Dietary Fat and Cholesterol Modulate the Plasma Lipoprotein Distribution and Production of Pigment or Cholesterol Gallstones in Hamsters1

Nutrient Metabolism

Dietary Fat and Cholesterol Modulate the PlasmaLipoprotein Distribution and Production of Pigment orCholesterol Gallstones in Hamsters1

K. C. HAYES2, PRAMOD KHOSLA, ANDREA KAISER,

VARTAN YEGHIAZARIANS ANDANDRZEJ PRONCZÃœK

Foster BiomédicalResearch Laboratory, BrandéisUniversity, Waltham, MA 02254

ABSTRACT To evaluate the impact of key dietaryfactors on plasma lipoproteins and gallstone induction,male Syrian hamsters were fed either cholesterol andfat-supplemented purified diets containing glucose orlactose, or cholesterol and fat-free diets with or withoutfiber, for 13 wk. Fat-supplemented hamsters werehyperlipidemic in comparison to those fed fat-free diets.The former group revealed a greatly expanded VLDLfraction, whereas a normal HDL.2pool predominated inthe latter group. Plasma fatty acids indicated that hamsters fed the fat-free diet were essential fatty add deficient whilst the hamsters fed the fat-supplemented dietwere subnormal in essential fatty acids. Ninety-threepercent of the hamsters fed the fat-supplemented diethad gallstones (mostly cholesterol), whereas 62% ofhamsters fed the fat-free diet had gallstones (almost allpigmented). Lactose increased cecal weight andprevented pigment stone formation in the fat-supplemented hamsters, whereas adding fiber to the fat-free diet contributed essential fatty acids, eliminatedcholesterol stones and enhanced pigment stone formation. Thus, diets containing casein, rice flour, glucoseand fiber with minimal essential fatty acids but no cholesterol promoted pigment stones (associated with anormal lipoprotein profile); a low fat diet limited in essential fatty acids but containing cholesterol and lactose, promoted cholesterol gallstones (associated withan expanded VLDLpool). J. Nutr. 122: 374-384, 1992.

INDEXING KEY WORDS:

•hamsters •gallstones•plasma lipoproteins •hepatic llplds•biliary llplds

Cholesterol-rich gallstones in hamsters have beentraditionally induced by two different dietary means,usually in weanling hamsters. One diet represents ano fat, fiber or cholesterol regimen (1); the secondcontains a modest level of butterfat with >0.3% (2) orwithout (3) cholesterol. A serious drawback of thefirst diet is the abnormal physiology that results fromthe associated essential fatty acid deficiency (EFAD)3

and produces an unpredictable diarrhea (referred to as"wet tail") and death as the large bowel flora arenegatively affected (4). The second diet model isbetter tolerated by the hamster as long as large bowelmetabolism is sustained and liver function is notimpaired by excessive absorption of dietary cholesterol.

Various dietary fibers, replacement of glucose withlactose or addition of polyunsaturated fat have had aprotective effect against cholesterol gallstone induction in the EFAD model (1, 5, 6), whereas complexcarbohydrate or lactose included in the formulation ofthe second diet is generally sufficient to sustain largebowel function and normal physiology, especially ifthe hamsters are sexually mature (7).

In a continuing effort to assess the nutritional andmetabolic aspects of these models, these two disparate nutritional paradigms were examined with specific questions in mind: When cholesterol gallstonesare induced by cholesterol and fat feeding (2), wouldlactose augment cecal flora activity to provide a protective effect as it did against pigment stones (7)?Could we, like Dam (1), induce cholesterol gallstoneswith a fat-free diet model and demonstrate the sameprotective effect of fiber? Finally, during induction ofcholesterol gallstones by these opposing dietarymanipulations, could a predictive lipoprotein profilebe identified in association with gallstone induction?

MATERIALS AND METHODS

Animals and diets. Thirty-two male Syrian hamsters (Lakeview strain, Charles River, Wilmington,

Supported by National Institutes of Health Grant DK 35375and the National Dairy Board.

2To whom correspondence should be addressed.

Abbreviations used: EFAD, essential fatty acid deficiency; FCR,fractional catabolic rate; IDL, intermediate density lipoprotein.

0022-3166/92 $3.00 <S>1992 American Institute of Nutrition. Received 31 January 1991. Accepted 23 July 1991.

374

by guest on March 2, 2015

jn.nutrition.orgD

ownloaded from

GALLSTONES IN HAMSTERS 375

MA) were obtained at an initial weight of 90-100 g.They were assigned (in groups of eight) to four different semipurified diets (Table 1) that were fed for aperiod of 13 wk. Of the four diets, the first two dietscontained both cholesterol (0.4 g/100 g) and butterfat(5 g/100 g). These two diets differed solely in theirsugar content, which was provided as 20 g/100 gglucose or lactose, to examine the potentially protective effect of lactose against gallstone induction.The third and fourth diets were devoid of cholesteroland fat, similar to the original Dam formula (1). Although these diets were eucaloric, diet 3 was alsofiber-free, whereas diet 4 contained 25 g fiber/100 gdiet (10 g cellulose plus 15 g wheat bran/100 g diet),which was included at the expense of rice flour andglucose. Hamsters were housed four per cage with a12-h light:dark cycle in which 0300 h was the beginning of the dark period. Hamsters had fiéeaccessto fresh food and water (provided daily). Body weightsand food intake were monitored routinely. All protocols and procedures were approved by the institutional Animal Use Committee.

Triglycéridesecretion rates. After 5 wk of feedingthe diets when lithogenesis was assumed to be inprogress, triglycéridesecretion rates were determinedas an index of lipoprotein metabolism after food waswithheld overnight as described previously (8). Hamsters were fasted individually in wire-bottomed cagesbefore being anesthetized lightly with a gaseousmixture of 50:50 CO2:U2 and injected with 250 mg/kgbody wt of Triton WR1339 via the jugular vein. Bloodsamples (200 jiL) were collected immediately (time 0)from the jugular vein and at 90 and 180 min post-injection by cardiac puncture. Each time hamsterswere under anesthesia for 1-2 min and then returnedto their cages. All animals continued to consumetheir respective diets until analysis of lipoproteinsand gallstones 8 wk later.

Plasma was obtained from blood samples collectedin EDTA, and triglycérideconcentrations were determined using an enzymatic kit (number 336, SigmaDiagnostics, St. Louis, MO). For plasma volume determinations, the Triton concentration was determinedby extracting the plasma with 10 volumes of iso-propanol and measuring the absorbance at 280 nm todetermine the dilution of the injected Triton (8). Apositive correlation (r2 = 0.995) was observed betweenbody weight and plasma volume for hamstersweighing between 90 and 160 g with plasma volumemeasuring 3.80 mL/100 g body wt. The triglycéridesecretion rate was calculated and expressed as milligrams per 100 g per hour. The basal circulating massof triglycérideat the time of Triton injection wasexpressed as milligrams per 100 g of body weight. Thefractional catabolic rate (FCR) for triglycéridewasexpressed as the fraction of the circulating poolcleared per hour, which was derived by assuming thatthe clearance rate equals the secretion rate in thefasting steady state (8-10).

TABLE 1

Composition of hamster purified diets1

Diets

Fat plus cholesterol

Glucose Lactose

No fat orcholesterol

No fiber Fiber

CaseinRiceFlourGlucoseLactoseCelluloseWheat

Bran2Butter3Minerals4Vitamins4Choline

C12Cholesterol2033.520—10554.61.20.30.4g/1002033.5_2010554.61.20.30.4g

diet18.52055.4—4.61.20.3—18.51040.4—1015^_4.61.20.3—

'Diets were fed as flour gels, prepared by withholding 60 g riceflour/kg mix from the formulation and premixing it with 800 mL ofsimmering water to form a gel to which the remaining ingredientswere added.

2The fatty acid composition of wheat bran was (g/100 g totalfatty acids): 12:0-0.2, 14:0-0.1; 16:0-16.4; 16:1-0.3; 18:0-1.3; 18:1-18.6; 18:2-55.4; 18:3-4.0.

*The fatty acid composition of the butterfat was (g/100 g totalfatty acids): 12:0-7.2, 14:0-9.9, 16:0-12.9, 18:0-4.1, 18:1-10.8, 18:2-2.5.

4The composition of the salt mix and vitamin mix was detailed

previously (7).

Necropsy and gallstone evaluation. After 13 wk,hamsters were assessed for plasma lipoproteins andgallstones. Following an overnight fast, hamsters wereexsanguinated under anesthesia by cardiac punctureinto EDTA-wetted syringes. The cecum (from theileo-cecal junction to the tip) and liver were excised,blotted and weighed. Portions of liver were removedfor analysis of cholesterol. Bile was aspirated from thegallbladder and kept for subsequent analysis. The gallbladder was opened under an Olympus Ml 00 dissecting microscope (Marcan Instrument Co., Norwood, MA) with mounted camera, and any gallstonespresent were examined under regular and polarizedlight prior to being photographed. Upon drying, representative gallstones of both cholesterol and pigmentvarieties were weighed and analyzed by HPLC fortheir cholesterol content (11).

Lipoprotein analysis. Plasma was isolated from theterminal blood samples, and its cholesterol and triglycérideconcentrations were determined in order topool similar plasma from 2-3 hamsters for lipoproteinisolation by sequential ultracentrifugation (12) inQuick seal tubes®(Beckman, Palo Alto, CA) using aTi 70.1 rotor. Five fractions were isolated, VLDL (d <

by guest on March 2, 2015

jn.nutrition.orgD

ownloaded from

376 HAYES ET AL.

1.006 kg/L), intermediate density lipoproteins (IDL)(1.006 < d < 1.019 kg/L), LDL (1.019 < d < 1.055 kg/L),HDL2 (1.055 < d < 1.125 kg/L) and HDL3 (1.125 < d <1.21 kg/L). Each isolated lipoprotein fraction waswashed once at its appropriate density. With the exception of VLDL and IDL, all lipoprotein fractionswere dialyzed against 0.15 mol/L NaCl/1 mmol/LEDTA, pH 7.4, at 4'C, for 24-^36 h. Cholesterol and

triglycérideconcentrations were determined using enzymatic kits (numbers 352 and 336 respectively,Sigma Diagnostics, St. Louis, MO). Protein concentration was determined by a modification of theLowry procedure (13).

Hepatic cholesterol analysis. Following lipid extraction of liver samples with chloroform:methanol(2:1 v/v) (14), cholesterol (free and esterified) was extracted by a double solvent procedure [isopropanol-0.75 mol/L NaOH (33:17, v/v), followed by n-octane]for HPLC analysis (11). Briefly, the individualcholesteryl esters were separated using a WatersRadial-Pak Resolveâ„¢C-18 column (Waters Associates, Milford, MA), eluted isocratically withacetonitrile-isopropanol (45:55, v/v) at 2 mL/min witha Beckman HOB solvent delivery module (BeckmanInstruments, Palo Alto, CA). The absorbance of theeluate was measured at 210 nm using a Waters model480 LC spectrophotometer (Waters Associates).

Analysis of bile. Fifty microliters of gallbladder bilewas subjected to lipid extraction (14) in 3 mL ofchloroform-methanol (2:1, v/v) and washed once with750 nL of 0.15 mol/L KCl solution. The lower organicphase was evaporated to dryness and reconstituted in250 pJLof isopropanol. Biliary cholesterol was determined in a 20-(iL aliquot by HPLC (15). Biliary phos-pholipids were determined in a lOO-jiL aliquot following evaporation and digestion with 500 (iL of 6.97mol/L perchloric acid at 180*C for 2 h prior to

analysis for phosphorus (16). Bile salts were measuredin 100 \iL of the methanol layer using the enzyme3a-hydroxysteroid dehydrogenase (EC 1.1.1.50) (17).The lithogenic index was calculated according to published procedures (18), based on the relative molarratios of the lipid components and total lipid using acomputerized version of cholesterol solubility (19).

Statistical analysis. The Statview 512â„¢statisticalsoftware package (Brain Power, Inc., Calabasca, CA)for the Macintosh was used for statistical analysis.Student's unpaired t test was used for comparisons

between glucose and lactose within the cholesteroland fat-supplemented diets and between no fiber vs.fiber within the cholesterol and fat-free diet groups.To determine the overall effect of fat and cholesterol,the data were then collapsed within the two major fatcategories and the t test again applied. The 5-wk and13-wk data (body weights, plasma cholesterol andtriglycérideconcentrations) were analyzed by Student's paired t test. In all cases statistical significancewas set at P < 0.05.

RESULTS

Growth and plasma llpids: 5 wk

After 5 wk the lipid secretion rates were determined. One of the glucose-fed hamsters died duringthis procedure and a second died 3 d later. At 5 wkthe body weights of the glucose-fed hamsters weregreater than those fed lactose, whereas all hamstersreceiving the fat-free diets were similar in weight tothe lactose group. The plasma lipid concentrationswere decidedly lower in the hamsters fed the fat-freediet at this time in comparison with the hamsters fedthe fat-supplemented diet. Plasma cholesterol waslowest in the hamsters fed the fat-free diet withoutfiber. Hamsters fed glucose had the highest triglycérideconcentrations, which were significantlyelevated compared with lactose-fed hamsters (Table2).

Triglycéridesecretion data

Carbohydrate effect. The marked expansion of thetriglycéridepool in the glucose-fed hamsters (P < 0.02)was associated with a tendency (P < 0.10) for a lowerfractional catabolism of triglycéride,whereas the triglycéridesecretion rates were similar between thetwo groups (Table 2).

Fiber effect. The triglycéridesecretion rates were22% lower (P < 0.05) and the fractional catabolic ratewas 30% lower (P < 0.02) in hamsters fed fiber. As aconsequence of these parallel decreases, the circulating triglycéridepool size was not significantly affected by dietary fiber (Table 2).

Fat vs. no fat To compare the effects of cholesteroland fat-supplemented vs. cholesterol and fat-free dietson triglycéridesecretion, the data from the two dietary subgroups (carbohydrate or fiber) within each fatcategory were combined. Using this approach it wasclear that, despite a lower secretion rate, the elevatedplasma triglycéridepool in the fat-supplementedhamsters was attributable to a decreased FCR (Table2).

Terminal body weights and plasma llpids: 13 wk

Body weights. These were not significantly different between hamsters fed glucose or lactose. Similarly, body weights of hamsters fed the two fat-freediets did not differ, although they tended to beslightly lighter than fat-supplemented hamsters(Table 3).

Plasma cholesterol. The average circulating cholesterol concentration in glucose-fed hamsters was

by guest on March 2, 2015

jn.nutrition.orgD

ownloaded from

GALLSTONES IN HAMSTERS 377

TABLE 2

Body weights, plasma lipidi and triglycéride secretion data in hamsters fed fat-supplemented or fat-free diets for S wk1

DietsFat

pluscholesterolGlucoseNumberBody

weight,gInitial5

-weeksPlasma

lipids,mmol/LCholesterolTriglycérideTriglycéride

dynamics2Pool

size,mg/100g bodywtSecretion

rate,mg-100g-i-h-1Fractional

catabolicrate,pools/h1001398.05.117.811.80.86±

1±6"±

0.9±1.1«±

3.8"±

2.0±

0.2Lactose897

±121±6.2

±2.3±7.6

±10.7

±1.6

±243"0.70.4«1.2«1.60.4No

fat orcholesterolNo

fiber892

±1116±32.4

±0.08a1.9±0.16.4

±1.319.1

±1.5*3.0

±0.2»Fiber901214.2188.114.92.18±

1+3±

0.4«±

0.4±

1.4±

1.2"±

0.3«SummaryFat14991297.08J12.011.21.3±

1«±4b±

0.6b±0.6b±

2.2b±

1.2b±

0.2bNofats16911193.32.17.317.02.6±

1«±2b±

0.3b±0.2b±

0.8b±

l.lb±

0.2b

'Values are mean ±SEM.«Valuesin a row with a common superscript were significantly different (P < 0.05), Student's unpaired t test.2Measured following injection of Triton WR1339.

higher than that in lactose-fed hamsters, but thisdifference was not significant due to wide variationamong hamsters. Plasma cholesterol concentrationfor fat-supplemented hamsters was markedly higherthan that for fat-free diet fed hamsters (Table 3).Dietary fiber significantly raised the depressed plasmacholesterol of the fat-free group at both 5 and 13 wk(Tables 2 and 3), and (with the exception of the fibergroup) the mean plasma cholesterol concentrationwas significantly higher after 13 wk compared with 5wk.

Plasma triglycérides.The concentration of triglycéridesin glucose-fed hamsters was about twice ashigh as that in lactose-fed hamsters and both groupsincreased between 5 and 13 wk (Tables 2 and 3). Thepresence of dietary fiber had no significant effect onthe plasma triglycérideconcentration in hamsters fedthe two fat-free diets, although values tended to riseover time in the group fed the fiber-free diet. Likecholesterol, the plasma triglycérideconcentration wasmuch higher in fat-supplemented hamsters comparedwith animals fed the fat-free diet.

Dietary effect on plasma llpoproteln profile

From lipoprotein composition data presented inTable 4, it is apparent that the effect of glucose didnot differ from that of lactose, with the possible exception that glucose tended to increase the VLDLtriglycérideand cholesterol concentrations and signif

icantly increased LDL cholesterol, triglycérideandprotein concentrations. The fiber vs. no fiber comparison showed only modest differences in IDL triglycéridesand LDL cholesterol. However, summarycomparison of fat-supplemented with fat-free diet-fedhamsters revealed that fat and cholesterol feedingsignificantly increased cholesterol, triglycérideandprotein in all lipoprotein fractions with the exceptionof triglycéridein the HDL fraction.

When the relative distribution of the various components in each lipoprotein fraction was calculatedfor the four dietary groups, the main dietary effectwas found between the fat-containing vs. the fat-freeformulas. Neither lactose nor dietary fiber causedappreciable differences within the two major dietaryfat categories. Among hamsters fed the cholesteroland fat-supplemented diets, the majority oflipoprotein cholesterol (40-55%) was found in VLDL.Of the remainder, 25^35% occurred in the HDL2fraction, 10% in both DDLand LDL, and the rest (3%)in the HDL3 fraction. For triglycéride,77-84% waslocated in the VLDL fraction, 6% in both DDLandLDL, 3% in HDL2 and only trace amounts in HDLg(<0.2%). For both the glucose and lactose groups,53-55% of lipoprotein protein was in HDLj, 18-24%in VLDL, 11-14% in HDL3 and 6% in both DDLandLDL.

Among the hamsters fed the cholesterol and fat-free diets, only 10-14% of the cholesterol was foundin the VLDL fraction, 6% in each of the DDLand LDLfractions, 66-69% in HDL2 and 8-9% in HDLa. The

by guest on March 2, 2015

jn.nutrition.orgD

ownloaded from

378 HAYES ET AL.

TABLE 3

Body, liver and cecina weights, plasma lipid concentrations, and gallstone incidence in hamsters fed fat-supplemented or fat-freediets for 13 wk1

DietsFat

pluscholesterolNumberBody

weight, gLiver weight, g/100 g body wtCecum weight, g/100 g bodywtTotal

plasma cholesterol,mmol/LPlasma

triglycéride,mmol/LGallstones2CholesterolPigment

MixedGlucose6152

±85.1±0.31.4

±0.117.6

±3.014.5±4.13:61:6

2:63Lactose8148

±65.4 ±0.32.8

±0.414.4

±2.47.0±1.27:8—No

fat orcholesterolNo

fiber8140

±43.6 ±0.11.3

±0.23.3

±0.2a3.4

±0.91:82:8

1:8Fiber8137

±43.4 ±0.10.9

±0.13.9

±0.2a2.7

±0.3—6:8SummaryFat14150

±5a5.3 ±0.2«2.1

±0.3a15.5

±1.8b10.3±2.1»10:141:14

2:14No

fats16139

±3a3.4 ±0.1a1.1

±0.1"3.6

±O.lb3.0±0.5a1:168:16

1:16Values are means ±SEM.abValues in a row with a common superscript were significantly different (P < 0.05), Student's unpaired t test.^Gallstone incidence; number of hamsters with gallstones:number of hamsters examined.3Mostly cholesterol (see text).

TABLE 4

Lipoprotein lipid and protein concentrations in hamsters fed fat-supplemented or fat-free diets for 13 weeks1

Diets

Fat plus cholesterol No fat or cholesterol Summary

Glucose Lactose No fiber Fiber Fat No fats

Number

VLDLCholesterol, mmol/LTriglycérides,mmol/LProtein, g/L

EDL2Cholesterol, mmol/LTriglycérides,mmol/LProtein, g/L

LDLCholesterol, mmol/LTriglycérides,mmol/LProtein, g/L

HDL2Cholesterol, mmol/LTriglycérides,mmol/LProtein, g/L

HDL3Cholesterol, mmol/LTriglycérides,mmol/LProtein, g/L

6.0 ±1.25.53 ±0.690.75 ±0.23

1.2 ±0.20.42 ±0.020.17 ±0.02

3.9 ±1.53.39 ±1.230.36 ±0.13

0.9 ±0.20.35 ±0.020.13 ±0.02

1.5 ±0.05a0.12±0.003a0.35±0.02a2.5

±0.30.20±0.021.21±0.111.2

±0.03"0.10±0.009a0.27±0.00a2.5

±0.010.16±0.021.07±0.09

0.3 ±0.003 0.3 ±0.01<0.01 <0.010.32 ±0.02 0.27 ±0.01

0.3 ±0.10.76 ±0.180.14 ±0.07

0.1 ±0.0050.26 ±0.03a0.06 ±0.003

0.3 ±0.03b0.08 ±0.010.11 ±0.01

1.3 ±0.080.14 ±0.010.75 ±0.04

0.2 ±0.030.59 ±0.080.05 ±0.01

0.15 ±0.030.16 ±0.02a0.06 ±0.01

0.5 ±0.03b0.07 ±0.010.14 ±0.01

1.7 ±0.10.15 ±0.020.89 ±0.07

0.2 ±0.01 0.2 ±0.02<0.01 <0.010.20 ±0.007 0.24 ±0.03

5.0 ±1.0a4.46 ±0.79a0.56 ±0.15a

1.1 ±0.1"0.38 ±0.02b0.15 ±0.01a

1.3 ±O.lc0.11 ±0.01b0.31 ±0.0b

2.5 ±0.1a0.18 ±0.021.14 ±0.07a

0.3 ±0.05a0.68 ±0.10a0.10 ±0.04a

0.0a0.03b

0.10 ±0.04"

0.10.21

0.40.07

0.05C0.01b

0.13 ±0.01"

1.5 ±0.1a0.15 ±0.010.82 ±0.05a

0.3 ±0.01" 0.2 ±0.02a<0.01 <0.010.30 ±0.0a 0.22 ±0.02a

'Values are means ±SEM.For each determination, lipoproteins were isolated from the pooled plasma of 2-3 hamsters. Recovery of plasmacholesterol, based on the cholesterol recovered in the five lipoprotein fractions, averaged 85% ±2%, with no difference between dietarygroups. Values in a row with a common superscript differ significantly (P < 0.05) by Student's unpaired t test.

2IDL - intermediate density lipoprotein.

by guest on March 2, 2015

jn.nutrition.orgD

ownloaded from

GALLSTONES IN HAMSTERS 379

distribution was not affected by the presence of fiber.The triglycéridedistribution was similar to that seenin hamsters fed the cholesterol and fat-supplementeddiets with the bulk in VLDL. As for protein, 63-69%was in HDL2, 17-19% in HDL3/ 4-11% in VLDL and4-5% in both IDL and LDL. Thus, addition of fat andcholesterol to the diet redistributed an expanded poolof cholesterol from HDL2 to VLDL and increased thepools of VLDL triglycérideand protein. As a consequence, the cholesterohtriglyceride ratio inlipoproteins of the hamsters fed the cholesterol andfat-supplemented diets was fourfold higher in theVLDL and twofold higher in the LDL and HDL2 thanthis ratio in these lipoproteins of the fat-free hamsters.

Hepatic and biliary llplds

Liver cholesterol. Liver weights obtained at 13 wkwere 50% greater in the hamsters fed the fat-supplemented diets when compared with hamstersfed fat-free diets (Table 3). Among hamsters fed thefat-supplemented diets, liver weights were not affected by the type of dietary carbohydrate. Nor didfiber influence liver size in the fat-free dietary groups.However, cecal weights from hamsters fed lactosewere twofold heavier than those from hamsters fedglucose (Table 3). When data from hamsters fed thefat-free diets were pooled for comparison with datafrom hamsters fed the fat-supplemented diets, totaland free hepatic cholesterol and cholesteryl esters(Table 5) were significantly greater in the fat-supplemented hamsters (13-fold, 1.6-fold and 65-fold,respectively). Whereas cholesteryl esters accountedfor 89% of the total liver cholesterol in fat-supplemented hamsters, they represented only 18%of the hepatic cholesterol in hamsters fed the fat-freediet. In all dietary groups, cholesteryl oleate was themajor (55-75%) ester present (data not shown).

Biliary lipid composition. Due to the presence ofgallstones, adequate samples of gallbladder bile werenot obtained from all animals, especially those fedglucose (Table 5). Because of this, a glucose-lactosecomparison was not done. Among hamsters fed thefat-free diet the bile phospholipid concentration wasdepressed in the absence of dietary fiber and theirlithogenic index was 50% higher as a consequence. Inthe fat vs. no fat comparison a threefold cholesterolenrichment of bile (to 8 mol/100 mol) was attributedto inclusion of fat and cholesterol. In addition, hamsters fed the fat-supplemented diet experienced relatively higher biliary phospholipid concentrations andlower bile acids, causing their mean lithogenic indexto exceed 1.7, 2.5-fold higher than that in hamstersfed the fat-free diet.

Gallstones. Examination of gallbladders after 13wk revealed the presence of spherical, white choles-

FIGURE 1 Gallstones are depicted in situ in Syrian hamsters fed purified diets for 13 wk. Top: white, sphericalcholesterol gallstones (87%-91% pure) predominated inhamsters fed diets containing fat and cholesterol. Below.dark-green to black pleomorphic pigment stones (<2% cholesterol) prevailed in hamsters fed no fat or cholesterol.

terol gallstones almost exclusively in hamsters fedthe cholesterol and fat-supplemented diets, whereaspigment gallstones were found primarily in hamstersfed the fat-free diets (Fig. 1, Table 3). Five of the sixhamsters fed the glucose diet and seven of eight hamsters fed lactose had cholesterol gallstones rangingfrom as many as 50 small stones (0.1-1.0 mm) to asfew as one or two large stones (1-3 mm). Althoughthe glucose group included two hamsters with mixedstones, one of these had more than 50 cholesterolstones with two pigment stones and the other hadone large cholesterol stone in which cholesterol waspacked around a pigment core and the surroundingbile was supersaturated with cholesterol crystals. Inthe lactose group, five of seven revealed well-formedcholesterol stones, whereas two of seven had mini-stones of cholesterol monohydrate crystals embedded

by guest on March 2, 2015

jn.nutrition.orgD

ownloaded from

380 HAYES ET AL.

TABLE5

Hepatic cholesterol and biliary lipid composition in hamsters fed fat-supplemented or fat-free diets for 13

DietsFat

pluscholesterolNumberGlucose2Lactose7No fat orcholesterolNo

fiber5Fiber8SummaryFat9No fats13

mg/g liver

LivercholesterolTotalUnesterifiedEsterifiedBiliary

lipidsBileacidsPhospholipidsCholesterol18.52.016.5942312±

2.0±0.1±2.0±

2±0.4±

223.0

±2.6±20.4

±170

±23±16

±4.50.24.431431.5

±1.3±0.2

±155

±7±4

±0.21"

1.90.03"1.60.1*

0.3mmol/L10

1852*190.4

5±

0.2*±0.1*±O.I4±

21±32'±

121.0

±2.3±18.5

±154

±23±15

±0.8bO.lb0.7b263b2«1.7

±1.4±0.3

±173

±14±5

±0.03b0.03b0.01b143b0.4«mol/100molBiliarylipids2Bile

acidsPhospholipidsCholesterolTotal

lipidsLithogenicindex372.818.09.2691.6±

1.5±0.2±1.6±

1±0.380.5

±11.3±8.2

±108

±1.8

±1.70.81.1190.293.0

±4.3±2.7

±83

±1.0

±1.4»

88.61.3"9.10.1

2.3g/L6

1070.1

0.6±

1.0«±1.0«±

0.3±

12±

0.178.8

±12.8±8.4

±99

±1.7

±1.7bl.lb0.9«150.2«90.3

±7.2±2.5

±98

±0.7

±1.0bL#0.2«80.1«

'Values are means ±S£Mof the number of hamsters indicated. Because the glucose group comprised only two hamsters, no statisticalcomparisons were made between the glucose-lactose groups. «Valuesin a row with a common superscript differ significantly (P < 0.05),Student's unpaired t test.

^Calculated by summing the values for bile acids, phospholipids and cholesterol (in mmol/L) and expressing each relative to the total.3The Lithogenic index (mol% cholesterol/maximum solubilizable cholesterol) was calculated according to published procedures, detailed

in Materials and Methods.

in a viscous, mucin gel molded to the shape of thegallbladder. The crystals readily polarized to revealtheir characteristic rhomboid shape with notchedcomer. No pigment stones were seen in lactose-fedhamsters. Analysis of cholesterol gallstones revealed acholesterol content between 87-91%. The incidenceof pleomorphic, dark green-to-black pigment gallstones was six of eight among hamsters fed the fat-free, fiber-supplemented diet, and no cholesterolcrystals were seen. In hamsters fed the fat-free dietwithout fiber, one of eight had formed cholesterolstones, two had pigment stones, and one had mixedstones. The pigment stones ranged from clusters ofsand-like granules too numerous to count, to one-to-ten larger (0.1-1.2 mm) pleomorphic bodies with irregular but smooth nodular protrusions. Compositional analysis typically revealed <2% cholesterol.Bile in gallbladders from the latter hamsters wasalmost never gelatinous or viscous but fluid, translucent and amber-colored.

Fatty acid profiles

Both plasma and adipose tissue were analyzed fortheir fatty acid profiles but only plasma fatty acids arereported (Table 6). In the cholesterol and fat-freegroups, 20:3(n-9) was elevated by comparison to fatand cholesterol-supplemented hamsters, but wasreduced toward normal by fiber supplementation. Surprisingly, 20:4(fl-6) was higher in hamsters fed thefat-free diet than in those fed fat, and highest in thefiber-supplemented group, which also revealed thehighest 18:2(n-6). The traditional EFAD index basedon the 20:3/20:4 ratio was elevated above the deficiency index of 0.4 (to 0.44) in the hamsters fed fiber-free diet but was reduced to 0.22 by the fiber supplement. This ratio was similar and only 0.16 (glucose)and 0.15 (lactose) in the fat and cholesterol-supplemented groups, even though the fat source wasbutter. The level of 18:l(n-9) was significantly lowerin the hamsters fed the fat-free diet. By comparison to

by guest on March 2, 2015

jn.nutrition.orgD

ownloaded from

GALLSTONES IN HAMSTERS 381

TABLE 6

Plasma fatty acid composition in hamsters fed fat-supplemented or fat-free diets for 13 wk1

Diets

Fat pluscholesterolNumber14:016:016:l(n-9)18:018:l(n-9)18:2(n-6)

18:3(n-3)20:3(n-9)20:4(n-6)22:6(n-3)Glucose51.8

±0.423.0±0.57.6±0.74.0±0.145.0±1.510.1±0.7

0.2 ±0.10.3 ±0.021.9 ±0.20.7 ±0.2Lactose61.5

±0.222.0±0.65.9±0.64.6±0.347.5±1.89.9±0.9

0.2 ±0.10.3 ±0.12.0 ±0.30.7 ±0.2No

fatNo

fiber6g/1000.9

±0.122.6±0.85.5±0.36.7±0.540.1±1.19.9±0.4"

0.3 ±0.11.4 ±0.1"3.2 ±0.4"0.8 ±0.2or

cholesterolFiber7g

fattyacids0.9±0.123.6±0.95.6±0.65.6±0.535.9±1.912.6±0.5"

0.3 ±0.10.9 ±0.1"4.1 ±0.4"1.3 ±0.2SummaryFat111.6

±0.2"22.4

±0.46.7±0.54.3±0.2"46.4±1.2"10.0±0.6

0.2 ±0.40.3 ±0.04b2.0 ±0.2b0.7 ±0.1"No

fats130.9

±0.1"23.1

±0.65.5±0.36.1±0.4"37.8±1.3"11.4±0.5

0.3 ±0.11.1 ±O.lb3.7 ±0.3b1.1 ±0.0"

1Values are means ±SEMof the number of hamsters indicated.Student's unpaired t test.

Values in a row with a common superscript differ significantly (/' < 0.05),

other hamsters fed similar diets with a higher polyenecontent (unpublished data), the sum of the (n-9) fattyacids for hamsters in this study was approximatelydoubled and the (n-6) fatty acids were less than halfthe normal values.

DISCUSSION

The novel aspect of this study is its comparison ofdiet-induced changes in lipoprotein metabolism inhamsters fed two disparate diets, each known toinduce cholesterol gallstones. Because it is unclearhow lactose and fiber alter lithogenesis in Syrianhamsters (7)or how dietary fat and cholesterol impactcholesterol metabolism to modulate gallstone formation in this model, we fed these diets to sexuallymature hamsters, incorporating a positive andnegative control for each diet. For example, in thefirst experiment the fat and cholesterol-supplementeddiets compared glucose with lactose, the latter ofwhich has previously protected against both cholesterol and pigment gallstones (1, 7). In a second experiment fat and cholesterol-free diets were fed toexamine the protective effect of dietary fiber (1, 5, 6)against cholesterol gallstones. These comparisonsdemonstrate that the sexually mature male Syrianhamster (Lakeview strain) is highly susceptible toexperimental induction of either cholesterol orpigment gallstones, the outcome depending on thedietary manipulation.

Lactose. In the first comparison lactose actuallyseemed to insure the formation (88% incidence) ofpure cholesterol gallstones by precluding pigment

stone formation, whereas in the glucose group all sixsurviving hamsters developed gallstones, but three ofthese included some pigment stones. In his originalstudies, Dam (1, 4) demonstrated the protective effectof lactose against cholesterol gallstones when feedinga fat-free EFAD diet to weanling hamsters and attributed the protection to the enhanced activity of thelarge bowel flora. He also found older animals (suchas ours) to be less apt to develop cholesterol stonesand more apt to have pigment stones. On the otherhand, we previously found that lactose protectedagainst pigment stone formation in sexually maturehamsters when estrogen supplementation of a highglucose diet was the inducing regimen (7).

In all these situations lactose seems to have exerted its influence by augmentation of large bowelmetabolism. In the present study, lactose-fed hamsters had twice the cecal size of the glucose-fedanimals, but this contrast between dietary carbohydrates was not generally reflected in lipoprotein composition or metabolism. Lactose did tend to lowerboth circulating VLDL cholesterol and triglycéridesand to increase the FCR of triglycérides.In our previous study (7), lactose also decreased plasma triglycérides.In that study the concentration of gallbladderbile acids also seemed to be enhanced by lactose.Limited bile samples in the glucose group of thepresent study prevent a definitive statement on thispoint. Therefore, in the sexually mature malehamster fed butterfat and cholesterol, lactose seemedto prevent pigment stones but not cholesterol gallstones, and the mechanism of its influence was notevident from the lipoprotein or bile lipid profiles generated.

by guest on March 2, 2015

jn.nutrition.orgD

ownloaded from

382 HAYES ET AL.

Fiber and essential fatty acid deficiency. Oursecond dietary comparison examined the effect offiber in hamsters fed a cholesterol and fat-free dietsimilar to the original Darn diet (1). However, 20% ofhis all-glucose formula was replaced with rice flour tolimit the likelihood of "wet tail" (1, 4, 7), and no

diarrhea or deaths occurred; the cholesterol gallstonesof Dam's weanling model were almost totallyreplaced by pigment stones in our more mature ham-sters. In fact, only two of the hamsters fed the fat-freediet developed any cholesterol stones, whereas eightothers had pigment stones. This was essentially consistent with Dam's observation that mature animalsfed a fat-free diet are more apt to develop pigmentstones and that substantial replacement of glucosewith crude carbohydrate contributes to this effect.

When we added 25% fiber to the fat- and fiber-freeformula, the incidence of gallstones actually increasedfrom 50 to 75%, but all were pigment stones. Dietaryfiber tended to normalize the moderately depressedplasma cholesterol without affecting plasma triglycérides, although it reduced the extraordinary triglycéridesecretion rate seen in the group fed the fat- andfiber-free diet. Adding fiber also decreased substantially elevated triglycérideFCRs by approximatelyone-third, but comparison of the lipoprotein profilesbetween the two fiber groups was uninformative.

The mechanism of the fiber effect (i.e., reducingcholesterol stones and increasing pigment stones) isunclear, but two aspects of this relationship deservemention. First, the fact that added fiber was associated with a slightly smaller cecum was surprising,but the observation supports the counter argumentthat an enlarged cecum with enhanced metabolism ofits flora (vis-à-vis lactose) may protect againstpigment stones. Second, the 15% wheat bran contributed about 300 mg fat (mostly polyenes) per 100 gdiet and noticeably improved the essential fatty acidstatus (20:3/20:4 ratio and 18:2) of the plasma fattyacid profile in conjunction with a reduced triglycéridesecretion rate. In this context it is noteworthy that anelevated triglycéridesecretion rate is characteristic ofEFAD in rats (20). Furthermore, the percentage (mol/100 mol) of bile phospholipids more than doubledwhen fiber was added to the fat- and fiber-free diet,suggesting an increased availability of the polyenes(18:2) required for synthesis of hamster bile phospholipids (1, 21). The result was a 30% lower lithogenicindex.

The overall lithogenic index for the two fat andcholesterol-free groups was <1.0 and lower than thattypically found by Dam in his weanling hamsters fedsuch diets (1), suggesting that our older hamsters hadacquired a better reserve of EFA than that present intypical weanlings. Whether the small amount of poly-unsaturated fat contributed by wheat bran in ourstudy actually modulated gallstone formation awaitsfurther investigation, but Dam found that as little as

2% dietary polyunsaturated fat largely prevented cholesterol gallstones in his EFAD model (1).

One should be aware that the fibers fed in thisstudy (cellulose and wheat bran) were selected fortheir relative lack of influence on plasma lipids andlipoproteins (22). Whether soluble fibers would exertspecific effects on lipoproteins and lithogenesis inthese dietary models awaits investigation.

Fat and cholesterol. The most revealing comparison was that between fat-supplemented and fat-free diet-fed hamsters. Although not strictly validbecause multiple changes were introduced in fat, cholesterol, carbohydrate and fiber, it is obvious from theother comparisons already discussed that the dietaryfat and cholesterol difference represented the majordietary manipulation effecting the profound differences observed in lipid metabolism and gallstone incidence. Whereas 10 of 14 hamsters fed the fat andcholesterol diets had cholesterol gallstones, only 1 of16 of the fat and cholesterol-free animals was soafflicted; yet 8 of 16 of the latter had pigment stones,whereas only 1 of 14 of the former did. Thus, our fatand cholesterol diets (especially with lactose) essentially assured induction of cholesterol gallstones,whereas the fat and cholesterol-free diets (especiallywith added wheat bran) favored production ofpigment stones, although at an overall lower incidence rate. Finally, the livers from hamsters fed thefat and cholesterol-free diets did not accumulate cholesterol.

hi fact, a major difference between cholesterol andpigment gallstone induction was the concentration ofliver cholesterol. When the liver was cholesterolladen, bile was supersaturated and cholesterol gallstones predominated. From this and other studies, thebile lithogenic index in hamsters seems to be directlycorrelated with the hepatic cholesterol concentrationwhen cholesterol is supplemented (23), but not in thefat-free, all-glucose diet model (23-25), in which thelithogenic index seems inversely related to essentialfatty acid status (1, 26-28). Dietary polyunsaturatedfat and cholesterol both affect EFA status, andlithogenicity in the EFAD model seems to bemitigated by the metabolic contribution of a large,active cecum and colon (4, 7). On this basis it wouldseem that dietary history (e.g., affecting adipose storesof polyunsaturates or the growth and development ofthe large bowel flora) could influence both biliarylipid secretion and bile lithogenicity (29).

Neither the propensity of certain hamsters to develop pigment gallstones when fed relatively balancedpurified diets nor the protective roles of polyunsaturates and the cecum against lithogenic bile arewell understood. It has been posited that polyunsaturates and cecal floral activity enhance lecithinand bile acid synthesis and secretion into bile, reducing the relative cholesterol saturation (27). Wefound evidence for increased bile phospholipid se-

by guest on March 2, 2015

jn.nutrition.orgD

ownloaded from

GALLSTONES IN HAMSTERS 383

cretion in the presence of dietary fiber (contributing asmall amount of polyenes) associated with an improved lithogenic index. The putative protection ofdietary polyenes (or fiber) against gallstones maypertain to humans as well, as a protective associationbetween vegetable oil and protein against gallstoneswas recently noted in women (30).

Lipoproteins. Another major point concernsplasma lipoproteins and lithogenesis. Although thelipoprotein profiles observed in hamsters bearingpigment or cholesterol gallstones tended to differ appreciably, both types of stones were found in hamsters expressing disparate lipoprotein profiles. Greatlyexpanded pools of VLDL and IDL cholesterol and areduced pool in HDL2 predominated in hamsters withcholesterol gallstones. In a like manner, expandedVLDL production and decreased HDL, primarilyHDL2, are common findings in obese, hyperinsuli-nemic humans with gallstones (31, 32). By assessingthe postabsorptive dynamics of hepatic lipid secretionand clearance from plasma using triton blockade, wefound that hamsters developing cholesterol gallstonestended to have lower triglycéridesecretion rates thanhamsters developing pigment stones. However, thefractional catabolism of triglycérideswas lower incholesterol-stone formers. Together, these datasuggest a decreased triglycérideproduction ratecoupled with impaired clearance of the greatly expanded pool of triglycérideassociated with the VLDL-IDL fractions. This impaired clearance was coupledwith the marked accumulation of hepatic cholesterol,which has been correlated with secretion of lithogenic bile in hamsters (23). The predominance of theHDL2 cholesterol pool observed in the fat andcholesterol-free hamsters is characteristic of thenormal lipoprotein profile in this species (unpublished data) and provides no clue as to lipoproteininvolvement in pigment stone formation, except toemphasize that pigment stones can form when totalplasma lipids are low, or normal, and the lipoproteinprofile is unremarkable.

LITERATURE CITED

1. Dam, H. (1969) Nutrition aspects of gallstone formation withparticular reference to alimentary production of gallstones inlaboratory animals. World Rev. Nutr. Diet. 11: 200-239.

2. Cohen, B. I., Matoba, N., Mosbach, E. H. &. McSherry, C.(1989) Dietary induction of cholesterol gallstones in hamstersfrom three different sources. Lipids 24: 151-156.

3. Tanimura, H. (1965) Experimental studies on the etiology ofcholelithiasis. Arch. Jpn. Chir. 34: 1160-1179.

4. Snog-Kjaer, A., Prange, I., Christensen, F. & Dam, H. (1963)Alimentary production of gallstones in hamsters. 12. Studieswith rice starch diets with and without antibiotics. Z. Er-naehrungswiss 4: 14-25.

5. Bergman, F. & van der Linden, W. (1975) Effect of dietary fibreon gallstone formation in hamsters. Z. Ernaehrungswiss 14:217-223.

6. Rotstein, O. D., Kay, R. M., Wayman, M. & Strasberg, S. M.(1981) Prevention of cholesterol gallstones by Ugni n and lac-tulose in the hamster. Gastroenterology 81: 1098-1103.

7. Hayes, K. C., Stephen, Z. F., Pronczuk, A., Lindsey, S. &.Verdón, C. (1989) Lactose protects against estrogen-inducedpigment gallstones in hamsters fed nutritionally adequate purified diets. J. Nutr. 119: 1726-1736.

8. Nicolosi, R. J., Hererra, M. G., El Lozy, M. & Hayes, K. C.(1976) Effect of dietary fat on hepatic metabolism of [14C]oleic

acid and very low density lipoprotein triglycéridein the gerbil.J. Nutr. 106: 1279-1285.

9. Wong, E. K., Nicolosi, R. J., Low, P. A., Herd, J. A. &.Hayes, K.C. (1980) Lecithin influence on hyperlipemia in rhesus monkeys. Lipids 15: 428-^33.

10. Bird, M., Williams, M. A. &. Baker, N. (1984) Triacylglycerolsecretion in rats: validation of a tracer method employingradioactive glycerol. J. Nutr. 114: 1978-1985.

11. Kim, J. C. &. Chung, T. H. (1984) Direct determination of freecholesterol and individual cholesteryl esters in serum byHPLC. Korean J. Biochem. 16: 69-77.

12. Havel, R. J., Eder, H. A. &. Bragdon, J. H. (1955) The distribution and chemical composition of ultracentrifugally isolatedlipoproteins in human serum. J. Clin. Invest. 34: 1345-1353.

13. MarkweU, M. A., Haas, S. M., Bieber, L. L. &. Tolbert, N. E.(1978) A modification of the Lowry procedure to simplifyprotein determination in membrane and lipoprotein samples.Anal. Biochem. 87: 206-210.

14. Folch, J., Lees, M. & Stanley, G.H.S. (1957) A simple methodfor the isolation and purification of total lipids from animaltissue. J. Biol. Chem. 226: 497-509.

15. Stephan, Z. F., Lindsey, S. &. Hayes, K. C. (1987) Taurineenhances low density lipoprotein binding internalization anddegradation by cultured Hep G2. J. Biol. Chem. 262:6069-6073.

16. Bartlett, G. R. (1957) Phosphorous assay in column chromatog-raphy. J. Biol. Chem. 234: 466-468.

17. Turley, S. D. &. Dietschy, J. M. (1978) Re-evaluation of the3a-hydroxysteroid dehydrogenase assay for total bile acid inbile. J. Lipid. Res. 19: 924-928.

18. Carey, M. C. &. Small, D. M. (1978) Physical chemistry ofcholesterol solubility in bile: relationship to gallstone formation and dissolution in man. J. Clin. Invest. 61: 998-1026.

19. Carey, M. C. (1978) Critical tables for calculating the cholesterol saturation of native bile. J. Lipid Res. 19: 945-955.

20. Huang, C. H. & Williams, M. A. (1980) Essential fatty aciddeficiency and plasma triglycéride turnover in rats. Am. J.Phys. 238: E499-E505.

21. Kajiyama, G., Kubota, S., Sasaki, H., Kawamoto, T. &.Miyoshi,A. (1981) Lipid metabolism in the development of cholesterolgallstones in hamsters. IH. The effect of dietary cholesterol onits bilary concentration and lithogenecity. Hiroshima J. Med.Sci. 30: 141-148.

22. Judd, P. A. &. Truswell, A. S. (1985) Dietary fibre and bloodlipids in man In: Dietary Fibre perspectives, (A. R. Leeds, ed.),pp. 23-39, John Libbey & Co., London, UK.

23. Anderson, J. M. & Cook, L. R. (1986) Regulation of gallbladdercholesterol concentration in the hamster. Role of hepatic cholesterol level. Biochim. Biophys. Acta 875: 582-592.

24. Robins, S. J. & Fasulo, J. (1973) Mechanism of lithogenic bileproduction: studies in the hamster fed an essential fatty acid-deficient diet. Gastroenterology 65: 104—114.

25. lijima, Y., Yamazaki, M. &. Maruyama, M. (1979) Effects ofdietary fatty acids on hepatic HMG-CoA reducíase activity inhamsters fed a high-glucose diet. Arch. Biochem. Biophys. 196:265-269.

26. Dam, H., Prange, L &. Christensen, F. (1965) Alimentary production of gallstones 17. Influence of butter fat and linoleicacid-rich margarine. Z. Ernaehrungswiss 6: 97-106.

27. Shioda, R. (1965) Experimental studies on gallstone formation.Arch. Jpn. Chir. 34: 571-585.

by guest on March 2, 2015

jn.nutrition.orgD

ownloaded from

384 HAYES ET AL.

28. Tanimura, H. (1965) Experimental studies on the etiology ofcholelithiasis. Arch. Jpn. Chir. 34: 1160-1179.

29. Cohen, B. L, Mosbach, E. H. & McSherry, C. K. (1989) Effect ofprevious nutritional status on the formation of cholesterolgallstones in the prairie dog. Lipids 23: 798-803.

30. Maclure, K. M., Hayes, K. C., Colditz, G. A., Stampfer, M. J. &.Willet, W. C. (1990) Dietary predictors of symptom-associated

gallstones in middle-aged women. Am. J. Clin. Nutr. 52:916-922.

31. Howard, B. V. (1986) Obesity, cholelithiasis, and lipoprotcinmetabolism in man. Atheroscler. Rev. 15: 169-186.

32. Scragg, R.K.R., Clavert, G. D. &. Oliver, J. R. (1984) Plasmalipids and insulin in gallstone disease: a case-control study.Brit. Med. J. 289: 521-525.

by guest on March 2, 2015

jn.nutrition.orgD

ownloaded from