N. Under both conditions of reduced fat absorption, a

Transcription

1 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 256,,No. 6, Issue of March 25, pp , 1981 Printed tn U.S.A. (Received for publication, October 1, 1980) Herbert G. Windmueller and Ai-Lien Wu From the Laboratory of Nutrition and Endocrinology, National Institute of Arthritis, Metabolism, and Digestive Diseases, National Institutes of Health, Bethesda, Maryland Using two complementary methods developed and applied earlier in rats absorbing triacylglycerols, we havenowdeterminedtheintestinalcontribution of individualplasmaapolipoproteinswhentherateof intestinal lipid absorption is very low. After all dietary fat was withheld for h, the intestine produced -50% of the plasma apolipoprotein A-I (apo-a-i) and apo-a-n, similar to its contribution during fat absorption. When rats were fed a fat-free enteric lymph. In Method 11, the pattern of apolipoprotein diet for 12 days and, in addition, biliary fat was diverted synthesis, defined as the relative amounts of [3H)leucine infrom the intestine for 16 h, the intestine still produced corporated into the individual proteins, is fiist determined -60% of the apo-a-1 and only slightly less of the apo-a- N. Under both conditions of reduced fat absorption, a separately in isolated, perfused livers, in isolated intestinal larger proportion of the intestinally-derived apo-a-i segment preparations, and in intact rats. Then the relative and apo-a-iv bypassed the mesenteric lymph and was organ contribution needed to account for the pattern observed released directly into intestinal venous blood. The in- in intact rats is calculated by multiple regression analysis. testinal apo-b contribution, 16% of the total in fat-fed Application of these methods in animals absorbing large rats, was reduced to -5% when dietary fat was with- amounts of fat, mostly triacylglycerols, showed that -50% of held. Intestinal apo-b was released entirely into the the plasma apo-a-i and apo-a-iv and 16% of the apo-b are lymph. Intestine produced only small amounts of apo- synthesized by the small intestine (8). Since apo-a-i and apo- C and little or no apo-e under all conditions. A-IV are the predominant protein constituents of the plasma The results indicate that production of apo-a-i and high density lipoproteins (7), the intestine plays an important apo-a-iv by the small intestine is not regulated by the role, at least under these conditions, in the biogenesis of this rate of intestinal triacylglycerol transport. Lipids in the circulation are transported as lipoprotein complexes containing a heterogenous but specific group of proteins (apolipoproteins) whose structure and function are being widely investigated (2,3). The principal apoproteins in the rat are apo-a-i, apo-a-iv, apo-b, apo-c peptides, and apo-e (4-7). Analogous apolipoproteins are found in most other animals and in man. In the rat, liver and small intestine are the sole sources of plasma apolipoproteins (8). Although the types of lipids released into the circulation by these two organs are similar, the patterns of apolipoprotein synthesis differ greatly. While apo-e is the major product of the liver (9, lo), apo-a-i (11-13) and apo-a-iv (13) are the chief products of intestine, and both tissues produce apo-b in abundance (9, 11, 13). We have described two complementary methods for quantifying the relative hepatic and intestinal contributions to individual plasma apolipoproteins in the rat (8). In Method I, intestine and liver are preferentially exposed to leucine labeled with 3H and 14C, respectively. The relative organ contribution of any plasma apolipoprotein is then indicated by its r value, * Presented in part at the annual meeting of the American Society of Biological Chemists, New Orleans, Louisiana, June 1 to 5, 1980, and published in abstract form (I). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The abbreviation is: apo, apolipoprotein the 3H/ 4C ratio of the protein normalized for the 3H/ 4C ratio of the same animal s plasma albumin, a protein made only in the liver. Thus, an r value of 1.00 reflects a uniquely hepatic origin for the apolipoprotein. The larger the r value, the greater the relative contribution from the intestine. The r value expected of a plasma protein made exclusively by intestine is estimated from the observed r values of intestinal reference proteins, selected apolipoproteins isolated from mes- lipoprotein class. To examine the relationship between fat transport and apolipoprotein production by the small intestine, we have now applied the same methods under two conditions of fat deprivation: (a) after all dietary fat was withheld for h; (b) after dietary fat was withheld for 12 days and, in addition, most endogenous lipid was excluded from the intestine for h by bile diversion. EXPERIMENTAL PROCEDURES Rats and Treatments All rats ( g) were Osborne-Mendel males. Treatment of the animals is summarized in Table I. The protocol used for Group A (fat-fed) has been described in detail (8). A similar schedule was used for Group B except that the intraduodenal fat emulsion preceding and during the experiments was replaced by a continuous glucosesaline infusion. Group C animals received a diet virtually free of fat, had biliary fat diverted from the intestine, and also received no fat emulsion. The animal preparations and procedures used were described previously (8) and are summarized below. Relative Organ Contribution, Method I: Differential Double Labeling Rats were surgically prepared with portal vein and duodenal infusion cannulae and, in some cases, a cannula for mesenteric lymph drainage. Additionally with Group C rats, bile, but not pancreatic secretions, was diverted from the intestine by tying one end of a polyethylene cannula (outer diameter, 0.61-mm) into the common bile duct close to the liver and passing the other end outside the abdomen through a small stab wound. Also in these rats, 8.3 mm sodium taurocholate was added to the intraduodenal infusion solution, to maintain the enterohepatic circulation of bile salts. Treatment of

2 TABLE I Animal treatments Rats were surgically prepared with various cannulae under light ether anesthesia (see Experimental Procedures ) and placed in restraining cages for h for recovery. Duodenal infusion was begun immediately after surgery and continued throughout the experiment as indicated. Time interval Group A (fatfed) Preceding surgical Stock diet6 implantation of cannulae During h recovery period glucose-salined During experiment Intraduodenal fat emulsion Plasma Apolipoprotein Biosynthesis by Intestine 3013 Group B (no fat l2 (no fat, Group days; bile diverh) sion) Stock diet glucose-saline sionl Some of the results from this group have been published (8) and are included here for purposes of comparison. NIH-07 open formula stock diet containing not less than 5% fat. For 12 days the rats received a semisynthetic diet containing 55% sucrose, 15% cornstarch, 20% purified casein, 5% fiber, 0.3% DL-methionine, plus adequate amounts of vitamins and minerals (Diet AIN- 76 (14) in which sucrose replaced the 5% corn oil and menaquinone, or Vitamin K, was increased 10-fold to provide 500 pg/kg of diet). A solution containing 10% glucose, 0.9% NaC1, and 0.04% KC1 was infused at 2.4 rnl/h. e As in d above, except 8.3 mm sodium taurocholate was included in the solution. A lipid emulsion (Intralipid, Vitrum AB, Stockholm) providing 450 &mol of fatty acid/h was infused continuously for 2 h before and, with intact rat experiments, 1 (Method 11) or 2 h (Method I) after labeled leucine administration (see Experimental Procedures ). the animals during the h recovery period was as shown in Table I. During the experiment, 1 mci of ~-[4,5-~H]leucine was administered via the duodenal cannula and, simultaneously, 0.2 mci of L-[U- ~C]- leucine was infused via the portal vein cannula. Only Group A rats received a fat emulsion intraduodenally starting 2 before h the labeled leucine dose. Two hours after label administration, blood was drawn by aortic puncture. The apolipoproteins from plasma and from the collected mesenteric lymph were isolated and the 3H and I4C contents were determined. Relative Organ Contribution, Method II: Zsolated Organs Versus Intact Rat Isolated Intestine-Fbts prepared with duodenal and mesenteric lymph cannulae and, for Group C, also a bile cannula, were maintained for h as described above (Table I). During the experiment, a segment of mid-jejunum was isolated for study in situ (13, 15). ~-[4,5-3H]leucine (1 mci) was administered into the lumen of the segment from which all venous blood and lymph were collected. This ensured that only the segment was exposed to the isotope. Labeled apolipoproteins, unequivocally of intestinal origin, were isolated from the mesenteric lymph and venous blood collected from the preparation for 3 h. Isolated Perfused Livers-Liver donor rats were prepared with duodenal cannulae and allowed to recover on the infusion schedules shown in Table I. Blood donor rats were fed stock diet and used without further preparation. During the experiment, livers were per- fused in situ (16) with 70 ml of recycling perfusate composed of defibrinated rat blood diluted with an equal volume of albumincontaining buffer (8). ~-[4,5-~H]leucine (1 mci) was injected into the portal vein cannula 5 min after perfusate recycling was begun. Perfusate was sampled 1-4 h later, and plasma apolipoproteins were isolated and counted. Intact Rats-Cannulae were secured in the duodenum and right jugular vein and the animals were allowed to recover with the infusion schedules shown in Table I. During the experiment, I mci of ~-14,s- 3H]leucine was administered via the jugular cannula. Blood was drawn 1 h later, and the plasma apolipoproteins were isolated and counted. Isolation and Counting of Lipoproteins and Apolipoproteins Lipoproteins from lymph and from blood and perfusate plasma were isolated as a single fraction (d < 1.21 g/d) by ultracentrifugation (8) to reduce losses of apolipoproteins. Following dialysis and partial delipidation of the lipoproteins, the apoproteins were separated by polyacrylamide gel electrophoresis, stained, sliced from the gels, and counted (8). The protein fraction of d > 1.21 g/ml following dtracentrifugation of mesenteric lymph was also dialyzed and similarly electrophoresed on polyacrylamide gels. Gel regions corresponding to the mobility of lymph apolipoproteins were sliced out and counted. The Fat-free diet radioactivity of total lipoproteins was determined in washed trichloroacetic acid precipitates (17) of the dialyzed samples. Plasma albumin was isolated and counted as previously described (8). Intraduodenal Intraduodenal lntraduodenal Materials glucose-salinetaurocholate plus bile diver- Sodium taurocholate, free of deoxycholate, was from Pierce. The source, purity, and specific radioactivity of other chemicals were as previously indicated (8). RESULTS Relative Orgun Contribution, Method I-The observed r values are shown in Table 11. For proteins made by the liver, the r value is 1.00, and for proteins made by the intestine (I) the r value was estimated from the r value of intestinal reference proteins, intestinally-made apolipoproteins isolated from mesenteric lymph draining from the intestine. Apolipoproteins in the plasma of animals without lymph drainage were produced partly by liver and partly by intestine, and therefore had r values between 1.00 and the r value of I, the exact value being determined by the relative organ contribution. The intestinal reference proteins had similar r values for all animal treatments. Any effect of treatment on the relative organ contribution of an apolipoprotein should therefore be reflected by a change in the r value of that protein in the plasma. The r values for plasma apolipoproteins in Group B animals did not differ significantly from those in Group A. For Group C, the r value for apo-a-i remained unchanged; however the r values for apo-a-iv, apo-b, and apo-c were slightly reduced. Relative Organ Contribution, Method II-The various animal treatments did not significantly affect the amount of TABLE I1 3H/14C ratio of apolipoproteins relative to albumin Animals received ~-[4,5- H]leucine intraduodenally and L- [U- 4C]leucine intraportally. Apolipoproteins were isolated from mesenteric lymph collected for 2 h or, in other animals without lymph drainage, from the plasma after 2 h. Each r value was calculated by dividing the 3H/ 4C ratio of the protein by the 3H/ 4C ratio of albumin isolated from the same animal. All values are means f S.E. for 4 rats. Unless otherwise indicated, results between groups were not significantlv different ( a > 0.05). Reference proteins Albumin I Plasma apolipoproteind APO-A-I APO-A-IV APO-B APO-C ADO-E r value (1.00) (1.00) (1.00) 4.95 f f f f f f f f f f f f 0.03d 1.11 f f f f f f 0.02 Intestinal reference proteins: the mean. r value for apo-a-i, apo- A-IV, and apo B from mesenteric lymph. From rats without lymph drainage. Groups A and C significantly different ( p e 0.05). Groups A and C significantly different ( p = 0.05).

3 3014 Plasma Apolipoprotein Biosynthesis by Intestine TABLE 111 L3H/Leucine incorporation into total lipoproteins and individual apolipoproteins ~-[4,5-~H]leucine was given into the portal cannula of the perfused livers, into the lumen of isolated intestinal segments, and via.the jugular vein cannula of intact rats. Lipoproteins from perfusate, lymph, and plasma were isolated at the time intervals indicated and apolipoprotein isolation and radioactivity determination were as described under Experimental Procedures. Animal groups A, B, and C are described in Table I. Values are means for the indicated number of animals. Individual values differed from the mean by 15% or less. Isolated liver Isolated intestineb Intact rat., plasma Group A Group B Group Ad Group B Group C Group A Group B Group C % dose 0.30 Total 0.36 lipoproteins % total 3H Apolipoproteins APO-A-I APO-A-IV APO-B APO-C APO-E (Total) (100) (loo) (100) (100) (100) (100) (100) Sixteen samples from 7 livers (Group A) and 6 samples from 3 livers (Group B) were collected 1-4 h after [3H]leucine administration. The distribution of 3H among the apolipoproteins did not change with time (see also, Ref. 8). * Three rats each for Groups A and B and 1 rat for Group C. Data include apolipoproteins recovered from the lipoproteins of venous blood and mesenteric lymph collected for 3 h and also the estimated recoveries from the fraction of the lymph, d z 1.21 g/ml (see Experimental Procedures ). e Six rats in Group A and 4 each in Groups B and C. Plasma was obtained 1 h following [ Hlleucine administration. Previous data (8) were corrected to include the estimated recovery of labeled apolipoproteins in the d > 1.21 g/mi lymph fraction (see Experimental Procedures ) TABLE IV Proportion of newlysynthesized apolipoproteins released by intestinal segments directly into the blood ~-[4,5-~H]leucine was administered into the lumen of isolated intestinal segments from which all lymph and venous blood was then collected for 3 h. The 3H content of individual apolipoproteins isolated from the lymph and plasma lipoproteins was determined (see Experimental Procedures ). For each apolipoprotein, the combined radioactivity in lymph and blood lipoproteins was considered as total recovery. Proportion recovered in blood lipoproteins Total plasma apolipoprotein Apolipoprotein Group B (no Group C (no fat, pool GreOdU;PnA~~ fat, h R 12 days; bile di- = 3) version; n = 1) Plasma apolipoprotein labeled leucine incorporated into total lipoproteins with either the isolated organs or the intact animals (Table 111). This indicates that the overall rate of apolipoprotein synthesis was little affected, since the size of tissue leucine pools would not be expected to differ appreciably. Table I11 also shows the observed patterns of apolipoprotein synthesis. Fat deprivation had only small effects on the pattern of hepatic apolipoprotein production. In determining the patterns observed with isolated intestine, apolipoproteins released into the blood as well as into the lymph were considered. The proportion of the total newly-synthesized apo-a-i, apo-a-iv, and apo-c released directly into the blood was increased when the transport of fat was diminished (Table IV). Newly-synthesized apo-b, on the other hand, always appeared exclusively in the lymph (see also, Refs. 11 and 13). The absence of fat also increased the proportion of labeled protein recovered in the lymph fraction of d > 1.21 g/d. Much of this newly-synthesized protein is apolipoprotein, TABLE V Contribution by intestine to plasma apolipoproteins (Method 11: isolated organs versus intact rat) The calculations werebased on the data in Table 111 and the principle and equations described inref.8. The per cent hepatic contribution is equal to 100 minus the per cent of total contributed by the intestine. Per cent of total contributed by intestine Group A (fat-fed) Group (no fat h) % total Apo-A-I 45.0 ( ) 47.4 ( ) APO-A-IV 48.4 ( ) 58.0 ( ) APO-A-I APO-B 16.2 ( ) 5.3 ( ) Apo-A-IV ( ) 3.2 Apo-C 2.5 ( ) APO-B Apo-E ( ) ( ) APO-C Values in parentheses are the upper and lower limits for the Much of the plasma apo-a-iv was apparently lost during ultra- intestinal contribution as determined from Method I using the r centrifugation; the analytical gel used for plasma apolipoprotein sep- values in Table I1 (see Ref. 8 for the method of calculation). aration showed less apo-a-iv and the protein in the fraction, d > 1.21 g/ml contained more radioactivity than similar samples from Group B. based on its apparent molecular weight and its capacity to associate with added chylomicrons (13). To reduce the error that would result by ignoring this differential loss, the recovery of labeled apolipoproteins in the lymph fraction of d > 1.21 g/ ml was estimated by gel electrophoresis and included in the values for isolated intestine shown in Table 111. Recovery in this fraction did not exceed 25% of the total synthesized for any apolipoprotein except apo-a-iv, of which more than 80% was recovered there in the absence of fat (Groups B and C). With these best-estimate corrections, the patterns of intestinal synthesis reveal a decreased relative production of apo-a-iv and particularly apo-b in the absence of fat (Groups B and C versus Group A). In the intact rats, there was an apparent decrease in fractional production of apo-a-iv during fat deprivation. This reflects a similar response in both liver and intestine, exaggerated perhaps by a greater loss of apo-a-iv to the plasma fraction of d > 1.21 g/d. The relative contributions by liver and intestine were then calculated by multiple regression analysis from the Table 111 results for Groups A and B, where complete data were ob-

4 tained (Table V). With minor exceptions, the values calculated by Method I1 fell within the upper and lower limits calculated from the r values in Method I. The results show that of the total [3H]leucine incorporated by the whole animal into apolipoproteins, 17.7% was incorporated by intestine in fat-fed rats and 12.2% when no fat was fed. Intestine in the absence of dietary fat continued to produce -50% of the circulating apo-a-i and apo-a-iv and -3% of the apo-c. Without fat, the intestinal apo-b contribution, however, was reduced from 16 to 5% of the total. DISCUSSION Our earlier work showed that during periods of fat absorption, the small intestine produced -50% of the newly-synthesized apo-a-i and apo-a-iv, 16% of the apo-b, and 3% of the apo-c that entered the blood. The remainder of each was produced by the liver. These values were derived using data from two complementary methods (8). With similar techniques, we have now shown that when fat transport is greatly diminished or even virtually abolished, intestinal production of apo-a-i and apo-a-iv continues relatively unchanged while the production of apo-b is reduced substantially. The most reliable and convincing results were those obtained by Method I, the double-label approach. An inability to quantitate the preferential isotope uptake by liver or intestine prevents a precise estimate of the intestinal apolipoprotein contribution; however, the derived r values can be used to calculate upper and lower limits for such measurements (Table V). More importantly, the r values themselves serve as a direct comparison of the intestinal contribution of an apolipoprotein under different physiological conditions, provided that the r values for the intestinal reference proteins remain constant, a condition fulfilled in the present study (Table 11). The r value obtained for a plasma apolipoprotein will be influenced neither by the route taken by an intestinally-made protein in reaching the blood nor by losses of apolipoprotein during ultracentrifugal lipoprotein isolation. These are important advantages for this method as presently employed since the route of delivery and losses during ultracentrifugation were both affected by the various animal treatments used. The r value for apo-a-i was similar for all animal treatments (Table 11), indicating that a constant proportion of the plasma apo-a-i was intestinally derived, irrespective of the extent of intestinal fat transport. Even the elimination of dietary fat for 12 days and, in addition, the elimination of endogenous, biliary fat from the intestine for 16 h failed to reduce the intestinal apo-a-i contribution. For apo-a-iv, the r values showed that the intestinal contribution was slightly reduced in the bilediverted rats on the fat-free diet (Group C), but that it was not reduced by withholding only dietary fat for 16 h (Group B). Since the total amount of these apolipoproteins synthesized by intestine plus liver in intact rats remained relatively unchanged (Table 111), available data indicate that the absolute rates of intestinal biosynthesis of apo-a-i and apo-a-iv are approximately the same during fat transport as at 16 h after withdrawal of dietary fat. For proteins with r values approaching 1.00 and derived largely from the liver, such as apo-b, apo-c, and apo-e, small changes in fractional organ contribution are difficult to measure. A reduced intestinal contribution of apo-b in the bile-diverted animals was apparent, however. Results with Method I1 corroborated the findings with Method I, once best-estimate corrections were made for the greater loss of apolipoproteins during ultracentrifugation. Segments of small intestine in situ from fat-deprived rats incor- Plasma Apolipoprotein Biosynthesis by Intestine 3015 porated approximately the same amount of labeled leucine into apolipoproteins as did segments transporting large amounts of fat, and -80% of the radioactivity was incorporated into apo-a-i and apo-a-iv (Table 111). The fractional incorporation of leucine into apo-b was reduced in the absence of fat. It should be noted that in addition to being essentially fat-free, the purified diet fed to Group C rats differed in many other respects from the stock diet initially fed to Group B animals (Table I). Still, the results for the two groups were similar, indicating little response to the other dietary variables. Multiple regression analysis of the data from the isolated organs and intact rats indicates that -50% of the plasma apo- A-I and apo-a-iv was still intestinally derived in animals h after all fat was removed from the diet (Table V). Withholding fat reduced the intestinal apo-b contribution from 16 to -5% of the total in the plasma. Consistent with our conclusions, the administration of a fat meal to previously fasted rats or human subjects produces little or no increase in the plasma concentrations of apo-a-i (12, 18) or apo-a-iv (19), measured immunochemically. Meanwhile, other immunochemical findings (12, 18-24) have been taken to suggest an increased intestinal biosynthesis of these apolipoproteins during fat absorption. Alternative inter- pretations are, however, possible. Firstly, fat feeding was shown to increase the output of apo-a-i in the mesenteric lymph of rats (12,20) and the output of apo-a-i (21) and apo- A-IV (19) in the urine of human subjects with chyluria. The apo-a-i output into lymph was increased less than 2-fold, while the triacylglycerol output was increased 20-fold (20). Since fat absorption reduces markedly the direct release of newly-synthesized intestinal apolipoproteins into the blood and increases the proportion appearing in mesenteric lymph (Table IV), the observed greater output into lymph does not necessarily indicate greater intestinal production. In other studies with cultured human duodenojejunal mucosa, the rate of apo-a-i release into the medium, but not necessarily apo- A-I biosynthesis, was accelerated by the addition of a micellar lipid solution (22). The damaging effects of the bile salts in the solution possibly increased the rate of protein loss from the tissue. Finally, an increase in cellular immunofluorescence in histological studies with specific antisera have suggested that there are larger amounts of apo-b (18, 23, 24), apo-a-i (12, 18), and apo-a-iv (19) within intestinal mucosal cells of man acd rats within minutes after feeding fat to previously fasted subjects. The heightened immunofluorescence may be due in part, however, to the observed changes in intracellular apolipoprotein distribution (see also, Ref. 25) or to altered apolipoprotein association with the increased amounts of intracellular fat. Quantification by radioimmunoassay suggests that fat feeding does, in fact, augment the cellular content of apo-a-i (181, but similar methods have indicated both an increase (18) and a decrease (26) in cellular apo-b. In any case, an intracellular accumulation of apolipoproteins would not require an increase in the rate of their biosynthesis. Thus, the results of the immunochemical studies are not incompatible with our biochemical findings. The increased rate of intestinal apo-b production accompanying fat absorption is in accord with the apparent role of this apolipoprotein in transporting esterified lipids out of cells and into the circulation, a role deduced from findings in humans with an inherited absence of apo-b and a defect in transporting lipids from both liver and intestine (27). In perfused rat livers, however, Witztum and Schonfeld (28) found no correlation between the net amounts of triacylglycerols and apo-b released into the recycling perfusate. At least two forms of apo-b different in apparent molecular weight are found in the circulation. Intestine produces only the smaller

5 3016 Plasma Apolipoprotein Biosynthesis by Intestine of the apo-b variants' (29, 30), but in the rat, both forms are produced by the liver? It wili be important to determine what physiological factors regulate the production of each variant by the liver and what metabolic properties are conferred on a lipoprotein by the presence of each of the apo-b variants. In contrast to apo-b, apo-a-i and apo-a-iv production by intestine does not appear to be regulated by the rate of fat transport, indicating, perhaps, that these two proteins do not play a functional role in chylomicron assembly or release. A similar dissociation can be observed in the liver by feeding the animal with orotic acid, which totally inhibits release of apo- B and triacylglycerols (31), but has no effect on the production of apo-a-i (32). Apo-A-I and apo-a-jv newly synthesized by intestine become associated with all lipoprotein classes in mesenteric lymph (13). During fat transport, a high proportion of both proteins is associated with lymph chylomicrons. Soon after these particles reach the blood, the apo-a-i and apo-a-iv are transferred to the high density lipoproteins (33-36). In the absence of intestinal fat transport, a greater proportion of both proteins bypasses the lymph and enters the blood directly from the small intestine (Table IV). Whether they penetrate the blood capillaries as free proteins or as high density lipoproteins, possibly in discoidal form (37), remains unknown. Conceivably, they serve as vehicles whereby small amounts of lipophilic substances gain rapid access from intestine to the blood. This could help to explain the low rate of long-chain fatty acid transport remaining across intestine of patients with abetalipoproteinemia (27). Apo-A-IV is a prominent protein in rat high density lipoproteins (7) and in the lipoprotein-free portion of rat plasma (38). An analogous protein is found in human and dog mesenteric lymph (39) and in human plasma (40,41), where most of it appears to be unassociated with lipoproteins (19,40). The same holds true for rat mesenteric 1-ymph, particularly lymph devoid of chylomicrons. A physiologicalroleforapo-a-iv remains unknown. Apo-A-I is the predominant protein of rat and human high density lipoproteins and is a cofactor for lecithin-cholesterol acyltransferase, the cholesterol esterifying enzyme in plasma (42, 43). Recent studies have identified an important role for the small intestine in the metabolism of plasma glutamine (44). The present studies reveal another metabolic function of the small intestine apparently not associated with the digestion and absorption of nutrients, the continuous production of two of the plasma apolipoproteins. Acknowledgments-We thank Albert E. Spaeth for his expert technical assistance and Dr. Mones Berman for the use of his computer. REFERENCES 1. Wu, A.-L., and Windmueller, H. G. (1980) Fed. Proc. 39, Schaefer, E. J., Eisenberg, S., and Levy, R. I. (1978) J. Lipid Res. 19, Smith,%. C., Pownd, H. J., and Gotto, A. M., Jr. (1978) Annu. Rev. Biochem. 47, Bersot, T. P., Brown, W. V., Levy, R. I., WindmueUer, H. G., Fredrickson, D. S., and LeQuire, V. S. (1970) Biochemistry 9, Koga, S., Bolis, L., and Scanu, A. M. (1971) Biochim. Biophys. Acta 236, Herbert, P. N., Windmueller, H. G., Bersot, T. P., and Shulman, R. S. (1974) J. Biol. Chem. 249, Swaney, J. B., Braithwaite, F., and Eder, H. A. (1977) Biochem- * Wu, A,-L., and Windmueller, H. G. (1981) J. Biol. Chem. 256, in press. istry 16, Wu, A.-L., and Windmueller, H. G. (1979) J. Biol. Chem. 254, Marsh, J. B. (1976) J. Lkid Res. 17, Felker, T. E., Fainaru, M., Hamilton, R. L., and Havel, R. J. (1977) J. Lipid Res. 18, Windmueller, H. G., Herbert, P. N., and Levy, R. I. (1973) J. Lipid Res. 14, Glickman, R. M., and Green, P. H. R. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, Wu,A.-L., and Windmueller, H. G. (1978) J. Biol. Chem. 253, Report of the American Institute of Nutrition Ad Hoc Committee on Standards for Nutritional Studies (1977) J. Nutr. 107, Windmueller, H. G., and Spaeth, A. E. (1975) Arch. Biochem. Biophys. 171, Mortimore, G. E. (1963) Am. J. Physwl. 204, Windmueller, H. G., and Spaeth, A. E. (1972) J. Lcpid Res. 13, Schonfeld, G., Bell, E., and Alpers, D. H. (1978) J. Clin. Inuest. 61, Green, P. H. R., Glickman, R. M., Riley, J. W., and Quinet, E. (1980) J. Clin. Invest. 65, Imaizumi, K., Havel, R. J., Fainaru, M., and Vigne, J.-L. (1978) J. Lipid Res. 19, Green, P. H. R., Glickman, R. M., Saudek, C. D., Blum, C. B., and Tall, A. R. (1979) J. Clin. Invest. 64, Rachmilewitz, D., and Fainaru, M. (1979) Metabolism 28, Glickman, R. M., Khorana, J., and Kilgore, A. (1976) Science 193, Glickman, R, M., Green, P. H.R., Lees, R. S., Lux, S. E., and Kilgore, A. (1979) Gastroenterology 76, Schonfeld, G., Grimme, N., and Alpers, D. (1980) J. Cell Biol. 86, Rachmilewitz, D., Albers, J. J., and Saunders, D. R. (1976) J. Clin. Invest. 57, Herbert, P. N., Gotto, A. M., and Fredrickson, D. S. (1978) in The Metabolic Basis of Inherited Disease (Stanbury, J. B., Wyngaarden, J. B., and Fredrickson, D. S., eds) pp , MC- Graw-Hill, New York 28. Witztum, J. L., and Schonfeld, G. (1978) Diabetes 27, Krishnaiah, K. V., Ong, J., Walker, L. F., Borensztajn, J., and Getz, G. S. (1978) Fed. Proc. 37, Kane, J. P., Hardman, D. A., and Paulus, H. E. (1980) Proc. Natl. Acad. Sci. U. S. A. 77, Windmueller, H. G., and Levy, R. I. (1967) J. Biol. Chem. 242, Chao, Y.-S., Windler, E. E., Chen, G. C., and Havel, R. J. (1979) J. Biol. Chem. 254, Rodbell, M., Fredrickson, D. S., and Ono, K. (1959) J. Biol. Chem. 234, Schaefer, E. J., Jenkins, L. L., and Brewer, H. B., Jr. (1978) Biochem. Biophys. Res. Cornmun. 80, Robinson, S. F., and Quarfordt, S. H. (1978) Biochim. Biophys. Acta 541, Tall, A. R., Green, P. H. R., Glickman, R. M., and Riley, J. W. (1979) J. Clin. Invest. 64, Green. P. H. R.. Tall, A. R., and Glickman, R. M. (1978) J. Clin. Invest. 61, Roheim. P. S.. Edelstein, D., and Pinter. G. G. (1976) Proc. Natl. Acad.' Sci. U. S. A. 73, Weisgraber, K. H., Bersot, T. P., and Mahley, R. W. (1978) Biochem. Biophys. Res. Commun. 85, Beisiegel, U., and Utermann, G. (1979) Eur. J. Biochem. 93, Utermann, G., and Beisiegel, U. (1979) Eur. J. Biochem. 99, Fielding, C. J., Shore, V. G., and Fielding, P. E. (1972) Biochem. Biophys. Res. Commun. 46, Soutar, A. K., Garner, C.W., Baker, H. N., Sparrow, J. T., Jackson, R. L., Gotto, A. M., and Smith, L. C. (1975) Biochem- istv 14, Windmueller, H. G., and Spaeth, A. E. (1974) J. Biol. Chem. 249,