1. Acting upon the idea on occurrence of diabetes published previously by Okamoto, we conducted our studies with 87 sorts of chemicals selected mainly from zinc-affine reagents and also from substances similar to these in structural formulae and closely related chemically to these and have arrived at the following results: (1) We have discovered 17 new substances effective in producing diabetes in rabbits, namely, 8-hydroxyquinaldine, 2, 4-dimethyl-8-hydroxyquinoline, 13 substances with phenylazo bases (phenylazo, tolylazo, hydroxyphenylazo, dimethylaminophenylazo, acetaminophenylazo and aminophenylazo bases) in the position 5 of oxine and 8-hydroxyquinaldine, diphenylthiocarbazide (which, however, apparently changed into dithizone while preparing the injection solution) and nicotine. Except nicotine, all these compounds being insoluble in water, were used dissolved in alkaline solvents or hydrochloric acid under room temperature or by applying heat. (2) The minimum dosis required for inducing diabetes has been estimated to range from 5mg./kg. in the smallest to 45mg./kg. in the highest for quinoline compounds, from several mg./kg. to 10mg./kg. for nicotine and for diphenylthiocarbazide 10-20cc./kg. of saturated solution in saturated aqueous solution of lithium carbonate. (3) These substances injured the β-cells of Langerhans islets and induced permanent diabetes by one single intravenous injections except nicotine which gradually cause severe lesions to β-cells upon 3 or 4 administrations at 5-15 days' intervals in doses of several mg./kg. and finally induces diabetes. In the cases of severe diabetization, the β-cells of the Langerhans islets were reduced to below 12 (below 1/3 of the normal), and in most cases to below 8 (below 1/5 of the normal) in number per islet. (4) At the earliest stage of administration of a reagent, excepting nicotine, a marked fluctuation was seen in blood sugar content. There are 3 major types of this fluctuation.(a) In the majority of cases, the blood sugar content temporarily rose at first (about 3, 6 or 21 hours after administration), then dropped widely (about 3-10, 6-12 or 12-24 hours after administration), rose to permanent high glycemia thereafter and passed over into diabetes. Most of the reagents cause such a type of fluctuation of blood sugar content.(b) Some reagents do not cause the temporary heightening of blood sugar content, but a strong fall of it in 1-4 hours after the administration and then the permanent high glycemia soon sets in. Diabetes is induced in a short time in these cases. 5-(p-hydroxyphenylazo)-8-hydroxyquinoline, 5-(p-dimethylaminophenylazo)-8-hydroxyquinoline and 5-(p-acetaminophenylazo)-8-hydroxyquinoline belong to this type of reagents.(c) The three phases of fluctuation is indeed complete following administration of a few other reagents, but the stage of low blood sugar content is not so distinct here. When the drop of blood sugar content was marked, the following diabetes was also severe. The rise and fall of blood sugar content runs abreast with the lesion of β-cells, the time of abnormal drop of blood sugar content just coinciding with the time of widest necrosis of β-cells. This seems to prove that this drop of blood sugar content is attributable to a sudden release of insulin. (5) 5-methyl-8-hydroxyquinoline, 8-hydroxy-2-(a-methyl) quinoline-methanole, 4-methylthiouracil, diphenylcarbazide, sodium diethyldithiocarbamate and guanidine carbonate are toxic against the β-cells but the toxicity does not reach threshold of diabetization. That is, the reduction of the number of β-cells due to the injury is limited to the extent of dropping to 2/3-1/3 of the normal count and this is not sufficient to cause diabetes.
1. Eighty-six varieties of chemical substances were tested for their efficacy in preventing alloxan diabetes in white rats and 37 items among them were found effective. 33 of the tested materials were zinc-affine substances, of which. 24 were found diabeto-preventive. Of the remaining 53, which were non-zinc-affine, 14 proved to be effective in prevention. 2. Among the substances proved diabeto-preventive, we found some familiar to us in daily life and physiologically important, including 8 vitamins, mainly in B group, 6 varieties of cholic acids, theobromine, sodium bicarbonate and 7 kinds of amino acids. 3. The minimum effective doses and the duration of the effect were determined for all the preventive substances. The results showed that vitamin B12 had the strongest preventive effect, followed by benzimidazole, salicylaldoxime, sodium thiosulphate, BAL, nicotinamide, cholic acids and dopa. 4. The mechanism of diabeto-prevention was studied using 29 representative varieties of the substances found preventive, and the preventive action was found to be accomplished either intra-pancreatically, extra-pancreatically or both intra-and extra-pancreatically. 5. Of the preventive substances, the zinc-affine varieties were found to act intra-pancreatically, probably intra-insularly, without exception-a finding apparently in support of Okamoto's zinc theory of diabetes mellitus. 6. The non-zinc-affine preventive substances (substances containing SH radicals and the majority of the effective vitamins) showed mechanism of extra--pancreatic prevention in most cases. The SH radicaled substances seemingly prevent diabetes by disposing of and detoxicating alloxan before it reached the pancreas. Such considerations also led us to doubt the tenability of Lazarow's glutathione theory of diabetogenesis. 7. With 12 of the preventive reagents as material, we conducted experiments for prevention of alloxan, dithizone and 8-hydroxyquinaldine diabetes in rabbits and found that the preventive effect was considerably impaired in some of the reagents. In general, the preventive action against alloxan diabetes is weaker in rabbits than in white rats. In some cases, the result may be explained by the lower doses of preventive reagents in ratio of the augmented doses of alloxan. Dithizone diabetes was found more difficult to prevent than the diabetes induced by the two other chemicals, which fact may be interpreted as due to the stronger affinity of dithizone with zinc. 8. Eight varieties of the substances found effective in preventing diabetes in white rats were found also effective in preventing experimental diabetes induced by oxine in guinea pigs in most cases. 9. Many substances with similar chemical formula and belonging to the same family of compounds were found to differ in their action-some being diabetogenic, some diabeto-preventive and the others neither diabetogenic nor preventive. In such cases, their manner of combination with zinc, that is, the fact that they are zinc-affine or not, was found to determine their reaction in regard to diabetes.
Male and female rabbits were made alloxan-diabetic, paired among themselves a certain time after the diabetization and the yotings begotten then were named rabbits F1. These F1 rabbits were similarly made diabetic upon attaining maturity, paired and made to breed. Such a procedure was repeated for five generations and the resulting F1-F5 rabbits were made the subject of our study. The results were as follows: 1. In increase of bodyweight after birth, little difference was found between the normal and the F, F5 rabbits, except that the F3 rabbits were slightly inferior in this respect and the F4 and F5 rabbits somewhat better, especially around 60-120 days after birth. 2. The β-cells in the Langerhans islets in the pancreas of F rabbits were found suffering from hypoplasia. It was evidenced by the decrease both in number and in size of β-cells per islet seen on the sections, observable in all ages from 30 days downwards, and in fact this reduction of cells progressed step by step from generation to the next. For example, the number of β-cells per islet was 31.4, 27.2, 22.5, 16.0 and 13.3 and their size 163, 142, 130, 129 and 104 μ2 respectively in P, F1, F2, F3 and F4 rabbits 120 days after birth. 3. The a-cells in F rabbits showed the tendency of slowly decreasing in size from generation to generation (at 120 days after birth, the respective areas showed the values of 148, 129, 126, 120, 111 μ2 in P and F1-F4, rabbits), but their number did not change appreciably. 4. The size of Langerhans islets was always smaller in F rabbits than in normal rabbits from 30 days after birth downwards, and this reduction in size progressed from generation to generation. For example, the mean area of the islets at 120 days after birth was 79.3×, 74.2×, 68.6×, 62.8× and 47.1×100μ2 respectively in P and F1-F4 rabbits. We found no marked change in the number of Langerhans islets. 5. Based upon the assumption that the Langerhans islets are spherical in shape and utilizing the data of measurements of the size and the number of the islet cells and the number of islets in rabbits 120 days of age, we made a tentative calculation of the volume occupied by the cells in 1mm3 of pancreas, and have found that the total volume of β-cells per mm3 of pancreas decrease by leaps and bounds. Thus, the value stands at 82×100, 000μ3 normal rabbits but drops to 49×100, 000μ3 in F1, to 36× in F2, to 14× in F3 and to 11 (or 8)× 100, 000μ3 in F4 rabbits. In F4 rabbits, the volume drops to about 1/8-1/10 of normal rabbits, and this value is presumed to be dangerously near or approximately equal to the threshold value of incidence of diabetes caused by injury in β-cells. The total volume of a-cells showed slight reduction but much less remarkable than that of β-cells. 6. The double charge sugar test conducted on normal and F rabbits showed that while all cases of normal, F1 and F2 rabbits resulted positively in Staub's effect, some F3 and the majority of F4 rabbits showed negative results. 7. By a continued observation of the rabbits for more than 120 days after birth, we found no spontaneous occurrence of diabetes (diabetes without giving diabetogenic reagents or taking any external means) in F1, F2 and F3 rabbits, not to speak of the normal rabbits. In F4 and F5 rabbits we found sure signs of spontaneous occurrence of diabetes. 18 of the 18 F4 rabbits and 7 of the 7 F5 rabbits fell to diabetes, between 61 to 425 days after birth.