Journal of Toxicology and Environmental Health
ISSN: 0098-4108 (Print) (Online) Journal homepage: http://www.tandfonline.com/loi/uteh19
Assessment of diabetogenic drug activity in the rat: Diphenylhydantoin Carl R. Mackerer , Robert N. Saunders , Janet R. Haettinger & Myron A. Mehlman To cite this article: Carl R. Mackerer , Robert N. Saunders , Janet R. Haettinger & Myron A. Mehlman (1976) Assessment of diabetogenic drug activity in the rat: Diphenylhydantoin, Journal of Toxicology and Environmental Health, 2:1, 139-151, DOI: 10.1080/15287397609529422 To link to this article: http://dx.doi.org/10.1080/15287397609529422
Published online: 19 Oct 2009.
Submit your article to this journal
View related articles
Citing articles: 1 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=uteh19 Download by: [Monash University Library]
Date: 09 November 2015, At: 23:40
ASSESSMENT OF DIABETOGENIC DRUG ACTIVITY IN THE RAT: DIPHENYLHYDANTOIN Carl R. Mackerer, Robert N. Saunders, Janet R. Haettinger
Department of Biological Research, Searle Laboratories, Chicago, Illinois
Downloaded by [Monash University Library] at 23:40 09 November 2015
Myron A. Mehlman
National Institutes of Health, Bethesda, Maryland An animal model is described that can be used to detect drugs that may exacerbate or ameliorate diabetes. The design of this model is based upon the finding that hyperglycemia caused by intravenous administration of streptozotocin to rats is inversely correlated with pancreatic insulin concentration when expressed as total pancreatic insulin (ng) per body weight (g). Drugs that alter this relationship may be classified, in a preliminary manner, as ameliorative or diabetogenic. The diabetogenic activity of diphenylhydantoin (DPH) via oral administration was assessed in both normal and streptozotocin diabetic rats. Rats were fed powdered chow diet, with or without 0.2% (w/w) DPH, for 19 days. Food consumption and rat weight were recorded daily; whole blood glucose concentrations were determined at the start of the study and at the midpoint. At sacrifice liver and pancreas were excised and blood samples were collected. Protein, glycogen, and lipid levels were determined in liver; insulin in pancreas; and insulin, ketone bodies, glucose, and Iipid in blood. DPH treatment did not affect growth or food consumption. The drug dose, calculated from the food consumption data, was 139 mg/kg-day for the normal rats and 311 mg/kg-day for the diabetic rats. DPH increased liver weight and lipid content in both normal and diabetic rats and lowered blood serum triglyceride concentration in normal rats. However, the concentrations of whole blood glucose, blood serum insulin, and pancreatic insulin were not altered by DPH treatment. The results indicate that the diabetogenic side effect of DPH cannot be observed in rats when the drug is administered orally.
INTRODUCTION Diphenylhydantoin (DPH) has been used since its introduction in 1938 by Merritt and Putnam (1938a,b) as an anticonvulsant agent for the symptomatic treatment of epilepsy. In 1965 Belton et al. reported that DPH produces hyperglycemia in rabbits; subsequently, similar effects were described in dogs (Sanbar et al., 1967) and humans (Goldberg and Sanbar, 1969; Said et al., 1968; Treasure and Toseland, 1971). The effects of DPH on carbohydrate metabolism may be mediated through a direct effect on secretion. In this regard Levin et al. (1970, 1972) and Kizer et al. (1970) have shown that DPH blocks glucose-stimulated insulin secretion in the The authors appreciate the technical assistance of Daniel Becker and Thomas Tobin. Requests for reprints should be sent to Carl R. Mackerer, Pharmacodynamics Section, Department of Biological Research, Searle Laboratories, P.O. Box 5110, Chicago, Illinois 60680. 139 Journal of Toxicology and Environmental Health, 2:139-151,1976 Copyright © 1976 by Hemisphere Publishing Corporation
140
C. R. MACKERER ET AL.
perfused rat pancreas. Malherbe et al. (1972) and Fariss and Lutcher (1971) have shown that therapeutic dosages of DPH impair insulin secretion in human patients. Although DPH inhibits insulin secretion in the perfused rat pancreas, a diabetogenic effect has not been reported in the rat. In the present study we have investigated the effects of subacute administration of DPH on several biochemical parameters of normal and moderately diabetic rats in an attempt to detect diabetogenic activity.
Downloaded by [Monash University Library] at 23:40 09 November 2015
METHODS Animals Male rats were selected from our stock colony of Charles River rats (CD strain). They were individually housed and fed commercial pelleted diet [Rockland mouse/rat diet (complete)] ad libitum until their use.
Diabetes Rats were anesthetized with diethyl ether and made diabetic by administration of streptozotocin via the tail vein. The streptozotocin was dissolved in 0.05 M citrate buffer, pH 4.5, and used within 5 min. Diphenylhydantoin Administration Normal rats were paired according to their body weight and fed powdered diet or diet containing 0.2% DPH. Food intake was controlled by a pair-feeding regimen, whereby after 1 day of ad libitum feeding, each control rat received an amount of diet equal to that consumed by the paired test rat over the preceding 24-hr period. After administration of 75 mg/kg streptozotocin, diabetic rats were permitted to equilibrate for 1 wk; during this time they received pelleted diet ad libitum. At the end of the equilibration period blood was collected for analysis from a scalpel cut in the tip of the tail, and the rats were paired according to their body weight and whole blood glucose concentration. Rats with blood glucose concentrations below 200 mg/100 ml were discarded. For each rat pair the initial values for blood glucose and body weight did not deviate from the median by more than 5%. DPH was administered in the diet as described above. Rat weight and food consumption were recorded daily. Blood was collected from the tail at the approximate midpoint of the study and analyzed for whole blood glucose. The rats were killed by decapitation, and blood was collected from the wound for analysis of glucose, fatty acids, and cholesterol. Livers were frozen in liquid nitrogen and later weighed and analyzed for protein, cholesterol, fatty acids, phospholipid, triglyceride, total lipid, and glycogen.
DIABETOGENIC ACTIVITY OF DPH
141
Downloaded by [Monash University Library] at 23:40 09 November 2015
Analysis Urine and whole blood glucose (Slein, 1965); whole blood acetoacetate (Mellanby and Williamson, 1965) and 0-hydroxybutyrate (Williamson and Mellanby, 1965); whole blood and liver fatty acid (Dalton and Kowalski, 1967), cholesterol (Levine et al., 1968), and triglyceride (Noble and Campbell, 1970); and liver phospholipid (Zilversmit and Davis, 1950) and glycogen (Montgomery, 1957) were determined by the cited procedures. Pancreata were homogenized immediately after removal, and extracts were prepared for analysis of insulin (Wright et al., 1968). Phadebas insulin kits were used to determine insulin levels in blood serum and pancreas extracts.
Materials Sodium diphenylhydantoin was obtained from Pfaltz and Bauer, Inc., Stamford, Conn.; streptozotocin from Upjohn Co., Kalamazoo, Mich.; rats from Charles River Breeding Laboratories, Inc., Wilmington, Mass.; and Phadebas insulin test kits from Pharmacia AB, Uppsala, Sweden. Purified rat insulin, rather than the procine insulin supplied with the kits, was used as the standard. RESULTS
Diabetes Model Streptozotocin is a diabetogenic agent that acts in the rat by a cytotoxic effect on pancreatic j3-cells (Junod et al., 1967; Rakieten et al,, 1963), resulting in a rapid necrosis (Junod et al., 1967). The effects of streptozotocin are dose related and diabetogenicity has been shown to occur between intravenous doses of 25 and 100 mg/kg (Junod et al., 1969). The effects of a single intravenous administration of streptozotocin on several parameters indicative of diabetes is presented in Table 1. The treated rats showed increased concentrations of urine glucose and whole blood glucose, /3-hydroxybutyrate, and acetoacetate. Rat weight gain was decreased, although food consumption was increased (data not shown), and this resulted in a decreased food efficiency. Rats that received 75 mg/kg streptozotocin displayed a moderate level of diabetes with low concentrations of ketone bodies in the urine as indicated by ketostix reagent strips (Ames Co., Elkhart, Ind.). Of the several biochemical parameters measured, the whole blood glucose level showed the least variability and was, therefore, most highly correlated with severity of the diabetic state. The (3-ceII cytotoxicity of streptozotocin causes the insulin secretory rate to decrease because both the synthetic and storage capacities of the islets are destroyed. In this regard streptozotocin administration caused a
Downloaded by [Monash University Library] at 23:40 09 November 2015
TABLE 1. Rat Growth, Whole Blood Levels of Ketone Bodies and Glucose, and Urine Glucose Concentration after Intravenous Administration of Streptozotocin 0
Rat weight (g) Streptozotocin (mg/kg)
to
0(14) 25(15) 50(14) 75(11)
Initial 192+ 3 1 191 ± 2 2 192 ± 2 3 186 + 3*
Final 419 402 379 294
±9 1 ' 2 t 4* ± 11"3 ± 161>3>4
Blood glucose (mg/lOOml) 91 92 147 321
± ± ± ±
2.4 1 ' 2 3.2 3 ' 4 24.2 1 ' 3 ' 5 15.0 2 ' 4 ' 5
Urine glucose (mg/100ml)
Blood acetoacetate (nmol/ml)
Blood (3-hydroxybutyrate (nmol/ml)
36±3.3l 37 ± 3.42 1920± 998 3 7819 ± 3 5 8 " 2 ' 3
67+ 6 1 54 ± 4 ' 61 ± 6 3 535 ± 3 5 8 " 2 ' 3
139 + 2 2 ' 54 ± 4 1 ' 2 106±212 490 ± 1 7 5 " 2
Rats were injected with streptozotocin as described in "Methods," fed pelleted diet for 37 days, and killed (without prior fasting) by decapitation. Blood was collected from the wound and analyzed for levels of whole blood glucose and ketone bodies. Urine was collected for the 24 hr preceding sacrifice and analyzed for glucose concentration. Pancreata were removed, weighed and analyzed for insulin content. Values represent the mean ± SEM. Numbers in parentheses refer to the number of animals used. Values in each column with the same superscript are significantly different (p < 0.05) by Student's t test.
DIABETOGENIC ACTIVITY OF DPH
143
Downloaded by [Monash University Library] at 23:40 09 November 2015
TABLE 2. Rat Blood Serum Insulin Level and Pancreas Weight and Insulin Content after Intravenous Administration of Streptozotocin
Streptozotocin (mg/kg)
Blood serum insulin (/ig/ml)
0(14) 25(15) 50(14) 75(11)
5.7 ± 0.401'2 5.9 ± 0.353'4 4.6±0.29" 3 3.9 ± 0.272'4
Pancreas weight (g)
1.03 1.01 1.10 1.12
± 0.0361 ±0.043 2 ± 0.0743 ±0.150"
Pancreas insulin (Mg/g) 100.1 83.0 37.5 2.5
± 3.6' ± 2.8 1 ±4.9' ±0.8'
Rats were the same as those used to derive the data for Table 1. Values represent the mean + SEM. Numbers in parentheses refer to the number of animals used. Values in each column with the same superscript are significantly different (p < 0.05) by Student's t test.
dose-dependent decrease of serum and pancreatic insulin concentrations (Table 2; Fig. 1), but total pancreas weight was unchanged. In order for streptozotocin to affect the biochemical parameters indicative of diabetes, the pancreatic insulin content had to be markedly reduced (Fig. 1). Most likely, destruction of the )3-cells by lower doses of streptozotocin resulted in a reduction of the capacity for both storing and synthesizing insulin although the remaining viable cells produced enough insulin to maintain metabolic parameters within normal limits. However, at higher doses of streptozotocin, when pancreatic insulin was considerably reduced, maximal
£ 10
0
200 400 blood glucose,mg/iooml
urine glucose,g/100ml
blood ketones.ju moles/ml
FIGURE 1. Relationship between pancreatic insulin concentration and whole blood glucose (a), urine glucose (b), and whole blood ketone (acetoacetate + (3-hydroxybutyrate) (c) concentrations in the rat. Pancreatic insulin content was reduced by administration of streptozotocin at various concentrations (see footnote of Table 1). Data for (a) were obtained from 98 rats, for (b) from 97 rats, and for (c) from 54 rats.
Downloaded by [Monash University Library] at 23:40 09 November 2015
144
C. R. MACKERER ETAL.
rates of insulin synthesis could not meet the bodily demand and symptoms of diabetes were observed. It was probable that the pancreatic insulin content of the frankly diabetic rats represented insulin in the process of being transported and released rather than a pool of stored insulin; thus in rats with diabetic symptomatology, pancreatic insulin content may be an index of the maximal rate of insulin synthesis. The high correlation between blood glucose and pancreatic insulin content at low levels (Fig. 1) (Mackerer et al., 1975) suggested that a diabetes test model could be developed that would involve evaluating the effects that an agent of unknown activity has on the relationship between these two parameters. An alteration of this relationship would not only identify an agent as having altered the diabetic state but would also yield information useful for pursuing studies relating to mechanism of action. An agent that affects the severity of diabetes in the streptozotocin diabetic rat would be likely to alter the relationship between blood glucose concentration and pancreatic insulin content according to one of the patterns shown in Table 3. The particular pattern that is observed indicates, in a preliminary manner, whether the mechanism can be attributed to altered j3-cell function. The correlation between blood glucose and pancreatic insulin content was evident in all of our experiments regardless of the weight of the rats chosen for study. However, in performing tests of this nature two effects had to be considered when rats of different initial weight or age were compared. As shown in Fig. 2a, pancreas weight increased as rat age and weight increased with ad libitum feeding, but the pancreas weight:body weight ratio declined while the insulin content per gram pancreas increased (Fig. 2b). However, the insulin content per total pancreas per 100 g body weight was constant across the weight range of 150-450 g, which was TABLE 3. Changes in the Relationship between the Concentrations of Pancreatic Insulin and Whole Blood Glucose after Administration of a Hyper- or Hypoglycemic Agent Changes Blood glucose Decreased
Pancreatic insulin Increased Decreased Unchanged
Increased
Increased Decreased Unchanged
Anticipated location of activity0 Pancreatic Pancreatic and extrapancreatic Extrapancreatic Pancreatic and extrapancreatic Pancreatic Extrapancreatic
Pancreatic effects refer to those directed at the /J-cells.
145
DIABETOGENIC ACTIVITY OF DPH
1.5
150 ug insulin/g pancreas
g pancreas
Downloaded by [Monash University Library] at 23:40 09 November 2015
100
0.5
50
450
150
450
body wl,g
FIGURE 2. Age concentration
(body
weight)-related changes of
pancreas weight
(a) and pancreatic
insulin
(b) in normal male rats fed ad libitum.
perhaps the most appropriate way of expressing pancreatic insulin content. We also observed that blood glucose concentration of normal rats (fed Rockland pelleted diet ad libitum) decreased (Fig. 3a) and serum insulin concentrations tended to increase (Fig. 3b) with age. However, in untreated streptozotocin diabetic rats (data not shown) age-related changes of blood glucose were not consistent. Most rats had unchanged levels for at least several weeks although rats occasionally showed worsening of the diabetic state while others spontaneously improved. The incidence of time-related spontaneous improvement in the diabetic condition was far lower than we commonly observe with alloxan diabetic rats. Diphenylhydantoin Treatment: Growth Parameters As shown in Table 4 , DPH administration did not alter food consumption or weight gain. However, as expected, the diabetic rats ate considerably more food than the normal rats but still failed to gain weight. The daily drug dose, calculated from the average daily food consumption and average rat weight at the midpoint of the study, was 139 mg/kg for the normal rats and 311 mg/kg for the diabetic rats. Blood Chemistry DPH administration did not affect the blood glucose concentration (Table 5) or cholesterol concentration (Table 6) of either the normal or
C. R.MACKERERETAL.
146
10
100
E o O
Downloaded by [Monash University Library] at 23:40 09 November 2015
E
5 5
50
a
260
160
360
0L
460 " body wt,g
260
160
360
460
FIGURE 3. Age (body weight)-related changes of whole blood glucose (a) and blood serum insulin (b) in normal male rats fed ad libitum, (a) y = - 0.0597A" + 117 (r = 0.58); (b) y = Q.0Q1X + 2.33 (/• = 0.33).
diabetic rats. In the normal rats, however, DPH lowered the blood serum triglyceride concentration by 30%. Liver Composition In both the normal and the diabetic rat DPH increased the liver weight, total lipid content, and triglyceride concentration (Table 7). in the TABLE 4. Growth, Food Consumption, and Food Efficiency of Normal and Diabetic Rats Fed Diets Containing 0.2% DPH0
Rat
Initial weight
Final weight
Food consumed
treatment
(g)
(g)
(g)
Normal (12)c +0.2% DPH (12)C
218 ±6 227 ±6 (NS)
312 ± 10 318 ± 11 (NS)
403 ± 11 436 ± 14 (NS)
0.23 ± 0.009 0.21 + 0.14 (NS)
Diabetic (6)c +0.2% DPH (6)c
162 ± 8 170 ±9 (NS)
162 ±7 190 + 13 (NS)
521 ±27 532 ± 18 (NS)
_ — -
Food efficiency
Rats were fed for 19 days. Values represent the mean ± SEM. The level of statistical significance is indicated in parentheses; (NS) means not significant [p < 0.05) by Student's f test. Food efficiency is the weight gain:food consumed ratio. Numbers in parentheses refer to the number of animals used.
DIABETOGENIC ACTIVITY OF DPH
147
TABLE 5. Whole Blood Glucose Levels of Normal and Diabetic Rats Fed Diet Containing 0.2% DPHJ
Downloaded by [Monash University Library] at 23:40 09 November 2015
Glucose (mg/100 ml)6 treatment
Initial
Midpoint
Normal (12) +0.2% DPH (12)
97 ± 3 94 ± 2 (NS)
101 ± 4 99 ± 3 (NS)
101 ±2 96 ±2 (NS)
Diabetic (6) +0.2% DPH (6)
454+ 18 439 ± 20 (NS)
487 ± 30 443 ± 24 (NS)
400 ± 36 384 ±9 (NS)
Final
°See footnotes of Table 4. Blood samples were obtained on day 1 (initial) and day 11 (midpoint) of the study from the tail (see "Methods"). On day 20 (final) blood was collected from the wound after decapitation.
diabetic rat liver levels of free fatty acids and phospholipid were also significantly elevated. DPH treatment did not alter the levels of glycogen or protein of either the normal or the diabetic rats. Pancreas Weight and Insulin Level DPH treatment did not affect the pancreas weight or serum insulin concentration of the normal rat, nor the pancreas weight and pancreas insulin content of the diabetic rat (Table 8). DISCUSSION As shown by the results, subacute administration of DPH at high levels did not produce diabetic symptomatology such as hyperglycemia in the TABLE 6. Blood Serum Cholesterol and Triglyceride Concentrations of Normal and Diabetic Rats Fed Diet Containing 0.2% DPH0 Rat treatment
Cholesterol (mg/100 ml)
Triglyceride (mg/100 ml)
Normal (12) +0.2% DPH (12)
86 ±4 82 ± 2 (NS)
72 ±4 50 ±2 «0.05)
Diabetic (6) +0.2% DPH (6)
104 ± 18 102 ±9 (NS)
163±57 179 ±27 (NS)
°Blood samples were obtained after decapitation. See footnotes of Tables 4 and 5.
Downloaded by [Monash University Library] at 23:40 09 November 2015
TABLE 7. Liver Weight and Liver Concentrations of Lipid, Glycogen, and Protein in Normal and Diabetic Rats Fed Diet Containing 0.2% DPH"
Rat
Liver weight
treatment
(g)
Normal (12) +0.2% DPH (12)
11.6 ± 0.6 13.8 ±0.7 (
ISSN: 0098-4108 (Print) (Online) Journal homepage: http://www.tandfonline.com/loi/uteh19
Assessment of diabetogenic drug activity in the rat: Diphenylhydantoin Carl R. Mackerer , Robert N. Saunders , Janet R. Haettinger & Myron A. Mehlman To cite this article: Carl R. Mackerer , Robert N. Saunders , Janet R. Haettinger & Myron A. Mehlman (1976) Assessment of diabetogenic drug activity in the rat: Diphenylhydantoin, Journal of Toxicology and Environmental Health, 2:1, 139-151, DOI: 10.1080/15287397609529422 To link to this article: http://dx.doi.org/10.1080/15287397609529422
Published online: 19 Oct 2009.
Submit your article to this journal
View related articles
Citing articles: 1 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=uteh19 Download by: [Monash University Library]
Date: 09 November 2015, At: 23:40
ASSESSMENT OF DIABETOGENIC DRUG ACTIVITY IN THE RAT: DIPHENYLHYDANTOIN Carl R. Mackerer, Robert N. Saunders, Janet R. Haettinger
Department of Biological Research, Searle Laboratories, Chicago, Illinois
Downloaded by [Monash University Library] at 23:40 09 November 2015
Myron A. Mehlman
National Institutes of Health, Bethesda, Maryland An animal model is described that can be used to detect drugs that may exacerbate or ameliorate diabetes. The design of this model is based upon the finding that hyperglycemia caused by intravenous administration of streptozotocin to rats is inversely correlated with pancreatic insulin concentration when expressed as total pancreatic insulin (ng) per body weight (g). Drugs that alter this relationship may be classified, in a preliminary manner, as ameliorative or diabetogenic. The diabetogenic activity of diphenylhydantoin (DPH) via oral administration was assessed in both normal and streptozotocin diabetic rats. Rats were fed powdered chow diet, with or without 0.2% (w/w) DPH, for 19 days. Food consumption and rat weight were recorded daily; whole blood glucose concentrations were determined at the start of the study and at the midpoint. At sacrifice liver and pancreas were excised and blood samples were collected. Protein, glycogen, and lipid levels were determined in liver; insulin in pancreas; and insulin, ketone bodies, glucose, and Iipid in blood. DPH treatment did not affect growth or food consumption. The drug dose, calculated from the food consumption data, was 139 mg/kg-day for the normal rats and 311 mg/kg-day for the diabetic rats. DPH increased liver weight and lipid content in both normal and diabetic rats and lowered blood serum triglyceride concentration in normal rats. However, the concentrations of whole blood glucose, blood serum insulin, and pancreatic insulin were not altered by DPH treatment. The results indicate that the diabetogenic side effect of DPH cannot be observed in rats when the drug is administered orally.
INTRODUCTION Diphenylhydantoin (DPH) has been used since its introduction in 1938 by Merritt and Putnam (1938a,b) as an anticonvulsant agent for the symptomatic treatment of epilepsy. In 1965 Belton et al. reported that DPH produces hyperglycemia in rabbits; subsequently, similar effects were described in dogs (Sanbar et al., 1967) and humans (Goldberg and Sanbar, 1969; Said et al., 1968; Treasure and Toseland, 1971). The effects of DPH on carbohydrate metabolism may be mediated through a direct effect on secretion. In this regard Levin et al. (1970, 1972) and Kizer et al. (1970) have shown that DPH blocks glucose-stimulated insulin secretion in the The authors appreciate the technical assistance of Daniel Becker and Thomas Tobin. Requests for reprints should be sent to Carl R. Mackerer, Pharmacodynamics Section, Department of Biological Research, Searle Laboratories, P.O. Box 5110, Chicago, Illinois 60680. 139 Journal of Toxicology and Environmental Health, 2:139-151,1976 Copyright © 1976 by Hemisphere Publishing Corporation
140
C. R. MACKERER ET AL.
perfused rat pancreas. Malherbe et al. (1972) and Fariss and Lutcher (1971) have shown that therapeutic dosages of DPH impair insulin secretion in human patients. Although DPH inhibits insulin secretion in the perfused rat pancreas, a diabetogenic effect has not been reported in the rat. In the present study we have investigated the effects of subacute administration of DPH on several biochemical parameters of normal and moderately diabetic rats in an attempt to detect diabetogenic activity.
Downloaded by [Monash University Library] at 23:40 09 November 2015
METHODS Animals Male rats were selected from our stock colony of Charles River rats (CD strain). They were individually housed and fed commercial pelleted diet [Rockland mouse/rat diet (complete)] ad libitum until their use.
Diabetes Rats were anesthetized with diethyl ether and made diabetic by administration of streptozotocin via the tail vein. The streptozotocin was dissolved in 0.05 M citrate buffer, pH 4.5, and used within 5 min. Diphenylhydantoin Administration Normal rats were paired according to their body weight and fed powdered diet or diet containing 0.2% DPH. Food intake was controlled by a pair-feeding regimen, whereby after 1 day of ad libitum feeding, each control rat received an amount of diet equal to that consumed by the paired test rat over the preceding 24-hr period. After administration of 75 mg/kg streptozotocin, diabetic rats were permitted to equilibrate for 1 wk; during this time they received pelleted diet ad libitum. At the end of the equilibration period blood was collected for analysis from a scalpel cut in the tip of the tail, and the rats were paired according to their body weight and whole blood glucose concentration. Rats with blood glucose concentrations below 200 mg/100 ml were discarded. For each rat pair the initial values for blood glucose and body weight did not deviate from the median by more than 5%. DPH was administered in the diet as described above. Rat weight and food consumption were recorded daily. Blood was collected from the tail at the approximate midpoint of the study and analyzed for whole blood glucose. The rats were killed by decapitation, and blood was collected from the wound for analysis of glucose, fatty acids, and cholesterol. Livers were frozen in liquid nitrogen and later weighed and analyzed for protein, cholesterol, fatty acids, phospholipid, triglyceride, total lipid, and glycogen.
DIABETOGENIC ACTIVITY OF DPH
141
Downloaded by [Monash University Library] at 23:40 09 November 2015
Analysis Urine and whole blood glucose (Slein, 1965); whole blood acetoacetate (Mellanby and Williamson, 1965) and 0-hydroxybutyrate (Williamson and Mellanby, 1965); whole blood and liver fatty acid (Dalton and Kowalski, 1967), cholesterol (Levine et al., 1968), and triglyceride (Noble and Campbell, 1970); and liver phospholipid (Zilversmit and Davis, 1950) and glycogen (Montgomery, 1957) were determined by the cited procedures. Pancreata were homogenized immediately after removal, and extracts were prepared for analysis of insulin (Wright et al., 1968). Phadebas insulin kits were used to determine insulin levels in blood serum and pancreas extracts.
Materials Sodium diphenylhydantoin was obtained from Pfaltz and Bauer, Inc., Stamford, Conn.; streptozotocin from Upjohn Co., Kalamazoo, Mich.; rats from Charles River Breeding Laboratories, Inc., Wilmington, Mass.; and Phadebas insulin test kits from Pharmacia AB, Uppsala, Sweden. Purified rat insulin, rather than the procine insulin supplied with the kits, was used as the standard. RESULTS
Diabetes Model Streptozotocin is a diabetogenic agent that acts in the rat by a cytotoxic effect on pancreatic j3-cells (Junod et al., 1967; Rakieten et al,, 1963), resulting in a rapid necrosis (Junod et al., 1967). The effects of streptozotocin are dose related and diabetogenicity has been shown to occur between intravenous doses of 25 and 100 mg/kg (Junod et al., 1969). The effects of a single intravenous administration of streptozotocin on several parameters indicative of diabetes is presented in Table 1. The treated rats showed increased concentrations of urine glucose and whole blood glucose, /3-hydroxybutyrate, and acetoacetate. Rat weight gain was decreased, although food consumption was increased (data not shown), and this resulted in a decreased food efficiency. Rats that received 75 mg/kg streptozotocin displayed a moderate level of diabetes with low concentrations of ketone bodies in the urine as indicated by ketostix reagent strips (Ames Co., Elkhart, Ind.). Of the several biochemical parameters measured, the whole blood glucose level showed the least variability and was, therefore, most highly correlated with severity of the diabetic state. The (3-ceII cytotoxicity of streptozotocin causes the insulin secretory rate to decrease because both the synthetic and storage capacities of the islets are destroyed. In this regard streptozotocin administration caused a
Downloaded by [Monash University Library] at 23:40 09 November 2015
TABLE 1. Rat Growth, Whole Blood Levels of Ketone Bodies and Glucose, and Urine Glucose Concentration after Intravenous Administration of Streptozotocin 0
Rat weight (g) Streptozotocin (mg/kg)
to
0(14) 25(15) 50(14) 75(11)
Initial 192+ 3 1 191 ± 2 2 192 ± 2 3 186 + 3*
Final 419 402 379 294
±9 1 ' 2 t 4* ± 11"3 ± 161>3>4
Blood glucose (mg/lOOml) 91 92 147 321
± ± ± ±
2.4 1 ' 2 3.2 3 ' 4 24.2 1 ' 3 ' 5 15.0 2 ' 4 ' 5
Urine glucose (mg/100ml)
Blood acetoacetate (nmol/ml)
Blood (3-hydroxybutyrate (nmol/ml)
36±3.3l 37 ± 3.42 1920± 998 3 7819 ± 3 5 8 " 2 ' 3
67+ 6 1 54 ± 4 ' 61 ± 6 3 535 ± 3 5 8 " 2 ' 3
139 + 2 2 ' 54 ± 4 1 ' 2 106±212 490 ± 1 7 5 " 2
Rats were injected with streptozotocin as described in "Methods," fed pelleted diet for 37 days, and killed (without prior fasting) by decapitation. Blood was collected from the wound and analyzed for levels of whole blood glucose and ketone bodies. Urine was collected for the 24 hr preceding sacrifice and analyzed for glucose concentration. Pancreata were removed, weighed and analyzed for insulin content. Values represent the mean ± SEM. Numbers in parentheses refer to the number of animals used. Values in each column with the same superscript are significantly different (p < 0.05) by Student's t test.
DIABETOGENIC ACTIVITY OF DPH
143
Downloaded by [Monash University Library] at 23:40 09 November 2015
TABLE 2. Rat Blood Serum Insulin Level and Pancreas Weight and Insulin Content after Intravenous Administration of Streptozotocin
Streptozotocin (mg/kg)
Blood serum insulin (/ig/ml)
0(14) 25(15) 50(14) 75(11)
5.7 ± 0.401'2 5.9 ± 0.353'4 4.6±0.29" 3 3.9 ± 0.272'4
Pancreas weight (g)
1.03 1.01 1.10 1.12
± 0.0361 ±0.043 2 ± 0.0743 ±0.150"
Pancreas insulin (Mg/g) 100.1 83.0 37.5 2.5
± 3.6' ± 2.8 1 ±4.9' ±0.8'
Rats were the same as those used to derive the data for Table 1. Values represent the mean + SEM. Numbers in parentheses refer to the number of animals used. Values in each column with the same superscript are significantly different (p < 0.05) by Student's t test.
dose-dependent decrease of serum and pancreatic insulin concentrations (Table 2; Fig. 1), but total pancreas weight was unchanged. In order for streptozotocin to affect the biochemical parameters indicative of diabetes, the pancreatic insulin content had to be markedly reduced (Fig. 1). Most likely, destruction of the )3-cells by lower doses of streptozotocin resulted in a reduction of the capacity for both storing and synthesizing insulin although the remaining viable cells produced enough insulin to maintain metabolic parameters within normal limits. However, at higher doses of streptozotocin, when pancreatic insulin was considerably reduced, maximal
£ 10
0
200 400 blood glucose,mg/iooml
urine glucose,g/100ml
blood ketones.ju moles/ml
FIGURE 1. Relationship between pancreatic insulin concentration and whole blood glucose (a), urine glucose (b), and whole blood ketone (acetoacetate + (3-hydroxybutyrate) (c) concentrations in the rat. Pancreatic insulin content was reduced by administration of streptozotocin at various concentrations (see footnote of Table 1). Data for (a) were obtained from 98 rats, for (b) from 97 rats, and for (c) from 54 rats.
Downloaded by [Monash University Library] at 23:40 09 November 2015
144
C. R. MACKERER ETAL.
rates of insulin synthesis could not meet the bodily demand and symptoms of diabetes were observed. It was probable that the pancreatic insulin content of the frankly diabetic rats represented insulin in the process of being transported and released rather than a pool of stored insulin; thus in rats with diabetic symptomatology, pancreatic insulin content may be an index of the maximal rate of insulin synthesis. The high correlation between blood glucose and pancreatic insulin content at low levels (Fig. 1) (Mackerer et al., 1975) suggested that a diabetes test model could be developed that would involve evaluating the effects that an agent of unknown activity has on the relationship between these two parameters. An alteration of this relationship would not only identify an agent as having altered the diabetic state but would also yield information useful for pursuing studies relating to mechanism of action. An agent that affects the severity of diabetes in the streptozotocin diabetic rat would be likely to alter the relationship between blood glucose concentration and pancreatic insulin content according to one of the patterns shown in Table 3. The particular pattern that is observed indicates, in a preliminary manner, whether the mechanism can be attributed to altered j3-cell function. The correlation between blood glucose and pancreatic insulin content was evident in all of our experiments regardless of the weight of the rats chosen for study. However, in performing tests of this nature two effects had to be considered when rats of different initial weight or age were compared. As shown in Fig. 2a, pancreas weight increased as rat age and weight increased with ad libitum feeding, but the pancreas weight:body weight ratio declined while the insulin content per gram pancreas increased (Fig. 2b). However, the insulin content per total pancreas per 100 g body weight was constant across the weight range of 150-450 g, which was TABLE 3. Changes in the Relationship between the Concentrations of Pancreatic Insulin and Whole Blood Glucose after Administration of a Hyper- or Hypoglycemic Agent Changes Blood glucose Decreased
Pancreatic insulin Increased Decreased Unchanged
Increased
Increased Decreased Unchanged
Anticipated location of activity0 Pancreatic Pancreatic and extrapancreatic Extrapancreatic Pancreatic and extrapancreatic Pancreatic Extrapancreatic
Pancreatic effects refer to those directed at the /J-cells.
145
DIABETOGENIC ACTIVITY OF DPH
1.5
150 ug insulin/g pancreas
g pancreas
Downloaded by [Monash University Library] at 23:40 09 November 2015
100
0.5
50
450
150
450
body wl,g
FIGURE 2. Age concentration
(body
weight)-related changes of
pancreas weight
(a) and pancreatic
insulin
(b) in normal male rats fed ad libitum.
perhaps the most appropriate way of expressing pancreatic insulin content. We also observed that blood glucose concentration of normal rats (fed Rockland pelleted diet ad libitum) decreased (Fig. 3a) and serum insulin concentrations tended to increase (Fig. 3b) with age. However, in untreated streptozotocin diabetic rats (data not shown) age-related changes of blood glucose were not consistent. Most rats had unchanged levels for at least several weeks although rats occasionally showed worsening of the diabetic state while others spontaneously improved. The incidence of time-related spontaneous improvement in the diabetic condition was far lower than we commonly observe with alloxan diabetic rats. Diphenylhydantoin Treatment: Growth Parameters As shown in Table 4 , DPH administration did not alter food consumption or weight gain. However, as expected, the diabetic rats ate considerably more food than the normal rats but still failed to gain weight. The daily drug dose, calculated from the average daily food consumption and average rat weight at the midpoint of the study, was 139 mg/kg for the normal rats and 311 mg/kg for the diabetic rats. Blood Chemistry DPH administration did not affect the blood glucose concentration (Table 5) or cholesterol concentration (Table 6) of either the normal or
C. R.MACKERERETAL.
146
10
100
E o O
Downloaded by [Monash University Library] at 23:40 09 November 2015
E
5 5
50
a
260
160
360
0L
460 " body wt,g
260
160
360
460
FIGURE 3. Age (body weight)-related changes of whole blood glucose (a) and blood serum insulin (b) in normal male rats fed ad libitum, (a) y = - 0.0597A" + 117 (r = 0.58); (b) y = Q.0Q1X + 2.33 (/• = 0.33).
diabetic rats. In the normal rats, however, DPH lowered the blood serum triglyceride concentration by 30%. Liver Composition In both the normal and the diabetic rat DPH increased the liver weight, total lipid content, and triglyceride concentration (Table 7). in the TABLE 4. Growth, Food Consumption, and Food Efficiency of Normal and Diabetic Rats Fed Diets Containing 0.2% DPH0
Rat
Initial weight
Final weight
Food consumed
treatment
(g)
(g)
(g)
Normal (12)c +0.2% DPH (12)C
218 ±6 227 ±6 (NS)
312 ± 10 318 ± 11 (NS)
403 ± 11 436 ± 14 (NS)
0.23 ± 0.009 0.21 + 0.14 (NS)
Diabetic (6)c +0.2% DPH (6)c
162 ± 8 170 ±9 (NS)
162 ±7 190 + 13 (NS)
521 ±27 532 ± 18 (NS)
_ — -
Food efficiency
Rats were fed for 19 days. Values represent the mean ± SEM. The level of statistical significance is indicated in parentheses; (NS) means not significant [p < 0.05) by Student's f test. Food efficiency is the weight gain:food consumed ratio. Numbers in parentheses refer to the number of animals used.
DIABETOGENIC ACTIVITY OF DPH
147
TABLE 5. Whole Blood Glucose Levels of Normal and Diabetic Rats Fed Diet Containing 0.2% DPHJ
Downloaded by [Monash University Library] at 23:40 09 November 2015
Glucose (mg/100 ml)6 treatment
Initial
Midpoint
Normal (12) +0.2% DPH (12)
97 ± 3 94 ± 2 (NS)
101 ± 4 99 ± 3 (NS)
101 ±2 96 ±2 (NS)
Diabetic (6) +0.2% DPH (6)
454+ 18 439 ± 20 (NS)
487 ± 30 443 ± 24 (NS)
400 ± 36 384 ±9 (NS)
Final
°See footnotes of Table 4. Blood samples were obtained on day 1 (initial) and day 11 (midpoint) of the study from the tail (see "Methods"). On day 20 (final) blood was collected from the wound after decapitation.
diabetic rat liver levels of free fatty acids and phospholipid were also significantly elevated. DPH treatment did not alter the levels of glycogen or protein of either the normal or the diabetic rats. Pancreas Weight and Insulin Level DPH treatment did not affect the pancreas weight or serum insulin concentration of the normal rat, nor the pancreas weight and pancreas insulin content of the diabetic rat (Table 8). DISCUSSION As shown by the results, subacute administration of DPH at high levels did not produce diabetic symptomatology such as hyperglycemia in the TABLE 6. Blood Serum Cholesterol and Triglyceride Concentrations of Normal and Diabetic Rats Fed Diet Containing 0.2% DPH0 Rat treatment
Cholesterol (mg/100 ml)
Triglyceride (mg/100 ml)
Normal (12) +0.2% DPH (12)
86 ±4 82 ± 2 (NS)
72 ±4 50 ±2 «0.05)
Diabetic (6) +0.2% DPH (6)
104 ± 18 102 ±9 (NS)
163±57 179 ±27 (NS)
°Blood samples were obtained after decapitation. See footnotes of Tables 4 and 5.
Downloaded by [Monash University Library] at 23:40 09 November 2015
TABLE 7. Liver Weight and Liver Concentrations of Lipid, Glycogen, and Protein in Normal and Diabetic Rats Fed Diet Containing 0.2% DPH"
Rat
Liver weight
treatment
(g)
Normal (12) +0.2% DPH (12)
11.6 ± 0.6 13.8 ±0.7 (
