- Email: [email protected]
The effect of intermittent carbon monoxide exposure on experimental atherosclerosis in the rabbit
527
Atherosclerosis, 24 (1976) 527-536 @ Elsevier Scientific Publishing Company,
Amsterdam
- Printed
in The Netherlands
THE EFFECT OF INTERMITTENT CARBON MONOXIDE EXPOSURE ON EXPERIMENTAL ATHEROSCLEROSIS IN THE RABBIT
R.F. DAVIES
*, D.L. TOPPING
** and D.M. TURNER
Tobacco Research Council Laboratories, (Received 22nd December, 1975) (Revised, received 29th, February, (Accepted 29th February, 1976)
***
Harrogate, Yorkshire (Great Britain)
1976)
---
Summary (1) Twenty-four female New Zealand White rabbits were fed commercial diet plus 2% cholesterol. Twelve of these animals were exposed to carbon monoxide for 4 hours per day, seven days per week for 10 weeks. The carbon monoxide exposure was such that the mean blood carboxy-haemoglobin was raised to approximately 20% during each exposure period. Twelve control animals breathed atmospheric air under the same conditions of confinement as the carbon monoxide-exposed group. (2) No significant differences in the plasma levels of cholesterol, triglycerides or glutamate oxalacetate transaminase were observed between the two groups during the experiment. (3) When the animals were sacrificed at the end of the experiment no significant differences were observed between the two groups in the aortic content of triglycerides, cholesterol or phospholipids. (4) The extent of coronary artery atherosclerosis was statistically significantly higher in the carbon monoxide group than in the control group. (5) Ultracentrifugal analysis of plasma lipoproteins revealed that there was significantly more cholesterol in the d < 1.006 fraction from the CO-exposed rabbits. (6) These findings, are discussed with particular reference to the claim that
* Present address: The Post Office Occupational Health Service, Royal Mail House. 29 Wellington Street, Leeds, LSl IDA. Great Britain. * * Present address: Isotope Unit, Division of Biochemistry, Royal Veterinary College. London, NW1 OTU, Great Britain. *** Address for correspondence: Department of Biochemistry and Drug Metabolism. Ha&ton Laboratories Europe Ltd., Otley Road, Harrogate, Great Britain.
528
the causal agent in tobacco ide. Key words:
Atherosclerosis
smoke associated
- Carbon monoxide
arterial
disease is carbon monox-
- Cholesterol - Rabbit
Introduction There is much epidemiological evidence to link heavy cigarette smoking with deaths from coronary heart disease [1,2] and there is growing opinion that the agent in cigarette smoke responsible for this link is carbon monoxide (CO). Experimental results in support of this causal relationship have mainly involved the effect of CO exposure on the cholesterol-fed rabbit [ 31. Astrup and colleagues [4] have described the differential effects of continuous and intermittent exposure of CO on aortic cholesterol accumulation. They reported a five times higher accumulation as a result of exposure for 4 hours a day compared to 24 hours a day. To our knowledge only Birnstingl et al [5] have claimed to confirm Astrup’s observations. Their claim, however, was based on a visual assessment of arterial lesions and did not included any plasma or arterial wall cholesterol estimations. Other species have been studied and these include the White Carneau pigeon [ 61 and the squirrel monkey [ 71. Before embarking on a large programme to study the possible relationship between CO and arterial disease in a number of experimental animal species, we decided to perform a more detailed study of the effects of CO exposure of the hypercholesterolaemic rabbit. Our experiment would be designed to repeat, as closely as practical, the basic rabbit experiment of Astrup and we chose to employ the intermittent rather than the continuous exposure to CO for two reasons: firstly, Astrup reported the most marked effect with this regime and secondly, it is more comparable to the exposure of the human cigarette smoker. In view of previous findings with the New Zealand white rabbit [8] it seemed reasonably certain that the addition of 2% cholesterol to the diet, would result in very high plasma cholesterol levels. Nevertheless we were obliged to use such a cholesterol level if our experiment were to resemble, in design, that used by Astrup and his colleagues [ 31. Materials and Methods Animals
Twenty-four female New Zealand White rabbits weighing 1500 g at the start of the experiment (February, 1974) were obtained from H & E Rabbitry, Chadderton, Lancashire, U.K. They were housed three per cage when not in the exposure chambers. Diet and experimental
procedure
Pellet diet SGI (Oxoid Ltd., London) was fed ad libitum to the animals during a 10 days period of acclimatisation. Subsequently, all animals were offered 150 g daily of SGI + 2% cholesterol, prepared by coating the pellets with cho-
529
lesterol dissolved in melted lard and 4% lard in the diet. After randomly divided into 2 groups ly the same average rise in plasma Exposure
to give a final concentration of 2% cholesterol 2 weeks cholesterol feeding the rabbits were of 12 animals, each group having approximatecholesterol.
chambers
The CO exposure chamber and identical sham exposure chamber have been described in detail by Armitage et al. [ 61. Group I animals were placed in the chamber daily for 10 weeks and exposed to a concentration of approximately 250 ppm CO from 10.00 to 14.00 h which raised the mean group carboxyhaemoglobin (COHb) level to approximately 20%. Group II animals were handled and transferred to the second chamber, to breathe atmospheric air, in a similar way. After completion of the 4 h exposure period the animals were returned to their conventional wire mesh cages. The animals were not maintained permanently in the exposure chambers since the dimensions were insufficiently large to allow free movement and adequate hygiene. Concentrations of CO in the chamber were monitored regularly throughout each day be means of Drager tubes (Drager Normalair Ltd., Kitty Brewster, Northumberland, U.K.). Serial laboratory
studies
Blood samples (2.5 ml) were obtained, after an overnight fast, from a marginal ear vein immediately before starting the cholesterol-enriched diet, after 2 weeks feeding, and at weekly intervals during the 10 week exposure period immediately after the end of the CO exposure. They were immediately placed in heparinised containers, on ice, and plasma was separated by centrifugation at 2000 rpm for 10 min. The samples obtained were analysed immediately or stored for no more than 24 h at 4°C prior to assay. Plasma cholesterol and triglycerides were simultaneously estimated by the method of Kessler and Lederer [9], using a Technicon AA11 Autoanalyser. COHb was assayed by the method of Commins and Lawther [lo] and haemoglobin by the method of Drabkin and Austin [ll]. Plasma glutamate oxalacetate transaminase (GOT) was assayed by the method of Kessler et al. [12]. All rabbits were weighed weekly immediately after venepuncture. Terminal
laboratory
studies
At the completion of the 10 week experiment all rabbits were killed by an intravenous injection of 5 ml of pentobarbitone sodium (Abbot Laboratories, Queenborough, Kent, U.K.). The aortae and hearts were removed for subsequent biochemical and microscopical analysis. Whole aortae were taken into ice-cold physiological saline carefully stripped of fat and adhering connective tissue, blotted and weighed. For lipid analysis they were puverised under liquid nitrogen and extracted using the Folch procedure [ 131. For lipoprotein analysis blood samples (5 ml) were taken from four representative non-fasted animals in each group and collected into tubes containing EDTA [14]. For each sample, the plasma was separated and layered under saline (d = 1.006) containing 1 mM EDTA and lipoproteins were separated ultracentrifugally into the density fractions d < 1.006, d = 1.006-1.019, and d > 1.019. The isolated lipoproteins were extracted with chloroform:methanol (2:1, v/v) and, after
530
partition of the solvent, assayed for cholesterol and triglycerides. Cholesterol and triglyceride levels of aortic or lipoprotein extracts were analysed as described above for plasma [9]. Phospholipids were assayed, after digestion, as inorganic phosphate by the method of Chen, Toribara and Warner v51. The heart was weighed and then fixed in form01 calcium and cut into six blocks. The subsequent method of preparation of stained frozen and paraffin wax sections and the evaluation of the extent of coronary artery atherosclerosis are described by Armitage et al. [ 61. Statistical methods
All data are expressed, where appropriate, comparisons made using Student’s t-test.
as means f SEM and statistical
Results Body weights
The mean body weights for the two groups were almost identical over the first five weeks of CO or sham exposure. After that time the CO-exposed rabbits continued to increase in weight, whereas the mean body weight of the sham-exposed rabbits fell over the next three weeks. Thereafter, the mean weights rose in this latter group. At no time, however, were the group differences statistically significant. Mean terminal body weights of 3224*97 g and 3024289 g were recorded for the CO- and sham-exposed rabbits respectively. Carboxyhaemoglobin
and haemoglobin
Both parameters were measured at intervals over the course of experiment and the results are shown in Table 1. After the first week of exposure the mean haemoglobin concentration was significantly different in the two groups of rabbits, and this was maintained throughout the rest of the experiment. Blood
TABLE 1 HAEMOGLOBIN AND COHb LEVELS IN NEW ZEALAND WHITE RABBITS DURING 10 WEEKS INTERMITTENT EXPOSURE TO CARBON MONOXIDE (GROUP I). CONTROL (GROUP II) DATA ARE ALSO SHOWN Results are expressed as mean + SEM. Time (weeks)
Haemoglobin (g/100
Carboxyhaemoglobin
ml)
(% COW)
Group I
Group II
Group I
Group II
1 2 5
12.4 k 0.1 13.6 f 0.2 a 11.8 f 0.4 b
12.1 + 0.3 12.7 + 0.2 9.6 t 0.3
14.5 * 1.6 18.2 t 1.8 18.6 f 1.9
0.1 ?r0.4 1.0 2 0.2 2.0 r 0.8
6 8 10
9.9 + 0.3 b 10.4 f 0.2 b 9.4 + 0.2 c
8.8 t 0.4 8.4 +_0.4 8.1 f 0.5
20.4 f 1.8 18.1 * 2.0 19.8 + 1.8
0.6 f 0.3 1.1 f 0.6 0.3 f 0.1
a P < 0.05. b P < 0.006. = P < 0.025.
Before diet Beforeexposw .e 1 2 3 4 5 6 7 8 9 10
(weeks)
Time
57.5 f 1486.8 f 1548.3 + 1936.7 + 2163.5 f 2335.7 f 2786.5+ 2749.4 f 2976.5 f 2895.3 + 3026.4 f 3002.7 f
Group1
5.4 144.4 150.2 128.6 128.0 98.0 135.9 162.0 179.9 196.7 205.3 293.9
Group I 62.7 f 53.5 f 61.3 f 70.3 f 103.2 f 85.1 + 95.9 + 133.7 + 215.5 + 194.3 f 219.6 f 253.2 +
55.5 + 4.2 1520.7 f 121.0 1840.7 + 123.3 1902.8 + 74.0 1907.6 It138.3 2087.7 f 125.1 2469.8 + 184.2 2468.4 + 243.2 2636.9 f 315.8 2628.3 f 285.5 2819.5 f 251.0 2786.3 f 324.7
3.9 4.6 4.6 7.2 8.8 7.2 10.7 14.1 27.0 31.3 40.2 50.9
27.5 f 3.0 20.0 + 1.6 27.5 f 3.3 24.4 f 9.4 48.4 + 8.0 53.0 f 9.7 70.5? 23.2 73.3 f 12.7 60.6 f 6.3 67.5 f 7.8 66.6 f 6.2 80.0 f 12.6
63.6 f 64.7 f 62.3 f 62.2 t 86.9 f 89.7 f 76.3 f 140.1 + 231.1 f 175.5 f 214.1 f 249.6 f
2.5 14.0 9.1 9.4 15.1 14.7 14.0 14.1 70.8 41.3 74.3 61.1
Group1
GOT(mU/ml)
Group II
Triglycerides (mg/lOO ml)
two weeks elapsedbetweenthestartofcholesterolfeedingandinitialexposureto CO.
38.4 f 7.8 19.6 + 2.0 39.6 f 5.5 34.4 f 4.7 52.4 t 7.7 63.4+ 9.2 50.2 + 6.4 79.5k12.2 91.1 f 21.5 84.0 + 10.9 98.7 f 22.9 112.9 f 29.9
Group II
INCONTROLRABBITS(GROUPII)ANDRABBITSINTERMITTENTLYEXPOSEDTO
Group II
Cholesterol(mg/lOOmI)
Resultsareexpressedasmean f SEM. A period of
CO(GROUP1)
PLASMALEVELSOFCHOLESTEROL,TRIGLYCERIDESANDGOT
TABLE2
532
from the sham-exposed rabbits always contained low levels of COHb. The levels of COHb attained by the CO breathing animals varied considerably individual variation tended to be as great as the group variation. Plasma lipids Table 2 shows the changes in plasma cholesterol and triglycerides over the period of experiment. A marked rise in plasma cholesterol occurred over the two week period during which the rabbits were given the cholesterol-enriched diet but before CO exposure started. Over the 10 weeks of CO or sham exposure plasma levels continued to rise and in general, the mean level of the CO breathing animals was higher than that of the sham-exposed animals. At no time, however, was this difference statistically significant. Plasma triglycerides showed a gradual rise in both groups over the course of the CO exposure period, and at the end of the experiment the levels were approximately four times greater than normal. There was no statistically significant difference, however, between CO- and sham-exposed animals at any period during the experiment. Plasma GOT In both experimental groups the GOT levels rose during the exposure period, as shown in Table 2, concomitant with the rise in the plasma cholesterol. There was no significant difference between the two groups at any time. Plasma lipoproteins Plasma lipoproteins from four non-fasted animals of each group were separated ultracentrifugally into the density fraction d < 1.006 (VLDL), d = 1.006-1.019 (intermediate density lipoprotein, IDL) and d > 1.019 (LDL + HDL). Table 3 shows the results obtained from rabbits of both groups. The major proportion of the lipoprotein cholesterol was found to be in the d < 1.006 density class with rather less in the IDL fraction and only some 10% of the total in the d > 1.019 fraction. Although cholesterol concentrations were raised in both VLDL and IDL fractions in CO-exposed rabbits, only in the former lipoprotein class was the effect significant statistically. Aortic lipids Table 4 shows the results of arterial lipid analysis in rabbits from both groups. In none of the lipid classes was the effect of statistically significant. TABLE
3
LIPOPROTEIN MITTENTLY Results
LIPIDS EXPOSED
are expressed
Lipoprotein
IN
PLASMA
TO
as means
CO
Cholesterol
class
IDL LDL
a
(d < 1.006)
2123
(d = 1.006-1.019) + HDL
P < 0.05.
(d >
1.019)
CONTROL I) FOR
RABBITS
(GROUP
II) AND
RABBITS
10 WEEKS
r SEM.
Group VLDL
FROM
(GROUP
(mg/lOO
Triglycerides
ml) Group
I f 257
a
Group
II
1563C
86
835+
14
683
f 107
386+
62
336
f
97
195
I
(mg/lOO
ml)
Group
f 333
113+
6
55?
15
II
154
+ 18
123
k 16
63*
4
INTER-
533 TABLE 4 AORTIC
LIPID CONTENT
OF RABBITS
IN BOTH TREATMENT
GROUPS
Results are expressed as means ?: SEM per g wet weight of tissue. Figures in parenthesis indicate the number of animals.
Triglycerides mg/g Cholesterol mglg Phospholipid fig/g
Group I (CO)
Group II (air)
(11)
(10)
69.2 + 11.6 20.4 k 2.3 302.4 f 29.2
62.4 f 12.9 19.6 f 2.3 344.2 f 53.1
Histopathology Three rabbits, 1 CO- and 2 sham-exposed, died during the experiment. One (CO) died of a cerebellar abscess secondary to acute purulent otitis media associated with ear mites. Another fractured a lumbar vertebra during venepuncture and was killed immediately. The third rabbit died of a lung abscess. Data obtained from the heart and aorta of these animals are not included in the results. Hearts weights were not significantly different between the two groups of rabbits, mean values being 8.1 + 0.3 g and 7.7 + 0.3 g for the CO- and shamexposed groups, respectively. Coronary artery atherosclerosis data are shown in Table 5. It may be seen that there was a statistically highly significant increase in the percentage of coronary arteries affected in the CO-exposed group in comparison to the shamexposed group. The percentage of stenosis of affected arteries was, on the other hand, rather higher in the sham-exposed group. This difference was not, however, quite significant at the 95% confidence level (P = 0.075). There was no statistically significant evidence of any correlation between either of these indicators of coronary artery atherosclerosis and plasma cholesterol levels at any time in the CO-exposed group, nor in the sham-exposed group. Discussion [ 1,2] and Correlations have been made between smoking and atherosclerosis it has been stated that the principal agent in tobacco smoke responsible for this correlation is carbon monoxide [16]. This is based largely on studies where rabbits were given relatively high levels of dietary cholesterol and exposed to TABLE 5 THE EXTENT OF CORONARY ARTERY ATHEROSCLEROSIS IN HYPERCHOLESTEROLAEMIC RABBITS INTERMITTENTLY EXPOSED TO CO (GROUP I) OR SHAM-EXPOSED (GROUP II) Results are expressed as means * SEM. Figures in parenthesis indicate the number of animals.
B Coronary arteries affected % Stenosis
Group I
Group II
(11)
(10)
Significance Group I vs Group II
50.6 f 2.:
35.5 f 2.2
P <
63.0 f 2.2
68.6 ? 1.9
n.s.
0.001
534
CO [4,17]. The present studies, where both coronary artery atherosclerosis and the accumulation of cholesterol and lipids in the aorta were measured, conflict with these earlier findings. The experimental design used to test the hypothesis was that which has been reported [4] to produce the greatest accumulation of cholesterol in the aortic wall. The main difference in the experimental procedure has been the use of a different strain of rabbit. Astrup and colleagues used Danish albino rabbits and in the present experiments we used New Zealand White. We have been unable to confirm that such an intermittent exposure regime results in any significant accumulation of cholesterol in the aortae of the rabbits, but we have observed an increase in coronary artery atherosclerosis in the CO-exposed animals as measured by the extent of internal lesions. This latter observation would agree with other workers’ findings in a primate species [ 181. We have shown, using the White Carneau pigeon, that the enhancing effect of CO exposure on coronary atherosclerosis is not maintained [6 ] and, therefore, the effects recorded in our much shorter rabbit experiment may represent an acute response to CO exposure. If strain difference, however, accounts for the difference between our results and the published data, then extrapolation of such results to man must only be made with caution. The finding that there was no statistically significant correlation within animals in our experiments between coronary artery atherosclerosis and plasma cholesterol levels is perhaps not surprising. All the rabbits in the experiment were fed the same amount of cholesterol and the relative variation that was observed between animals in plasma cholesterol levels was very much smaller than than observed between cholesterol-fed and non-cholesterol-fed rabbits. Thus it would have needed a very strong relationship to exist for such a correlation to have been identified in groups of 12 rabbits only. Both groups showed a fall in haemoglobin which was less marked in the CO group, probably representing an attempt to compensate for the reduced ,oxygen carrying capacity of the blood in these animals. It is unlikely that the anaemia was related to weekly withdrawal of 2.5 ml portions of blood, but more likely to an accumulation of cholesterol in the bone marrow resulting in depression of erythropoiesis. It is known that cholesterol is deposited in all major organs during cholesterol feeding to the rabbit [8]. We have demonstrated in this experiment that, in postabsorptive rabbits, most of the plasma cholesterol is to be found in the density i.e. in the regions where “remnant” fractions d < 1.006 and d = 1.006-1.019, particles [19] may be expected on ultracentrifugal isolation. Moreover, the concentrations of cholesterol in these fractions were raised by exposure to CO. The relative increase in the VLDL and IDL cholesterol levels in CO-exposed rabbits may have been due to a reduction in the ability of the liver to metabolize cholesterol. Hepatic oxygen consumption is reduced in the presence of carboxyhaemoglobin [20] and, if the metabolism of lipoprotein cholesterol depends on oxygen supply, it may have been reduced in the CO-exposed rabbits. Alternatively, if VLDL triglyceride incorporation is curtailed in the presence of carboxyhaemoglobin, the triglyceride content of VLDL and chylomicrons released by the gut in the CO group may have been relatively reduced. It must, however, be borne in mind that the effects of CO alone on, for example, lipid metabolism cannot necessarily be assumed to be similar to the effects of tobac-
535
co smoke. We have shown in the squirrel monkey [21] that, even though the degree of exposure may be the same, the effects on lipid metabolism of CO alone and in tobacco smoke are quite different. The relationship of moderate chronic CO exposure to the development of arterial disease still remains open. We have been unable to confirm Astrup’s [3,4] observations of an enhancement of aortic lipid accumulation, although we have shown an effect on coronary atherosclerosis. The mean plasma cholesterol levels in our experiment are higher than those found by Astrup et al. and it may be that a possible CO-enhanced aortic lipid accumulation could have been masked. If such a dose-related effect occurs, then extrapolation of these and similar data to the human tobacco smoker must, again, be applied with much caution, and the role of substances other than carbon monoxide should continue to be investigated. Acknowledgements The authors skilled technical
wish to thank assistance.
Mrs. J. Bedwell
and Miss J. O’Brien
for their
References 1 Smoking and Health Now - A new report and summary on smoking and its effects on health, from the Royal College of Physicians of London. Pitmsn Medical and Scientific Publishing Co. Ltd., London, 1971. 2 The Health Consequences of Smoking, U.S. Department of Health, Education and Welfare, 1973. 3 Astrup. P.. Kjeldsen. K. and Wan&up. J., Enhancing influence of carbon monoxide on the development of atheromatosis in cholesterol fed rabbits, J. Atheroscler. Res.. 7 (1967) 343. 4 Astmp, P.. Some physiological and pathological effects of moderate carbon monoxide exposure, Brit. Med. J.. 2 (1972) 447. 5 Bimstingl. M., Hawkins, L. and McEwen. T., Experimental atherosclerosis during chronic exposure to carbon monoxide, Europ. J. Surg. Res., 2 (1970) 92. 6 Armitage, A.K., Davies, R.F. and Turner, D.M., The effects of carbon monoxide on the development of atherosclerosis in the White Cameau pigeon, Atherosclerosis, 23 (1976) 333. 7 Webster, W.S., Clarkson. T.B. and Lofland. H.B.. Carbon monoxide aggravated atherosclerosis in the squirrel monkey, Exp. Mol. Path., 13 (1968) 36. 8 Prior, J.T., Kurtz, D.M. and Ziegler, D.D.. The hypercholesterolaemic rabbit, Arch. Path., 71 (1961) 672. 9 Kessler, G. and Lederer. L., in: L.T. Skeggs (Ed.), Automation in Analytical Chemistry, Mediad Inc., 10 11 12 13 14 15 16 17 18
New York, N.Y., 1965, P. 341. Commins. B.T. and Lawther, P.J., A sensitive method for the determination of carboxyhaemoglobin in a finger prick sample of blood, Brit. J. Ind. Med., 22 (1965) 139. Drabkin, D.L. and Austin, J.H.. Spectrophotometric studies. Part 1 (Spectrophotometric constants for common haemoglobin derivatives in human, dog and rabbit blood), J. Biol. Chem.. 98 (1932) 719. Kessler, G., Rush, R.L., Leon, L., Decca, A. and Cupiola. R., Automated 340 nm measurement of SGOT. SGPT and LDH, Clin. Chem., 16 (1970) 530. Folch. J., Lees, M.. and Sloan-Stanley, G.H.. A simple method for the isolation and purification of total lipids from animal tissues, J. Biol. Chem., 22 (1957) 497. Havel, R.J., Eder, H.A. and Bragdon. J.P.. Distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum, J. Clin. Invest., 34 (1955) 1345. Chen, P.S., Toribara, T.Y. and Warner, H.. Microdetermination of phosphorus, Anal. Chem., 28 (1956) 1756. Kjeldsen, K., Astrup. P. and Wan&up, J., Ultrastructural internal changes in the rabbit aorta after moderate carbon monoxide exposure, Atherosclerosis, 16 (1972) 67. Astrup, P.. Kieldsen, K. and Wanstrup. J., Effects of carbon monoxide exposure on the arterial wslls, Ann. N.Y. Acad. Sci., 174 (1970) 294. Thomsen, H.K., Carbon monoxide-induced atherosclerosis in primates, Atherosclerosis, 20 (1974) 233.
536
19 20 21
Zilversmit. D.B., A proposal finking atherogenesis to the interaction of endothelial lipoprotein lipase with triglyceride rich lipoproteins, Circ. Res., 33 (1973) 633. Topping, D.L., The acute effects of carbon monoxide on the metabolism of perfused rat liver, Biothem. J., 152 (1975) 425. Turner, D.M. and Topping. D.L.. The effects of tobacco smoke and some of its constituents on triglyceride secretion in the squirrel monkey, Res. Comm. Chem. Path. Pharmac., 12 (1975) 85.
Atherosclerosis, 24 (1976) 527-536 @ Elsevier Scientific Publishing Company,
Amsterdam
- Printed
in The Netherlands
THE EFFECT OF INTERMITTENT CARBON MONOXIDE EXPOSURE ON EXPERIMENTAL ATHEROSCLEROSIS IN THE RABBIT
R.F. DAVIES
*, D.L. TOPPING
** and D.M. TURNER
Tobacco Research Council Laboratories, (Received 22nd December, 1975) (Revised, received 29th, February, (Accepted 29th February, 1976)
***
Harrogate, Yorkshire (Great Britain)
1976)
---
Summary (1) Twenty-four female New Zealand White rabbits were fed commercial diet plus 2% cholesterol. Twelve of these animals were exposed to carbon monoxide for 4 hours per day, seven days per week for 10 weeks. The carbon monoxide exposure was such that the mean blood carboxy-haemoglobin was raised to approximately 20% during each exposure period. Twelve control animals breathed atmospheric air under the same conditions of confinement as the carbon monoxide-exposed group. (2) No significant differences in the plasma levels of cholesterol, triglycerides or glutamate oxalacetate transaminase were observed between the two groups during the experiment. (3) When the animals were sacrificed at the end of the experiment no significant differences were observed between the two groups in the aortic content of triglycerides, cholesterol or phospholipids. (4) The extent of coronary artery atherosclerosis was statistically significantly higher in the carbon monoxide group than in the control group. (5) Ultracentrifugal analysis of plasma lipoproteins revealed that there was significantly more cholesterol in the d < 1.006 fraction from the CO-exposed rabbits. (6) These findings, are discussed with particular reference to the claim that
* Present address: The Post Office Occupational Health Service, Royal Mail House. 29 Wellington Street, Leeds, LSl IDA. Great Britain. * * Present address: Isotope Unit, Division of Biochemistry, Royal Veterinary College. London, NW1 OTU, Great Britain. *** Address for correspondence: Department of Biochemistry and Drug Metabolism. Ha&ton Laboratories Europe Ltd., Otley Road, Harrogate, Great Britain.
528
the causal agent in tobacco ide. Key words:
Atherosclerosis
smoke associated
- Carbon monoxide
arterial
disease is carbon monox-
- Cholesterol - Rabbit
Introduction There is much epidemiological evidence to link heavy cigarette smoking with deaths from coronary heart disease [1,2] and there is growing opinion that the agent in cigarette smoke responsible for this link is carbon monoxide (CO). Experimental results in support of this causal relationship have mainly involved the effect of CO exposure on the cholesterol-fed rabbit [ 31. Astrup and colleagues [4] have described the differential effects of continuous and intermittent exposure of CO on aortic cholesterol accumulation. They reported a five times higher accumulation as a result of exposure for 4 hours a day compared to 24 hours a day. To our knowledge only Birnstingl et al [5] have claimed to confirm Astrup’s observations. Their claim, however, was based on a visual assessment of arterial lesions and did not included any plasma or arterial wall cholesterol estimations. Other species have been studied and these include the White Carneau pigeon [ 61 and the squirrel monkey [ 71. Before embarking on a large programme to study the possible relationship between CO and arterial disease in a number of experimental animal species, we decided to perform a more detailed study of the effects of CO exposure of the hypercholesterolaemic rabbit. Our experiment would be designed to repeat, as closely as practical, the basic rabbit experiment of Astrup and we chose to employ the intermittent rather than the continuous exposure to CO for two reasons: firstly, Astrup reported the most marked effect with this regime and secondly, it is more comparable to the exposure of the human cigarette smoker. In view of previous findings with the New Zealand white rabbit [8] it seemed reasonably certain that the addition of 2% cholesterol to the diet, would result in very high plasma cholesterol levels. Nevertheless we were obliged to use such a cholesterol level if our experiment were to resemble, in design, that used by Astrup and his colleagues [ 31. Materials and Methods Animals
Twenty-four female New Zealand White rabbits weighing 1500 g at the start of the experiment (February, 1974) were obtained from H & E Rabbitry, Chadderton, Lancashire, U.K. They were housed three per cage when not in the exposure chambers. Diet and experimental
procedure
Pellet diet SGI (Oxoid Ltd., London) was fed ad libitum to the animals during a 10 days period of acclimatisation. Subsequently, all animals were offered 150 g daily of SGI + 2% cholesterol, prepared by coating the pellets with cho-
529
lesterol dissolved in melted lard and 4% lard in the diet. After randomly divided into 2 groups ly the same average rise in plasma Exposure
to give a final concentration of 2% cholesterol 2 weeks cholesterol feeding the rabbits were of 12 animals, each group having approximatecholesterol.
chambers
The CO exposure chamber and identical sham exposure chamber have been described in detail by Armitage et al. [ 61. Group I animals were placed in the chamber daily for 10 weeks and exposed to a concentration of approximately 250 ppm CO from 10.00 to 14.00 h which raised the mean group carboxyhaemoglobin (COHb) level to approximately 20%. Group II animals were handled and transferred to the second chamber, to breathe atmospheric air, in a similar way. After completion of the 4 h exposure period the animals were returned to their conventional wire mesh cages. The animals were not maintained permanently in the exposure chambers since the dimensions were insufficiently large to allow free movement and adequate hygiene. Concentrations of CO in the chamber were monitored regularly throughout each day be means of Drager tubes (Drager Normalair Ltd., Kitty Brewster, Northumberland, U.K.). Serial laboratory
studies
Blood samples (2.5 ml) were obtained, after an overnight fast, from a marginal ear vein immediately before starting the cholesterol-enriched diet, after 2 weeks feeding, and at weekly intervals during the 10 week exposure period immediately after the end of the CO exposure. They were immediately placed in heparinised containers, on ice, and plasma was separated by centrifugation at 2000 rpm for 10 min. The samples obtained were analysed immediately or stored for no more than 24 h at 4°C prior to assay. Plasma cholesterol and triglycerides were simultaneously estimated by the method of Kessler and Lederer [9], using a Technicon AA11 Autoanalyser. COHb was assayed by the method of Commins and Lawther [lo] and haemoglobin by the method of Drabkin and Austin [ll]. Plasma glutamate oxalacetate transaminase (GOT) was assayed by the method of Kessler et al. [12]. All rabbits were weighed weekly immediately after venepuncture. Terminal
laboratory
studies
At the completion of the 10 week experiment all rabbits were killed by an intravenous injection of 5 ml of pentobarbitone sodium (Abbot Laboratories, Queenborough, Kent, U.K.). The aortae and hearts were removed for subsequent biochemical and microscopical analysis. Whole aortae were taken into ice-cold physiological saline carefully stripped of fat and adhering connective tissue, blotted and weighed. For lipid analysis they were puverised under liquid nitrogen and extracted using the Folch procedure [ 131. For lipoprotein analysis blood samples (5 ml) were taken from four representative non-fasted animals in each group and collected into tubes containing EDTA [14]. For each sample, the plasma was separated and layered under saline (d = 1.006) containing 1 mM EDTA and lipoproteins were separated ultracentrifugally into the density fractions d < 1.006, d = 1.006-1.019, and d > 1.019. The isolated lipoproteins were extracted with chloroform:methanol (2:1, v/v) and, after
530
partition of the solvent, assayed for cholesterol and triglycerides. Cholesterol and triglyceride levels of aortic or lipoprotein extracts were analysed as described above for plasma [9]. Phospholipids were assayed, after digestion, as inorganic phosphate by the method of Chen, Toribara and Warner v51. The heart was weighed and then fixed in form01 calcium and cut into six blocks. The subsequent method of preparation of stained frozen and paraffin wax sections and the evaluation of the extent of coronary artery atherosclerosis are described by Armitage et al. [ 61. Statistical methods
All data are expressed, where appropriate, comparisons made using Student’s t-test.
as means f SEM and statistical
Results Body weights
The mean body weights for the two groups were almost identical over the first five weeks of CO or sham exposure. After that time the CO-exposed rabbits continued to increase in weight, whereas the mean body weight of the sham-exposed rabbits fell over the next three weeks. Thereafter, the mean weights rose in this latter group. At no time, however, were the group differences statistically significant. Mean terminal body weights of 3224*97 g and 3024289 g were recorded for the CO- and sham-exposed rabbits respectively. Carboxyhaemoglobin
and haemoglobin
Both parameters were measured at intervals over the course of experiment and the results are shown in Table 1. After the first week of exposure the mean haemoglobin concentration was significantly different in the two groups of rabbits, and this was maintained throughout the rest of the experiment. Blood
TABLE 1 HAEMOGLOBIN AND COHb LEVELS IN NEW ZEALAND WHITE RABBITS DURING 10 WEEKS INTERMITTENT EXPOSURE TO CARBON MONOXIDE (GROUP I). CONTROL (GROUP II) DATA ARE ALSO SHOWN Results are expressed as mean + SEM. Time (weeks)
Haemoglobin (g/100
Carboxyhaemoglobin
ml)
(% COW)
Group I
Group II
Group I
Group II
1 2 5
12.4 k 0.1 13.6 f 0.2 a 11.8 f 0.4 b
12.1 + 0.3 12.7 + 0.2 9.6 t 0.3
14.5 * 1.6 18.2 t 1.8 18.6 f 1.9
0.1 ?r0.4 1.0 2 0.2 2.0 r 0.8
6 8 10
9.9 + 0.3 b 10.4 f 0.2 b 9.4 + 0.2 c
8.8 t 0.4 8.4 +_0.4 8.1 f 0.5
20.4 f 1.8 18.1 * 2.0 19.8 + 1.8
0.6 f 0.3 1.1 f 0.6 0.3 f 0.1
a P < 0.05. b P < 0.006. = P < 0.025.
Before diet Beforeexposw .e 1 2 3 4 5 6 7 8 9 10
(weeks)
Time
57.5 f 1486.8 f 1548.3 + 1936.7 + 2163.5 f 2335.7 f 2786.5+ 2749.4 f 2976.5 f 2895.3 + 3026.4 f 3002.7 f
Group1
5.4 144.4 150.2 128.6 128.0 98.0 135.9 162.0 179.9 196.7 205.3 293.9
Group I 62.7 f 53.5 f 61.3 f 70.3 f 103.2 f 85.1 + 95.9 + 133.7 + 215.5 + 194.3 f 219.6 f 253.2 +
55.5 + 4.2 1520.7 f 121.0 1840.7 + 123.3 1902.8 + 74.0 1907.6 It138.3 2087.7 f 125.1 2469.8 + 184.2 2468.4 + 243.2 2636.9 f 315.8 2628.3 f 285.5 2819.5 f 251.0 2786.3 f 324.7
3.9 4.6 4.6 7.2 8.8 7.2 10.7 14.1 27.0 31.3 40.2 50.9
27.5 f 3.0 20.0 + 1.6 27.5 f 3.3 24.4 f 9.4 48.4 + 8.0 53.0 f 9.7 70.5? 23.2 73.3 f 12.7 60.6 f 6.3 67.5 f 7.8 66.6 f 6.2 80.0 f 12.6
63.6 f 64.7 f 62.3 f 62.2 t 86.9 f 89.7 f 76.3 f 140.1 + 231.1 f 175.5 f 214.1 f 249.6 f
2.5 14.0 9.1 9.4 15.1 14.7 14.0 14.1 70.8 41.3 74.3 61.1
Group1
GOT(mU/ml)
Group II
Triglycerides (mg/lOO ml)
two weeks elapsedbetweenthestartofcholesterolfeedingandinitialexposureto CO.
38.4 f 7.8 19.6 + 2.0 39.6 f 5.5 34.4 f 4.7 52.4 t 7.7 63.4+ 9.2 50.2 + 6.4 79.5k12.2 91.1 f 21.5 84.0 + 10.9 98.7 f 22.9 112.9 f 29.9
Group II
INCONTROLRABBITS(GROUPII)ANDRABBITSINTERMITTENTLYEXPOSEDTO
Group II
Cholesterol(mg/lOOmI)
Resultsareexpressedasmean f SEM. A period of
CO(GROUP1)
PLASMALEVELSOFCHOLESTEROL,TRIGLYCERIDESANDGOT
TABLE2
532
from the sham-exposed rabbits always contained low levels of COHb. The levels of COHb attained by the CO breathing animals varied considerably individual variation tended to be as great as the group variation. Plasma lipids Table 2 shows the changes in plasma cholesterol and triglycerides over the period of experiment. A marked rise in plasma cholesterol occurred over the two week period during which the rabbits were given the cholesterol-enriched diet but before CO exposure started. Over the 10 weeks of CO or sham exposure plasma levels continued to rise and in general, the mean level of the CO breathing animals was higher than that of the sham-exposed animals. At no time, however, was this difference statistically significant. Plasma triglycerides showed a gradual rise in both groups over the course of the CO exposure period, and at the end of the experiment the levels were approximately four times greater than normal. There was no statistically significant difference, however, between CO- and sham-exposed animals at any period during the experiment. Plasma GOT In both experimental groups the GOT levels rose during the exposure period, as shown in Table 2, concomitant with the rise in the plasma cholesterol. There was no significant difference between the two groups at any time. Plasma lipoproteins Plasma lipoproteins from four non-fasted animals of each group were separated ultracentrifugally into the density fraction d < 1.006 (VLDL), d = 1.006-1.019 (intermediate density lipoprotein, IDL) and d > 1.019 (LDL + HDL). Table 3 shows the results obtained from rabbits of both groups. The major proportion of the lipoprotein cholesterol was found to be in the d < 1.006 density class with rather less in the IDL fraction and only some 10% of the total in the d > 1.019 fraction. Although cholesterol concentrations were raised in both VLDL and IDL fractions in CO-exposed rabbits, only in the former lipoprotein class was the effect significant statistically. Aortic lipids Table 4 shows the results of arterial lipid analysis in rabbits from both groups. In none of the lipid classes was the effect of statistically significant. TABLE
3
LIPOPROTEIN MITTENTLY Results
LIPIDS EXPOSED
are expressed
Lipoprotein
IN
PLASMA
TO
as means
CO
Cholesterol
class
IDL LDL
a
(d < 1.006)
2123
(d = 1.006-1.019) + HDL
P < 0.05.
(d >
1.019)
CONTROL I) FOR
RABBITS
(GROUP
II) AND
RABBITS
10 WEEKS
r SEM.
Group VLDL
FROM
(GROUP
(mg/lOO
Triglycerides
ml) Group
I f 257
a
Group
II
1563C
86
835+
14
683
f 107
386+
62
336
f
97
195
I
(mg/lOO
ml)
Group
f 333
113+
6
55?
15
II
154
+ 18
123
k 16
63*
4
INTER-
533 TABLE 4 AORTIC
LIPID CONTENT
OF RABBITS
IN BOTH TREATMENT
GROUPS
Results are expressed as means ?: SEM per g wet weight of tissue. Figures in parenthesis indicate the number of animals.
Triglycerides mg/g Cholesterol mglg Phospholipid fig/g
Group I (CO)
Group II (air)
(11)
(10)
69.2 + 11.6 20.4 k 2.3 302.4 f 29.2
62.4 f 12.9 19.6 f 2.3 344.2 f 53.1
Histopathology Three rabbits, 1 CO- and 2 sham-exposed, died during the experiment. One (CO) died of a cerebellar abscess secondary to acute purulent otitis media associated with ear mites. Another fractured a lumbar vertebra during venepuncture and was killed immediately. The third rabbit died of a lung abscess. Data obtained from the heart and aorta of these animals are not included in the results. Hearts weights were not significantly different between the two groups of rabbits, mean values being 8.1 + 0.3 g and 7.7 + 0.3 g for the CO- and shamexposed groups, respectively. Coronary artery atherosclerosis data are shown in Table 5. It may be seen that there was a statistically highly significant increase in the percentage of coronary arteries affected in the CO-exposed group in comparison to the shamexposed group. The percentage of stenosis of affected arteries was, on the other hand, rather higher in the sham-exposed group. This difference was not, however, quite significant at the 95% confidence level (P = 0.075). There was no statistically significant evidence of any correlation between either of these indicators of coronary artery atherosclerosis and plasma cholesterol levels at any time in the CO-exposed group, nor in the sham-exposed group. Discussion [ 1,2] and Correlations have been made between smoking and atherosclerosis it has been stated that the principal agent in tobacco smoke responsible for this correlation is carbon monoxide [16]. This is based largely on studies where rabbits were given relatively high levels of dietary cholesterol and exposed to TABLE 5 THE EXTENT OF CORONARY ARTERY ATHEROSCLEROSIS IN HYPERCHOLESTEROLAEMIC RABBITS INTERMITTENTLY EXPOSED TO CO (GROUP I) OR SHAM-EXPOSED (GROUP II) Results are expressed as means * SEM. Figures in parenthesis indicate the number of animals.
B Coronary arteries affected % Stenosis
Group I
Group II
(11)
(10)
Significance Group I vs Group II
50.6 f 2.:
35.5 f 2.2
P <
63.0 f 2.2
68.6 ? 1.9
n.s.
0.001
534
CO [4,17]. The present studies, where both coronary artery atherosclerosis and the accumulation of cholesterol and lipids in the aorta were measured, conflict with these earlier findings. The experimental design used to test the hypothesis was that which has been reported [4] to produce the greatest accumulation of cholesterol in the aortic wall. The main difference in the experimental procedure has been the use of a different strain of rabbit. Astrup and colleagues used Danish albino rabbits and in the present experiments we used New Zealand White. We have been unable to confirm that such an intermittent exposure regime results in any significant accumulation of cholesterol in the aortae of the rabbits, but we have observed an increase in coronary artery atherosclerosis in the CO-exposed animals as measured by the extent of internal lesions. This latter observation would agree with other workers’ findings in a primate species [ 181. We have shown, using the White Carneau pigeon, that the enhancing effect of CO exposure on coronary atherosclerosis is not maintained [6 ] and, therefore, the effects recorded in our much shorter rabbit experiment may represent an acute response to CO exposure. If strain difference, however, accounts for the difference between our results and the published data, then extrapolation of such results to man must only be made with caution. The finding that there was no statistically significant correlation within animals in our experiments between coronary artery atherosclerosis and plasma cholesterol levels is perhaps not surprising. All the rabbits in the experiment were fed the same amount of cholesterol and the relative variation that was observed between animals in plasma cholesterol levels was very much smaller than than observed between cholesterol-fed and non-cholesterol-fed rabbits. Thus it would have needed a very strong relationship to exist for such a correlation to have been identified in groups of 12 rabbits only. Both groups showed a fall in haemoglobin which was less marked in the CO group, probably representing an attempt to compensate for the reduced ,oxygen carrying capacity of the blood in these animals. It is unlikely that the anaemia was related to weekly withdrawal of 2.5 ml portions of blood, but more likely to an accumulation of cholesterol in the bone marrow resulting in depression of erythropoiesis. It is known that cholesterol is deposited in all major organs during cholesterol feeding to the rabbit [8]. We have demonstrated in this experiment that, in postabsorptive rabbits, most of the plasma cholesterol is to be found in the density i.e. in the regions where “remnant” fractions d < 1.006 and d = 1.006-1.019, particles [19] may be expected on ultracentrifugal isolation. Moreover, the concentrations of cholesterol in these fractions were raised by exposure to CO. The relative increase in the VLDL and IDL cholesterol levels in CO-exposed rabbits may have been due to a reduction in the ability of the liver to metabolize cholesterol. Hepatic oxygen consumption is reduced in the presence of carboxyhaemoglobin [20] and, if the metabolism of lipoprotein cholesterol depends on oxygen supply, it may have been reduced in the CO-exposed rabbits. Alternatively, if VLDL triglyceride incorporation is curtailed in the presence of carboxyhaemoglobin, the triglyceride content of VLDL and chylomicrons released by the gut in the CO group may have been relatively reduced. It must, however, be borne in mind that the effects of CO alone on, for example, lipid metabolism cannot necessarily be assumed to be similar to the effects of tobac-
535
co smoke. We have shown in the squirrel monkey [21] that, even though the degree of exposure may be the same, the effects on lipid metabolism of CO alone and in tobacco smoke are quite different. The relationship of moderate chronic CO exposure to the development of arterial disease still remains open. We have been unable to confirm Astrup’s [3,4] observations of an enhancement of aortic lipid accumulation, although we have shown an effect on coronary atherosclerosis. The mean plasma cholesterol levels in our experiment are higher than those found by Astrup et al. and it may be that a possible CO-enhanced aortic lipid accumulation could have been masked. If such a dose-related effect occurs, then extrapolation of these and similar data to the human tobacco smoker must, again, be applied with much caution, and the role of substances other than carbon monoxide should continue to be investigated. Acknowledgements The authors skilled technical
wish to thank assistance.
Mrs. J. Bedwell
and Miss J. O’Brien
for their
References 1 Smoking and Health Now - A new report and summary on smoking and its effects on health, from the Royal College of Physicians of London. Pitmsn Medical and Scientific Publishing Co. Ltd., London, 1971. 2 The Health Consequences of Smoking, U.S. Department of Health, Education and Welfare, 1973. 3 Astrup. P.. Kjeldsen. K. and Wan&up. J., Enhancing influence of carbon monoxide on the development of atheromatosis in cholesterol fed rabbits, J. Atheroscler. Res.. 7 (1967) 343. 4 Astmp, P.. Some physiological and pathological effects of moderate carbon monoxide exposure, Brit. Med. J.. 2 (1972) 447. 5 Bimstingl. M., Hawkins, L. and McEwen. T., Experimental atherosclerosis during chronic exposure to carbon monoxide, Europ. J. Surg. Res., 2 (1970) 92. 6 Armitage, A.K., Davies, R.F. and Turner, D.M., The effects of carbon monoxide on the development of atherosclerosis in the White Cameau pigeon, Atherosclerosis, 23 (1976) 333. 7 Webster, W.S., Clarkson. T.B. and Lofland. H.B.. Carbon monoxide aggravated atherosclerosis in the squirrel monkey, Exp. Mol. Path., 13 (1968) 36. 8 Prior, J.T., Kurtz, D.M. and Ziegler, D.D.. The hypercholesterolaemic rabbit, Arch. Path., 71 (1961) 672. 9 Kessler, G. and Lederer. L., in: L.T. Skeggs (Ed.), Automation in Analytical Chemistry, Mediad Inc., 10 11 12 13 14 15 16 17 18
New York, N.Y., 1965, P. 341. Commins. B.T. and Lawther, P.J., A sensitive method for the determination of carboxyhaemoglobin in a finger prick sample of blood, Brit. J. Ind. Med., 22 (1965) 139. Drabkin, D.L. and Austin, J.H.. Spectrophotometric studies. Part 1 (Spectrophotometric constants for common haemoglobin derivatives in human, dog and rabbit blood), J. Biol. Chem.. 98 (1932) 719. Kessler, G., Rush, R.L., Leon, L., Decca, A. and Cupiola. R., Automated 340 nm measurement of SGOT. SGPT and LDH, Clin. Chem., 16 (1970) 530. Folch. J., Lees, M.. and Sloan-Stanley, G.H.. A simple method for the isolation and purification of total lipids from animal tissues, J. Biol. Chem., 22 (1957) 497. Havel, R.J., Eder, H.A. and Bragdon. J.P.. Distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum, J. Clin. Invest., 34 (1955) 1345. Chen, P.S., Toribara, T.Y. and Warner, H.. Microdetermination of phosphorus, Anal. Chem., 28 (1956) 1756. Kjeldsen, K., Astrup. P. and Wan&up, J., Ultrastructural internal changes in the rabbit aorta after moderate carbon monoxide exposure, Atherosclerosis, 16 (1972) 67. Astrup, P.. Kieldsen, K. and Wanstrup. J., Effects of carbon monoxide exposure on the arterial wslls, Ann. N.Y. Acad. Sci., 174 (1970) 294. Thomsen, H.K., Carbon monoxide-induced atherosclerosis in primates, Atherosclerosis, 20 (1974) 233.
536
19 20 21
Zilversmit. D.B., A proposal finking atherogenesis to the interaction of endothelial lipoprotein lipase with triglyceride rich lipoproteins, Circ. Res., 33 (1973) 633. Topping, D.L., The acute effects of carbon monoxide on the metabolism of perfused rat liver, Biothem. J., 152 (1975) 425. Turner, D.M. and Topping. D.L.. The effects of tobacco smoke and some of its constituents on triglyceride secretion in the squirrel monkey, Res. Comm. Chem. Path. Pharmac., 12 (1975) 85.