Atherosclerosis,

335

22 (1975) 335-348

0 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands

THE LIUM

EFFECT

OF

STUDIED

J. B. SILKWORTH,

HYPERCHOLESTEROLEMIA

ON

AORTIC

ENDOTHE-

EN FACE

B. MCLEAN

AND

W. E. STEHBENS*

Department qf Pathology and Specialized Center of Research on Arteriosclerosis, Albany Medical College of Union University and Veterans Administration Hospital, Albany, N. Y. 12208 (U.S.A.)

(Received September 6th, 1974) (Accepted June 9th, 1975)

SUMMARY

The aortic endothelium of rabbits fed a stock diet containing 1 or 2 % cholesterol for intervals of 1 day to 21 weeks was compared with that of normal rabbits on a stock diet without cholesterol. The endothelium was examined en face using the Hautchen technique. The aortic endothelium of the experimental animals revealed an increase in stigmata and stomata, multinucleated giant cells and leukocytes particularly about sites of branching. These changes were also observed in the endothelium overlying lipid deposits but in addition, there was an increase in mitoses, an alteration their

in the orientation

argyrophilic

increasing leukocytes.

number

Key words:

staining of foam

of endothelial

properties.

Closely

cells which

cells and their nuclei, and a change

in

related

appeared

to the endothelium was an to be derived from monocytic

Aortic endothelium - Endothelial permeability - Hiiutchen preparation Hypercholesterolemia - Rabbit

INTRODUCTION

Bizarre atherosclerotic

multinucleated cells are known to be prevalent in the endothelium of aortae in man 1,s. Therefore the effect of dietary-induced hypercho-

This work was supported by U.S. Public Health Service grants HL 14177, HL 15484andresearch funds of the Veterans Administration (No. 8117-01). * Present address: Department of Pathology, Wellington Clinical School, Wellington, 2, New Zealand. Address for reprints: Dr. W. E. Sthebens, Wellington Clinical School, Wellington, 2, New Zealand.

336

J. B. SILKWORTH, B. MCLEAN, W. E. STEHBENS

lesterolemia (Hautchen)

on the endothelium

of rabbit

aortae

was investigated

using the en face

technique.

MATERIALS AND METHODS

rabbits

Hautchen preparations3 were made of the aortae of 45 New Zealand white male weighing approximately 21/a kg at the commencement of the experiments.

These animals comprised 3 groups (Table I>. Group 1 consisted of 19 control rabbits maintained

for varying intervals

of up to

22 weeks on a diet of Wayne rabbit pellets without cholesterol or corn oil. The 23 rabbits of Group 2 were fed a diet of Wayne rabbit pellets (Allied Chemical) with 1 % cholesterol and sufficient corn oil to permit adequate mixing. They were maintained on this diet until sacrificed at varying intervals from 1 day to 21 weeks (Table 1). The three rabbits in Group 3 were fed a 2% cholesterol diet similarly prepared and were sacrificed 7, 8 and 9 weeks after the commencement of the diet. The serum cholesterol for each animal was determined on a sequential multiple autoanalyzer (Technicon SMA 12/60) prior to initiation of the diets and again at the time of sacrifice. When each animal was killed, it was injected intravenously with 1000 units of Sodium Heparin (Liquaemin Sodium, Organon) followed one minute later by a lethal

dose of Sodium

Nembutal.

The neck was shaved

and the right and left

common carotid arteries were exposed immediately. The right common carotid artery was ligated and the left cannulated with fine polyethylene tubing. The femoral arteries were exposed quickly and transected to facilitate drainage of the perfusate. Three control

animals

from Group

1 were perfused

in a retrograde

direction

through

cannulae inserted in the femoral arteries, drainage being effected by transection of the common carotid arteries. Each animal was perfused successively with 200-300 ml of freshly prepared 5.0 % glucose at a rate of 1 ml/set, 100 ml of 0.25 % silver nitrate delivered in 30-40 set, 200-300 ml of 5.0 % glucose and 2 1 of 10 % neutral buffered formalin delivered in approximately one hour from a reservoir. Perfusion usually commenced within 15 min following the overdose of Nembutal. TABLE DETAILS

1 OF EXPERIMENTAL

ANIMALS

Figures in parentheses indicate the number of rabbits subjected to the diet for the respective time intervals. Only one rabbit was maintained for the remaining time intervals. Group

Number

1

of rabbits

Diet

Duration

of diet before sacrifice

19

control

2

23

1% cholesterol

3

3

2 % cholesterol

1 day; 1 (5), 2 (4), 5, 6, 8, 9, 10 (2), 17, 19 and 22 weeks 1, 2, 3, 4, 5, 7, 9 and 11 days; 2 (2), 3 (2), 4 (2); 5, 7, 9, 11, 12, 14, 16, 18 and 21 weeks 7, 8 and 9 weeks

AORTIC

ENDOTHELIUM

The aorta to prevent

337

IN HYPERCHOLESTEROLEMIA

was exposed,

the entrance

carefully

dissected

and ligated proximally

of blood from the heart into the fixed aorta.

to the arch, The perieortic

tissues were carefully dissected away from the entire length of the aorta which was periodically flushed with distilled water to prevent a reflux of blood and collapse of the vessel. The aorta (from the ascending

segment

1 cm distal to the

to approximately

aortic bifurcation) was tied to a stainless steel rack to prevent retraction and removed. It was immersed in distilled water, cleaned of any remaining periaortic fat, opened longitudinally entomological

along the ventral surface, pinned out on a polyethylene strip with pins and stored in 10 % neutral buffered formalin. The aortae of Groups

2 and 3 and 7 aortae of Group 1 were stained for lipid with Fat Red 7B (Sigma) and photographed. The remaining 12 aortae of Group 1 were washed in distilled water for 10 min, quickly rinsed in 50% ethanol and immersed in 70% ethanol for 15 min, rinsed in distilled water and then returned to formalin for further fixation. This procedure was used because it seemed to facilitate the separation of the endothelium from the underlying

tissues in the Hgutchen

with Harris’ hematoxylin

and dehydrated

was cut into 6 to 10 segments and to indicate

the direction

then made and examined

preparation. in ascending

and labelled

to identify

of flow. HButchen

All aortae were then stained grades of alcohol. its position

preparations

with special attention

Each aorta

along the vessel

of the endothelium

were

to the areas of lipid deposition.

RESULTS

Serum

cholesterol

values

Prior to the commencement of the experimental diet, the cholesterol level of each rabbit was less than 100 mg/lOO ml serum. All the control rabbits had normal serum cholesterol levels when sacrificed. The serum cholesterol levels exhibited some individual variability but at the time of sacrifice all animals of Group 2 on the cholesterol diet had developed with one exception

hypercholesterolemia

within 7 days and thereafter

(840 mg/lOO ml at 11 weeks) had serum levels exceeding

100 ml (with a range of 12OCM320 mg/lOO ml). The three animals 2% cholesterol (Group the time of sacrifice. Control aortae from

3) had serum

cholesterol

all animals 1200 mg/

on a diet containing

levels of 22004780

mg/lOO ml at

Group 1

Only 7 aortae from the animals in Group 1 were stained for fat and no demonstrable lipid was present macroscopically. No lesions suggestive of lipid deposition were seen macroscopically in the remaining 12 aortae. Histologically most of the endothelium consisted of a sheet of elongated cells with a tear-drop configuration; the cells measured an average of 96 p in length and 11 p in their greatest width as delineated by argyrophilic lines. They contained round or oval nuclei. The endothelial cells unrelated to branches were usually oriented with their longitudinal axis parallel to the direction of blood flow and most had the nucleus and the bulk of the cytoplasm distal (i.e. downstream) to the midpoint of their longi-

J. B. SILKWORTH,

338 tudinal

axes (Fig. 1). Stigmata,

stomata,

mitotic

B. MCLEAN,

W. E. STEHBENS

figures and multinucleated

cells were

not often seen but in 3 old control animals their incidence rose somewhat. Endothelial cells around orifices of branches were often polygonal in outline and displayed stigmata, divider branch

stomata

and heavy silver staining

of individual

cells especially

about

the flow

which extended like a cresent about the downstream aspect of each aortic (Fig. 2). The argyrophilic lines of the endothelium along this flow divider

were at times more faintly cleate cells, a variable

staining

number

than adjoining

of leukocytes,

cells. There were also a few binu-

and a rare mitotic

figure.

Aortae from cholesterol-fed rabbits (Groups 2 and 3) Macroscopically no lipid was present in the aortae of rabbits on the cholestero!containing diet for 1 to 5 days. Thereafter lipid was found macroscopically in all other animals maintained on the diet for 7 or more days. The distribution of the lipid accumulation in the aorta was similar to that previously described495. In brief, the lipid had a remarkable predilection for sites of branching.

Initially

lipid occurred

at the face and dorsum5

but subsequently

extended

towards the apex along either side of the crescentic flow divider which for a time remained free of fat (the flow divider is the crescentic ridge extending from the face across the apex to the dorsum of the branch6 and it divides the bloodstream into two channels). The fat-staining areas extended to form a triangular shaped extension downstream from the flow divider. Lipid also appeared at the lateral angle of many but not all the branching sites and confluence of the lipid-staining regions resulted

Fig. 1. HIutchen preparation

of normal aortic endothelium.

Blood flow was from left to right. x 580.

339

AORTIC ENDOTHELIUM IN HYPERCHOLESTEROLEMIA

Fig. 2. HPutchen preparation of normal aortic endothelium showing site of small branch. Note crescentic flow divider (arrows). Blood flow was from left to right. x 78.

ultimately in complete encirclement of the orifices. Stippling of the aorta at sites unrelated to branches was observed as early as nine days and occurred particularly in the proximal part of the thoracic aorta. The lesser curvature of the aortic arch immediately distal to the origins of the great vessels of the neck remained relatively free of lipid for at least 14 weeks. In the descending thoracic aorta, there was more lipid accumulation on the right than the left side proximally but the reverse applied to the distal segment. In the aortae with moderate the intimal surface overlying the fat-staining A few severely affected aortae exhibited

and considerable accumulations of lipid, areas was distinctly uneven or granular.

non-uniform

yellowish-brown

pigmentationr.

Microscopically the aortae of rabbits on the cholesterol-containing diet for one to seven days all exhibited a noticeable increase in the prevalence of small stigmata (Fig. 3) and stomata. They were mostly along the argyrophilic lines although stomata were observed within the endothelial cytoplasm also. They occurred in particular in those areas corresponding to sites of known predilection for lipid accumulation. At three days leukocytes (both polymorphonuclear and round cells) were more numerous than in control aortae (Fig. 4). They were not specifically related to stigmata and stomata and it was not possible to determine whether they were on the surface of or deep in the endothelium. At seven days non-endothelial cells closely applied to the endothelium were more numerous than at three days. These aggregations of cells were located at sites of lipid accumulation (Figs. 5 and 6). Some of the cells were polymorphonuclear leukocytes. Others were small round cells as in the aortae at three

340

J. B. SILKWORTH,

B. MCLEAN,

W. E. STEHBENS

Fig. 3. HButchen preparation showing numerous small stigmata (arrows). Also note binucleated cells. Rabbit fed cholesterol for 7 days. x 430.

Fig. 4. Appearance of leukocytes (polymorphonuclear and round cells) closely related to the endothelium of a rabbit fed cholesterol for 3 days. x 580.

AORTIC ENDOTHELIUM

IN HYPERCHOLESTEROLEMIA

341

Fig. 5. Cellular infiltration related to facial portion of flow divider at aortic bifurcation. Many of the cells are lipophages. Rabbit on cholesterol diet for 2 weeks. Blood flow from left to right. x 58.

Fig. 6. Cellular infiltration of face and dorsum at origin of small branch. Lipid was present in both areas. Flow from left to right. x 58.

J. B. SILKWORTH, B. MCLEAN, W. E. STEHBENS

342 days. However,

some of these cells had ovoid

and had abundant In rabbits

granular

nuclei,

were occasionally

binucleate

or foamy cytoplasm.

on the experimental

diet for nine days to four weeks, stigmata

and

stomata were still prevalent (Fig. 7), and large stomata were particularly frequent in endothelium overlying the lipid-containing areas. Moreover in such regions the endothelial cells were slightly pattern

was not as regular

be a slight increase

increased

in the number

In lipid-containing lesions and rarely polymorphous.

in size, particularly

as in the control

aortae

of mitotic

in width, and the endothelial (Fig. 8). There also appeared

figures and binucleate

endothelial

to

cells.

the prevalent, darkly staining nuclei were round or oval These cells had a more abundant and darker blue-gray

cytoplasm than did the endothelial cells and appeared to be predominantly foam cells. The argyrophilic lines overlying lipid deposits were often more faintly stained than those in the non-lipid areas. There was a tendency for silver granules to be related to the foam cells. In those rabbits of Group 2 on the diet for more than four weeks the endothelial changes were considerably more prominent than in those killed before four weeks and the endothelium overlying lipid deposits exhibited greater variation in size and shape and in the location

and orientation

often noticeably

pale and occasionally

and moderately silver deposition

large stomata of endothelial

of the nuclei.

had disappeared

The argyrophilic completely

lines were

(Fig. 9). Stigmata

were fairly numerous as was the heavy cytoplasmic cells (Fig. 10). Multinucleated endothelial cells were al-

Fig. 7. HButchen preparation showing stigmata (STI), stoma (arrow), binucleated cells (BC), and a mitotic figure (MF). Rabb’t fed cholesterol for 2 weeks. x 445.

AORTIC ENDOTHELIUM IN HYPERCHOLESTEROLEMIA

343

Fig. 8. Foam cells closely related to endothelium overlying lipid-containing area in rabbit fed cholesterol for 4 weeks. There is some change in shape and size of the endothelial cells. Flow from left to right. x 580.

Fig. 9. Aggregates of lipophages applied to endothelium the pale staining silver borders. x 240.

in a rabbit on cholesterol for 9 weeks. Note

344

J. B. SILKWORTH, B. MCLEAN, W. E. STEHBENS

Fig. 10. Hautchen preparation showing stigmata (STI), large stomata (STO), heavy cytoplasmic silver deposition (SD), a giant binucleated cell (CC), and several foam cells (FC). Rabbit fed cholesterol for 18 weeks. x 44.5.

Fig. 11. Abrupt margin between normal endothelium (above) and that overlying a lipid-containing lesion below. There are some stigmata and stomata along the argyrophilic lines near the border but endothelial cells below have lost their staining properties for silver nitrate. Numerous other cells (leukocytes and liphophages) are superimposed on the endothelium overlying the lesions. Y 240.

AORTIC ENDOTHELIUM IN HYPERCHOLESTEROLEMIA

so of increased

frequency.

shape, and occasionally

They were mostly

had abnormally

345

binucleate,

often

atypical

in size and

shaped nuclei (Fig. IO). These endothelial

cells

were however never as bizarre nor as large as those in human atherosclerosisr*a or in veins subjected to arteriovenous shunts7. Endothelial mitoses appeared to be more frequent

in the advanced

The change the normal

endothelium

was considerably with

hematoxylin.

appeared

lesions than in rabbits

from the atypical

endothelium

was quite abrupt

increased

exposure

to the diet.

the lipid containing

lesions to

(Fig. 11). The accumulation

and they were often in clumps

On occasions

to have displaced

with a shorter

overlying

a foam

the nucleus

Fig. 12. Foam cells displacing endothelial

cell superimposed

of the endothelial

of foam cells

and more intensely

stained

on the endothelium

cell (Fig.

nuclei of a rabbit fed cholesterol

12). Cells with

for 9 weeks. x 2400.

Fig. 13. Darkly stained nucleus of unidentified cell located within a large stoma displacing endothelial nucleus from a rabbit fed cholesterol for 14 weeks. x 980.

the

346

J. B. SILKWORTH, B. MCLEAN, W. E. STEHBENS

darkly staining nuclei were often present in the center of large stomata but it was difficult to be certain of their identity (Fig. 13). Leukocytes were present but not numerous. DISCUSSION

It has been long known that the application of aqueous silver nitrate to fresh endothelium washed free of excess chloride produces a network of dark brown lines which delineate endothelial cell boundaries 3,s. On account of the extreme susceptibility of endothelium to injury9 and to obviate traumatic artefacts, the aortae in the present experiments were all stained with silver nitrate and fixed in situ before removal. Descriptions of the endothelium from control aortae will reported elsewhere. The orientation and shape of the endothelial cells appear to be determined to some extent by hemodynamic forces*. Poole ef al3 demonstrated that endothelial cells overlying lipid-containing lesions contained considerable quantities of lipid in droplet form. The lipid droplets were bipolar in location. It was not possible to ascertain the presence or absence of lipid in the endothelial preparations of our hypercholesterolemic rabbits because a vacuolated cytoplasm occurs in control rabbits as well as in experimental animals and seems to be related to the technical preparation of the Hautchen. The preparations also demonstrated the occasional displacement of endothelial nuclei by lipophages as in Fig. 12. Poole rt al.3 demonstrated that in conventional histological preparations, lipid-laden macrophages or lipophages were very frequently in close apposition with the endothelium, sometimes lying on the luminal surface, and sometimes on the intimal surface, while at other times they appeared to pass through interendothelial cell gaps or stomata. On scanning electron microscopylO the gross irregularity of the endothelium overlying these experimental lesions may be attributed to the presence of liphophages either on the luminal surface or in the subendothelial space elevating the endothelium. Cells related to large stomata (Fig. 13) in the present experiments could not be identified with certainty but in all likelihood they were lipophages. on the endothelial response to hyperPoole et al3 made no other observations cholesterolemia, as their purpose was to determine whether or not lipid occurred in the endothelial cells and to define the relation of the lipophages to the endothelium. en face in hypercholNo other investigations have been conducted of endothelium esterolemia. In the present investigation the endothelium at sites of lipid deposition in hypercholesterolemic rabbits displays (1) an increase in stigmata and stomata, (2) an increase in mitotic figures and multinucleated giant cells, (3) a loss of axial orientation of the endothelial cells and their nuclei, (4) change in the staining characteristics of endothelial cells overlying lipid-containing areas and (5) increasing numbers of non-endothelial cells adhering to the endothelium. Stigmata and stomata have been shown to represent localized intercellular gaps throughout the endothelium, through which tracer particles and leukocytes can gain access to the subendothelial space it. They also occur at the equator of dividing endo-

AORTIC

ENDOTHELIUM

thelial

cellsis

and

341

IN HYPERCHOLESTEROLEMIA

at sites of endothelial

injury g. Moreover

enhanced

endothelial

permeability to protein-bound dyes has been demonstrated, particularly over lipidcontaining lesionslsJ4. The presence of stigmata and stomata has been reported15 in areas which are sites of predilection to lipid accumulation. display enhanced endothelial permeability to protein-bound

Such areas are known to dyes alsoi6. The increase

in stigmata and stomata in the hypercholesterolemic rabbits of the present investigation may therefore be attributed in part to an enhanced turnover of endothelial cells. Furthermore,

hypercholesterolemic

serum induces increased

mitotic activity and DNA

synthesisl7,1*. Some toxiceffect ofcholesterol administration which Dufflsassumed was an important feature of experimental cholesterol atherosclerosis in rabbits or the toxicity of impurities in commercial preparations of cholesterol20 may also be the cause. The enhanced turnover of endothelial cells in hypercholesterolemiai7J* could account for the increased number of mitoses and giant cells. Endothelial regeneration is characteristically associated with a prevalence of giant cells though their significance is uncertain. In endothelial regeneration, human atherosclerosis and hemodynamically injured endothelium, the endothelial giant cells are frequently very large and bizarre, often containing several nuclei. By contrast, in the present experiments, the aortic endothelium even in the aortae with gross lesions rarely contained more than two nuclei per endothelial cell. This may represent a difference in the degree of injury only or possibly some crucial difference in the etiology between human atherosclerosis and the endothelial response to hypercholesterolemia. There is also the possibility that hypercholesterolemia may aggravate the injurious effects of hemodynamic stress which is held responsible for the enhanced turnover of endothelial cells, the presence of mitoses, stigmata and stomata, and the variation in endothelial cell topography about aortic branches in control rabbits. The abrupt change in endothelial pattern at the margin of lipid-containing lesions is remarkably similar to that observed in areas of post-traumatic regeneration3. The initial increase in the number of leukocytes related to the endothelium of the cholesterol-fed rabbits appears to be an exaggeration of the changes present in control animals. The apparent transformation of some round cells, possibly monocytes, to lipid-laden macrophages (Figs. 4 and 8) suggests that this could be one mechanism by which the experimental arterial lesion acquires lipophages. However, it is also known that lipid accumulates in the smooth muscle cells of the arterial wall in hypercholesterolemic rabbits5 and they may also be a source of lipophages. No muscle was observed in any of our endothelial preparations and consequently transformation of muscle cells to lipophages could not be documented in this study.

REFERENCES 1 LAUTSCH, E. V., MCMILLAN, G. C. AND DUFF, G. L., Techniques for the study of the normal and atherosclerotic arterial intima from its endothelial surface, Lab. Invesf., 2 (1953) 397407. 2 COTTON, R. AND WARTMAN, W. B., Endothelial patterns in human arteries, A&. P&ho/., 71 (1961) 3-12.

348

J. B. SILKWORTH, B. MCLEAN, W. E. STEHBENS

3 POOLE, J. C. F., SANDERS,A. G. AND FLOREY, H. W., Regeneration

4 5 6 7 8 9 10 11

I2 13 14 15 16 17 18 19 20

of aortic endothelium, J. Puthol. Bacterial., 75 (1958) 133-143. ANITSCHKOW,N., Experimental arteriosclerosis in animals. In: E. V. COWDRY (Ed.), Arteriosclerosis: A Survey ofthe Problem, The Macmillan Company, New York, N.Y., 1933. pp. 271-317. STEHBENS,W. E., The renal artery in normal and cholesterol-fed rabbits, Amer. J. Puthol., 43 (1963) 969-985. STEHBENS,W. E., Medial defects of the cerebral arteries of man, J. Patho/. Bucfeviol., 78 (1959), 179-185. FALLON,J. T. AND STEHBENS,W. E., Venous endothelium of experimental arteriovenous fistulas in rabbits, Circ. Rex, 31 (1972) 546-556. FLOREY,H. W., POOLE,J. C. F. AND MEEK, G. A., Endothelial cells and “cement” lines, J. Puthol. Bacterial., 77 (1959) 625-636. STEHBENS,W. E., Reaction of venous endothelium to injury, Lab. Invest., I4 (1965) 449-459. WEBER,G. AND Tosr, P., Some observations with the scanning electron microscope on the development of cholesterol aortic atherosclerosis in the rabbit, Pathol. Europ., 6 (1971) 407-114. MAJNO, G. AND PALADE.G. E., Studies on inflammation. The effect of histamine and serotonin on vascular permeability - An electron microscopic study, J. Biophys. Biochem. Cytol., 11 (1961) 571-605. STEHBENS,W. E., Endothelial cell mitosis and permeability, Quart. J. Exp. Physiol., 50 (1965) 9&92. SOMER,J. B., EVANS, G., AND SCHWARTZ,C. J., Influence of experimental aortic coarctation on the pattern of aortic Evans blue uptake in vivo, Atherosclerosis, 16 (1972) 127-133. VERESS,B., BALINT,A., KOCZE, A., NAGY, Z. AND JELLINEK,H., Increasing aortic permeability by atherogenic diet, Atherosclerosis, 11 (1970) 369-371. GEISSINGER,H. D., MUSTARD,J. F. AND ROWSELL,R. C., The occurrence of microthrombi on the aortic endothelium of swine, Cunad. Med. Ass. J., 87 (1962) 405-408. CAPLAN, B. A. AND SCHWARTZ,C. J., Increased endothelial cell turnover in areas of in vivo Evans Blue uptake in the pig aorta, Atherosclerosis, 17 (1973) 401-417. THOMAS,W. A., FLORENTIN,R. A., NAM, S. C., KIM, D. N., JONES, R. M. AND LEE, K. T., Preproliferative phase of atherosclerosis in swine fed cholesterol, Arch. Pathol., 86 (1968) 621-643. FLORENTIN,R. A., NAM, S. C., LEE, K. T., LEE, K. J. AND THOMAS,W. A., Increased mitotic activity in aortas of swine, Arch. Puthol., 88 (1969) 463-469. DUFF, G. L., Experimental cholesterol arteriosclerosis and its relationship to human atherosclerosis, Arch. Puthol., 20 (1935) 81-123 and 259-304. IMAI, H., LEE, K. T., TAYLOR,C. B. AND WERTHESSEN, N. T., Angiotoxicity of methanol-soluble fraction in altered cholesterol, Circulation, Suppl. 4, 48 (1973) 42.

The effect of hypercholesterolemia on aortic endothelium studied en face.

The aortic endothelium of rabbits fed a stock diet containing 1 or 2% cholesterol for intervals of 1 day to 21 weeks was compared with that of normal ...
3MB Sizes 0 Downloads 0 Views