Narrow superficial injury to rabbit aortic endothelium

Atherosclerosis, 43 (1982) 233-243 Elsevier/North-Holland Scientific

233 Publishers,

Ltd.

Narrow Superficial Injury to Rabbit Aortic Endothelium The Healing

Margaret Department

Process

as Observed

M. Ramsay,

by Scanning

Electron

Microscopy

Louise N. Walker and David E. Bowyer

of Pathology, Unioersity of Cambridge,

Tennis Court Road, Cambridge (Great Britain)

(Received 24 August, 1981) (Revised, received 30 November, 198 1) (Accepted 8 December, 1981)

Summary A study was made of the healing of aortic endothelium in rabbits following the production of a defined superficial injury. This was induced using a fine nylon filament which removed the endothelial cells without producing significant damage to underlying structures. The morphology of the injury and subsequent repair was observed using light microscopy and scanning and transmission electron microscopy. Two forms of injury were produced (a) a longitudinal injury along the full length of the aorta which was 50-80 pm wide (about 5-8 cell widths), (b) a circumferential injury approximately 80 pm wide (about 2 cell lengths). Thirty minutes after injury the exposed tissue was almost devoid of adherent cells, but after 4 h became covered by a sparse monolayer of platelets. Occasional leukocytes were also present from 7 h after injury. Injury tracks were found to repair very quickly; re-endothelialisation being complete by 48 h and there being no sign of injury by 7 days. Key words:

Aortic endothelium - Atherosclerosis - Endothelial Rabbit - Scanning electron microscopy

injury and repair -

Introduction The arterial endothelium acts as a barrier, separating the circulating the deeper layers of the vessel wall, and controlling transport of material Financial

support

was supplied

OOZl-9150/82/0000-0000/$02.75

by May and Baker Ltd., Dagenham, 0 1982 Elsevier/North-Holland

blood from between the

Essex, U.K.

Scientific

Publishers.

Ltd.

234

two. Its structural and functional integrity is vital for maintenance of normal conditions in the arterial wall. Derangement of the intimal and medial environments may follow injury to the endothelium, whether that injury is slight, causing altered permeability of the endothelial cells, or more severe, actually removing these cells. Damage to the endothelial cells in vivo can be caused by several factors such as hypercholesterolaemia [ 11, immunological damage [2], hypertension [3], haemodynamic stress [4-61, endotoxin [7] and noradrenaline [S]. In experimental studies of the effect of damage, widespread use has been made of mechanically induced injury. For example by a brass rod with a roughened tip [9], a diamond-coated cutting catheter [lo] and embolectomy catheters [ 1 l-161. Many of these methods for producing mechanical injury have the disadvantage that the damage is not confined to the intima and is not sufficiently precise to permit measurement of the rate of repair. More recently, catheters which produce a more defined injury confined to the intima have been developed [ 17,181. In the studies reported here we have used a modified version of the catheter described by Reidy and Schwartz [18] to determine the time course of endothelial repair in the rabbit.

Materials and Methods Animals Fifteen Redfern rabbits (an inbred strain of the New Zealand White rabbits, kept in the Department) of mixed sex, aged 14-20 weeks were used in this study. The rabbits were killed at time intervals of 30 min, 2, 4, 7, 13, 20, 30, 48, 53 h, 5 and 7 days after induction of the injury. Catheterisation The injuries were made using a nylon filament, diameter 0.7 mm, inserted into an outer nylon catheter, outside diameter 1.2 mm (Portex Ltd., Hythe, Kent). For the procedure, animals were sedated with Hypnorm@ (Janssen Pharmaceutics, Crown Chemical Co. Ltd., Lamberhurst, Kent) 0.3 ml/kg administered intramuscularly, and anaesthetised with Althesin@ (Glaxo Laboratories, Greenford, U.K.) 0.05 ml/kg administered intravenously. The catheter, with the filament positioned inside the catheter, was inserted into the aorta via the right femoral artery; on reaching the thoracic region the filament was extruded from the tip of the catheter (Fig. 1). To produce a longitudinal injury the whole assembly was pulled down the aorta. To produce a circular injury the filament was rotated whilst the outer catheter was kept stationary. After catheterisation the femoral artery was sutured to restore blood flow. Evans Blue dye (0.5%, w/v in saline) was administered intravenously. 2 ml/kg, 20 min after catheterisation. Fixation and staining Hypnorm@ and Althesin@ were again used to achieve anaesthesia.

The method

of

Fig. I. Nylon catheter position of filament:

with nylon

filament

used to produce

(a), on insertion of catheter:

longitudinal

(h). on withdrawing

and circular catheter.

injuries,

showing

staining and pressure fixation used was that described by Bowyer [19]. Solutions were introduced into the aorta at 100 mm Hg (13.33 kPa) via a cannula inserted into the left carotid artery, the left femoral artery was cut to allow slow outflow of the perfusing fluids. The animals were killed at this point with an overdose of Euthatal@ (May and Baker Ltd., Dagenham, Essex), 0.7 ml/kg administered intravenously. The solutions introduced were: (i) Wash - isotonic sucrose solution (8.75% v/v) buffered with N -2hydroxyethylpiperazine-N’-ethanesulphonic acid (HEPES), 20 mM, adjusted to pH 7.4, at 37’C. This was perfused until draining fluid from the femoral artery was clear of blood. (ii) Stain100 ml of silver nitrate solution (0.2% w/v in wash), at 37°C. This was present in the aorta for approximately 1 min. (iii) Wash-As above, for approximately 1 min. (iv) Fixative-2% glutaraldehyde in Sdrensen’s phosphate buffer, 0.15 M, pH 7.4. After flushing the aorta with fixative for 2 min, the femoral artery was ligated and fixation was allowed to proceed at 100 mm Hg for 1 h. The aorta was then dissected out and placed in fixative for a minimum of 12 h. Preparation of tissue for microscopy (a) Scanning electron microscopy (SEM)

The aorta

was cleaned

of adventitial

tissue,

opened

longitudinally,

and loosely

236

sutured onto a strip of polyethylene sheeting. The tissue was then briefly washed in deionised water, dehydrated in AnalaR acetone (2 X 5 min in 50%; 2 X 5 min in 70%; 3 X 5 min in 95%; 3 X 10 min in 100%) and critical point dried in a Polaron E3000 critical point drier (Polaron Equipment Ltd., Watford, U.K.) using 3 flushes of liquid carbon dioxide. Pieces of dried tissue were mounted on aluminium stubs using high-conductivity silver paint (Electrodag 916, Achesons Colloids Ltd., Prince Rock, Plymouth, U.K.), coated with 20 nm of gold in a Polaron E5000 sputtering unit and viewed in a Stereoscan S600 electron microscope (Cambridge Instruments) at 15 kV. (h) Transmission electron microscopy (TEM) Tissue was post-fixed for 2 h in 1% osmium tetroxide, dehydrated in acetone, and embedded in Spurr low viscosity resin. Thin sections were viewed at 80 kV in a Philips EM 300 electron microscope. (c)Light microscopy (LM) Wet tissue (immersed in fixative) or dried tissue (mounted in DPX) were viewed en face in a Leitz Orthoplan light microscope.

Results The nylon filament produced an injury of reproducible depth and width. Endothelial cells were removed from the vessel wall without damage to the internal elastic lamina for both the longitudinal and circular injuries (Fig. 2).

Fig. 2. Depth of injury showing

undamaged

internal

elastic lamina and adherent

platelets:

X 12ooO TEM.

231

Longitudinal injury The longitudinal injury produced was 5-8 cells in width (Fig. 3) and parallel to the direction of blood flow. For time periods up to 4 h after injury, endothelial cells at the edge of the injured area were loosely attached to the underlying tissue which was probably basement membrane (Fig. 3); thereafter cells proximal to the injury appeared to be more firmly attached and became enlarged. These enlarged cells formed an approximately straight edge to the line of injury (Fig. 4), and gradually covered the injured area. These enlarged migrating cells started to develop prominent nuclei at 7 h after injury and, at 13 h, cells at the edge of the endothelial sheet were seen to be undergoing division (Fig. 5). Endothelial cell cover was fully restored over the injured zone at 48 h (Fig. 6), the cells being narrowed with prominent nuclei. At 5 days this region of increased density over the injured area was still visible; 7days after injury there was no morphological evidence of the longitudinal injury as observed by SEM. Circular injury The circular injury produced was approximately 2 cell lengths in width (‘Fig. 7a). The repair process of the circular injury followed that of the longitudinal injury; cells adjacent to the injury became enlarged (Fig. 7b), developed prominent nuclei, and migrated over the qenuded area. At 31 h regions of circular injury were almost

Fig. 3. The longitudinal injury at 30 min. Damaged stained with silver; X 400 LM.

endothelial

cells at the edge of the injury are heavily

238

Fig. 4. The aorta 13 h after longitudinal visible: X 700 SEM.

Fig. 5. A cell undergoing

division

injury. Enlarged,

13 h after longitudinal

migrating

injury;

cells with prominent

X 1200 SEM

nuclei are clr

239

Fig. 6. The longitudinal nuclei; X 700 SEM.

injury viewed at 48 h, the injured

area is covered

Fig. 7. (0): The circular injury at 30 min: X250 SEM. (h): The circular the injury are firmly attached and enlarged; X 3 IO SEM.

by narrow

cells with prominent

injury at 7 h, cells at the edge of

240

Fig. 8. (u): Detail of adherent platelets on the longitudinal injury at 4 h. no aggregates can be seen: X 5800 SEM. (h): Detail of adherent leukocytes on the circular injury at 24 h; *: 2900 SEM.

completely re-endothelialised, detected by SEM.

5 days after injury

the circular

injury

could

not be

Platelet and leukocyte adherence No difference was observed between the longitudinal and circular injuries as regards the degree of platelet and leukocyte adherence to the exposed subendothelial material. Thirty minutes after injury very few adherent platelets were observed (Fig. 3); thereafter a sparse monolayer of platelets was found to adhere (Fig. 8a) which persisted until re-endothelialisation was achieved. At none of the time intervals studied were platelet aggregates or fibrin deposits found. Leukocytes were also found to adhere to the denuded areas from about 7 h after injury until re-endothelialisation (Fig. 8b); these reached maximum numbers at about 30 h after injury but did not adhere in sufficient number to form aggregates.

Discussion Many studies have been carried out to investigate the response of an artery to endothelial denudation. Major differences have been reported in terms of the time taken for re-endothelialisation, the degree of platelet adherence in denuded areas,

241

and the extent of smooth muscle cell migration into the intima. With embolectomy catheters used in rabbits, it has been reported that some areas remain uncovered by endothelial cells for more than 14 months [20], the luminal surface being covered by a pseudo-endothelium composed of smooth muscle cells. In rats, using a similar method for achieving aortic denudation, re-endothelialisation was complete 2 months after injury [21]. These differences may be due to (i) variations in the extent of damage caused to the vessel wall, (ii) the response to injury may also be speciesdependent, The results of our studies can be directly compared with those of Reidy and Schwartz [22,23], who produced injuries of similar dimensions in the rat aorta. These workers have reported that circular injuries were repaired within 8 h by migration of endothelial cells; longitudinal injuries were re-endothelialised at 48 h by both cell migration and division. Our results are consistent with these observations in that endothelial cover was rapidly restored over small areas of injury. We did not find that circular injuries were covered by endothelial cells very much more rapidly than longitudinal injuries of a similar size, nor that cover of circular injuries occurred solely by cellular migration. Morphological evidence of cell division in the endothelial sheet immediately adjacent to the injured areas was observed for both the circular and longitudinal injuries (see Fig. 5). These results show that small areas of endothelial denudation in the rabbit, which do not involve damage to the media, are covered primarily by lateral spreading of adjacent endothelial cells. These cells subsequently divide to form a region of temporarily increased cell density. The relative importance of these processes of migration and division will depend on the size of the denuded area-very small defects in the endothelial sheet (such as might occur due to haemodynamic stress) could possibly be repaired very rapidly by migration alone. It is conceivable that single cells or small groups of cells may desquamate without exposure of the subendothelium [24]. Exposure of subendothelial material caused a sparse monolayer of platelets to adhere with no evidence of platelet thrombi being formed. Previous studies have also noted that platelets form a monolayer on subendothelial material, with thrombi occurring only where the basement membrane was damaged [17,25]. Investigations of platelet interactions with the basal lamina of blood vessels in vitro have shown that platelets adhere and form a ‘pavement’ but do not aggregate [26,27]. These observations are not consistent with the idea that when an endothelial cell is removed, massive platelet reactions occur [28]. Platelets are probably involved to a greater extent in the later stages of the development of an atherosclerotic lesion where the collagen and other connective tissue components of the media has been exposed [28], rather than as initiators of the process following endothelial damage and desquamation. These present studies support the view that the normal response to endothelial desquamation in vivo is rapid repair of the defect by migrating cells from the adjacent endothelial sheet, without significant platelet reactions or smooth muscle cell proliferation. Thus endothelial desquamation may not give rise to atherosclerotic lesions; it may be that other factors such as hypercholesterolaemia, hypertension and

242

aging, lead to distortion of the repair processes [7]. The nylon filament catheter proved to be a satisfactory tool for producing injury, as it was both simple to use and resulted in the formation of injuries of consistent width and depth.

Acknowledgements We are grateful to Mrs. C. Pye for technical assistance, Mr. M. Pollard and Mr. R. Burgess for animal husbandry, Mr. K. Thurley and the late Mr. W. Moue1 of the Department of Anatomy, University of Cambridge for assistance with electron microscopy, Mrs. L.M. Wright for typing this manuscript.

References 1 Bondjers, G., Brattsand, R., Bylock, A., Hansson, G.K. and Bjorkerud, S., Endothelial integrity and atherogenesis in rabbit with moderate hypercholesterolaemia, Artery, 3 (1977) 395. 2 Minick, C.R., Immunologic arterial injury and atherogenesis. In: A.M. Gotto, Jr., L.C. Smith and B. Allen (Eds.), Atherosclerosis V, Springer-Verlag, New York, 1980, p. 330. 3 Gabbiani, G., Elemer, G.. Guelpa, C., Vallotton, M.B., Badonnel, M.-C. and Htittner, I., Morphologic and functional changes of the aortic intima during experimental hypertension, Amer. J. Path., 96 (1979) 399. 4 Reidy, M.A. and Bowyer, D.E., Scanning electron microscopy of arteries. The morphology of aortic endothelium in haemodynamically stressed areas associated with branches, Atherosclerosis, 26 (1977) 181. 5 Fry, D.L.. Acute vascular endothelial changes associated with increased blood velocity gradients, Circulat. Res., 22 (1968) 165. 6 Bergel, D.H., Nerem, R.M. and Schwartz, C.J., Fluid dynamic aspects of arterial disease, Atherosclerosis, 23 (1976) 253. 7 Reidy, M.A. and Bowyer. D.E., Distortion of endothelial repair - The effect of hypercholesterolaemia on regeneration of aortic endothelium following injury by endotoxin, Atherosclerosis, 29 (1978) 459. 8 Christensen, B.C., Repair in arterial tissue - A scanning electron microscopic (SEM) and light microscopic study on the endothelium of the rabbit thoracic aorta following noradrenaline in toxic doses. Virch. Arch. A. Path. Anat., 363 (1974) 33. 9 Poole, J.C.F., Sanders, A.G. and Florey, H.W., The regeneration ol aortic endothelium, J. Path. Bact., 75 (1958) 133. 10 Bjbrkerud, S., Reaction of the aortic wall of the rabbit after superficial. longitudinal, mechanical trauma, Virch. Arch. A. Path. Anat., 347 (1969) 197. 11 Stemerman, M.B., Spaet, T.H., Pitlick, F.. Cintron, J., Lejnieks, I. and Tiell, M.L., Intimal healing The pattern of reendothelialisation and intimal thickening, Amer. J. Path., 87 (1977) 125. 12 Christensen, B.C. and Garbarsch, C., Repair in arterial tissue - A scanning electron microscopic (SEM) and light microscopic study on the endothelium of rabbit thoracic aorta following a single dilatation injury, Virch. Arch. A. Path. Anat., 360 (1973) 93. 13 Christensen, B.C., Chemnitz, J., Tkocz, I. and Blaabjerg, 0.. Repair in arterial tissue - The role of endothelium in the permeability of a healing intimal surface. Vital staining with Evans Blue and silver staining of the aortic intima after a single dilatation trauma, Acta Path. Microbial. Stand., 85 (1977) 297. 14 Schwartz, S.M., Stemerman, M.B. and Benditt, E.P., The aortic intima, Part 2 (Repair of the aortic lining after mechanical denudation), Amer. J. Path., 81 (1975) 15.

243

15 Haudenschild, C. and Studer, A., Early interactions between blood cells and severely damaged rabbit aorta, Europ. J. Clin. Invest., 2 (1971) 1. 16 Haudenschild, C.C. and Schwartz, SM., Endothelial regeneration, Part 2 (Restitution of endothelial continuity), Lab. Invest., 41 (1979) 407. 17 Hirsch, E.Z. and Robertson, Jr., A.L., Selective acute arterial endothehal injury and repair, Part 1 (Methodology and surface characteristics), Atherosclerosis, 28 (1977) 271. 18 Reidy, M.A. and Schwartz, SM., A new procedure for in vivo wounding of arterial endothelium. Fed. Proc., 38 (1979) 998. 19 Bowyer. D.E., Assessment of endothelial integrity by scanning electron microscopy (SEM) - A summary of appropriate methods for avoiding artefacts. In: W. Auerswald, H. Sinzinger. C. Leithner and W. Feigt (Eds.), Atherogenese 3, Wilhelm Maudrich, Vienna. 1978, p. 12. 20 Spaet, T.H., Sternerman, M.B., Friedman, R.J. and Burns, E.R., Arteriosclerosis in the rabbit aorta-Long-term response to a single balloon injury, Ann. N.Y. Acad. Sci., 275 (1976) 76. 21 Schwartz, S.M., Stemerman, M.B. and Benditt, E.D., The aortic intima, Part 2 (Repair of the aortic lining after mechanical denudation). Amer. J. Path., 81 (1975) 15. 22 Reidy, M.A., Blood and arterial endothehal cells, Fol. Angiol., 27 (1980) 50. 23 Reidy, M.A. and Schwartz, S.M.. Endothelial regeneration, Part 3 (Time course of intimal changes after small defined injury to rat aortic endothelium). Lab. Invest., 44 (1981) 301. 24 Schwartz, SM., Role of endothelial integrity in atherosclerosis, Artery, 8 (1980) 305. 25 Sheppard, B.L. and French, J.E., Platelet adhesion in the rabbit abdominal aorta following the removal of the endothelium - A scanning and transmission electron microscopical study, Proc. Roy. Sot. B, 176 (1971) 427. 26 Huang, T.W., Lagunoff, D. and Benditt, E.P., Nonaggregative adherence of platelets to basal lamina in vitro, Lab. Invest.,31 (1974) 156. 27 Huang. T.W. and Benditt, E.P., Mechanisms of platelet adhesion to the basal lamina, Amer. J. Path., 92 (1978) 99. 28 Ross, R., The arterial wall and atherosclerosis, Ann. Rev. Med., 30 (1979) 1.