3T~ whom correspondence should be addressed.

Contribution of apob-48 and apob-1 triglyeriderih lipoproteins (TRL) to postprandial inreases in the plasma onentration of TRL triglyerides and retinyl esters Jeffrey S. Cohn,'.* Elizabeth J. Johnson,' John S. Millar: Susan D. Cohn,' Ross W. Milne,',t Yves L. Marel,'.t Robert M. Russell,' and Ernst J. Shaefer3.* Lipid Metabolism Laboratory and Gastrointestinal Nutrition Laboratory,' USDA Human Nutrition Researh Center on Aging at Tufts University, 711 Washington Street, Boston, MA 2111, and Clinial Researh Institute of Montreal,t 11 Pine Avenue West, Montreal, Quebe H2W 1R7, Canada Abstrat After the ingestion of a fat-rih meal, there is a postprandial inrease in the plasma onentration of both apolipoprotein B-48- and apob-1-ontaining triglyeride-rih lipoproteins (apob-48 and apob-1 TRL). In order to determine the ontribution of these lipoproteins to postprandial lipemia, the onentration of triglyerides ('E) and retinyl esters (RE) was measured in apob-48 and apob-1 TRL after an oral fat load. Six normolipidemi male subjets were fed heavy ream (1 g fat per kg body weight) ontaining vitamin A (3 retinol equivalents). TRL were isolated by ultraentrifugation from plasma samples obtained at regular intervals after the meal, and apob-1 TRL were separated from apob-48 TRL by affinity hromatography using monolonal antibodies. Postprandial inrease in plasma TG onentration was due to an inrease in TG in the TRL fration, whih in turn was predominantly (82 f 4%) due to an inrease in TG in apob-48 TRL. Contribution of apob-1 TRL to postprandial inrease in TRL TG was 3-27% in individual subjets. ApoB-1 TRL remained a signifiant arrier of total plasma triglyeride in the fed state, as refleted by similar apob-1 and apob-48 TRL TG onentrations at 2, 4, and 6 h after the fat meal. Retinyl esters were regularly deteted in apob-1 TRL. Seventy-five (f 9) perent of the inrease in TRL-RE was due to RE in apob-48 TRL and 25 + 9% was due to RE in apob-1. These data suggest that RE in plasma are not always assoiated with apob-48- ontaining lipoproteins. Furthermore, we onlude that apob-1 TRL, as well as apob-48 TRL, make a signifiant ontribution to postprandial trig1yeridemia.-cohn, J. s., E. J. Johnson, J. S. Millar, S. D. Cohn, R. W. Milne, Y. L. Marel, R. M. Russell, and E. J. Shaefer. Contribution of apob-48 and apob-1 triglyeride-rih lipoproteins (TRL) to postprandial inreases in the plasma onentration of TRL triglyerides and retinyl esters. J. Lipid Res. 1993. 34: 233-24. Supplementary key words postprandial lipemia - atherogenesis It has been proposed that the interation of triglyeriderih lipoproteins (TRL) Of and liver Origin with lipoprotein lipase and ells of the artery wall onstitutes an atherogeni proess (1, 2). This onept is supported by studies showing that partially delipidated very low density lipoproteins (VLDL) and hylomirons are able to load ultured smooth musle ells and marophages with esterified holesterol (3-5). The doumented in vitro atherogeniity of postprandial lipoproteins has led to numerous in vivo studies doumenting the qualitative and quantitative hanges of plasma lipoproteins in the fed state (as reviewed in refs. 6 and 7). We have shown in previous studies that after the ingestion of a fat-rih meal there is a postprandial inrease in the plasma onentration of both apob-48- and apob-1-ontaining TRL (8). These lipoprotein speies ontribute to both early (-6 h postprandially) and late (6-12 h) hanges in plasma triglyeride (Z) onentration (9). The rate of TRL apob-1 prodution is also inreased in the fed ompared to the fasted state, as measured by the in vivo rate of inorporation of intravenously administered [D,]-L-leuine into apob-1 in the d < 1.6 g/ml fration of plasma (1). In order to substantiate the onept that both apob-48 and apob-1 TRL ontribute to postprandial triglyeridemia, we have assessed the ontribution of these lipoproteins to postprandial inreases in the plasma onentration of TRL TG. Subjets in the present study were also fed vitamin A in order to measure the distribution of retinyl esters (RE) Abbreviations: TRL, triglyeride-rih lipoproteins; VLDL, very low density lipoproteins; LDL, low density lipoproteins; HDL, high density lipoproteins; TG, triglyeride; RE, retinyl ksters. 'Present address: Clinial Researh Institute of Montreal, 11 Pine Avenue West, Montreal Quebe H*W lr77 Canada. *Present address: University of Ottawa Heart Institute, 4th Floor, ROO^ H453, 153 Carling Avenue, Ottawa Ontario K1Y 4E9, Canada. 3T~ whom orrespondene should be addressed. Journal of Lipid Researh Volume 34, 1993 233

between apob-48 and apob-1 TRL. Retinyl esters in the irulation have previously been used to identify the presene in plasma of lipoproteins of intestinal origin (hylomirons and hylomiron remnants). The rationale for this approah (11-14) is based on the onept that dietary vitamin A is esterified in the intestine and is inorporated into the ore of hylomiron partiles. These lipoproteins are sereted into intestinal lymph and their omponent TG are hydrolyzed by lipoprotein lipase in the irulation. It is believed that the RE remain assoiated with hylomirons during lipolysis and are taken up by the liver within hylomiron remnants via a reeptormediated proess. Evidene suggests that the liver does not reserete these RE, and they are either stored or resereted as unesterified retinol bound to retinol-binding protein. As plasma exhange of RE between lipoproteins is believed to be minimal and RE are not resereted into the irulation, it is postulated that RE in plasma are appropriate markers for lipoproteins ontaining apob-48 of intestinal origin. Cirumstantial evidene has been presented, however, suggesting that plasma RE are not always assoiated with apob-48-ontaining lipoproteins (15). Postprandial hanges in the plasma onentration of TRL apob-48 are not always mimiked by hanges in TRL RE onentration. Furthermore, in the fasting state and 12 h after a fat-rih meal, a signifiant proportion of plasma RE in normolipidemi subjets is found in the low density lipoprotein (LDL) and high density lipoprotein (HDL) frations, apparently unassoiated with apob-48. The seond purpose of the present experiments was, therefore, to diretly examine the extent of RE assoiated with apob-48 in the TRL fration. Subjets METHODS Six healthy male subjets (3 k 1 years, mean k SEM) who were of average height and weight were reruited. They had normal fasting lipid levels (holesterol: 4.4.32 mmolll; triglyerides:.86 f.9 mmol/l; HDL holesterol: 1.24 k.12 mmol/l). They were not taking vitamin A or mediations known to affet plasma lipids. All studies were onduted in the Metaboli Researh Unit at the U.S. Department of Agriulture Human Nutrition Researh Center on Aging at Tufts University. Informed onsent was obtained from all volunteers under the guidelines established by the Human Investigation Review Committee of the New England Medial Center and Tufts University. Fat-feeding protool After a 12-h overnight fast, subjets were given an oral fat load (1 g fat/kg body weight) in the form of heavy ream, together with 3 retinol equivalents of vitamin A (three times the reommended daily allowane). Th vitamin was given as retinyl palmitate in orn oil (PlMO, a gift from Hoffman LaRohe In., Nutley, NJ). Blood samples (1 ml) were obtained via a small forearm indwelling atheter prior to the fat load and at 2, 4, 6, 9, and 12 h thereafter. Blood samples were olleted in tubes ontaining ethylenediaminetetraaetate (EDTA) to give a final onentration of.1% EDTA, and were proteted from light with the use of aluminium foil. Water, but no food, was allowed during the ourse of the study. Lipoprotein separation Plasma was separated from red blood ells by ultraentrifugation at 1 g for 15 min at 4OC. TRL frations (ontaining hylomirons, very low density lipoproteins, and intermediate density lipoproteins) were isolated from 5 ml of plasma by a single ultraentrifugal spin (39, rpm, 18 h, 4OC) at d 1.19 g/ml in a Bekman 5.3 Ti rotor. Fration volumes were adjusted to 3. ml with normal saline and were assayed for protein by the method of Lowry et al. (16), using bovine serum albumin as a standard. Turbidity was leared with hloroform. Freshly isolated TRL frations were subjeted to affinity hromatography, using speifi apob-1 monolonal antibodies 4G3 and 5El1, whih do not rossreat with apob-48 (17, 18). These antibodies have previously been used to haraterize the very low density lipoprotein fration of Type I11 and Type IV hyperlipoproteinemi subjets (19, 2). Antibodies (mixed together in a 1:l protein ratio) were oupled to ativated Sepharose 4B (Pharmaia Fine Chemials, In., Uppsala, Sweden) as desribed previously (17). Routinely, 16 mg of 4G3 protein and 16 mg of 5Ell protein were oupled to 8 ml of ativated Sepharose. Binding apaity of the gel was.75-1. mg TRL protein/ml gel. Separation of apob-48 and apob-1 frations was arried out in 1.5-ml Eppendorf tubes. Sepharose gel (1.2 ml) in phosphate-buffered saline (PBS, ph 7.4,.2% NaN,) was aliquoted into tubes and entrifuged for 5 min at 1, rpm. The supernate was aspirated and TRL (3 pg protein) was added to.6 ml of paked gel. Tubes were vortexed gently, roked for 1 h at room temperature on an osillating platform, and entrifuged, and the supernate was olleted (approximately.5 ml was retrieved). The gel was washed with 1. ml PBS, roked for 15-2 min, and after entrifugation, the supernate was aspirated and added to the first supernate. The washing proedure was repeated with a seond volume of PBS (final volume of fration: 2.5 ml). This sample (the unbound fration) ontained the apob-48 TRL. Lipoproteins bound to the gel were dissoiated by the addition of 3 M Na thioyanate (1. ml). Samples were roked for 1 h at room temperature, entrifuged, and the supernate was olleted. The gel was washed with two further 1.-ml aliquots of Na thioyanate, and these washes were added to the first 234 Journal of Lipid Researh Volume 34, 1993

thioyanate supernate to give the bound fration, ontaining apob-1 TRL. Reovery of initial TRL added to the affinity gel was 92 k 4% as estimated by measuring the reovered protein in the bound and unbound frations. The adequay of lipoprotein separation was assessed by SDS polyarylamide gel eletrophoresis using 4-22.5 % arylamide gradient gels as previously desribed (8). Nondelipidated lipoprotein samples (3 pg protein), redued in SDS sample buffer ontaining 3% meraptoethanol, were routinely loaded onto the 1.5-mm thik vertial slab gels. The apob-48 (unbound) frations were onentrated before eletrophoresis to allow for adequate visualization of the apob-48 protein band. The apob-1 band of TRL omigrated with the major stainable protein of a narrow-ut LDL fration (1.4 < d < 1.5 g/ml). The apob-48 band of TRL omigrated with the major high moleular weight band of lymph hylomirons. The identity of the apob bands was onfirmed by Western blotting analysis using a polylonal antibody that was known to reat with both apob-48 and apob-1. Lipid and lipoprotein analyses Plasma and lipoprotein frations were assayed for total holesterol and TG with an Abbott Diagnostis ABA-ZOO bihromati analyzer using enzymati reagents (21). Assays were standardized through partiipation in the Centers for Disease Control - National Heart, Lung and Blood Institute Standardization program. All lipid measurements (inluding separated TRL frations) were arried out in dupliate. HDL holesterol was quantitated by analyzing the supernate obtained by the preipitation of a plasma aliquot with dextran-mg*+ (22). Retinyl esters were measured under red lights by high-performane liquid hromatography using retinyl aetate as an internal standard (23). Using this method, all the major plasma RE (retinyl palmitate, stearate, and oleate) oeluted together and appeared as a single peak on the hromatograms. Statistis Paired t-tests were used to determine the statistial signifiane between lipid measurements obtained from fasting ( h) and postprandial samples. Postprandial inrease in TRL TG and RE was measured by planimetry as the area under the onentration-time urve, where the baseline was the zero-time (fasting) onentration. Area measurements are expressed in the units of mmol h/l. RESULTS Mean (k SEM) plasma holesterol, TG, and RE onentrations before and after the ingestion of the fat load are shown in Fig. 1. Mean plasma holesterol onentration did not hange signifiantly during the ourse of the experiment. This is onsistent with our previous studies I 1 1 1 1 1 1, 1 1 1 1 1 I 2 4 6 8 1 1 2 Time After Ingestion of Cream (hours) - 3-2 - 1 Fig. 1. Postprandial hanges in the plasma onentration of holesterol, triglyerides, and retinyl esters. Signifiantly different from fasting onentration by paired t-test: *P <.5; **P <.1. (24) and those of others (6). Plasma TG onentration peaked either 2 or 4 h after the fat load in individual subjets, and the mean plasma TG onentration was signifiantly (P <.5) elevated above fasting onentration 2 and 4 h postprandially. Mean plasma RE onentration peaked at 6 h and was signifiantly greater than baseline at all postprandial timepoints. As shown previously (9, 15) and as refleted by the different time ourse of the urves for the mean data (Fig. l), hange in plasma RE onentration ourred after hange in plasma TG in all subjets. TRL frations (d < 1.19 g/ml) were isolated by ultraentrifugation and were separated into apob-48 and apob-1 TRL speies by affinity hromatography. The apolipoprotein omposition of bound and unbound frations is shown in Fig. 2. Very adequate separation of apob-48- and apob-1-ontaining frations was ahieved. ApoB-48 was not detetable in the bound (apob-1-ontaining) frations. A faint apob-1 band was often observed in the unbound frations (see right hand gel of Fig. 2). Minor ontamination of the unbound fration has been noted previously using these antibodies (19, 2). Additional hromatography was not arried out to remove ontaminating apob-1 material, as this was found to adversely affet reovery. Laser sanning densitometry showed that less than 1% of the apolipoprotein omposition of apob-48 TRL was apob-1 (assuming equal hromogeniity of apolipoprotein bands). Two perent of apob-48 TRL protein was apob-48, 15% was albumin, 8% was 6-2-glyoprotein l, 21% was apoa-iv, 2% was apoe, 19% was apoa-i, and 34% was C- apolipoproteins. In apob-1 TRL, 35% of total protein by gel sanning was apob-1, 15% was albumin, 1% Cohn et af. Postprandial ontribution of apob-48 and apob-1 TRL 235

Albumin - - ApoA-IV - ApoE ApoA-l - APOCS -- Fig. 2. Apolipoprotein omposition of TRL frations separated by afinity hromatography using monolonal antibodies against apor-1. The hound fration ontaininq apor-1 is on the left, and the unbound fration ontaining apor-48 is on the right. Round TRL ontained relatively more apoe and less apoa-iv and epoa-i than unbound TRL. The unbound frartion was onentrated before loading onto the gel. Trae amounts of apor-1 were routinely deteted in the unbound fration. was /3-2-glvoprotein 1, 1% was apoa-iv, 9% was apoe, and 31% was C-apolipoproteins. In the fasting state (at zero time), mean plasma TG onentration was.86 +.9 mmol/l and mean TRL TG onentration was.54.11 mmol/l (representing 61 + 6% of total plasma triglyeride). In the fed state at A z E E v.- L E L ) V 1. -.8 -.6 -.4 - the timepoint of maximal TG inrease (2 h postprandially), mean plasma TG onentration was 1.26 k.19 mmol/l and mean TRL TG onentration was 1. k.19 mmol/l. A mean 77 k 3% of plasma TG was thus ontained in the TRL fration 2 h after the fat load. Mean TRL TG was signifiantly inreased above fasting onentration 2 and 4 h after the fat load (Fig. 3) in ' aord with the total plasma TG data. These results are omparable to those obtained in male subjets of different ages, reported previously (8). The plasma TG onentration of the apor-48 and apor-1 frations before and after the fat load is shown in Fig. 3. Postprandial inrease in plasma TG onentration was predominantly due to inrease in apor-48 TRL, whih was signifiantly inreased 2, 4, and 6 h after the meal. A smaller but signifiant inrease was observed in apor-1 TRL TG. The ontribution of apor-48 and apor-1 TRL frations to postprandial inrease in TRL TG (Table 1) was quantitated by measuring the area under individual TG response urves (fasting [ h] onentration taken as baseline). Inrease in apor-48 TRL TG was five times greater than inrease in apor-1 TRL TG (2.44 k.45 vs..47 k.15 mmol h/l). Eighty-two (k 4) perent of the postprandial inrease in TRL TG was thus due to apor-48 TRL (range: 73-97%). The ontribution of apor-1 TRL was not insignifiant, aounting for 3-27% of total inrease in TRL TG in individual subjets (mean k SEM: 18 k 4%). Total area (as opposed to inremental area) was also measured under the TRL TG response urves (zero onentration used as baseline). The apor-48 TRL TG total area was 3.96 k.48 mmol h/l and the apob-1 TRL TG area was 4.64 k.75 mmol h/l. This means that although apor-48 TRL were most responsible for the inrease in TRL TG onentration after the fat load, in absolute terms apob-1 TRL were the predominant arriers of plasma TG during the total 12 h postprandial period. RE were measured in TRL frations isolated at different postprandial timepoints. It is signifiant that we deteted RE in nearly all apor-1 TRL, as well as apob-48 TRL samples. The mean plasma RE onentrations of TABLE 1. Contribution of apor-1 TRL and apor-48 TRL to postprandial inreases in the plasma onentration of trixlyeride and retinyl ester.2 - lnrrraw in Conentration' ApoR-IO TRI. ApoR-4R TRL mmol- ha. I 2 4 6 8 112 Time After Ingestion of Cream (hours) Fig. 3. Postprandial hanges in the triglyeride onentration of TRL frations. Signifiantly different from fasting onentration by paired I- test: 'P <.5; **P <.1. Triglyeride.47 i.15 (18%Ib 2.44 i.45 (82%)b Retinyl ester 333 i 129 (25%) 889 i 169 (75%)' 'Measured as the area between the postprandial response urves and a baseline drawn through the fasting onentration. 'Values in parentheses represent inreases expressed as a perentqe of the total. 236 Journal of Lipid Researh Volume 34, 1993

the total, apob-48, and apob-1 TRL frations before and after the fat load are shown in Fig. 4. Total TRL RE onentration was signifiantly elevated 2, 4, 6, and 9 h after the fat load, whih mimiked the postprandial hanges in total plasma RE onentration (Fig. 1). Inreases in both apob-48 TRL and apob-1 TRL RE were responsible for the hange in total TRL RE. RE onentration peaked on average at 6 h in the apob-48 TRL fration, whereas the maximum RE onentration in apob-1 TRL tended to our at later timepoints (at 9 h for the mean data). The ontribution of apob-1 TRL and apob-48 TRL to postprandial inreases in plasma TRL RE onentration was quantitated by planimetry and mean area measurements are shown in Table 1. Inrease in RE onentration postprandially tended to be 2- to 3-times greater in apob-48 TRL than in apob-1 TRL. Thus 75 * 9% of TRL RE inrease was due to apob-48 TRL and 25 f 9% was due to apob-1 TRL. It is worth noting that the ontribution of apob-1 TRL varied from one subjet to another, and was quite signifiant in some individuals (range: 6-56%). Cholesterol onentrations were also measured in the different TRL frations and these results are shown in Fig. 5. A postprandial inrease in total TRL holesterol was not observed in every subjet (as we have observed in previous studies (8)). In this group of subjets, who had relatively low plasma holesterol onentrations, a derease in the mean total TRL holesterol below fasting onentration was, in fat, observed at later timepoints. This derease was predominantly due to a fall in the postprandial onentration of apob-1 TRL holesterol. In h 25 -\ E" E. 2 C.- 2-15 E 8 1 G v) w -.- )r 5 2 T 1 1 1 1 1 1 1 1 1 1 1 1 1 2 4 6 8112 Time After Ingestion of Cream (hours) Fig. 4. Postprandial hanges in the retinyl ester onentration of TRL frations. Signifiantly different from fasting onentration by paired f- test: 'P <.5..6 4 1.5.4.3.2.1. 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 4 6 8 1 1 2 Time After Ingestion of Cream (hours) Fig. 5. Postprandial hanges in the holesterol onentration of TRL frations. Signifiandy different from fasting onentration by paired t- test: 'P <.5. ontrast, mean apob-48 TRL holesterol was signifiantly inreased 2 and 4 h after the fat load. DISCUSSION Speifi monolonal antibodies for apob-1 have been used before (19, 2) to isolate and haraterize the very low density lipoprotein (VLDL) fration of dyslipidemi subjets in the fasted state. The same antibodies were used in the present study to isolate the apob-48- and apob-1-ontaining TRL of plasma obtained from subjets in the fed state. In order to obtain reliable quantitative data with these antibodies, we have separated single sample aliquots in individual Eppendorf tubes (see Methods), ontaining antibodies bound to Sepharose, rather than eluting frations from affinity hromatography olumns, as desribed previously (19). In addition, sodium thioyanate rather than itri aid was used to separate the bound TRL fration from the affinity hromatography gels. These modifiations resulted in aeptable reovery (85-95%) of TRL samples and allowed for the plasma onentration of different frations to be estimated. Adequay of separation was also ontrolled as losely as possible by monitoring the apolipoprotein omposition of isolated frations by polyarylamide gel eletrophoresis (Fig. 2). Some ontamination of apob-48 frations was noted (see Methods), leading to a small underestimation of apob-1 TRL lipid onentrations. This underestimation, if taken into aount, would only Cohn et al. Postprandial ontribution of apob-48 and apob-1 TRL 237

inrease the ontribution of apob-1 TRL and enhane the signifiane of the present results. We have shown that inrease in plasma triglyeride onentration after an oral fat load is predominantly due to triglyeride ontained within apob-48-ontaining lipoproteins in the TRL fration of plasma. In normolipidemi male subjets, we have found that 82 f 4% of the postprandial inrease in TRL TG was attributable to triglyeride in apob-48 TRL. This is onsistent with the established onept that dietary fat is inorporated into large triglyeride-rih hylomirons in the intestine that have apob-48 as their major strutural apolipoprotein (25). These lipoproteins are sereted into intestinal lymph and then into the irulation where their triglyerides are hydrolyzed by lipoprotein lipase. ApoB-48-ontaining lipoproteins in the irulation, therefore, represent a spetrum of partially atabolized lipoproteins of intestinal origin, whih make a signifiant ontribution to postprandial triglyeridemia. Although the inrease in plasma TRL 7G onentration in the fed state was predominantly due to an inrease in apob-48 TRL, the ontribution of apob-1 TRL was not insignifiant; 18 f 4% of the total TRL TG inrease following the oral fat load was attributable to apob-1 TRL, and this varied among individual subjets (range: 3-27%). In absolute terms, apob-1 TRL were responsible for transporting a signifiant proportion of plasma X in the postprandial state, as refleted by the relatively similar onentrations of TG in apob-1 and apob-48 TRL at 2, 4, and 6 h after the fat load (Fig. 3), and as refleted by the total area under the TRL response urves for the 12-h period after the meal (4.64.75 vs. 3.96 f.48 mmol - h/l for apob-1 TRL and apob-48 TRL, respetively). Similar data have reently been presented by Shneeman et al. (26). These results suggest that apob-1 TRL make a signifiant ontribution to the total plasma triglyeride onentration in the fed as well as in the fasted state. There are at least three possible explanations for the inrease in the onentration of apob-1 TRL triglyeride in postprandial plasma. I) The plasma learane of apob-1 TRL is inhibited by the influx into plasma of postprandial hylomirons; 2) the ingestion of a fat-rih meal stimulates the synthesis and seretion of apob-1 TRL from the intestine; and 3) hepati apob-1 TRL are sereted postprandially in response to lipid of dietary origin reahing the liver via the portal vein or via hylomiron transport. The first possibility is supported by the studies of Brunzell et al. (27) showing that hylomirons and VLDL are atabolized by a ommon pathway. Triglyerides in both hylomirons and VLDL are hydrolyzed by lipoprotein lipase at the apillary endothelial surfae and these lipoproteins an ompete for enzyme-mediated lipolysis. Redued availability of lipolyti ativity due to the postprandial presene of apob-48- ontaining hylomirons ould explain the observed inrease in apob-1 TRL triglyeride. An inrease in the prodution of apob-1 TRL is, however, an equally possible explanation. We have previously found that in the fed state there is an inrease (5% on average) in the plasma onentration of apob-1 in the TRL fration (8), as well as an inrease in the rate of prodution of TRL apob-1 (1). We have suggested that this represents a postprandial inrease in postprandial synthesis and seretion of TRL by the liver, as it is normally assumed that apob-1 in plasma is of hepati origin. Data from a reent study in whih apob-1 epitopes were determined in human subjets before and after liver transplantation (28) support the hepati origin of irulating apob-1. Other lines of evidene, however, have suggested that the human intestine has the apaity to synthesize apob-1 (29, 3), and it annot be ruled out that in the fed state apob-1 in TRL is partly of intestinal origin. Irrespetive of the soure of apob-1 TRL in the fed state, and irrespetive of the mehanism for the inreased presene of apob-1 TRL in postprandial plasma, our studies suggest an additional reason why the postprandial state is potentially atherogeni. Zilversmit (1) originally proposed that atherogenesis was a postprandial phenomenon beause it involved the binding of hylomirons to the arterial surfae, the hydrolysis of triglyeride by arterial lipoprotein lipase and the subsequent internalization of holesterol-enrihed hylomiron remnants by arterial smooth musle ells. Postprandial apob-1 TRL may, however, be of equal signifiane, espeially sine a ertain proportion of TRL ontaining apob-1 are the preursors of potentially atherogeni LDL (31) and partially degraded apob-1 TRL remnants have themselves been shown to be atherogeni (2). An important finding of the present study is that RE were routinely deteted in apob-1 TRL. The presene of RE in apob-1 TRL was of greater signifiane at later postprandial timepoints (Fig. 4), and for the 12-h period as a whole 25 9% of TRL RE was ontributed by RE in apob-1 TRL. These data lend support to the irumstantial evidene that we and others have presented previously, suggesting that plasma RE are not always assoiated in plasma with apob-48-ontaining lipoproteins (12, 15, 32). There are three possible explanations for the presene of RE in apob-1 TRL. 1) RE are sereted by the intestine within hylomirons ontaining apob-1 or possibly within smaller VLDL-sized apob-1 partiles; 2) RE are sereted by the liver in VLDL ontaining apob-1; or 3) RE are transferred to apob-1 TRL in the irulation from other plasma lipoproteins. As disussed before, evidene has been presented showing that the intestine has the apaity to synthesize apob-1 (29, 3). The quantitative signifiane of this synthesis and whether it results in the seretion of a mature protein remains, however, to be substantiated. As far as hepati 238 Journal of Lipid Researh Volume 34, 1993

seretion of RE is onerned, there are data from isolated ell studies (33) and from experimental animals (34) that suggest that RE are not sereted by the liver. Nevertheless, it annot be totally ruled out that the human liver in vivo is able to serete RE inorporated into apob-1- ontaining lipoproteins. The third possibility is supported by evidene that small amounts of RE are able to transfer between plasma lipoproteins (35, 36); however several investigators onsider this transfer to be quantitatively insignifiant (12, 14, 37). Finally, plasma RE have been used as a measure of the duration and extent of postprandial lipemia in order to establish the role of postprandial lipoproteins (speifially intestinal lipoproteins) in the etiology of atheroslerosis (38-4). As suggested before, however (15), studies traing the presene of RE in plasma for relatively long periods of time after meal-feeding need to be interpreted with aution. Small amounts of RE not assoiated with the TRL fration (assoiated with LDL and HDL at later timepoints) are probably not assoiated with remnant lipoproteins. In addition, RE within the TRL fration are not always assoiated with apob-48-ontaining lipoproteins and they are therefore not neessarily indiative of TRL of intestinal origin. I Supported by grant HL 39326 from the National Institutes of Health and ontrat 53-3K-6 from the U.S. Department of Agriulture Researh Servie. Manusript reeived 19 Jub 1991, in revisedjonn 26 May 1993, and in re-revised fom 21 J.4 1993. REFERENCES 1. Zilversmit, D. B. 1979. Atherogenesis: a postprandial phenomenon. Cirulation. 6: 473-485. 2. Gianturo, S. H., and W. A. Bradley. 1988. Lipoproteinmediated ellular mehanism for atherogenesis in hypertriglyeridemia. Semin. Thmmb. Hemostasis. 14: 165-169. 3. Floren, C-H., J. J. Albers, and E. L. Bierman. 1981. Uptake of hylomiron remnants auses holesterol aumulation in ultured human arterial smooth musle ells. Biohim. Biophys. Ata. 663: 336-349. 4. Van Lenten, B. J., A. M. Fogelman, R. L. Jakson, S. Shapiro, M. E. Haberland, and P. A. Edwards. 1985. Reeptor-mediated uptake of remnant lipoproteins by holesterol-loaded human monoyte-marophages. J. Biol. Chem 26 8783-8788. 5. Huff, M. W., A. J. Evans, C. G. Sawyez, B. M. Wolfe, and P. J. Nestel. 1991. Cholesterol aumulation in 5774 marophages indued by triglyeride-rih lipoproteins. Comparison of very low density lipoprotein from subjets with type 111, IV, and V hyperlipoproteinemias. Arteriosler Thromb. 11: 221-233. 6. Tall, A. R. 1986. Metabolism of postprandial lipoproteins. MethodF Enzymol. 129: 469-482. 7. Miesenboek, G., and J. R. Patsh. 1992. Postprandial hyperlipidemia: the searh for the atherogeni lipoprotein. Cum Opin. Lipidol. 3: 196-21. 8. Cohn, J. S., J. R. MNamara, S. D. Cohn, J. M. Ordovas, and E. J. Shaefer. 1988. Plasma apolipoprotein hanges in the triglyeride-rih lipoprotein fration of human subjets fed a fat-rih meal. J. Lipid Res. 29: 925-936. 9. Cohn, J. S., J. R. MNamara, S. D. Krasinski, R. M. Russell, and E. J. Shaefer. 1989. Role of triglyeride-rih lipoproteins from the liver and intestine in the etiology of postprandial peaks in plasma triglyeride onentration. Metabolism. 38: 484-49. 1. Cohn, J. S., D. A. Wagner, S. D. Cohn, J. S. Millar, and E. J. Shaefer. 199. Measurement of very low density and low density lipoprotein apolipoprotein (apo) B-1 and high density lipoprotein apoa-i prodution in human subjets using deuterated leuine. Effet of fasting and feeding. J. Clin. Invest. 85: 84-811. 11. Hazard, W. R., and E. L. Bierman. 1976. Delayed learane of hylomiron remnants following vitamin A- ontaining oral fat loads in broad-beta disease (type 111 hyperlipoproteinemia). Metabolism. 25: 777-81. 12. Wilson, D. E., I-F. Chan, and M. Ball. 1983. Plasma lipoprotein retinoids after vitamin A feeding in normal man: minimal appearane of retinyl esters among low density lipoproteins. Metabolism. 32: 514-517. 13. Blomhoff, R. M., M. H. Green, T. Berg, and K. R. Norum. 199. Transport and storage of vitamin A. Siene. 25: 399-43. 14. Berr, F., and F. Kern, Jr. 1984. Plasma learane of hylomirons labeled with retinyl palmitate in healthy human subjets. J. Lipid Res. 25: 85-812. 15. Krasinski, S. D., J. S. Cohn, R. M. Russell, and E. J. Shaefer. 199. Postprandial plasma vitamin A metabolism in humans: a reassessment of the use of plasma retinyl esters as markers for intestinally derived hylomirons and their remnants. Metabolism. 39: 357-365. 16. Lowry,. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent.j. Bid. Chem. 193: 265-275. 17. Marel, Y. L., M. Hogue, R. Theolis, Jr., and R. W. Milne. 1982. Mapping of antigeni determinants of human apolipoprotein B using monolonal antibodies against LDL. J. Biol. Chem. 257: 13165-13168. 18. Milne, R. W., R. Theolis, Jr., R. B. Verdery, and Y. L. Marel. 1983. Charaterization of monolonal antibodies against LDL. Arterioslerosis. 3: 23-3. 19. Milne, R. W., P. K. Weeh, L. Blanhette, J. Davignon, P. I. Alaupovi, and Y. L. Marel. 1984. Isolation and haraterization of apolipoprotein B-48 and apob-1 very low density lipoproteins from type I11 hyperlipoproteinemi subjets. J. Clin. Invest. 73: 816-823. 2. Tere, F., R. W. Milne, P. K. Weeh, J. Davignon, and Y. L. Marel. 1985. Apolipoprotein B-48 and B-1 very low density lipoproteins. Comparison in dysbetalipoproteinemia (type 111) and familial hypertriglyeridemia (type IV). Arterioslerosis. 5: 21-211. 21. MNamara, J. R., and E. J. Shaefer. 1987. Automated enzymati standardized lipid analyses for plasma and lipoprotein frations. Clin. Chim. Ata. 166: 1-8. 22. Warnik, G. R., J. Benderson, and J. J. Albers. 1982. Dextran sulfate-mgz+ preipitation proedure for quantitation of high density lipoprotein holesterol. Clin. Chem. 28: 1379-1388. 23. Bankson, D. D., R. M. Russell, and J. A. Sadowski. 1986. Determination of retinyl esters and retinol in serum or Cohn et al. Postprandial ontribution of apob-48 and apob-1 TRL 239

plasma by normal-phase liquid hromatography: method and appliations. Clin. Chem. 32: 35-4. 24. Cohn, J. S., J. R. MNamara, S. D. Cohn, J. M. Ordovas, and E. J. Shaefer. 1988. Postprandial plasma lipoprotein hanges in human subjets of different ages. J. Lipid Res. 29: 469-479. 25. Bisgaier, C. L., and R. M. Glikman. 1983. Intestinal synthesis, seretion and transport of lipoproteins. Annu. Rev. Phyriol. 45: 625-636. 26. Shneeman, B. O., L. Kotite, K. M. Todd, and R. J. Havel. 1993. Relationships between the responses of triglyeriderih lipoproteins in blood plasma ontaining apolipoproteins B-48 and B-1 to a fat-ontaining meal in normolipidemi humans. Pro. Natl. Aad. Si. USA. 9: 269-273. 27. Brunzell, J. D., W. R. Hazzard, D. Porte, Jr., and E. L. Bierman. 1973. Evidene for a ommon, saturable, triglyeride removal mehanism for hylomirons and very low density lipoproteins in man. J. Clin. Invest. 52: 1578-1585. 28. Linton, M. E, R. Gish, S. T. Mubl, F. Butler, C. Esquivel, W. I. Bry, J. K. Boyles, M. R. Wardell, and S. G. Young. 1991. Phenotypes of apolipoprotein B and apolipoprotein E after liver transplantation. J Clin. Invest. 88: 27-281. 29. Hughes, T. E., W. V. Sasak, J. M. Ordovas, T. M. Forte, S. Lamon-Fava, and E. J. Shaefer. 1987. A novel ell line (Cao-2) for the study of intestinal lipoprotein synthesis. J Biol. Chem. 262: 3762-3767. 3. Hoeg, J. M., D. D. Sviridov, G. E. Tennyson, S. J. Demosky, Jr., M. S. Meng, D. Bojanovski, I. G. Safonova, V. S. Repin, M. B. Kuberger, V. N. Smirnov, K. Higuhi, R. E. Gregg, and H. B. Brewer, J. 199. Both apolipoproteins B-48 and B-1 are synthesized and sereted by the human intestine.j. Lipid Res. 31: 1761-1769. 31. Sigurdsson, G., A. Nioll, and B. Lewis. 1975. Conversion of very low density lipoprotein to low density lipoprotein. A metaboli study of apolipoprotein B kinetis in human 32. 33. 34. 35. 36. 37. 38. 39. 4. subjets. J. Clin. Znuest. 56: 1481-149. Berr, F. 1992. Charaterization of hylomiron remnant learane by retinyl palmitate label in normal humans.,/. Lipid Res. 33: 915-93. Lenih, C. M., and A. C. Ross. 1987. Chylomiron remnant-vitamin A metabolism by the human hepatoma ell line HepGZ. J Lipid Res. 28: 183-194. Thompson, K. H., L. B. Hughes, and D. B. Zilversmit. 1983. Lak of seretion of retinyl esters by livers of normal and holesterol-fed rabbits. J. Nutr. 113: 1995-21. Wilson, D. E., and I-F. Chan. 1983. In vitro transfer of hylomiron retinol and retinyl esters. Biohem. Biophys. Res. Commun. 115: 958-964. Zilversmit, D. B., R. E. Morton, L. B. Hughes, and K. H. Thompson. 1982. Exhange of retinyl and holesteryl esters between lipoproteins of rabbit plasma. Biohim. Biophys. Ata. 712: 88-93. Spreher, D. L., S. L. Knauer, D. M. Blak, L. A. Kaplan, A. A. Akeson, M. Dusing, D. Lattier, E. A. Stein, M. Rymaszewski, and D. A. Wiginton. 1991. Chylomironretinyl palmitate learane in type I hyperlipidemi families. J. Clin. Inuest. 88 985-994. Weintraub, M. S., S. Eisenberg, and J. L. Breslow. 1987. Dietary patterns of postprandial lipoprotein metabolism in normal, type IIa, type 111, and type IV hyperlipoproteinemi individuals. J Clin. Inuest. 79: 111-1119. Simpson, H. S., C. M. Williamson, T. Oliverona, S. Pringle, J. Malean, A. R. Lorimer, F. Bonnefous, Y. Bogaievsky, C. J. Pakard, and J. Shepherd. 199. Postprandial lipemia, fenofibrate and oronary artery disease. Atheroslerosis. 85: 193-22. Groot, P. H. E., W. A. H. j. van Stiphout, X. H. Krauss, H. Jansen, A. van Tol, E. van Ramshorst, S. Chin-On, A. Hofman, S. R. Cresswell, and L. Havekes. 1991. Postprandial lipoprotein metabolism in normolipidemi men with and without oronary artery disease. Arterzosler. Thromb. 11: 653-662. 24 Journal of Lipid Researh Volume 34, 1993