- Email: [email protected]
Early changes in the arterial wall of chickens fed a cholesterol diet
393
Atherosclerosis, 24 (1976) 393-405 0 Elsevier Scientific Publishing Company,
Amsterdam
EARLY CHANGES IN THE ARTERIAL CHOLESTEROL DIET
M. CHVAPIL, P.L. STITH, CD. ESKELSON
L.M. TILLEMA,
- Printed
in The Netherlands
WALL OF CHICKENS
E.C. CARLSON,
FED A
J.B. CAMPBELL
and
Departments of Surgery and Anatomy, Arizona Medical Center University of Arizona, and Veterans Administration Hospital, Tucson, Ariz. (U.S.A.) (Received 29th October, 1975) (Revised received 16th March, 1976) (Accepted 16th March, 1976)
Summary A total of 160 1-2 day old chickens were fed a 2% cholesterol diet for a period of 8 to 42 days and compared with an equal number of controls. Aortas were analyzed for various indexes of reactivity of connective tissue, cholesterol content and scanning electron microscopy (SEM) characteristics of the endothelial lining. Cholesterol feeding for a period up to 6 weeks resulted in doubling the level of serum cholesterol. It was, however, without effect on the activity of prolyl hydroxylase, lysyl oxidase, collagenase and collagen content in the aortic wall. As early as 3 weeks of feeding significant changes occurred in total and esterified cholesterol content. At the same time endothelial cells were characteristically contracted with several long cytoplasmic elongations and protrusions. A significant decrease of activity of the above enzymes was found in aortic tissue with increased age of the chicken. Collagen content in aortas increased with age of chickens. It is concluded that cholesterol as an atherogenic agent induces marked changes in endothelial cells and lipids of chicken aorta at earlier periods, prior to the activation of connective tissue. Key words: -_
_~
This
work
Address Aria.
Arteriosclerosis - Chicken aorta - Cholesterol diet - Cholesterol free, esterified - Collagen - Collagenase - Endothelial cells - Lysyl oxidase - Prolyl hydroxylase - Scanning electron microscopy
was supported
correspondence
85724.
U.S.A.
in part to:
M.
by
N.I.H.
Chvapil,
grant M.D..
HI,-16386. Drpartmr*nt
oC Surgery.
Arizona
Medical
Crntl*r.
Tucson,
394
Introduction There are several reasons to believe that changes in connective tissue components in aorta are of crucial importance in the general pathology of the developing atherosclerosis. The abundance of connective tissue in arteries, the fact that mesenchymal tissue in general is considered the most reactive biological structure in situations of injury and repair processes [l-3], as well as long chain of evidence on collagen and elastin synthesizing capacity of smooth muscle cells of the media (for example see [4,5 ] ) were indicative of a possible primary role of connective tissue reactivity in the development of arteriosclerosis. In several studies evidence was presented on the enhanced activity of connective tissue components after various atherogenic stimuli [ 6-81. In recent years prolyl hydroxylase (PH), and enzyme synthesizing hydroxyproline from proline linked in collagen polypeptide chains has been shown by various studies to be elevated in the process of atherogenesis [g-12]. This simply indicated the activation of fibrogenic cells, although sometimes incorrectly PH was considered an index of the rate of collagen biosynthesis. Only one study failed to show increased PH in arteriosclerotic lesions [ 131. we Our research interest was to test when in the dynamics of arteriosclerosis can detect the earliest changes in connective tissue reactivity, and how these changes relate to other aortic components, such as endothelial lining, and eventually to the lipid content. As a model of arteriosclerosis we selected chicken, fed a 2% cholesterol diet. Some structural and functional aspects of connective tissue in chicken aortas were already studied by Rucker et al. [ 141. Our intention to study early changes in connective tissue reactivity required selection of sensitive parameters reflecting tissue reactivity. Thus, we used measurements of enzymes such as PH, lysyl oxidase and collagenase indicating the activity or readiness of fibrogenic cells to synthesize collagen, to crosslink collagen by more stable bonds and finally to degrade collagen. In addition, we measured the content of collagen in aortic tissue. This study will show that during 42 days feeding of 2% cholesterol diet no changes occurred in any of the above parameters, while striking changes were present in endothelial lining and lipid content in chicken aorta. Materials and Methods Animals were purchased from Demler Chick Farms in California. A total of 320 Shaver Star-cross-288 White Leghorn cockerels arrived at one or two days of age. Half of these chickens (160) were placed on a 2% cholesterol diet while the controls received regular chick starter diet (CO-OP Chick Starter Mash). The 2% cholesterol diet was composed of CO-OP Chick Starter Mash and cholesterol (nonreagent grade) purchased from Nutritional Biochemicals (Ohio). Animals were weighed at weekly intervals. At given time, i.e., 8, 15, 22 and 42 days of feeding the diet the animals were sacrificed by decapitation. Serum was collected for cholesterol analysis. There were at least 25 chickens in each group at each time interval. Aorta was dissected under steromicroscope at 6 times magnification. Aortas used for biochemical analysis were cut longitudi-
39.5
nally, washed in ice-gold saline, blotted and weighed. Specimens used for scanning electron microscopy were excised from the chicks, of either control or cholesterol fed group, perfused with saline, 37°C warm at low pressure to avoid any distension of the vessel. Aortae were then rinsed gently at room temperature with dilute glutaraldehyde-paraformaldehyde containing 0.1% procaine hydrochloride (pH 7.4), and immersed in Karnovsky’s [15] fixative for 18 h. Aortic lumenal surfaces were exposed, rinsed with 0.2 M sodium cacodylate buffer and post-fixed 1 h in cacodylate buffered 2% 0~0~. After brief rinsing in distilled water, tissue samples were dehydrated in a series of graded acetones and. critical point dried in CO*. Aortas were mounted on aluminum specimen stubs, coated with thin layers of evaporated carbon and gold and observed in a Hitatchi HiScan scanning electron microscope at original magnifications of 175-10,000 diameters. Prolyl hydroxylase assay Freshly dissected aortas were washed in iced saline, blotted and weighed. We homogenized the tissue in a Brinkman Polytron for 30 set in a medium containing 0.23 M sucrose, 0.014 M Tris-HCl buffer (pH 7.5), 50 pg/ml phenylmethylsulfonylfluoride and 50 PM dithiothreitol. After we centrifuged the homogenate at 4°C for 20 min at 15,000 X g, we assayed the supernatant for PH activity. Using the method of Lowry et al. [16] we determined the protein content. DNA content was measured according to Burton [ 171. The procedure for assaying the activity of PH was essentially that of Hutton et al. [18] with modifications by Chvapil and Ehrlich [ 191. In this assay, we used [3,4-‘Hlprolinelabelled collagen synthesized by &day-old chick embryo tissue in the presence of 2,2’-dipyridyl(l r&J) as a substrate. Two ml of incubation medium contained 0.1 ml supernatant, 500,000 dis/min of 3H substrate, 0.05 M Tris-HCl buffer (pH 7.5), 4 mg catalase, 15 mg bovine serum albumin, 1 mM ascorbic acid, 0.5 mM cr-ketoglutarate and 0.1 mit4 ferrous ammonium sulfate. We ran blanks similarly without cu-ketoglutarate. Incubation lasted 30 min at 30” with shaking in unstoppered distillation tubes. Trichloroacetic acid (TCA) added to a final concentration of 0.2 M stopped the reaction. We vacuum-distilled and counted released 3H in Aquasol in a Beckman LS-250 scintillation counter. Determination of extractibility and activity of lysyl oxidase Dissected and washed aortas were homogenized by a Polytron (Brinkmann Instruments, Westburg, N.Y.) in a 0.1 M phosphate-O.16 M NaCl buffer, pH 7.7, in a ratio of 1 : 2.5 (w/v). The homogenate was centrifuged for 20 min at 12,000 X g at 4°C and the supernatant filtered through glass wool to remove insoluble lipids. An equal volume of saturated ammonium sulfate was added to the filtrate and it was stirred for 60 min at 4°C. The pellet obtained after centrifugation (20,000 X g for 20 min at 4°C was resuspended in the original buffer to an original volume, mixed and dialyzed for 3-4 h, first against 0.01 M NaH,P04-0.16 M NaCl, pH 7.7, and then against a tenfold higher molarity of the same buffer for another 16 h at 4°C. The protein content [16] and lysyl oxidase activity (phosphate extractable) was determined in aliquots. The pellet left after phosphate-NaCl extraction was similarly treated with 6 M urea. The pellet was rehomogenized every time with urea solutions in a Poly-
396
tron and was stirred for 3-4 h at 4°C. Further steps were similar to those described above for the phosphate-extractable fractions of lysyl oxidase. The residual pellet was also assayed for the activity of lysyl oxidase. Preparation of the substrate for lysyl oxidase We followed the procedure by Pinnell and Martin [20] with several modifications. Aortas from 17-day-old chick embryos were incubated with [6-‘H] DL-lysine (spec. act. 30.3 ci/mmole, New England Nuclear) in the presence of @uninoproprionitrile (Aldrich Chemicals) for 24 h at 37°C. Lyophilized aortas (2.5 mg) were weighed and homogenized by hand in small, conical glass centrifuge tubes in 1 ml of 0.5 N hydrochloric acid. After being spun at 12,000 X g for 15 min the pellet was washed twice with a 0.1 M phosphate-O.16 M NaCl buffer, pH 7.7, and finally resuspended in 0.5 ml of the same buffer. Assay for lysyl oxidase activity One ml of enzyme extract (see above) with 1 drop of toluene was incubated in stoppered test tubes with 0.5 ml of substrate for 4 h at 37°C under slight shaking. The reaction was stopped by freezing the sample. Determination of the released tritium by distillation was carried out as described by Pinnell and Martin [20]. The total activity of lysyl oxidase represents the activity present in 6 M urea, whether related to weight unit of the total protein in extracts or to the mg of collagen in the sponge. Collagenase activity Collagenase activity was measured in aortas homogenized in a mortar with washed and ignited sand (Mallinckrodt) under cooling in a medium of 0.5% Triton X-100 in distilled water. The procedure is principally that of Ryan and Woesner [ 211 with several modifications as suggested by F. Woessner (personal communication). Finely ground tissue homogenate was equally divided into two portions into ultracentrifuge tubes, spun at 105,000 X g for 30 min at 4°C and the supernate discarded. The pellet of one aliquot was resuspended in 0.01 M CaCl,, 0.04 M Tris-buffer (pH 7.5) and 0.15 M NaCl in a large screw-capped culture tube. The other half served as control (blank) and was resuspended in 1 mM NazEDTA and 0.04 M Tris-buffer and 0.15 M NaCl. Resuspended tissue was warmed up to 37°C in a shaking waterbath, trypsin added in the amount of 0.1 mg trypsin/2 ml sample and incubated for another 5 min. The reaction was stopped with soy bean trypsin inhibitor, 0.36 mg/2 ml sample. Samples were capped and incubated while shaking for 48 h at 37°C. Ampicillin was added to the original Tris-NaCl buffer in the amount of 12 mg/lOO ml buffer medium. The incubation was stopped by cooling and spinning the samples at 25,000 X g for 30 min at 4°C. Supernate was separated into teflon-lined screw cap tubes and hydrolyzed in 6 N HCl, 105°C 16 h. The pellet was hydrolyzed similarly in the original incubation tubes. Hydrolyzed samples were decolorized with activated charcoal evaporated to dryness and redissolved in distilled water. Hydroxyproline in the supernate hydrolysate was assayed according to Woessner’s procedure [22] using benzine extraction; hydroxyproline in the pellet hydrolysate was determined by Technicon Autoanalyzer standard procedure. Collage-
397
nase activity is expressed as percent of hydroxyproline released per 48 h from the pellet of experimental samples after subtracting the value of the blank. Collagen concentration of the sort?. was measured in dried tissue, hydrolyzed in 6 N HCl, at 105°C for 16 h, as total hydroxyproline by Technicon Autoanalyzer standard procedure. Cholesterol determination
The aortas were frozen and maintained in a sealed tube at -80°C until analyzed 2 weeks following their removal from the chicks. Each aorta was weighed and homogenized in 2.5 ml of isopropanol using a Polytron homogenizer. The homogenizer was rinsed once with 2.5 ml of isopropanol and the rinse solution added to its original homogenate. The homogenate was thoroughly mixed and allowed to stand overnight at room temperature before analysis. The extraction method is essentially one developed by Naito and Lewis [23] for extracting total hepatic cholesterol. Following centrifugation 2 ml of the extract and 2 ml of the standard cholesterol solutions were taken to dryness in a water bath maintained at 70°C and using a stream of air to evaporate the solvent. Total and free cholesterol and color development for cholesterol were done according to the method of Zak [24]. Using the above methods on rat aorta a complete recovery of added cholesterol was achieved. The standard cholesterol solutions were treated in an identical manner as that described for the tissue extracts. Serum cholesterol was determined by standardized Technicon Autoanalyzer procedure. Results Body and aorta weights were not affected during 6 weeks of feeding a 2% cholesterol diet (Table 1). Not only that average values of control and experimental chicken were almost identical, but the variability of results was also similar. Serum cholesterol (Fig. 1) in cholesterol-fed chickens was more than twice as high as in control diet fed chickens during 7 to 35 days of treatment. During this period the serum cholesterol content in control chickens was very stable and amounted in average to 176 mg/lOO ml. TABLE
3
EFFECT Days
OF
FEEDING
of
2%
Body
treatment
CHOLESTEROL
weight
DIET
ON
BODY
AND Aorta
(g)
AORTA wet
weight
WEIGHT
OF
(mg)
a Control
b
Cholesterol-fed
1
42+
2
37*
1
9
61+
7
60+
6
14
120i
3
115t
21
180+
8
163
f 17
4
b
controlb 66% 115+
5 9
Cholesterol-fed 64+113
2 L 11
128
f 28
123
t 24
197
? 38
174
i 19
28
268
+ 34
254
* 29
290
r 77
291
+ 81
35
461
+ 34
466
* 21
275
+ 13
295
* 21
a Corresponds b Data
CHICKENS
presented
with
the
as x
age of chickens. + SEM.
b
398
7
Serum Cholesterol in Cholesterol-fed
1 I
0
I
1
Chickens
1
I
I
I
2
3
4
5
Weeks on 2% Cholesterol Diet Fig. 1. Effect of 2% cholesterol diet on the serum cholesterol in young chickens. Data presented as X f SEN. Each value refers to 4 to 6 analyses.
Prolyl hydroxyhse actiuity is commonly considered a sensitive and early index of activation of fibroblasts by various injurious agents [ 111. In this study, however, the activity of this enzyme was not affected by cholesterol feeding in any time period of the experiment. As shown in Table 2, this conclusion is valid irrespective of the reference basis the activity of prolyl hydroxylase is related to, i.e. weight of aorta, or mg of protein in the analyzed tissue sample. The only significant change, found in both control and experimental groups is the decreasing activity of aortic prolyl hydroxylase with increasing age of the chicken. Lysyl oxidase activity measured in 6 M urea extract of aortic tissue represents almost total activity of this enzyme in the analyzed tissue. While only 10% lysyl oxidase activity was found in phosphate buffer extract, 42% of the activity was extracted by 2 M urea and almost 95% extracted into 6 M urea. The problem of various extractable forms of lysyl oxidase is, however, not the object of this study. Here again, cholesterol feeding was without any affect on this enzyme (Table 3). A significant decrease was found, however, between enzyme activity in S- and 22-day-old aortas, irrespective of treatment. The observed age TABLE 2 EFFECT OF FEEDING CHICKEN AORTAS Days of treatment a
8 15 22 42
(3) c (6) (3) (4)
2% CHOLESTEROL
DIET
______ control b
ON PROLYL
HYDROXYLASE
ACTIVITY
Cholesterol b ___~_
cpmlmg protein X lo*
cpm/g wet wt. x 103
cpmlmg protein X lo*
cpm/g wet wt. x 103
137 96 81 21
610 485 548 297
161 f 9.9 78 i 9.0 22 f 3.2
750 471 485 326
+ + * f
8.6 7.1 4.7 3.7
f. r f f
70 37 61 30
a Refers both to age of chickens as well as to days of cholesterol feeding. b Data presented as mean + SE. C Gives number of animals tested in each group.
c f + f.
86 21 52 29
IN
399
TABLE
3
EFFECT -__
OF FEEDING
Days of treatment
2% CHOLESTEROL
a
ON LYSYL
OXIDASE
ACTIVITY
control group b
Cholesterol-fed
57.3 * 3.7 43.5 + 2.7
60.5 45.6
IN CHICKEN
AORTAS
b -___
8 (4) c 22 (6)
+ 3.1 + 2.1
a Embryonated eggs of White Leghorn chickens were processed at given days and 8 days after hatching. b Refers to total activity of lysyl oxidase given as cpm/g wet weight X 103. ’ Gives number
TABLE
of animals tested in each group.
4
ACTIVITY OF LYSYL NATAL PERIOD
OXIDASE
IN CHICKEN
Lysyl oxidase
Age a
cpm/g Prenatal
activity
x 103
AORTA
cpm/mg
protein
AND
EARLY
POST-
55.6 i 4.5 52.3 + 3.0
40.4 -
f 0.06
18 20
60.4 + 3.1 60.7 i 4.1
43.2 49.9
t 0.28 f 0.58
8
60.5 + 3.7
44.0
t 0.98 -___
X lo2
a Embryonated eggs of White Leghorn chickens were processed b Refers to Mean f SE, based on 4 independent analyses.
TABLE
PRENATAL
b
14 16
Postnatal
DURING
at given days and 8 days after hatching.
5
EFFECT AORTAS -~-.
OF
FEEDING
2%
CHOLESTEROL
DIET
ON
COLLAGENASE
ACTIVITY
IN CHICKEN
__--
Days of treatment
a
15 (6) c 22 (5) 42 (6)
Control group b (% digested substrate)
Cholesterol-fed b (% digested substrate)
43.7 + 1.8 23.8 + 6.2 15.7 F 0.95
41.1 f 1.1 30.5 ?; 2.1 17.2 + 1.6 ~-
a Embryonated eggs of White Leghorn chickens were processed at given days and 8 days after hatching. b See Methods, refers to endogenous collagen, digested during 48 h of incubation and measured as released hydroxyproline. ’ Gives number of animals tested in each group. TABLE
6
CONTENT OF TOTAL FED CHICKENS Age (days) duration --
HYDROXYPROLINE
IN AORTAS
Total hydroxyproline
(mg/g
FROM
CONTROL
AND
CHOLESTEROL-
dry weight)
of ferding
~~~~~~-~~~.- ~-
14 21 35 There were 6 chickens
Control .~~_._________.____~_~_._~__ 9.4 f 0.5 9.8 f 0.4 12.8 i 0.2 in each group.
Cholesterol-fed _~__ 9.5 + 0.4 10.3 + 0.2 12.7 +_0.7
400 TABLE EFFECT
I OF
CHOLESTEROL
DIET
ON
CHOLESTEROL
Control Total cholesterol (mg/g) Free cholesterol (mg/g) Esterified cholesterol (mg/g)
(6)
2.06 * 0.19 a 0.81 r 0.06 1.24 t 0.23 a
COMPOSITION Cholesterol-fed
IN CHICKEN
AORTA
(8)
3.32 + 0.32 0.89 + 0.10 2.49 + 0.32
Data refer to 3 weeks old chickens fed for the same time a special diet. Number is given in parenthesis. Results presented per g weight of aorta as 2 + SEM. a Refers to P < 0.01.
of animals
in each group
affect was followed further in the direction to embryonic period of chicken development as shown in Table 4. No change in total lysyl oxidase activity was found in aortas of chicken embryos 8 to 18 days old. Collagenme activity in aortic tissue significantly decreases with age of the animals equally in control as well as experimental groups studied. No affect of cholesterol diet on this enzyme activity was observed (Table 5). Total collagen was determined in aortas of chickens fed a cholesterol diet for 2,3 and 5 weeks. Although no differences were found in concentration of total collagenous hydroxyproline in tissue from control and cholesterol-fed animals, a significant increase (P < 0.001) in collagen concentration was found between 2-week- and 5-week-old chicken aortas in both groups (Table 6). Cholesterol content in chicken aortas was determined only at 22 days of feeding cholesterol diet (Table 7). Statistically significant (P < 0.01) increase in total cholesterol due to increased esterified cholesterol was found. Content of free cholesterol was the same in aortas of either group. Endo the/id lining. At low magnification the luminal surface of thoracic aortae from chickens fed a control diet show prominent folds or ridges parallel to the longitudinal axis of the vessel (Fig. 2). Surface contours of these ridges (Fig. 3) may indicate the general position of endothelial cells, but individual cell boundaries are indistinct. The relatively evenly distributed endothelial surface projections (Fig. 3), appear at higher magnification to consist of small knob-like protrusions interspersed with low ridges along the cell surfaces (Fig. 4). At very low magnification many areas of the aortic surface from cholesterolfed chickens appear similar to those from animals fed a control diet. In other areas, however, the coarse ridges of the lumenal lining become less distinct (Fig. 5). A clear and consistent difference was observed in the lumenal surface of aortas from control and cholesterol-fed chicks, comparing samples from corresponding segments of aorta. Large round or spindle-shaped bulges, possibly consisting of individual endothelial cells, protrude into the lumen. Although the areas between the mounds exhibit few surface specializations, many short knoblike projections crown these endothelial colliculi, and long filamentous structures form bridges between them (Fig. 6). Recent transmission electron microscopic data (Carlson, 1975, unpublished observations) show that knob-like projections, low ridges and filamentous projections on endothelial surfaces are composed of endothelial cell cytoplasm.
Fig. 2. Low power scanning electron micrograph chick fed control diet for 28 days. The lumenal long axis of the aorta. X 165.
of the lumenal surface of thoracic aorta section from lining exhibits a series of ridges or folds parallel to the
Fig. 3. Higher magnification of area similar to Fig. 2. Surface contours may indicate position of endothelial cells, but cellular boundaries are indistinct. Endothelial surface exhibit a relatively even distribution of short cytoplasmic projections. X 1,000. Fig. 4. High power scanning electron micrograph of control aorta. Surface specializations consist primarily of short knob-like projections (P) and low ridges (R) along endothelial cell surfaces. X 12,700.
Discussion This study indicates that feeding newborn chickens a diet with 2% cholesterol for 6 weeks produced no changes in connective tissue biochemistry while at half of this time period striking changes were observed in aortic lipids and endothelial cells. Connective tissue in general as part of mesencymal tissues is commonly considered a very reactive structure [l-3]. There may be several reasons for the inertness of connective tissue in this study. It may well be that cholesterol feeding for 42 days is not an adequate atherogenic stimulus (as far as duration and magnitude of its effect) for chickens to initiate the sequence of reaction8 characterizing the process of atherogenesis. The documented change in the content of aortic cholesterol was a more logical consequence of the cholesterol diet. It may be that we assumed the activation of aortic connective tissue by cholesterol too early after starting the diet. It has been well documented (for review see ]25]) that extensive fibrosis develop8 usually in advanced arteriosclerosis associated with foam cell disintegration, lipid gruel formation, calcification and medial destruction or thrombus organization. We reasoned, how-
402
Fig. 6. Low power scanning electron micrograph of the lumenel surface of thoracic aorata from a chick fed high cholesterol diet for 28 days. Lumenal folds are less distinct than in control samples. Spindleshaped protrusions of the lumenal lining (EC) exhibit a variety of surface structures, while intervening spaces (arrows) are relatively smooth. X 1,300. Fig. 6. Higher magnification scanning electron micrograph of area in Fig. 5. Spindle-shaped endothelial colliculi (EC) are crowned with an array of short projections. Long Mamentous extensions of cytoplasm give a matted fibrillar appearance to some areas of the lumenal lining and occasionally from intercellular bridges (B) of cytoplasm. X 3.600.
ever, that because we selected very sensitive parameters reflecting connective tissue reactivity, we should see eventual changes much sooner than those detected by morphological evaluation of fibrosis or by determination of density of collagenous structures. The documented changes in endothelial lining as well as in aortic cholesterol indicate, however, that connective tissue will react later, secondary to other existing changes in the aorta. It has been shown that prolyl hydroxylase activity is enhanced in early periods after various pathogenic stimuli [ll]. Contrary to some authors who consider this enzyme as indicative of the rate of collagen biosynthesis we believe that the activity of prolyl hydroxylase merely shows the functional state of fibrogenic cells, the potential of the tissue to synthesize collagen. Furthermore, the existence of a rather large pool of inactive prolyl hydroxylase in many tissues [26] and a prompt activation by lactate [27,28], ascorbate [29] or ferrous iron chelators [30] suggests a possible shift in functional pools of this enzyme. As documented by this study, cholesterol diet had no effect on aortic prolyl hydroxylase activity. This is in agreement with a recent report by St. Clair et al. [13] who also did not find changes in this enzyme activity associated with de-
403
velopment of cholesterol-aggrevated atherosclerotic lesions in pigeons. On the other side, the fact that only 4 days after administering epinephrine and thyroxin, Langner and Fuller [lo] found a significant increase in aortic prolyl hydroxylase activity supports our view that the nature of the atherogenic stimulus is decisive for the activation of fibroproliferative reaction in atherosclerotic damage. Keeping in mind that in the early stages of atherogenesis serum proteins flux into the aortic wall, and that platelets adhere to gaps in the endothelial lining [31], we thought we might find some changes in the activity of collagenase. Serum a,-macroglobulines are known inhibitors of collagenase [ 321, while platelets contain this enzyme [33]. Neither of these hypotheses were substantiated by our results as we did not see any changes in the activity of collagenase during the 6 weeks of cholesterol diet feeding. Lysyl oxidase total activity has been measured in aortic tissue for the first time. Previous reports on the activity of this enzyme in aorta [34,35] dealt with only a small portion of the enzyme extractable into phosphate buffer: the biological meaning of the phosphate-extractable form is unknown. Not seeing any effect of cholesterol on aortic lysyl oxidase activity we found a definite decrease in the activity of the enzyme in aorta in early posthatch periods, while during embryonic stages, the total activity of this enzyme remained rather the same. As this enzyme reflects the capacity to form stable covalent crosslinks within collagenous and elastin structures, it is tempting to suggest that quite early in embryogenesis (10 days) there is already a highly active enzyme, activity of which declines at early posthatch developmental stage when enough structural stability in supporting tissue of the vascular system was established. Determination of total collagen is a rather insensitive, static index of collagen metabolism. Because the rate of collagen synthesis in the aortic tissue was also not affected by cholesterol feeding augmented by immunization of chickens by BSA [ 361, we believe that all findings indicate no response of connective tissue in chicken aorta to cholesterol during 6 weeks of feeding. While in vivo cholesterol did not affect connective tissue reactivity, in in vitro systems of chicken fibroblasts it was shown that hypercholesterolemic sera stimulated the synthesis of collagen, but not the synthesis of non-collagenous proteins ]371. In contrast to the documented inertness of aortic connective tissue to high cholesterolemia, striking changes in esterified cholesterol in the arterial wall were seen in 22-day-old cholesterol fed chickens. Results similar to our finding on a significant increase of the concentration of only esterified cholesterol (P < 0.01) have been reported by St. Clair et al. [38] and Day and Proudlock [39] and suggested as early stages in the dynamics of atherosclerosis. In support of our view on the specificity of atherogenic stimulus for the type and sequence of events in arterial wall it is worth mentioning that after epinephrine and thyroxin, Lambert and Fuller [lo] have not seen changes in lipids or rat aorta while prolyl hydroxylase was already activated. Based on the work of several authors (for review see [40,41]) we expected to see definite changes in endothelial cells due to cholesterol feeding in rather early period. The character of changes, as shown in our scanning electron-micrographs, however, exceeded our expectation. Numerous long cytoplasmic ex-
404
trusions of endothelial cells, reminding one of a fibrilar meshwork superimposed on contracted endothelial cells, were impressive. We speculate that these changes indicate irritation or hyperactivity of these cells by the dietary regimen. References 1 Hauss. W.H., Uber die Rolle des Mesenchyms in der Genese der Arteriosklerose. Virchows Arch. Abt. A. Path. Anat., 359 (1973) 135. 2 Hauss. W.H., Jung+Hulsing, G. and Gerlach. U., Die unspezifische Mesenchymreaktion. Georg Thieme Verlag. Stuttgart, 1968. 3 Bertelsen, S.. Chemical studies on the arterial wall in relation to atherosclerosis, Ann. N.Y. Acad. Sci.. 149 (1968) 643. 4 Ross, R. and Klebanoff, S.J., The smooth muscle cell, Part 1 (In viva synthesis of connective tissue proteins), J. Cell Biol.. 50 (1971) 159. 5 Ross, R., The smooth muscle cell, Part 2 (Growth of smooth muscle in culture and formation of elastic fibers), J. Cell Biol., 50 (1972) 172. 6 Levene, C.I., The collagen content of normal and atherosclerotic human aortic intima, Brit. J. EXP. Patti., 43 (1962) 469. 7 Anastassiades, T., Anastassiadis. P. and Denstedt, O.F. Changes in the connective tissue of atherosclerotic intima and media of the aor& Biochim. Biophys. Acta, 261 (1972) 418. 8 Jarmolych, J., Daoud. A.S., Landau, J.. Fritz, K.E. and McElvene, E., Aortic media explants -Cell proliferation and production of mucopolysaccharides. collagen and elastic tissue, Exptl. Molec. Pathol., 9 (1968) 171. 9 Langner, R.O. and Fuller, G.C.. Demonstration of proline hydroxylase activity in rabbit aorta, Biothem. Biophyr Res. Commun., 36 (1969) 559. 10 Langner. R.O. and Fuller, G.C.. Collagen synthesis in thoracic aortas of rabbits with epinephrine-thyroxine induced arteriosclerosis, Atherosclerosis, 17 (1973) 463. 11 Fuller, G.C. and Langner, R.O., Elevation of aortic proline hydroxylase - A biochemical defect in experimental arteriosclerosis, Science, 168 (1970) 987. 12 Crossley. H.L.. Johnson. A.R., Mauger, K.K., Wood, N.L. and Fuller, G.C., Aortic proline hydroxylase in hypoxia induced arterio lerosis in rabbits, Life Sci., 11 (1972) 869. 13 St.Clair. R.W., Toma, J.J.%n d Lofland. H.B., Proline hydroxylase activity and collagen content of pigeon aortas with naturally-occurring and cholesterol-aggravated atherosclerosis, Atherosclerosis, 21 (1975) 165. 14 Rucker, R.B. and Goettlich-Riemann, W., Isolation and properties of soluble elastin from copper deficient chicks, J. Nutr., 102 (1972) 663. 15 Kamovski, M.J.. A formaldehyde-glutaraldehyde fixation of high osmolality for use in electron microscopy, J. Cell. Biol.. 27 (1965) 137. 16 Lowry, O.H.. Rosebrough. N.J., Farr. A.L. and Randall, R.J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 266. 17 Burton, K.. Study of condition and mechanism of diphenylamine reaction for colorimetric estimation of deoxyribonucleic acid, Biochem. J.. 62 (1966) 315. 18 Hutton, Jr., J.J., Tappel, A.L. and Udenfriend, S., A rapid assay for collagen proline hydroxylase. Anal. Biochem., 16 (1966) 384. 19 Chvapil. M. and Ehrlich. E.. Effect of increased oxygen on hydroxyproline synthesis in the cuticle and body wall of Ascaris lumbricoides, Biochim. Biophys. Acta, 208 (1970) 467. 20 Pinnell. S.R. and Martin, G.R., The cross-linking of collagen and elastin -Enzymatic conversion of IYsine in peptide linkage to o-amino-6 semisldehyde (allysine) by an extract from bone, Proc. Natl. Acad. Sci.. 61 (1968) 708. 21 Ryan. J.N. and Woessner, Jr., J.F., Mammalian collagenase -Direct demonstration in homogenates of involuting rat uterus, Biochem. Biophys. Res. Commun., 44 (1971) 144. 22 Woessner. Jr., J.F., The determination of hydroxyproline in tissue and protein samples containing small proportions of this amino acid, Arch. Biochem. Biophys., 93 (1961) 440. 23 Naito, H.K. and Lewis, L.A.. Rapid simplified method for measuring total bepatic cholesterol. Clin. Chem., 18 (1972) 911. 24 Zak. B., Methods of Clinical Chemistry, Vol. 5, Academic Press, New York, N.Y.. 1966. P. 79. 25 Constantinides, P.. Experimental Atherosclerosis. Elsevier. Amsterdam, 1965. 26 McGee, J.O’D.. Langness. U and Udenfriend, 8. Immunological evidence for an inactive precursor of collagen proline hydroxylase in cultured fibroblasta, Proc. Natl. Acad. Sci., 68 (1971) 1586. 27 Comstock, J.P., Gribble. T.J. and Udenfrlend. S.. Further study on the activation of collagen proline hydroxylase in cultures of L-929 fibroblasta. Arch. Biochem. Biophys., 137 (1970) 115.
405 28 Grlbble, T.J., Comstock, J.P. and Udenfrlend. S., Collagen chain formation and peptidyl proline hydroxylation in monolayer tissue cultures of L-929 fibroblasta, Arch. Biochem. Biophys., 129 (1969) 308. 29 Stassen, F.L., Cardinale, G.J. and Udenfriend, S.. Activation of prolyl hydroxylase in L-929 fibroblasts by ascorbic acid, Proc. NatL Acad. Sci. 70 (1973) 1090. 30 Chvapll. M.. McCarthy, D.. Madden, J.W. and Peacock, Jr., E.E., Effect of 1.10~phenanthroline and desferrioxamine in viva on prolyl hydroxylase hydroxylation of collagen in various tissues of rats, Biochem. Pharmacol.. 23 (1974) 2165. 31 Baumgartner, H.R., Thrombosis. In: Risk Factors and Diagnostic Approaches, Schattauer, Stuttgart, 1972, p. 161. 32 Werb, Z.. Burleigh, M.C., Barrett, A.J. and Starkey, P.M., The interaction of o2-macroglobulin with proteinsses -Binding and inhibition of mammalian collagenases and other metal proteinases, Biothem J.. 139 (1974) 369. 33 Chesney, C.M., Harper, E. and Coman. R.W.. Human platelet collagenase, J. Clin. Invest., 53 (1974) 1647. 34 Harris, E.D., Gonnerman, W.A., Savage, J.E. and O’Dell, B.L., Connective tissue aimine oxidase, Part 2 (Purification and partial characterization of lysyl oxidase from chick aorta). Biochim. Biophys. Acta. 341(1974) 332. 36 Rucker, R.B.. Roensch, L.F.. Savage, J.E. and O’Dell, B.L.. Oxidation of peptidy Wine by an amine oxidase from bovine aorta, Biochem. Biophys. Res. Commun., 40 (1970) 1391. 36 Chvapil, M., Sith, P., Tillema, L.. Carlson. EC., Campbell, J.B. and Eskelson, C.D., Reaction of chicken aorta to the effect of cholesterol diet and lmmunication. Unpublished observation. 37 Ronnemaa, T.. Juva. K. and Kulonen. E.. Effect of hyperlipidemic rat serum on the synthesis of collagen by chick embryo fibroblasts, Atherosclerosis, 21 (1976) 316. 38 St.Clair, R.W., tofland. H.B. and Clarkson, T.B., Influence of duration of cholesterol feeding on esterlflcation of fatty acids by cell-free preparation of pigeon aorta, Circulation Res., 27 (1970) 213. 39 Day, A.J. and Proudlock, J.W., Changes in aortic cholesterol-esterifying activity in rabbits fed cholesterol for three days, Atherosclerosis, 19 (19741 253. 40 Shimatnoto. T., Hyperreactive arterial endothelial cells in the atherogenesls and cyclic AMP phosphodiesterase inhibitor in prevention and treatment of atherosclertic disorders, Jap. Heart J., 16 (19751 76. 41 Shimamoto, T., Injury and repair in arterial tissue, Angiology, 25 (1974) 682.
Atherosclerosis, 24 (1976) 393-405 0 Elsevier Scientific Publishing Company,
Amsterdam
EARLY CHANGES IN THE ARTERIAL CHOLESTEROL DIET
M. CHVAPIL, P.L. STITH, CD. ESKELSON
L.M. TILLEMA,
- Printed
in The Netherlands
WALL OF CHICKENS
E.C. CARLSON,
FED A
J.B. CAMPBELL
and
Departments of Surgery and Anatomy, Arizona Medical Center University of Arizona, and Veterans Administration Hospital, Tucson, Ariz. (U.S.A.) (Received 29th October, 1975) (Revised received 16th March, 1976) (Accepted 16th March, 1976)
Summary A total of 160 1-2 day old chickens were fed a 2% cholesterol diet for a period of 8 to 42 days and compared with an equal number of controls. Aortas were analyzed for various indexes of reactivity of connective tissue, cholesterol content and scanning electron microscopy (SEM) characteristics of the endothelial lining. Cholesterol feeding for a period up to 6 weeks resulted in doubling the level of serum cholesterol. It was, however, without effect on the activity of prolyl hydroxylase, lysyl oxidase, collagenase and collagen content in the aortic wall. As early as 3 weeks of feeding significant changes occurred in total and esterified cholesterol content. At the same time endothelial cells were characteristically contracted with several long cytoplasmic elongations and protrusions. A significant decrease of activity of the above enzymes was found in aortic tissue with increased age of the chicken. Collagen content in aortas increased with age of chickens. It is concluded that cholesterol as an atherogenic agent induces marked changes in endothelial cells and lipids of chicken aorta at earlier periods, prior to the activation of connective tissue. Key words: -_
_~
This
work
Address Aria.
Arteriosclerosis - Chicken aorta - Cholesterol diet - Cholesterol free, esterified - Collagen - Collagenase - Endothelial cells - Lysyl oxidase - Prolyl hydroxylase - Scanning electron microscopy
was supported
correspondence
85724.
U.S.A.
in part to:
M.
by
N.I.H.
Chvapil,
grant M.D..
HI,-16386. Drpartmr*nt
oC Surgery.
Arizona
Medical
Crntl*r.
Tucson,
394
Introduction There are several reasons to believe that changes in connective tissue components in aorta are of crucial importance in the general pathology of the developing atherosclerosis. The abundance of connective tissue in arteries, the fact that mesenchymal tissue in general is considered the most reactive biological structure in situations of injury and repair processes [l-3], as well as long chain of evidence on collagen and elastin synthesizing capacity of smooth muscle cells of the media (for example see [4,5 ] ) were indicative of a possible primary role of connective tissue reactivity in the development of arteriosclerosis. In several studies evidence was presented on the enhanced activity of connective tissue components after various atherogenic stimuli [ 6-81. In recent years prolyl hydroxylase (PH), and enzyme synthesizing hydroxyproline from proline linked in collagen polypeptide chains has been shown by various studies to be elevated in the process of atherogenesis [g-12]. This simply indicated the activation of fibrogenic cells, although sometimes incorrectly PH was considered an index of the rate of collagen biosynthesis. Only one study failed to show increased PH in arteriosclerotic lesions [ 131. we Our research interest was to test when in the dynamics of arteriosclerosis can detect the earliest changes in connective tissue reactivity, and how these changes relate to other aortic components, such as endothelial lining, and eventually to the lipid content. As a model of arteriosclerosis we selected chicken, fed a 2% cholesterol diet. Some structural and functional aspects of connective tissue in chicken aortas were already studied by Rucker et al. [ 141. Our intention to study early changes in connective tissue reactivity required selection of sensitive parameters reflecting tissue reactivity. Thus, we used measurements of enzymes such as PH, lysyl oxidase and collagenase indicating the activity or readiness of fibrogenic cells to synthesize collagen, to crosslink collagen by more stable bonds and finally to degrade collagen. In addition, we measured the content of collagen in aortic tissue. This study will show that during 42 days feeding of 2% cholesterol diet no changes occurred in any of the above parameters, while striking changes were present in endothelial lining and lipid content in chicken aorta. Materials and Methods Animals were purchased from Demler Chick Farms in California. A total of 320 Shaver Star-cross-288 White Leghorn cockerels arrived at one or two days of age. Half of these chickens (160) were placed on a 2% cholesterol diet while the controls received regular chick starter diet (CO-OP Chick Starter Mash). The 2% cholesterol diet was composed of CO-OP Chick Starter Mash and cholesterol (nonreagent grade) purchased from Nutritional Biochemicals (Ohio). Animals were weighed at weekly intervals. At given time, i.e., 8, 15, 22 and 42 days of feeding the diet the animals were sacrificed by decapitation. Serum was collected for cholesterol analysis. There were at least 25 chickens in each group at each time interval. Aorta was dissected under steromicroscope at 6 times magnification. Aortas used for biochemical analysis were cut longitudi-
39.5
nally, washed in ice-gold saline, blotted and weighed. Specimens used for scanning electron microscopy were excised from the chicks, of either control or cholesterol fed group, perfused with saline, 37°C warm at low pressure to avoid any distension of the vessel. Aortae were then rinsed gently at room temperature with dilute glutaraldehyde-paraformaldehyde containing 0.1% procaine hydrochloride (pH 7.4), and immersed in Karnovsky’s [15] fixative for 18 h. Aortic lumenal surfaces were exposed, rinsed with 0.2 M sodium cacodylate buffer and post-fixed 1 h in cacodylate buffered 2% 0~0~. After brief rinsing in distilled water, tissue samples were dehydrated in a series of graded acetones and. critical point dried in CO*. Aortas were mounted on aluminum specimen stubs, coated with thin layers of evaporated carbon and gold and observed in a Hitatchi HiScan scanning electron microscope at original magnifications of 175-10,000 diameters. Prolyl hydroxylase assay Freshly dissected aortas were washed in iced saline, blotted and weighed. We homogenized the tissue in a Brinkman Polytron for 30 set in a medium containing 0.23 M sucrose, 0.014 M Tris-HCl buffer (pH 7.5), 50 pg/ml phenylmethylsulfonylfluoride and 50 PM dithiothreitol. After we centrifuged the homogenate at 4°C for 20 min at 15,000 X g, we assayed the supernatant for PH activity. Using the method of Lowry et al. [16] we determined the protein content. DNA content was measured according to Burton [ 171. The procedure for assaying the activity of PH was essentially that of Hutton et al. [18] with modifications by Chvapil and Ehrlich [ 191. In this assay, we used [3,4-‘Hlprolinelabelled collagen synthesized by &day-old chick embryo tissue in the presence of 2,2’-dipyridyl(l r&J) as a substrate. Two ml of incubation medium contained 0.1 ml supernatant, 500,000 dis/min of 3H substrate, 0.05 M Tris-HCl buffer (pH 7.5), 4 mg catalase, 15 mg bovine serum albumin, 1 mM ascorbic acid, 0.5 mM cr-ketoglutarate and 0.1 mit4 ferrous ammonium sulfate. We ran blanks similarly without cu-ketoglutarate. Incubation lasted 30 min at 30” with shaking in unstoppered distillation tubes. Trichloroacetic acid (TCA) added to a final concentration of 0.2 M stopped the reaction. We vacuum-distilled and counted released 3H in Aquasol in a Beckman LS-250 scintillation counter. Determination of extractibility and activity of lysyl oxidase Dissected and washed aortas were homogenized by a Polytron (Brinkmann Instruments, Westburg, N.Y.) in a 0.1 M phosphate-O.16 M NaCl buffer, pH 7.7, in a ratio of 1 : 2.5 (w/v). The homogenate was centrifuged for 20 min at 12,000 X g at 4°C and the supernatant filtered through glass wool to remove insoluble lipids. An equal volume of saturated ammonium sulfate was added to the filtrate and it was stirred for 60 min at 4°C. The pellet obtained after centrifugation (20,000 X g for 20 min at 4°C was resuspended in the original buffer to an original volume, mixed and dialyzed for 3-4 h, first against 0.01 M NaH,P04-0.16 M NaCl, pH 7.7, and then against a tenfold higher molarity of the same buffer for another 16 h at 4°C. The protein content [16] and lysyl oxidase activity (phosphate extractable) was determined in aliquots. The pellet left after phosphate-NaCl extraction was similarly treated with 6 M urea. The pellet was rehomogenized every time with urea solutions in a Poly-
396
tron and was stirred for 3-4 h at 4°C. Further steps were similar to those described above for the phosphate-extractable fractions of lysyl oxidase. The residual pellet was also assayed for the activity of lysyl oxidase. Preparation of the substrate for lysyl oxidase We followed the procedure by Pinnell and Martin [20] with several modifications. Aortas from 17-day-old chick embryos were incubated with [6-‘H] DL-lysine (spec. act. 30.3 ci/mmole, New England Nuclear) in the presence of @uninoproprionitrile (Aldrich Chemicals) for 24 h at 37°C. Lyophilized aortas (2.5 mg) were weighed and homogenized by hand in small, conical glass centrifuge tubes in 1 ml of 0.5 N hydrochloric acid. After being spun at 12,000 X g for 15 min the pellet was washed twice with a 0.1 M phosphate-O.16 M NaCl buffer, pH 7.7, and finally resuspended in 0.5 ml of the same buffer. Assay for lysyl oxidase activity One ml of enzyme extract (see above) with 1 drop of toluene was incubated in stoppered test tubes with 0.5 ml of substrate for 4 h at 37°C under slight shaking. The reaction was stopped by freezing the sample. Determination of the released tritium by distillation was carried out as described by Pinnell and Martin [20]. The total activity of lysyl oxidase represents the activity present in 6 M urea, whether related to weight unit of the total protein in extracts or to the mg of collagen in the sponge. Collagenase activity Collagenase activity was measured in aortas homogenized in a mortar with washed and ignited sand (Mallinckrodt) under cooling in a medium of 0.5% Triton X-100 in distilled water. The procedure is principally that of Ryan and Woesner [ 211 with several modifications as suggested by F. Woessner (personal communication). Finely ground tissue homogenate was equally divided into two portions into ultracentrifuge tubes, spun at 105,000 X g for 30 min at 4°C and the supernate discarded. The pellet of one aliquot was resuspended in 0.01 M CaCl,, 0.04 M Tris-buffer (pH 7.5) and 0.15 M NaCl in a large screw-capped culture tube. The other half served as control (blank) and was resuspended in 1 mM NazEDTA and 0.04 M Tris-buffer and 0.15 M NaCl. Resuspended tissue was warmed up to 37°C in a shaking waterbath, trypsin added in the amount of 0.1 mg trypsin/2 ml sample and incubated for another 5 min. The reaction was stopped with soy bean trypsin inhibitor, 0.36 mg/2 ml sample. Samples were capped and incubated while shaking for 48 h at 37°C. Ampicillin was added to the original Tris-NaCl buffer in the amount of 12 mg/lOO ml buffer medium. The incubation was stopped by cooling and spinning the samples at 25,000 X g for 30 min at 4°C. Supernate was separated into teflon-lined screw cap tubes and hydrolyzed in 6 N HCl, 105°C 16 h. The pellet was hydrolyzed similarly in the original incubation tubes. Hydrolyzed samples were decolorized with activated charcoal evaporated to dryness and redissolved in distilled water. Hydroxyproline in the supernate hydrolysate was assayed according to Woessner’s procedure [22] using benzine extraction; hydroxyproline in the pellet hydrolysate was determined by Technicon Autoanalyzer standard procedure. Collage-
397
nase activity is expressed as percent of hydroxyproline released per 48 h from the pellet of experimental samples after subtracting the value of the blank. Collagen concentration of the sort?. was measured in dried tissue, hydrolyzed in 6 N HCl, at 105°C for 16 h, as total hydroxyproline by Technicon Autoanalyzer standard procedure. Cholesterol determination
The aortas were frozen and maintained in a sealed tube at -80°C until analyzed 2 weeks following their removal from the chicks. Each aorta was weighed and homogenized in 2.5 ml of isopropanol using a Polytron homogenizer. The homogenizer was rinsed once with 2.5 ml of isopropanol and the rinse solution added to its original homogenate. The homogenate was thoroughly mixed and allowed to stand overnight at room temperature before analysis. The extraction method is essentially one developed by Naito and Lewis [23] for extracting total hepatic cholesterol. Following centrifugation 2 ml of the extract and 2 ml of the standard cholesterol solutions were taken to dryness in a water bath maintained at 70°C and using a stream of air to evaporate the solvent. Total and free cholesterol and color development for cholesterol were done according to the method of Zak [24]. Using the above methods on rat aorta a complete recovery of added cholesterol was achieved. The standard cholesterol solutions were treated in an identical manner as that described for the tissue extracts. Serum cholesterol was determined by standardized Technicon Autoanalyzer procedure. Results Body and aorta weights were not affected during 6 weeks of feeding a 2% cholesterol diet (Table 1). Not only that average values of control and experimental chicken were almost identical, but the variability of results was also similar. Serum cholesterol (Fig. 1) in cholesterol-fed chickens was more than twice as high as in control diet fed chickens during 7 to 35 days of treatment. During this period the serum cholesterol content in control chickens was very stable and amounted in average to 176 mg/lOO ml. TABLE
3
EFFECT Days
OF
FEEDING
of
2%
Body
treatment
CHOLESTEROL
weight
DIET
ON
BODY
AND Aorta
(g)
AORTA wet
weight
WEIGHT
OF
(mg)
a Control
b
Cholesterol-fed
1
42+
2
37*
1
9
61+
7
60+
6
14
120i
3
115t
21
180+
8
163
f 17
4
b
controlb 66% 115+
5 9
Cholesterol-fed 64+113
2 L 11
128
f 28
123
t 24
197
? 38
174
i 19
28
268
+ 34
254
* 29
290
r 77
291
+ 81
35
461
+ 34
466
* 21
275
+ 13
295
* 21
a Corresponds b Data
CHICKENS
presented
with
the
as x
age of chickens. + SEM.
b
398
7
Serum Cholesterol in Cholesterol-fed
1 I
0
I
1
Chickens
1
I
I
I
2
3
4
5
Weeks on 2% Cholesterol Diet Fig. 1. Effect of 2% cholesterol diet on the serum cholesterol in young chickens. Data presented as X f SEN. Each value refers to 4 to 6 analyses.
Prolyl hydroxyhse actiuity is commonly considered a sensitive and early index of activation of fibroblasts by various injurious agents [ 111. In this study, however, the activity of this enzyme was not affected by cholesterol feeding in any time period of the experiment. As shown in Table 2, this conclusion is valid irrespective of the reference basis the activity of prolyl hydroxylase is related to, i.e. weight of aorta, or mg of protein in the analyzed tissue sample. The only significant change, found in both control and experimental groups is the decreasing activity of aortic prolyl hydroxylase with increasing age of the chicken. Lysyl oxidase activity measured in 6 M urea extract of aortic tissue represents almost total activity of this enzyme in the analyzed tissue. While only 10% lysyl oxidase activity was found in phosphate buffer extract, 42% of the activity was extracted by 2 M urea and almost 95% extracted into 6 M urea. The problem of various extractable forms of lysyl oxidase is, however, not the object of this study. Here again, cholesterol feeding was without any affect on this enzyme (Table 3). A significant decrease was found, however, between enzyme activity in S- and 22-day-old aortas, irrespective of treatment. The observed age TABLE 2 EFFECT OF FEEDING CHICKEN AORTAS Days of treatment a
8 15 22 42
(3) c (6) (3) (4)
2% CHOLESTEROL
DIET
______ control b
ON PROLYL
HYDROXYLASE
ACTIVITY
Cholesterol b ___~_
cpmlmg protein X lo*
cpm/g wet wt. x 103
cpmlmg protein X lo*
cpm/g wet wt. x 103
137 96 81 21
610 485 548 297
161 f 9.9 78 i 9.0 22 f 3.2
750 471 485 326
+ + * f
8.6 7.1 4.7 3.7
f. r f f
70 37 61 30
a Refers both to age of chickens as well as to days of cholesterol feeding. b Data presented as mean + SE. C Gives number of animals tested in each group.
c f + f.
86 21 52 29
IN
399
TABLE
3
EFFECT -__
OF FEEDING
Days of treatment
2% CHOLESTEROL
a
ON LYSYL
OXIDASE
ACTIVITY
control group b
Cholesterol-fed
57.3 * 3.7 43.5 + 2.7
60.5 45.6
IN CHICKEN
AORTAS
b -___
8 (4) c 22 (6)
+ 3.1 + 2.1
a Embryonated eggs of White Leghorn chickens were processed at given days and 8 days after hatching. b Refers to total activity of lysyl oxidase given as cpm/g wet weight X 103. ’ Gives number
TABLE
of animals tested in each group.
4
ACTIVITY OF LYSYL NATAL PERIOD
OXIDASE
IN CHICKEN
Lysyl oxidase
Age a
cpm/g Prenatal
activity
x 103
AORTA
cpm/mg
protein
AND
EARLY
POST-
55.6 i 4.5 52.3 + 3.0
40.4 -
f 0.06
18 20
60.4 + 3.1 60.7 i 4.1
43.2 49.9
t 0.28 f 0.58
8
60.5 + 3.7
44.0
t 0.98 -___
X lo2
a Embryonated eggs of White Leghorn chickens were processed b Refers to Mean f SE, based on 4 independent analyses.
TABLE
PRENATAL
b
14 16
Postnatal
DURING
at given days and 8 days after hatching.
5
EFFECT AORTAS -~-.
OF
FEEDING
2%
CHOLESTEROL
DIET
ON
COLLAGENASE
ACTIVITY
IN CHICKEN
__--
Days of treatment
a
15 (6) c 22 (5) 42 (6)
Control group b (% digested substrate)
Cholesterol-fed b (% digested substrate)
43.7 + 1.8 23.8 + 6.2 15.7 F 0.95
41.1 f 1.1 30.5 ?; 2.1 17.2 + 1.6 ~-
a Embryonated eggs of White Leghorn chickens were processed at given days and 8 days after hatching. b See Methods, refers to endogenous collagen, digested during 48 h of incubation and measured as released hydroxyproline. ’ Gives number of animals tested in each group. TABLE
6
CONTENT OF TOTAL FED CHICKENS Age (days) duration --
HYDROXYPROLINE
IN AORTAS
Total hydroxyproline
(mg/g
FROM
CONTROL
AND
CHOLESTEROL-
dry weight)
of ferding
~~~~~~-~~~.- ~-
14 21 35 There were 6 chickens
Control .~~_._________.____~_~_._~__ 9.4 f 0.5 9.8 f 0.4 12.8 i 0.2 in each group.
Cholesterol-fed _~__ 9.5 + 0.4 10.3 + 0.2 12.7 +_0.7
400 TABLE EFFECT
I OF
CHOLESTEROL
DIET
ON
CHOLESTEROL
Control Total cholesterol (mg/g) Free cholesterol (mg/g) Esterified cholesterol (mg/g)
(6)
2.06 * 0.19 a 0.81 r 0.06 1.24 t 0.23 a
COMPOSITION Cholesterol-fed
IN CHICKEN
AORTA
(8)
3.32 + 0.32 0.89 + 0.10 2.49 + 0.32
Data refer to 3 weeks old chickens fed for the same time a special diet. Number is given in parenthesis. Results presented per g weight of aorta as 2 + SEM. a Refers to P < 0.01.
of animals
in each group
affect was followed further in the direction to embryonic period of chicken development as shown in Table 4. No change in total lysyl oxidase activity was found in aortas of chicken embryos 8 to 18 days old. Collagenme activity in aortic tissue significantly decreases with age of the animals equally in control as well as experimental groups studied. No affect of cholesterol diet on this enzyme activity was observed (Table 5). Total collagen was determined in aortas of chickens fed a cholesterol diet for 2,3 and 5 weeks. Although no differences were found in concentration of total collagenous hydroxyproline in tissue from control and cholesterol-fed animals, a significant increase (P < 0.001) in collagen concentration was found between 2-week- and 5-week-old chicken aortas in both groups (Table 6). Cholesterol content in chicken aortas was determined only at 22 days of feeding cholesterol diet (Table 7). Statistically significant (P < 0.01) increase in total cholesterol due to increased esterified cholesterol was found. Content of free cholesterol was the same in aortas of either group. Endo the/id lining. At low magnification the luminal surface of thoracic aortae from chickens fed a control diet show prominent folds or ridges parallel to the longitudinal axis of the vessel (Fig. 2). Surface contours of these ridges (Fig. 3) may indicate the general position of endothelial cells, but individual cell boundaries are indistinct. The relatively evenly distributed endothelial surface projections (Fig. 3), appear at higher magnification to consist of small knob-like protrusions interspersed with low ridges along the cell surfaces (Fig. 4). At very low magnification many areas of the aortic surface from cholesterolfed chickens appear similar to those from animals fed a control diet. In other areas, however, the coarse ridges of the lumenal lining become less distinct (Fig. 5). A clear and consistent difference was observed in the lumenal surface of aortas from control and cholesterol-fed chicks, comparing samples from corresponding segments of aorta. Large round or spindle-shaped bulges, possibly consisting of individual endothelial cells, protrude into the lumen. Although the areas between the mounds exhibit few surface specializations, many short knoblike projections crown these endothelial colliculi, and long filamentous structures form bridges between them (Fig. 6). Recent transmission electron microscopic data (Carlson, 1975, unpublished observations) show that knob-like projections, low ridges and filamentous projections on endothelial surfaces are composed of endothelial cell cytoplasm.
Fig. 2. Low power scanning electron micrograph chick fed control diet for 28 days. The lumenal long axis of the aorta. X 165.
of the lumenal surface of thoracic aorta section from lining exhibits a series of ridges or folds parallel to the
Fig. 3. Higher magnification of area similar to Fig. 2. Surface contours may indicate position of endothelial cells, but cellular boundaries are indistinct. Endothelial surface exhibit a relatively even distribution of short cytoplasmic projections. X 1,000. Fig. 4. High power scanning electron micrograph of control aorta. Surface specializations consist primarily of short knob-like projections (P) and low ridges (R) along endothelial cell surfaces. X 12,700.
Discussion This study indicates that feeding newborn chickens a diet with 2% cholesterol for 6 weeks produced no changes in connective tissue biochemistry while at half of this time period striking changes were observed in aortic lipids and endothelial cells. Connective tissue in general as part of mesencymal tissues is commonly considered a very reactive structure [l-3]. There may be several reasons for the inertness of connective tissue in this study. It may well be that cholesterol feeding for 42 days is not an adequate atherogenic stimulus (as far as duration and magnitude of its effect) for chickens to initiate the sequence of reaction8 characterizing the process of atherogenesis. The documented change in the content of aortic cholesterol was a more logical consequence of the cholesterol diet. It may be that we assumed the activation of aortic connective tissue by cholesterol too early after starting the diet. It has been well documented (for review see ]25]) that extensive fibrosis develop8 usually in advanced arteriosclerosis associated with foam cell disintegration, lipid gruel formation, calcification and medial destruction or thrombus organization. We reasoned, how-
402
Fig. 6. Low power scanning electron micrograph of the lumenel surface of thoracic aorata from a chick fed high cholesterol diet for 28 days. Lumenal folds are less distinct than in control samples. Spindleshaped protrusions of the lumenal lining (EC) exhibit a variety of surface structures, while intervening spaces (arrows) are relatively smooth. X 1,300. Fig. 6. Higher magnification scanning electron micrograph of area in Fig. 5. Spindle-shaped endothelial colliculi (EC) are crowned with an array of short projections. Long Mamentous extensions of cytoplasm give a matted fibrillar appearance to some areas of the lumenal lining and occasionally from intercellular bridges (B) of cytoplasm. X 3.600.
ever, that because we selected very sensitive parameters reflecting connective tissue reactivity, we should see eventual changes much sooner than those detected by morphological evaluation of fibrosis or by determination of density of collagenous structures. The documented changes in endothelial lining as well as in aortic cholesterol indicate, however, that connective tissue will react later, secondary to other existing changes in the aorta. It has been shown that prolyl hydroxylase activity is enhanced in early periods after various pathogenic stimuli [ll]. Contrary to some authors who consider this enzyme as indicative of the rate of collagen biosynthesis we believe that the activity of prolyl hydroxylase merely shows the functional state of fibrogenic cells, the potential of the tissue to synthesize collagen. Furthermore, the existence of a rather large pool of inactive prolyl hydroxylase in many tissues [26] and a prompt activation by lactate [27,28], ascorbate [29] or ferrous iron chelators [30] suggests a possible shift in functional pools of this enzyme. As documented by this study, cholesterol diet had no effect on aortic prolyl hydroxylase activity. This is in agreement with a recent report by St. Clair et al. [13] who also did not find changes in this enzyme activity associated with de-
403
velopment of cholesterol-aggrevated atherosclerotic lesions in pigeons. On the other side, the fact that only 4 days after administering epinephrine and thyroxin, Langner and Fuller [lo] found a significant increase in aortic prolyl hydroxylase activity supports our view that the nature of the atherogenic stimulus is decisive for the activation of fibroproliferative reaction in atherosclerotic damage. Keeping in mind that in the early stages of atherogenesis serum proteins flux into the aortic wall, and that platelets adhere to gaps in the endothelial lining [31], we thought we might find some changes in the activity of collagenase. Serum a,-macroglobulines are known inhibitors of collagenase [ 321, while platelets contain this enzyme [33]. Neither of these hypotheses were substantiated by our results as we did not see any changes in the activity of collagenase during the 6 weeks of cholesterol diet feeding. Lysyl oxidase total activity has been measured in aortic tissue for the first time. Previous reports on the activity of this enzyme in aorta [34,35] dealt with only a small portion of the enzyme extractable into phosphate buffer: the biological meaning of the phosphate-extractable form is unknown. Not seeing any effect of cholesterol on aortic lysyl oxidase activity we found a definite decrease in the activity of the enzyme in aorta in early posthatch periods, while during embryonic stages, the total activity of this enzyme remained rather the same. As this enzyme reflects the capacity to form stable covalent crosslinks within collagenous and elastin structures, it is tempting to suggest that quite early in embryogenesis (10 days) there is already a highly active enzyme, activity of which declines at early posthatch developmental stage when enough structural stability in supporting tissue of the vascular system was established. Determination of total collagen is a rather insensitive, static index of collagen metabolism. Because the rate of collagen synthesis in the aortic tissue was also not affected by cholesterol feeding augmented by immunization of chickens by BSA [ 361, we believe that all findings indicate no response of connective tissue in chicken aorta to cholesterol during 6 weeks of feeding. While in vivo cholesterol did not affect connective tissue reactivity, in in vitro systems of chicken fibroblasts it was shown that hypercholesterolemic sera stimulated the synthesis of collagen, but not the synthesis of non-collagenous proteins ]371. In contrast to the documented inertness of aortic connective tissue to high cholesterolemia, striking changes in esterified cholesterol in the arterial wall were seen in 22-day-old cholesterol fed chickens. Results similar to our finding on a significant increase of the concentration of only esterified cholesterol (P < 0.01) have been reported by St. Clair et al. [38] and Day and Proudlock [39] and suggested as early stages in the dynamics of atherosclerosis. In support of our view on the specificity of atherogenic stimulus for the type and sequence of events in arterial wall it is worth mentioning that after epinephrine and thyroxin, Lambert and Fuller [lo] have not seen changes in lipids or rat aorta while prolyl hydroxylase was already activated. Based on the work of several authors (for review see [40,41]) we expected to see definite changes in endothelial cells due to cholesterol feeding in rather early period. The character of changes, as shown in our scanning electron-micrographs, however, exceeded our expectation. Numerous long cytoplasmic ex-
404
trusions of endothelial cells, reminding one of a fibrilar meshwork superimposed on contracted endothelial cells, were impressive. We speculate that these changes indicate irritation or hyperactivity of these cells by the dietary regimen. References 1 Hauss. W.H., Uber die Rolle des Mesenchyms in der Genese der Arteriosklerose. Virchows Arch. Abt. A. Path. Anat., 359 (1973) 135. 2 Hauss. W.H., Jung+Hulsing, G. and Gerlach. U., Die unspezifische Mesenchymreaktion. Georg Thieme Verlag. Stuttgart, 1968. 3 Bertelsen, S.. Chemical studies on the arterial wall in relation to atherosclerosis, Ann. N.Y. Acad. Sci.. 149 (1968) 643. 4 Ross, R. and Klebanoff, S.J., The smooth muscle cell, Part 1 (In viva synthesis of connective tissue proteins), J. Cell Biol.. 50 (1971) 159. 5 Ross, R., The smooth muscle cell, Part 2 (Growth of smooth muscle in culture and formation of elastic fibers), J. Cell Biol., 50 (1972) 172. 6 Levene, C.I., The collagen content of normal and atherosclerotic human aortic intima, Brit. J. EXP. Patti., 43 (1962) 469. 7 Anastassiades, T., Anastassiadis. P. and Denstedt, O.F. Changes in the connective tissue of atherosclerotic intima and media of the aor& Biochim. Biophys. Acta, 261 (1972) 418. 8 Jarmolych, J., Daoud. A.S., Landau, J.. Fritz, K.E. and McElvene, E., Aortic media explants -Cell proliferation and production of mucopolysaccharides. collagen and elastic tissue, Exptl. Molec. Pathol., 9 (1968) 171. 9 Langner, R.O. and Fuller, G.C.. Demonstration of proline hydroxylase activity in rabbit aorta, Biothem. Biophyr Res. Commun., 36 (1969) 559. 10 Langner. R.O. and Fuller, G.C.. Collagen synthesis in thoracic aortas of rabbits with epinephrine-thyroxine induced arteriosclerosis, Atherosclerosis, 17 (1973) 463. 11 Fuller, G.C. and Langner, R.O., Elevation of aortic proline hydroxylase - A biochemical defect in experimental arteriosclerosis, Science, 168 (1970) 987. 12 Crossley. H.L.. Johnson. A.R., Mauger, K.K., Wood, N.L. and Fuller, G.C., Aortic proline hydroxylase in hypoxia induced arterio lerosis in rabbits, Life Sci., 11 (1972) 869. 13 St.Clair. R.W., Toma, J.J.%n d Lofland. H.B., Proline hydroxylase activity and collagen content of pigeon aortas with naturally-occurring and cholesterol-aggravated atherosclerosis, Atherosclerosis, 21 (1975) 165. 14 Rucker, R.B. and Goettlich-Riemann, W., Isolation and properties of soluble elastin from copper deficient chicks, J. Nutr., 102 (1972) 663. 15 Kamovski, M.J.. A formaldehyde-glutaraldehyde fixation of high osmolality for use in electron microscopy, J. Cell. Biol.. 27 (1965) 137. 16 Lowry, O.H.. Rosebrough. N.J., Farr. A.L. and Randall, R.J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 266. 17 Burton, K.. Study of condition and mechanism of diphenylamine reaction for colorimetric estimation of deoxyribonucleic acid, Biochem. J.. 62 (1966) 315. 18 Hutton, Jr., J.J., Tappel, A.L. and Udenfriend, S., A rapid assay for collagen proline hydroxylase. Anal. Biochem., 16 (1966) 384. 19 Chvapil. M. and Ehrlich. E.. Effect of increased oxygen on hydroxyproline synthesis in the cuticle and body wall of Ascaris lumbricoides, Biochim. Biophys. Acta, 208 (1970) 467. 20 Pinnell. S.R. and Martin, G.R., The cross-linking of collagen and elastin -Enzymatic conversion of IYsine in peptide linkage to o-amino-6 semisldehyde (allysine) by an extract from bone, Proc. Natl. Acad. Sci.. 61 (1968) 708. 21 Ryan. J.N. and Woessner, Jr., J.F., Mammalian collagenase -Direct demonstration in homogenates of involuting rat uterus, Biochem. Biophys. Res. Commun., 44 (1971) 144. 22 Woessner. Jr., J.F., The determination of hydroxyproline in tissue and protein samples containing small proportions of this amino acid, Arch. Biochem. Biophys., 93 (1961) 440. 23 Naito, H.K. and Lewis, L.A.. Rapid simplified method for measuring total bepatic cholesterol. Clin. Chem., 18 (1972) 911. 24 Zak. B., Methods of Clinical Chemistry, Vol. 5, Academic Press, New York, N.Y.. 1966. P. 79. 25 Constantinides, P.. Experimental Atherosclerosis. Elsevier. Amsterdam, 1965. 26 McGee, J.O’D.. Langness. U and Udenfriend, 8. Immunological evidence for an inactive precursor of collagen proline hydroxylase in cultured fibroblasta, Proc. Natl. Acad. Sci., 68 (1971) 1586. 27 Comstock, J.P., Gribble. T.J. and Udenfrlend. S.. Further study on the activation of collagen proline hydroxylase in cultures of L-929 fibroblasta. Arch. Biochem. Biophys., 137 (1970) 115.
405 28 Grlbble, T.J., Comstock, J.P. and Udenfrlend. S., Collagen chain formation and peptidyl proline hydroxylation in monolayer tissue cultures of L-929 fibroblasta, Arch. Biochem. Biophys., 129 (1969) 308. 29 Stassen, F.L., Cardinale, G.J. and Udenfriend, S.. Activation of prolyl hydroxylase in L-929 fibroblasts by ascorbic acid, Proc. NatL Acad. Sci. 70 (1973) 1090. 30 Chvapll. M.. McCarthy, D.. Madden, J.W. and Peacock, Jr., E.E., Effect of 1.10~phenanthroline and desferrioxamine in viva on prolyl hydroxylase hydroxylation of collagen in various tissues of rats, Biochem. Pharmacol.. 23 (1974) 2165. 31 Baumgartner, H.R., Thrombosis. In: Risk Factors and Diagnostic Approaches, Schattauer, Stuttgart, 1972, p. 161. 32 Werb, Z.. Burleigh, M.C., Barrett, A.J. and Starkey, P.M., The interaction of o2-macroglobulin with proteinsses -Binding and inhibition of mammalian collagenases and other metal proteinases, Biothem J.. 139 (1974) 369. 33 Chesney, C.M., Harper, E. and Coman. R.W.. Human platelet collagenase, J. Clin. Invest., 53 (1974) 1647. 34 Harris, E.D., Gonnerman, W.A., Savage, J.E. and O’Dell, B.L., Connective tissue aimine oxidase, Part 2 (Purification and partial characterization of lysyl oxidase from chick aorta). Biochim. Biophys. Acta. 341(1974) 332. 36 Rucker, R.B.. Roensch, L.F.. Savage, J.E. and O’Dell, B.L.. Oxidation of peptidy Wine by an amine oxidase from bovine aorta, Biochem. Biophys. Res. Commun., 40 (1970) 1391. 36 Chvapil, M., Sith, P., Tillema, L.. Carlson. EC., Campbell, J.B. and Eskelson, C.D., Reaction of chicken aorta to the effect of cholesterol diet and lmmunication. Unpublished observation. 37 Ronnemaa, T.. Juva. K. and Kulonen. E.. Effect of hyperlipidemic rat serum on the synthesis of collagen by chick embryo fibroblasts, Atherosclerosis, 21 (1976) 316. 38 St.Clair, R.W., tofland. H.B. and Clarkson, T.B., Influence of duration of cholesterol feeding on esterlflcation of fatty acids by cell-free preparation of pigeon aorta, Circulation Res., 27 (1970) 213. 39 Day, A.J. and Proudlock, J.W., Changes in aortic cholesterol-esterifying activity in rabbits fed cholesterol for three days, Atherosclerosis, 19 (19741 253. 40 Shimatnoto. T., Hyperreactive arterial endothelial cells in the atherogenesls and cyclic AMP phosphodiesterase inhibitor in prevention and treatment of atherosclertic disorders, Jap. Heart J., 16 (19751 76. 41 Shimamoto, T., Injury and repair in arterial tissue, Angiology, 25 (1974) 682.