Metabolism of mevalonate in rats and man not leading to sterols.

C-5 of mevalonate appears as C0-2 in the breath of rats and men almost immediately after administration either by injection or by mouth. Adult rats exhaled up to 6.5% of a dose of RS-[5-14C]mevalonate (13% of the utilizable R-enantiomer) in the breath in 100 min. The 14-C02 was not derived either from the matabolism of cholesterol biosynthesized from [5-14C]mevalonate or from the metabolism of the unnatural S-enantiomer of mevalonate. The amount of 14-C02 expired in the breath was the same whether the [5-14C]mevalonate was given intravenously or in a drink of water to man. One normocholesterolemic man dissipated 12%, a mildly non-familial hypercholesterolemic man dissipated 10%, and a familial hypercholesterolemic man dissipated 7% of a dose of [5-14C]mevalonate in 24 hours (calculated as a per cent of the R-enantiomer). The observations support the hypothesis of the existence of a metabolic shunt of intermediates of sterol biosynthesis, derived from mevalonate, not leading to sterols.

SUMMARY C-5 of mevalonate appears as CO, in the breath of rats and men almost immediately after administration either by injection or by mouth.
Adult rats exhaled up to 6.5 % of a dose of RS-[5-Wlmevalonate (13 % of the utilizable R-enantiomer) in the breath in 100 min.
The 14C02 was not derived either from the metabolism of cholesterol biosynthesized from [5J4C]mevalonate or from the metabolism of the unnatural S-enantiomer of mevalonate. The amount of 14C02 expired in the breath was the same whether the [5-14C]mevalonate was given intravenously or in a drink of water to man. One normocholesterolemic man dissipated 12%, a mildly nonfamilial hypercholesterolemic man dissipated lo%, and a familial hypercholesterolemic man dissipated 7 % of a dose of [5-14C] mevalonate in 24 hours (calculated as a per cent of the R-enantiomer).
The observations support the hypothesis of the existence of a metabolic shunt of intermediates of sterol biosynthesis, derived from mevalonate, not leading to sterols.
It has been a generally held view over the last 15 years that mevalonate had no metabolic fate other than the provision of carbon atoms for polyisoprenoid biosynthesis, cholesterol being quantitatively the most important end product of its metabolism in animals (1). Christophe and Popjsk (2) first drew attention in 1961 to a pathway for the metabolism of allylic prenyl pyrophosphate intermediates of sterol biosynthesis, dcrived from mevalonate, not leading to sterols. It was shown that in liver homogenates, t)he prenyl pyrophosphates were hydrolyzed by a microsomal phosphatase to the free prenols which then were converted irreversibly in two steps of dehydrogenation to the corresponding carboxylic acids. Edmond and PopjBk (3) have shown recently that C-2 of mevalonate was transferred to n-fatty acids (palmitic and stearic) in young rats and have invoked a hypothesis first proposed by Popjak (4,5) in order to explain the appearance of mevalonate carbons in n-fatty acids. The cardinal feature of this hypothesis was that 3,3-dimethylacrylic acid-derived by dehydrogenations of 3,3-dimethylallyl alcohol, released from dimethylallyl pyrophosphate-would be converted to its cocnzyme A derivative and hence merge with an identical metabolite of leucine. The conversion of 3,3-dimethylacrylyl-CoA via tram-3-methylglutaconyl-Cob to 3-hydroxy&methylglutaryl-CoA is an established metabolic path (6). Cleavage of 3-hydroxy-3methylglutaryl-Coil by its lyase (6) generates free acetoacetate from its C-3, C-3', C-4, and C-5 and acetyl-CoA from its C-l and C-2. By the hypothesis proposed, it was mandatory that C-4 and C-5 of mevalonate be converted into C-2 and C-l, respectively, of acetyl-CoA, and C-3', C-3, and C-4 of mevalonate to become, respectively, C-4, C-3, and C-2 of free acetoacetate.' From the known transformations of mevalonate int#o squalene and sberols (1) it can bc predicted that [5-%]mevalonate will give lanosterol labeled in positions 2, 6, 11, 12, 15, and 23 and cholesterol labeled in the same positions (Fig. 1). None of these labeled positions can be lost in oxidizable products during the formation of bile acids. The only known reaction by which cholesterol biosynthesixed from [5-14C]mevalonate could generate an oxidizable product and give 14C02 in the breath is the formation of pregnenolone whereby the cholesterol side-chain is cleaved and C-23 becomes the /3 carbon atom of isocaproyl-CoA.
We report now that C-5 of mevalonate appears in the breath as CO2 within 1 rnin after the administration of [5-%]mevalonate to adult rats or man in amounts far greater than could be accounted for by the transformation of cholesterol into pregnenolone and explicable only by the hypothesis proposed (3)(4)(5)  the filter paper and Hyamine solution were transferred to scintillation fluid for counting. Serum Cholesterol and Triglycerides-These were determined in the Clinical Laboratory of the University of California at Los Angeles Hospital by methods previously referred to (14).
Animals-Adult male Sprague-Dawley rats were used. The methods of tissue analysis were described in a previous publication (3). RESULTS

'*CO2 in Breath after Administration of [6J*C]MevaZonate to
Ruts-For these experiments, three 250-g male rats were injected intramuscularly with 21 PCi (1.78 I.cmol) of RS-[5-14C]mevalonate and were placed immediately into the plastic cage of the 14COz-monitor. Radioactive COZ appeared in the expired air within 1 min after the injection and rose to a maximum at about 45 min after the injection followed by an exponential decline over several hours. By the integration of the records, we found that the rats expired in 100 min up to 6.5% of the l*C of the total dose corresponding to 13% of the 1% contained in the R-enantiomer (Fig. 2). In order to exclude the possibility that the 14C02 was generated by the degradation of sterols synthesized by the animals, we gave by mouth to a 260-g rat 9.07 PCi (0.97 mg) of ['*Clcholesterol biosynthesized from [5-i*C]mevalonate (cf. "Methods and Materials") and dissolved in 0.1 ml of 20% ethanol in olive oil. The breath of the animal then was monitored for 7 hours. No trace of r*C could be detected in the expired air during that time. Analysis of the organs for ['*C]cholesterol 7 hours after the feeding showed that the cholesterol was well absorbed. Table I  The possibility remained that the '*CO* arose by the metabolism of the unnatural S-enantiomer of the mevalonate. Hence, we injected intramuscularly 10 /.Si of S-[5-i*C]mevalonate (cf. "Materials and Methods") into a 250-g rat and monitored its breath for 2 hours. Only a trace of radioactivity appeared in the breath, too small to be quantitated, and which could have arisen from the metabolism of the small amount of the Renantiomer present in the injected specimen. The presence of some R-[B-14C]mevalonate in the specimen of S-mevalonate was indicated by the incorporation of label into the sterols in the liver and kidneys of the animal: these organs together contained a total of 2.57 X lo* dpm of i*C in sterols as compared to 6.03 x lo6 dpm after the injection of 10.6 PCi of RS-[5-W]mevalonate.
14C02 in Breath and Blood-HCOs after Administration of [6J*Cj-Mevalonate to Man-The present experiments and those of Edmond and Popjak (3), and some further studies to be reported later, support the thesis that a part of the intermediates derived from mevalonate normally are diverted onto a metabolic pathway not leading to sterols. As impairments of such a shunt could lead to substantial changes in cholesterol biosynthesis (cf. "Discussion"), we explored the existence of such a metabolic shunt in man.
Informed consent for study was obtained from four men who volunteered for experiments with the [EiJ*C]mevalonate.
The first subject was a 74-year-old man who was given intravenously 300 CrCi (25.4 pmol) of RS-[5-14C]mevalonate and whose breath was monitored continuously with the ionization chamber (cj. "Methods and Materials"). l*C appeared in the breath almost immediately after the injection and continued to rise for about 100 min. The calibrated ionization chamber used for the experiments with the rats was out of order on this occasion and another uncalibrated instrument had to be used. For this reason, we were unable to calculate the total 14C expired. The experiment confirmed, nevertheless, the existence in man also of a metabolic path of mevalonate not leading to sterols.
In three further subjects, the intermittent sampling of breath was employed (cf. "Materials and Methods") over 24 hours after administration of RS-[5-r4C]mevalonate. The study was made in the Clinical Research Center of University of California at Los Angeles Hospital.
Two tests, 6 weeks apart, were made on the next subject, a 42-year-old male heteroeygote familial hypercholesterolemic whose serum cholesterol levels varied between 400 and 500 mg/lOO ml (serum triglycerides, 150 mg/lOO ml) and who had severe atherosclerosis. On the first occasion he was given injecting RS-[6-14C]mevalonate to rats One 260-g rat was given by mouth 9.07 &i (0.97 mg) of [2,6,11, 12,15, Breath and blood samples were taken at frequent intervals for analysis. The specific activities of the CO:! in the breath and of the radioactivity in blood cholesterol are shown in Fig. 3. The specific activities of the blood bicarbonate gave a curve almost superimposable over that given by the breath-COz.
Six weeks later, after base-line determinations and under the same conditions as in the first test, the same dose of mevalonate as above was given to this man in a drink of water. A swab of the mouth and base of the tongue showed only traces of 14C immediately after the dose was washed down with further amounts of water. Breath and blood were analyzed as before; the data obtained are included on Fig. 3. From the integration of the CO2 specific activity-time curves and by the assumption that adult man produces hourly 9 mmol of COZ per kg body weight (15), we calculated that this man on both occasions exhaled in 24 hours 7% of the dose of R-mevalonate as l*COZ.
Two further persons were studied similarly: a 50-year-old man with atherosclerosis and nonfamilial hypercholesterolemia (serum cholesterol, 285 mg/lOO ml; serum triglycerides, 137 mg/lOO ml) was given 26 &i (2.2 pmol), and a 39-year-old normocholesterolemic man (serum cholesterol, 225 mg/lOO ml; serum triglycerides, 117 mg/lOO ml) was given in a drink of water 13 PCi (1.1 pmol) of RS-[5-14C]mevalonate. In both men 14C02 appeared in the breath in less than 5 min after the drink, and in the 24 hours they excreted 10 and 12y0, respectively, of the 14C contained in the R-enantiomer as 14C02.
Possible Bacterial Metabolism of [P4C]MevaZonate to 14CO~-In order to exclude the possibility that bacterial action on the orally administered mevalonate might have been responsible for some of the 14C02 excretion, we incubated a heavy inoculum of nine species of bacteria and yeast (Lactobacillus species, Lactobacillus plantarum, Bacteroides oralis, Fusobacterium nucleatum, Bacteroides melaninogenicus, Candida albicans, Streptococcus saliva&s, Streptococcus lactis, and Streptococcus mitis) isolated from human upper small intestine for 3 hours at 37" with 0.5 PCi of RS- [5-14C]mevalonate and trapped the CO2 over the cultures in Hyamine. The amount of 1% trapped in the Hyamine, 940 dpm, was so small that bacterial action cannot account for our observations in man.
The observations presented demonstrate that a pathway for the metabolism of mevalonate, or of intermediates derived from it, not leading to polyisoprenoids and sterols exists not only in newborn animals (3), but also in the adult rat and man. The rapid excretion of C-5 of mevalonate in breath as COz is in full accord with the metabolic shunt ("the trun.s-methylglutaconate shunt") postulated previously (3)(4)(5). Further experiments made in animals with mevalonate and other substrates labeled in various positions with 3H, 1%) and 14C, to be reported elsewhere, give also full support for the existence of this metabolic pathway. The elimination of C-5 of mevalonate as CO2 is probably proportional to the amount of mevalonate not used for sterol synthesis, but can give only the lowest limit for the extent of the metabolic shunt. It was shown previously (3) that significant amounts of mevalonate carbon appeared also in n-fatty acids. Since respiratory COz is in equilibrium with urea, significant amounts of radioactivity, derived from C-5 of mevalonate, must have been excreted in the urine with urea also.
Our data demonstrate that a pathway for the metabolism of mevalonate and intermediates derived from it and not leading to sterols exists in the rat and man, a pathway not generally recognized before.
The significance of this shunt of mevalonate metabolism is that a defect in it could explain some of the human hypercholesterolemias. For example, the approximate normal synthesis of 3 mmol of cholesterol per day in the liver of an adult man represents 18 mmol of mevalonate. An individual who metabolizes 12% of the mevalonate synthesized daily by an alternative pathway would produce daily about 20.5 mmol and dissipate 2.5 mmol. If 20.5 mmol of mevalonate were the average daily production of this substance but only 7% of it were metabolized on a "shunt," only 1.4 mmol would be dissipated and an extra 1.1 mmol would become available for cholesterol synthesis. Such a change would lead to an excess production of 0.18 x 365 mmol = 25.4 g of cholesterol in 1 year, more than twice the amount of cholesterol circulating in the blood of a normocholesterolemic individual.
We do not wish to imply from our data that familial hypercholesterolemia, for example, results from an impairment of the metabolic shunt of mevalonate, as the extent and variations of this metabolic pathway remain to be defined in a statistically significant population of normals and abnormals, including persons with familial hypercholesterolemia and other metabolic abnormalities.