Journal of Clinical Investigation Vol. 44, No. 11, 1965

Studies on the Influence of Dietary on Cholesterol Metabolism in the Isotopic Steady State in Man *

JEAN D. WILSON t AND CHARLES A. LINDSEY, JR. (From the Department of Internal Medicine, The University of Texas Southwestern Medical School, Dallas, Texas)

The relation between dietary cholesterol and it can quantitatively account only for a degree of the concentration of cholesterol in serum has been protection equal to that amount ordinarily syn- the subject of intense interest since the original thesized in the liver, whereas the serum choles- demonstration that both hypercholesterolemia and terol remains virtually stable despite the in- atherosclerosis could be induced in the rabbit by gestion of quantities of cholesterol many times the feeding of diets rich in cholesterol (2). On this value. Second, the ingestion of cholesterol the other hand, in man virtually all studies have could cause an increase in the rate of degradation demonstrated that rather wide variations in the of cholesterol to bile acids. This phenomenon cholesterol intake result in less marked effects has been clearly documented in the rat (12, 13); on the serum cholesterol (3-8). Even when the in this species increased bile acid production oc- intake of dietary cholesterol greatly exceeds the curring with very effective cholesterol absorption normal turnover rate, the concentration of choles- is the major means by which serum cholesterol terol in the serum almost never rises more than is stabilized. Finally, it is obvious that any effect 100 mg per 100 ml (8), and indeed, in careful of exogenous cholesterol on cholesterol metabo- balance studies it is necessary to compare cho- lism is dependent on its rate of absorption. Be- lesterol diets with diets totally devoid of choles- cause cholesterol is both absorbed and secreted terol before any significant changes can be across the wall of the intestine, the best means detected (4-7). This ability of the serum choles- of quantifying cholesterol absorption in the intact terol of man to stabilize at only moderately animal involves the use of a complicated double elevated levels despite enormous intake resembles isotopic steady state technique (13), and, conse- the situation in such animals as the rat (9) and quently, very little is known about net cholesterol is in marked contrast to the profound degrees of absorption in various species. hypercholesterolemia produced in the rabbit by To evaluate in the human the possible roles of cholesterol feeding (2). these three factors-the inhibition of cholesterol At least three compensatory mechanisms could synthesis, increased bile acid production, and a account for this phenomenon. First, dietary cho- limited ability to absorb dietary cholesterol-five lesterol suppresses the hepatic synthesis of cho- balance studies have been performed in two nor- lesterol in the human (10) as well as in other mal men who were fed varying quantities and species (11) via a negative feedback system. types of cholesterol-4-14C until an isotopic steady However, the suppression of hepatic synthesis state was attained. The results clearly demon- the compensatory mechanism since cannot be only strate that dietary cholesterol, whether fed in the * Submitted for publication April 29, 1965; accepted form of eggs or in a purified state, has little July 15, 1965. effect on the total rate of endogenous cholesterol Supported by grants from the U. S. Public Health to Service and the American Heart Association. synthesis or of cholesterol degradation bile The results have been published in abstract form (1). acids. We have concluded, therefore, that in tWork performed during tenure as an Established In- man the major mechanism that stabilizes the vestigator of the American Heart Association. serum cholesterol during large variations in the Address requests for reprints to Dr. Jean D. Wilson, is a limited ca- Dept. of Internal Medicine, The University of Texas intake of exogenous cholesterol 'Southwestern Medical School, Dallas, Texas 75235. pacity for cholesterol absorption. 1805 1806 JEAN D. WILSON AND CHARLES A. LINDSEY, JR. Methods position of the tip of the tube as evidenced by a mercury marker was checked by fluoroscopy just before the col- Two normal men, ages 31 and 23, were studied during lection of the samples, and consequently, although tele- five different feeding periods varying from 21 to 40 days scoping of bowel around the tubes can occur, the position- over an 8-month interval. During the periods of study ing of the tubes was thought to be reliable. the subjects were ambulatory outpatients who ate only a Serum cholesterol analysis. The serum cholesterol constant weighed quantity of formula diet, vitamins, specific activity and content were measured as follows: Hawk-Oser salt mix, and clear liquids. After a steady 2 ml of serum was mixed with 0.4 ml of 10 N KOH and state had been approached, stools were collected into re- saponified in an autoclave for 30 minutes at 15 pounds frigerated, tared aluminum containers for periods of time pressure. An equal volume of ethanol was then added, that varied from 4 to 7 days; the stool samples. were and the solution was brought to a boil on a steam bath. pooled and treated as a single sample for the purpose of Neutral were then extracted into a tenfold excess the subsequent analyses. of ethyl ether, and after backwashing the ether extracts Two different formula diets were utilized in these were dried. After dissolving the residue in acetone-alco- studies. For the periods in which crystalline cholesterol hol (1:1) cholesterol digitonides were then formed, was added to the diets the basic formulas consisted of washed, and dried by the method of Sperry and Webb triolein,l vitamin-free casein,1 gelatin, sucrose, vanilla ex- (16). The cholesterol digitonides were dissolved in tract, egg white, and methyl cellulose; the analyzed ; one aliquot was added to 0.4% diphenyloxazole content of this basic formula was less than 5 mg of digi- in toluene and assayed for 14C in a Packard liquid scintil- tonin precipitable material per day. Varying amounts of lation spectrometer, and one aliquot was assayed for cho- crystalline cholesterol-4-1C were added to this basic diet lesterol content (16). Each determination was performed by dissolving the cholesterol in a portion of the triolein in duplicate. Specific activity determinations performed and feeding it directly in either two or three portions. per in this way varied less than 5% when rechecked on differ- day. The cholesterol-4-14C used in these studies was ob- ent days. tained by adding cholesterol-4-14C 2 that had been purified Bile acid and neutral sterol content of . To the by chromatography on silica gel in benzene: ethyl acetate weighed feces samples was added 2 vol of water, and the (4:1) (14) to cholesterol that had been regenerated from mixture was shaken on a paint can shaker for 5 minutes the dibromide and then recrystallized five times (15). and then transferred to a Waring blendor and homogenized This mixture was then recrystallized three times from for 5 minutes. One- to 2-g samples of the liquid mixture ethanol. were then carefully weighed in triplicate, mixed with 100 For the formula diets in which the cholesterol source ml chloroform: methanol (2: 1), and refluxed for 5 min- was egg yolk, the eggs were obtained as follows: eight utes. The chloroform: methanol extracts were filtered by laying hens were each injected with 1.2 X 108 cpm of washing into 250-ml Erlenmeyer flasks, dried under ni- cholesterol-4-j4C dissolved in ethanol, and all the eggs trogen by boiling, mixed with 3.5 ml of 10 N NaOH and were collected for 1 month. The pooled, frozen yolks (6 10 ml H20, and autoclaved for 3 hours at 15 pounds dozen) were then diluted with 140 dozen fresh egg yolks. pressure and 1300 C. Two vol of ethanol was then added, This mixture was stirred repeatedly in two large batches and the samples were refluxed on a hot plate for 5 min- and frozen in weighed aliquots. The formula for this diet utes. Neutral sterols were extracted three times from consisted of sucrose, vanilla extract, gelatin, and egg the mixture by shaking with pentane; the pentane extracts yolk that were whipped together and then frozen. The were combined, backwashed with 10 ml of ethanol: water total daily caloric intakes for each of the study periods (1: 1), and dried under nitrogen. After dissolving the and the calculated distribution of the calories are given residue in methanol one aliquot was assayed for 14C, and in Table I. In each instance the cholesterol content and one was saved for thin-layer chromatography. The wa- the cholesterol specific activity were measured directly ter layer remaining after the pentane extraction was acidi- by analysis of the entire diet. fied to pH 2 to 3 by titrating with concentrated H2SO4, At the end of the first feeding period a study of the and the bile acid fraction was extracted into ethyl ether. change in specific activity of the dietary cholesterol at The ethyl ether extracts were dried under nitrogen and various levels in the was performed dissolved in 50 ml acetone. The acetone extract was then in each of the subjects. A Levine tube was allowed to passed through a Florisil column (1 X 10 cm), and the pass overnight into the duodenum (95 cm from the bile acids were eluted with 50 ml of acetone: methanol: mouth), and then a second tube was passed into the stom- acetic acid (80: 20: 1) (17). The eluate was. then dried ach (65 cm from the mouth). Fasting samples of gas- on a steam bath, and the residue was dissolved in metha- tric juice and succus entericus mixed with bile were nol. One aliquot was assayed for 14C, and one was chro- then aspirated, and during the next 24 hours the tubes matographed by thin-layer chromatography. In every were allowed to pass to the jejunum (150 cm) and the instance the samples were reassayed for 14C after the ad- ileum (250 cm), respectively, and fasting samples were dition of an internal standard, and appropriate corrections again collected the next morning. In each instance the for quenching were made. To separate cholesterol from bacterial transformation 'Nutritional Biochemicals, Cleveland, Ohio. products, the neutral sterol fractions were chromato- 2 New England Nuclear Corp., Boston, Mass. graphed on silica gel plates in benzene: ethyl acetate DIETARY CHOLESTEROL IN MAN 1807

t- 00 ) 0 NO -4 -t - U)U)d U)C e - , 0 0U)to- WU)i)0U) 0 Ut )-4C U) VUN 0-W C-40 CD C- to 40 to)00 CU a-ICVc 06'lro S N0%0U)ItO0 10 0 00 V U) - 0% - d4 0 00 00 0 U) C-U 0% 0*- OOa-a-u) C a-. 0004 C-I4U~) U)+4 V%O 00a--0% 0 0 C-a0(- Vin 0% 0% m - ., VO 00ON C-i -VO U) - C14 4

Q ~o C.) r. 0C-4C-4 0 -*-.U).. 00O000 U) e14 C-a C--C--- U) V I' C-a c"I (OItU'-I4 4.) C.) ) VU)oo--) t. 40tU 0% 000 '-C I. C-I 00 'a QimO c d4in moo iCU o C-I orU)oo 'a 0 e) U) 0 11 C-4 000CA C--I - sf -- C-U)_)_U)) _u) to to o m 'a '-a 'a 00 0 C-4 -4 'a C--0 a C-a 0% ~4 00U) 0)Q0000 U) ,bOd -.4 C-I Cf)00 %a-- .. Vc W) oO 4 :4 ok01 I. U) a)W

U) tn a) 'a ) '-a 44 -at r. cou I-4 CtJ 4- ;rb C--)co 'I .. Ca 'a HaU~ ._0c0 'a-' r) -0Ca (U

'a 'a a) °NdtDo°~~~~Cd u aa 4-)

'a (n .C4)0 'a S 4- (A fI- 'a a)Ut -'a 0 - U)

(0: .a ._ ; E , b4 E Un . - 7L.J Cd -o ~0(U .4) *-, *ot; a)4

.o . 'a NoCa Ca 4z' 0 la U 0Uc)" (Uo 0o r:*WM.E°. 1808 JEAN D. WILSON AND CHARLES A. LINDSEY, JR.

0 0 x 100

F. _ too- CRYSTALLINE CHOLESTEROL DIET (2.281g/dfty) IALANCE PERIOD 1 Q,' EGG CHOLESTEROL DIET (2.78,51day) ,ALAANCE PERIOD 2 4- LOW CHOLESTEROL DIET (o.10 I5ALANCE PERIOD 3 u Q 75 V g/day) -1 tz

A. -.1 00 50H 0Q: 200 L4j -4j I'- W, N. ,. i *^ c i 25 tt100A /.*O t4o- M0j-I-I I .000 (t Q!; 01 0 Q 10 20 30 40 10 20 30 40 D A VS DAYS FIG. 1. THE INFLUENCE OF HIGH AND LOW CHOLESTEROL-4-'4C DIETS ON THE SPECIFIC ACTIVITY (A) AND THE CONCENTRATION (B) OF SERUM CHOLESTEROL IN SUBJECT A.

(2: 1) (14), and after spraying with Rhodamine G the the necessary corrections were made. With this system cholesterol area was scraped and eluted into acetone: the recovery of cholesterol-4-."C and cholic acid-4-4C ethanol (1: 2). Cholesterol digitonides were formed and added to feces samples was checked; the recoveries aver- assayed for 14C and chemical content as before. To in- aged 91% for cholesterol and 96% for cholic acid. sure that each bile acid fraction was free of neutral sterol Analysis of diet and succus entericus. The weighed the ethyl ether extracts were also chromatographed on formula intake along with the daily portions of salt mix, silica gel plates in benzene: ethyl acetate (4: 1) and clear liquids, vitamins, and methyl cellulose were homoge- sprayed with Rhodamine G; in this system the bile acids nized in a Waring Blendor in 2 L of chloroform: methanol remained near the origin, and any contaminating choles- (2: 1). The mixture was filtered, and 5-ml aliquots in terol or coprostanol moved near the front; in some in- triplicate were dried and assayed as described for the stances a small amount of radioactivity could be located stool samples for cholesterol content and 14C. The as- in the cholesterol and coprostanol areas, and consequently pirates from the gastrointestinal tract (averaging 10 ml

0 0

k)c O."-S5 EGG CHOLESTEROL DIET(2.90/d84y) 3ALANCE PERIOD 2 > Z> 100 a_ LOW CHOZESTEROL DIET(0.06c/day) 3ALANCE PERIOD 1

l< < -.4 ki Q 40 km

-4 0 ,oO-'O C)--O-_o.--O0 4 on '44 0x~0 -J4 % Q C,) LA 2f '44, 4 dj QI 40 DA Y S DAYS

FIG. 2. THE INFLUENCE OF HIGH AND LOW CHOLESTEROL-4-14C DIETS ON THE SPECIFIC ACTIVITY (A) AND THE CONCENTRATION (B) OF SERUM CHOLESTEROL IN SUBJECT B. DIETARY CHOLESTEROL IN MAN 1809

in volume) were mixed with- ' vol of 10 N NaOH and jects after the feeding of semisynthetic diets very autoclaved for 3 hours at 15 pounds pressure and 250° C. low in cholesterol content (20 and 35 mg per An equal volume of ethanol was then added, and neutral sterols were extracted with pentane and chromatographed 100 ml) are very similar to the results observed on plates of silica gel. The cholesterol areas were by Conner, Hodges, and Bleiler (5). In these scraped and eluted, and the digitonides were assayed for two subjects the differences between several weeks "C and cholesterol content as before. of feeding diets containing either very low or very high concentrations of cholesterol were 50 Results and 55 mg per 100 ml in the case of Subject A The influence of feeding two normal subjects and 50 mg per 100 ml in the case of Subject B. that these two amounts of cholesterol- It can be concluded, therefore, diets containing varying subjects exhibited a typical response of serum 4-14C on the specific activity and the content of cholesterol concentration to variations in cho- serum cholesterol is illustrated in Figures 1 and lesterol ingestion. 2. In each instance, the specific activity of the The results of the balance data obtained from dietary cholesterol was directly measured by anal- the five study periods in these subjects are illus- ysis of the total daily dietary intake during the trated in Table I. Because of unpalatability, the study. The lengths of the five feeding periods for these two subjects varied from 21 to 40 days, caloric intake of these semisynthetic diets was and in each of these periods an isotopic steady only 1,600 to 2,700 kcal per day. Nevertheless, as state was approached by the end of the study. illustrated by the weights, the maximal weight In the case of the low cholesterol diets (0.10 and change was only 3 kg for a 39-day study period, 0.08 g per day) a steady state was attained within and in each instance the major weight change occurred during the first 2 weeks of the study. 10 days. When diets high in cholesterol were fed, however, an isotopic steady state was not Under the section labeled sterol balance are listed the results of the chemical and isotopic balances. approached until after 20 to 30 days of feeding. cholesterol diets The cholesterol intake and the neutral sterol, bile In Subject A who was fed high each day are listed in either acid, and total 14C excretion which the source was crystalline choles- in columns A, B, C, and D. The per cent re- terol or egg, the ratios of the specific activity of that of the in the covery of the ingested radioactive cholesterol is the serum cholesterol to diet listed in column E; it can be seen that the re- isotopic steady state (18 and 24%) were similar, indicating that the source of the cholesterol had covery varied from 95 to 110%o, and consequently in the little influence on the degree of equilibration it can be concluded that errors subsequent The of be- calculations due to imperfections in collection achieved. percentage equilibration In tween diet and serum in these studies (2.5 and technique cannot be more than 10%o. these for the low cholesterol diets and and studies the total neutral sterol excretion, repre- 4%o 18, 24, the of cholesterol and 38%o for the high cholesterol diets) is similar to senting sum coprostanol, the results obtained by Taylor, Patton, Yogi, and was determined as follows: total neutral sterol Cox (18), and by Kaplan, Cox, and Taylor (19). excretion (grams per day) = neutral sterol out- This confirmation is striking in view of the fact put (counts per minute per day)/fecal cholesterol that in the latter studies the specific activity of specific activity (counts per minute per milligram) x The of are dietary cholesterol was a calculated and not a di- 1,000. results these determinations rectly determined value. The results mean that listed in column J. As would be expected, total paralleled even when the ingested cholesterol grossly exceeds neutral sterol excretion closely choles- endogenous production, no more than 25 to 40%o terol intake, but in each of the five studies total of the serum cholesterol in these two subjects is excretion slightly exceeded the intake. Bile acid derived from the diet, regardless of whether the excretion in the isotopic steady state can be cal- source is cholesterol in a naturally occurring culated as follows: bile acid excretion (grams = (egg) or purified form. per day) bile acid output (counts per minute Furthermore, the falls in serum cholesterol 8 Since the molecular weights of cholesterol and bile acid (Figures 1 and 2, sections B) in these two sub- are similar but not identical, this formulation should only 1810 JEAN D. WILSON AND CHARLES A. LINDSEY, JR. per day) /average serum cholesterol specific ac- the two pathways of cholesterol elimination from tivity (counts per minute per milligram) x 1,000. the body. It is striking that when cholesterol The bile acid excretions for these balances are was fed in large quantities the endogenous sterol listed in column K. The bile acid excretion in fell only from 0.88 to 0.53 and 0.77 g per day these two subjects was markedly different; it in Subject A and from 0.49 to 0.41 g per day in averaged 0.26 g per day in Subject B and 0.97 Subject B. Therefore, despite the operation in g per day in Subject A. However, it is quite clear man of the hepatic negative feedback (11), some that in these subjects bile acid excretion was cholesterol continues to be synthesized. This surprisingly constant regardless of the type or finding confirms results obtained by Taylor and quantity of dietary cholesterol; under circum- co-workers using D20 (20) and clearly means stances in which dietary cholesterol was increased either that the suppression of cholesterol syn- 30-fold, bile acid excretion was 0.83 and 1.07 g thesis in the liver is incomplete or that a major per day in Subject A and 0.23 and 0.28 g per nonhepatic source for the synthesis of endogenous day in Subject B. Within the limits of this cholesterol is present in the human. This con- methodology it is very likely that these slight clusion is supported by the determinations of net changes are not significant. Furthermore, it is sterol balance (column N): net sterol balance certain that even if such small changes are real, = cholesterol intake (grams per day) - [total increased bile acid excretion cannot be the major neutral sterol excretion (grams per day) + bile protective device since they could only account acid excretion (grams per day)]. In every in- for a trivial portion of the ingested cholesterol. stance of both high and low intake the net sterol Several other calculations were made from these balance is negative; furthermore, the degree of data. The net dietary cholesterol excreted (col- negative balance is affected slightly if at all by umn L) was determined as follows: net dietary the quantity of cholesterol ingested. It is obvious, continues cholesterol excreted (grams per day) = neutral then, that cholesterol synthesis during intake. The net sterol balance sterol output (counts per minute per day)/dietary high cholesterol on the low cholesterol diets (- 1.70 g per day cholesterol specific activity (counts per minute in Subject A and - 0.71 g per day in Subject B) per milligram) X 1,000. This value represents should equal, furthermore, cholesterol turnover in the portion of the fecal neutral sterol that was the steady state. It is interesting that these values derived ultimately from the diet; it is a net are similar to values of cholesterol turnover rate figure and does not indicate whether the choles- in the human reported by Chobanian, Burrows, terol passed through the gastrointestinal tract and Hollander (21) by the analysis of disappear- unabsorbed or was absorbed and re-excreted. As ance curves after a single injection of cholesterol- would be predicted in the isotopic steady state 4-14C. This remarkable similarity (their mean this value in each instance closely parallels the value being 1.68 + 0.98 g per day) is even more intake. It is also possible to determine the en- striking in view of the fact that dietary choles- dogenous neutral sterol excretion under these terol was uncontrolled in their study and that conditions (column M): endogenous neutral the intake of dietary cholesterol might increase sterol excretion (grams per day) = total neutral the turnover rate. sterol excretion (grams per day) - net dietary Cholesterol turnover rate can be approximated cholesterol excreted (grams per day). This value in another way from these data (column 0): represents the amount of cholesterol synthesized cholesterol turnover (grams per day) = bile acid and excreted as neutral sterol from the body each excretion (grams per day) + endogenous neutral day; it is not the same as the total amount syn- sterol excretion (grams per day) + [percentage thesized per day since a portion of the total is of serum cholesterol derived from diet X endoge- catabolized to bile acids. However, it does repre- nous neutral sterol excretion (grams per day)]. sent the total endogenous contribution to one of This formulation adds a correction factor for the cholesterol absorbed and re-excreted into the be regarded as a close approximation; e.g., the molecular weights of cholesterol, deoxycholic acid, and cholic acid feces as neutral sterol in the steady state. The are 387, 392, and 409. values obtained from this estimation agree rea- DIETARY CHOLESTEROL IN MAN 1811

a sonably well, in the case of the low cholesterol RX diets, with the estimate from the net sterol bal- ance. In two of the three cases the estimate for I-. l1 turnover by this method is slightly higher in ,; 100_ the case of the high cholesterol diets. I, AS SUBJECT A PERIOD I (HIGH CHOLESTEROL DIET) / The calculation of cholesterol turnover on the I.izQ high intake makes possible one other approxima- iZ 75 .%' SUBJECT 8 PERIOD 2 / k, (LOW CHOLESTEROL DIET) tion from these balance figures. If, in the iso- tn -4 tu topic steady state the specific activity of serum a cholesterol represents the specific activity of the ~44 L 50 _ entire miscible pool, then a maximal value for 44 -4 44 net cholesterol absorption can be determined: 0 cholesterol absorption (grams per 16 maximal net 25 _ 0 day) = percentage of serum cholesterol derived 44 I.. from the diet X turnover rate of cholesterol ------0 44Q_I.- (grams per day). It must be emphasized that this value is a maximal value for cholesterol ab- DUODENUM JEJUNAUM ILEUM FECES sorption and cannot be assumed to be equal to FIG. 3. THE CHANGE IN SPECIFIC ACTIVITY OF DIETARY the absolute absorption rate for two reasons: CHOLESTEROL-4-14C DURING THE POSTABSORPTIVE CONDITION radioactive cholesterol can exchange across the AT VARIOUS LEVELS IN THE GASTROINTESTINAL TRACT IN gut wall without representing net movement, and THE ISOTOPIC STEADY STATE. cholesterol can be excreted and reabsorbed via an enterohepatic circulation. However, with these more, if, in addition, cholesterol not in equilibrium limitations in mind the value for maximal net with the blood is secreted from the wall of the absorption is of interest; in these two individuals gastrointestinal tract, then one would expect a the maximal absorption of dietary cholesterol was further fall of the specific activity of the lumenal 0.32 g per day (Subject A) and 0.34 g per day cholesterol in the steady state at sites distal to (Subject B). This means that at the most only the common bile duct. In the studies shown in 10 to 15%So of the ingested cholesterol was ab- Figure 3 the cholesterol content of the postab- sorbed in this study, and it is very likely that sorption gastric aspirate was too low for an the real net value is considerably less than this. accurate determination of the, specific activity; Quite clearly, this limited ability to absorb dietary at all other levels sampled in the gastrointestinal cholesterol is quantitatively the most important tract adequate samples were obtained for this of the devices that protect against dietary choles- analysis. In both studies maximal dilution of terol-induced hypercholesterolemia in the human. the specific activity was obtained in the duodenal To obtain some insight into the site of the aspirate, and this specific activity was in each equilibration of dietary cholesterol with endoge- instance almost equal to that of the serum; in nous cholesterol in the isotopic steady state, the the subject on the low cholesterol intake (Sub- specific activity of the cholesterol was determined ject B), the specific activity remained virtually at various levels of the gastrointestinal tract at unchanged throughout the remainder of the gas- the end of the first period of study in Subjects trointestinal tract; in this individual cholesterol A and B (Figure 3). It is obvious that if the turnover (0.74 g per day) enormously exceeded specific activity of the serum cholesterol never cholesterol intake (0.08 g per day), and it may equals that of the diet in the steady state, ingested be concluded that complete mixing and dilution cholesterol will be diluted with endogenous cho- occurred by the time bile mixed with the exogen- lesterol as it passes along the gastrointestinal ous sterol. This is certainly not surprising in tract. If the bile is the source for mixing en- view of the fact that in the human bile is poured dogenous and exogenous cholesterol, then a fall out from the gall bladder just at the appropriate in specific activity of dietary cholesterol would time to mix with the exogenous sources. The be expected to occur in the duodenum; further- fact that no further dilution occurs down the 1812 JEAN D. WILSON AND CHARLES A. LINDSEY, JR. gastrointestinal tract, furthermore, implies that were estimated to approximate about 0.3 g per if secretion of cholesterol not in equilibrium with day when 3 g of cholesterol was fed. These blood occurs across the intestinal wall it must results are similar to estimates obtained for maxi- be quantitatively insignificant. The situation in mal cholesterol absorption (0.3 to 0.5 g per day) the study of the specific activity along the intestine by another application of the steady state tech- of the subject on the high cholesterol intake is nique (19) and differ markedly from earlier es- slightly more complex. Again maximal dilution timates of this capacity at 1 to 3 g per day (22- occurred by the time the duodenum was reached, 24). The actual direct estimation of this ca- and this specific activity was approximately that pacity in an intact organism can best be made of the serum. However, in this study choles- by use of a double isotopic steady state (13). terol intake (2.28 g per day) grossly exceeded Although these results have been interpreted the turnover rate (1.60 g per day) and the net as constituting strong evidence that a limitation in excretion of endogenous sterol (0.53 g per day), the ability to absorb cholesterol is the major and down the gastrointestinal tract the specific means by which the human is protected from the activity rose slightly in the small intestine and development of hypercholesterolemia during high then increased significantly to approximately 90% cholesterol intake, they clearly do not mean that of the dietary cholesterol by the time the feces either an inhibition of cholesterol synthesis or an were sampled. In other words, food containing increase in bile acid excretion may not be of im- far more cholesterol than is produced each day portance in the steady state. If the human does passes through the duodenum in a wave and is indeed absorb 300 mg or less of cholesterol per diluted slightly. As it passes on through the day then, eventually, hypercholesterolemia would gastrointestinal tract the area proximal to it result unless one or both of these mechanisms approaches equilibrium with blood. Again, these were operative. The suppression of hepatic cho- data do not mean that secretion of cholesterol lesterol synthesis by the feeding of exogenous may not occur across the intestinal wall, but they cholesterol to man (10) is entirely adequate to do suggest that dilution occurring by the time compensate for this quantity of cholesterol ab- the bile is mixed with the diet is entirely suf- sorption; whether increased bile acid formation ficient to account for the drop in the specific actually occurs as a second compensatory mecha- activity along the length of the gastrointestinal nism cannot be determined at present. tract even in the postabsorptive state when dilu- The use of the isotopic steady state also per- tion of secreted cholesterol would be expected to mits an analysis of the biosynthetic origin of be minimal. serum cholesterol in man. If the liver were, in- deed, the principal source of endogenous serum Discussion cholesterol in man (25) and if the feeding of cholesterol to man causes a shutoff of hepatic cho- These studies clearly demonstrate that, of the lesterol synthesis (10), then in the steady state possible mechanisms that protect the human from the specific activity of serum should approach developing extreme hypercholesterolemia during that of the diet even when small net amounts of the excessive intake of dietary cholesterol, a lim- cholesterol are absorbed. In this study as in the ited ability to absorb dietary cholesterol is quan- experiments of Kaplan, Cox, and Taylor (19) titatively more important than inhibition of hepatic a maximum of only about 40%o of the serum cholesterol synthesis or acceleration of cholesterol cholesterol was derived from dietary sources un- degradation to bile acids. The human thus differs der circumstances of high cholesterol intake. Two in this regard from the rat; in this species serum possible explanations for this phenomenon can cholesterol is kept relatively stable even during be offered. First, the suppression of hepatic high cholesterol absorption by increasing bile synthesis may be incomplete in the intact animal; acid output (13). Furthermore, although quan- in view of the demonstration that in man (10), as titative estimates of cholesterol absorption rates well as in other species (11), the inhibition of cannot be determined from the studies reported cholesterol synthesis in liver by cholesterol feed- here, maximal values for net absorption rates ing is virtually complete, this possibility seems DIETARY CHOLESTEROL IN MAN 1813 unlikely. Second, it is clear that the only tissue biliary obstruction continued to excrete endoge- in which the negative feedback has been proven nous cholesterol in the feces (29); however, in to be operative is the liver (26), and it is pos- the normal human the cholesterol reaching the sible that other tissues such as intestine (27) feces from the intestinal wall may be in equilib- might contribute ultimately to the circulating pool. rium with that of blood. The results of the present study, along with the demonstration by Taylor and associates that cho- Summary lesterol synthesis, as indicated by the incorpora- the tion of D20, continues in the cholesterol-fed sub- The technique of analysis utilizing isotopic steady state has been applied to assess the effect ject (20), are compatible with either possibility. of varying cholesterol intakes on cholesterol syn- A third implication of this study is that puri- thesis, excretion and degradation, and absorp- fied cholesterol when fed in a soluble form has tion in man. Five balance studies were per- the same quantitative effect on endogenous cho- formed on two normal subjects who had been lesterol metabolism as does a similar amount of fed diets containing varying amounts of choles- cholesterol in a naturally occurring state such as terol-4-14C until an isotopic steady state was at- egg. Indeed, when two separate studies were tained. It has been concluded from these studies performed in the same individual similar equili- that a limited ability to absorb dietary cholesterol bration with the serum was obtained. This find- is more important in protecting the human from ing implies that the relative inability of the human hypercholesterolemia when cholesterol is fed than intestine to absorb cholesterol under the conditions is either the inhibition of cholesterol synthesis or of the present study is not due to variations in the acceleration of bile acid excretion. In addi- the lipoprotein form in which it is ingested. tion, the biosynthetic origin of serum cholesterol Whether in the human this absorptive ability has been assessed under these circumstances; ap- may be influenced by variations in the type of proximately 60 to 80% of the serum cholesterol dietary fat or carbohydrate cannot be determined is derived from endogenous sources when a diet from this study. However, the lack of any differ- high in cholesterol is fed. ence in effect on serum cholesterol level between purified and lipoprotein-bound cholesterol is Acknowledgments hardly surprising in view of the fact that cho- We wish to express our appreciation to Dr. John F. lesterol are split to free cholesterol before Fordtran for assistance in obtaining the aspirates from the cholesterol absorption (28). This conclusion gastrointestinal tract. The technical assistance of Joanne probably only applies in the case of cholesterol Sherwood and George T. Crowley in these studies is that is fed in a dissolved state since cholesterol gratefully acknowledged. fed in a crystalline, undissolved state is absorbed very poorly in most species (12) and, in addi- References tion, this lack of any difference in effect on cho- 1. Lindsey, C. A., Jr., and J. D. Wilson. Effect of lesterol metabolism may apply only to the cir- dietary cholesterol on cholesterol metabolism in cumstances in which formula diets are fed. man (abstract). Clin. Res. 1965, 13, 64. Finally, the results of this study may be inter- 2. Anitschkow, N. Experimental arteriosclerosis in animals in Arteriosclerosis, E. V. Cowdry, Ed. as no support preted furnishing for the suggestion New York, MacMillan, 1933, p. 271. that cholesterol not in equilibrium with the blood 3. Beveridge, J. M. R., W. F. Connell, H. L. Haust, and cholesterol may be secreted into the intestinal G. A. Mayer. Dietary cholesterol and plasma lumen in the human as well as in the rat (13) cholesterol levels in man. Canad. J. Biochem. in the isotopic steady state. The dilution of 1959, 37, 575. dietary cholesterol with that of bile appears to 4. Connor, W. E., R. E. Hodges, and R. E. Bleiler. of upon serum in be sufficient to of Effect dietary cholesterol explain the entire excretion man. J. Lab. clin. Med. 1961, 57, 331. endogenous sterol. It is certain that the bile is 5. Connor, W. E., R. E. Hodges, and R. E. Bleiler. The not the only endogenous source of fecal cholesterol serum lipids in men receiving high cholesterol and in the human since the human with complete cholesterol-free diets. J. clin. Invest. 1961, 40, 894. 1814 JEAN D. WILSON AND CHARLES A. LINDSEY, JR.

6. Connor, W. E., D. B. Stone, and R. E. Hodges. The 18. Taylor, C. B., D. Patton, N. Yogi, and G. E. Cox. interrelated effects of dietary cholesterol and fat Diet as source of serum cholesterol in man. Proc. upon human serum levels. J. clin. Invest. Soc. exp. Biol. (N. Y.) 1960, 103, 768. 1964, 43, 1691. 19. Kaplan, J. A., G. E. Cox, and C. B. Taylor. Choles- 7. Erickson, B. A., R. H. Coots, F. H. Mattson, and terol metabolism in man. Arch. Path. 1963, 76, A. M. Kligman. The effect of partial hydrogena- 359. tion of dietary fats, of the ratio of polyunsaturated 20. Taylor, C. B., B. Mikkelson, J. A. Anderson, D. T. to saturated fatty acids, and of dietary cholesterol Forman, and G. Franz. Human serum cholesterol upon plasma lipids in man. J. clin. Invest. 1964, synthesis measured with the deuterium label (ab- 43, 2017. stract). J. Lab. clin. Med. 1964, 64, 1009. 8. Steiner, A., E. J. Howard, and S. Akgun. Importance 21. Chobanian, A. V., B. A. Burrows, and W. Hollander. of dietary cholesterol in man. J. Amer. med. Ass. Body cholesterol metabolism in man. II. Meas- 1962, 181, 186. urement of the body cholesterol miscible pool and 9. Cook, R. P., and R. 0. Thomson. The absorption of turnover rate. J. clin. Invest. 1962, 41, 1738. fat and of cholesterol in the rat, guinea-pig, and 22. Cook, R. P., D. C. Edwards, and C. Riddell. Cho- rabbit. Quart. J. exp. Physiol. 1951, 36, 61. lesterol metabolism. Cholesterol absorption and 10. Bhattathiry, E. P. M., and M. D. Siperstein. Feed- excretion in man. Biochem. J. 1956, 62, 225. back control of cholesterol synthesis in man. J. 23. Ivy, A. C., E. Karvinen, T. M. Lin, and E. R. Ivy. clin. Invest. 1963, 42, 1613. Some parameters of sterol metabolism in man on 11. Siperstein, M. D., and M. J. Guest. Studies on the a sterol- and fat-free diet. J. appl. Physiol. 1957, site of the feedback control of cholesterol synthesis. 11, 1. clin. J. Invest. 1960, 39, 642. 24. Karvinen, E., T. M. 12. D. of Lin, and A. C. Ivy. Capacity of Wilson, J. Influence dietary cholesterol on human intestine to absorb exogenous cholesterol. excretion of cholesterol-4-C1' by the rat. Amer. J. Physiol. 1962, 202, 1073. J. appl. Physiol. 1957, 11, 143. 13. Wilson, J. D. The quantification of cholesterol ex- 25. Hotta, S., and I. L. Chaikoff. The role of the liver cretion and degradation in the isotopic steady state in the turnover of plasma cholesterol. Arch. Bio- in the rat: the influence of dietary cholesterol. J. chem. 1955, 56, 28. Lipid Res. 1964, 5, 409. 26. Dietschy, J. M., and M. D. Siperstein. Personal 14. Avigan, J., D. S. Goodman, and D. Steinberg. Thin- communication. layer chromatography of sterols and steroids. J. 27. Lindsey, C. A., Jr., and J. D. Wilson. Evidence for Lipid Res. 1963, 4, 100. a contribution by the intestinal wall to the serum 15. Fieser, L. F., and M. Fieser. Steroids. New cholesterol of the rat. J. Lipid Res. 1965, 6, 171. York, Reinhold, 1959, p. 51. 28. Swell, L., E. C. Trout, Jr., R. Hopper, H. Field, Jr., 16. Sperry, W. M., and M. Webb. A revision of the and C. R. Treadwell. The mechanism of choles- Schoenheimer-Sperry method for cholesterol de- terol absorption. Ann. N. Y. Acad. Sci. 1958- termination. J. biol. Chem. 1950, 187, 97. 1959, 72, 813. 17. Spritz, N., E. H. Ahrens, Jr., and S. Grundy. Sterol 29. Cheng, S. H., and M. M. Stanley. Secretion of cho- -balances in man as plasma cholesterol concentra- lesterol by intestinal mucosa in patients with com- tions are altered by exchange of dietary fats. J. plete common bile duct obstruction. Proc. Soc. exp. clin. Invest. 1965, 44, 1482. Biol. (N. Y.) 1959, 101, 223.