Neonatal Treatment With Beta-Cell Stimulatory Agents Reduces the Incidence of Diabetes in BB Rats

The aim of the study was to investigate whether various beta-cell stimulatory drugs, given neonatally, influence the incidence of diabetes in BB rats. Newborn BB rats were treated twice daily for 6 days and diabetes development was observed during the following 200-day study period. Compared to a diabetes incidence of 63.8% in 163 control BB rats which received saline or were untreated, the percentage of experimental BB rats that developed diabetes was as follows in the different subgroups: arginineglucose: 47% (n= 73, p < 0.02); glucagon: 37% (n = 93, p < 0.0001); tolbutamide-glucose: 36% (n = 58, p < 0.0005); and theophylline-glucose: 39% (n = 41, p < 0.005). A long-term arginine-glucose treatment was not superior to the shorter neonatal treatment. Histological examination revealed a higher degree of insulitis in diabetic than in non-diabetic animals but no difference according to the kind of treatment was observed. Finally, we found that the diabetes incidence in BB rats was higher in the first litter compared to subsequent litters (p = 0.04). Thus, neonatal treatment with various beta-cell stimulatory agents reduces diabetes incidence in BB rats. The theory behind the study, that the treatment accelerates beta-cell maturation leading to increased immunological tolerance towards beta cells, is discussed.


e et al., 198
).

Normally, beta cells are immature at birth (Freinkel et al., 1984; Grasso et al., 1973;  Otonkoski et al., 1988) whereas in the diabetic pregnancy the beta cells mature earlier (Heding   et al., 1980).Since diabetes is an autoimmune disease resulting from a break of self tolerance, we wanted to investigate whether neonatal beta- cell stimulation, and thereby early maturation of the beta cells like in the diabetic pregnancy with fetal hyperglycemia, could lower the diabetes incidence.Normally, neonatal beta cells are insensitive to glucose during the first two weeks of life (Otonkoski et al., 1988) but sensitive to arginine, glucagon, and theophylline (Otonkos- ki, 1988).Therefore, in this study we used the three mentioned agents as well as tolbutamide; all these molecules are non-toxic to mammalians and they have different mechanisms of action.The optimal duration of the treatment is unknown.Therefore, besides short-term stimu- lation which we arbitrarily chose to be twice daily for six days, we also performed a pro- longed experiment with treatment weekly throughout the entire study.


METHODS

Neonatal Stimulation of Beta Cells BB rats (Mollegaard, L1.Skensved, Denmark)  were paired in our Institute.No diabetic BB rats were used as fathers, or as mothers during the pregnancy and the weaning period.From the day of bir h and during the 5 following days the newborn rats were treated twice daily (morning and evening) with intraperitoneal or subcuta- neous injections of 100 tl sterile water or saline to which had been added: 7.5 mg (1.1 g/kg body weight) L-arginine (Sigma, St. Louis, USA) and 2.5mg (0.4g/kg BW) glucose (n=73 BB rats); 0.01mg(1.4mg/kgBW)glucagon (Novo Nor- disk, Bagsvaerd, Denmark) and 5 mg (0.7 g/kg BW) glucose (n=43); 0.02mg (2.8mg/kg BW) glucagon zinc protamine (Novo Nordisk) (n 50);1.25 mg (180mg/kg BW) tolbutamide (Hoechst, Frankfurt, Germany) and 10 mg (1.4g/kg BW) glucose (n=58); or 0.140mg (20mg/kg BW) theophylline and 10mg (1.4g/ kg BW) glucose (n=41).The dosage of the different compounds was arbitrarily chosen within the limits of possible pharmacological treatment.Addition of glucose was performed in order to avoid any risk of hypoglycemia and the amount was arbitrary as well.Control BB rats were given saline i.p. (n=46), saline s.c. (n 49), or were untreated (n 68).

The animals were weaned after 3 weeks and allowed free access to chow and water

were observed daily, and weighed a
d examined weekly for glucosuria (TesTape, Lilly, Indianapolis, USA).Diabetes was diagnosed if at least ++ glucosuria (1/4%)with > 10% weight loss (compared to the previous week) was observed.When this occurred the animals were sacrificed by CO2 anesthesia and the blood glucose values were measured.The study was terminated at day 200.In a long-term arginine-glucose treatment group (n=31), L-arginine (75mg/ml)-glucose (25 mg/ml) was given i.p. twice a day from the day of birth and during the 5 following days, afterwards once a week to the end of the study at a dose of 100 tl/6 g BW up to a maximum dose of 1000 tl.Control BB rats (n =36) were corre- spondingly given saline.


Histological Studies

The paraffin-embedded pancreases were cut into 4 tm sections and stained with hematoxylin and eosin.An area of 11.4mm 2 (11,400,000 tm 2) in sections from each BB rat was scored in a Leitz Wetzler dialux microscope equipped with a Microvid image analysis system for the degree of insulitis.Insulitis was scored semiquantitatively with Grade 0 denoting morphologically unaffected islets, Grade 1 a very disperse intra- insular infiltration of mononuclear cells, Grade 2 a more prono nced intra-insular infiltration with either disperse or small clusters of mono- nuclear cells and, finally, Grade 3 denoting islets dominated by infiltrating mononuclear cells.In Grade 3 the islets often show an altered architecture and pyknotic cell nuclei.Islets with peri-insular infiltration were placed in group 2 or 3 depending on the intra-insular infiltration of mononuclear cells.Ten islets were scored in each section and the average was calculated and considered the final score of insulitis.

Statistics ences between the various treated groups, and their incidence curves were comparable.On the other hand, all the experimental groups each displayed a substantially lower diabetes inci- dence than the 163 control BB rats of which 63.8% developed diabetes during the 200 days observation period.

In

igure 2 the rats are
ivided according to sex.The curves for the male and female rats were similar for both the treated and control animals, respectively, apart from a slightly lower incidence of diabetes for the female BB rats at the end of the study (treated BB rats: 41.1% vs. 38.6%,p>0.10, control BB rats: 67.6% vs. 60.9%,p > 0.10).

Figure 3 shows the result of the extended arginine-glucose treatment.The diabetes inci- dence was significantly lower (p=0.04)for the long-term treated animals than for the control group.

The cumulative diabetes incidence observed in each test group was calculated by Kaplan-Meier estimation and statistical significance was eval- uated by the log-rank (Peto) test.Differences in diabetes incidence between groups of litters were compared by means of t-test after a variance stabilizing angle transformation of the frequency within litters (Hald, 1957).Other differences were estimated by the Mann-Whit- ney test.Spearman's rank correlation test was used for the calcul tions of the coefficient of correlation.Data are presented as mean 4-SEM.


RESULTS


Diabetes Incidence


Histological Studies

The result of the examination of the islets for insulitis is shown in Table II.All the groups of diabetic rats had significantly higher scores than the non-diabetic rats.Ther were no differences between the various experimental groups.How- ever, it should be noticed that overall the degree of insulitis was modest and massive insulitis was only rarely seen.Control rats with early diabetes development (before day 100) showed similar degrees of insulitis as rats that became diabetic after day 100 (1.44-0.1 vs. 1.14-0.1,(p > 0.10).Likewise, no differences were seen for treated diabetic animals (1.2 4-0.2 vs. 1.1 4-0.1 insulitis score).

The results of the different neonatal treatments can be seen in Fi ure 1 and Table I.The experimental BB rats received either arginine, glucagon, tolbutamide or theophylline, and the diabetes incidences were found to remain between 30% and 47%.There were no differ-


Studies of Litters and Sex

Litters contained more female than male BB rats in both the experimental and control groups (42.3% and 43.6% males, respectively).Litter sizes were similar in the experimental and the  control group (6.6 4-0.5 and 6.5 4-0.7 rats, re- spectively).There was no association between litter size and diabetes incidence in any of the

ter size was the same
in the different litter generations.

To investigate the correlation between litter number and diabetes incidence, the 25 control litters with a total of 163 offspring were divided into a group of 12 first litters and a group of 13 second or later litters (Tab.III).The overall diabetes incidences were 57/77=0.74

and 47/86=0.54,respectively.By comparison of these incidences after angle transformation the difference was found statistically significant (p=0.04).The mean litter number was 1.754-0.20 for the experimental BB rats and 1.964-0.24for the controls, which was not significantly different.,,I FIGURE 2 Cumulative diabetes incidence f r male and female BB rats in the collective neonatally treated group (M F and in the control group (M m; F ).   Significance tests refer to the result of non-diabetic control BB rats Non-diabetic rats Diabetic rats

Control BB rats Treated BB rats

Long-term treated B

rats 0.6 + 0.1 (70) 0.84-0.
(158) p<0.005 0.7 + 0.2 (9) n.s.

1.2 4-0.1 (87) p < 10 -7 1.1 4-0.1 (88) p < 10 -6 1.1 4-0.2 (15) p<0.005


DISCUSSION

The present investigations have shown in largescale experiments that the diabetes incidence in BB rats can be reduced by neonatal injections of various agents known to stimulate insulin secretion by different mechanisms.These agents include arginine, glucagon, and for the first time tolbutamide and theophylline.Also, for longterm treatment with arginine-glucose a reduc- tion in diabetes incidence is found, but this treatment is not superior compared with the proper neonatal one.

Until now only a few other studies have dealt with the concept of neonatal beta-cell stimula- tion.In a barrier-confined colony of BB rats with a high diabetes incidence, this was moderately but significantly reduced and the onset of diabetes delayed after neonatal treatment with arginine and glucose combined, but given only once a day (Hansen et al., 1993).In NOD mice the diabetes

cidence has
been found by our group to be reduced by neonatal treatment with glucose (Bock et al., 1991).We have as yet no explanation for the discrepancy between NOD mice and BB rats in which glucose treatment per se does not significantly alter the diabetes incidence (Buschard et al., 1990).Interestingly, female NOD mice treated with neonatal injections the first 6 days of life with I mg arginine and 0.5 mg glucose combined displayed diabetes earlier and with a higher incidence (Senecat et al.,  1994).Obviously, the outcome of arginine on the one hand and of pure glucose on the other is different, and the effect of these compounds on beta-cell maturation should be studied with molecular biological methods.In another approach of early beta-cell stimulation, 4-week old NOD female mice were fed with 1 to 1.75g glucose orally per day until the age of 30 weeks; at that time 37% of the experimental mice had developed diabetes compared to 86% of the controls (Vardi and Buschard, 1995).Regarding tolbutamide, treatment starting at the time of weaning has resulted in a reduced diabetes incidence in NOD mice (Williams et al., 1993); although different in time of treatment this study may be comparable with our neonatal one.

The precise mechanism behind the observed reduction of diabetes incidence and delayed onset of the disease in BB rats is unknown.The theory behind the study is that the antigenic self of the beta cells is less well established because the cells are immature at birth and not fully developed until later in life; antigenic differ- ences should then exist between young and adult beta cells, which has indeed been con- firmed.A 38 kDa protein, reacting with sera from diabetic BB rats, is present in adult but not in neonatal BB rat islet cells (Ko et al., 1994).Furthermore, adult but not neonatal islets are destroyed after transplantation to diabetic BB rats (Ihm et al., 1991).Another study measured in vitro cytotoxicity mediated by mononuclear spleen cells from newly diabetic BB rats and found a significantly lower cytotoxicity towards islet cells from BB r ts < 4 days old than towards islet cells from adult insulitis-free BB rat islets (Ekblond et al., 1995).Islet cell maturation in BB rats-as evidenced by sensitivity to cytotoxi- city-was not seen before the age of 21 days after birth (Ekblond et al., 1995).

The finding in the present study-that the litter sequence number seems to influence the diabetes incidence in BB rats-is to our knowl- edge a new observation which may reflect that, among diabetes-prone rats, the risk of developing diabetes in the offspring decreases with maternal age; this parallels findings in humans (Warram et al., 1991).

It has been suggested that the protection of the diabetic pregnancy against development of diabetes in the offspring may be due to a direct immunological mechanism (Dosch et al., 1993).This might be mediated by placental transfer of autoantibodies which in the fetus, according to the immunological network theory, might raise specific regulator cells.However, offspring of Type 2 diabetic mothers also have lower incidence of Type 1 diabetes compared to children of Type 2 diabetic fathers (Green et al.,  1994) and ype 2 diabetic mothers do not display autoantibodies.This may support the mechanism involving early stimulation of the beta-cells as suggested in this study.

FIGURE

FIGURECumulative diabetes incidence for the different treated groups (arginine-glucose glucagon


FIGURE 3
3
FIGURE 3 Cumulative diabetes incidence for long-term arginine-gluc se treated BB rats


TABLE

Number of animals and cumulative diabetes incidence at 200 days of age in the different treatment groups Number of BB rats Diabetes incidence at 200 days of age(%)
Difference from controlsArginine-glucose7347p < 0.02Glucagon-glucose4342p < 0.02Glucagon zink protamine5030p < 0.00005Tolbutamide-glucose5836p < 0.0005Theophylline-glucose4139p < 0.005Controls16364

TABLE II
II
Average degree of insulitis 4-SEM in the different groups.Number of investigated animals is indicated in brackets.


TABLE III
III
Litter size and diabetes incidence for the first and second-fifth litter numbers, respectively, among the diabetic control BB rats
Litter number