Effect of mild diabetes and dietary fructose on very-low-density lipoprotein triglyceride turnover in rats

Effect of Mild Diabetes

and Dietary Fructose on Very-Low-Density Triglyceride Turnover in Rats

Lipoprotein

Gen Yoshino, Masayuki Matsushita, Masahide Iwai, Munetaka Morita, Kohji Matsuba, Kohichi Nagata, Eiichi Maeda, Seiichi Furukawa, Tsutomu Hirano, and Tsutomu Kazumi Very-low-density lipoprotein (VLDL) triglyceride turnover was examined in mildly streptozotocin (25 mg/kg)-diabetic rats, using Triton WR1339. Diabetic rats fed standard rat chow showed mild hyperglycemia and suppressed levels of plasma insulin. Their triglyceride secretion was significantly suppressed despite an elevated level of plasma free fatty acids. However, the plasma triglyceride level of these diabetic rats was significantly elevated compared with nondiabetic controls. This suggested that the removal of triglyceride from the circulation, as well as its entry into the circulation, was impaired in mildly insulin-deficient rats. Glucose or fructose supplementation (10% in drinking water for 14 days) significantly increased the triglyceride secretion rate of diabetic rats. Especially, fructose supplementation increased plasma insulin to normal levels, but resulted in markedly elevated plasma triglyceride levels (three times higher than glucose-supplemented or chow-fed diabetic rats) despite similar triglyceride secretion rates between the two types of sugar-supplemented diabetic rat groups. This suggested an impairment of triglyceride removal by dietary fructose. The result obtained from chow-fed diabetic rats indicates that mild but significant insulin deficiency resulted in mild hypertriglyceridemia, linked to impaired triglyceride removal rather than to an overproduction of VLDL-triglyceride, despite elevated levels of plasma free fatty acids. Furthermore, fructose feeding induced prominent hypertriglyceridemia not only by stimulating triglyceride secretion, but also by suppressing triglyceride removal from the circulation of mildly diabetic rats. Copyright 8 7992 by W.B. Saunders Company

T

HE CAUSE of human diabetic hypertriglyceridemia is unclear, but there is substantial evidence that a defect in lipoprotein clearance is an important contributory factor.‘.* Increased production of plasma very-low-density lipoprotein (VLDL) may also occur.3.4 In experimental diabetes, the mechanisms underlying the hyperlipidemia of insulin deficiency also remain unclear. Reaven and Reaven’ found that the entry rate of triglyceride into the plasma from the liver or intestine was decreased in streptozotocindiabetic rats, thereby eliminating increased production rate as a cause of the hypertriglyceridemia. Van To16 also reported that the fractional clearance rate of radioactive triglyceride was severely reduced in streptozotocin-diabetic rats. Our previous observation revealed that rats with streptozotocin-induced diabetes of 2 week’s duration were normotriglyceridemic despite decreased triglyceride secretion rate, suggesting reduced triglyceride removal.’ A recent observation by Okabayashi et al8 demonstrated that rats treated with a small dose of streptozotocin demonstrated mild hyperglycemia and suppressed plasma insulin levels in response to an oral glucose load. It is of interest to know how triglyceride production and secretion are modified by mild hyperglycemia and mild insulin deficiency. However, very few reports are currently available concerning triglyceride turnover in a mildly insulindeficient animal model. Thus, the present study was con-

From the Second Depatiment of Internal Medicine, Kobe University School of Medicine, Chuo-ku, Kobe, Japan; the First Department of Internal Medicine, Shown University School of Medicine, Shinagawaku, Tokyo, Japan; and the Division of Endocrinology and Metabolhm, Department of Medicine, Hyogo Medical Center for Adults, Kitaohjicho, Akashi, Japan. Address reprint requests to Gen Yoshino, MD, Second Department of Internal Medicine, Kobe University School of Medicine, Chuo-ku, Kobe 650 Japan. Copyright 0 I992 by W.B. Saunders Company 0026-0495/92/4103-0002$03.00/0 236

ducted to examine triglyceride turnover in mildly streptozotocin-diabetic rats. It is known that fructose feeding induces insulin resistance9,” and hypertriglyceridemia.1’-‘3 Previous observations have shown that an elevation of plasma triglyceride in fructose-fed, nondiabetic rats was due to an impaired rate of removal of VLDL triglyceride and to an overproduction of hepatic VLDL triglyceride.‘3.‘4 Therefore, we also examined whether fructose or glucose feeding can modify triglyceride turnover in mildly streptozotocin-diabetic rats. MATERIALS Animals

AND METHODS

and Diet

Eight-week-old, male Wistar rats weighing 260 g were used in this study. All animals were housed in a ventilated (15 times per hour), temperature- and humidity-controlled room (22 f 2°C. SO% to 60% humidity) with a 12-hour light/dark (6:00~~/6:00 PM) cycle. Animals had ad libitum access to rat chow (Oriental MF, Oriental Yeast, Tokyo, Japan) and drinking water, except during streptozotocin injection and sampling for the measurement of triglyceride secretion rates. Three groups of rats were made diabetic by intravenous (IV) administration of 25 mg/kg streptozotocin (Wako Pure Chemicals, Osaka, Japan). Streptozotocin was dissolved in citrate buffer, pH 4.5, and then immediately injected into the tail vein of rats that had been fasted overnight and placed under light ether anesthesia. The rats with a plasma glucose level higher than 200 mg/dL 2 weeks after streptozotocin injection (n = 56) were used for the measurement of triglyceride secretion rate. Two of the diabetic groups received sugar supplied as a 10% solution in drinking water. In one group (n = 18), the sugar was fructose, and in the other (n = 18), glucose. The third diabetic group (n = 20) received drinking water without sugar and served as a diabetic control. The fourth group (n = 20) was injected with citrate buffer alone, received normal drinking water and served as a nondiabetic control. The amount of sugar and chow consumed by each rat was measured daily. Secretion

of VLDL

Fourteen days after streptozotocin injection, the triglyceride secretion rate was measured in each rat. Chow was removed 16 Metabolism,

Vol41, No 3 (March), 1992: pp 236-240

237

MILD DIABETES AND TRIGLYCERIDE TURNOVER

hours before the measurement, but free access to drinking water was allowed until the end of the experiments. The triglyceride

Initial and Final Body Weight

secretion rate was determined by the Triton method as described by Steiner et alI5and by our previous work.16The suitability of this procedure has been previously discussed and fully described.7”,” The Triton WR1339 (Nakarai Chemicals, Kyoto, Japan) used in this study has been demonstrated to completely block removal of triglyceride from the plasma and to produce a linear increase in triglyceride for at least 90 minutes. Triton was dissolved in distilled water (300 mg/mL) and injected into the tail veins (600 mglkg) of rats under light ether anesthesia. Blood was withdrawn from the tails of conscious rats before the injection. Ninety minutes later, blood was withdrawn from the abdominal aorta with EDTA (1.5 mg/mL blood) under pentobarbital anesthesia. Triglyceride secretion rate was calculated from the increment of plasma triglyceride level per minute multiplied by the plasma volume of the rat, and the results were expressed as mg/min.”

Chow

and Sugar Intake

VLDL Composition Since Triton potently suppresses VLDL catabolism by inhibiting peripheral lipase activity,15 more than 96% of VLDL particles accumulated 90 minutes after Triton injection represent newly secreted VLDL particles. I’ In addition, the VLDL present after Triton injection more closely resembled nascent VLDL than circulating VLDL.” Thus, our Triton method may provide information on the VLDL synthetic response” in a more physiological condition than can be achieved with the liver perfusion system. In order to obtain newly secreted VLDL particles, the postTriton samples were extensively dialyzed against phosphatebuffered saline at 4°C overnight. The VLDL fraction (d < 1.006) was separated from the samples by ultracentrifugation at 39,000 rpm for 16 hours in sealed tubes, using a Beckmann LS-55 ultracentrifuge (Beckman Instruments, Fullerton, CA). The top fraction, corresponding to VLDL, was washed once, sliced off, and used for lipid assay and apolipoprotein (apo) B subfractionation.

50 (Kcaliday)

n

Cl83 fructosesupplemented diabetic

(183

(201

1201

glucosesupplemented diabetic

chowfed diabetic

nondiabetic control

Fig 1. Body weight end chow end suger intake of the four experimental groups. Vertical bars indicate meens 2 SE with the number of rats given in brackets. (F) and (G) indicate fructose and glucose suger consumption, respectively.

Chemical Analysis Apo B subfractions in the VLDL were separated by 0.1% sodium dodecyl sulfate (SDS)-3.3% polyacrylamide gel electrophoresis (PAGE), as described previously.‘O Cholesterol, phospholipid, and triglyceride in plasma, and the lipoprotein fraction were measured enzymatically, using commercially available kits (Determiner-736TC-S, Kyowa Medix, Tokyo, Japan, for cholesterol; Determiner-736PL-S, Kyowa Medix, for phospholipid and Triglyceride-HRII; Wako Pure Chemicals, Osaka, Japan, for triglyceride). Correction for the presence of free glycerol can be made simultaneously by this triglyceride assay system. Plasma free fatty acid level was also estimated enzymatically, using commercially available kits (NEFA-SS ‘Eiken,’ Eiken Kagaku, Tokyo, Japan). Plasma glucose and insulin were measured by the glucose oxidase method and by a double-antibody radioimmunoassay using rat standard insulin.” The data are expressed as means 2 SE. ANOVA was used to determine whether there were significant differences between groups. The two-tailed P values were then calculated by Bonferroni’s multiple comparison procedure.= RESULTS

There was no significant difference in the weight gained by any of the experimental groups during the 2 weeks of sugar supplementation (Fig 1). There was also no significant difference in sugar consumption between the two sugar-supplied groups, and the total daily energy intake was similar in each group. All three diabetic groups showed

significant hyperglycemia, and the fructose-supplemented diabetic group had the highest value (Fig 2). The plasma insulin levels of the diabetic controls and the glucosesupplemented diabetic group were significantly suppressed, while those of the fructose-supplemented diabetic group were not. Plasma free fatty acid levels were elevated in the three diabetic groups, and the degree of difference from the nondiabetic control value was significant in fructosesupplemented diabetic and chow-fed diabetic controls (Fig 3). Phospholipid and cholesterol concentrations in the plasma of all three diabetic groups tended to increase, but were not significantly different from values in the nondiabetic controls. Fructose-supplemented diabetic rats showed prominent hypertriglyceridemia. Diabetic rats without sugar supplementation also showed mild but significant hypertriglyceridemia (Fig 4). Triglyceride secretion rates of the two sugar-supplemented diabetic groups were increased compared with the nondiabetic control value. In contrast, diabetic rats without sugar supplementation had significantly suppressed secretion rates compared with all other groups. In the post-Triton samples, cholesterol content (percent of total lipid mass) of the VLDL fraction was significantly decreased in fructose-supplemented diabetic rats (Table 1).

YOSHINO

238

Plasma

Glucose

Plasma

Insulin

ET AL

significantly suppressed triglyceride secretion rates compared with nondiabetic control rats. As this diabetic group was mildly hypertriglyceridemic, it was suggested that there might be impaired triglyceride removal in mildly insulindeficient rats. Thus, overproduction of VLDL triglyceride was excluded as a cause of the hyperlipidemia seen in this mildly diabetic rat animal model, On the other hand, the plasma triglyceride concentration in fructose-fed, mildly diabetic rats was three times higher than that in glucosesupplemented diabetic rats. However, triglyceride secretion rates were similar in the two sugar-supplemented groups. This disparity between the changes in triglyceride concentration and triglyceride secretion rate suggests that dietary fructose interferes with triglyceride removal from the circulation in mildly diabetic rats. A recent reportzJ indicates that lipoprotein lipase and hepatic lipase are not involved in the reduced VLDLtriglyceride removal observed in fructose-fed, nondiabetic rats. Therefore, there must be other mechanisms responsible for the impaired triglyceride removal induced by fructose. Although mild diabetes had almost no effect on lipid composition of newly secreted VLDL (Table l), cholesterol and phospholipid contents (percent of total lipid mass) of this fraction from fructose-fed, mildly diabetic rats were

600 (w/de) 400 -

T

Plasma

Free

Fatty

Acid

.% glUCOCE!chowsupplemented supplemented fed diabetic diabetic diabetic

fructme-

nondiabetic control

2.0

(mEq/J) 1.0

Fig 2. Plasma glucose and insulin levels in the four experimental groups. Vertical bars indicate means + SE. Significant difference; lP < .05, l*P < .005 v nondiabetic control and fructose-supplemented diabetic groups. §P < .05 v all other groups.

Plasma

Phospholipid

Plasma

Cholesterol

Phospholipid content was also suppressed in this group. There was no significant difference in the B48 to B95+ BlOO ratio of the newly secreted VLDL fraction among the four experimental groups.

DISCUSSION

One purpose of this study was to examine the effect of mild insulin deficiency on triglyceride turnover in rats. The plasma triglyceride level reflects the steady-state of triglyceride turnover. Since triglyceride removal from the circulation was blocked by Triton WR1339,” it was not possible to determine both triglyceride secretion into, and triglyceride removal from, the circulation in the same rat. However, as we measured the basal triglyceride level in each rat, inferences can be made about the rate constant of triglyceride removal. In support of this, the VLDL-triglycerideremoval capacity estimated by the Triton method and by tracer methods using radiolabeled glycerol are in paralle1.20~2’ In this study, diabetic rats fed chow only showed

wppknented

supplaented

diatetlc

dub&z

fed dmbettc

dlabetlc control

Fig 3. Free fatty acid, phospholipid, and cholesterol in the plasma of four experimental groups. Vertfcal bars indicate means k SE. *A signfficant difference from nondiabetfc control group, lP < .05.

MILD DIABETES AND TRIGLYCERIDE TURNOVER

Plasma

Triglyceride

239

Triglyceride

Secretion

glucose chowsupplementedsupplementedfed diabetic diibk diabetic

Rate

and the significantly elevated plasma glucose levels of fructose-fed, mildly diabetic rats, compared with both glucose- and chow-fed mildly diabetic rats, may indicate fructose-induced insulin resistance. Further study will be necessary to elucidate the precise mechanisms for this impaired VLDL catabolism by fructose. In this study, diabetic rats fed chow only showed significant hyperglycemia and lower plasma insulin levels compared with nondiabetic control rats. The triglyceride secretion rate of this group was suppressed slightly but significantly in the presence of an elevated plasma free fatty acid level. In our previous work,” exogenous hyperinsulinemia stimulated VLDL-triglyceride production in the presence of a suppressed plasma free fatty acid level. Although substrate availability is an important condition for triglyceride production and secretion, this new finding supports our earlier proposal that circulating insulin plays a key role in triglyceride production and secretion in rats.” In support of this, we also found that Zucker fatty rats with endogenous hyperinsulinemia had an increased triglyceride secretion rate.25 In contrast, the two sugar-supplemented diabetic groups had an increased triglyceride secretion rate. The supplied sugars might be utilized as substrates for de novo synthesis of triglycerides in the liver under mildly diabetic conditions. However, in our previous observation, glucose and fructose supplementation failed to stimulate triglyceride secretion from severely insulin-deficient rats.’ This may also support the role of circulating insulin in triglyceride production and secretion. Three forms of apo B have been found in the rat VLDL.26 Apo B48, B95, and BlOO are metabolized independently of each other, and B48 is the major apoprotein in normal rats and facilitates speedier clearance than the other two. Although the relative proportion of these apo B subspecies can be changed by different feeding conditions,” neither mild streptozotocin-diabetes nor sugar supplementation influenced this proportion in newly secreted VLDL in rats. Thus, we conclude that mild insulin deficiency resulted in hypertriglyceridemia in rats fed standard chow, and it was not caused by increased production of VLDL triglyceride, in spite of an elevated plasma free fatty acid level. Instead, a defect in the removal of VLDL triglyceride appears as the most likely cause for this hypertriglyceridemia. Glucose supplementation stimulated triglyceride secretion in mildly diabetic rats without stimulating insulin secretion. Excess supply of glucose might modify the relationship between plasma insulin level and triglyceride secretion in this rat

1

nondiabetic control

Fig 4. Plasma triglyceride and triglyceride secretion rate in the four experimental groups. Vertical bars indicate means )_ SE. Significant difference; lP < .005 Y nondiabetic control group; §P < 95 v all other groups.

significantly lower than corresponding values from all other experimental groups. At present, we do not know whether this compositional abnormality has a direct influence on the clearance of VLDL from the circulation. It is also a unique finding that dietary fructose increased plasma insulin levels of mildly diabetic rats but interfered with the rate of triglyceride removal from their circulations. Another possibility is that fructose-induced insulin resistance’,“’ caused additional impairment of triglyceride removal from the circulation in mildly diabetic rats. The increased (compared with the nondiabetic control value) plasma insulin levels

Table 1. Effect of Fructose or Glucose on Lipid Composition (Percent of Total Lipid Mass) and Apolipoprotein

848 to 885 + BlOO Ratio

of VLDL Fraction 90 Minutes After Injection of Triton WR133g in Mildly Diabetic Rats

Fructose-supplemented

Triglyceride(%)

Phospholipid(%)

848 to 695 + BlOORatio

5.16 -’ 0.35*

83.32 ‘- 0.39

11.52 f 0.23*

3.0 k 0.9

5.95 z 0.22

82.62 2 1.11

13.13 f 0.38

3.3 + 1.0

Diabetic control

7.67 2 0.36

79.04 z!z1 .oo

14.26 2 0.44

2.7 r 1.2

Nondiabetic control

6.78 2 0.27

80.34 r 0.85

13.61 -t 0.28

3.6 -t 0.8

Glucose-supplemented

diabetic

Cholesterol(%)

diabetic

NOTE. Data are expressed as means f SE. *Significant difference from other three groups, P < .05 or less.

240

YOSHINO ET AL

group. On the other hand, fructose supplementation produced prominent hypertriglyceridemia in mildly diabetic rats by interfering with triglyceride removal, as well as by stimulating triglyceride secretion. Although lipoprotein metabolism is different in rats and humans, our findings draw attention to the possibility of

reevaluating the advisability of fructose consumption hyperlipidemic patients with mild diabetes.

in

ACKNOWLEDGMENT

We arc indebted to Dr Frederick J. Haynes of Toronto, Ontario, Canada for his valuable advice in preparing the manuscript.

REFERENCES

1. Biermann EL, Bagdade J, Porte D Jr: Concept of pathogenesis of diabetic lipemia. Trans Assoc Am Physicians 74:348-360, 1966 2. Greenfield M, Kolterman 0, Olefsky J, et al: Mechanism of hypertriglyceridemia in diabetic patients with fasting hyperglycemia. Diabetologia 18:441-446,198O 3. Abrams JJ, Ginsberg H, Grundy SM: Metabolism of cholesterol and plasma triglyceride in nonketotic diabetes mellitus. Diabetes 31:903-910,1982 4. Kissebah AH, Alfarsi S, Evans DJ, et al: Integrated regulation of very low density lipoprotein triglyceride and apolipoprotein-B kinetics in non-insulin-dependent diabetes mellitus. Diabetes 31: 217-225, 1982 5. Reaven EP, Reaven GM: Mechanism of development of diabetic hypertriglyceridemia in streptozotocin-treated rats: Effect of diet and duration of insulin deficiency. J Clin Invest 54:11671178,1974 6. Van Tol A: Hypertriglyceridemia in the diabetic rat. Defective removal of serum very-low-density lipoproteins. Atherosclerosis 261117.128,1977 7. Yoshino G, Iwai M, Kazumi T, et al: Effect of dietary fructose on triglyceride turnover in streptozotocin-diabetic rats. Atherosclerosis 79:41-46.1989 8. Okabayashi Y, Ohki A, Sakamoto C, et al: Relationship between the severity of diabetes mellitus and pancreatic exocrine dysfunction in rats. Diabetes Res Clin Pratt 1:21-30, 1985 9. Hill R, Baker N, Chaikoff IL: Altered metabolic patterns induced in the normal rat by feeding an adequate diet containing fructose as sole carbohydrate. J Biol Chem 209:705-716,1954 10. Tobay TA, Mondon CE, Zavaroni I, et al: Mechanism of insulin resistance in fructose-fed rats. Metabolism 31:608-612,1982 11. Nikkila EA, Ojala K: Induction of hypertriglyceridemia by fructose in the rat. Life Sci 4:937-943, 1965 12. Bar-On H, Stein Y: Effect of glucose and fructose administration on lipid metabolism in the rat. J Nutr 94:95-105,1968 13. Kazumi T, Vranic M, Steiner G: Triglyceride kinetics: Effects of dietary glucose, sucrose or fructose alone or with hyperinsulinemia. Am J Physiol250:E325-E330. 1986 14. Hirano T, Mamo JCL, Poapst M, et al: Very-low-density lipoprotein triglyceride kinetics in acute and chronic carbohydratefed rats. Am J Physiol255:E236-E240, 1988

15. Steiner G, Heynes FJ, Yoshino G, et al: Hyperinsulinemia and in vivo very-low-density lipoprotein-triglyceride kinetics. Am J Physiol246:E187-E192, 1984 16. Yoshino G, Kazumi T, Iwai M, et al: Effects of CS-514 (pravastatin) on VLDL-triglyceride kinetics in rats. Atherosclerosis 73:191-196.1988 17. Kazumi T. Vranic M, Bar-On H, et al: Portal v peripheral hyperinsulinemia and very-low-density lipoprotein-triglyceride kinetics. Metabolism 35:1024-1028,1986 18. Ishikawa T, Fidge N: Changes in the concentration of plasma lipoproteins and apoproteins following the administration of Triton WR1339 to rats. J Lipid Res 20:254-264, 1979 19. Guettet C, Mathe D, Navarro N, et al: Effects of chronic glucagon administration on rat lipoprotein composition. Biochim Biophys Acta 1005:233-238,1989 20. Hirano T, Mamo JCL, Poapst M, et al: Impaired very low-density lipoprotein-triglyceride catabolism in acute and chronic fructose-fed rats. Am J Physiol256:E559-E565. 1989 21. Morgan CR, Lazarow A: Immunoassay of insulin: Two antibody system. Plasma insulin levels of normal, subdiabetic and diabetic rats. Diabetes 12:115-118.1963 22. Glanz SA: Primer of Biostatistics. New York. NY, McGrawHill, 1981 23. Furukawa S, Hirano T, Mamo JCL, et al: Catabolic defect of triglyceride is associated with abnormal very-low-density lipoprotein in experimental nephrosis. Metabolism 39:101-107, 1990 24. Mamo JCL, Hirano T. Szeto L, et al: VLDL-Tg metabolism in carbohydrate fed rats. Proceedings of the 8th International Symposium on Atherosclerosis, Rome, Italy, October 1988, p 566 25. Kazumi T, Yoshino G, Matsuba K, et al: Effect of dietary glucose or fructose on the secretion rate and particle size of triglyceride-rich lipoproteins in Zucker fatty rats. Metabolism 40:962-966, 1991 26. Krishnaiah KV, Walker LF. Borensztajn J, et al: Apolipoprotein B variant derived from rat intestine. Proc Nat1 Acad Sci USA 77:3806-3810,198O 27. Wu A-L, Windmueller HG: Variant forms of plasma apolipoprotein B: Hepatic and intestinal biosynthesis and heterogenous metabolism in the rat. J Biol Chem 256:3615-3618,198l