Michael Somogyi

Blood Diastase as an Indicator of Liver Function

The method developed in this laboratory for the estimation of diastase in biological material satisfies 2 essential requirements: it yields accurate quantitative results, and is sufficiently sensitive to permit the determination of very small quantities of the enzyme. The quantity of the enzyme we express as the amount of reducing matter, in terms of glucose, which is produced by a known amount of the enzyme-bearing material under standardized conditions. As applied to blood, when we state that the diastase value of human blood serum is on an average 120, this means that 100 cc. of serum, incubated with 1.5% starch paste for 1/2 hour at 40°, produces a quantity of reducing matter which in regard to reducing capacity is equivalent to 120 mg. of glucose. The determination is actually carried out with 1 cc. of plasma (or serum), which is incubated with 5 cc. of starch paste and 2 cc. of a 1% NaCl solution for 30 minutes, and subsequently deproteinized by our copper method. The reduction value determined in the filtrate, minus the original sugar content of the plasma, represents the diastase value. From the normal average figure of 120 considerable deviations are found in either direction. The lowest value is, with very few exceptions, 80, the highest about 180. But, while individual variations spread over a considerable range, the blood diastase of one and the same individual shows a remarkably constant level; it is, moreover, largely independent of nutritional factors and does not change even over periods of months and years. During the past 2 1/2 years we have been running blood diastase determinations on nearly 100 healthy persons and several hundred hospital patients, in order to gather information in regard to possible deviations from the normal values.

Relationship Between Blood Amylase and Urinary Amylase in Man

Somogyi 1 has shown that the amylase content in the blood is rather constant for the individual, while the variations from individual to individual are considerable. The study of the amylase content of the urine, however, reveals great irregularity in the same individual at various periods of the day, without any apparent regularity in the variations. The analytical methods used were those described by Somogyi. 1 , 2 , 3 We found it very important to take into consideration the optimum pH (6.8-7.4) and the optimum salt concentration (0.25-0.4% NaCl). In healthy human beings the concentration of the enzyme in the urine is greater than in the blood, the ratio of urine amylase to blood amylase being usually between 2:1 and 6:1. These fluctuations in ratio occur in the same person, often in the course of a single day. Great as these variations are, we find that on the whole there is a parallelism between the blood and urinary amylase concentrations; i. e., high blood amylase usually goes with a high urinary amylase, and low blood amylase with a low urinary amylase. In pancreatic injury, as acute pancreatitis, obstruction of pancreatic ducts or trauma of the gland, there is a tremendous increase in both amylases, but without a shift in the relative concentrations. Since the urinary amylase remains high for a period of about 24 hours longer than the blood amylase, it may be preferable to determine the urinary amylase for the diagnosis of acute pancreatitis. In conditions where the blood amylase is low, as in severe toxemias of pregnancy, pneumonia, liver abscess, many cases of cholecystitis, obstructive jaundice and diabetes, 1 the urinary amylase is proportionately low. Both occasionally may be zero.

A Method for the Preparation of Blood Filtrates for Analysis.

The use of zinc salts for the precipitation of blood proteins furnishes a method of preparing filtrates in which fermentable sugar (glucose) is the only substance to reduce alkaline copper solutions. The reducing substances other than sugar—responsible for the “residual reduction” of fermented blood filtrates—are precipitated along with the proteins so that sugar determinations in zinc filtrates give true blood sugar values. In a simple method of deproteinization we employ the following 2 reagents: Reagent I. 12.5 g. of zinc sulphate (ZnSo47H2O) are dissolved in water, 125 cc. of 0.25 normal sulfuric acid are added, and made up with water to one liter. Reagent II is a 0.75 normal solution of sodium hydroxide. The 2 reagents have to be in such a definite relation that when Reagent I is titrated against Reagent II in the presence of phenol phtalein, 25 cc. of I should require 3.35 to 3.40 of II to produce a permanent pink color. Procedure: In order to prepare a 1:10 filtrate, lake one volume of blood in 8 volumes of Reagent I, add one volume of Reagent II, stopper tightly, shake well, filter after a few minutes through a dry filter paper. Centrifuge before filtration if a maximum yield of filtrate is wanted. For the preparation of filtrates of concentrations other than 1:10, we use a somewhat different technique and pair of reagents. Reagent A is a 19.5% zinc sulphate solution, Reagent B a normal sodium hydroxide. One volume of blood requires one half a volume of each of these reagents for deproteinization. If, for instance, a 1:5 filtrate is desired, lake one volume of blood in 3 volumes of water, introduce one half a volume of Reagent A, mix, then add one half a volume of Reagent B, stopper tightly, shake well, filter after a few minutes.

Hyperglycemic Response to Hypoglycemia in Diabetic and in Healthy Individuals

Summary 1. In numerous healthy individuals and normal experimental animals hypoglycemia entails hyperglycemia in the post-absorptive state, irrespective of the experimental conditions which initiated the hypoglycemia. The phenomenon is interpreted as the result of a delay in the adjustment between the reaction velocities of glycogen breakdown and glycogen formation in the liver. 2. In the diabetic the same physiologic process takes place on a greatly magnified scale. Hypoglycemias, caused by overdoses of insulin, entail in the diabetic patient excessive degrees of hyperglycemia and glycosuria. Recurrence of this sequel over considerable periods of time progressively increases the instability of the patient and aggravates the disease.

Changes in Ketonemia and Ketenuria During Hypoglycemia

Summary Hypoglycemia, either induced by hyperglycemia or occurring spontaneously (“hyperinsulinism”?), is frequently accompanied by ketosis (ketonemia and ketonuria). Increased rates of formation of ketone acids, known to occur in the liver, are coincident with increased rates of hepatic glycogenolysis.

Effect of insulin on carbohydrate tolerance of nondiabetic individuals.

One of us (Somogyi) 1 has shown that in diabetic patients the continued overdosage of insulin, i. e., the administration of quantities sufficient to cause hypoglycemias, gradually diminishes the carbohydrate tolerance of the patient and aggravates the diabetic condition. In this paper we call attention to the fact that insulin when administered to nondiabetic individuals, exerts a similar effect. This observation was made on nondiabetic tuberculous patients, who were treated with insulin with the purpose of inducing an increase in their body weight. The procedure in this work differed from others described in the literature in 2 respects. In the first place, the amount of insulin used was very small. In 2 cases only 10 units were given daily, injected in 5-unit doses before the noon and evening meals, while the remaining patients received at the beginning 10, and subsequently 20 units daily, always administered in 2 equal doses. In the second place, the duration of the treatment in this experiment was much longer than in others reported; the observations extended in 5 cases over a period of 6 months, and in one case over 12 months. In each case glucose tolerance test was performed a short time before the beginning of the treatment and a second test after insulin injections had been continued for approximately 6 months; in one case a third test was carried out after 12 months of uninterrupted treatment. In the tolerance tests uniformly 100 gm. of glucose was given orally and the blood sugar was determined at hourly intervals by the Shaffer-Hartmann method, using copper reagent No. 50 of Shaffer and Somogyi. 2 The blood samples were deproteinized with Somogyis zinc method 3 which removes nonfermentable reducing substances so that the analysis furnishes true sugar values (which are about 20 mg. % lower than the apparent sugar values obtained with the Folin-Wu method).

Diastase in milk.

Béchamp in 1883 was the first to recognize the presence of diastase in human milk; at the same time in cows milk he found no trace of this enzyme. 1 Bouchut 2 and Moro 3 , 4 , 5 confirmed the findings of Béchamp. During the last decade a number of investigators have attempted to establish a quantitative diastase test as a means of detecting whether or not a milk had been pasteurized. Namely, diastase would be entirely or partly inactivated during pasteurization, the extent of its destruction depending on the temperature and the duration of heating. The methods used by these workers, while claiming to yield quantitative results, are quite crude in comparison to the qualitative methods of Béchamp and Bouchut. The more recent workers find diastase in the milk of practically all the mammals 6 examined and are able to determine diastatic activity in the presence of lead 7 , 8 , 9 and even mercury salts. 10 The latter fact is characteristic of the unreliability of these methods. We approached the problem with analytical procedures, which in the instance of blood and urine proved to be adequate for the determination of very low as well as of high diastase values. The method is in brief as follows: a 1.5% starch paste is prepared of pure commercial corn- or rice-starch (but not of soluble starch); 10 cc. of this starch paste and 4 cc. of a 1% NaCl solution are measured into a test tube, the mixture is warmed to 40°, and then 2 cc. of diluted milk are added. The extent of the dilution (usually 1:10 to 1:40) depends on a preliminary test, which is based upon the amyloclastic activity of the milk.

Studies on blood diastase.

Diastase is known to be a normal constituent of human blood, but available information as to its quantity is inconsistent. The variety of measuring units and the multiplicity of methods render it impossible to correlate results and to explain conflicting conclusions. Even one and the same technique yields different results in the hands of different workers. One factor responsible for this situation is the use of soluble starch as substrate in measuring the enzyme action. It is practically impossible to prepare 2 identical batches of soluble starch, and consequently each will give different results with the same quantities of the enzyme, the discrepancies frequently being quite considerable. Moreover, investigators in general fail to consider the intricacies of the kinetics of diastase action and carry out their determinations under conditions where the amount of reaction products determined is not in linear proportion to the amount of enzyme. A study of substrates convinced us that starch pastes prepared from various refined natural starches by boiling at atmospheric pressure furnish adequate substrates. Pastes of 0.5 to 2% concentration, prepared from well washed rice, corn, wheat, potato or arrowroot starches, yield identical amounts of reducing sugars if incubated with identical amounts of enzyme under standardized conditions. It is remarkable that with glycogen the same results are obtained as with starch pastes. In another approach to the quantitative determination of diastase, in the measurement of the rate of cleavage of starch to the point where it no longer gives blue color with iodine, all starch pastes studied yield reproducible results, while every batch of soluble starch differs from the other. Thus the use of starch paste as substrate eliminates one variable in the study of diastatic activity.

Effect of the Fat Content of Diets on Blood Sugar

Summary 1. Diets containing high fat rations and restricted amounts of carbohydrates increase the postabsorptive (fasting) blood sugar level. 2. This change takes place in healthy and diabetic individuals alike, with the difference that it is much more accentuated and obvious in the diabetic organism. Thus, the physiologic processes in the healthy and in the diabetic individual are qualitatively identical, but in the diabetic they are greatly exaggerated (Claude Bernard); in other words, the difference between the normal and the diabetic individual, in regard to the physiologic process described, is only of degree but not of kind.

Changes in Blood Ketone Acids during Artificial Fever.

Summary Artificial hyperthermia in man increases ketosis. This, according to available evidence, is due to an increased rate of ketogenesis in the liver. Our observations on human subjects are in accord with the findings of previous investigators, who demonstrated on perfused liver, on liver slices and on living animals, that an inverse relationship exists between the glycogen content of the liver and the rate of ketone formation. The data presented indicate, moreover, that an increased rate of hepatic glycogenolysis per se suffices to increase ketosis, without any appreciable diminution of the glycogen reserves of the liver. An increased ketonemia in artificial fever can be entirely forestalled by the continuous injection of glucose so as to ensure the maintenance of blood sugar at hyperglycemic levels, a condition which enhances glycogen deposition in the liver and at the same time prevents an increase in the rate of glycogenolysis.