Plexus muscularis profundus and associated interstitial cells. I. Light microscopical studies of mouse small intestineкод для вставкиСкачать
THE ANATOMICAL RECORD 203:115-127 (1982) Plexus Muscularis Profundus and Associated Interstitial Cells. I. Light Microscopical Studies of Mouse Small Intestine JURI JOHANNES RUMESSEN AND LARS THUNEBERG Anatomy Department C, University of Copenhagen, 1 Universitetsparken, DK-2100 Copenhagen 0,Denmark ABSTRACT The zinc iodidelosmic acid (ZIO)method was used in a modification that selectively stained nerves and associated interstitial cells of Cajal (ICC ) of muscularis externa. Due to its selectivity the method allowed a detailed stereoscopical analysis of whole mounts with respect to the topography and morphology of these elements. The method thus assisted and expanded our ultrastructural studies. The ZIO staining allowed a distinction of four morphologically different interstitial cell types (ICC-I-IV)confined to four compartments. The stained components were: (1)A rich plexus of highly ramified intestitial cells (ICC11) in the subserous layer. (2) Auerbachs plexus with an associated extensive plexus of interstitial cells (ICC-I)in close contact with tertiary fasciculi. (3)Nerve fasciculi of the outer division of the circular muscle layer. These formed a nerve plexus in a well-definedplane in the outermost cell layers (plexusmuscularis superficialis),with few fasciculi located internal to this plexus. A few bipolar interstitial cells (ICC-IV)were associated with nerve fasciculi of this region. (4)A nerve plexus located in the region between the two subdivisions of the circular muscle, plexus muscularis profundus (PMP).PMP was revealed throughout the small intestine as a continuous network of elongated, circularly oriented meshes. The pattern of connections between PMP and the other enteric plexuses was studied stereoscopically. Ganglion cells intrinsic to PMP occurred widely scattered. Interstitial cells associated with PMP (ICC-111)were arranged in a plexifonn manner; their morphology and relations to nerves were investigated in great detail. A selective innervation of ICC-I11via axons of PMP was strongly supported. Plexus muscularis profundus (PMP)is a welldefined nerve plexus, sandwiched between two subdivisions of the circular muscle layer of the small-intestinal muscularis externa: an outer, thick layer and an inner, very thin layer (Cajal, 1893,1911; Li, 1937,1940; Taxi, 1965; Gabella, 1972, 1974; Duchon et al., 1974; Rumessen et al., 1982; Thuneberg, 1982). A nerve plexus at this specific location has been identified in all mammals examined. Its function is unknown, but the potential importance of PMP as a significant component of the enteric nervous system was clearly demonstrated by Gabella (1972): In guinea-pig ileum PMP contains almost two-thirds of all vesiculated axon terminals counted within the circular layer. In recent reviews it has been stressed that the inner division of the circular muscle is richly innervated by PMP (Gabella, 1979). the 0003-276W82/2031-0115$04.00 0 1982 Alan R. Liss. Inc. branches of which are “running between the innermost layers of muscle cells” (Furness and Costa, 1980). In contrast, our ultrastructural studies (Rumessen et al. 1982; Thuneberg, 1982) on mouse intestine show conclusively that PMP specifically innervates the outer, main layer of circular muscle. Furthermore, the innervation is mediated by interstitial cells of Cajal (type 111:ICC-111; Rumessen et al., 1982; Thuneberg, 19821, identical with the “hybrid cells” of dog intestine (Duchon et al., 1974; Daniel, 1977). In accordance with the latter authors we have found that PMP preferentially innervates ICC-111, which are coupled by numerous gap junctions with cells of the outer, main division only. This specific arrangement stresses the potential importance of ICC-I11as Received July 2,1981; accepted December 2, 1981. 116 J.J. RUMESSEN AND L. THUNEBERG intercalated, regulatory cells in circular muscle function (Daniel,1977).The “interstitialcells of Cajal” (ICC) were described by Cajal (1911)in two locations related to the intestinal muscularis externa: associated with (1)PMP and (2) Auerbach’s plexus. In a previous light and electron microscopical study (Thuneberg, 1982)we have accumulated evidence in favor of a role as intestinal pacemaker cells of ICC (type I: ICCI) associated with Auerbach’s plexus. The study of the organization of a network of highly ramified ICC-I was carried out by selective staining of ICC-I with supravital methylene blue. In order to expand our ultrastructural observations on ICC-111 (which do not take up methylene blue), we have developed a modification of the zinc iodide/osmic acid (ZIO)method, which selectively stains the intestinal nerves and associated ICC, including ICC-111. The study of whole mounts of the muscularis, stained with ZIO, and observed with stereoscopical methods, has allowed a precise and thorough description of the general organization of ICC and nerves throughout the small intestine. The present investigation provides a basis for our accompanying ultrastructural study (Rumessen et al., 1982) and expands the observations made by Cajal (1911), Li (1940). and Taxi (1965). MATERIALS AND METHODS Adult albino mice of either sex were used. After cervical dislocation, segments of duodenum, jejunum, and ileum were excised and processed for light microscopy. The excised segments were immediately transferred to a modified Tyrode solution (Caprilli e t al., 1970). Then papaverine (1 mM) was added. The segments were immersed for 5-10 minutes, at an initial temperature of 37 “C, declining to room temperature. This procedure caused an apparent complete relaxation of smooth muscle cells, providing comparable material from different parts of the intestine. The segments were flattened and slightly distended by strips of dental wax (TruwaxDentsply) introduced into the lumen. Subse quently, 1-cm pieces were cut and immersed in a freshly prepared ZIO mixture comprising 0.4%OsO,, 2.4% ZnIz (Merck)(Maillet, 1959). The segments were immersed for 4, 10, 24,48, 72 hours, and some for 7 and 14 days at room temperature. Some preparations were not subjected to either papaverine or slight distension, and others were subjected to only one of these procedures. Following ZIO-staining/fixation whole mounts of the muscularis externa mea- suring about 15-20 mm2 were cautiously dissected under a stereomicroscope by means of small-caliber hypodermic needles. The isolated muscle layers were mounted with Aquamount (Gurr biol. reagent, Searle). The preparations were left to harden for 1-2 days at room temperature before examination. For optimal microscopical examination of the inner circular muscle layer most preparations were mounted with this layer upward. Sphincter regions were not included in the material. Further examination was carried out, and all photos taken, on a Zeiss photomicroscope equipped with differential interference contrast (Nomarski) optics. Nerves and ICC were studied by stereoscopical methods. A polarizing filter system (Michel, 1967)which converts the ordinary binocular microscope to a stereomicroscope allowed true stereoscopical view a t all magnifications. Limitations due to interpretation of sectioned material were thus overcome. Stereo-pairs(Figs.4-7) should be examined by means of an ordinary binocular stereo-viewer or naked-eye stereopsis. Nakedeye stereopsis with inversion (convergent visual axes) will produce a reversal of desired perspective. RESULTS Our previous investigations (Thuneberg, 1982) have revealed four main types of interstitial cells of Cajal (ICC) related to the muscularis externa of mouse small intestine. ICC-I are situated between the main muscle layers, associated with Auerbach’s plexus (AP) in a plexiform manner, ICC-I1 form a plexus in the subserous layer, ICC-I11 are associated with PMP, and ICC-IV are disposed along intra-muscular nerve fasciculi in the outer, main layer of circular muscle. In initial experiments, it was observed that preincubation of the intestinal segments in papaverine (see Materials and Methods) prevented a heavy, unspecific staining of smooth muscle cells, which otherwise occurred in contracted segments, so that nerves and ICC were totally obscured. Subsequent to this preincubation the ZIO mixture selectively stained nervous elements in all regions of the muscularis externa and of the submucous layer, with a distinct outline of single axons and varicosites. ZIO staining of the different elements followed a characteristic time scale. First to appear (incubation times 2-3 hours) was AP, followed by intramuscular axons, PMP, and the submucous plexus (PS) (4-10 hours). ICC generally appeared later (10-24 INTERSTITIAL CELLS IN PLEXUS MUSCULARIS PROFUNDUS hours), and no further ZIO staining occurred after incubation times of more than 72 hours. Nervous elements AP, confined to the space between the main muscle layers, exhibited a plexiform appearance of ganglia and of primary, secondary, and tertiary fasciculi. The tertiary plexus was generally irregularly arranged, and appeared little or variably developed. In a well-defined plane in the outer part of the outer circular muscle layer was revealed a plexus of parallel, regularly spaced, and largely unbranching nerve fasciculi (plexusmuscularis (strati circularis) superficialis (PMS), Thuneberg, 1982). These nerve fasciculi extended from primary or secondary fasciculi of A P to the interstices of the circular muscle layer, thus invariably located at the inner aspect of AP. The fasciculi were very seldom obliquely or transversely interconnected, but showed bifurcations of single fasciculi continuing as two parallel fasciculi, separated 12-18 pm from each other. The width of the fasciculi was 3-4 pm (Figs. 1, 5, 6) . A prominent PMP was confined to a welldefined plane, separated from the submucosa by the layer of one to two muscle cells of the inner subdivision of the circular muscle layer (Fig. 2). The plexus was encircling the intestinal wall in its whole circumference. No difference in overall arrangement was observed in different parts of the small intestine. The main constituents of PMP were larger (primary),circularly oriented fasciculi separated 30-40 pm and measuring 4-7 pm in width. These fasciculi were thus more widely separated and of a larger width than fasciculi of PMS and of approximately the same width as true secondary fasciculi of A P (Figs. 1, 2). Via oblique or transverse interconnections these fasciculi formed the largest meshes, generally elongated in the direction of the main axis of the circularmuscle cells and with sizes varying between 40 X 200 pm and 10 X 15 pm. The large (primary) meshes were often subdivided by finer fasciculi or single axons. Higher magnifications of fasciculi revealed preferentially stained varicose axons. Varicosities (1-4 pm in diameter) were separated 1 pm to several pm from one another within a single axon. Schwann cells related to PMP could be distinguished with certainty from ICC-111. Schwann cells were stained only very occasionally and usually only empty swellings within the fasciculi indicated the location of Schwann cell nuclei (Fig.2). When present, the 117 staining of Schwann cells was very characteristic, as the nucleus was invariably the most colored part, the cytoplasm being only faintly shadowed. In a small number of preparations ganglionlike formations were observed as a part of PMP, clearly separated from submucosa by the muscle cells comprising the inner subdivision of the circular muscle layer. As neurons were most often unstained by the ZIO mixture, their presence in PMP was indicated by the occurrence of the typical basketlike networks of stained axons surrounding groups of two to five or single (unstained) ganglion cells. The neurons inside the ganglionlike formations were identified at higher magnification, using differential interference contrast (DIC)optics, which revealed their rounded nuclei and large spherical nucleoli (Fig. 3). very similar to ganglion cells of AP. Stereoscopy a t high magnifications established the location of ganglion cells and pericellular networks with certainty to the level of PMP, as neurons intrinsic to this plexus. The main part of connections to PMP appeared to be derived from AP, and f a r fewer connections were observed between PMP and PS. Perivascular connections originating in the zone of mesenteric attachment were not investigated. The stereoscopic investigations revealed connections between AP and PMP, with a perpendicular course in relation to the circular muscle layer (Figs. 4,5,7).These connections originated from primary or secondary elements of AP and often gave off branches to supply PMS (Fig. 5 ) . The only connections between PMP and PS were revealed by a few of the large perpendicular connections between AP and PMP continuing to connect with PS (Fig. 4). Infrequent, perpendicular connections between PMP and PMS were revealed (Fig. 6). Occasionally fasciculi left the plane of PMS to traverse the circular muscle coat for a long distance nearly parallel to the muscle fibers, and ultimately merge with PMP. Thus, PMS, PMP, and the connections mentioned constituted all intramuscular nerves. Nerves were neither observed within the longitudinal muscle, nor in the subserous layer. Interstitial cells associated with PMP (ICC-111) ICC-I11 exhibited a characteristic staining pattern, as their nuclei were only faintly stained, although sometimes with nucleoli standing out in negative contrast. Cytoplasm of ICC-I11 was well delimited and invariably of 118 J.J. RUMESSEN AND L. THIJNEBERG INTERSTITIAL CELLS IN PLEXUS MUSCULARIS PROFUNDUS a darker shade than that of the nuclei. Precise discrimination of axon bundles and cell processes of ICC-I11 was possible, since ICC-I11 cytoplasm was of different shades of gray in contrast to the deep black, varicose axons. ICC-I11were demonstrated in all parts of the small intestine, along its whole circumference, constituting a plexus of somewhat irregular outlines, with nuclei oriented in the same plane, and spaced with regular intervals. Thus ICCI11 formed a flat, two-dimensional cellular plexus in the same plane as PMP, intermingling with this (Fig. 8). The nuclei of ICC-I11showed some variation in size and shape. ICC-I11 had oval or rounded nuclei, flattened in the radial direction, while measuring about 13-25 pm X 6 pm in the tangential plane. The longest axis of the nucleus was oriented in the direction of the muscle cells of the circular layer. Nucleoli (three to five) often were arranged in a longitudinal row (Fig. 11). Cells with larger nuclei most often were bipolar with more abundant perinuclear cyto- Abbreviations A, Single axon of PMP AP, Auerbach’s plexus DIC. Differential interference contrast (Nomarski) F, Large (primary) fasciculus of PMP N, Nucleus of ICC-111 (unless specified) PMP, Plexus muscularis profundus; deep plexus of circular muscle PMS. Plexus muscularis superficialis; superficial plexus of circular muscle Fig. 1. PMS (arrows)located in theouter part of theouter subdivision of the circular muscle layer. Fasciculi are regularly spaced in a well-defined plane at the luminal aspect of AP. of which a primary fasciculus is labeled. Doubleheaded arrow: main axis of smooth muscle cells in the circular layer. X 610. (All figures show nerve plexuses and interstitial cells as they appear in the tangential plane of ZIOstained whole mounts.) Fig. 2. PMP (arrows) confined to the plane between the thick outer and thin inner subdivision of the circular muscle layer. Fasciculi are interconnected, forming elongated meshes. Fasciculi of PMP are thicker and more widely spaced as compared with fasciculi of PMS (Fig. 1).S : location of nucleus of Schwann cell (unstained).Double-headed arrow: a s in Figure 1. X 610. Fig. 3. Group of four ganglion cells constituting a part of PMP. Nuclei and large rounded nucleoli (arrows) are seen. Ganglion cells are surrounded by a network of axons. DIC. X 1,760. 119 plasm, whereas cells with smaller nuclei often exhibited two to four and occasionally five primary processes in combination with a marked scarcity of perinuclear cytoplasm. In comparison, nuclei of smooth muscle cells averaged 25 X 6 pm. ICC-I11 showed considerable variation in their pattern of ramification. The primary processes extended from the sparse perinuclear cytoplasm and were slender (2-3 pm) or had a broader, laminar appearance. They gave off secondary and tertiary processes, not infrequently terminating in a complex arborizing formation (Figs. 9, 10).Almost all ICC-111 had very fine, short “twigs,” either at the most distal parts of the processes, or at several sites along the larger processes, as well as extending from the perinuclear cytoplasm. Thus, the pattern of branching could be described as a mixture of dichotomic (main processes) and monopodic (“twigs”).ICC-I11 most often had their main axis oriented in the direction of the muscle cells of the circular layer (Fig. 8), but multipolar ICC with primary and secondary processes extending for a considerable distance in all directions in the plane of PMP were often observed (Figs. 9, 13). Some main processes of ICC-111, crossing the outer circular-muscle cells at an angle, showed small expansions closely following the contours of the smooth muscle cells (Fig. 13). Occasionally processes of ICC-111 were followed deep into the outer circular layer. ICC-I11 could be seen to be in apparent contact with each other in perinuclear areas (Fig. 1 l),or by means of distal processes overlapping each other for a distance (Fig. 8).Some ICC-I11 seemed to be enveloped by secondary processes of neighboring ICC111. Relations to nerves. In preparations with well-developed staining of nerves and ICC, all ICC-I11encountered were in close contact with axonal varicosities, at some point or another. ICC-111 with nuclei apposed to larger fasciculi of PMP, often seemed to supply the latter with an incomplete sheath of primary processes, which often followed the fasciculi for several hundred microns. ICC-I11 with no apparent perinuclear association to axons, always contacted larger fasciculi via primary or secondary processes (Fig.9).Single axons were often seen to follow closely the contours of ICC-I11 processes and perinuclear cytoplasm (Figs. 13,14). This close relationship of ICC-111 to varicose axons could not be explained as a passive close apposition of the involved structures in a 120 J.J. RUMESSEN AND L. THUNEBERG Fig. 4 . Stereo-pair showing connection (arrowhead) between a submucous ganglion (G) and PMP (F). This connection can be followed (arrows)toward AP (out of focus).X 440. Fig. 6. Stereo-pair showing connection (small arrows) be. tween afasciculusof PMS(broad arrow) and PMP(F).X 370. Fig. 5. Stereo-pair of the same area as Figure 4. Focus is a t AP. The fasciculus connecting PMP and AP contacts a fasciculus of PMS (small arrow) and merges with a ganglionated strand of AP (broad arrow). X 440. Fig. 7. This stereo-pair shows a perpendicularly oriented connection (arrow head) between PMP (focus) and AP. G: location of a ganglion cell in AP. X 740. Fig.8. Survey of PMPand ICC-111. ICC-111show anoval nucleus and two to five primary processes. The finer cell processes form a complex pattern. Double-headed arrow: main axis of smooth muscle cells of the circular layer. X 550. Fig. 9. A multipolar ICC-111 spanning between two larger fasciculi of PMP. This cell exhibits a very complex pat- tern of ramification, and an extensive development of cytoplasmic "twigs" (arrows). DIC. X 1,200. Fig. 10. Abipolar ICC-111.showing anextensivedevelopment of cytoplasmic "twigs" (arrows). Preterminal varicosities are in close contact with the cell (arrowheads). DIC. X 1.530. 122 J.J. RUMESSEN AND L.THUNEBERG INTERSTITIAL CELLS I N PLEXUS MUSCULARIS PROFUNDUS restricted space, because it very often seemed that larger fasciculi of P M P gave off a single or a few axons to pursue, and selectively supply, ICC-I11 with terminal (Figs. 11, 12, 16)and/or preterminal (Figs. 13, 14, 15)varicosities, in a manner very suggestive of innervation. It was observed that varicosities of larger size (approximately 3-4 pm) more often were in apparent selective contact with ICC-I11 than could be expected from their overall distribution (Figs. 11,12, 15).Nomarski microscopy indicated that contacts to ICC-I11 revealed as single axons actually were single and not accompanied by other unstained axons (Figs. 14, 16). Other single axons leaving fasciculi and terminating among smooth muscle cells but not associated with ICC-I11 were very rare, although they have been observed. Smooth muscle cells of the inner subdivisions of the circular muscle layer (Fig. 17)sometimes showed a forklike branching not observed in the outer circular muscle cells, but finer branching was never observed. Fig. 11. Two ICC-I11 apparently contacting each other (arrowheads). The cell to the right appears to be innervated by the terminal part of a varicose single axon (arrows).A row of nucleoli is seen in negative contrast in the samecell. X 960. Fig. 12. ICC-I11 with terminal varicosities abutting on its cytoplasm (arrow).The varicosities are larger than most others seen. This cell has three primary processes, which are delicately branched. DIC. X 960. Fig. 13. An ICC-I11 with four main processes. Two processes are closely followed by a varicose single axon (arrows). The other processes, which traverse the muscle cells, are slightly expanded and appear to follow the contours of the faintly stained smooth muscle cells (arrowheads). X 960. Fig. 14. ICC-I11 in close contact with single axons and preterminal varicosities. These are disposed along the perinuclear cytoplasm (arrowheads) and along a primary process (arrow). X 1,530. Fig. 15. Multipolar ICC-111 apparently innervated by preterminal varicosities (small arrows). The varicosities impress ICC-I11 cytoplasm. Broad arrows: ICC-I11 processes. X 1,380. Fig. 16. Bipolar ICC-I11 apparently innervated by terminal varicosities on a single axon leaving a larger fasciculus of PMP. The selective nature of the contact is evident (small arrows). Broad arrows: ICC-I11 processes. DIC. X 1,370. Fig. 17. A smooth muscle cell of the inner division of the circular layer. The cell branches in a forklike fashion (arrows).DIC. X 550. 123 Relations to other types of ICC (I, 11, IV). In most preparations a rich plexus of ICC-I confined to the space between the main muscle layers was revealed. ICC-I were stained with fine detail of processes including the finest and most distal ramifications. An intimate association of the tertiary elements of AP with ICC-I was observed (Fig. 18). Nuclear features of ICC-I were much like nuclei of ICC-111. Although no contacts between ICC-I and ICCI11 were seen, processes of both cell types were traced in the bulk of the circular muscle. ICC-IV were only occasionally stained. They were bipolar cells with very long, largely unbranching processes, closely apposed fasciculi of PMS,and other intramuscular fasciculi (Fig. 20). No contacts between ICC-IV and ICC-I11 were seen. The subserous plexus of ICC-I1 was also demonstrated, showing a lower cellular density than the plexus of ICC-I (Fig. 19).The cells had three to five long primary processes, branching in a largely dichotomic pattern, apparently interconnecting with processes of other ICC-11. Contact points between individual ICC-I1were difficult to distinguish. The main axis of ICC-I1 was oriented roughly in the direction of the muscle cells of the longitudinal layer. Nuclear features were similar to other types of ICC. No contacts between ICC-I1 and other ICC or nerves were observed. DISCUSSION Osmic acidliodide staining of peripheral autonomic nerves was introduced by Champy (1913) and modified and improved by Maillet (1959) (osmic acidlzinc iodide). Initially it was claimed (Coujard, 1943) that the osmic acidiodide methods rendered adrenergic nerves selectively visible, but this was later questioned and discredited (e.g., Hillarp, 1959). The problems concerning the specificity of the ZIO methods are still not solved (see Pellegrino de Iraldi, 1977). At the light microscopical level, ZIO staining is highly selective for unmyelinated nerves, using incubation times of 3-4 hours (Maillet, 1963, 1968/69;present report). With prolonged incubation time a wider spectrum of cell types is stained, most notable in this context the ICC. The preincubation of smooth muscle tissue in papaverine as in the present study, seems to further increase the specificity of ZIO methods, preventing heavy unspecific staining of contracted smooth muscle cells. In the present report, all interstitial elements stained with ZIO methods are referred to as 124 J.J. RUMESSEN AND L.THUNEBERG INTERSTITIAL CELLS IN PLEXUS MUSCULARIS PROFUNDUS ICC, but other cell types may occasionally be stained. Indeed, cells interpreted as fibroblasts have been demonstrated with ZIO methods (Stach, 1963; Maillet, 1968/69; Stockinger and Graf, 1965; Garret, 1965)as have macrophages (Stach, 1963, 1969; Clara et al., 1968; Maillet, 1968/69).Both macrophagelike cells and fibroblastlike cells have been identified electron microscopically in muscularis externa of mouse small intestine (Rumessen et al., 1982; Thuneberg, 1982). With respect to the identification if ICC-I and ICC-I1 the results obtained in the present study with ZIO methods are in accordance with supravital methylene blue studies (Thuneberg, 1982).With respect to the level of PMP it is reasonable to assume that those ZIO-positive cells which are preferentially supplied with varicose axons are identical with the innervated ICC-I11 identified ultrastructurally (Rumessen et al., 1982; Thuneberg, 1982). In the following our results will be discussed, as they relate to four distinct regions of the musculature. Longitudinal muscle layer and serosa The absence of nerves within the subserous compartment and the longitudinal muscle layer confirms results obtained with methods other than ZIO (Thuneberg, 1982). Apart from the present results, light microscopical illustrations of ICC-I1 have been provided only by Richardson (1960, silver impregnation), and Thuneberg (1982, methylene blue). The latter method seems to be unsatisfactory as applied to ICC-11, since only occasional single or small groups of ICC-I1 are stained. Electron microscopy of ICC-I1 (Thuneberg, 1982) reveals a fibroblastlike Fig. 18. The plexus of ICC-I at the level of AP. The cells are extensively branched and exhibit close relations t o tertiary elements of A P (arrows).N: nuclei of ICC-I. X 290 and X 720. Fig. 19.a. The subserous plexus of ICC-11. This plexus has a lower cellular concentration than the plexus of ICC-I (Fig. 18a). Note that nerves are absent. X 290. b. A larger magnification of a single ICC-11, branching in a dichotomic pattern. The main axis of the cell is roughly in the longitudinal direction. N: nucleus. X 610. Fig. 20. An ICC-IV with nucleus (N) and processes (arrows) associated with a fasciculus of PMS. ICC-IV are generally bipolar with long unbranching processes. X 720. 125 ultrastructure. The nature of these cells is obscure at the moment. The region between the main muscle layers As a consequence of the simultaneous ZIO staining of nerves and ICC, we have observed that the finest strands of Auerbachs plexus (tertiary fasciculi) exhibit very close contacts with the extensive plexus of ICC-I;the nature of these contacts deserves further investigation, as they may be of importance to the concept of a role of ICC-I in the origin or modulation of electrical impulses in mouse small intestine, as suggested by Thuneberg (1982). Outer part of the circular muscle layer Regularly spaced axon bundles, oriented parallel to the muscle cells of the circular layer and situated invariably at the inner aspect of the main plexus of Auerbach (Schabadasch, 1930; Stohr, 1957) have generally been regarded as a component of the secondary plexus of Auerbach. In mouse small intestine evidence exists (Thuneberg, 1982) that these fasciculi are actually located within the interstices of the outermost cell layer of the circular muscle coat (hence the term plexus muscularis (strati circularis) superficialis) (PMS).The present investigation provides further evidence of a very regular and well-developednerve plexus at this location. Apart from the present findings, interstitial cells (ICC-IV)apposed to PMS and to oblique or perpendicular fasciculi within the outer division of the circular muscle layer have been investigated only electron microscopically (Thuneberg, 1982). ICC-IV exhibit a fibroblastlike ultrastructure. Nothing is known of their function. Region between outer and inner subdivision of the circular muscle coat Variable definitions of the term “plexus muscularis profundus” occur in the literature. Cajal (1893, 1911) described a rich plexus of nerves and ICC situated at the internal surface of muscularis externa of the mammalian gastrointestinal tract (“plexussous-musculeux ou musculaire profound). That Cajal referred to this location only is evident from his drawings of Golgi-impregnated guinea pig intestine (Cajal, 1911).SinceCajaldidnotdividethecircular layer of muscularis externa of mammalian small intestine in two separate layers (aslater reported by Li, 1937, 1940), he did not distinguish clearly between small and large in- 126 J.J. RUMESSEN AND L. THUNEBERG testine with respect to PMP. No local plexiform concentration of nerve fibers has been reported within the circular muscle layer of mammalian large intestine, but a plexus of nerves and ICC perhaps corresponding to “plexus sous-musculeux” of Cajal (1911) has been described just luminal to the circular muscle of colon in guinea pig and rat (Stach, 1969, 1972).Stach refers to this plexus as “plexus entericus (submucosus) extremus” and regards it as the outermost extension of plexus submucosus. This nerve plexus may be homologous to PMP of small intestine, with which it seems to share some similarities, regarding morphology and relations to ICC. It would be of interest to investigate the relation between PMP and its possible homologue in colon with respect to presumptive transitions and ultrastructural similarities. Consequently, the term plexus muscularis (strati circularis) profundus should be restricted to the nerve plexus situated between the two divisions of the circular muscle layer of small intestine (Cajal, 1893, 1911; Li, 1937, 1940; Taxi, 1965; Gabella, 1972,1974; Duchon et al., 1974; Rumessen et al., 1982; Thuneberg, 1982), and the designation of PMP to all intramuscular nerve bundles of the circular muscle coat in all regions of the digestive tract (Schabadasch, 1930; Ohkubo, 1937a,b; Stohr, 1957; Stach, 1969, 1972; Schardt and van der Zypen, 1974) should be abandoned. Similarly, our findings disagree with a concept of PMP as situated in the interstices of the innermost layer of muscle cells (Furnessand Costa, 1980). Our investigations have revealed PMP as an extensive nerve plexus, confined to the space between the outer and inner subdivisions of the circular muscle layer of mouse small intestine. PMP is equally developed along the entire length of the small intestine. Most nerve fibers present in PMP seem to be derived from AP, and connections to PS are sparse. Ganglion cells are occasionally seen, but are very widely scattered, which explains why they up to now have escaped both light and electron microscopic characterization. The arrangement of nerves in the circular muscle layer in two well-defined planes (PMS and PMP respectively) should be considered in studies of the nervous regulation of the muscle cells of the circular layer. The unequal distribution of fasciculi suggests that PMS and PMP maintain different functions. Our investigations indicate that PMS and other intramuscular fasciculi are engaged with innervation of the outermost layers of the circular muscle (Thuneberg, 1982),whereas PMP preferentially innervates ICC-I11(Rumessen et al., 1982; Thuneberg, 1982). I t has been suggested that PMPprimarily innervates the inner subdivision of the circular muscle layer (Gabella, 1974,1979),but other ultrastructural reports have emphasized the role of associated interstitial cells (ICC-III),as anatomically intercalated between nerve terminals and the outer division of the circular muscle layer. ICC-I11 establish gap junctions to muscle cells of the outer division of the circular layer as well as to each other (Duchon et al., 1974; Rumessen et al., 1982; Thuneberg, 1982). Their close relations to vesicle-containing nerve terminals are well documented (Taxi, 1965;Ono, 1967; Duchon et al., 1974; Yamamoto, 1977; Rumessen et al., 1982; Thuneberg, 1982), and we have found synapselike formations (Rumessen et al., 1982;Thuneberg, 1982). The present study shows that the relation of nerves to ICC-I11 is not a passive or random phenomenon, but that ICC-I11 very often are selectively contacted by axons of PMP and, furthermore, that this arrangement is a general feature throughout the small intestine. A concept of ICC-I11 as primary regulatory cells is further developed and strongly supported by our accompanying ultrastructural study (Rumessen et al., 1982). LITERATURE CITED Cajal. S.R. (1893)Sur les ganglions et plexus nerveux de fintestin. C.R. SOC. Biol. (Paris),5r217-223. Cajal, S.R. (1911)Histologiedu systemenerveuxdel’homme e t des vertebres. Tome 11, Maloine A, Paris, pp. 925-928. Caprilli, R., L. Onori, M. Tonini. and G. Zappono (1970)Slow waves and mechanical activity in the cat colon circular muscle. Rend. R. Gastroenterol.. 283-89. Champy, C. (1913)Granules e t substances rMuisant l’iodure dosmium. J. Anat. Physiol. (Paris). 49r323-343. Clara, C., C. David, and M. Maillet (1968)Ultrastructure des fibres nerveuses amyeliniques fixees par le reactif de Champy-Maillet. 2. Mikrosk. Anat. Forsch., 79:292-308. Coujard, R. (1943) Essais sur la signification chimique de quelques methodes histologiques. 1. Reactions phenoliques, colorations nerveuses et mitochondriales. Bull. Microsc. Appl., 2018:161- 17 3. Daniel, E.E. (1977)Nerves and motor activity of the gut. In: Nerves and the Gut. F.P. Brooks and R.W. Evers, eds. Charles B. Slack, New York, pp. 154-196. Duchon, G., R. Henderson, and E.E. Daniel (1974)Circular muscle layers in the small intestine. In: Roc. 4th Int. Symp. Gastrointestinal Motility. E.E. Daniel, K. Bowes, J.A.L. Gilbert, B. Schofield, T.K. Schnitka and G. Scott, eds. Mitchell Press, Vancouver, pp. 635-646. Furness. J.B., and M. Costa (1980) Types of nerves in the enteric nervous system. Neuroscience, 5rl-20. Gabella. G. (1972) Innervation of the intestinal muscular coat. J. Neurocytol.. 1:341-362. Gabella, G. (1974)Special muscle cells and their innervation in the mammalian small intestine. Cell Tissue Res., 153:63-77. INTERSTITIAL CELLS IN PLEXUS MUSCULARIS PROFUNDUS Gabella. G. (1979) Innervation of the gastrointestinal tract. Int. Rev. Cytol.. 59:129-193. Garrett, J.R. (1965)Electron microscopicalevaluation of the osmic acid-sodium iodide nerve staining technique on salivary tissue. Acta Anat.. 62325-333. Hillarp. N.A. (1959) On the histochemical demonstration of adrenergic nerves with the osmic acid-sodium iodide technique. Acta Anat., 38:379-384. Li. P.L. (1937) Neue Beobachtungen uber die Struktur des Zirkularmuskels im Dunndarm bei Wirbeltieren. Eine vergleichende Studie. Z. Anat. Entw. Gesch., 107:212222. Li, P.L. (1940)The intramural nervous system of the small intestine with special reference to the inner subdivision of its circular muscle. J. Anat.. 74:348-359. Maillet, M. (1959)Modifications de la techniquede Champy au tetrozyde dosmium-iodure de potassium. Resultats de son application a l'etude des fibres nerveuses. C.R. Seanc. SOC. Biol., I53:939-940. Maillet, M. (1963)Le reactif au tetraoxyde d'osmium-iodure du zinc. Z. Mikrosk. Anat. Forsch.. 70.397-425. Maillet. M. (1968169)Etudecritiquedes fixations au tetraoxyde dosmium-iodure. C.R. Assoc. Anat.. 53233-394. Michel, K. (1967) Die Mikrophotografie. 3rd ed. Springer, Wien, New York, pp. 303-304. Ohkubo K. (1937a) Studien uber das intramuralen Nervensystem des Verdauungskanals. 11.Die Plexus myentericus und Plexus subserosus des Meerschweinchens. Jpn. J. Med. Sci. Biol., 6:21-37. Ohkubo, K. (193713)Studien uber das intramurale Nervensystem des Verdauungskanals. 111. Affe und Mensch. Jpn. J. Med. Sci. Biol.. 6:219-247. Ono. M. (1967) Electron microscopic observations on the ganglia of Auerbachs plexus and autonomic nerve endings in muscularis externa of the mouse small intestine. Sapporo Med. J.. 32:56-74. Pellegrino de Iraldi, A. (1977) Significance of the Maillet method (ZIO) for cytochemical studies of subcellular structures. Experientia. 33:l-10. Richardson, K.D. (1960) Studies on the structure of autonomic nerves in the small intestine, correlating the silver-impregnated image in light microscopy with the per- 127 manganate-fixed ultrastructure in electronmicropy. J Anat.. 94:457-472 Rumessen, J.J.. L. Thuneberg, and H.B. Mikkelsen (1982) Plexus muscularis profundus and associated interstitial cells. I1 Ultrastructural studies of mouse small intestine Anat. Rec.. 203:129-146. Schabadasch, A. (1930) Intramurale Nervengeflechte des Darmrohrs. Z. Zellforsch. Mikrosk. Anat., I0:320-385. Schardt, M., and E. van der Zypen (1974) Enzymhistochemische und quantitative Untersuchungen uber die regionalen Unterschiede des intramuralen Nervensystems im Magen-Darm-Kana1 des weissen Laboratoriumsmaus. Acta Anat., 90:403-430. Stach. W. (1963) Die Mailletsche Osmiumsaure-Zinkjodatum-technik als Routinemethode zur darstellung spezieller Strukturen (vege-vegetative Endformation, Endigungsformen des zerebrospinalen Nervensystems, Golgi-apparate, Mitochondrien) in Kurs- und Demonstrationspraparaten. Z. Med. Labortech, 4:7-20. Stach, W. (1969)Neurohistologische Untersuchungen an den Nervengeflechten der Dickdarmwand. Ein Beitrag zur Innervation des Magen-Darmkanals. Thesis, Rostock. Stach. W. (1972) Der Plexus Entericus Extremus des Dickdarmes und seine Beziehungen zu den interstitiellen Zellen (Cajal). Z. Mikrosk. Anat. Forsch., 85:245-272. Stockinger, L.. and J. Graf (1965) Elektronenmikroskopische Analyse der Osmium-Zinkjodid Methode. Mikroskopie, 2016-35. Stohr, P. (1957) Mikroskopische des vegetativen Nervensystems. In: Handbuch der mikroskopischen Anatomie des Menschen. W. Bargmann, ed. Springer, BerlinGottingen-Heidelberg, Vol. 4, Part 5, pp. 1-678. Taxi, J. (1965) Contribution a l'etude dea connexions des neurones moteurs du systeme nerveux autonome. Ann. Sci. Naturelles Zool., 7:413-674. Thuneberg, L. (1982)Theinterstitial cells of Cajal: intestinal pacemaker cells? Adv. Anat. Embr. Cell Biol., 71:1-130. Yamamoto, M. (1977)Electron microscopical studies on the innervation of the smooth muscle and the interstitial cells of Cajal in the small intestine of the mouse and rat. Arch. Histol. Jpn.. 40:171-201.