FUNDAMENTAL AND APPLIED TOXICOLOGY 4, 819-826 (1984)

Hepatic Steatosis in Rats Fed Diets with Varying Concentrations of Sucrose

BRUCE R. BACON, C. H. PARK,* ELVIN M. FOWELL, AND CHRISTINE E. MCLAREN

Departments of Medicine and *Pathology, Case Western Reserve University School of Medicine at Cleveland Metropolitan General Hospital, Cleveland, Ohio 44109

Hepatic Steatosis in Rats Fed Diets with Varying Concentrations of Sucrose. BACON, B. R., PARK, C. H., FOWELL, E. M., AND MCLAREN, C. E. (1984). Fundam. Appl. Toxicol. 4, 819- 826. The role of dietary sucrose concentrations in the development of hepatic steatosis in rats was investigated. Twelve groups of weanling male Sprague-Dawley rats received semipurified diets with different sucrose concentrations ranging from 20 to SW (w/w); one group received a cereal-based chow diet. Rats were sacrificed after 3 weeks and body weight, liver/body weight ratio, plasma alanine aminotransferase concentration, hepatic triglyceride concentration, and liver morphology (light and electron microscopy) were determined. Body weight and liver/body weight ratio were decreased in rats receiving 40-50 or 25-3596 dietary sucrose compared to tats receiving 20% sucrose or chow. Plasma alanine aminotransfemse concentrations were within normal limits. Hepatic triglyceride concentration was significantly increased in rats receiving 40- 50 and 25-351 dietary sucrose compared to rats receiving 20% dietary sucrose or chow. Light microscopy showed hepatic steatosis in a periportal distribution at all concentrations of dietary sucrose. Both the frequency and the severity of the steatosis were increased with increasing dietary sucrose concentrations. Electron microscopy from selected livers with increased hepatic triglyceride concentrations revealed increased lipid spheres and increased smooth endoplasmic reticulum without prominent Golgi apparatus or GERL complex. It is concluded that high dietary sucrose concentrations are responsible for the development of hepatic steatosis. Semipurified diets with high dietary sucrose concentrations such as the AIN-76A diet (50% sucrose) should not be used in animal studies in which increased triglyceride deposition could influence experimental outcome. 8 1984 Society of Toxicology.

Carefully defined purified diets with known amounts of nutrients, vitamins, and minerals have been recommended for use in animal studies of toxicity and carcinogenicity. The National Research Council Committee on Laboratory Animal Diet has emphasized the need to use such defined diets in order to make valid comparisons of studies in different lab- oratories (Newberne et al., 1978). For this rea- son, the American Institute of Nutrition (AIN) Committee on Standards for Nutritional Studies has recommended a purified diet

I To whom correspondence and requests for reprints should be sent: Division of Gastroenterology, Cleveland Metropolitan General Hospital, 3395 Scranton Rd., Cleveland, Ohio 44 109.

composed of commercially refined proteins (casein), carbohydrates (sucrose, cornstarch), fat (corn oil), and fiber (cellulose) with added mineral and vitamin mixtures and supple- mental choline and methionine (AIN-76A Purified diet) (Bieri et al., 1977; Bieri, 1980).

Since the introduction of this diet there have been reports of excessive bleeding (Roebuck et al., 1979; Medinsky et al., 1982), nephro- calcinosis (Nguyen and Woodard, 1980; Med- insky et al., 1982), and fatty liver (Medinsky et al., 1982; Hamm et al., 1982) in rats fed the AIN-76A diets. Excessive bleeding has been prevented by increasing the level of vi- tamin K added to the diet (Bieri, 1980). Prob- lems with the development of fatty liver appear to be related to the high sucrose concentration

819 0272X)590/84 $3.00 Copyci&t Q 1984 by the society of Toxicology. All rights of reproduction in any fom esemecl.

820 BACON ET AL.

used in the diet (Hamm et al., 1982), and make the use of such a diet questionable in animal studies that require normal hepatic morphology and/or function. The current study was designed to determine more pre- cisely the role of dietary sucrose concentration in the development of hepatic triglyceride ac- cumulation.

METHODS

Diets. Twelve experimental diets were prepared and kindly donated by ICN Nutritional Biochemicals (Cleve- land, Ohio). The formulation of these 12 experimental diets is shown in Table 1. Purina Lab Chow (Ralston Purina Company, Chicago, Ill.) was the closed formula cereal-based diet used as the control diet.

Animals. One hundred forty-eight weanling (50-75 g) male rats of the Sprague-Dawley strain were purchased from &&Miller Laboratories, Inc. (Allison Park, Pa.). Rats were housed in polyethylene cages with stainless steel wire tops and were allowed diet and tap water ad lib. Rats were weighed at the beginning of the study and at weekly intervals thereafter. Sixteen rats were sacrificed at the be- ginning of the study period to provide baseline data. Them were 10 rats in each of the 12 experimental groups and 12 in the group which received chow (control group). Rats

TABLE 1

FORMULATION OF EXPERIMENTAL DIETS (W)

Group Sucrose Cornstarch Casein Alphacel

1 50 15 20 2 45 15 20 3 40 15 20 4 40 20 20 5 40 20 25 6 40 25 20 7 35 15 20 8 35 20 25 9 30 15 20

10 25 15 25 11 20 20 30 12 20 20 25 13 Purina Lab Chow

5 10 15 10 5 5

20 10 25 25 20 25

The remaining 10% of each diet consisted of corn oil-51, AIN mineral mix-3.5% AIN vitamin mix- 1.0% DL-methionine-O.396, and choline bitartrate- 0.2%. Diet No. 1 is the recommended formulation of the American Institute of Nutrition (AIN-76A) (Bieri et al.. 1977; Bieri, 1980).

were fed the diets for 21 days and then fasted overnight prior to sacrifice, which was by exsanguination via cardiac puncture while under light ether anesthesia. The liver was excised, weighed, and portions from the right lobe were saved for light microscopy, electron microscopy, and analysis of triglyceride concentration.

Lighf microscopy. Blocks of liver were fixed in 10% (w/v) neutral buffered fotmalin, dehydrated with a graded series of ethyl alcohol, and embedded in paraffin. Sections (5 pm) were prepared and stained with hematoxylin and eosin. Frozen sections of selected livers were stained with oil red 0 to assess hepatic neutral lipid (Luna, 1968).

The histologic sections from all rat livers were evaluated blindly by one of the investigators (C.H.P.). Rat liver his- tology was evaluated for the presence or absence of hepatic steatosis and for the presence of any other hepatic ab normality. For the purpose of this study, we have graded the presence of steatosis as follows: none-no macroves- icular fat seen anywhere within the entire section of liver tissue, mild-less than 10% replacement with fat, mod- erate- 10 to 25% replacement with fat, and severe-greater than 25% replacement with fat.

Electron microscopy. Blocks of selected livers were fixed (4-6 hr) in 2.5% (w/v) glutaraldehyde, postfrxed in osmium tetroxide, dehydrated in a graded series of acetone, and embedded in Araldite (R. P. Cargille Laboratories, Inc., Cedar Grove, N.J.). Sections (1 pm) were stained with methylene blue and viewed by light microscopy. Repre- sentative areas were further sectioned (ultrathin) for elec- tron microscopy. Ultrathin sections were stained with uranyl acetate and lead citrate.

Hepatic triglyceride concentration. Liver was stored fro- zen (-20C) until the assay was performed. Hepatic lipids were extracted with chloroform/methanol (2: 1) according to the method described by Folch et al. (1954) with the modification of Bligh and Dyer (1959). Phospholipids were separated and removed by batch chromatography with silicic acid (Schotz and Recknagel, 1960). Triglycerides were determined by glycerol analysis alter saponification of the nonphospholipid fraction (Pinter et al.. 1967) and are expressed as milligrams triglyceride per gmm liver (wet wt).

Plasma alanine aminotransferase. Blood was drawn by

fuged. The plasma was assayed for alanine aminotrans- ferase (ALT) concentration according to the method of I&men et al. (1955), with the modification of Henry et al. (1960) using a Beckman Trace III spectrophotometer (Beckman Instruments, Inc., Fullerton, Calif.).

Statistical methods. Data were analyzed using BMDP computer programs (Dixon, 1977). For each variable, one way analysis of variance was used to test for statistically significant differences between the means of the 13 diet groups ( 12 experimental, 1 control) and between the means of 4 groups determined by dietary sucrose concentration (40-50% sucrose, 25-3596 sucrose, 20% sucrose, and chow). When the hypothesis of equality of variances was rejected

HEPATIC STEATOSIS DUE TO DIETARY SUCROSE 821

TABLE 2

EFFECT OF DIETARY SUCROSE CONCENTRATION ON RAT GROWTH, PLASMA ALT, AND HEPATIC TRIGLYCERIDE CONCENTRATIONS

Group

I 2 3 4 5 6 7 8 9

10 11 12 13

a Means f SE.

n

IO 10 10 10 10 10 9

10 10 10 10 8

12

MY weight at

sacrifice (g)

199.8 * 8.6 196.8 f 4.3 206.2 f 4.0 194.5 f 7.3 201.5 + 2.4 196.3 zk 5.6 207.0 + 8.5 202.9 + 7.3 226.0 + 6.4 198.5 2 6.8 216.8 f 3.6 238.1 k 4.4 227.6 + 8.7

Liver/body weight (%)

3.50 k 0.16 3.20 f 0.05 3.40 1 0.09 3.59 f 0.09 3.40 f 0. IO 3.35 2 0.06 3.60 f 0.09 3.76 f 0.14 3.39 + 0.04 3.58 + 0.14 3.79 zk 0.1 I 3.73 f 0.10 4.08 f 0.10

Plasma Hepatic ALT triglyceride

(IU/liter) (mg/g liver)

45.4 * 2.2 21.9 f 5.7 37.9 + 2.2 25.8 + 4.1 35.3 + 2.1 30.1 f 5.3 38.6 f 1.1 20.3 k 3.6 43.5 + 2.5 20.9 + 4.1 40.8 + 2.1 24.2 + 4.3 36.8 + 2.8 12.9 f 2.9 31.2 + 2.1 27.9 + 6.9 28.6 f 2.9 20.7 + 5.8 23.3 + 1.6 20.8 + 5.7 33.8 f 3.4 7.0 f 1.5 33.2 2 2.6 10.1 + 2.4 41.9 f 3.1 8.0 f 1.5

at the 0.05 significance level by Levenes test, the Brown- Forsythe analysis of variance statistic was used. To assure an overall significance level of 0.05, the Bonferoni multiple comparisons procedure was applied to test for pairwise differences between the means (Ahfi and Azen, 1979).

RESULTS

During the course of the 3-week study pe- riod two rats died (one each from Groups 7 and 12) and one rat (from Group 12) had lost considerable weight by the time of sacrifice and was found to have pneumonia.

For the sixteen weanling rats that were sac- rificed at the beginning of the study period the following results were obtained (mean f SE): body weight, 59.4 f 1.3 g; liver/body weight, 3.20 f 0.05%; plasma ALT, 28.4 + 3.3 III/liter; and hepatic triglyceride concentra- tion, 5.2 + 0.5 mg/g. No hepatic steatosis was seen by light microscopy.

The mean (*SE) for each variable (body weight, liver/body weight ratio, plasma ALT, and hepatic triglyceride concentration) for rats from all 13 diet groups is shown in Table 2. One-way analysis of variance showed that the means of the 13 diet groups differed signifi-

I- P

a22 BACON ET AL.

- p

HEPATIC STEATOSIS DUE TO DIETARY SUCROSE 823

FIG. 3. Light microscopy of rat liver. This photomicrograph (hematoxylin and eosin stain) is of liver tissue from a rat that had been fed a 40% sucrose diet for 3 weeks. Macrovesicular fat can be seen in a periportal distribution (P = portal triad, C = central vein). X105.

creased triglyceride concentrations showed electron-lucent spaces (Fig. 5). Mitochondria numerous large electron-lucent lipid spheres showed marked variability in size and shape distributed throughout the cytoplasm. Some (Fig. 5) and smooth endoplasmic reticulum lipid spheres were heterogenous containing was increased with both vesicular and non- areas of finely granular osmiophilic material vesicular forms present (Fig. 6). Primary and thought to be triglyceride (Peterson, 1974) and secondary lysosomes were scarce. Golgi zones

PIG. 4. Light microscopy of rat liver. This photomicrograph (hematoxylin an eosin stain) is of liver tissue from a rat that had been fed a 40% sucrose diet for 3 weeks. Both microvesicular (arrows) and macrovesicular fat can be seen in this mid- and centrilobular zone (C = central vein). X340.

824 BACON

FIG. 5. Electron microscopy of rat hepatocyte. This electron micrograph is from a rat that had been fed a 25% sucrose diet for 3 weeks. Multiple heterogenous lipid spheres can be seen containing finely granular osmiophilic material (triglyceride, arrows) and electron-lucent spaces. Mitochondria showed marked variability in sixe and shape x3700.

were not prominent and vacuoles containing very low density lipoprotein (VLDL) within Golgi zones were not found. No saccules con-

ET AL.

taming VLDL were found in the area of the sinusoidal plasma membrane.

DISCUSSION

In the present study, we have demonstrated both biochemical and morphological evidence of increased hepatic steatosis in male Sprague- Dawley rats fed the AIN-76A diet (containing 50% sucrose) and similar semipurified diets containing lesser amounts of sucrose. Only when the dietary sucrose concentration was reduced to 20% did the hepatic triglyceride concentration approximate that found in rats which were fed the cereal-based control diet (Fig. 2). This suggests that a dietary sucrose concentration of 20% may be the upper limit allowable in a semipurified diet to avoid sig- nificant hepatic steatosis. Even at this dietary concentration of sucrose (20%), 6 of 18 rats still had a histologically identifiable slight in- crease in hepatic steatosis. This level (20% su- crose) is far lower than the amount recom- mended by the Ad Hoc Committee on Stan- dards for Nutritional Studies (Bieri et al.,

FG 6. Electron microscopy of rat hepatocyte. This electron micrograph is from a rat that had been fed a 25% sucrose diet for 3 weeks. Electron-lucent lipid spheres can be seen in close proximity to mitochondria. Also, there is an increased amount of vesicular smooth endoplasmic reticulum (arrows). X 13,500.

HEPATIC STEATOSIS DUE TO DIETARY SUCROSE 825

1977). Our study confirms and extends the findings reported by Medinsky et al. ( 1982) who demonstrated increased hepatic steatosis in a periportal distribution in male Fischer- 344 rats fed three different lots of the AIN- 76A diet prepared by two different suppliers. Rats received the diets for up to 8 weeks and progressively increasing amounts of periportal lipidosis were observed with increasing lengths of time that the rats received the diets. Ad- ditionally, it has been shown that if sucrose is replaced entirely by cornstarch the hepatic steatosis does not appear (Hamm et al., 1982).

We observed slight but significant differ- ences in body weight and liver/body weight ratio between groups, such that with decreas- ing dietary sucrose concentration there was increasing body weight and liver/body weight ratio after the 3-week feeding period (Fig. 1A and B). These findings are consistent with what was found by the Ad Hoc Committee on Standards for Nutritional Studies. In their re- port, the body weight and the liver/body weight ratio were greater in male Sprague- Dawley rats fed the cereal-based chow diet than in rats fed the AIN-76A diet for 2 1 days (Bieri et al., 1977). This is in contrast to what was observed by Medinsky et al. (1982), who noted no change in body weight between male Fischer-344 rats that were fed either the AIN- 76A diet or a control chow diet. The liver/ body weight ratio however was greater for rats fed the AIN-76A diet compared to controls (Medinsky et al., 1982). These differences be- tween studies may be related to changes in rat strains or perhaps to different lengths of time the rats were receiving the various diets. In our own study, the differences in body weight and liver/body weight ratio were sta- tisticallly significant between the dietary groups (Fig. 1 A and B) but the relative changes were actually quite small.

Hepatic steatosis can be caused by many alterations in hepatic triglyceride synthesis and secretion. An increase in synthesis without a concomitant increase in secretion can result in lipid accumulation in hepatocytes. Peri- portal hepatic steatosis can be seen in Kwash-

iorkor, a human condition which results from severe protein deficiency. A major cause of fatty liver in Kwashiorkor is thought to be defective hepatic removal of very low density lipoprotein (VLDL), presumed to be due to decreased synthesis of VLDL apoproteins (Hoyumpa et al., 1975). Also, dietary defi- ciencies of methionine and other lipotropic factors can cause hepatic steatosis. Adequate protein concentration (20-30% casein) and methionine and choline supplementation were provided in all of our experimental diets, thus these deficiencies could not be considered to be contributory to the development of the he- patic steatosis.

The mechanism whereby a high dietary su- crose concentration causes periportal hepatic steatosis in unknown. High sucrose diets in conjunction with hepatotoxins and in genet- ically obese Zucker rats have resulted in the formation of hepatic steatosis. Novikoff et al. (1974) have demonstrated that hepatic stea- tosis (two times control triglyceride concen- tration) could be produced by feeding 0.25% chlorophenoxyisobutyrate (clofibrate) to rats which were receiving a diet containing 60% sucrose. This increase in hepatic triglyceride concentration however, may have resulted from the high dietary sucrose concentration alone. In contrast, severe hepatic steatosis (9 to 20 times control) was produced when 1 .O% erotic acid was added to the high sucrose diet. The addition of erotic acid clearly led to sig- nificant hepatic steatosis. Ultrastructural studies showed the presence of prominent Golgi apparatus with numerous saccules and vacuoles containing VLDL, increased smooth endoplasmic reticulum, and increased lyso- somes forming the GERL complex. These ul- trastructural findings suggested that there was accumulation of both triglyceride and VLDL without an abnormality in the secretory ap- paratus. In our study with increased dietary sucrose we did not find evidence of prominent Golgi apparatus or the GERL complex. Smooth endoplasmic reticulum however, was increased, suggesting an increase in synthesis of triglyceride without a concomitant increase

826 BACON ET AL.

in secretion. Finally, genetically obese Zucker DIXON, W. J. (1977). BMDP: Biomedical Computer Pro- rats (homozygous recessive, ff) developed grams. Univ. of California Press, Los Angeles. marked hepatic steatosis after receiving a diet FOLCH, J., LEES, M., AND SLOANE, S. G. H. (1954). A

containing 60% sucrose for 1 week, whereas simple method for the isolation and purification of total

lean litter-mates (FF and/or Ff) did not develop lipids from animal tissues. J. Biof. Chem. 226, 497- 509.

the heDatic steatosis (Novikoff, 1977). HAMM, T. E., RAYNOR, T., AND CAVISTON, T. (1982). In conclusion, the major significance of our

findings is that the AIN-76A semipurified diet is not suitable for toxicologic studies which may depend on normal hepatocellular integ- rity. It appears from our work and from that of Hamm et al. (1982) that the high sucrose concentration in the diet is responsible for the development of the hepatic steatosis. Satis- factory growth and liver/body weight ratios apparently are insufficient parameters for use in judging the adequacy of new purified or semipurified diets. At the very least, histologic evaluation of body tissues should be included in order to identify obvious histopathologic alterations that may be related to dietary com- position.

ACKNOWLEDGMENTS

The authors thank Dr. Anthony S. Tavill for critically reviewing the manuscript, Joan Skiba for expert secretarial assistance, and Pat Griffin and Linda Hrabak for skilled technical assistance. This research was supported in part by the Cuyahoga County Hospital Foundation and by U.S. Public Health Service Grants ROl AM 31505 and T32 HL 07147. Dr. Bacon was a Postdoctoral Research Fellow of the American Liver Foundation and is a Teach- ing and Research Scholar of the American College of Physicians.

REFERENCES

AF~FI, A. A., AND AZEN, S. P. (1979). Statistical Analysis. A Computer Oriented Approach. Academic Press, New York.

BIERI, J. G. (1980). Second Report of the Ad Hoc Com- mittee on Standards for Nutritional Studies. J. Nutr. 110, 1726.

BIERI, J. G., STOEWSAND, G. S., BRIGGS, G. M., PHILLIPS, R. W., WOODARD, J. C., AND KNAPKA, J. J. (1977). Report of the American Institution of Nutrition Ad Hoc Committee on Standards for Nutritional Studies. J. Nutr. 107, 1340-1348.

BLIGH, E. G., AND DYER, W. J. (1959). A rapid method of total lipid extraction and purification. Canad. J. Biochem. Physiol. 37, 9 1 l-9 17.

Unsuitability of the AIN-76A diet for male F-344 and CD rats and improvement by substituting starch for sucrose. Lab. Anim. Sci. 32, 414-415 (abstract).

HENRY, R. T., CHIAMORI, N., GOLUB, 0. J., AND BERK- MAN, S. (1960). Revised spectrophotometric method for the determination of ghrtamic oxaloacetic ttans- aminase, glutamic pyruvate transaminase and lactic acid dehydrogenasc. Amer. J. Clin. Pathol. 34, 381-398.

HOYUMPA, A. M., GREENE, H. L., DUNN, G. D., AND SCHENKER, S. (1975). Fatty liver: Biochemical and clinical considerations. Dig. Dis. 20, 1142-l 170.

KARMEN, A., WROBLEWSKI, F., AND LADUE, J. E. (1955). Transaminasc activity in human blood. J. Clin. Invest. 34, 125-133.

LUNA, L. G. ( 1968). Manual of Histologic Staining Meth- ods of the Armed Forces Institute ofPathology, 3rd ed., pp. 140-141. McGraw-Hill, New York.

MEDINSKY, M. A., POPP, J. A., HAMM, T. E., AND DENT, J. G. (1982). Development of hepatic lesions in male Fischer-344 rats fed AIN-76A purified diet. Toxicol. Appl. Pharmacol. 62, 11 I- 120.

NEWBERNE, P. M., BIERI, J. G., BRIGGS, G. M., AND NESHEIM, M. C. (1978). Control of Diets in Laboratory Animal Experimentation. A Report of the Committee on Laboratory Animal Resources, Assembly of Life Sciences, National Research Council. National Academy of Sciences, Washington, D.C.

NGUYEN, H. T., AND WOODARD, J. C. (1980). Intrane- phrenic calculosis in rats. An ultrastructural study. Amer. J. Pathol. 100, 39-56.

NOVIKOFF, P. M. (1977). Fatty liver induced in Zucker fatty (8) rats by a semisynthetic diet rich in sucrose. Proc. Nat/. Acad. Sci. USA 74, 3038-3042.

NOVIKOFF, P. M., ROHEIM, P. S., NOVIKOFF, A. B., AND EDELSTEIN, B. S. (1974). Production and prevention of fatty liver in rats fed clofibrate and erotic acid diets containing sucrose. Lab. Invest. 30, 732-750.

PETERSON, P. (1974). Lipid droplets in fatty liver. Acta Puthoi. Microbial. Stand. Sect. A 82, 255-263.

PINTER, J. K., HAYASHI, J. A., AND WA-~X~N, J. A. (1967). Enzymatic assay of glycerol, dihydroxyacetone and glyceraldehyde. Arch. Biochem. Biophys. 12h404-414.

ROEBUCK, B. D., WILPONE, S. A., FIFIELD, D. S., AND YAGER, J. D. (1979). Hemorrhagic deaths with AIN- 76 diet. J. Nutr. 109, 924-925.

SCHOTZ, M. C., AND RECKNAGEL, R. 0. (1960). Rapid increase of rat liver triglycerides following carbon tet- rachloride poisoning. B&him. Biophys. Ada 41, 15 l- 152.