dietary fish oil increases conversion of very low density...

Dietary Fish Oil Increases Conversionof Very Low Density Lipoprotein Apoprotein B

to Low Density Lipoprotein

Murray W. Huff and Dawn E. Telford

Dietary fish oils, which are rich in omega-3 fatty acids, are known to produce amarked lowering of very low density lipoprotein (VLDL) trlglycerlde concentrations,but they have a less marked effect on low density lipoprotein (LDL) cholesterol. Ourprevious apollpoproteln (apo) B kinetic studies in miniature pigs demonstrated thatconversion of VLDL apo B to LDL apo B accounted for 15% to 20% of total VLDL apo Bcatabolism. In addition, 75% to 80% of LDL apo B was derived independent of plasmaVLDL or intermediate density lipoprotein (IDL) apo B catabollsm. The present studieswere carried out to determine If fish-oil diets Influenced: 1) the conversion of VLDLto LDL, and 2) the pathways of LDL apo B synthesis. Autologous 12SI-VLDL and131I-LDL were Injected Into four pigs after both a corn-oil (30 g/day for 18 days) and aMaxepa (30 g/day for 18 days) dietary period. Analysis of apo B specific activitycurves demonstrated that fish oil reduced the VLDL pool size by 38% (p<0.05) dueto an Increase in fractional catabollc rate (0.83±0.13 vs. 0.48±0.03 hr~1), as thesynthesis rate was unaffected. However, the proportion of VLDL apo B converted toLDL Increased significantly (56±7% vs. 17±3%, p<0.01) whereas the proportioncleared directly decreased (46±5% vs. 83±3%, p<0.005). Fish oil reduced total LDLapo B synthesis (0.6±0.1 vs. 1.1 ±0.2 mg/hr/kg, p<0.05). LDL-B derived Independent ofVLDL catabollsm was reduced by 90% (0.1 ±0.04 vs. 0.9±0.2 mg/hr/kg, p<0.01),whereas VLDL derived synthesis increased significantly (0.5±0.08 vs. 0.1 ±0.01,p<0.01). Although LDL apo B fractional catabollc rate decreased 22% (p<0.01), thepool size decreased 20% (p<0.05) due to the larger decrease In synthesis. Total llpldprofiling revealed no major differences in the percent composition of the main llpldclasses present In VLDL, IDL, and LDL. Thus, the fish-oil diet resulted In the secretionof a VLDL particle that Is preferentially converted to LDL This may explain theInconsistent and variable effects of fish oil on LDL concentrations observed In otherstudies. Whether LDL concentrations are Increased would depend on other factorsregulating LDL concentrations. In the present study, the LDL-B pool size was re-duced, due entirely to the marked reduction In VLDL Independent synthesis.(Arteriosclerosis 9:58-66, January/February 1989)

It is well documented that the intake of marine (fish) oils,which are rich in omega-3 polyunsaturated fatty acids,

significantly reduces plasma very taw density lipoprotein(VLDL) triglyceride concentrations.1-4 This effect is greaterthan with equal amounts of omega-6-rich vegetable oils.1-34

Nestel et al. have shown in both in vivo3 and in vitro56

studies that the reduction in VLDL concentrations associ-ated with fish-oil consumption is due to reduced synthesisof VLDL triglyceride and apoprotein (apo) B.

Since the majority of low density lipoprotein (LDL) apo Bin normal humans is derived from VLDL apo B,78 it mightbe expected that fish-oil treatment would decrease LDL

From trie Department of Medicine, University Hospital, and theRoberts Research Institute, University of Western Ontario, Lon-don, Ontario, Canada.

This work was supported by a grant from the Heart and StrokeFoundation of Ontario. Dr. Huff is a Research Scholar of theHeart and Stroke Foundation of Ontario.

Address for reprints: Dr. Murray W. Huff, Department ofMedicine, The University of Western Ontario, 4-16 RobartsResearch Institute, London, Ontario, Canada N6A 5K8.

Received November 12, 1987; revision accepted July 20,1988.

cholesterol and apo B levels. However, the effect on LDLcholesterol and apo B is inconsistent.149 Illingworth etal.10 demonstrated a reduction of LDL cholesterol and apoB with fish oil that was, indeed, due to a reduced rate ofLDL synthesis. On the other hand, in studies in whichtriglyceride-lowering effects were observed, no change orincreases in LDL cholesterol and apo B were reported.29

The reasons for this are not known, but they may berelated to the variable effects of fish oils on the complexmechanisms regulating LDL concentrations.

In normal humans, the proportion of VLDL apo B fluxconverted to LDL is variable.7811 For example, in hyper-triglyceridemic subjects, compared to normals,78 the pro-portion converted to LDL is decreased and the proportioncleared directly is increased. Thus, a reduction in VLDLapo B synthesis (flux) by fish-oil treatment could resultin a decreased conversion of VLDL apo B to LDL, adecreased direct removal of VLDL apo B, or both. It ispossible that an effect of fish oil is a reduction in directVLDL removal and that the rise in LDL observed in somestudies29 was due to an increase in VLDL converted toLDL. This possibility has not been tested.

58

by guest on July 1, 2018http://atvb.ahajournals.org/

Dow

nloaded from

http://atvb.ahajournals.org/

FISH OIL ENHANCES CONVERSION OF VLDL TO LDL Huff and Telford 59

In several animal species, including primates,12 rabbits,13

and miniature pigs,14 a large proportion of VLDL apo B iscleared directly from the circulation without conversion toLDL. Using apo B kinetic studies, we have also demon-strated in miniature pigs,14 as others have shown inprimates,12 the metabolic heterogeneity of LDL formation.We found that over 80% of LDL apo B was synthesizeddirectly, the remainder being derived from VLDL conver-sion. The increased secretion of LDL from perfusedanimal livers after cholesterol feeding15'16 indicated thatthe source of direct LDL synthesis was hepatic and thusmay be related to the availability of hepatic cholesterol fortransport. Kinetic studies have demonstrated direct LDLsynthesis in both familial hypercholesterolemic homo-zygous17 and heterozygous subjects,18 as well as infamilial combined hyperiipoproteinemic subjects19 and, toa lesser extent, in normal humans.20 Treatment of familialcombined hyperiipoproteinemic subjects reduced directLDL synthesis,19 and weight loss in hypertriglyceridemicsubjects increased LDL direct synthesis,21 indicating thatthis pathway could be regulated. We have also demon-strated that the LDL direct synthesis pathway in miniaturepigs could be regulated.22 Enhanced cholesterol excretionby cholestyramine treatment and inhibition of cholesterolsynthesis by mevinolin treatment inhibited the direct LDLsynthesis pathway.22

The present experiments carried out in miniature pigswere designed to answer two questions. First, would dietscontaining fish oil inhibit VLDL apo B synthesis (flux) andsubsequently influence the proportion of VLDL apo Bconverted to LDL, compared to that removed directly?Second, since it is known that fish oil inhibits VLDLsynthesis, we wanted to determine if fish oil would alsoinhibit the direct synthesis of LDL.

MethodsAnimals and Diets

Miniature pigs (20 to 35 kg) were obtained from a localsupplier (Hyde Park Farms, Hyde Park, Ontario). After a1-week acclimatization, an indwelling silastic catheter(0.078" ID) was surgically implanted in each externaljugular vein under halothane anesthesia. Ketamine wasused as a pre-anesthetic. The catheters were tunneledunder the skin and externalized in the middle of the back.Three-way stop cocks were attached and held in placewith a bandage and elastic netting. The catheters, keptpatent by filling them with 7% EDTA-Na2, allowed for easeof sample injection as well as blood sampling throughoutthe study in unanesthetized animals. This protocol wasapproved by the animal care committee of the Universityof Western Ontario.

A crossover design in which each of four pigs receiveda diet containing either corn oil or Maxepa (R. P. Scherer,Windsor, Canada) for 18 to 20 days before lipoproteinturnover studies was used. Each animal was then switchedto the other diet (either Maxepa or corn oil) for a further 18to 20 days, followed by a second turnover study. Eachanimal received 750 g/day of Purina pig chow (16%protein, 5% fat, Ralston Purina, Longeuil, Quebec) towhich 30 g of com oil or 30 g of Maxepa were mixed just

before feeding. The total dietary fat content was 9% wt/wt.The Maxepa diet provided 9.45 g/day of omega-3 fattyacids (5.04 g/day of eicosapentaenoic acid, 20:5 w-3;3.63 g/day docosahexaenoic acid, 22:6 w-3), 3.54 g/dayof omega-6 fatty acids (3.48 g/day linoleic acid, 18:2,w-6), and 110 mg/day of cholesterol (contained in Maxepaoil). The corn-oil diet provided 17.0 g/day of omega-6 fattyacids (16.8 g/day, linoleic acid 18:2, w-6) and no omega-3fatty acids. Cholesterol was dissolved in the corn oil toequal that present in the Maxepa diet. Maxepa oil, whichcontained <100 IU vitamin A/g, no vitamin D, and toco-pherol (1 lU/g) as an antioxidant, was aliquoted in dailydoses and stored under N2 in sealed vials at 4°C until justbefore feeding. Maxepa and com oil were then mixed withthe feed, which was consumed within 1 hour.

Lipoprotein Turnover StudiesLipoprotein turnover studies were conducted essen-

tially as described previously.14-22 Autotogous VLDL (Sf 20to 400) and LDL (Sf 0 to 12) for radiolabeling were iso-lated from 100 ml of plasma obtained after a 16-hour fast.Radioiodination was performed using the iodine mono-chloride method as modified by Fidge and Poulis23 asdescribed previously for pigs.22 Lipoproteins were steril-ized by the addition of 100 uglml gentamycin sulfate(Schering, Pointe Claire, Quebec) and checked for sterilityand pyrogenicity. VLDL contained less than 2% free iodine,24% to 28% of the label was bound to lipid, and 25% to35% of the protein-bound iodine was bound to apo B. LDLcontained less than 1% free iodine, 20% to 24% lipidlabeling; 80% to 90% of the LDL protein labeled was boundto apo B. After each dietary period, each of four pigsreceived 20 /iCi of autologous 125Mabeled VLDL apo B and15 jiCi of autologous 131l-labeled LDL apo B. Animals werefasted for 16 hours before each study. After injection,blood samples (12 ml) were obtained at 5,15, 30, and 45minutes and at 1, 1.5, 2, 3, 4, 6, 12, 24, 30, 48, and 72hours. Pigs received no food until after the 12-hoursample, at which time one-half of the daily food was given.Animals were given their full diet after the 24- and 48-hoursamples. This procedure limited the contribution of intes-tinally derived particles, resulting in constant concentra-tions of VLDL, IDL, and LDL apo B.22 VLDL (d<1.006),IDL (d=1.006 to 1.019), and LDL (d=1.019 to 1.063) wereseparated from plasma obtained from blood samplescollected in tubes containing EDTA-Na2 (1.5 mg/ml).22 Inthree of four animals, as part of both turnover studies, LDLwas subfractionated into LDL, (d=1.019 to 1.040) andLDL2 (d=1.040 to 1.063). The density of total LDL waslowered to d=1.019 by the addition of d=1.006. LDL, wasseparated by ultracentrifugation in a Beckman 50 Ti rotor,12°C, 50K, for 24 hours. The infranatant was raised tod=1.063 by the addition of d=1.34 and was recentrifugedunder the same conditions to give LDL2. Apo B wasisolated from each lipoprotein fraction by isopropanolprecipitation.22 Specific activities were calculated aftercounting the washed apo B pellet22 and subsequentdetermination of its protein content by a modified Lowryprocedure.24 The plasma concentration of apo B in eachlipoprotein fraction was determined by subtracting theprotein value of the first precipitation supernatant from the


Dow

nloaded from


60 ARTERIOSCLEROSIS VOL 9, No 1, JANUARY/FEBRUARY 1989

Table 1. Plasma Lipld and Llpoproteln Concentrations In Miniature Pigs Consuming Corn- and Flsh-oll Diets

Diet

Com oil

Fish oil

Significance

Trigtyceride

Total

33±6

17±2

p<0.05

VLDL

18±5

7±4

p<0.05

Total

111±4

92±5

p<0.025

Cholesterol

VLDL

2±0.9

1.5+0.9

NS

•

LDL

48±3

43±4

NS

HDL

59±1

48±2

p<0.01

Apolipoprotein B

VLDL

0.59+0.06

0.37±0.06

p<0.01

LDLf20±0.3

15±1.0

p<0.05

Values are expressed as mg/dl. Each value is the mean±SEM of eight determinations, two samples from each of four animals.Significance was determined by paired f test NS=not significant.

*VLDL cholesterol was determined after ultracentrifugation at d<1.006. HDL cholesterol was determined after precipitation ofapo-B-containing lipoproteins from plasma. LDL cholesterol was calculated as total cholesterol minus the sum of VLDL cholesterol andHDL cholesterol.

tDetermined in the LDL fraction (d=1.019 to 1.063) separated by ultracentrifugation.

total lipoprotein protein concentration.25 For apo B con-centration determinations, apo B was not measured directlyon the precipitated pellet due to the known nonspecificlosses of apo B during the washing and delipidation of thepellet in order to assay it for protein.

Kinetic AnalysesThe transport of apo B in VLDL and LDL was calculated

from the apo B specific activity curves.24 As foundpreviously,22 both curves were best described by a bi-exponential curve, and curve parameters were calculatedby a computer with a nonlinear, least squares technique.27

Kinetic parameters yielded values for irreversible frac-tional catabolic rate (FCR) and the mass of apo B (poolsize) in the largest, most rapidly turning over pool. Trans-port rates or flux were calculated by multiplying the FCRby the apo B mass (pool size). Details of the calculationshave been published previously.3728 The 12Sl-labeledapo B specific activity curves for VLDL, IDL, and LDL werecompared, which allowed for the calculation of precursor-product relationships between lipoprotein fractions.Precursor-product relationships were assessed by twomethods: 1) the peak product specific activity methoddescribed by Zilversmit29 and applied to apo B kinetics asreported previously,714-22>28 and 2) the area under thespecific activity curves as described by Goldberg et al.12

and applied previously to apo B kinetics.121422 The ratio-nale and assumptions involved in using these methodshas been discussed previously.22

AnalysesTwice during each dietary period (immediately before

and 5 days after each turnover study) plasma was obtainedfor lipid analysis. Total cholesterol and triglyceride, VLDLcholesterol and triglyceride, and HDL cholesterol concen-trations were determined. VLDL were obtained after ultra-centrifugation at d<1.006,22 and HDL was obtained afterprecipitation of other lipoproteins by heparin-manganesechloride.14 LDL cholesterol was calculated by the differ-ence. Enzymatic methods described in kits obtained fromBoehringer-Mannheim (Montreal, Canada) were used fortriglyceride (Peridochrom, GPO-PAP) and cholesterol(CHOD-PAP C-system) analyses. In addition, VLDL, IDL,LDL (d=1.019 to 1.063) and HDL (d=1.063 to 1.21) wereisolated by ultracentrifugation22 and were analyzed for lipidsby total lipid profiling by using gas-liquid chromatography.30

Total lipids were extracted by using chloroform/methanol

(2:1, vol/vol) after dephosphorylatjon of the choline con-taining phospholipids by phospholipase-C30 Dietary fattyacid composition was determined by gas-chromatographyusing a 2 meter column (SP 2330, liquid phase) on aVarian 6000 gas chromatograph. Lipoprotein proteinwas determined by the method of Markwell et al.24

Differences between control and treatment values wereanalyzed by paired f analysis.

ResultsConcentrations of plasma triglyceride and cholesterol

were both significantly reduced by the fish-oil diet com-pared to the corn-oil diet (Table 1). The fall in plasmatriglyceride concentration (48%, p<0.05) was due primar-ily to a 65% (p<0.05) drop in VLDL triglyceride concen-trations. The decline in total cholesterol (17%, p<0.025)was related to a significant drop in HDL cholesterolconcentrations (p<0.01). There was a trend towardreduced VLDL and LDL cholesterol concentrations. Apo-lipoprotein B concentrations were also significantly reducedby the fish-oil diet. VLDL and LDL apo B levels declined by37% (p<0.01) and 25% (p<0.05), respectively.

Autologous iodinated VLDL and LDL were simulta-neously injected into each pig at the end of each dietaryperiod. Apo B specific activity curves for VLDL, IDL, andLDL from one animal after injection of VLDL are shown inFigure 1. Examination of precursor-product relationshipsduring the corn-oil period (Figure 1A) demonstrated thatthe maximal value for LDL apo B was reached well beforeit crossed the specific activity curves of its precursors, IDLand VLDL. We have interpreted this as indicating that asubstantial proportion of LDL was derived independent ofIDL and VLDL catabolism. The fraction of LDL synthesisderived from the catabolism of VLDL and IDL was deter-mined by two methods as described previously.22 In thefirst method, described initially by Zilversmit,29 this frac-tion was derived by calculating the ratio of the peak LDLspecific activity value to the IDL specific activity value atthe same time. Subsequently, the ratio of the peak IDLspecific activity to the VLDL specific activity value wascalculated. In the second method described by Goldberget al.12 the fraction of LDL not derived from VLDL or IDLwas determined by calculating the dilution factor: 1 -(areaunder the LDL specific activity curve/area under the VLDLor IDL specific activity curve). This latter method does notrequire estimation of the exact point of peak specific


Dow

nloaded from



103!CORNOL

10'

10 12 14 16

mcoo18 20 22 24

Figure 1. Apoprotein B precursor-product relationships betweenVLDL, IDL, and LDL fractions after the injection of radiolabeledVLDL. The results are from animal #1. A was obtained during thecorn-oil dietary period and B, during the fish-oil (Maxepa) period.

activity values. We have previously applied this method toapo B metabolism in miniature pigs.22

Figure 1B depicts apo B specific activity curves obtainedfrom the same animal during the fish-oil dietary period.The peak LDL apo B specific activity value occurred closeto the point where it crossed the IDL, and subsequently,the VLDL apo B curves. We have interpreted this asindicating that the independent synthesis of LDL apo Bwas dramatically reduced, suggesting that most LDL B isderived from VLDL/IDL catabolism.

The fish-oil diet significantly reduced the pool size ofVLDL apo B by 38% (1.13±0.14 vs. 1.82±0.39 mg/kg,p<0.05) as shown in Table 2. This was primarily due to anincrease in irreversible fractional catabolic rate (0.83±0.13vs. 0.48±0.03 h r \ p<0.05) as the total flux or productionrate was not altered (0.95±0.14 vs. 0.85±0.17 mg/hr/kg).Compared to the corn-oil diet, the fish-oil diet also reducedthe LDL apo B pool size, but to a lesser extent than forVLDL (Table 3). LDL apo B declined by 20% (13.3+0.6vs. 16.5±1.2 mg/kg, p<0.05). This was primarily due to adecrease in flux or production rate by 40% (0.62±0.11 vs.

1.13±0.19 mg/hr/kg, p<0.05) despite a significant fall inthe irreversible fractional catabolic rate (0.046+0.006 vs.0.059±0.006, h r \p<0 .01) .

The fish-oil diet had a striking effect on the source ofLDL apo B. The percent of LDL derived from VLDLincreased from 13%±1.9% to 78%±6% during the fish-oilperiod (Table 3). Also, the fraction of LDL derived from thedirect synthesis pathway was markedly reduced in areciprocal fashion (22%±6% vs. 87%±2%). This wascalculated by the area under the curve method, which onaverage differed from the peak specific activity method by4.8%. Knowing the fraction of total LDL apo B flux derivedfrom VLDL and the total LDL apo B flux, we can calculatethat the flux of LDL derived from VLDL increased signifi-cantly (0.49±0.08 vs. 0.13±0.01 mg/hr/kg, p<0.01) onthe fish-oil diet (Table 3). In contrast, the independentsynthesis of LDL apo B was significantly reduced(0.13±0.04 vs. 0.99±0.19 mg/hr/kg, p<0.025). Of thetotal VLDL apo B flux, we know the amount converted toLDL, since this is equivalent to the amount of LDL apo Bderived from VLDL. Subtracting this value from the totalVLDL apo B ftux, we determined the flux of VLDL apo Bleaving the circulation directly (Table 2). Fish oil reducedthe flux of VLDL apo B removed by this pathway by 40%,but due to the variability between animals this was notsignificant. However, as a percent of total VLDL apo Bflux, the amount removed by this pathway was reduced by44%, p<0.005. The fractional catabolic rate of VLDL apoB cleared directly did not change with the fish-oil diet(0.39±0.16 vs. 0.40±0.05 hr1 ) . This was calculated as:the flux of VLDL apo B not converted to LDL divided by theVLDL apo B pool size. This was interpreted as indicatingthat fish oil decreased the capacity for VLDL apo B directremoval. One would have expected that, in view of thedecreased VLDL apo B pool size, an increased FCRwould have been observed if the removal capacity had notbeen decreased.

Thus, compared to the corn-oil diet, fish oil did notaffect total VLDL apo B flux (synthesis). However, theproportion of total flux converted to LDL was significantlyincreased, whereas the proportion cleared directly wasdecreased (Table 2). The fish-oil diet significantly reducedthe entry of total apo B into plasma (1.05±0.18 mg ofapo B/mg/kg vs. 1.84±0.35, p<0.05). The latter was cal-culated as: total VLDL apo B synthesis plus direct LDLapo B synthesis.

To determine if these marked changes in apo B metab-olism were related to changes in lipoprotein particlecomposition, total lipid profiles of VLDL, IDL, LDL, andHDL were analyzed as shown in Table 4. No majordifferences between the fish-oil and corn-oil diet wereobserved for any of the parameters measured. The per-centage of omega-3 fatty acids in phospholipids, choles-terol esters, and triglycerides increased on the fish-oil diet(data not shown). The amount of apo B as a percent oftotal protein in VLDL, IDL, and LDL was not altered.

In three of the four animals, the effect of the fish-oildiet on the metabolism of LDL subtractions, LDL,(d=1.019 to 1.040) and LDL2 (d=1.040 to 1.063) wasdetermined as shown in Figure 2. The peak specificactivity of both fractions occurred well before they crossed


Dow

nloaded from



Table 2. Metabolism of Very Low Density Llpoproteln Apollpoproteln B In Miniature Pigs Fed Diets ContainingCorn ON and Fish ON

Animal

1

2

3

4

Mean±SE

P<

Diet

C

M

C

M

C

M

C

M

C

M

Pool size*(mg/kg)

2.23

1.10

2.71

1.46

1.12

0.80

1.20

1.16

1.82±0.39

1.13±0.14

0.05

FCRf(hr1)0.38

1.17

0.48

0.54

0.49

0.85

0.55

0.74

0.48±0.03

0.83±0.13

0.05

Flux total(mg/hr/kg)

0.89

1.31

1.29

0.79

0.54

0.68

0.66

0.86

0.85±0.17

0.91 ±0.14

NS

Flux to LDLt(mg/hr/kg)

0.13

0.70

0.17

0.53

0.14

0.33

0.11

0.38

0.13±0.01

0.49±0.08

0.01

(15)

(54)

(11)(77)

(26)

(49)

(17)

(45)

17.3±3

56.3±7

0.01

Flux

0.76

0.61

1.12

0.26

0.40

0.35

0.55

0.48

0.71 ±0.

direct removal§(mg/hr/kg)

(85)

(46)

(89)

(33)

(74)

(51)

(83)

(55)

16 82.7±3.2

0.43±0.08 46.3±4.8

NS 0.005

'Pool size refers to the mass of apo B in the largest pool (pool 1).fFCR=irreversible fractional catabolic rate.tFlux of VLDL apo B to LDL apo B is equivalent to the flux of LDL apo B derived from VLDL apo B. The latter is determined as described

in the footnotes to Table 3. Values In brackets are the percent of total flux.§Calculated by subtracting the VLDL apo B flux to LDL from the total VLDL apo B flux.C=com-oil diet, M=Maxepa (fish-oil) diet. Probability determined by paired f test.

Table 3. Metabolism of Low Density Upoproteln Apollpoproteln B In Miniature Pigs Fed Diets Containing CornOH and Fish ON

Animal

1

2

3

4

Mean±SE

P<

Diet

C

M

C

M

C

M

C

M

C

M

Pool size*(mg/kg)

19.6

15.2

26.7

12.9

15.9

13.2

13.6

13.2

16.5±1.2

13.3±0.6

0.05

FCRf(hr')0.080

0.064

0.055

0.044

0.053

0.035

0.051

0.040

0.059 ±0.006

0.046±0.006

0.01

Flux total(mg/hr/kg)

1.56

0.95

1.45

0.56

0.84

0.46

0.68

0.52

1.13+0.19

0.62±0.11

0.05

Flux from \/LDI4(mg/hr/kg)

0.13

0.70

0.17

0.53

0.14

0.33

0.11

0.38

0.13±0.01

0.49±0.08

0.01

(8)(74)

(12)

(95)

(16)

(71)

(16)

(73)

13±2

78±6

0.005

Flux direct synthesis§(mg/hr/kg)

1.43

0.25

1.28

0.03

0.70

0.13

0.57

0.14

0.99±0.19

0.13±0.04

0.025

(92)

(26)

(88)

(5)(84)

(29)

(84)

(27)

87±2

22±6

0.005

*Pool size refers to the mass of apo B in the largest pool (pool 1).fFCRHrreversible fractional catabolic rate.^Calculated by subtracting the LDL apo B flux (from the 131I-LDL apo B curve) derived by direct synthesis from total LDL apo B flux.

The latter was determined from the 131 I-LDL apo B specific activity curve. Values in brackets are a percent of the total flux.§Calculated by multiplying the total LDL apo B flux (from the13' I-LDL apo B curve) by the proportion of LDL apo B not derived from VLDL

(direct synthesis). The latter, termed the dilution factor, was calculated as: 1 -(area under the 12SI-LDL apo B specific activity curve/areaunder the 12SI-VLDL apo B specific activity curve). Values in brackets are a percent of total flux.

C=com-oil diet, M=Maxepa (fish-oil) diet. Probability determined by paired nest.

their respective precursors, and the areas under theLDL, and LDL2 curves were substantially lower than theVLDL curve suggesting dilution of both the LDL, andLDL2 by unlabeled apo B. As calculated from the dilutionfactor, a mean of 17%±4% of LDL, was derived fromVLDL, and 73%±6% of LDL2 was derived from LDL,.During the fish-oil period, the specific activities of LDL,and LDL2 peaked closer to the point they crossed theirrespective precursors. As calculated from the dilutionfactors, LDL, derived from VLDL increased to 85%±6%and LDL2 derived from LDL, increased to 88%±4% ofthe total LDL2 flux. This indicates that the fish-oil dietsignificantly inhibited the independent synthesis of bothLDL, and LDL2. The mass of LDL2 apo B accounted for

85%±5% of the total plasma LDL apo B (which did notchange with the diets).

DiscussionThe experiments reported in this paper were designed

to determine whether, in the miniature pig, the lowering ofVLDL concentrations by a diet rich in fish oil (omega-3)fatty acids was related to a decreased synthetic rate andwhether this diet would influence the degree of conversionof VLDL to LDL. Since plasma VLDL is a precursor ofplasma LDL, it has been assumed that the degree ofconversion of VLDL to LDL must be a key component inthe regulation of LDL concentrations.21 However, direct


Dow

nloaded from



Table 4. Effects of FIsh-OII Diet on Percent LIpki Composition of Plasma LIpoprotelns

Unesterffied cholesterol

Cholesterol ester

Phospholipid

Trig lyce ride

Cholesterol ester:trlglyceride

Unesterifled cholesterol:phospholipid

Apo B (percent of totalprotein)

Estimated averagediameter (A0)

C

3.5

11.5

26.8

58.8

0.261

0.266

23

318

VLDL

M

3.6

12.1

24.0

60.1

0.273

0.301

20

374

NS

C

9.1

23.6

33.1

34.0

0.91

0.555

21

210

IDL

M

7.6

27.6

33.0

31.5

1.28

0.465

17

210

NS

C

9.9

54.9

32.4

3.2

47.9

0.62

84

210

LDL

M

9.8

51.1

34.8

4.6

41.2

0.58

89

198

NS

C4.7

41.7

52.7

1.9

28.7

0.234

128

HDL

M

5.2

46.5

47.5

2.9

21.1

0.231

149

NS

C=corn-oil diet, M=Maxepa (fish-oil) diet, NS=not significant.Values are the means from four animals.

secretion of LDL1'17-22 might also regulate steady-stateplasma LDL concentrations. We have shown that theminiature pig possesses a substantial direct LDL synthe-sis pathway,14'22 which allowed us to assess the effect offish oil on its regulation.

Our experiments have confirmed the triglyceride-lowering effects of omega-3 fatty acids (Maxepa) com-pared to omega-6 fatty acids (corn oil), reported in studiesin humans1-4910 and in rats.56 Although the effect wasless marked, we observed lower LDL cholesterol andapo B concentrations with the fish-oil diet (Table 1). Thisis consistent with previous reports in normal humans,41031

but differs from the rise or lack of change in LDL choles-terol and apo B reported by others.1-2-3'9

The reduced VLDL triglyceride concentrations on thefish-oil diet in the present study were associated with areduced VLDL apo B pool size (Table 2). This was relatedto an increased fractional catabolic rate, since flux orsynthesis was not changed. Nestel et al.3 also observedthat the VLDL apo B FCR was markedly increased innormal men fed fish oil, but in contrast to our results, theyshowed a substantial reduction in VLDL apo B synthesis.The reason for this difference is not known, but may bedue to the fact that the already low VLDL apo B syntheticrates during the corn-oil period, (which were lower thanobserved in control animals previously) could not betowered further by the fish-oil diet.

In the present study, we determined that fish oil has amarked effect on the catabolic fate of VLDL apo Bcompared to the corn-oil diet. The proportion of VLDLapo B converted to LDL increased fourfold, whereas theproportion removed directly decreased by 50% (Table 2).To our knowledge, the effect of fish oil on the metabolicchanneling of VLDL apo B has not been reported. Ourunique observation may explain why in some studies inhumans, LDL concentrations are not changed or mayactually rise,1239 even though VLDL apo B productionwas reduced.3

The shift from VLDL removal in favor of conversion toLDL in our studies may be related to a lower hepatic LDL(B/E) receptor activity (as discussed below), since hepatic

uptake of VLDL remnants is thought to occur via thisreceptor.32 The lack of change in FCR for VLDL apo Bremoved directly, in spite of a lower VLDL apo B pool size,is consistent with this interpretation.

The increased conversion of VLDL apo B to LDL duringthe fish-oil diet could also be related to the production of asmaller VLDL particle of altered lipid composition. Pack-ard et al.33 reported that smaller VLDL particles arepreferentially converted to LDL. However, results from thelipid analyses by total lipid profiling (Table 4) do notsupport this idea. No change in estimated particle diam-eter or proportions of the major lipid classes were observed.Sullivan et al.9 reported that men fed fish oil had reducedVLDL particle size; however, the control diet, unlike ourexperiments, was not based on polyunsaturated fattyacids of the omega-6 series. In our experiments, theincreased proportions of omega-3 fatty acids in the majorVLDL lipid classes may have enhanced the conversion toLDL. Although not determined in the present study, differ-ences in VLDL small molecular weight apoprotein com-position, such as reduced apo E content, may have beena factor. Yamada et al.13 have demonstrated in rabbitsthat VLDL particles poor in apo E are preferentiallyconverted to LDL.

In the present studies, the increased conversion ofVLDL apo B to LDL by fish oil did not result in anincreased LDL apo B pool size. LDL pool size actuallydecreased by 20% (Table 3), which was due primarily tothe decrease in LDL apo B direct synthesis being muchlarger than the increase in LDL apo B derived from VLDL.This net decrease in total LDL apo B synthesis is consis-tent with the work of Illingworth et al.,10 who found adecrease in LDL apo B synthesis in normal men fedfish-oil diets. The observed fall in LDL apo B directsynthesis is not consistent with the idea proposed byNestel et al.3 that a decreased production rate of VLDLtriglyceride may lead to an increased secretion oftriglyceride-poor particles with a density of LDL.

The direct synthesis of LDL in the corn-oil fed pigs wasrelated to both LDL, and LDL2 density classes. The kineticdata indicate that fish oil inhibited the direct synthesis of


Dow

nloaded from



103

102

CORN OL

VLDt- (d< 1.006)LDL-1 <d 1.019-1.040)LDL-2 (d 1.040-1.083)

B

VLDL (d< 1.006)LDL-1 (d. 1.019-1.040)LDL-2 (d 1.040-1.063)

10

10 12 14T I E (hot*)

16 18 20 22 24

Figure 2. Apoprotein B precursor-product relationships betweenVLDL, LDL-1, and LDL-2 fractions after the injection of radiola-beled VLDL. The results are from animal #4. A was obtainedduring the corn-oil dietary period and B, during the fish-oil(Maxepa) period.

both LDL, and LDL2. The results with the animals fed cornoil differ somewhat from our previous studies. In pigs fedonly chow, no direct LDL, synthesis was observed (HuffMW, Telford DE, unpublished observations). Althoughpaired experiments have not been carried out to confirmthis finding, it is possible that LDL synthesized directly incorn-oil-fed animals is lighter in density and thus a greaterproportion enters the circulation as LDL,. The physiolog-ical significance of an elevated direct synthesis of LDLapo B, particularly LDL2 is not known. However, it ispossible that the metabolism of directly synthesized LDLdiffers from that of LDL derived from VLDL and, if asso-ciated with high levels of plasma LDL, may contribute tothe accelerated atherosclerosis observed in patients withhypercholesterolemia and familial combined hyperiipo-proteinemia.171819 We have proposed that the directsynthesis pathway of LDL formation is related to theamount of hepatic cholesterol relative to triglyceride des-tined for hepatic secretion.22 Previously, we demonstratedthat this pathway was inhibited by cholestyramine and/or

mevinolin treatment.22 It is possible that fish oil reduceshepatic cholesterol concentrations and reduces apo Bsynthesis. Rats fed fish oil have lower liver cholesterolconcentrations, cholesterol synthesis, and an increasedrate of biliary cholesterol excretion compared to rats fedsafflower oil.34 Omega-3 fatty acids have been shown toinhibit the synthesis of apo B by cultured liver cells (HepG2).35 It is possible that the inhibition of direct LDLsynthesis observed in our studies with fish oil is due tothe inhibition of apo B and cholesterol destined forassembly and secretion with LDL-like particles.

The concept of direct LDL apo B synthesis remainscontroversial even though other investigators have inter-preted their kinetic data in terms of direct LDL apo Bsynthesis.3'12'14'17-22 An alternate explanation for directsynthesis of LDL is the secretion from the liver of VLDL-like particles that are rapidly converted to LDL, but whoseplasma pool size within the VLDL density range is verysmall and thus are not labeled with plasma VLDL norcontribute to the VLDL apo B mass for specific activitydeterminations. It is possible that the fish-oil diet selec-tively inhibited the synthesis of this hypothetical VLDLsubpopulation. Our studies could not rule out this possi-bility. It should be pointed out that in our studies totalapo B synthesis (total VLDL plus LDL direct synthesis)was reduced by 4 1 % with fish-oil treatment.

The observed decrease in LDL apo B FCR and theincreased conversion of VLDL apo B to LDL could beinterpreted in terms of a decrease in hepatic LDL receptoractivity. Wong et al.35 have reported lower LDL (B/E)receptor activity in Hep G2 cells after preincubation witheicosapentaenoic acid. Roach et al.38 have demonstrateda decreased LDL receptor activity in liver membranesisolated from rats fed fish oil. Illingworth et al.10 demon-strated no consistent difference in LDL apo B FCR innormal men fed fish oil, although the FCR decreased infive of seven subjects studied.

Witztum et al.37 have demonstrated that alterations inLDL composition can result in a reduced LDL apo B FCRdue to a poor interaction of LDL with altered compositionwith the LDL receptor. In the present experiments, plasmaLDL during the corn-oil period, which was derived mainlyfrom direct LDL synthesis, and LDL during the fish-oilperiod, which was derived mainly from VLDL catabolism,did not differ in terms of size or percent composition oflipid classes. Thus, once in the circulation, LDL derivedfrom either source had similar composition of major lipkJs,although the percent of omega-3 fatty acids was higher onthe fish-oil diet (Table 4). Parks et al.38 reported changesin plasma LDL composition and size in monkeys fed fishoil. However, the control monkeys were fed a lard dietrather than a diet rich in omega-6 fatty acids.

The findings in the present study clearly demonstratethat fish oil reduces VLDL apo B concentrations; however,the proportion converted to LDL is increased. LDL apo Bconcentrations are not elevated due to the greater reduc-tion in the direct LDL synthesis pathway. The increasedconversion of VLDL apo B to LDL apo B combined withreduced fractional catabolic rates for LDL apo B suggest adown-regulation of hepatic LDL receptor activity. Whetherthis mechanism is responsible for the variable and incon-


Dow

nloaded from



sistent effect of fish oil on LDL cholesterol concentrationsin humans requires further investigation.

Acknowledgments

We thank Kim Woodcraft for expert technical assis-tance, Jay Engle and Chris Lipohar for performing thesurgery, and Lynn Thomson for manuscript preparation.The Maxepa was generously provided by Dave Hutchin-son, R. P. Scherer Canada, Windsor, Ontario. We aregrateful to W. Carl Breckenridge, Dalhousie University,Halifax, Nova Scotia, for performing the total lipid profiling.

References

1. Ptilllipson B, Rothrock DW, Connor WE, Harris WS,Illlngworth DR. Reduction of plasma liplds, lipoproteins, andapoproteins by dietary fish oils In patients with hypertriglyc-eridemia N Engl J Med 1985;312:1210-1216

2. Simons LA, Hlckie JB, Balasubramanlam S. On theeffects of dietary n-3 fatty acids (maxepa) on plasma lipkJsand lipoproteins in patients with hyperlipidemia. Atheroscle-rosis 1985;54:75-88

3. Nestel PJ, Connor WE, Reardon MF, et al. Suppression bydiets rich in fish oil of very-low-density lipoprotein productionin man. J Clin Invest 1984;74:82-89

4. Harris WS, Connor WE, McMurry MP. The comparativereductions of the plasma lipids and lipoproteins by dietarypolyunsaturated fats: salmon oil vs vegetable oils. Metabo-lism 1983;32:179-184

5. Wong SH, Reardon M, Nestel PJ. Reduced triglycerideformation from long-chain polyenotc fatty adds in rat hepa-tocytes. Metabolism 1985;34500-905

6. Wong SH, Nestel PJ, Trimble RP, Storer GB, Illman RJ,Topping DL. The adaptive effects of dietary fish and saf-flower oil on lipid and lipoprotein metabolism in perfused ratliver. Biochim Biophys Acta 1984;792:103-109

7. Reardon MF, Fldge NH, Nestel PJ. Catabolism of very lowdensity lipoprotein B apoprotein in man. J Clin Invest 1978;62:850-860

8. Slgurdsson Q, Nlcholl A, Lewis B. Conversion of very lowdensity lipoprotein to low density lipoprotein. A metabolicstudy of apolipoprotein B kinetics in human subjects. J ClinInvest 1975;56:1481-1490

9. Sullivan DR, Sanders TAB, Trayner IM, Thompson GR.Paradoxical elevation of LDL apoprotein B levels in hyper-triglyceridaemic patients and normal subjects ingesting fishoil. Atherosclerosis 1986:61:129-134

10. Illlngworth DR, Harris WS, Connor WE. Inhibition of lowdensity lipoprotein synthesis by dietary omega-3 fatty acidsin humans. Arteriosclerosis 1984:4:270-275

11. Cortes* C, Levy Y, Janus ED, et al. Modes of action of lipidlowering diets in man: studies of apolipoprotein B kinetics inrelation to fat consumption and dietary fatty adds composi-tion. Eur J Clin Invest 1983;13:79-85

12. Goldberg IJ, Le NA, Ginsberg HN, Patemttl JR Jr, BrownVW. Metabolism of apoprotein B in cynomokjus monkey:evidence for independent production of low-density lipopro-tein apoprotein B. Am J Physiol 1983;244:E196-E201

13. Yamada N, Shames DM, Stoudemlre JB, Havel RJ.Metabolism of lipoproteins containing apolipoprotein B-100in blood plasma of rabbits: heterogeneity related to thepresence of apolipoprotein E. Proc Natl Acad Sci USA1986:83:3479-3483

14. Huff MW, Telford DE. Direct synthesis of low-density lipo-protein B in the miniature pig. Metabolism 1985;34:36-42

15. Guo LS, Hamilton RL, Ostwald R, Havel RJ. Secretion ofnascent lipoproteins and apolipoproteins by perfused liversof normal and cholesterol-fed guinea pigs. J LJpkJ Res 1982;23:543-555

16. Johnson FL, St Clalr RW, Rudel LL. Studies on theproduction of low density lipoproteins by perfused livers fromnonhuman primates. J Clin Invest 1983:72:221-236

17. Soutar AK, Myant NB, Thompson GR. Simultaneous mea-surements of apolipoprotein B turnover in very low and lowdensity lipoproteins in familial hypercholesterolemia. Athero-sderosls 1977;28:247-256

18. Janus ED, Nlcholl A, Wootton R, Turner PR, Maglll PJ,Lewis B. Quantitative studies of very low density lipopro-tein conversion to low density lipoprotein in normal controlsand primary hyperiipoproteinemic states and the role ofdirect secretion of low density lipoprotein in heterozygousfamilial hypercholesterolemia. Eur J Clin Invest 1980;10:149-159

19. Klssebah AH, Atfarsl S, Evans DJ. Low density lipoproteinmetabolism in familial combined hyperlipidemia. Mechanismof the multiple lipoprotein phenotypic expression. Arteriosde-rosis 1984 ;4:614-624

20. Kesanleml YA, Beta WF, Grundy SM. Comparisons ofmetabolism of apolipoprotein B in normal subjects, obesesubjects and patients with coronary heart disease. J ClinInvest 1985:76:586-595

21. Ginsberg HN, Le NA, Gibson JC. Regulation of the pro-duction and catabolism of plasma low density lipoproteins inhypertriglyceridemic subjects. Effect of weight loss. J ClinInvest 1985;75:614-623

22. Huff MW, Terford DE, Woodcroft K, Strong WLP. Mevln-olin and cholestyramlne inhibit the direct synthesis of lowdensity lipoprotein apolipoprotein B in miniature pigs. J LipidRes 1985;26:1175-1186

23. Fldge NH, Poulls P. Studies on the radioiodination of verylow density lipoprotein obtained from different mammalianspedes. Clin Chim Acta 1974;52:15-26

24. Markwell MAK, Haas SM, Bieber LL, Tolbert NE. A mod-ification of the Lowry procedure to simplify protein determi-nation in membrane and lipoprotein samples. Anal Biochem1978;87:206-210

25. Egusa G, Brady DW, Grundy SM, Howard BV. Isopropanolpredpitation method for the determination of apolipoprotein Bspecific activity and plasma concentrations during metabolicstudies of very low density lipoprotein and low density lipopro-tein apolipoprotein B. J Lipid Res 198324:1261-1267

26. Gurplde E, Mann J, Sandberg E. Determination of kineticparameters in a two-pool system by administration of one ormore tracers. Biochemistry 1964:3:1250—1255

27. Bard Y. Nonlinear parameter estimation. New York: Aca-demic Press, 1974:1-145

28. Nestel PJ, Reardon MF, Fldge NH. Very low densitylipoprotein B-apoprotein kinetics in human subjects. Rela-tionships between pool size, flux and removal rate. Circ Res1979;45:35-41

29. Zllversmlt DB. The design and analysis of isotope experi-ments. Am J Med 1960;29:832-848

30. Kuksls A, Myher JJ, Maral L, Geher K. Determination ofplasma lipid profiles by automated gas chromatography andcomputerized data analysis. J Chromatogr Sd 1975;13:423-430

31. Sanders TB, Vlckers M, Halnes AP. Effect on blood lipidsand haemostasis of a supplement of cod liver oil, rich ineicosapentaenoic and docosahexaenoic adds, in healthyyoung men. Clin Sd 1981:61:317-324

32. Goldstein JL, Brown MS. Progress in understanding theLDL receptor and HMG-CoA reductase, two membraneproteins that regulate the plasma cholesterol. J Lipid Res1984:25:1450-1461

33. Packard CJ, Munro A, Lorimer AR, Gotto AM, ShepherdCJ. Metabolism of apolipoprotein B in large triglyceride-richvery low density lipoproteins of normal and hypertriglyceri-demic subjects. J Clin Invest 1984;74:2178-2192

34. Balasubramanlam S, Simons LA, Chang S, Hlckie JB.Reduction in plasma cholesterol and increase in biliarycholesterol by a diet rich in n-3 fatty adds in the rat. J LJpidRes 1985^6:684-689


Dow

nloaded from



35. Wong S, Nestel PJ. Eicosapentaenok; acid inhibits the secre-tion of traJcytglycerol and of apoprotein B and the binding ofLDL in Hep G2 cells. Atherosclerosis 1987:64:139-146

36. Roach PD, Kambouris AM, Trimble RP, Topping DL,Nestel PJ. The effects of dietary fish oil on hepatic highdensity and low density lipoprotein receptor activities in therat. FEBS Lett 1987;222:159-162

37. Wttztum JL, Young SG, Elam RL, Caraw TE, Fisher M.Cholestyramine-induced changes in low density lipoproteincomposition and metabolism. I. Studies in the guinea pig. JUpid Res 1985^6:92-103

38. Parka JS, Bullock BC. Effect of fish oil versus lard diets onthe chemical and physical properties of low density lipopro-telns of nonhuman primates. J LJptid Res 1987^8:173-182

Index Terms: fish oils • Maxepa • VLDL • LDL • apolipoprotein B metabolism


Dow

nloaded from


M W Huff and D E Telforddensity lipoprotein.

Dietary fish oil increases conversion of very low density lipoprotein apoprotein B to low

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 1989 American Heart Association, Inc. All rights reserved.

Avenue, Dallas, TX 75231is published by the American Heart Association, 7272 GreenvilleArteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/01.ATV.9.1.581989;9:58-66Arterioscler Thromb Vasc Biol.

http://atvb.ahajournals.org/content/9/1/58World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

http://www.lww.com/reprints

Information about reprints can be found online at: Reprints:

document.Permissions and Rights Question and AnswerFurther information about this process is available in theis being requested is located, click Request Permissions in the middle column of the Web page under Services.Clearance Center, not the Editorial Office. Once the online version of the published article for which permission

can be obtained via RightsLink, a service of the CopyrightArteriosclerosis, Thrombosis, and Vascular Biology Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:


Dow

nloaded from

http://atvb.ahajournals.org/content/9/1/58

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://atvb.ahajournals.org//subscriptions/


dietary fish oil increases conversion of very low density...

Documents