Intraarterial cushions of the rat uterine arteryA scanning electron microscope evaluation utilizing vascular casts.код для вставкиСкачать
THE ANATOMICAL RECORD 203:19-29 (1982) lntraarterial Cushions of the Rat Uterine Artery: A Scanning Electron Microscope Evaluation Utilizing Vascular Casts RANDY H. KARDON, DONNA B. FARLEY, PAUL M. HEIDGER, JR., DIANNA E. VAN ORDEN AND Departments of Obstetrics and Gynecology, Pharmacology, and Anatomy, University of Iowa, Iowa City, Iowa 52242 ABSTRACT The intraarterial cushions present in the rat at the points of branching of the main uterine artery have been studied by means of scanning electron microscopy. Such studies confirmed the three-dimensional concept of these structures gained from previous light microscopic serial section reconstructions as incomplete, raised, asymmetric ridges which encompass the branch orifice. The examination of methacrylate corrosion casts of the uterine vasculature with the scanning electron microscope provided a means for evaluating the relative protrusion or retraction of the cushion structures within the vessel lumen, and thus for assessing their role in regulating uterine blood flow in various physiologic states. Cushions were studied in this manner at the stages of the estrous cycle, in castrated animals, and in animals receiving pharmacologic doses of an alpha adrenergic agonist, phenylephrine. Evaluation of the relative depth of the impression left upon the vascular casts by cushions permitted the followingconclusions. The cushions protrudcd maximally (and thus impeded flow most effectively) in castrated animals and in animals treated with the vasoconstrictor, phenylephrine. In contrast, the cushions protruded less in animals in proestrus and estrus. These data suggest that the cushions do respond, either actively, by virtue of the contractile state of the smooth muscle within the cushion, or passively, as a function of overall vessel geometry, to alpha adrenergic stimulation. The contrast in cushion protrusion between the castrated state, and proestrus and estrus, suggests that ovarian hormones exert an influence over the functional morphology of the cushions in a manner which promotes maximal uterine perfusion during those periods of the estrous cycle which are documented as periods of uterine hyperemia. These studies thus provide evidence for the dynamic role of intraarterial cushions in the regulation of uterine blood flow. Intraarterial cushions, variously termed “intimal cushions,” “subendothelial cushions,” or “arterial branch pads (polsters),”have been observed within arteries in a variety of organs (Shanklin and Azzam, 1963; Rosen, 1967; Moffat, 1969; Pogorzelski, 1964; Menschik and Dovi, 1965; Takayanagi et al., 1972; Moffat, 1952; Moffat and Creasey, 1971). In microscopic section, the structures resemble valves within arteries. However, three-dimensional reconstructions of the cushions (Velican and Velican, 1977) have confirmed that the cushion is not a paired structure, as suggested in section, but rather is an elevated, asymmetric ring of tissue which encompasses the lumen of the artery at sites where the vessel gives rise to a collateral branch. The subendothelial cushion 0003-276X/82/2031-0019$03.500 1982 Alan R. Liss. Inc. contains bundles of longitudinally arranged smooth muscle, the orientation of which suggests that it could provide sphincterlike control over blood flow to the collateral branch (Moffat, 1959). The function of the cushion may be particularly important in vascular beds which undergo regular changes in blood flow. In this respect, the uterus is a suitable model for investigating the role of cushions in the regulation of blood flow, owing to the cyclical changes in blood flow which result from the influence of steroid hormones. Received April 6, 1981; accepted November 10,1981, Direct correspondence and reprint requests to: Dr.Paul M. Heidger, Jr., Depertment of Anatomy, University of Iowa, College of Medicine, Iowa City. Iowa 52242. 20 R.H. KARDON, D.B. FARLEY,P.M. HEIDGER.ANDD.E.VANORDEN In addition to their proposed influence on blood flow, intraarterial cushions of the uterus may subserve another function. Numerous right-angle branches occur along the length of the rodent’s main uterine artery, which should give rise to plasma skimming, with consequent lowering of the hematocrit in the microcirculation. However, Fourman and Moffat (1961) have suggested that this probably is not the case in the uterus. These investigators r e ported that the hemoglobin content of uterine blood in the arterial branch is greater than in the parent trunk. They suggested that this finding may be explained by the presence of intraarterial cushions within the uterine artery in that these structures might provide a means for deflecting red blood cells from the axial stream of blood into the right-angle branch. Supporting this hypothesis was their observation that the hematocrit was decreased at rightangle branches of the intestinal artery, where cushions are absent. Aside from their proposed role in the control of blood flow and hematocrit, the function of intraarterial cushions within the uterine vasculature is not well understood. In the present study, we have attempted to correlate the configuration, extent of projection, and position of intraarterial cushions in the arterial wall with the degree of uterine blood flow (low, moderate, or high) to determine if cushions may function in regulating the size and configuration of the branch orifice. Detecting such changes by light microscopy would require serial sections taken from numerous branch sites of the uterine artery, an approach that is both technically difficult and time consuming. The three-dimensional shape of intraarterial cushions may be studied directly utilizing scanning electron microscopy (Yohro and Burnstock, 1973); however, it is technically difficult to expose the cushion and view its shape in its entirety. In order to circumvent these problems we have used scanning electron microscopy to examine the cushion imprints left on corrosion casts of the uterine vasculature. This technique involves infusing a low-viscosity casting medium into the uterine vasculature and allowing it to polymerize. Digestion of the tissue leaves a replica, or mold, of the entire circulatory system which can be viewed with the scanning electron microscope (Kardon and Kessel, 1979; Kessel and Kardon, 1979; Kardon and Kessel, 1980; Murakami, 1978).The shape of the cushion can be evaluated by viewing the impression left by the cushion upon the cast. In the present study, we have used this technique to assess the shape of uterine intraarterial cushions in rats during different stages of the estrous cycle, in pregnant rats, in castrated rats, and in rats preinfused with the alpha agonist, phenylephrine, with the objective of correlating changes in cushion structure with changes in uterine hemodynamics. MATERIALS AND METHODS Animals Female Sprague-Dawley rats (Bio-Labs, Madison, WI) weighing between 175 and 200 gm were housed six to eight per cage under conditions of constant temperature and humidity; a 12-hour lightldark cycle was maintained, with the lights being turned on at 0600 hours. Animals received Purina Formulab Chow and water, ad libitum. Direct SEM observation of intraarterial cushions Sites of branching of the uterine artery were dissected from Sprague-Dawley rats following perfusion fixation with 3% glutaraldehyde in cacodylate buffer. These specimens were dehydrated, dried by the critical-point method, and coated for routine scanning electron microscopy. Estrous cycle group To confirm that rats were undergoing a normal estrous cycle, vaginal smears were taken between 0800 and 0900 hours and graded according to the criteria of Long and Evans (1922).Only those rats completing at least two consecutive estrous cycles were used for the study. Two rats from each stage of the cycle (diestrus, proestrus, estrus, and metestrus) were anesthetized with 50 mglkg pentobarbital midmorning and prepared for vessel casting. Pregnancy group Two female Wistar inbred rats (University of Iowa, Iowa City, IA) that were in the 20th day of pregnancy were anesthetized and prepared for vessel casting. Castrated group In this group, four rats were bilaterally ovariectomized under ether anesthesia and sacrificed at 2 and 4 weeks postovariectomy and prepared for vessel casting. Phenylephnne-treated group Two castrated rats were given a maintenance dose of 0.1 pglkg estradiol benzoate sub- INTRAARTERIAL CUSHIONS 21 solution and casting medium into the vascular system. The flow was set at the maximal flow rate during the infusion of Ringer solution. This flow (40 mllminute) produced an intraarterial infusion pressure of 75 mmHg. During the infusion of Ringer solution, the polymerization of the casting medium was initiated by the addition to the casting mixture of 18 drops of promoter from the Batson’s corrosion kit. Following 15 seconds of mixing, a 60-ml disposable syringe was filled with the medium. Vascular casts The syringe was then connected to a three-way stopcock after the infusion of 40 ml of Ringer The detailed procedures followed in preparsolution and the medium infused at a rate of ing vascular casts for evaluation with the approximately 2 ml/min.; this produced an inscanning electron microscope have been tra-arterial infusion pressure within the physireported in earlier publications (Kardon and Kessel, 1979; Kessel and Kardon, 1979). In or- ologic range (75-100 mmHg). The infusion was stopped when the casting medium began to der to insure a viscosity of the casting medium low enough to fill all divisions of the micro- polymerize as evidenced by an increase in infusion pressure at constant flow. Typically, 16 vasculature, the following modification of Batml of casting medium was introduced into the son’s corrosion compound was employed. The casting medium was prepared by combining vascular system. Polymerization of the casting medium to hardness occurred in approximatethe following chemicals: 10 ml of methyl methacrylate monomer (Aldrich Chemical), 10 ml of ly 20 minutes. The uterus was then excised and placed in a monomer base, 5 ml of catalyst, and dye. The Petri dish containing Tyrode-Ringer solution. monomer base, catalyst, and dye are constitFor light microscopy, the two horns of the uents of Batson’s No. 17 corrosion compound kit (Polysciences).The components were mixed uterus were dissected free; one horn was prewith a magnetic stirring bar in a beaker cov- served in formaldehyde, dehydrated, and emered with Parafilm; all mixing of components bedded in Epon. One-micrometer serial secof the casting medium and their infusion were tions were cut and stained with toluidine blue. performed under a well-ventilated hood owing The second horn was digested and processed to the hazardous nature of the fumes. with the remaining tissues for scanning electron microscopy of the casts. The cast uteri Using a dissecting microscope, a cannula of polyethylene tubing (PE50) was placed retro- were macerated in 6.0 M potassium hydroxide grade to flow within the right external iliac ar- at 60°C. The solution was changed at least tery of each anesthetized rat. Lidocaine (1%) once a day and the casts were rinsed with diswas applied to the outer surface of the artery to tilled water between changes. A total macerafacilitate cannulation. The tubing, which was tion time of 1-2 days was usually required to prefilled with heparinized TyrodeRinger solu- free the casts of all tissue. Following maceration, was connected to a pressure transducer tion, casts of the main uterine artery with its and recorder to monitor arterial pressure dur- primary branches were usually dissected free ing the infusion of casting medium. The left ex- from the rest of the uterine cast, although ternal iliac artery was then cannulated as some specimens were left intact to facilitate above with polyethylene tubing for the infu- study of the entire uterine vascular bed. The sion of the casting medium. After cannulations casts were then dehydrated in ethanol, dried in CO, by the critical-point method, and mounted were completed, a 60-ml syringe was filled with 40 ml of heparinized Tyrode-Ringer solution on aluminum specimen holders using adhesive and connected via a three-way stopcock to the copper tape. Specimens were rendered electron infusion cannula which previously had been conductive by sputter coating with gold (apfilled with heparinized Tyrode-Ringer solution. proximately 300 A thick) and subsequently At the onset of infusion of casting medium into viewed in a JEOL 35C scanning electron microscope operated at an accelerating voltage of the iliac artery, the aorta was ligated below the 16 KV. The specimens were coded so that they origin of the renal arteries. The inferior vena cava was then cut to provide a route for could be randomly viewed, photographed, and drainage. A Harvard constant flow infusion evaluated in an unbiased manner with respect pump was used to introduce the Tyrode-Ringer to the shape and prominence of the imprints of cutaneously on the seventh day postcastration. On the 14th day postcastration, they were anesthetized and perfused with 40 ml of Tyrode-Ringer solution; thereafter, they were perfused with 40 ml of Tyrode-Ringer containing phenylephrine (1 mg/ml). The phenylephrine solution was infused at the same rate as was the Tyrode-Ringer flush and the intravascular pressure was allowed to increase as the vasoconstriction proceeded. 22 R.H.KARDON,D.B.FARLEY,P.M. HEIDGER,ANDD.E.VANORDEN Fig. 1. Scanning electron micrograph of endothelial surfaceof uterine artery a t point of branching. Theorificeof the collateral is bounded by an asymmetric, elliptical ridge, the intraarterial cushion. Note that the region of asymmetry indicated by the arrow is tapered; the point of the taper is directed retrograde to arterial blood flow. The arrowhead indicates the site of a protruding endothelial cell nucleus. X 500. Fig. 2. Scanning electron micrograph of vascular cast of uterine artery and collateral, from region similar to that indicated by arrow in Figure 3. The indentation in the cast a t the base of the collateral was formed by the intraarterid cushion. Comparision of Figures 1 and 2 facilitates the iden- tification of corresponding structures in scanning electron microscopic preparations of cushions in situ, and in vascular casts. The m o w s in each indicate the tapered “quill point” portion; the arrowheads indicate the bulges produced by endothelial cell nuclei (Fig. I), and their “negative image” preserved in the cast (Fig. 2). X 250. Fig. 3. Scanning electron micrograph of cast of uterine vasculature. The uterine artery (A)and vein (V)lie a t the base of the preparation. Collaterals arise from the artery (arrow) and give rise to the profuse myornetrial vascular bed (MVB). X 16. INTRAARTERIAL CUSHIONS cushions preserved within the vascular casts. RESULTS Morphology of casts Direct visualization of the intraarterial cushion with scanning electron microscopy was accomplished using specimens fixed in situ by vascular perfusion. In those fortuitous preparations in which the point of branching within the artery was exposed during the processing procedures, en face views of the cushion structure were afforded in such preparations (Fig. l),an elliptical ridge protruded into the lumen of the main artery, encircling the opening of the side branch. However, the ridge was incomplete at the proximal side of the branch point. Thus, the cushion assumed the overall configuration of a horseshoe, with the open end of the structure directed retrograde to blood flow. Such specimens also depicted the bulging of endothelial nuclei into the lumen of the vessel. By means of vascular casts, a more extensive overview of the complex ramifications of the uterine vasculature was obtained (Figs. 2,3).As depicted in Figure 3, arterial, capillary, as well as venous channels were cast using the technical procedures described. Casts of arteries were easily distinguished from those of adjacent veins by their smaller diameter, nature of branching, round cross-sectional a p pearance, and characteristic fusiform surface depressions (Fig. 2). These latter features correspond to the impressions left by the endothelid cell nuclei which projected into the vessel lumen. Casts of the main uterine artery revealed that the vessel lumen was expanded at the origin of each branch. Furthermore, the base of each branch was surrounded by a groove formed by the raised intraarterial cushion (Fig. 2). Two configurations of the cushion impressions were observed. In the first, the cushion impressions in the casting medium correlated well with the structure observed by direct SEM of the vessel. These were pyriform in shape with the tapering point directed retrograde to the flow of blood (Figs. 2 , 4 , 5 , 9 , 10). The second configuration was that of a sphincteric ring surrounding the orifice of the arterial branch and lacking a tapering point (Figs. 6,7). Both of these configurations varied in the extent to which they projected into the vessel lumen. Light microscopic examination of 1-pm sections of uterine arteries filled with casting m e dium corroborated the observations made using scanning electron microscopy with respect 23 to the degree of projection and positioning of the intraarterial cushion. In cross section, the edges of the cushion resembled valves. In some cases, the cushion ridge projected from the branch site inward, toward the lumen of the main uterine artery. Such cushions would produce an impression on the surface of the main uterine artery cast (Fig 4 and inset). In other specimens, the ridge of the cushion projected perpendicularly into the lumen of the branch vessel (Fig. 6 and inset). Cushions of this configuration would produce an impression at the branch site of the cast resembling a constricting ring. Thus, the prominence of the cushion ridge, as well as the angle and extent to which it projects from the branch site, would appear to determine the configuration of impression left upon a vascular cast. Changes with estrous cycle and pregnancy Intraarterial cushions were observed within the microcirculation of the uterine horn at branch sites of small arteries and arterioles in each of the hormonal states investigated. However, only those cushions at the branch site of the main uterine artery were considered in this investigation. The cushion impressions were evaluated as to whether they appeared deep, indicative of decreased orifice diameter, or shallow, indicating an enlarged orifice. Within the uterine arteries of each experimental group, variability existed as to the depth of cushion impression. The proportions of deep and shallow impressions observed in each group are summarized in Table 1. As shown by the table and Figures 6,7, and 8, the frequency of shallow cushion impressions was increased in casts from animals in states of increased uterine blood flow-proestrus, estrus, and pregnancy. Castrated animals Casts of vessels from this group of animals generally revealed deeply indented cushion impressions (Table 1, and Fig. 9). The impressions left upon the casts by endothelial cell nuclei resembled closely those observed in estrus animals (cf., Fig. 7). Phenylephrine-treated animals As shown in Figure 10, deeply indented cushion impressions were observed in all arterial branches studied in this group. Also, the endothelial cell nuclei produced deep irregular creases on the vascular cast, rather than the oval-shaped depressions observed in nontreated animals. Marked constrictions were 24 R.H.KARD0N.D.B. FARLEY.P.M.HEIDGER.ANDD.E.VANORDEN 25 INTRAARTERIAL CUSHIONS TABLE 1. The percentage of shallow and deep impressions made by intraarterial cushions upon uterine vascular casts in various animal Emups No. of cushion impressions observed Animal groups Cycling animals Diestrus Proestrus Estrus Metestrus Pregnant animals Castrated animals 12 15 8 13 9 23 observed along casts of the uterine artery and vein (Fig. 11).At certain sites along casts of these vessels, areas were identified where casting medium had passed from the lumen of the vessel into its wall, producing thin semicircular bands which surrounded the vessel casts (Fig. 11).This was not observed in casts of the main uterine artery or vein from any of the other experimental groups. DISCUSSION Scanning electron microscopy of corrosion casts has provided evidence for changes in the shape and orientation of intraarterial cushions under varying conditions of blood flow. Our data support that in states of relatively high Figs. 4-7. All figures are scanning electron micrographs of vascular casts of uterine arteries a t the point of branchingof a collateral exhibiting an arterial cushion. Fig. 4. Specimen from animal in diestrus. Note the deep, elliptical impression left by the arterial cushion. Such impressions result from the protrusion of the cushion ridge toward the lumen of the main vessel, as suggested in light microscopic preparations (inset). X 280. Inset: 0.5-pm Eponembedded section, toluidine blue stain. X 195. Fig. 5. Specimen from animal in metestrus. As in diestrus (Fig. 4). a prominent cushion impression lies at the base of the collateral. X 275. Fig. 6. Specimen from animal in proestrus. Note the a b sence of the elliptical cushion impression seen in Figures 4 and 5. The narrow stricture a t the base of the collateral r e flects that the cushion ridge was directed perpendicularly into the lumen of the branch vessel. Both profiles such as this, and shallow profiles as seen in Figure 7, were characteristic of specimens examined from animals in proestrus. X 200. Inset: 0.5ym Epon-embedded section, toluidine blue stain. X 195. Fig. 7. Specimen from animal in estrus. Note the shallow cushion impression a t the base of the collateral vessel. X 340. % Shallow % Deep 33 67 33 25 85 67 75 15 56 22 44 78 uterine blood flow (proestrus, estrus, pregnancy)the casts exhibited shallow cushion impressions, reflecting less protrusion of the cushion into the vessel lumen. In contrast, in states of low uterine blood flow (metestrus, diestrus, castrate, and phenylephrine-treated rats), intraarterial cushions tended to protrude further into the vessel lumen, making deep impressions upon the casts. Thus, a correlation appears to exist between blood flow and the degree to which the cushions project into the vessel lumen. By projecting into the branch vessel lumen, and thereby acting as a sphincter, the ridges of the cushion could effectively decrease the diameter of the opening from the main uterine artery into the collateral branch. Since the resistance to blood flow through vessels is directly proportional to the fourth power of the radius (Burton, 1965), small changes in the opening of each branch site of the main uterine artery could significantly influence blood flow. This concept of arterial cushions functioning in the regulation of blood flow gains support from the work of Harvey and Owen (1976),who measured changes in uterine blood flow during the estrous cycle in the rat, and who documented flows which are in concert with the above observations on uterine arterial cushions. A high percentage of shallow cushion impressions would be intuitively expected in a high blood flow state, such as pregnancy. However, our study detected an almost equal percentage of shallow and deep cushion impressions in arterial specimens from pregnant animals. We feel that one plausible explanation of this result is that regional perfusion of the pregnant uterus most likely represents a dynamic process in which blood flow to different segments of the uterine horn may change significantly over time. Indeed, Markee (1929) reported direct observations of dynamic 26 R.H.KARDON.D.B.FARLEY, P.M.HEIDGER,ANDD.E.VANORDEN INTRAARTERIAL CUSHIONS regional changes in blood flow to segments of the uterus over time. I t is unfortunate that such dynamics cannot be appreciated in casts of the uterine artery, and that resolution of the questions concerning the control of regional perfusion of the uterine vascular bed must await further investigation with other techniques. The degree to which a cushion projects into the vessel lumen may be determined by the contractile state of the smooth muscle within the cushion. At this time it is not known what factors may influence the smooth muscle at this location. Although direct innervation of the cushion smooth muscle has been found to be lacking, nerve fibers which show catecholamine fluorescence have been observed in the vessel adventitia (Falck et al., 1974).Alpha adrenergic receptors are apparently present in this area of the vessel since cushions project further into the vessel lumen after administration of the alpha adrenergic agonist, phenylephrine. These observations suggest that either the smooth muscle cells of the cushion possess adrenergic receptors that respond directly to the drug or that the change in cushion shape is an indirect effect resulting from changes in vessel geometry brought about by vasoconstriction. The presence of alpha adrenergic receptors on cushion smooth muscle cells lacking neuronal innervation may indicate that the cells could be responsive to bloodborne substances that can act on the receptors (e.g., circulating catecholamines, catechol estrogens). Changes in vessel geometry did result from phenylephrine treatment, as eviFigs. 8-10 are scanning electron micrographs of vascular casts of uterine arteries a t points of branching of collaterals. Fig, 8. Specimen from pregnant animal. Note the shallow cushion impression a t the base of the branch vessel. X 360. Fig. 9. Specimen from a castrated animal. In contrast to the pregnant, proestrus. and estrus animals, the cushions from castrated animals were often deep and asymmetric, as depicted here. X 385. Fig. 10. Specimen from phenylephrinetreated animal. Note the deep, asymmetric cushion a t the base of the collateral, and the elongated impressions left by the endothe l i d cell nuclei (arrow). X 335. Fig. 11. Specimen from phenylephrinetreated animal. A region of marked vasoconstriction filled with casting medium is seen in the lower arterial cast. Narrow bands of casting material, circumferentially disposed upon the surface of the vascular casts (arrows). were frequently observed in specimens from this treatment group. X 155. 27 denced in casts by the decreased diameter of both the main uterine artery and its branches. Thus, overall vascular constriction may have been a contributing factor to the enhanced projection of the cushion into the arterial lumen. Cushion position was also affected by the steroid milieu of the vessel. In the group of castrated animals studied, the cushions left a deep impression upon vascular casts, and the branch sites along the main uterine artery appeared to be very constricted. This finding was interesting not only because of the direct corre lation with decreased uterine blood flow that occurs in this state, but also because this represented the state of the cushion in the absence of ovarian hormones. Our study documented the presence of cushion structures at the bifurcation of small uterine arteries and arterioles, as well as at the right-angle branching points of the main uterine artery. We originally attempted to study the casts of the smaller arterial branches within the uterine horn. However, we could not consistently sample the smaller casted vessels because they were frequently obscured by dense capillary plexuses. Attempts to isolate these casted vessels by dissection could not be performed consistently. Only by studying vessels in which we could consistently observe the cushion impressions could an accurate assessment of cushion shape and position be made. The present investigation, therefore, was limited to the study of the cushions surrounding larger arterial vessels, the site upon which previous investigations of uterine arterial cushions have also focused. The correlative light microscopic findings have indicated that the intimal cushions vary not only in the extent of their projection into the vessel lumen, but also with respect to their angle of projection. This finding should not be attributed to perfusion artifact since all animals were perfused with casting medium within a physiologic range of pressure and the hardened casting medium prevented subse quent deformation of the cushions. A consideration of differences in the angle of projection of the cushion ridge appears to hold a plausible explanation for why some cushion impressions on vascular casts appear to result from the presence of a pyriform ridge projecting into the lumen of the main uterine artery, and others from a constricting ring of tissue surrounding the junction between the collateral branch and the main uterine artery. Dif- 28 R.H.KARDON,D.B. FARLEY. P.M. HEIDGER,ANDD.E.VANORDEN ferences in configuration of the cushion impression may be influenced by the angle of vessel branching from the main uterine artery. The existence of a number of differently oriented bands of smooth muscle within the cushion could account for movement of the cushion in more than one plane, thus contributing to changes in its configuration. Phenylephrine administration yielded two unusual findings apart from the effect on cushion shape, i.e., the appearance of semicircular bands of casting material surrounding cast vessels and the change in shape of the endothe l i d cell impressions. The strips of casting material observed to surround casts of large arteries and veins in the phenylephrine-treated group may reflect sites of increased permeability of the endothelium which allowed passage of the casting medium into the subendothelial space. Capillarylike structures, interpreted to represent increased permeability of the vessel wall, have been observed to surround casts of arteriolar divisions in a number of organs (Anderson and Anderson, 1978; Hodde, 1977; Reynolds and Kardon, 1981). However, such structures have not been described previously in association with casts of larger vessels such as we have observed in the phenylephrinetreated vessels. In order to verify that these strips of casting medium were not merely extravasated by high perfusion pressure, two additional animals were perfused at a decreased flow rate such that intravascular pressure did not exceed the physiological level of 75-100 mmHg; vessel casts revealed that the strips were present regardless of perfusion pressure employed. However, cushion impressions observed in this latter group of animals were of both the deep and shallow varieties. The changes in endothelid cell impression, which were supported by correlative light microscopic studies conducted, may reflect a direct effect of the drug on endothelial cells, or an indirect effect involving alterations in the configuration of the underlying vessel wall. These endothelial cell changes and the apparent alteration in perme ability of the endothelium following phenylephrine administration warrant further exploration. The results of these experiments have suggested a possible role for intraarterial cushions in the regulation of uterine blood flow. The techniques employed in this study have permitted an assessment of the dynamics of both shape and extent of projection of intraarterial cushions, which appear to be related to the hor- monal state of the animal. I t is hoped that future investigations into specific factors which may influence the shape of intraarterial cushions, as well as investigations into the quantitative effect that these structures have on blood flow, will contribute to the better understanding of the precise role of intraarterial cushions in uterine physiology. ACKNOWLEDGMENTS The authors wish to thank Linda Radde for her assistance in casting; Susan Wiltse and Jack Burke for preparation of one-pm sections; and Paul Reimann for photographic assistance. These studies were supported by a grant from NIH (HD06380). LITERATURE CITED Anderson, B.G., and W. Anderson (1978)Scanning electronmicroscopy of microcorrosion casts: Intra-cranial and abdominal microvasculature in domestic animals. Am. J. Anat.. 153: 523. Burton, A.C. (1965) Hemodynamics and the physics of the circulation. In: Physiology and Biophysics. T.C. Ruch and H.D. Patton, ed., W.B. Saunders, Philadelphia, p. 528. Falck. B.. S. Gardmark, G . Nybell, Ch. Owman, E. Rosengren, and N:O. Sjoberg (1974)Ovarian influence on the content of norepinephrine transmitter in guinea pig and rat uterus. Endocrinology, 94: 1475-1479. Fourman, J.. and D.B. Moffat (1961) The effect of intra-arterial cushions on plasma skimming in small arteries. J. Physiol. (Lond.). 158: 374-380. Harvey, C.A.. and D.A.A. Owen, (1976) Changes in uterine and ovarian blood flow during the oestrous cycle in rats. J. Endocrinol.. 71: 367-369. Hodde, K. (1977)Scanning electronmicroscopy of microcorrosion casts with special attention on arterial-venous differences. Johari, ed.. IITRIISEM. Chicago, pp. 477-484. Kardon, R.H., and R.G. Kessel(1979) SEM studies on vascular casts of the rat ovary. In: Scanning Electron Microscopy. SEM Inc., OHare, Vol. 111, pp. 743-749. Kardon, R.H., and R.G. Kessel(1980)Three dimensional organization of the hepatic microcirculation in the rodent as observed by scanning electron microscopy of corrosion casts. Gastroenterology, 79: 72-81. Kessel. R.G., and R. Kardon (1979) Tissues and Organs: A text-atlas of Scanning Electron Microscopy. W.H. Freeman, San Francisco. Long, J.A., and H.M. Evans (1922)The oestrus cycle in the rat and its associated phenomena. Mem. Univ. Calif., 6: 1-148. Markee. J.E. (1929)Rhythmicvariationin thevascularityof the uterus of the guinea-pig during the estrus cycle. Am. J. Obstet. Gynecol.. 17: 205-208. Menschik. Z.. and F.S. Dovi (1965) Normally occurring intraluminal projections in the arterial system of the mouse. Anat. Rec., 153: 265-274. Murakami, T. (1978) Methyl methacrylate injection replica method. In: Principles and Techniques of Scanning Electron Microscopy. M. Hayat, ed. Van Nostrand Reinhold, New York, Vol. 6, Ch. 6, pp. 159-169. Moffat, D.B. (1952) A regulatory mechanism in the posterior ciliary arteries of the dog. Nature, 169:1015-1016. Moffat, D.B. (1959)An intra-arterial regulating mechanism in the uterine artery of the rat. Anat. Rec.. 134: 107-123. INTRAARTERIAL CUSHIONS Moffat. D.B. i1969) Intra-arterial cusions in the arteries of the rat's eye. Acta Anat., 72: 1-11. Moffat, D.B., and M. Creasey (1971) The fine structure of the intraarterial cushions a t the origins of the juxtamedullary afferent arterioles in the rat kidney. J. h a t . , (Lond.) 110: 409-419. Pogorzelski, J.K. (1964) Intima cushions in the arteries of the choroid plexuses of the lateral cerebral ventricles. Folia Morphol. 23: 386-388. Reynolds, D.G., and R.H. Kardon (1981)Methods of studying the splanchnic microvascular architecture. In: Measurement of Blood Flow. D.N. Granger and G.B. Bulkley, eds. Williams and Wilkins, Baltimore, pp. 80-88. 29 Rosen, W.C. (1967) The morphology of valves in cerebral arteries of the rat. Anat. Rec., 157: 481-488. Shanklin, W.M.. and N.A. Azzam (1963) On the presence of valves in the rat cerebral arteries. Anat. Rec.. 146: 145- 148. Takayanagi, T., M.L. Rennels, and E. Nelson(1972)An electron microscopic study of intimal cushions in intracranial arteries of the cat. Am. J. Anat., 133: 415-430. Velican, C.. and D. Velican (1977) Studies of human coronary arteries. Acta Anat., 951. 337-385. Yohro, T., and G. Burnstock (1973) Fine structure of "intimd cushions" a t branching sites in coronary arteries of vertebrates. Z. Anat. Entwickl. Gesch.. 140: 187-202.