- Email: [email protected]
Measurement of Bile Acid Kinetics by Isotope Dilution in Man
GASTROENTEROLOGY 67 :314-323, 1974 Copy ri ght © 1974 by The Willi a ms & Wilki ns Co.
Vol. 67. ?'-/o . 2
Prin ted in U. S.A.
MEASUREMENT OF BILE ACID KINETICS BY ISOTOPE DILUTION IN MAN ALAN F . HOFMANN, M.D. , AND NEVILLE E. HOFFMAN, M.B. , B .S ., PH.D.
Gastro en t f!rolo f!,y Unit , May o Clinic and Mayo Foundation. Rochester, Minnesota
Measurement of bile acid synthesis rate and pool size by isotope dilution is a reliable technique if the proper isotope is used, the isotope is injected in the chemical form of the free acid, and the pool is sampled validly and for a sufficient time period. However, the accuracy of the method remains uncertain . By defining the relative synthesis rates of the two primary bile acids, one obtains the proportion of each synthesized from cholesterol-a fundamental hepatic property that is significantly altered in several disease conditions. Deoxycholic acid input from the intestine may also be defined by the isotope dilution technique, but the method gives no information on deoxycholic acid formation. Pool size represents the interaction of synthesis and intestinal conservation and may not be used to predict secretion. Specific recommendations are made, for isotope dilution studies in man, regardin g chemical form and location of label for the tracer, route and time of administration of tracer, measurement of specific activity in biliary bile acids, a nd expression of data. Measurement of bile acid synthesis rate and size of the exchangeable pool b y isotope dilution has now become a standard research technique in gastroenterology. The purpose of this article is to recall some historical highlights, to present some recommendations regarding technique, and to review the value as well as the limitations of the method. Several reviews comparing the isotope dilution method with other Received September 6, 1973 . Accepted D ecember 31, 1973 . Address requests for reprints to : Dr. Alan F. Hofmann, Mayo Clinic, Rochester, Minnesota 55901. This inv estigation was s upported in part by Researc h Gra nt AM-16770 from the N a tional Institutes of Health , Public Health Service, a nd grants -in-aid from Mead Johnson & Company , Eli Lilly and Company, the Quinn Fund, and th e Share Foundation. Dr. Nevill e Hoffman is a Fulbright-Hays Fellow. His present address is: Department of Medicin e, Univers ity of Melbourn e, Melbourn e, Australia. 314
techniques for characterizing bile acid metabolism in man have been published. 1 ' 2 History In 1953, Bergstrom et al. 3 prepared 24- 14 C-labeled bile acids which were used by Lindstedt, 4 who gave [24- 14 C]cholic acid to healthy volunteers. Bile was collected daily from the duodenum , with cholecystokinin used to contract the gallbladder. Cholic acid was isolated from hydrolyzed bile by the methods of reversed-phase partition chromatography that Sjovall 5 • 6 and Norman 7 had developed. The specific activity of the cholic acid decreased exponentially with time, and Lindstedt concluded that the bile acid pool behaved as a single well mixed compartment in a steady state. Using the paper chromatographic methods of Sjovall and Eriksson, s-u Lindstedt 4 determined biliary bile acid composition and used these data to calculate total bile acid pool size. By 1959, Bergstrom's group 12 had de-
August 1974
315
MEASUREMENT OF BILE ACID KINETICS
veloped combustion techniques for measuring 3H and had used 3 H-labeled bile acids for metabolic experiments in animals; in the early 1960's, [14 C ]cholic acid and [3 H]chenodeoxycholic acid were given to two human volunteers. 13 This showed that deoxycholic acid was derived solely from cholic acid and that lithocholic acid was derived solely from chenodeoxycholic acid, indicating that the primary bile acids have separate metabolic pathways. In addition, the data suggested different turnover rates of chenodeoxycholic acid and cholic acid, implying that any complete description of bile acid metabolism would require characterization of the metabolism of both primary bile acids. Lindstedt et al. 14 used the isotope dilu tion technique to test whether unsaturated fats increased bile acid excretion when they decreased serum cholesterol levels. In the meantime, several groups reported values for fecal excretion of bile acids, but none of these groups attempted to compare synthesis rates obtained in this manner with those obtained by isotope dilution. When Lack and Weiner 1 5 showed that the terminal ileum actively transported bile acids and that ileal resection caused profound bile acid malabsorption in the dog, 16 bile acid metabolism became a subject of interest to the gastroenterologist. Heaton et al. 17 used a modification of the Lindstedt procedure to establish the diagnosis of bile acid malabsorption in patients with ileal dysfunction. Wollenweber et al. 18 and Kottke 19 used the technique to define bile acid metabolism in patients with hyperlipidemia. However, only in the past few years has the isotope dilution technique provided new, unsuspected , information that appeared to shed new light on a common disease-cholelithiasis . Vlahcevic et al. 2 0 measured bile acid kinetics in patients with gallstones and healthy controls and showed that patients with gallstones had far smaller bile acid pools than did the control subjects. In subsequent studies, 21 they showed that chenodeoxycholic acid had a lower production rate than cholic acid in healthy man. More recently, they
have reported 22 • 23 application of the isotope dilution technique to patients with cirrhosis.
Method Principle When a tracer dose of radioactive bile acid is administered, it mixes with the bile acid pool-the pool being defined as that mass of bile acid that dilutes an injected dose of tracer. The pool may then be sampled from the duodenum after time has been allowed for adequate mixing. The following two assumptions have generally been made about the bile acid pool: (1) it behaves as a single compartment, and (2) it is in steady state (pool size is constant) . From conventional analysis of a onecompartment system in a steady state, it can be shown that SArn
=
SAro1e - kt
in which SA 11 1 and SAro1are specific activities (disintegrations per minute per mass) at time t and time 0, respectively, and k is the fractional turnover rate (time- 1). From this equation , it is clear that a plot of the natural logarithm of specific activity against time would have a slope of - k and an intercept of ln (SAo). From SA 10 1 and size of the injected dose, pool size can be calculated (pool size = dose injected/SA 10 1) and, because the enterohepatic circulation is in steady state, daily synthesis rate = pool size x fractional turnover rate Inherent in the assumption of steady state is an averaging of the events over many enterohepatic cycles. Fractional turnover rate has the unit of days - 1 and usually is measured over a period of 5 to 7 days . During each day there are six to 12 enterohepatic cycles . In this technique of isotope dilution, we avoid the influence of cycle-to-cycle variation by averaging over many cycles. This appears to be justifiable in health but may not be so in disease states such as severe bile acid malabsorption.
316
HOFMANN AND HOFFMAN
Route of Administration Accurate calculation of pool size depends on exact knowledge of the amount of radioactivity administered. Therefore, we think that the labeled bile acid should be given by intravenous injection , which avoids the uncertain absorption via the oral route. Time of Administration We think that it is rational and convenient to give the label before a meal. During digestion, the label is well mixed with the bile acid pool , so that sampling of the bile acid pool some time later ( >6 hr) gives a valid point on the specific activity decay curve. It often is convenient to administer bile acids before the evening meal and to begin bile collections the following day. Sampling the Bile Acid Pool For an estimate of t he average specific activity of the bile acid pool, one must sample a well mixed portion of the pool. This is done by recovering as much of the gallbladder contents as possible after an overnight fast. A duod enal tube is passed and vigorous gallbladder contraction is induced by intravenous administration of cholecystokinin-pancreozy min or its terminal octapeptide or by intraduodenal administration of divalent cations (such as Mg2+ or Ca 2 +) or of a glass of milk . As much bile as possible is collected and mixed well, a small sa mple (usually 1 to 2 ml) , which should contain less than 50 mg of bile acids, is saved for analysis, and the remainder is returned to the duodenum . The duodenal tube may be inserted each day or left in place for the duration of the study. In our opinion, at least five sa mples are necessary to establish a valid specific activity decay curve , and they usually are obtained over a period of 5 to 7 days. When the specific activity is d ecreasing rapidly , the samples may be obtained at shorter intervals over a 3-day period. A fairl y accurate estimate of pool size can be obtained from the decay curve constructed from multiple bile samples collected in the first 36 hr after administration of the isotope . A crude estimate of pool size can
Vol. 67, No.2
be obtained by determining a single early point on the specific activity decay curve and taking an arbitrary value for the fractional turnover rate (the slope of the curve). Despite its imprecision, such a one-point estimate of pool size would still distinguish most patients with gallstones and small pools from most healthy control subjects. The bile acid pool also may be sampled by using a capsule of dialysis tubing containing cholestyramine. 24 For 5 to 7 days a marked capsule is given each day with a meal ; the capsule traps conjugated bile acids during intestinal passage. The capsule is recovered from the feces, the bile acids are eluted and isolated chromatographically, and the specific activity of the appropriate steroid moiety is determined. The technique appears to yield data identical to t hose obtained by duodenal samplin g but with much less precision. However, it does offer a means of determinin g bile acid kin etics in children or in population studies when intubation is difficult. Its major advantage is that the capsules may be taken and the fecal samples collected at home, so that the patient merely mails the stool speci mens to the laboratory , using a convenient " fecal field ki t " re centl y described. 2 5
Choice of Tra cer Isotope. The three major hum an bile acids-cholic, chenodeoxycholic, and deoxyc holic acids-are commercially available labeled with ' 4 C in the C-24 position (the carboxyl carbon). Bile acids labeled with 3 H in the 2, 2', 4, and 4' positions are readily prepared b y enolic exchange 2 6 and [2 , 2', 4, 4'- 3 H ]cholic aci d is available commercially; the validity of isotope dilution measurements made with bile acids labeled in this manner has recently been established. 27 Choli c ac id labeled with 3 H in the 7{3 position is also easily synthesized 12 and has been used for kinetic studies in man. 28 Bile acids rand omly labeled with 3 H are also available commercially, but whether suc h bile acids give valid estimates of pool size and synthesis rate has not been studied. Recently, we have synthesized [11 , 12- 3 H ]chenode-
August 1974
MEASUREMENT OF BILE ACID KINETICS
oxycholic acid from an unsaturated precursor; this label appears to be completely stable during enterohepatic cycling in man (A. E. Cowen, A. F. Hofmann et al., unpublished observations). For the calculation of pool size and synthesis rate for a single bile acid, the 14 C label is preferable because it is more conveniently counted and its metabolic fate is more reliably followed. During intestinal passage and probably in association with dehydroxylation, 3 H is partially lost from 2, 2', 4, 4' -3H-labeled cholic or chenodeoxycholic acid, 29 precluding use of these compounds to define the biotransformation of the primary bile acids with validity. Since cholic and chenodeoxycholic acids have different turnover rates-cholic acid turnover being more rapid 13 • 21 • 30 -valid estimation of total bile acid synthesis requires that they both be labeled. Vlahcevic et al. 21 used [14 C ]cholic ncid and [14 C ]chenodeoxycholic acid simultaneously and separated the labeled chenodeoxycholic acid from labeled deoxycholic acid (the dehydroxylation product of cholic acid) by thin layer chromatography. Others have used one 3 H-labeled bile acid with the other primary bile acid labeled with 14 C. 13 • 19 • 30 This then requires only the simpler separation of trihydroxy from dihydroxy bile acids by column or thin layer chromatography followed by chemical analysis of the dihydroxy fraction. The isotope dilution technique also may be used with bile acids tagged with the stable isotopes, 13 C or 2 H. With such compounds, the decrease in atoms percent excess with time is determined by coupled gas chromatography-mass spectrometry. 31 Bile acids with 2 H at the 2,2',4,4', and 3(/3) positions are readily prepared and have been shown to give reasonably valid and precise estimates of pool size and synthesis rates when compared with simultaneous measurements carried out with 3 H-labeled bile acids. 32 13 C-labeled bile acids have been synthesized 33 but have not been used for biological studies. The use of bile acids tagged with stable isotopes for measurement of bile acid kinetics eliminates any radiation exposure, and this may permit studies that are otherwise impossi-
317
ble in children or pregnant women. In adults, the radiation exposure resulting from the use of 3 H- or 14 C-labeled bile acids in the isotope dilution technique is extremely low . Nonetheless, repeated measurements might be of value (for example, in monitoring drug effects) and could be done without radiation hazard if stable isotopes were used. Chenodeoxycholic acid labeled with 2 H at the 11 and 12 positions will soon be available commercially. However, the complex instrumentation required for measurement of stable isotope ratios in bile acid samples is still available in very few laboratories. In time, the development of improved instrumentation together with the availability of 2 H-labeled bile acids should permit extensive application of the isotope dilution technique in large patient groups; population studies appear feasible. Chemical form. Lindstedt 4 administered bile acid in the unconjugated form and determined the specific activity in that bile acid after alkaline saponification. Others have used labeled cholyltaurine or cholylglycine and determined the specific activity decay of that conjugate. These labeled species do not have the same fate in the body after administration. Cholic acid is conjugated with both glycine and taurine, and the fate of the label will be that of these two conjugated bile acids. Cholylglycine is hydrolyzed rather rapidly (daily fractional turnover rate ~ 1.0) and the cholic acid liberated is reconjugated, predominantly with glycine; cholyltaurine is hydrolyzed more slowly (daily fractional turnover rate ~ 0.4), and again most of the liberated cholic acid will be reconjugated with glycine. To deal with these complexities , we recently developed a multicompartmental model of bile acid metabolism, shown in part in figure 1, and used this to test the error when the isotope dilution technique was carried out with administration of different labeled bile acids and determination of the specific activity decay curve of that bile acid. 34 The Lindstedt procedure using free bile acid appears to be valid; the error associated with use of the labeled glycine conjugate is potentially larger; but
318
HOFMANN AND HOFFMAN Input 2
1
Cholylglycine
Free cholic acid
k13
3
Cholyltaurine
k31 ko3
FIG. 1. Multicompartmental model of metabolism of conjugated cholic ac id in ma n . Conjugat ion is indicat ed by the rate consta n ts, k,, and k ,; return to the liv er of free bil e acid form ed by bac terial deconjugati on is indicated by the constants, k, , and k ., . Free cholic acid pool refers to hypothetical intracellul a r precursor pool for bile acid conjugat ion in which bile ac ids both synthes ized endo genously and return ed are uniformly mixed. The transfer constants, k 02 and k 03 , represent loss from the system by either exc retion , chemical transformation ( su e ~ as dehydroxylat ion), or phys ical transformat ion (such as precipitation).
the potential error associated with use of the labeled taurine conjugate is very large if specific activity is estimated in the taurine conjugate. Measurement of Specific Activity
The adaptation of Talalay's 3-hydroxysteroid dehydrogenase assay of steroids to bile acids by Iwata and Yamasaki 35 was of immense importance. Thus , cholic acid can be isolated from hydrolyzed bile by thin layer chromatography and eluted from the adsorbent . One aliquot can be counted for radioactivity and a second aliquot can be analyzed by the steroid dehydrogenase method. This technique cannot be used for chenodeoxycholic acid unless it is separated from deoxycholic acid. However, Haslewood et a!. 36 recently described the isolation and use of a microbial 7a-hydroxysteroid dehydrogenase that is specific for 7a-hydroxylated bile acids. This method should permit direct assay of the chenodeoxycholic mass in the dihydroxy fraction. Specific colorimetric methods for chenodeoxycholic 37 or deoxycholic 3 8 acid measurement have been described but are used infrequently, and nonspecific spectrophotometric or fluorometric procedures are used with diminishing frequency. The bile acids also may be estimated by gas-liquid
Vol. 67, No. 2
chromatography with an internal standard. Hyodeoxycholic, nordeoxycholic, and hyocholic acids have been used as internal standards, but the last may be difficult to derivatize completely. 39 Hydroxyketo bile acids also have been used , but this procedure seems hazardous because these compounds have been identified in samples of bile from patients with gallstones . 4 0 Generally, bile acids are chromatographed as trifluoroacetates on QF-1 columns 41 or as acetates on QF-1 or AN-600 columns 4 2 ; trimethylsilyl ethers are well separated on HiEfffiB columns. 43 Good quantification of underivatized methyl esters has proved difficult to achieve for most workers . Expression of Data The primary data derived from regression analysis of the specific activity decay curve are pool size and daily fractional turnover rate. The product of these two is the daily synthesis rate. The half-life of label Ctv, ) is commonly reported in place of k, the fractional turnover rate. There is no point in reporting both since t ~~, = In 2/k. For a pool that remains constant in size, we prefer to use k. Some authors report pool size and daily synthesis in units of mass; others report molar units. We think that the latter is preferable, particularly if free and conjugated bile acids are to be compared. One further problem is the normalization of data. Should data be expressed in terms of actual body weight, ideal body weight, body surface area, or some other variable? Because the purpose of normalization is to decrease the variance within groups believed not to vary with respect to the factor under investigation , the correct method of normalization must be arrived at experimentally. Currently, insufficient data are available to allow clear choice . We recommend that authors publish bile acid kinetics in comparable terms , including data in micromoles per kilogram of body weight or micromoles per square meter of body surface area . Recommendations for calculating body surface area from height and weight have recently been published. 44
August 1974
MEASUREMENT OF BILE ACID KINETICS
Reports should include-or archive-full patient data (age, sex, weight, and height); this practice will permit the appropriate method of normalization to become apparent in time. Validation and Interpretation Validation of Pool Size The pool of bile acid was operationally defined as that mass of bile acids that dilutes a tracer. This should be equal approximately to the mass of bile acids drained by an acute bile fistula, but recent experiments of Mack and Dowling (personal communication) suggest that the mass of bile acid that dilutes a tracer is considerably larger than the mass of bile acids recovered from an acute bile fistula, at least in the rhesus monkey. If the estimation of pool size is valid, then estimation of cholic, deoxycholic, and chenodeoxycholic acid pool sizes by the isotope dilution technique in the same subject should produce pool size values in the same proportion as estimated from class distribution of bile acids by gas-liquid chromatography, and this has been proved true 45 (also, T. S. Low-Beer, personal communication). Validation of Synthesis Rate Synthesis can be measured by chemical assay of fecal bile acids (balance technique) because, in the steady state, fecal excretion is equal to synthesis. We can find no published reports in which isotope dilution measurements have been directly compared with fecal excretion measurements in the same patient, although L. Swell and Z. R. Vlahcevic (personal communication) report good agreement between the two methods. Miettinen 2 observed that synthesis values obtained by isotope dilution tend to be higher than those estimated by the balance technique. At present, it seems likely that the isotope dilution measurements give fairly valid but quite reliable estimates of bile acid synthesis. If it is shown that such measurements include a systematic error, the method will still probably be preferred to the balance technique because of the
319
complexity of the analytical methods and the uncertainties of sampling in the balance technique. 46 In patients with severe bile acid malabsorption, the isotope dilution technique cannot be assumed to be valid, and here measurement of fecal bile acids is the only correct method for estimation of bile acid synthesis.' Whether the one-compartment model is valid in patients with hepatobiliary obstruction is uncertain and should be tested by comparison with balance studies. Balance studies in such patients should measure both urinary and fecal bile acid excretion. 23 Relationship of Bile Acid Pool Size to Bile Acid Secretion The output of bile acids into the duodenum is defined as secretion and was the major measurement made by physiologists during the past century. With the development of isotope dilution techniques, secretion studies languished, especially because it is difficult to measure secretion in man without perturbing the enterohepatic circulation. Thureborn 47 and Isaksson and Thureborn, 48 in man, and Dowling et al., 49 in the monkey, developed methods for controlled interruption of the enterohepatic circulation; the latter used this technique to define the relationship between return to the liver and bile acid synthesis in a number of fundamental papers. 5°' 51 Because secretion is a product of pool size and recycling frequency and because pool size was easily measured, some workers assumed that it was reasonable to estimate secretion by multiplying pool size by a constant recycling frequency. To paraphrase, secretion was considered to be directly proportional to pool size. During this past year, biliary secretion has been measured directly, by duodenal perfusion techniques, by Northfield and Hofmann. 52 In studies carried out on a group of patients with various pool sizes, they found that recycling frequency was inversely proportional to pool size. Thus, when normalized for body weight, secretion was constant despite widely different pool sizes.
320
HOFMANN AND HOFFMAN
Based on Northfield and Hofmann's data, one is tempted to coin a "bile's law" (analogous to Boyle's gas law): CP
C'P'
=
in which C is recycling frequency and Pis pool size. This idea needs further experimental verification but has been argued from other lines of evidence by Low-Beer and Pomare. 53 Perhaps students of the enterohepatic circulation might have done well to discuss this problem with cardiologists, who never have attempted to predict cardiac output from blood volume. The functions of bile acids in the small intestine are more closely related to bile acid secretion than to pool size. Measurements of secretion would appear to have an important place in the characterization of bile acid metabolism in digestive disease or of the mode of action of agents that alter biliary lipid secretion. Value of Synthesis Measurements Bile acid synthesis data represent the most valuable information afforded by the isotope dilution technique, in our judg-
Vol. 67, No.2
ment. The measurement of greatest potential interest is the rate of cholic acid synthesis relative to chenodeoxycholic acid synthesis; this appears to be strikingly altered in a number of metabolic diseases (fig. 2). If labeled deoxycholic acid is injected, its
specific activity decay curve also decreases exponentially. 45 Thus, the deoxycholic acid pool behaves as a single compartment. The fractional turnover rate of deoxycholic acid seems to be identical to that of chenodeoxycholic acid, indicating similar intestinal conservation 45 • 54 ; few data are available, however. The origin of deoxycholic acid is bacterial 7a-dehydroxylation of cholic acid. In health, the deoxycholic acid input is about one-third of the cholic acid excretion (or synthesis). Thus, less than half of the deoxycholic acid that could be formed is absorbed from the intestine. One can define a fraction, fctehydrox , as the observed input of deoxycholic acid divided by the maximal amount of deoxycholic acid that could be formed. This is simply input of deoxycholic acid divided by synthesis of cholic acid. There is no meaningful way to estimate
1.0 8
0.8 0.6
XC f!j
X •
~ ·~ .. x
4§ 0 4
D D
_a.E.,_ D
8
4 4.44
~0
D
~
0.4
. .t
40c:P
--;:r-
•
• --•
4
440
g
0.2 0 Normal
Cholelithiasis
TypeTI hyperlipidemia
TypeN Cerebrohyper- tendinous xantholipidemia matosis
FIG. 2. Mole fraction of cholic acid in bile acid synthesis (cholic acid synthesis/total bile acid synthesis) in health and several disease states. Symbols indicate source of data: e, Vlahcevic et al. 21 ; x, Hepner et al. "· "; 0 • , Danzinger et al. 30 ; !;,., Kottke 19 ; and 0 , Einarsson and Hellstrom." In the studies by Danzinger et al.' 0 , kinetics were measured in the same patient before (D) and after 6 months . Data from patients with cerebrotendinous xanthomatosis ( A) were generously provided by Dr. Gerald Salen (unpublished observations). Horizontal lines indicate group means .
<•l
August 1974
MEASUREMENT OF BILE ACID KINETICS
actual deoxycholic acid formation, because it could occur in the rectum or after defecation. The value of fctehyctrax would approach 1 in those individuals in whom deoxycholic acid is the major bile acid but would be near 0 in individuals who, although making deoxycholic acid, have little in their bile. Patients with cirrhosis or patients ingesting cholestyramine are probably good examples of the latter. To measure fctehydrax , one should carry out conventional isotope dilution techniques using labeled cholic and deoxycholic acids. As previously noted, the decay of specific activity is due to the input of unlabeled bile acid. During bile acid feeding, the input will be the sum of synthesis and absorbed fed bile acid. As long as the fed dose is constant and the rate of absorption is approximately constant, a steady state will obtain; the specific activity decay will be exponential, and tlie usual kinetic analysis is appropriate. 30 If the exogenous dose is large relative to endogenous synthesis, then an estimate of the fractional absorption of the fed dose may be calculated.
The Future Measurement of bile acid synthesis rate and pool size by isotope dilution techniques appears to have come of age at last. The measurement of synthesis rate, especially of both primary bile acids, appears to be of great value. The value of pool size measurements is much less certain. Indeed, if recycling frequency and pool size are inversely related, then estimation of pool size may be of value chiefly for estimating recycling frequency which in turn may give a clue to altered gallbladder function. If the major value of the measurement of bile acid kinetics is estimation of primary bile acid synthesis, then the major use of bile acid kinetics will be in the area of intermediary metabolism. On the other hand, the role of bile acids in the pathogenesis of many hepatic and intestinal diseases is poorly understood and may be pivotal; measurement of bile acid kinetics is an essential part of complete characterization of bile acid metabolism. From this standpoint, we can expect that the determination of bile acid kinetics by isotope
321
dilution techniques will continue to have an important place among investigative procedures in gastroenterology. REFERENCES 1. Hofmann AF, Schoenfield W, Kottke BA, et al: Methods for the description of bile acid kinetics in man. Methods Med Res 12:149- 180, 1970 2. Miettinen TA: Clinical implications of bile acid metabolism in man, The Bile Acids: Chemistry, Physiology, and Metabolism, vol 2. Edited by PP Nair, D Kritchevsky. New York, Plenum Press, 1973, p 191-247 3. Bergstrom S, Rottenberg M, Voltz J: The preparation of some carboxylabelled bile acids: bile acids and steroids 2. Acta Chern Scand 7:481- 484, 1953 4. Lindstedt S: The turnover of cholic acid in man : bile acids and steroids 51. Acta Physiol Scand 40:1- 9, 1957 5. Sjovall J: Separation of bile acids by paper chromatography: bile acids and steroids 3. Acta Chern Scand 6:1552- 1553, 1952 6. Sjovall J: On the separation of bile acids by partition chromatography : bile acids and steroids 4. Acta Physiol Scand 29:232-,240, 1953 7. Norman A: Separation of conjugated bile acids by partition chromatography: bile acids and steroids 6. Acta Chern Scand 7:1413-1419, 1953 8. Sjovall J: Separation of conjugated and free bile acids by paper chromatography: bile acids and steroids 12. Acta Chern Scand 8:339-345, 1954 9. Eriksson S, Sjovall J: The absorption spectra of bile acids in sulfuric acid: bile acids and steroids 31. Arkh Kemi 8:303-310, 1955 10. Eriksson S, Sjovall J : The absorption spectra of conjugated bile acids in sulfuric acid: bile acids and steroids 32. Arkh Kemi 8:311-315, 1955 11. Sjovall J: Quantitative determination of bile acids on paper chromatograms: bile acids and steroids 33 . Arkh Kemi 8:317-324, 1955 12. Bergstrom S, Lindstedt S, Samuelsson B: Bile acids and steroids LXXXII: on the mechanism of deoxycholic acid formation in the rabbit. J Bioi Chern 234:2022-2025, 1959 13. Danielsson H, Eneroth P, Hellstrom K, et al: On the turnover and excretory products of cholic and chenodeoxycholic acid in man: bile acids and steroids 134. J Bioi Chern 238:2299-2304, 1963 14. Lindstedt S, Avigan J, Goodman DS, et al: The effect of dietary fat on the turnover of cholic acid and on the composition of the biliary bile acids in man . J Clin Invest 44:1754- 1765, 1965 15. Lack L, Weiner IM : In vitro absorption of bile salts by small intestine of rats and guinea pigs. Am J Physiol 200:313-317, 1961 16. Playoust MR, Lack L, Weiner IM: Effect of
322
17.
18.
19.
20.
21.
22.
23 .
24.
25.
26.
27 .
28. 29.
30 .
31.
HOFMANN AND HOFFMAN
intestinal resection on bile salt absorption in dogs. Am J Physiol 208:363-369, 1965 H eaton KW, Austad WI, Lack L, et al: Enterohepatic circulation of C " -la beled bil e salts in disorders of the distal s m a ll bowel. Gastroenterology 55 :5- 16, 1968 W ollenweber J , Kottke BA, Owen CA Jr: P ool size and turnover of bile aci ds in six hypercholesteremic patients with an d without adm inistration of nicotinic acid. J Lab Clin Med 69:584-593, 1967 Kottke BA: Differences in bile acid excretion: primary hypercholesterolemia compared to combined hy percholeste rolemia and hy pertriglyceridemia. Circul ation 40:13-20, 1969 Vla hcevic ZR, Bell CC Jr, Buhac I, et al: Diminish ed bile acid pool size in patients wi t h gallstones. Gastroenterology 59: 165-173, 1970 Vlahcevic ZR, Mill er JR, F arra r JT, et al: Kin etics and pool size of primary bile aci ds in man. Gastroenterology 61:85- 90, 1971 Vl a hcevic ZR, Buhac I, Farrar JT, et al: Bile acid metabolism in patients with cirrhosis. I. Kinet ic aspects of cholic acid metabolism. Gastroenterology 60: 491- 498, 1971 Vlahcevic ZR, Juttijudata P, Bell CC Jr, et al: Bile acid metabolism in p atients with cirrhosis. II . C holic and chenodeoxyc holic ac id metabolism. Gastroenterology 62:1174-1181, 1972 H offm an NE , LaRusso N F , Hofm ann AF: Noninvasive determination of bile aci d kinetics using a sequesteri ng capsule (abstr). Gastroe nt erology 64:744, 1973 H offman NE , LaRusso NF, Hofm ann AF: An improved met hod for faecal collection: the faeca l field-kit. Lancet 1:1422- 1423, 1973 H ofmann AF, Szczepanik PA , Klein PD : Ra pid preparation of tritium -labeled bile acids by enolic exchan ge on basic alumina containin g tritiated water. J Lipid Res 9:707- 713, 1968 LaRusso NF, Hoffm an NE, Hofmann AF: Validity of using 2, 4-' H-labe led bile acids to study bile acid kin etics in man . J Lab Clin Med (in press) Stahl E, Amesjo B: Ta uroc holate metabolism in man . Scand J Gastroenterol 7:559-566, 1972 H epner GW, Sturman JA , Hofman n AF, et al: M etabolism of steroid an d am ino acid moieties of conjugated bile acids in man . III. Choly ltaurine (ta urocholic acid) . J Clin Invest 52:433-440, 1973 D anzin ger RG, Hofmann AF, Thistle JL, et al: Effect of oral chenodeoxych olic acid on bil e acid kinetics and biliary lipid composition in women with cholelithiasis. J Clin Invest 52:2809-2821, 1973 Klein PD, Haumann JR, Eisler WJ: Gas chromatograph-mass spectrometer-acceleratin g voltage alternator system for t he m easurement of
32.
33.
34.
35.
36.
37.
38.
39 .
40.
41.
42.
43.
44.
45 .
Vol . 67, No.2
stable isotope ratios in organic molecules . Anal Chern 44:490- 493, 1972 H ofmann AF, Dan zin ger RG, Hoffman NE, et al: Validation an d comparison of deuterium/tritium and carbon-1 3/carbon-14 in clinical studies of bile ac id metabolism (USAEC conf 711115), Proceedings of a Seminar on the Use of Stable Isotopes in Clinical Pharmacology . Edited by PD Kl ein , LJ Roth. Oak Ridge, United States Atomic Energy Commission, 1972, p 37-51 Hachey DL, Szczepanik PA, Berngruber OW, et al: Syntheses of stable isotopes: synthesis of deuterium and "C labeled bile acids. J Labelled Compounds 9:703-719, 1973 Hoffman NE, Hofmann AF: M etabolism of steroid and amino acid moieties of conjugated bile acids in man. IV. Desc ription and validation of a multicompartmental model. Gastroenterology (in press) Iwata T, Yamasaki K: En zy matic determination and thin-layer chromatography of bile ac ids in blood. J Bioc hem (T okyo) 56:424- 431, 1964 Haslewood GAD, Murphy GM, Richardson JM: A direct enzymic assay for ?a -hydroxy bile acids and their conjugates. Clin Sci 44:95-98, 1973 Isaksson B: A method for spectrophoto metric determination of chenodesoxyc holic aci d in bile. Acta Chern Scand 8:889- 897, 1954 Bruusgaard A: Quantitative determination of the major 3-hydroxy bile acids in biological material after thin-layer chromatographic separation. Clin Chim Acta 28:495-504, 1970 Mott GE, Moore RW, Reiser R: Gas-liquid chromatography of hyocholic acid. J Lipid Res 12:117- 119, 1971 Hepner GW, Hofmann AF, Mal agelada JR, et al: Increased bacterial degradation of bile acids in chol ecystect omize d patients. Gastroenterology 66:556-564, 1974 Eneroth P, Sjovall J: E xtraction, purification, and chromatographic analysis of bile ac ids in biological materials, The Bile Acids: Chemistry, Phys iology, and Metabolis m, voll. Edited by PP Nair, D Kritchevsky . New York, Plenum Press, 1971, p 121-171 Roovers J , Evrard E , Vanderhaeghe H : An improved met hod for measuring human blood bile acids . Clin Chim Acta 19:449-457, 1968 Makita M, Wells WW: Quantitative analysis of fecal bile acids by gas-liquid chrom atography. Anal Biochem 5:523-530, 1963 Gehan EA, George SL : Estimation of human body surfa ce area from height and weight . Cancer Chemother Rep 54:225-235, 1970 Hepner GW, Hofm ann AF, Thomas PJ: Metabolism of steroid and a mino acid moieties of conjugated bile ac ids in man . II. Glycine-conjugated
August 1974
46.
47.
48.
49.
50.
MEASUREMENT OF BILE ACID KINETICS
dihydroxy bile acids. J Clin Invest 51:1898- 1905, 1972 Grundy SM, Ahrens EH Jr, Miettinen TA: Quantitative isolation and gas-liquid chromatographic analysis of total fecal bile acids . J Lipid Res 6:397- 410, 1965 Thureborn E: Human hepatic bile: composition changes due to a ltered enterohepatic circulation. Acta Chir Scand [Suppl] 303:1- 63, 1962 Isaksson B, Thureborn E : En metod for uppsamling av hepaticusgalla vid intakt enterohepatiskt kretslopp (abstr). Nord Med 66:1519, 1961 Dowling RH, MackE , Picott J, et al: Experimental model for the study of the enterohepatic circulation of bile in rhesus monkeys . J Lab Clin Med 72:169-176, 1968 Dowling RH, Mack E, Small DM: Effects of controlled interruption of the enterohepatic circulation of bile salts by biliary diversion and by ileal resection on bile salt secretion, synthesis, and pool size in the rhesus monkey. J Clin Invest
323
49:232-242, 1970 51. Dowling RH, Mack E , Small DM: Biliary lipid secretion and bile co mposition after acute and chronic interruption of the enterohepatic circulation in the rhesus monkey. IV. Primate biliary physiology. J Clin Invest 50:1917- 1926, 1971 52. Northfield TC, Hofmann AF: Biliary lipid secretion in gallstone patients . Lancet 1:747-748, 1973 53. Low-Beer TS, Po mare EW: Regulation of bile salt pool size in man . Br Med J 2:338-340, 1973 54. Quarfordt SH, Greenfield MF: Estimation of cholesterol and bile acid turnover in man by kinetic analysis. J Clin Invest 52:1937- 1945, 1973 55. Hepner GW, Hofmann AF, Thomas PJ : Metabolism of steroid and amino acid moieties of conjugated bile acids in man. I. Cholylglycine. J Clin Invest 51:1889-1897 , 1972 56. Einarsson K, Hellstrom K: The formation of bile acids in patients with three types of hyperlipoproteinaemia. Eur J Clin Invest 2:225- 230, 1972
Vol. 67. ?'-/o . 2
Prin ted in U. S.A.
MEASUREMENT OF BILE ACID KINETICS BY ISOTOPE DILUTION IN MAN ALAN F . HOFMANN, M.D. , AND NEVILLE E. HOFFMAN, M.B. , B .S ., PH.D.
Gastro en t f!rolo f!,y Unit , May o Clinic and Mayo Foundation. Rochester, Minnesota
Measurement of bile acid synthesis rate and pool size by isotope dilution is a reliable technique if the proper isotope is used, the isotope is injected in the chemical form of the free acid, and the pool is sampled validly and for a sufficient time period. However, the accuracy of the method remains uncertain . By defining the relative synthesis rates of the two primary bile acids, one obtains the proportion of each synthesized from cholesterol-a fundamental hepatic property that is significantly altered in several disease conditions. Deoxycholic acid input from the intestine may also be defined by the isotope dilution technique, but the method gives no information on deoxycholic acid formation. Pool size represents the interaction of synthesis and intestinal conservation and may not be used to predict secretion. Specific recommendations are made, for isotope dilution studies in man, regardin g chemical form and location of label for the tracer, route and time of administration of tracer, measurement of specific activity in biliary bile acids, a nd expression of data. Measurement of bile acid synthesis rate and size of the exchangeable pool b y isotope dilution has now become a standard research technique in gastroenterology. The purpose of this article is to recall some historical highlights, to present some recommendations regarding technique, and to review the value as well as the limitations of the method. Several reviews comparing the isotope dilution method with other Received September 6, 1973 . Accepted D ecember 31, 1973 . Address requests for reprints to : Dr. Alan F. Hofmann, Mayo Clinic, Rochester, Minnesota 55901. This inv estigation was s upported in part by Researc h Gra nt AM-16770 from the N a tional Institutes of Health , Public Health Service, a nd grants -in-aid from Mead Johnson & Company , Eli Lilly and Company, the Quinn Fund, and th e Share Foundation. Dr. Nevill e Hoffman is a Fulbright-Hays Fellow. His present address is: Department of Medicin e, Univers ity of Melbourn e, Melbourn e, Australia. 314
techniques for characterizing bile acid metabolism in man have been published. 1 ' 2 History In 1953, Bergstrom et al. 3 prepared 24- 14 C-labeled bile acids which were used by Lindstedt, 4 who gave [24- 14 C]cholic acid to healthy volunteers. Bile was collected daily from the duodenum , with cholecystokinin used to contract the gallbladder. Cholic acid was isolated from hydrolyzed bile by the methods of reversed-phase partition chromatography that Sjovall 5 • 6 and Norman 7 had developed. The specific activity of the cholic acid decreased exponentially with time, and Lindstedt concluded that the bile acid pool behaved as a single well mixed compartment in a steady state. Using the paper chromatographic methods of Sjovall and Eriksson, s-u Lindstedt 4 determined biliary bile acid composition and used these data to calculate total bile acid pool size. By 1959, Bergstrom's group 12 had de-
August 1974
315
MEASUREMENT OF BILE ACID KINETICS
veloped combustion techniques for measuring 3H and had used 3 H-labeled bile acids for metabolic experiments in animals; in the early 1960's, [14 C ]cholic acid and [3 H]chenodeoxycholic acid were given to two human volunteers. 13 This showed that deoxycholic acid was derived solely from cholic acid and that lithocholic acid was derived solely from chenodeoxycholic acid, indicating that the primary bile acids have separate metabolic pathways. In addition, the data suggested different turnover rates of chenodeoxycholic acid and cholic acid, implying that any complete description of bile acid metabolism would require characterization of the metabolism of both primary bile acids. Lindstedt et al. 14 used the isotope dilu tion technique to test whether unsaturated fats increased bile acid excretion when they decreased serum cholesterol levels. In the meantime, several groups reported values for fecal excretion of bile acids, but none of these groups attempted to compare synthesis rates obtained in this manner with those obtained by isotope dilution. When Lack and Weiner 1 5 showed that the terminal ileum actively transported bile acids and that ileal resection caused profound bile acid malabsorption in the dog, 16 bile acid metabolism became a subject of interest to the gastroenterologist. Heaton et al. 17 used a modification of the Lindstedt procedure to establish the diagnosis of bile acid malabsorption in patients with ileal dysfunction. Wollenweber et al. 18 and Kottke 19 used the technique to define bile acid metabolism in patients with hyperlipidemia. However, only in the past few years has the isotope dilution technique provided new, unsuspected , information that appeared to shed new light on a common disease-cholelithiasis . Vlahcevic et al. 2 0 measured bile acid kinetics in patients with gallstones and healthy controls and showed that patients with gallstones had far smaller bile acid pools than did the control subjects. In subsequent studies, 21 they showed that chenodeoxycholic acid had a lower production rate than cholic acid in healthy man. More recently, they
have reported 22 • 23 application of the isotope dilution technique to patients with cirrhosis.
Method Principle When a tracer dose of radioactive bile acid is administered, it mixes with the bile acid pool-the pool being defined as that mass of bile acid that dilutes an injected dose of tracer. The pool may then be sampled from the duodenum after time has been allowed for adequate mixing. The following two assumptions have generally been made about the bile acid pool: (1) it behaves as a single compartment, and (2) it is in steady state (pool size is constant) . From conventional analysis of a onecompartment system in a steady state, it can be shown that SArn
=
SAro1e - kt
in which SA 11 1 and SAro1are specific activities (disintegrations per minute per mass) at time t and time 0, respectively, and k is the fractional turnover rate (time- 1). From this equation , it is clear that a plot of the natural logarithm of specific activity against time would have a slope of - k and an intercept of ln (SAo). From SA 10 1 and size of the injected dose, pool size can be calculated (pool size = dose injected/SA 10 1) and, because the enterohepatic circulation is in steady state, daily synthesis rate = pool size x fractional turnover rate Inherent in the assumption of steady state is an averaging of the events over many enterohepatic cycles. Fractional turnover rate has the unit of days - 1 and usually is measured over a period of 5 to 7 days . During each day there are six to 12 enterohepatic cycles . In this technique of isotope dilution, we avoid the influence of cycle-to-cycle variation by averaging over many cycles. This appears to be justifiable in health but may not be so in disease states such as severe bile acid malabsorption.
316
HOFMANN AND HOFFMAN
Route of Administration Accurate calculation of pool size depends on exact knowledge of the amount of radioactivity administered. Therefore, we think that the labeled bile acid should be given by intravenous injection , which avoids the uncertain absorption via the oral route. Time of Administration We think that it is rational and convenient to give the label before a meal. During digestion, the label is well mixed with the bile acid pool , so that sampling of the bile acid pool some time later ( >6 hr) gives a valid point on the specific activity decay curve. It often is convenient to administer bile acids before the evening meal and to begin bile collections the following day. Sampling the Bile Acid Pool For an estimate of t he average specific activity of the bile acid pool, one must sample a well mixed portion of the pool. This is done by recovering as much of the gallbladder contents as possible after an overnight fast. A duod enal tube is passed and vigorous gallbladder contraction is induced by intravenous administration of cholecystokinin-pancreozy min or its terminal octapeptide or by intraduodenal administration of divalent cations (such as Mg2+ or Ca 2 +) or of a glass of milk . As much bile as possible is collected and mixed well, a small sa mple (usually 1 to 2 ml) , which should contain less than 50 mg of bile acids, is saved for analysis, and the remainder is returned to the duodenum . The duodenal tube may be inserted each day or left in place for the duration of the study. In our opinion, at least five sa mples are necessary to establish a valid specific activity decay curve , and they usually are obtained over a period of 5 to 7 days. When the specific activity is d ecreasing rapidly , the samples may be obtained at shorter intervals over a 3-day period. A fairl y accurate estimate of pool size can be obtained from the decay curve constructed from multiple bile samples collected in the first 36 hr after administration of the isotope . A crude estimate of pool size can
Vol. 67, No.2
be obtained by determining a single early point on the specific activity decay curve and taking an arbitrary value for the fractional turnover rate (the slope of the curve). Despite its imprecision, such a one-point estimate of pool size would still distinguish most patients with gallstones and small pools from most healthy control subjects. The bile acid pool also may be sampled by using a capsule of dialysis tubing containing cholestyramine. 24 For 5 to 7 days a marked capsule is given each day with a meal ; the capsule traps conjugated bile acids during intestinal passage. The capsule is recovered from the feces, the bile acids are eluted and isolated chromatographically, and the specific activity of the appropriate steroid moiety is determined. The technique appears to yield data identical to t hose obtained by duodenal samplin g but with much less precision. However, it does offer a means of determinin g bile acid kin etics in children or in population studies when intubation is difficult. Its major advantage is that the capsules may be taken and the fecal samples collected at home, so that the patient merely mails the stool speci mens to the laboratory , using a convenient " fecal field ki t " re centl y described. 2 5
Choice of Tra cer Isotope. The three major hum an bile acids-cholic, chenodeoxycholic, and deoxyc holic acids-are commercially available labeled with ' 4 C in the C-24 position (the carboxyl carbon). Bile acids labeled with 3 H in the 2, 2', 4, and 4' positions are readily prepared b y enolic exchange 2 6 and [2 , 2', 4, 4'- 3 H ]cholic aci d is available commercially; the validity of isotope dilution measurements made with bile acids labeled in this manner has recently been established. 27 Choli c ac id labeled with 3 H in the 7{3 position is also easily synthesized 12 and has been used for kinetic studies in man. 28 Bile acids rand omly labeled with 3 H are also available commercially, but whether suc h bile acids give valid estimates of pool size and synthesis rate has not been studied. Recently, we have synthesized [11 , 12- 3 H ]chenode-
August 1974
MEASUREMENT OF BILE ACID KINETICS
oxycholic acid from an unsaturated precursor; this label appears to be completely stable during enterohepatic cycling in man (A. E. Cowen, A. F. Hofmann et al., unpublished observations). For the calculation of pool size and synthesis rate for a single bile acid, the 14 C label is preferable because it is more conveniently counted and its metabolic fate is more reliably followed. During intestinal passage and probably in association with dehydroxylation, 3 H is partially lost from 2, 2', 4, 4' -3H-labeled cholic or chenodeoxycholic acid, 29 precluding use of these compounds to define the biotransformation of the primary bile acids with validity. Since cholic and chenodeoxycholic acids have different turnover rates-cholic acid turnover being more rapid 13 • 21 • 30 -valid estimation of total bile acid synthesis requires that they both be labeled. Vlahcevic et al. 21 used [14 C ]cholic ncid and [14 C ]chenodeoxycholic acid simultaneously and separated the labeled chenodeoxycholic acid from labeled deoxycholic acid (the dehydroxylation product of cholic acid) by thin layer chromatography. Others have used one 3 H-labeled bile acid with the other primary bile acid labeled with 14 C. 13 • 19 • 30 This then requires only the simpler separation of trihydroxy from dihydroxy bile acids by column or thin layer chromatography followed by chemical analysis of the dihydroxy fraction. The isotope dilution technique also may be used with bile acids tagged with the stable isotopes, 13 C or 2 H. With such compounds, the decrease in atoms percent excess with time is determined by coupled gas chromatography-mass spectrometry. 31 Bile acids with 2 H at the 2,2',4,4', and 3(/3) positions are readily prepared and have been shown to give reasonably valid and precise estimates of pool size and synthesis rates when compared with simultaneous measurements carried out with 3 H-labeled bile acids. 32 13 C-labeled bile acids have been synthesized 33 but have not been used for biological studies. The use of bile acids tagged with stable isotopes for measurement of bile acid kinetics eliminates any radiation exposure, and this may permit studies that are otherwise impossi-
317
ble in children or pregnant women. In adults, the radiation exposure resulting from the use of 3 H- or 14 C-labeled bile acids in the isotope dilution technique is extremely low . Nonetheless, repeated measurements might be of value (for example, in monitoring drug effects) and could be done without radiation hazard if stable isotopes were used. Chenodeoxycholic acid labeled with 2 H at the 11 and 12 positions will soon be available commercially. However, the complex instrumentation required for measurement of stable isotope ratios in bile acid samples is still available in very few laboratories. In time, the development of improved instrumentation together with the availability of 2 H-labeled bile acids should permit extensive application of the isotope dilution technique in large patient groups; population studies appear feasible. Chemical form. Lindstedt 4 administered bile acid in the unconjugated form and determined the specific activity in that bile acid after alkaline saponification. Others have used labeled cholyltaurine or cholylglycine and determined the specific activity decay of that conjugate. These labeled species do not have the same fate in the body after administration. Cholic acid is conjugated with both glycine and taurine, and the fate of the label will be that of these two conjugated bile acids. Cholylglycine is hydrolyzed rather rapidly (daily fractional turnover rate ~ 1.0) and the cholic acid liberated is reconjugated, predominantly with glycine; cholyltaurine is hydrolyzed more slowly (daily fractional turnover rate ~ 0.4), and again most of the liberated cholic acid will be reconjugated with glycine. To deal with these complexities , we recently developed a multicompartmental model of bile acid metabolism, shown in part in figure 1, and used this to test the error when the isotope dilution technique was carried out with administration of different labeled bile acids and determination of the specific activity decay curve of that bile acid. 34 The Lindstedt procedure using free bile acid appears to be valid; the error associated with use of the labeled glycine conjugate is potentially larger; but
318
HOFMANN AND HOFFMAN Input 2
1
Cholylglycine
Free cholic acid
k13
3
Cholyltaurine
k31 ko3
FIG. 1. Multicompartmental model of metabolism of conjugated cholic ac id in ma n . Conjugat ion is indicat ed by the rate consta n ts, k,, and k ,; return to the liv er of free bil e acid form ed by bac terial deconjugati on is indicated by the constants, k, , and k ., . Free cholic acid pool refers to hypothetical intracellul a r precursor pool for bile acid conjugat ion in which bile ac ids both synthes ized endo genously and return ed are uniformly mixed. The transfer constants, k 02 and k 03 , represent loss from the system by either exc retion , chemical transformation ( su e ~ as dehydroxylat ion), or phys ical transformat ion (such as precipitation).
the potential error associated with use of the labeled taurine conjugate is very large if specific activity is estimated in the taurine conjugate. Measurement of Specific Activity
The adaptation of Talalay's 3-hydroxysteroid dehydrogenase assay of steroids to bile acids by Iwata and Yamasaki 35 was of immense importance. Thus , cholic acid can be isolated from hydrolyzed bile by thin layer chromatography and eluted from the adsorbent . One aliquot can be counted for radioactivity and a second aliquot can be analyzed by the steroid dehydrogenase method. This technique cannot be used for chenodeoxycholic acid unless it is separated from deoxycholic acid. However, Haslewood et a!. 36 recently described the isolation and use of a microbial 7a-hydroxysteroid dehydrogenase that is specific for 7a-hydroxylated bile acids. This method should permit direct assay of the chenodeoxycholic mass in the dihydroxy fraction. Specific colorimetric methods for chenodeoxycholic 37 or deoxycholic 3 8 acid measurement have been described but are used infrequently, and nonspecific spectrophotometric or fluorometric procedures are used with diminishing frequency. The bile acids also may be estimated by gas-liquid
Vol. 67, No. 2
chromatography with an internal standard. Hyodeoxycholic, nordeoxycholic, and hyocholic acids have been used as internal standards, but the last may be difficult to derivatize completely. 39 Hydroxyketo bile acids also have been used , but this procedure seems hazardous because these compounds have been identified in samples of bile from patients with gallstones . 4 0 Generally, bile acids are chromatographed as trifluoroacetates on QF-1 columns 41 or as acetates on QF-1 or AN-600 columns 4 2 ; trimethylsilyl ethers are well separated on HiEfffiB columns. 43 Good quantification of underivatized methyl esters has proved difficult to achieve for most workers . Expression of Data The primary data derived from regression analysis of the specific activity decay curve are pool size and daily fractional turnover rate. The product of these two is the daily synthesis rate. The half-life of label Ctv, ) is commonly reported in place of k, the fractional turnover rate. There is no point in reporting both since t ~~, = In 2/k. For a pool that remains constant in size, we prefer to use k. Some authors report pool size and daily synthesis in units of mass; others report molar units. We think that the latter is preferable, particularly if free and conjugated bile acids are to be compared. One further problem is the normalization of data. Should data be expressed in terms of actual body weight, ideal body weight, body surface area, or some other variable? Because the purpose of normalization is to decrease the variance within groups believed not to vary with respect to the factor under investigation , the correct method of normalization must be arrived at experimentally. Currently, insufficient data are available to allow clear choice . We recommend that authors publish bile acid kinetics in comparable terms , including data in micromoles per kilogram of body weight or micromoles per square meter of body surface area . Recommendations for calculating body surface area from height and weight have recently been published. 44
August 1974
MEASUREMENT OF BILE ACID KINETICS
Reports should include-or archive-full patient data (age, sex, weight, and height); this practice will permit the appropriate method of normalization to become apparent in time. Validation and Interpretation Validation of Pool Size The pool of bile acid was operationally defined as that mass of bile acids that dilutes a tracer. This should be equal approximately to the mass of bile acids drained by an acute bile fistula, but recent experiments of Mack and Dowling (personal communication) suggest that the mass of bile acid that dilutes a tracer is considerably larger than the mass of bile acids recovered from an acute bile fistula, at least in the rhesus monkey. If the estimation of pool size is valid, then estimation of cholic, deoxycholic, and chenodeoxycholic acid pool sizes by the isotope dilution technique in the same subject should produce pool size values in the same proportion as estimated from class distribution of bile acids by gas-liquid chromatography, and this has been proved true 45 (also, T. S. Low-Beer, personal communication). Validation of Synthesis Rate Synthesis can be measured by chemical assay of fecal bile acids (balance technique) because, in the steady state, fecal excretion is equal to synthesis. We can find no published reports in which isotope dilution measurements have been directly compared with fecal excretion measurements in the same patient, although L. Swell and Z. R. Vlahcevic (personal communication) report good agreement between the two methods. Miettinen 2 observed that synthesis values obtained by isotope dilution tend to be higher than those estimated by the balance technique. At present, it seems likely that the isotope dilution measurements give fairly valid but quite reliable estimates of bile acid synthesis. If it is shown that such measurements include a systematic error, the method will still probably be preferred to the balance technique because of the
319
complexity of the analytical methods and the uncertainties of sampling in the balance technique. 46 In patients with severe bile acid malabsorption, the isotope dilution technique cannot be assumed to be valid, and here measurement of fecal bile acids is the only correct method for estimation of bile acid synthesis.' Whether the one-compartment model is valid in patients with hepatobiliary obstruction is uncertain and should be tested by comparison with balance studies. Balance studies in such patients should measure both urinary and fecal bile acid excretion. 23 Relationship of Bile Acid Pool Size to Bile Acid Secretion The output of bile acids into the duodenum is defined as secretion and was the major measurement made by physiologists during the past century. With the development of isotope dilution techniques, secretion studies languished, especially because it is difficult to measure secretion in man without perturbing the enterohepatic circulation. Thureborn 47 and Isaksson and Thureborn, 48 in man, and Dowling et al., 49 in the monkey, developed methods for controlled interruption of the enterohepatic circulation; the latter used this technique to define the relationship between return to the liver and bile acid synthesis in a number of fundamental papers. 5°' 51 Because secretion is a product of pool size and recycling frequency and because pool size was easily measured, some workers assumed that it was reasonable to estimate secretion by multiplying pool size by a constant recycling frequency. To paraphrase, secretion was considered to be directly proportional to pool size. During this past year, biliary secretion has been measured directly, by duodenal perfusion techniques, by Northfield and Hofmann. 52 In studies carried out on a group of patients with various pool sizes, they found that recycling frequency was inversely proportional to pool size. Thus, when normalized for body weight, secretion was constant despite widely different pool sizes.
320
HOFMANN AND HOFFMAN
Based on Northfield and Hofmann's data, one is tempted to coin a "bile's law" (analogous to Boyle's gas law): CP
C'P'
=
in which C is recycling frequency and Pis pool size. This idea needs further experimental verification but has been argued from other lines of evidence by Low-Beer and Pomare. 53 Perhaps students of the enterohepatic circulation might have done well to discuss this problem with cardiologists, who never have attempted to predict cardiac output from blood volume. The functions of bile acids in the small intestine are more closely related to bile acid secretion than to pool size. Measurements of secretion would appear to have an important place in the characterization of bile acid metabolism in digestive disease or of the mode of action of agents that alter biliary lipid secretion. Value of Synthesis Measurements Bile acid synthesis data represent the most valuable information afforded by the isotope dilution technique, in our judg-
Vol. 67, No.2
ment. The measurement of greatest potential interest is the rate of cholic acid synthesis relative to chenodeoxycholic acid synthesis; this appears to be strikingly altered in a number of metabolic diseases (fig. 2). If labeled deoxycholic acid is injected, its
specific activity decay curve also decreases exponentially. 45 Thus, the deoxycholic acid pool behaves as a single compartment. The fractional turnover rate of deoxycholic acid seems to be identical to that of chenodeoxycholic acid, indicating similar intestinal conservation 45 • 54 ; few data are available, however. The origin of deoxycholic acid is bacterial 7a-dehydroxylation of cholic acid. In health, the deoxycholic acid input is about one-third of the cholic acid excretion (or synthesis). Thus, less than half of the deoxycholic acid that could be formed is absorbed from the intestine. One can define a fraction, fctehydrox , as the observed input of deoxycholic acid divided by the maximal amount of deoxycholic acid that could be formed. This is simply input of deoxycholic acid divided by synthesis of cholic acid. There is no meaningful way to estimate
1.0 8
0.8 0.6
XC f!j
X •
~ ·~ .. x
4§ 0 4
D D
_a.E.,_ D
8
4 4.44
~0
D
~
0.4
. .t
40c:P
--;:r-
•
• --•
4
440
g
0.2 0 Normal
Cholelithiasis
TypeTI hyperlipidemia
TypeN Cerebrohyper- tendinous xantholipidemia matosis
FIG. 2. Mole fraction of cholic acid in bile acid synthesis (cholic acid synthesis/total bile acid synthesis) in health and several disease states. Symbols indicate source of data: e, Vlahcevic et al. 21 ; x, Hepner et al. "· "; 0 • , Danzinger et al. 30 ; !;,., Kottke 19 ; and 0 , Einarsson and Hellstrom." In the studies by Danzinger et al.' 0 , kinetics were measured in the same patient before (D) and after 6 months . Data from patients with cerebrotendinous xanthomatosis ( A) were generously provided by Dr. Gerald Salen (unpublished observations). Horizontal lines indicate group means .
<•l
August 1974
MEASUREMENT OF BILE ACID KINETICS
actual deoxycholic acid formation, because it could occur in the rectum or after defecation. The value of fctehyctrax would approach 1 in those individuals in whom deoxycholic acid is the major bile acid but would be near 0 in individuals who, although making deoxycholic acid, have little in their bile. Patients with cirrhosis or patients ingesting cholestyramine are probably good examples of the latter. To measure fctehydrax , one should carry out conventional isotope dilution techniques using labeled cholic and deoxycholic acids. As previously noted, the decay of specific activity is due to the input of unlabeled bile acid. During bile acid feeding, the input will be the sum of synthesis and absorbed fed bile acid. As long as the fed dose is constant and the rate of absorption is approximately constant, a steady state will obtain; the specific activity decay will be exponential, and tlie usual kinetic analysis is appropriate. 30 If the exogenous dose is large relative to endogenous synthesis, then an estimate of the fractional absorption of the fed dose may be calculated.
The Future Measurement of bile acid synthesis rate and pool size by isotope dilution techniques appears to have come of age at last. The measurement of synthesis rate, especially of both primary bile acids, appears to be of great value. The value of pool size measurements is much less certain. Indeed, if recycling frequency and pool size are inversely related, then estimation of pool size may be of value chiefly for estimating recycling frequency which in turn may give a clue to altered gallbladder function. If the major value of the measurement of bile acid kinetics is estimation of primary bile acid synthesis, then the major use of bile acid kinetics will be in the area of intermediary metabolism. On the other hand, the role of bile acids in the pathogenesis of many hepatic and intestinal diseases is poorly understood and may be pivotal; measurement of bile acid kinetics is an essential part of complete characterization of bile acid metabolism. From this standpoint, we can expect that the determination of bile acid kinetics by isotope
321
dilution techniques will continue to have an important place among investigative procedures in gastroenterology. REFERENCES 1. Hofmann AF, Schoenfield W, Kottke BA, et al: Methods for the description of bile acid kinetics in man. Methods Med Res 12:149- 180, 1970 2. Miettinen TA: Clinical implications of bile acid metabolism in man, The Bile Acids: Chemistry, Physiology, and Metabolism, vol 2. Edited by PP Nair, D Kritchevsky. New York, Plenum Press, 1973, p 191-247 3. Bergstrom S, Rottenberg M, Voltz J: The preparation of some carboxylabelled bile acids: bile acids and steroids 2. Acta Chern Scand 7:481- 484, 1953 4. Lindstedt S: The turnover of cholic acid in man : bile acids and steroids 51. Acta Physiol Scand 40:1- 9, 1957 5. Sjovall J: Separation of bile acids by paper chromatography: bile acids and steroids 3. Acta Chern Scand 6:1552- 1553, 1952 6. Sjovall J: On the separation of bile acids by partition chromatography : bile acids and steroids 4. Acta Physiol Scand 29:232-,240, 1953 7. Norman A: Separation of conjugated bile acids by partition chromatography: bile acids and steroids 6. Acta Chern Scand 7:1413-1419, 1953 8. Sjovall J: Separation of conjugated and free bile acids by paper chromatography: bile acids and steroids 12. Acta Chern Scand 8:339-345, 1954 9. Eriksson S, Sjovall J: The absorption spectra of bile acids in sulfuric acid: bile acids and steroids 31. Arkh Kemi 8:303-310, 1955 10. Eriksson S, Sjovall J : The absorption spectra of conjugated bile acids in sulfuric acid: bile acids and steroids 32. Arkh Kemi 8:311-315, 1955 11. Sjovall J: Quantitative determination of bile acids on paper chromatograms: bile acids and steroids 33 . Arkh Kemi 8:317-324, 1955 12. Bergstrom S, Lindstedt S, Samuelsson B: Bile acids and steroids LXXXII: on the mechanism of deoxycholic acid formation in the rabbit. J Bioi Chern 234:2022-2025, 1959 13. Danielsson H, Eneroth P, Hellstrom K, et al: On the turnover and excretory products of cholic and chenodeoxycholic acid in man: bile acids and steroids 134. J Bioi Chern 238:2299-2304, 1963 14. Lindstedt S, Avigan J, Goodman DS, et al: The effect of dietary fat on the turnover of cholic acid and on the composition of the biliary bile acids in man . J Clin Invest 44:1754- 1765, 1965 15. Lack L, Weiner IM : In vitro absorption of bile salts by small intestine of rats and guinea pigs. Am J Physiol 200:313-317, 1961 16. Playoust MR, Lack L, Weiner IM: Effect of
322
17.
18.
19.
20.
21.
22.
23 .
24.
25.
26.
27 .
28. 29.
30 .
31.
HOFMANN AND HOFFMAN
intestinal resection on bile salt absorption in dogs. Am J Physiol 208:363-369, 1965 H eaton KW, Austad WI, Lack L, et al: Enterohepatic circulation of C " -la beled bil e salts in disorders of the distal s m a ll bowel. Gastroenterology 55 :5- 16, 1968 W ollenweber J , Kottke BA, Owen CA Jr: P ool size and turnover of bile aci ds in six hypercholesteremic patients with an d without adm inistration of nicotinic acid. J Lab Clin Med 69:584-593, 1967 Kottke BA: Differences in bile acid excretion: primary hypercholesterolemia compared to combined hy percholeste rolemia and hy pertriglyceridemia. Circul ation 40:13-20, 1969 Vla hcevic ZR, Bell CC Jr, Buhac I, et al: Diminish ed bile acid pool size in patients wi t h gallstones. Gastroenterology 59: 165-173, 1970 Vlahcevic ZR, Mill er JR, F arra r JT, et al: Kin etics and pool size of primary bile aci ds in man. Gastroenterology 61:85- 90, 1971 Vl a hcevic ZR, Buhac I, Farrar JT, et al: Bile acid metabolism in patients with cirrhosis. I. Kinet ic aspects of cholic acid metabolism. Gastroenterology 60: 491- 498, 1971 Vlahcevic ZR, Juttijudata P, Bell CC Jr, et al: Bile acid metabolism in p atients with cirrhosis. II . C holic and chenodeoxyc holic ac id metabolism. Gastroenterology 62:1174-1181, 1972 H offm an NE , LaRusso N F , Hofm ann AF: Noninvasive determination of bile aci d kinetics using a sequesteri ng capsule (abstr). Gastroe nt erology 64:744, 1973 H offman NE , LaRusso NF, Hofm ann AF: An improved met hod for faecal collection: the faeca l field-kit. Lancet 1:1422- 1423, 1973 H ofmann AF, Szczepanik PA , Klein PD : Ra pid preparation of tritium -labeled bile acids by enolic exchan ge on basic alumina containin g tritiated water. J Lipid Res 9:707- 713, 1968 LaRusso NF, Hoffm an NE, Hofmann AF: Validity of using 2, 4-' H-labe led bile acids to study bile acid kin etics in man . J Lab Clin Med (in press) Stahl E, Amesjo B: Ta uroc holate metabolism in man . Scand J Gastroenterol 7:559-566, 1972 H epner GW, Sturman JA , Hofman n AF, et al: M etabolism of steroid an d am ino acid moieties of conjugated bile acids in man . III. Choly ltaurine (ta urocholic acid) . J Clin Invest 52:433-440, 1973 D anzin ger RG, Hofmann AF, Thistle JL, et al: Effect of oral chenodeoxych olic acid on bil e acid kinetics and biliary lipid composition in women with cholelithiasis. J Clin Invest 52:2809-2821, 1973 Klein PD, Haumann JR, Eisler WJ: Gas chromatograph-mass spectrometer-acceleratin g voltage alternator system for t he m easurement of
32.
33.
34.
35.
36.
37.
38.
39 .
40.
41.
42.
43.
44.
45 .
Vol . 67, No.2
stable isotope ratios in organic molecules . Anal Chern 44:490- 493, 1972 H ofmann AF, Dan zin ger RG, Hoffman NE, et al: Validation an d comparison of deuterium/tritium and carbon-1 3/carbon-14 in clinical studies of bile ac id metabolism (USAEC conf 711115), Proceedings of a Seminar on the Use of Stable Isotopes in Clinical Pharmacology . Edited by PD Kl ein , LJ Roth. Oak Ridge, United States Atomic Energy Commission, 1972, p 37-51 Hachey DL, Szczepanik PA, Berngruber OW, et al: Syntheses of stable isotopes: synthesis of deuterium and "C labeled bile acids. J Labelled Compounds 9:703-719, 1973 Hoffman NE, Hofmann AF: M etabolism of steroid and amino acid moieties of conjugated bile acids in man. IV. Desc ription and validation of a multicompartmental model. Gastroenterology (in press) Iwata T, Yamasaki K: En zy matic determination and thin-layer chromatography of bile ac ids in blood. J Bioc hem (T okyo) 56:424- 431, 1964 Haslewood GAD, Murphy GM, Richardson JM: A direct enzymic assay for ?a -hydroxy bile acids and their conjugates. Clin Sci 44:95-98, 1973 Isaksson B: A method for spectrophoto metric determination of chenodesoxyc holic aci d in bile. Acta Chern Scand 8:889- 897, 1954 Bruusgaard A: Quantitative determination of the major 3-hydroxy bile acids in biological material after thin-layer chromatographic separation. Clin Chim Acta 28:495-504, 1970 Mott GE, Moore RW, Reiser R: Gas-liquid chromatography of hyocholic acid. J Lipid Res 12:117- 119, 1971 Hepner GW, Hofmann AF, Mal agelada JR, et al: Increased bacterial degradation of bile acids in chol ecystect omize d patients. Gastroenterology 66:556-564, 1974 Eneroth P, Sjovall J: E xtraction, purification, and chromatographic analysis of bile ac ids in biological materials, The Bile Acids: Chemistry, Phys iology, and Metabolis m, voll. Edited by PP Nair, D Kritchevsky . New York, Plenum Press, 1971, p 121-171 Roovers J , Evrard E , Vanderhaeghe H : An improved met hod for measuring human blood bile acids . Clin Chim Acta 19:449-457, 1968 Makita M, Wells WW: Quantitative analysis of fecal bile acids by gas-liquid chrom atography. Anal Biochem 5:523-530, 1963 Gehan EA, George SL : Estimation of human body surfa ce area from height and weight . Cancer Chemother Rep 54:225-235, 1970 Hepner GW, Hofm ann AF, Thomas PJ: Metabolism of steroid and a mino acid moieties of conjugated bile ac ids in man . II. Glycine-conjugated
August 1974
46.
47.
48.
49.
50.
MEASUREMENT OF BILE ACID KINETICS
dihydroxy bile acids. J Clin Invest 51:1898- 1905, 1972 Grundy SM, Ahrens EH Jr, Miettinen TA: Quantitative isolation and gas-liquid chromatographic analysis of total fecal bile acids . J Lipid Res 6:397- 410, 1965 Thureborn E: Human hepatic bile: composition changes due to a ltered enterohepatic circulation. Acta Chir Scand [Suppl] 303:1- 63, 1962 Isaksson B, Thureborn E : En metod for uppsamling av hepaticusgalla vid intakt enterohepatiskt kretslopp (abstr). Nord Med 66:1519, 1961 Dowling RH, MackE , Picott J, et al: Experimental model for the study of the enterohepatic circulation of bile in rhesus monkeys . J Lab Clin Med 72:169-176, 1968 Dowling RH, Mack E, Small DM: Effects of controlled interruption of the enterohepatic circulation of bile salts by biliary diversion and by ileal resection on bile salt secretion, synthesis, and pool size in the rhesus monkey. J Clin Invest
323
49:232-242, 1970 51. Dowling RH, Mack E , Small DM: Biliary lipid secretion and bile co mposition after acute and chronic interruption of the enterohepatic circulation in the rhesus monkey. IV. Primate biliary physiology. J Clin Invest 50:1917- 1926, 1971 52. Northfield TC, Hofmann AF: Biliary lipid secretion in gallstone patients . Lancet 1:747-748, 1973 53. Low-Beer TS, Po mare EW: Regulation of bile salt pool size in man . Br Med J 2:338-340, 1973 54. Quarfordt SH, Greenfield MF: Estimation of cholesterol and bile acid turnover in man by kinetic analysis. J Clin Invest 52:1937- 1945, 1973 55. Hepner GW, Hofmann AF, Thomas PJ : Metabolism of steroid and amino acid moieties of conjugated bile acids in man. I. Cholylglycine. J Clin Invest 51:1889-1897 , 1972 56. Einarsson K, Hellstrom K: The formation of bile acids in patients with three types of hyperlipoproteinaemia. Eur J Clin Invest 2:225- 230, 1972