique, the glomeruli are severed from their vascular poles. By combining observations from these two different approaches, it is possible that after the two exiting arterioles pass through the Bowman’s capsule, they join together to become the single efferent arteriole found in most of the rat glomeruli in Murakami’s corrosion study. Of course, it is also possible that those two exiting arterioles are true double efferent arterioles. The isolation and reembedment of the glomeruli is a crucial step in obtaining its complete ultrastructural images. Serial tissue sections are only adequate in building glomerular models based on light microscopic images. The semithin “survey” tissue section may locate many glomeruli, but only after the top parts of their capillary tuft are removed. In order to be LARGE DISSECTABLE GLOMERULAR MODELS 435 Fig. 2. Glomerular model X1. a) Frontal view showing an afferent arteriolar opening (A) and a pair of efferent openings (E);b) Dorsal view showingthree lobules:two larger lateral lobules, a right (R) and a left (L), and a smaller central lobule (C). (Reduction: x 4.5). 436 GRACE C.H. YANG AND ASHTON B. MORRISON Fig. 3. Glomerular model #2. a) Frontal view showingthree openings:two (E) and one (A). b) Dorsal view showing three lobules: right (R), center (C), and left (L). LARGE DISSECTABLE GLOMERULAR MODELS 437 Fig. 4. Glomerular model X3. a) Frontal view showing three openings:two (E)and one (A).b) Dorsal view showing three lobules: right (R),center 0,and left (L). 438 GRACE C.H. YANG AND ASHTON B. MORRISON Fig. 5. Two wide-field electron micrographs chosen from glomerulus 6 3 . a) Seetion X128 showing af€erent opening (A); b) Section X331 showing double efferent openings (ELBowman’s capsules are indicated by arrows. 439 LARGE DISSECTABLE GLOMERULAR MODELS GLOMERULUSSl Center to Left Anastomoses Diam. 9.9pm 7.4 6.6 10.7 6.6 6.1 8.2 Center to Right Ana stornoses Lensth 11.5pm 17.9 21.7 9 9 1 7.9 14 Diam. 6.1p 93 86 Length 35.8p.m 64 16.9 GLOMERULUS S 2 9.6pn 9.6 5.9 2 9 6.6 1 1 . 3 ~ 1~m 4.9 9.7 10.1 38.6 20.7~ 20 9 6.1 39.1 1 GLOMERULUSB 3 14.9 6 . 9 ~ 7 14.lpm 16.6 7.4 @ ; .6 I5.5P xl$ .47 1.69 x ~ ~ 5 None Volume 1 0.9 10.9~ Fig. 6. Three simplified and stretched glomerular diagrams on which the following measurements are listed: the volume of left (L),right (R),and center (C)lobules; the diametersof afferent openings (El, and their major hranchings; and the diameters and the lengths of all the interlobular anastomoses. 440 GRACE C.H. YANG AND ASHTON B. MORRISON recorded, an ultrathin section has to be supported on a grid where the largest opening a t present is only 1 mm x 2 mm. Of course, a complete glomerulus might be encountered by very long serial sectioning. However, such a probability on a section of lmm x 2 mm through a tissue slice of 1.5 mm thickness is, indeed, very low, not to mention the wear-andtear on the diamond knife by aimless sectioning. The efficiency of sectioning is further improved by embedding three glomeruli in a row; by placing the glomeruli close to the cutting surface, and by lengthwise cross sectioning associated with oval glomeruli being placed on a flat surface. Styrofoam is an ideal material in building complicatedmodels. It has very low density and therefore can be built into a complex network without falling apart a t weak points. The finished models have a capacity for bouncing, i.e., a part of the lobule can be pushed away and it will bounce back. This is very helpful in locating the interlobular anastomoses. Styrofoam sheets are also cheaper and easier to work with than wood or other material. In Boyer’s model, the lobulation of the capillary tuft is indistinct, perhaps because the adhesion of the neighboring lobules is unavoidable since liquid plastic is poured into stacks of wax plates with many holes representing capillary profiles. In our glomerular models, the pattern of lobulation is quite distinct because there is no random adhesion of the neighboring lobules, even though they might be touching each other in the finished model. If the connection cannot be verified on the micrographs, no pin or glue is applied. The reason why the parietal layer of the Bowman’s capsule stays attached to the glomeruli in all cases using this technique (Fig. 5, Fig. lc) is that it has been reinforced by the polymerized Epon. When the fracture forces the plasticized cortex apart, the surrounding tubules separate, leaving the intact renal corpuscle exposed, The mapping of capillary network can be done with much assurance. Not only each connection has been checked on the original electron micrographs, but also the model is so large that even the smallest capillary of 2 pm in width has a diameter of 7.8 mm in our model. Though the construction of these models took very much time and effort (from January, 1975 to June, 19761, it nevertheless led to a better understanding of the vascular arrangement of the rat glomerulus, and it is the only way at present to measure and map the complex network of the glomerulus. ACKNOWLEDGMENTS It is a pleasure to thank Dr. Stephen Shea for suggestions and support during the work. The authors also wish to thank Miss Eloise Ten Eyck for her artistic help in the construction of models, and Miss Robin Paulmeno for her most diligent work in printing and tracing of hundreds of micrographs. This study was supported by U.S. Public Health Service grant No. AM-12495 (Principal Investigator: Dr. Shea) and CMDNJ-Rutgers Medical School General Research grant No. 9642. LITERATURE CITED Aeikens, B., A. Eenboom, W. Plate, L. Fronhold, and A. Bohle (1977)The construction of models intending spatial representation of complex vessel structures demonstrated by means of the mammalian glomerulus. (Ger.) Microsc. Acta, 79: 12CL-126. Boyer, C.C. (1956) The vascular pattern of the renal glomerulus as revealed by plastic reconstructions from serial sections. Anat. Rec., 125:43%441. Elias, H. (1957) De structura glomeruli renalis. Anat. Am., 104:26-36. Johnston, W.B. (1899 A reconstruction of the human kidney. Anat. Am., Bd. 16. Hall, B.V. (1955) The organization of the renal glomerulus into independent lobular systems of intercommunicating anastomosing capillaries. Anat. Rec., 121,433 (Abstr.). Lewis, O.J. (1958) The vascular arrangement of the mammalian renal glomerulus as revealed by a study of its development. J. Anat. (Lond.), 92,433. Murakami, T., M. Miyoshi, and T. Fujita (1971) Glomerular vessels of the rat kidney with special reference to double efferent arterioles. A scanning electron microscope study of corrosion casts. Arch. Histol. Jap., 33: 179-198. Shea, S.M. (1979) Glomerular hemodynamics and vascular structure. The pattern and dimensions of a single rat glomerular capillary network reconstructed from ultrathin section. Microvasc. Res., 17 (In press). Trabucco, A,, and F. Marquez (1952) Structure of the glomerular tuft. J. Urol., 67:235-255. Vimtrup, B.J. (1928) On the number, shape, structure, and surface area of the glomeruli in the kidneys of man and mammals. Am. J. Anat., 41;12%151. Wilmer, H.A. (1941)The arrangement ofthe capillary tuft of the human glomerulus. Anat. Rec., 80:507. Yang, G.C.H., and A.B. Morrison (1975) Wide-field electron microscopy. A rapid method for the study of histologic material that provides a bridge between light and electron microscopy. Am. J. Clin. Pathol., 64:648-654. Yang, G.C.H., and S.M. Shea (1975) The precise measurement of the thickness of ultrathin sections by a “resectioned section” technique. J. Microsc., 103:385392. Yang, G.C.H., A.B. Morrison, and S.M. Shea (1976)Three’ dimensional models of the glomerulus. XI Congress of the International Academy of Pathology. Washington, D.C., October 17-23 (Exhibit). Yang, G.C.H., and A.B. Morrison (1977) The application of wide-field electron microscopy.Microsc. Acta, 79:277-279. Zlabek, K. (1957) The arrangement of the intraglomerular blood vessels in the human kidney. Rev. Szech. Med. 111, 41:l-12. Zlabek, K. (1973)Uber dunne intercapillare anastomosen in nierenglomerulus der rat. Acta. Anat. (Basel), 85: 177189.