The effect of insulin on lysosomal acid cholesterol ester hydrolase activity was studied in liver, heart and fat pad preparations from rats and mice. Hyperinsulinemia was induced for a period of 6 days in rats by the subcutaneous administration of exogenous insulin by an osmotic minipump. The effect of more chronic endogenous hyperinsulinemia was studied using genetic strains of diabetic (db/db) mice at 12 weeks of age. Mouse liver and heart preparations were characterized as having an acid pH optimum of 4 .5 -5 for cholesterol ester hydrolase activity; a smaller but distinct pH optimum could also be observed at pH 7. In contrast, hydrolase activity in mouse fat pad preparations had only one distinct pH optimum of 6.5. Hyperinsulinemia in rats and mice resulted in a significant decrease in acid cholesterol ester hydrolase activity in heart preparations, but had no consistent effect on acid hydrolase activity observed in liver and fat pad preparations. This decrease in lysosomal acid cholesterol ester hydrolase activity in cardiac tissue due to hyperinsulinemia cannot be related to any changes in lipoprotein turnover caused by insulin or diabetes.
Introduction
An acid cholesterol ester hydrolase has been implicated in the lysosomal degradation of the cholesterol ester-component of internalized lipo- proteins in liver [1-3] and in peripheral (non- hepatic) tissues [4,5]. Recently, Wolinsky [6] has proposed that the dynamic clearance of lipopro- teins is linked to cellular metabolic activity such that levels of acid cholesterol ester hydrolase might determine the rate of removal of LDL from the circulation. For example, the induction of diabetes by the administration of streptozotocin to rats resulted in an increase in the fractional catabolic rate of 125I-labelled LDL and an increase in acid cholesterol ester hydrolase activity in liver and
Abbreviations: LDL, low density lipoprotein; Hepes, 4-(2-hy- droxyethyl)- 1-piperazineethanesulphonic acid.
kidney [7]. In contrast, diabetes resulted in a de- crease in acid hydrolase activity in rat aorta [7,8]. Therefore, changes in rates of clearance of LDL in response to insulin deficiency are reflected in changes in hydrolase activity at least in non-vascu- lar tissues. Insulin has also been shown to increase the binding and degradation of LDL in cultured fibroblasts [9, 10], and insulin therapy has been demonstrated to increase acid cholesterol ester hy- drolase activity in mononuclear cells from humans [11]; these findings are consistent with the ob- servation that plasma cholesterol levels are in- creased in some insulin deficient human diabetics [12,13]. Wolinsky et al. [7] have shown that the administration of insulin to diabetic rats reduced the fractional catabolic rate for LDL and reduced acid cholesterol ester hydrolase activities liver and kidney to control values. Since hyperinsulinemia has been implicated as a risk factor in the develop- ment of atherosclerosis [14], the objective of the
present investigation was to determine if hyperin- sulinemia would influence acid cholesterol ester hydrolase activity in liver, heart and epididymal fat pad preparations from rats and mice.
Experimental Procedures
Materials Materials were as listed in Ref. 15 with the
following additions. Regular pig insulin (U-500, Iletin R) was a gift from Eli Lilly (Canada), Scar- borough, Ontario; rat insulin was obtained from NOVO Research Institute, Copenhagen. Alzet R osmotic minipumps (Model 1701) were purchased from Alza Corp, Palo Alto, CA.
Animals Male Sprague-Dawley rats (150-175g) were
obtained from either breeding colonies at the Uni- versity of Calgary or from Canadian Breeding Farm and Laboratories, Ltd., Laprairie, Quebec. In order to study the effects of chronic hyperin- sulinemia, an osmotic minipump containing regu- lar pig insulin (diluted to 200 uni ts /ml with sterile saline) was implanted subcutaneously into 25 rats. Assuming a nominal flow rate of 1 /~l/h, each rat received approximately 5 units of insulin per day. Control rats underwent a sham operation but a minipump was not implanted. Rats were main- tained on a 12-h alternating light cycle (light from 0700 to 1900 h) and were allowed free access to standard laboratory chow and water containing 10% (w/v) glucose. After 6 days, the rats were killed by decapitation and a sample of blood was collected in heparinized tubes for subsequent de- terminations of plasma levels of glucose and im- munoreactive insulin. One of the rats implanted with the insulin-containing osmotic minipump was excluded from the study because of normal plasma glucose and insulin values. Livers, hearts and epi- didymal fat pads were individiually removed, frozen rapidly, and stored at - 8 0 ° C until use.
Studies on the characterization of cholesterol ester hydrolase from mouse tissues were performed with mice obtained from breeding colonies at the University of Calgary. Hyperinsulinemia in mice was studied with genetic strains of diabetic (db/db) mice. 25 diabetic mutant mice (C57BL/ Ksj-db +/db + ) at approx. 4 - 5 weeks of age
145
were obtained from The Jackson Laboratory, Bar Harbour, Maine. The same number of control litter mates that were either heterozygotes for the d iabe tes and mis ty coa t co lour alleles (C57BL/Ksj-db + / + m ) or homozygotes for the misty coat colour allele (C57BL/Ksj + re~m) were also obtained. The mice were maintained on a 12-h alternating light source and were allowed free access to water and standard laboratory chow. Mice were killed at 12 weeks of age, and plasma and tissues were collected as described above.
Tissue preparations and enzyme assays Liver and hearts were homogenized in 10 vol. isotonic sucrose buffer (0.25 M sucrose/1 mM E D T A / 1 0 mM Hepes, pH 7.5) at 4°C for 30-40 s with a Polytron PT-10 homogenizer (rheostat set- ting of 7); epididymal fat pads were homogenized in 3 vol. buffer. An aliquot of the liver and heart homogenates were removed, frozen and stored at - 8 0 ° C for DNA determinations. Cellular debris was removed by centrifugation for 20 min at 5 000 g and enzyme activities were routinely determined in this low-speed supernatant fraction after filtering through glass wool.
The preparation of the glycerol-dispersed cholesterol oleate substrate and determination of hydrolase activity was accomplished by the proce- dures as described by Severson and Fletcher [15]. A unit of cholesterol ester hydrolase activity was arbitrarily defined as that amount of enzyme which hydrolyzed 1 nmol cholesterol oleate in 1 h at 30°C. All assays were performed in duplicate at two incubation times (15 and 30 min). Hydrolase activity in low-speed supernatant fractions is ex- pressed as units per mg protein and as units per mg homogenate DNA, except in the case of fat pad preparations where activity is expressed only per mg protein.
The lysosomal marker enzyme N-acetylglu- cosaminidase was assayed by either a fluorometric assay [16] with preparations from rat tissues, or by a spectrophotometric procedure [17] with prepara- tions from mouse tissues, fl-Glucuronidase was assayed by the fluorometric method as described by Peters et al. [16]. A unit of enzyme activity for N-acetylglucosaminidase and fl-glucuronidase was arbitrarily defined as that amount of enzyme which hydrolyzed 1/~mol substrate per h at 30°C. Under
146
the experimental conditions of this investigation which used frozen tissues that were vigorously homogenized and involved storage of subcellular fractions at -80°C, no latency could be detected for the lysosomal marker enzymes and so assays were routinely determined in the absence of Triton X-100.
Other assays Protein alad DNA was measured by the fluoro-
metric techniques of Robrish et al. [18] and Thomas and Farquhar [19], respectively. Plasma glucose was determined by the glucose oxidase method using a Beckman glucose analyzer. Plasma levels of immunoreactive insulin were measured by ra- dioimmunoassay (using rat insulin as standard) in the laboratory of Dr. S. Ross, Department of Medicine, The University of Calgary.
Results
Effect of hyperinsulinemia on acid cholesterol ester hydrolase activity in rat liver, heart and fat pad preparations
After 6 days of insulin treatment, rats were characterized as being hypoglycemic (84 -+ 8mg/dl
compared to 141 mg/dl for controls; mean ---S.E.) and hyperinsulinemic (592- + 109 /~U/ml com- pared to 4 4 - 3 for controls). Similar effects of insulin delivered by an osmotic minipump have been reported by McCormick et al. [20]. For com- parisons of acid cholesterol ester hydrolase activity between control and insulin-treated rats, the livers, hearts and fat pads from eight animals were pooled for each preparation; each of these groups had identical mean plasma glucose and insulin levels.
Hyperinsulinemia resulted in a modest increase (approx. 20%) in the average wet weight of livers and hearts compared to tissue weights from con- trol rats; in contrast, the wet weight of epididymal fad pads was increased by 80% following insulin treatment. The effect of insulin on acid cholesterol ester hydrolase activity (measured at pH 5) in rat liver, heart and fat pad preparations is shown in Table I. Hyperinsulinemia resulted in a decrease in the specific activity of acid cholesterol hydrolase in heart preparations. Hyperinsulinemia had no ef- fect on the activity of the lysosomal marker en- zymes (N-acetylglucosaminidase and fl- glucuronidase) in liver and heart preparations, but specific enzyme activities were significantly re- duced in fat pad preparations (Table I).
TABLE I
EFFECT OF HYPERINSULINEMIA ON ACID CHOLESTEROL ESTER HYDROLASE, N-ACETYLGLUCOSAMINIDASE (NAGA) AND fl-GLUCURONIDASE ACTIVITY IN PREPARATIONS FROM RAT LIVER, HEART AND FAT PADS
Results are the mean -4- S.E. for three preparations. P values were calculated by the Student's t-test. Cholesterol ester hydrolase assays
were performed at pH 5.
Acid cholesterol ester hydrolase activity uni t s /mg protein uni t s /mg DNA
NA, GA Activity (uni ts /mg protein)
fl-Glucuronidase Activity (uni ts /mg protein)
A. Liver Control 5.99 -+ 0.79 Insulin-treated 5.36 --+ 0.63
B. Heart Control 6.60 ± 0.36 Insulin-treated 4.28-+ 0.09 b
C. Fat Pads Control 97.0 -+ 1.9 Insulin-treated 80.9 -+ 6.1
Comparison of acid cholesterol ester hydrolase activ- ity in liver, heart and fat pad preparations from control and diabetic mice
The observation (Table I) that hyperinsulinemia in rats did not alter hepatic acid cholesterol ester hydrolase activity is in contrast to the results ob- tained by Wolinsky et al. [7] in which administra- tion of insulin to diabetic rats for 8 weeks resulted in a reduction of acid cholesterol ester hydrolase in liver. Since it might be argued that 6 days of hyperinsulinemia may not be of sufficient duration to observe an effect on acid hydrolase activity, a more chronic animal model of hyperinsulinemia provided by genetic strains of diabetic mice was investigated. However, it was first necessary to determine the properties of cholesterol ester hy- drolase(s) in mouse tissue preparations. The effect of pH on cholesterol hydrolase activity in subcellu- lar fractions obtained by differential centrifuga- tion from mouse liver, heart and fat pad homo- genates is shown in Fig. 1. In the case of the liver and heart enzyme, a distinct acid pH optimum of 4.5-5 can be observed along with a smaller but distinct optimum at pH 7 (panels A and B). These results showing predominantly an acid pH opti- mum with mouse liver and heart preparations are very similar to those reported for cholesterol ester hydrolase activity in prep~irations from rat liver and heart [15]. The effect of pH on cholesterol
0
A Liver 10
B. Heart 30 f
C.
g e ") li d pH
F a t Pads
/\ /.,. 5 6 7 8 9
Fig. 1. Effect of pH on cholesterol ester hydrolase activity in subcellular fractions from mouse liver, heart and epididymal fat pads. Assays were performed at the indicated pH values with subcellular fractions obtained by differential centrifugation (17000Xg pellet, A; 100000× g pellet, R; 100000X g super- natant, @) from liver (panel A), heart (panel B), and with low-speed (5 000X g) supernatant fraction from epididymal fat pads (0; panel C - mean of three fat pad preparations).
147
ester hydrolase activity in a mouse fat pad pre- paration is also shown in Fig. 1 (Panel C). Al- though a shoulder of activity could be observed at pH 5, the only distinct pH optimum was observed at a pH value of 6.5. These results are in contrast to cholesterol ester hydrolase activity in rat epidi- dymal fat pad preparations, in which a distinct pH optima at 5 was observed [15].
The genetic strain of diabetic mice (C57BL/Ks db/db) is characterized by obesity, hyperglycemia and hyperinsulinemia [21-23]. Diabetic mice at 12 weeks of age exhibited an increased body weight (49---1 g compared to 26--1 g for controls) and increased plasma levels of glucose (481 ± 24 mg/d l compared to 161 ± 5 mg/d l for controls) and in- sulin (over 500/zU/ml compared to 23 ± 3/~U/ml for controls).
A comparison of cholesterol ester hydrolase activity and N-acetylglucosaminidase activity in liver, heart and fat preparations from control and diabetic mice is shown in Table II. Preparations consisted of pooled tissue samples from five animals. The average wet weight of livers removed from diabetic mice was approximately double that from control .animals; similar results have been reported previously [21,24]. Liver preparations from diabetic mice demonstrated an elevation in N-acetylglucosaminidase activity relative to con- trols, but acid cholesterol ester hydrolase was elevated only when activity was expressed per mg protein (Table II).
The wet weight of hearts removed from diabetic mice was not significantly different from the weight of hearts removed from age matched control litter mates; diabetes (hyperinsulinemia) resulted in a consistent decrease in acid cholesterol ester hydl'o- lase activity (Table II). N-Acetylglucosaminidase activity was not changed in heart preparations from diabetic mice.
The average wet weight of epididymal fat pads removed from diabetic mice was some 4 to 5-fold greater than the weight of fat pads removed from control mice; the excessive weight gain in diabetic mice has been shown to be due to the accumula- tion of fat in adipose tissue depots [21,24]. The activity of cholesterol ester hydrolase in fat pad preparations was determined at pH 6.5 only since a distinct pH optimum at a more acid pH value was not observed (Fig. 1). Diabetic (hyperin-
148
TABLE II
COMPARISON OF CHOLESTEROL ESTER HYDROLASE AND N-ACETYLGLUCOSAMINIDASE (NAGA) ACTIVITY IN LIVER, HEART AND FAT PAD PREPARATIONS FROM CONTROL AND DIABETIC MICE
Cholesterol ester hydrolase assays were performed at pH 5 for heart and liver preparations and at pH 6.5 for fat pad preparations. Results are the mean ± S.E. for five preparations: P values were calculated by the Student's t-test.
Cholesterol ester hydrolase activity
units/mg protein units/mg DNA
NAGA activity (units/mg protein)
A. Liver Control 3.96 Jr 0.12 Diabetic 4.49 ± 0.19 a
B. Heart Control 8.81 -+ 0.09 Diabetic (db/db) 5.07-+0.31 b
C. Fat Pads Control 108 -+7 Diabetic (db/db) 97 +7
214±21 1.52+0.02 233-- + 17 1.81 --+0.08 a
261 --+22 0.25--+0.01 157± 17 b 0.22+0.01
3.68±0.33 3.46+0.18
a p <0.05. b p <0.001.
sulinemia) did not result in significant alteration in either cholesterol ester hydrolase or N-acetylglu- cosaminidase (Table II).
Discusion
Insulin has been shown to affect lysosomal function as evidenced by the inhibition of proteol- ysis in a variety of rat tissues [25-28]. Wolinsky et al. [7] have shown that insulin deficiency in rats resulted in an increase in lysosomal acid cholesterol ester hydrolase activity in liver and kidney, and that the administrat ion of insulin to these diabetic rats reduced acid cholesterol ester hydrolase in liver and kidney to control levels [7]. In contrast, insulin has been shown to increase the degradat ion of L D L in fibroblasts and increase acid cholesterol ester hydrolase activity in mononuclear cells [9-11 ].
The present investigation was initiated in order to determine whether hyperinsulinemia would in- fluence acid cholesterol ester hydrolase activity in various tissues from rats and mice. Effects of hyperinsulinemia on acid hydrolase activity could be linked to changes in rates of L D L catabolism and therefore be relevant with respect to the suggestion that insulin (hyperinsulinemia) may ac- celerate the pathogenesis of atherosclerosis [14]. Two animal models of chronic hyperinsulinemia
were employed. The first utilized delivery of exog- enous insulin for 6 days using an osmotic mini- p u m p implanted subcutaneously in rats. The sec- ond was the endogenous hyperinsulinemia that is one of the characteristic features of a genetic strain of diabetic mice. The two animal models were similar with respect to the degree of hyperin- sulinemia (elevations of serum insulin levels of greater than 10 fold); an obvious difference be- tween the two models was the durat ion of the hyperinsulinemia (days versus weeks). In addition, administrat ion of exogenous insulin to rats re- sulted in hypoglycemia, whereas the endogenous hyperinsulinemia in the diabetic mice was associ- ated with hyperglycemia.
Acid cholesterol ester hydrolase activity was determined in preparat ions from liver, epididymal fat pads and heart. The liver is the major tissue site for the catabolism of LDL, and adipose tissue is an important extrahepatic site for L D L uptake and degradation [2,3]. Al though cholesterol esters f rom chylomicrons are taken up by the isolated perfused rat heart, this uptake process is non- saturable and receptor- independent and does not involve endocytosis or pinocytosis [29]. The uptake of cholesterol esters by perfused hearts was fol- lowed by a slow rate of hydrolysis [29] but no experiments were presented to indicate whether
this hydrolys is occur red in lysosomes. Therefore, any changes in acid cholesterol ester hydro lase in ca rd iac t issue due to hyper insu l inemia would not be expected to inf luence the turnover of l ipopro- teins, whereas a l te ra t ions in acid hydro lase act ivi ty in liver and fat pads could be l inked to the c learance of l ipopro te ins as p roposed by Wol insky [6]. The po ten t ia l for the ho rmona l regula t ion of ac id lysosomal cholesterol ester hydrolases has been es tabl ished by the observa t ion that adminis- t ra t ion of thyro id hormones to rats resul ted in an increase of acid hydro lase act ivi ty in l iver and fat p a d p repa ra t ions [15].
Wi th bo th models of hyper insu l inemia used in the present invest igat ion, no consis tent effect of insul in was observed on acid cholesterol ester hy- d ro lase act ivi ty in liver and fat pad prepara t ions . In contras t , Wol insky et al. [7] have repor ted that the admin i s t r a t ion of insul in to d iabet ic ( insulin- deficient) rats decreased hepat ic acid hydro lase ac t iv i ty bu t only to the same value as in cont ro l (non-d iabe t ic ) rat livers. These results can be rec- onci led only if it is suggested that l iver acid cholesterol ester hydro lase is not subject to regula- t ion by insul in levels in exess of cont ro l values. I f so, the cont ro l of acid hydro lase act ivi ty is m a r k e d l y di f ferent f rom other metabo l ic processes such as l ipogenesis, since chronic hyper insu l inemia as a consequence of imp lan t ed osmot ic min ipumps in rats or as a character is t ic feature of genetic s t ra ins of d iabe t ic mice has been demons t r a t ed to increase the act ivi ty of cer ta in hepa t ic l ipogenic enzymes [20,21,30].
In cont ras t to results with liver and fat pad p repara t ions , acid cholesterol ester hydro lase ac- t ivi ty was s ignif icant ly decreased in hear t p repara - t ions f rom insul in- t rea ted rats and d iabet ic mice. This effect a p p e a r e d to be specific since the activi- ties of the lysosomal marke r enzymes ( N - ace ty lg luosamin idase and f l -glucuronidase) were no t changed. As no ted above, any a l tera t ions in acid cholesterol ester hydro lase act ivi ty in card iac t issue in response to hyper insu l inemia would not be l inked to changes in the turnover of l ipopro- teins. However , the decrease in acid cholesterol ester hydro lase act ivi ty in hear t p repa ra t ions is consis tent with the inhib i t ion of card iac pro te inase act ivi ty by insul in [28,31,32].
149
Acknowledgements
Ms. M. Scace and Mr. G. Groves p rov ided skillful technical assistance. This work was sup- po r t ed by grants f rom the A lbe r t a Hear t F o u n d a - tion. D.L.S. is a Scholar of the Medica l Research
Counci l of Canada .
References
l Pittman, R.C., Attic, A.D., Carew, T.E. and Steinberg, D. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 5345-5349
2 Pittman, R.C., Attie, A.D., Carew, T.E. and Steinberg, D. (1982) Biochim. Biophys. Acta. 710, 7-14
3 Floren, C.-H. and Nilsson, A. (1977) Biochem. J. 168, 483-494
4 Goldstein, J.L. and Brown, M.D. (1977) Metabolism 26, 1257-1275
5 Goldstein, J.L. and Brown, M.S. (1977) Annu. Rev. Bio- chem. 46, 897-930
6 Wolinsky, H. (1980) Circ. Res. 47, 301-311 7 Wolinsky, H., Katz. D., Markle, R., Mills, J., Brem, S. and
