Department of Pathology, University of Washington School of Medicine, Seattle, Wash . .98 I 95

Implications of the Monoclonal Character of Human Atherosclerotic Plaques Hinweise auf den monoklonalen Charakter der atherosklerotischen Plaques beim Menschen E. P. BENDITT1) With

I

Figure and 3 Tables· Received March 15, 1976. Accepted March 30, 1976

Key words: Mutagens - Cholesterol derivatives - Hypertension - Genetics - Environment

Summary The evidence for the monoclonal nature of human atherosclerotic plaques is briefly reviewed. The interpretation of the cellular monotypy found in a large percentage (75 to 800/0) of discrete raised atherosclerotic plaques as being monoclonal in origin appear firm since the size of patches of cells of the same type appears to be very small (ca. 10-4 cu mm and 10 ± cells). Evidence for explanations other than single cell origins of each plaque do not appear, at the moment, to be compelling. If we assume then that there is reason to accept the monoclonal ,character of atherosclerotic plaques in human beings thn we are led to a search for the presence of factors that could "initiate" the monoclonal proliferation as well as factors that may promote the growth of the plaques. Evidence is presented that increased risk of atherosclerosis found with cigarette smoking is due to absorption from the lung and circulation in the blood of aryl hydrocarbons. Experiments show that these are preferentially carried in the same parts of the serum that transport cholesterol. The possibility of intrinsic initiators derived from cholesterol is discussed. A possible mechanism for the role of hypertension in promoting atherosclerosis is considered. Finally the evidence for a possible role of viruses in the pathogenesis of atherosclerosis is discussed.

1) Supported in Part by U.S.P.H.S. grants

# HL-oJI74, G.M.-I3543.

406 . E. P. Benditt

The commonest lesions of atherosclerosis seen in vessels of human beings at autopsy take the form of a smooth-surfaced mass raised above the level of surrounding nonatherosclerotic vascular intima. Such raised lesions varying in color from a pearl gray to faintly yellow gray are, on histological examination, composed of cells embedded in dense extracellular connective tissue. Selective stains identify collagen as the main connective tissue fibrillar constituent and elastin usually as a minor or variable constituent. Glycosaminoglycans are present in varying amounts in the extracellular matrix. These histologic findings are confirmed by electron microscopy. The electron microscope reveals, furthermore, that the cells embedded in and responsible for the production of the extracellular matrix substances have properties identifying them as smooth muscle (Haust et al., 1960; Geer et al., 1961; Daoud et al., 1964). Lipid stains and electron microscopic examination show that the smooth muscle cells may have very little fat in them, except in the deeper layers of the plaques. In some plaques lipid is present and substantial amounts of cholesterol can be found both in cells of the deeper layers of the plaque and in the atheromatous debris underlying the plaques. Smooth muscle cells in the deeper layers of plaques, adjacent to the atheromatous debris, frequently exhibit pathological fatty changes, indicative of cell injury, and cells in the deepest layers appear frequently to be dying and disintegrating. All of this mass of the plaque lies in the intima beneath the endothelium and entirely on the lumenal side of the internal elastica of the artery. The masses are usually sharply raised above the adjacent intima in the undistended vessel. Moreover, some features of the cellular mass and associated extracellular matrix set it off from the underlying media: Cells of plaques appear to be smaller than cells of normal media. The main extracellular material produced by plaque cells is collagen, whereas, elastin is a major item in the aortic media. The arrangement of cells of the plaque lacks the order found in the arterial media and the numbers of intercellular junctions appear to be reduced. These features, first clearly recognized in spontaneous lesions of chicken aorta, appear to us to be shared by plaques of human aorta and coronary arteries (Moss and Benditt, 1970). The impressive feature of the human lesions is the presence of an excessive and apparently useless mass of cells resembling in many respects, but differing from in the subtle ways indicated, the cells and structures comprising the media of arteries. It seems reasonable to emphasize the pathological features of the human atherosclerotic plaques because it is the origin of this lesion we are seeking. There is a tendency to confuse the lesions induced in animals by various manipulations with the real lesion. The animal experiments are

Human Atherosclerosis . 4 0 7

helpful in developing understanding of vascular responses to injury, but our ultimate goal is the answer to the question: "What is the nature and origin of human atherosclerosis?" Since the lesion of atherosclerosis has a mass of cells as a major feature, we can rephrase the question as follows: "What is the nature of the cellular proliferation involved in the 'new formation', the atherosclerotic plaque"; specifically, we can ask "Is it of multicellular or is it of monoclonal origin?" The importance of this distinction lies in the fact that many benign new growths have been found to be of monoclonal origin. On the other hand, many common cell proliferations seen in embryogenesis, maintenance and repair are multicellular in character. It is axiomatic in research that the form of the question posed will determine the direction of our search for causational factors in disease. Posing our questions in this way is reasonable to do now since there is currently a basis for obtaining an answer and the methods involved are applicable to human tissues and to human lesions. The method of analysis applicable to study of proliferated masses of cells and for distinguishing the origin of these from one or from many precursor cells requires individual organisms that are mosaics, that is comprise a mixture of two of more distinctive cell types. According to the concept of Mary Lyon (I968), all human females are mosaics, composed of two phenotypically distinct cell types. This is due to the fact that early in embryonic development there is a random inactivation in each somatic cell of one or the other of the two X-chromosomes. Once inactivation has occurred, each cell reproduces true to type and all daughter cells of a particular cell exhibit the activity of the single, same X-chromosoma genes. The stability of this state has been shown by cultivating single cells from connective tissues of human nosaic donors (Davidson et al., I963). Given a stable mosaic population it is possible to ask questions in regard to cell population origins in embryogenesis and whether a pathologic new formation is derived from one or from many cells. The enzyme glucose-6-phosphate dehydrogenase (G-6-PD) is a polymorphic, X-linked gene product (Beutler et al., I962). Of particular interest among the large number of variants now known to exist are two, the common B form and the A form migrating more rapidly in an electric field. The A form of G-6-PD is present in a substantial proportion of the black population. In our series 40% of I20 cases of black females are heterozygous and their tissues comprise two cell types each producing one or the other enzyme form. This property of females has been used to assess the origin of cell populations in a number of tumors (Fialkow, I974)

including benign smooth muscle tumors of the uterus (Linder and Gartler,

19 6 5).

408 . E. P. Benditt

In our first series atherosclerotic plaques from aortas of three women when analysed by this method revealed the fact that most discrete plaques appear to be monoclonal, i.e. they comprise a monotypic cell population (Benditt and Benditt, 1973). In the original series we presented data on 30 plaques from the three cases and 59 samples of uninvolved aorta (see Tab. I); 24 of the 30 plaques exhibited one enzyme type. Pearson et al. (1975) have recently published data on material examined in a similar way and report almost the identical result (Tab. I). The combined data indicate that 85% of 59 raised fibrous plaques are clearly monotypic whereas only 2.5010 of apparently normal aortic wall exhibit single isoenzyme patterns. In our recent studies the incidence of monotypic samples in more than 1,000 assays of normal artery wall has given a lower figure than this « 1010). In addition, further analysis of discrete plaques continue to show a high incidence of monotypic examples (75-80010). The fact that the bulk of raised atherosclerotic lesions are monotypic with regard to cell type, as indicated by isoenzyme pattern, does not immediately yield the strong inference that lesions of atherosclerosis are monoclonal. Such monotypic cell masses could arise in one of the following ways: I) If the 6-G-PD enzyme variant itself, or if some gene linked by location on the X -chromosome, imparted to cells a selective growth advantage (or disadvantage) then we could get monotypy due to stimulus to cell multiplication of any kind, e.g. local injury, increased nutrient supply, or growth promoting factor. 2) If the pattern of cell mixture in the artery wall were such that monotypic clusters (not necessarily clones) were present and of sufficient size then monotypy could result from localized pro-

Table 1. G-6-pd phenotypes of cell populations of human atherosclerotic plaques compared to non plaque artery samples

Non Plaque

AB

Benditt & Benditt (4 cases)

57

Pearson et al. r975 (r6 cases)

99

Totals

Plaques

A

B

Tot.

AB

0

2

59

6

A 8

ror

r60

B

Tot.

r6

30

9

9

-----~------------ - --

59 - - - -- - - -

-

Human Atherosclerosis . 4 0 9

liferative stimuli. 3) The third possibility is that during cycles of cell death and cell multiplication repetitive sampling of the population could lead to a drift toward a monotypic population. The question whether one or the other of the G-6-PD variants gives a selective advantage to the cell types is easily disposed of in the following manner: In all of the original three cases and in four of the 16 cases of Pearson et al. (1975) individual plaques of each of both types were found in the same vessels. Thus neither G-6-PD isoenzyme type nor linked genes appear in these cases to provide a major selective advantage. The problem of whether in some instances genes linked tol the G-6-PD locus can provide a selective advantage is a more subtle one and requires further analysis. However, the proliferative advantage must arise in other ways in the majority of cases. We have addressed ourselves to the question of patch size in the following way: We first attempted to ascertain the smallest patch size by examining very small samples of the order of .25 down to .03 cu mm in volume. The enzyme ratios appeared to show an expected increase in variance with reductions in sample size. However, when we attempted to examine larger samples to see if their variability was limited to that of methodological error we found a variation substantially larger than that of the methods error. This led to a systematic examination of the pattern of cell population variation in the artery wall. From this analysis emerges the following fact: The pattern of the variation showed unmistakable clusters of values, the within cluster variance is of the order of the error variance (Tab. 2).

Table 2.

Fragment of map of cell mixtures in aortic wall

--.-------

------

37

37

38

31

30

33

31

33

36

36

32

36

38

31

31

33

33

26

35

38

34

35

32

37

33

37

40

41

36

35

35

37

38

40

38

40

32

39

38

34

35

37

39

34

36

38

34

33

38

Values are % A type cells (as measured by 'A' enzyme proportion) in sample from thoracic aorta of 16 year old. Area approximately 1.5 X 1.5 cm. 27

Beitr. Path. Bd. 158

410 .

E. P. Benditt

The percentage of A cells, indicated by the percentage of A enzyme, showed a range of variation between clusters some five to seven times that of the error of variance. We interpret this clustering to be a manifestation of confined growth of a group of cells recruited during early embryonic development from the mesenchyme to form the arterial media. Irrespective of the interpretation of how these clusters arise, the fact of their presence becomes important in the analysis of the limiting patch size. An analysis of the sources of variation indicates that there should be three constituent factors contributing to the variance: 1) a variance contribution due to the draw from the embryonic pool, 2) a variance contribution due to true clonal proliferation plus accidental clustering of cells of the same type i.e. monotypic patches, and 3) a contribution from methodologic error. From the variation found in twenty one samples of volume, .03 to .1 cu mm, on which we could make satisfactory enzyme measurements, we found that almost the entire variance is accounted for by the contribution of the draw from the embryonic pool. This and the methods error leave only a very small fraction of the total variance attributable to the individual patches of monotypic character; the estimate of the monotypic patch size is of the order of 10-4 cu mm. The important point to be derived from this analysis is as follows: Any injury that allows access of nutrients or growth stimulants in order to produce coherent monotypic cell proliferation would have to operate on a very small group of cells, of the order of 10 in a space of the order of 10-4 cu mm. This seems a highly unlikely event to produce 80% of plaques that are clearly monotypic. Finally, to answer the question of selection based on population drift we have examined patch size in the non atherosclerotic aortic wall over an age span of 1 week to 91 years. There does not appear to be any substantial change in the character of the cell mixture in non atherosclerotic intima over this age range. This is so in spite of the fact that with age the intima becomes progressively thicker. This thickening is most certainly due to death of cells and their replacement resulting from the mechanical, chemical and microbial insults occurring with time. Taking all of the facts together the reasons are strong for believing that raised lesions are monoclonal. Since the plaques are almost invariably monoclonal the insudation model is no longer tenable. Organization of mural thrombi as the origination of atherosclerotic plaques in human beings seems equally untenable, since organization of thrombi appears to involve many cells (Poole et al., 1970). This is not to say that simple injury and repair

or insudation do not occur. However, the causes of monoclonal proliferation must be sought among a different set of factors. Prominent among

Human Atherosclerosis . 41 I

these is the role of chemical mutagens (or premutagens) derived from the environment; another possibility is that there are endogenous substances derived from or related to cholesterol and its metabolism. Also possible is the role of viral agents. With any formulation regarding the etiology and pathogenesis of a disease one asks whether there are inferences that can be tested, or does it offer explanation of facts already available. We wish now to examine several integrative and predictive features of this hypothesis that none of the other hypotheses so far introduced have in relation to several established risk factors for atherosclerosis. Cigarette smoking is well established as a "risk" factor. Burning of cigaretts produces aryl hydrocarbons and some of these (benzo(a)pyrene) are well known precarconogens. It is easy to conceive of the hydrocarbons from the smoke affecting the lung; can they also affect the arteries? The question is: Whether chemical substances from the smoke get into the blood and then to the artery wall. Evidence that the precarcinogens get into the blood is available. As shown in Table 3, the aryl hydrocarbon hydroxylase, an inducible enzyme system, is elevated enormously in the placentas of women who are heavy smokers and intermediate amounts in women of intermediate smoking habits, when compared with enzyme levels of non smokers (Juchau et aI., 1974). Clearly the route must be from the lung via the blood to the placenta.

Table 3. smokers

Aryl hydrocarbon hydroxylase activities of term placentas of smokers and non

Cigarettes: A v. Smoked/Day

No. of Cases

Mean value 3,4 Benzo(a)pyrene Hydroxylation + S.E. nmol/g. protein per 15 min

o 10

±

4

8

29 ±

6

10

4

5

20

243

± 19

> 25

2,160

± 213

8 - - - - - - -

Modified from Juchau et al. Drug metabolism and Disposition,

2 :

79, 1974.

4I2 .

E. P. Benditt

It becomes of interest to ask where in the blood plasma the aryl hydrocarbons travel. We would like specially to know whether or not they travel in the same plasma fractions that cholesterol travels. Evidence has been presented, as one might expect from the solubility characteristics of the aryl hydrocarbons, that they travel with the blood plasma lipoproteins (Margolis, 1969). We have begun to reexamine the problem as a preliminary to searching for the presence of these, or similar compounds, in the plasma of human beings. The comparison we have made is that of the distribution of cholesterol and of 3-methyl cholanthrene in samples of the same human serum. Cholesterol-C-I4, and 3-methyl cholanthrene-C-I4 were used for the comparison. Each labeled compound was dissolved in benzene. Two tenths of a ml. of the benzene solution of the compound was placed in a scintillation counting vial and the vial supported in a roller tube rack on its side. Using a gentle stream of nitrogen each of the compounds was evaporated in a very thin (invisible) layer on the sides of the tube. One ml. of human serum was placed into each vial and the vials rotated at 37 degrees C for 24 hours. The serum was then removed from the vial and centrifuged at 20 degrees C and 20,000 times gravity for 30 minutes to remove any solid that might have flaked off into the serum. By this means it was found that about 86% of the cholesterol and 82% of the methyl cholanthrene were dissolved in the plasma. Small aliquots of the plasma (15 microliters) were applied to filter paper strips and electrophoresed in a spinco-durrum electrophoresis apparatus using the standard barbital buffer system. From a given run separate strips were stained for protein (bromphenol blue), lipoproteins (oil red 0), DISTRIBUTION OF CHOLESTERO AND 3-M[-CHOLANTHRENE IN SERUM IlZl MeA D

n

Choleslerol

I,

· 11

l A seru~lprolelOs

IViI

Serum lipoproteins

r

I \I I

+I I

I

I I

I

~

I

IAI~

I

I

I

I

J

\

I

-20

\

0

20 40

60

Electrophoretic mi9rollon

Fig.

1

80

Human Ather05;clerosis . 413

and cut into sections 2 mm wide for radioactive isotope counting. The strips stained for protein and for lipid were scanned with a densitometer. The results of such an experiment are shown in Figure 1. It is readily apparent that the cholesterol travels with the lipoproteins, mainly the beta lipoproteins. Moreover, the aryl hydrocarbon travels in essentially the same portions of the lipids of the serum as the cholesterol, except that relatively less is found in the a lipoprotein portions of the electrophoretic serum protem spectrum. Further refinements of the analysis are being made and a method for extracting, detecting and measuring the amounts of compounds of this general character is being developed. With such a method we will be in a position to examine populations of human beings prone to and those not prone to development of atherosclerosis. Obviously the first population in which to look for aryl hydrocarbons is smokers and to compare the constituents present in the plasmas of smokers with those of nonsmokers. Assuming that we find these substances in the plasmas of smokers, it will be of great interest to see if any other individuals (nonsmokers) have these substances. There seems to us at the moment to be a real possibility that some populations are exposed to potentially mutagenic hydrocarbons from sources other than cigarette smoke and that we are only dimly or not at all aware of this exposure. Another interesting possibility that the monoclonal concept of the nature of atherosclerotic plaques suggests is the following: In addition to compounds, such as the aryl hydrocarbons, that are mutagenic, there may appear in the plasma of individuals substances not derived from environmental sources but derived from intrinsic substances via aberrations in the individuals own metabolism. We already know of at least one sterol derived from cholesterol that is oncogenic and presumably mutagenic and in fact is capable of causing tumors in animals, cholesterol -5a, 6a- epoxi de (Bischoff, 1969). This substance has already been found in human serum (Gray et aI., 1971). If we use an assay for mutagenic substances that is sufficiently sensitive then we should be able to find these substances in the proper serum extracts. Such a sensitive assay is already at hand, the reversion test of Ames using S. typhi murium (Ames et aI., 1973). This we are adapting to the problem. If, using this system, we find such compounds we can then go on to ascertain whether or not there are defects in the enzymes forming or degrading the compounds. There is already evidence that genetically controlled factors are involved in the propensity of some people to develop atherosclerosis at an accelerated rate. This being the case, one may find at least a part of the explanation, not simply in the presence of an elevated cholesterol, but in the fact that some enzymatic defect

414 . E. P. Benditt

permits the appearance in plasma of such injurious derivatives as the cholesterol 5a, 6a epoxide, a known mutagen. Hypertension is another important risk factor associated with a marked increase in the incidence of atherosclerosis. Studies of the effect of hypertension on the artery wall in animals are being carried out by Dr. Steven Schwartz in our laboratory. The first interesting fact that has appeared in these studies is that hypertension, even of brief duration, causes increased multiplication of the endothelial lining cells of arteries. An important question is whether the mechanical effects of the hypertension or associated chemical factors cause this and related proliferative changes in the arteries. Finally, we should consider the possible role of viruses in the production of therosclerosis. At the moment we have two pieces of evidence of a presumptive nature on this point. First, is the fact that young women, age 25 to 30, have been showing up with myocardial infarcts. These have been found to be due to atherosclerosis with thrombosis. We have had the opportunity to examine three cases. In each instance, the young women all had disseminated lupus erythematosus. The lesions are not arteritis, as was originally supposed. The plaques appear to fit appropriately the category of atherosclerosis. The current belief is that disseminated lupus erythematosus is a latent virus disease. If this is so, then we may have our first virus cause of atherosclerosis. The second piece of information has to do with estrogen treatment and with the spontaneous atherosclerosis of the chickens and estrogen treatment of men post myocardial infarcts. Dr. Moss and I, in repeating the work of the Katz group (Pick et al., 1962) treated some of the birds with estrogen. We found that half the animals died after several months, some of a virus lymphoma that chickens are prone to (Burmester, 1945). In addition, we noted in the survivors that the atherosclerotic plaques were worse. This was similar to what Pick et al. had noted. Furthermore, we found in one of the animals that there were typical C-type virus particles multiplying in the cells of the plaque and elsewhere in the artery wall. This observation seems specially important when we recall that in the recently completed Coronary Drug Project the administration of estrogen was associated not with a decreased but with an increased death rate from new M. I.'s. Furthermore, there was seen an increase in the incidence of cancer in the treated group. The effects of estrogens (especially diethylstilbesterol) in eliciting latent viruses and inducing expression of lymphomas in mice have now been observed (Fowler et al., 1972). Could this be the mechanism involved in the excess mortality found in the estrogen portion of the Coronary Drug Project?

Human Atherosclerosis . 4 15

The grouping of facts presented above highlights the proposition that we must look at the inner surface of arteries in the same way that we view the lining inside of the bronchi, the inner lining of the gastrointestinal tract and the surface of the skin. All three of these are more or less closely, and in different ways, in contact with the environment. Most of the material absorbed from the gastrointestinal tract, much of the gaseous material inhaled and some of the material applied to the skin enter the blood stream. The inner lining of the vessels thus become exposed to injurious environmental agents and, being a specialized, but otherwise ordinary tissue, responds by healing wounds, by producing inflammatory responses to infections and by tumor formation with the appropriate stimuli. Atherosclerosis has been spoken of as a "multifactorial" disease. However, the nature of the factors that may be involved has only been looked at in the most superficial manner and in the most simple terms. A modern approach to Pathology demands that we better utilize the large repetoire of ideas and of methods to look at the problem of the many factors that may be involved in the etiology and the pathogenesis of this most important disease process. The invention of a new question and the implications of the answer to that question, i.e. the monoclonal character of the proliferations that make up the plaques, provides a new and pertinent challenge for our imaginations, our analytical powers and our technical skills.

Zusammenfassung Es wird kurz iiber die monoklonale Natur der menschlichen atherosklerotischen Plaques berichtet. Die Interpretation, daB die zelluHire Monotypie, welche in einem groBen Prozentsatz der atherosklerotischen Plaques gefunden wird, monoklonalen Ursprungs sei, scheint gesichert. E~ stellt sich deshalb die Frage, welche Faktoren den AnstoB zu einer monoklonalen Proliferation und welche Faktoren das Wachstum dieser Zellplaques bewirken konnen. Der Zigarertenrauch, der als Risikofaktor bei der Atherosklerose bekannt ist, gelangt nach der Absorption durch die Lunge in die Blutzirkulation in Form der Arylhydrokarbone. Diese Molckiile werden vorwiegend durch denselben Anteil des Serums transportiert, der auch den Cholesteroltransport bewerkstelligt. Ein moglicher Mechanismus, welcher dem Hochdruck bei der Arteriosklerose eine wichtige Rolle zuschreibt, wird besprochen. SchlieBlich werden Hinweisc fUr eine mogliche Virusgenese im Rahmen der Atheroskleroseentstehung crortert.

References Ames, B. N., Lee, F. D., and Durston, W. D.: An improved bacterial test system for detection and classification of mutagens and carcinogens. Proc. nat. Acad. Sci. (Wash.) 7°,7 82-7 86 (1973)

416 . E. P. Benditt Benditt, E. P., and Benditt, J. M.: Evidence for a monoclonal origin of human atherosclerotic plaques. Proc. nat. Acad. Sci. (Wash.) 70, 1753-1756 (1973) Beutler, E., Yeh, M., and Fairbanks, V. F.: The normal human female as a mosaic of X-chromosome activity: Studies using the gene for G-6-PDdeficiency as a marker. Proc. nat. Acad. Sci. (Wash.) 48, 9-17 (1962) Burmester, B. R.: The incidence of lymphomatosis among male and female chickens. Poultry Sci. 24, 469 (1945) Bischoff, F.: Carcinogenic effects of steriods, Advanc. Lipid Res. 7, 165-244 (1969) Coronary Drug Research Group: The coronary drug project findings leading to discontinuation of the 2.5 mg/day estrogen group. J. Amer. med. Ass. 226, 652 (1973) Davidson, R. G., Nitowsky, H. M., and Childs, B.: Demonstration of two populations of cells in the human female heterozygous for glucose-6-phosphate dehydrogenase variants. Proc. nat. Acad. Sci. (Wash.) 5°,481-485 (1963) Daoud, A., Jarmolych, ]., Zumbo, 0., Fani, K., and Florentin, R.: "Preatheroma" phase of coronary atherosclerosis in man. Exp. molec. Path. 3, 475-484 (1964) Fialkow, P. J.: The origin and development of human tumors studied with cell markers. New Eng!. J. Med. 291, 26-35 (1974) Fowler, A. K., Reed, C. D., Todara, G. ]., and Hellman, A.: Activation of C-type RNA virus markers in mouse uterine tissue. Proc. nat. Acad. Sci. (Wash.) 69, 2254 (1972) Geer, J. c., McGill, H. C. Jr., and Strong, J. P.: The fine structure of human atherosclerotic lesions. Amer. J. Path. 38, 263-287 (1961) Gray, M. F., Lawrie, T. D. V., and Brooks, C. J. W.: Isolation and identification of cholesterol IX oxide and other minor sterols in human serum. Lipids 6, 836-843 (1971) Haust, M. D., More, R. H., and Movat, H. Z.: The role of smooth muscle cells in the fibrogenesis of atherosclerosis. Amer. J. Path. 37, 377-389 (1960) Juchau, M. R., Zachariah, P. K., Colson, J., Symms, K. G., Krasner, J., and Yaffe, S. J.: Studies on human placental carbon monoxide-binding cytochromes. Drug Metabo!' and Disposit. 2, 79-76 (1974) Linder, D., and Gartler, S. M.: Glucose-6-phosphate dehydrogenase mosaicism: Utilization as a cell marker in the study of leiomyomas. Science 150, 67-69 (1965) Lyon, M. F.: Chromosomal and subchromosomal inactivation. Ann. Rev. Genet. 2, 31 (1968 ) Margolis, S.: Structure of very low and low density lipoprot,eins. In: Structural and functional aspects of lipoproteins in living systems, edited by E. Trea, A. M. Scanu, p. 144. Academic Press, New York (1969) Moss, N. 5., and Benditt, E. P.: The ultrastructure of spontaneous and experimentally induced arterial lesions. II. The spontaneous plaque in the chicken. Lab. Invest. 23, 231-245 (1970) Pearson, T. A., Wang, A., Solez, K., and Heptinstall, R. H.: Clonal characteristics of fibrous plaques and fatty streaks from human aortas. Amer. J. Path. 81, 379-387 (1975 ) Pick, R., Katz, L. M., Century, D., and Johnson, P. J.: Production of ulcerated atherosclerotic lesions in cockerels. Circulation Res. II, 811 (1962) Poole, J. c. F., Cromwell, S. B., and Benditt, E. P.: Behavior of smooth muscle cells and the formation of extracellular structures in the reaction of arterial walls to injury. Amer. J. Path. 62, 391-414 (1971) Earl P. Benditt, M. D., Professor and Chairman, Department of Pathology SM-30, School of Medicine, University of Washington, Seattle, Wash. 98195