ON THE NATURE OF THE URINE SUGARS. BY H. S. EAGLE. (From the Chemical Division, Medical Clinic, the Johns Hopkins University, Baltimore.) (Received for publication, November 3, 1926.) INTRODUCTION. In spite of an extensive literature, and numerous methods of attack, the nature of the urine sugars is still in doubt. Following the early work of Baisch (I), Pavy and Siau (22), Breul (S), and others, it had seemed clearly established that the normal daily output of urine contained from 0.3 to 1.9 gm. of non-nitrogenous reducing substances, of which glucose constituted from 30 to 60 per cent. That glucose was present seemed definitely proven by the optical rotating powers of urine, by its fermentat,ion with yeast, with the production of carbon dioxide, by the formation of benzoyl glucose when urine was shaken with benzoyl chloride, and the formation of an osazone in all respects identical with glucosazone (1). The close agreement in the value for glucose as obtained by fermentation and polariscopic determinations seemed conclusive. That the urine does contain non-nitrogenous reducing sub- stances simulating sugars, is recognized by almost all recent workers (5, 11, 13, 15, 16, 18, 21). The concentration of these substances has been found to vary from 0.01 to 0.2 per cent, representing a daily excretion of from 0.25 to 1.6 gm. Whether even a portion of this is true glucose, however, has become very doubtful. For the optical rotation measured by Baisch (1) is of very small magnitude, and in view of the complexity of urine, it is impossible to ascribe such rotation to glucose. H&t (15) has been unable to form glucosazone by heating urine with phenyl- hydrazine, using a method sensitive to 0.05 per cent of glucose. Osazones are formed, it is true, but they differ in size, shape, 481 by guest on May 25, 2018 http://www.jbc.org/ Downloaded from
Transcript
ON THE NATURE OF THE URINE SUGARS.
BY H. S. EAGLE.
(From the Chemical Division, Medical Clinic, the Johns Hopkins University, Baltimore.)
(Received for publication, November 3, 1926.)
INTRODUCTION.
In spite of an extensive literature, and numerous methods of attack, the nature of the urine sugars is still in doubt. Following the early work of Baisch (I), Pavy and Siau (22), Breul (S), and others, it had seemed clearly established that the normal daily output of urine contained from 0.3 to 1.9 gm. of non-nitrogenous reducing substances, of which glucose constituted from 30 to 60 per cent. That glucose was present seemed definitely proven by the optical rotating powers of urine, by its fermentat,ion with yeast, with the production of carbon dioxide, by the formation of benzoyl glucose when urine was shaken with benzoyl chloride, and the formation of an osazone in all respects identical with glucosazone (1). The close agreement in the value for glucose as obtained by fermentation and polariscopic determinations seemed conclusive.
That the urine does contain non-nitrogenous reducing sub- stances simulating sugars, is recognized by almost all recent workers (5, 11, 13, 15, 16, 18, 21). The concentration of these substances has been found to vary from 0.01 to 0.2 per cent, representing a daily excretion of from 0.25 to 1.6 gm. Whether even a portion of this is true glucose, however, has become very doubtful. For the optical rotation measured by Baisch (1) is of very small magnitude, and in view of the complexity of urine, it is impossible to ascribe such rotation to glucose. H&t (15) has been unable to form glucosazone by heating urine with phenyl- hydrazine, using a method sensitive to 0.05 per cent of glucose. Osazones are formed, it is true, but they differ in size, shape,
color, and arrangement from glucosazone. The formation of carbon dioxide upon incubating urine with yeast is negatived by Lund and Wolf (16).
Recently, however, Benedict (6) has found that the ingestion of carbohydrates is followed by an increased sugar output in the urine, and that the consumption of glucose causes a corresponding increase in the fermentable fraction of urine sugar. He has therefore suggested that the term glycosuria, implying the sudden appearance of a sugar hitherto absent, be replaced by the term glycuresis, signifying merely a quantitative change, an increase in sugar excretion. It is evident that he considers glucose nor- mally present in urine, and glucuresis, an increase in glucose excretion following its ingestion, to be a physiological occurrence, a specific example of the more general term glycuresis. Accord- ingly, the sugar in diabetic urine would represent, not a pathologi- cal breakdown in an absolute tolerance, but only a quantitative exaggeration of a preexisting normal excretion. To safeguard the kidney, he proposes prophylactic limitation of carbohydrate intake even in normal people.
The validity of this conception has been greatly discussed within the last few years. Folin and Berglund (11) have taken a diametrically opposed position: there is a true glycuresis after every carbohydrate meal, consisting entirely of foreign carbo- hydrates present in grains, nuts, and fruits, and of decomposition products formed in process of preparing food. Glucose is nor- mally not present in urine, and appears only when a certain critical level of tolerance has been exceeded. Glucose ingestion up to 200 gm. fails to break down this barrier and does not cause a perceptible increase in the excretion of urine sugar. In view of these findings they are inclined to discount the significance of yeast fermentation. Benedict (4), in an analysis of their data, calls attention to a slight increase in the excretion of urine sugars following glucose intake, and points out that such an increase is proof for the presence of glucose normally, and the validity of his concept of glycuresis. But H&t (15) as well as Greenwald, Gross, and Samet (13) also finds that the ingestion of glucose causes no increase in t,he excretion of urine sugar and both conclude that the so called urine sugars represent by products of met’abolism. Yet Blatherwick and her coworkers (7) confirm Benedict’s find-
ings, and ascribe the non-fermentable fraction of urine sugar to by products of protein metabolism.
The results of fermentation experiments are no less confusing, to a great extent due to lack of uniformity in method. Shaffer and Hartmann (23) report no decrease in sugars following in- cubation with yeast, but give no details as to method. Green- wald, Gross, and Samet (13) obtain variable results upon incubat- ing at 37-38” for 24 hours, and place no reliance in the results obtained by fermentat,ion methods. Most workers find a 20 to 70 per cent decrease in amount of reducing substances upon incubation with yeast, a decrease which is usually ascribed to the presence of glucose. Yet as already noted, Lund a.nd Wolf (16) are unable to demonstrate formation of carbon dioxide, and H&t finds that while glucose is fermented under his conditions in 24 to 48 hours, the decrease in urine sugars is not completed in less than 48 to 72 hours. And very recently, Patterson (21), by the addition of toluene to the urine inhibits what she takes t,o be a bacterial decomposition, and obtains no fermentation under conditions in which glucose is completely fermented within 17 hours.
I. Scope qf Present Investigation.
The present experiments grew out of some work upon blood sugar fermentation, in the course of which we were able to confirm the finding of Hiller, Linder, and Van Slyke (14) that, using proper methods, the fermentation of blood sugar by yeast is completed within 20 minutes. In the light of this finding, it was somewhat surprising to note that as far as could be ascertained, none of the previous work upon urine had involved ferment,ation for less than 17 hours, many workers fermenting urine for as long as days, by which time bacterial decomposition is quite evident. As will appear, we were able to show that any glucose present in urine might, under proper conditions, be completely fermented within 40 minutes. Using this method as the’ basis for glucose determinations, the effect of foods and especially of glucose upon the excretion of urine sugars was then investigated.
In these determinations, Benedict’s recent copper and tung- state reagents (2) were used. They were found to yield consistent results over a wide range, the values for urine sugar being slightly
lower than those obtained by using the Folin-Wu reagents (12). The green color of the blank which makes their use impossible in blood fermentation, here, because of the lesser dilution (1:4 instead of 1: 10) makes readings difficult only for solutions con- taining less than 0.01 per cent sugar. The new alkaline copper tartrate and molybdic acid reagents suggested by Folin (9) for use in blood and urine were found to be useful only for blood sugar determination. In making determinations of urine sugar by this method, fully 20 to 30 per cent of sugar added to urine is not recovered. An extensive series of determinations has shown that the loss is due to some substances dissolved from the Lloyd’s reagent use’d, substances which, as Benedict has suggested, probably remove hydroxides and carbonates from solution when the alkaline reagent is added.’
II. Fermentation of Glucose in Pure Solution.
To each of a series of flasks were added 5 cc. of glucose solutions of varying concentration and 3 cc. of the yeast suspension described by Hiller and Van Slyke in blood sugar fermentation;2 the mixture was shaken and incubated at 3738” for var.ying intervals. Upon withdrawal from the incubator, each mixture was diluted to 20 cc. and shaken with Lloyd’s alkaloidal reagent ih order to secure a water-clear filtrate. Determinations were made upon 2 cc. of filtrate, using appropriate standards. The results are tabulated in Table I.
It is quite evident that glucose within the physiological range of concentration is completely fermented within 40 minutes. Two factors are of great importance in determining the speed of fermentation, and explain the failure of previous workers to obtain complete fermentation in less than 17 hours. The first of these is the use of inadequate quantities of yeast. The dilution used in
1 In his last communication, published after this study had been com- pleted (Folin, O., and Svedberg, A., J. Bid. Chem., 1926, lxx, 405), Folin suggests a method of overcoming this difficulty.
2 One cake of Fleischmann’s yeast suspended in 20 cc. of water. There are thus obtained 32 cc. of suspension. The water content of a cake of yeast
being 9 cc. (14, 17), 3 cc. of the suspansion represent & X (20 + 9) = 2.72 cc. of water. In the final dilution of 20 cc. this represents a percentage error of 1.4 per cent, which is well within limits of accuracy of the method.
the foregoing experiment corresponds to 16.4 per cent by weight of yeast cake.
None of the earlier workers (1, 8, 22) gives the exact quantity of yeast used. MacLean (17) simply mixed urine well with yeast. Oppler (20) used 1 cc. of yeast to 100 cc. of urine-a maxi- mum of 0.63 per cent by weight of yeast cake. Host (15) added yeast to 5 per cent, and Patterson (21) to 0.5 per cent. Benedict (3) and most of the subsequent workers (7, 13, 19, 24) have used quite sufficient quant,ities; viz., t cake of yeast to 20 cc. of urine, 17.5 per cent of yeast by weight. It is difficult to see why none of the workers who have been using yeast in quantities suggested by Benedict, a quantity sufficient for rapid fermentation, has commented on the discrepancy in the relative speeds of fer- mentation of glucose and of urine.
TABLE I.
Glucose is completely fermented in 40 minutes at 37-38”.
But just as important as the amount of yeast used is the volume of fluid incubated. For, within 10 to 15 minutes after being shaken, most of the yeast has settled as a compact layer in the bottom of the flask, covered by a relatively slightly cloudy sus- pension. The bulk of fermentation takes place in the lower layers of fluid to which yeast has gravitated. Obviously, in the large volumes of urine usually used, doubling the quantity of yeast will not speed up fermentation as much as halving the volume of fluid.
III. Fermentation of Urine Sugar.
When the method used for pure glucose solutions was applied to urine, however, difficulty was encountered. 5 cc. of urine were shaken with 3 cc. of the yeast suspension and.incubated at
3738”. Following incubation there were added 7 cc. of water and 5 cc. of 0.1 N H&SOh, and the mixture shaken with 1.5 gm. of Lloyd’s alkaloidal reagent for 2 minutes, a procedure suggested by Folin for the removal of most of the interfering substances in urine (10). As appears in Table II, far from an immediate decrease, there is a gradual increase in the amount of reducing substances extending over several hours, followed by an even slower decrease over several days. It is this decrease which most workers have ascribed to a true fermentation of glucose. Yet, confused as results were by this initial increase due to a puzzling
TABLE II.
Incubation of urine with yeast leads to slow formation of reducing substances, obscuring the fact that glucose is completely fermented within 30 to 40 minutes, with a subsequent slow decrease over 48 hours.
Reducing substances. 20 min. 40 min. 2 hrs. 24 hrs. 48 hrs.
reaction between urine and yeast, it was apparent that any glucose added to urine was completely fermented within 40 minutes, as shown in Table II.
It was obviously impossible to decide upon the presence or absence of glucose until this extensive formation of reducing substances could be eliminated. A pH to either side of the neutral point of phenolphthalein retarded, but did not prevent, the urine-yeast reaction. It was finally found that if the urine was shaken with Lloyd’s reagent preliminary to fermentation, those substances in the urine responsible for the confusing reaction were entirely removed.
To 5 cc. of urine were added 10 cc. of water and 5 cc. of 0.1~ sulfuric acid and the mixture shaken with 1.5 gm. of Lloyd’s reagent (Folin (10)). To 10 cc. of the 13 to 14 cc. of filtrate thus obtained were added 3 cc. of yeast suspension and the whole incubated at 37-38”. Determinations were made upon 2 cc. of filtrate from the mixture. As before, the addition of 0.5 gm. of Lloyd’s reagent prevents the passage of yeast through the filter paper. To allow for the dilution with yeast suspension, the results obtained are multiplied by 1.27 (3 cc. of yeast suspension = 2.72 cc. of water).
TABLE III.
No fermentation occurs in the filtrates from urines shaken with Lloyd’s alkaloidal reagent, although glucose added is quantitatively fermented within 40 minutes.
It is evident from Table III that the preliminary shaking with Lloyd’s reagent has removed from urine whatever substance was responsible for the production of reducing substances upon incuba- tion with yeast. There is now found an extremely slow decrease in apparent value of urine sugar over a period of days, a decrease which cannot be attributed to glucose fermentation, for any glucose added to urine is quantitatively fermented within 40 minutes. Yet the same urines which, after treatment with Lloyd’s re- agent, fail to show any fermentation in 40 minutes, will, when incubated directly with yeast, show a decrease of 20 to 70 per cent in 24 hours, as shown in Table IV. Such a decrease is,
therefore, not due to glucose fermentation, occurs even in the Lloyd’s filtrate on prolonged incubation, and, as has been suggested by recent workers, is possibly due to bacterial decomposition.
IV. E$ect of Ingestion of Pure Glucose upon Urine Sugar Excretion.
100 gm. of glucose in 500 cc. of water, flavored with coffee to prevent nausea, were given to eight patients, the last previous meal having been eaten on the preceding night, and blood obtained as in tolerance test, after 3 hour, 1 hour, and 2 hours. Urine was obtained just before glucose ingestion and every half hour
TABLE IV.
Changes in reducing value of urine upon incubation with yeast over long intervals do not represent true fermentation of glucose.
* Figures represent mg. of reducing substances per 100 cc. of urine.
thereafter over several hours. To facilitate diuresis, 60 cc. of water were given by mouth every half hour. Of the subjects two were known diabetics, two were normal, and four were patients with chronic arthritis, none of whom, so far as was known, had any difficulty in metabolizing carbohydrates. Of the arthritic patients, two had normal blood sugar curves, and two showed examples of the so called “prediabetic” curve. The results are tabulated on the following pages.
In all the normal (and arthritic) patients, the fasting urine specimen was absolutely free of glucose. In some of these it
remained so throughout the experiment (Table V). In others (Table VI) no sugar appeared until a certain critical level of
TABLE V.
No appearance of glucose in urine after ingestion of 100 gm. of glucose by arthritic subject.
Blood. I Ud ne
Time.
9.30 10.00 10.30 11.00 11.30 12.00
per cent cc. per cent per cent 0.082 9.25 68* 0.084 0.082 0.153 9.55 30 0.095 0.097 0.183 10.25 30 0.053 0.050 0.148 10.55 210 0.018 0.018
11.25 120 0.020 0.019 0.090 11.55 125 0.019 0.020
Total. Non- FtP
ermcnt~ ment- able. able.
per cent
Sugar. -
fl
-
-
Non- ,rment.- able.7
mg.
28.5 28.5 15.9 37.8 24.8 23.8
ab1e.t
* 2 hours excretion. t Mg. excreted per half hour. Glucose ingested at 9.31.
TABLE VI.
Temporary overflow of glucose into urine after ingestion of 100 gm. of glucose in normal subject.
Blood. T Time. sugar.
per cent 10.16 0.082 10.46 0.126 11.30 0.102
12.16 0.090
Time.
10.19 10.49 11.19 11.49 12.19
* Excretion per half hour. t 3 hours excretion. Glucose taken at 10.17.
Urine.
Volume. Total.
___- cc. per cent
200t 0.020 33(?) 0.022 48 0.103 48 0.033 76 0.020
f
-
-
Sugar.
NOW erment-
able.
per cent
0.021 0.020 0.026 0.021 0.022
Fer- NOW ment- able. fe;?c~
per cent nag.
6.4 7.2
0.075 12.5 0.012 10.8
16.8
_
-
Fer ment- able.’
mg.
36 5.8
blood sugar was reached, whereupon there was a sudden appear- ance of fermentable sugar in the urine, in some cases in large quantities, in others only in traces. Following its initial appear-
ante, the excretion of glucose continued even after the blood sugar had fallen below its original critical level, but in all the patients, sugar had disappeared from the urine by the 3rd hour.
In the diabetic patients, glucose was found in urine even at the fasting level of blood sugar as shown in Table VII. It is to be remembered, however, that this is not necessarily true of dia- betics, some of whom develop even an increased tolerance for sugar.
In most of the patients tested there was noted, during t,he 2nd hour after ingestion of glucose, an increase in the excretion of non-fermentable carbohydrate. Whether this is due to varia-
TABLE VII.
Constant overflow of glucose into urine before and after ingestion of 100 gm. of glucose by a diabetic patient.
* Excretion per half hour. Glucose taken at 10.02.
tions in the rate of urine secretion which become evident on taking specimens over short intervals of time, whether it is due to for- eign carbohydrates supplied by the coffee used to flavor the sugar solution, or whether the mobilization of carbohydrates in blood affects the non-fermentable as well as the fermentable fraction, we are unable to say.
It is this increase which has led Benedict to believe that glu- cose is normally excreted in quantities which may be increased by it,s ingestion. A faulty method of fermentation apparently confirmed the fact. It has been seen, however, that the increase is altogether in the non-fermentable fraction, and can by no means be ascribed to glucose excretion.
V. Effect of Commercial Glucose and an Ordinary Meal upon Excretion of Urine Sugar.
Objections have been raised (4) to such experiments as the foregoing that the shock caused the organism by the ingestion of very large quantities of carbohydrates may well prevent a glu- curesis which would occur from the feeding of more moderate quantities of glucose. To test this, two normal individuals took small amounts of commercial glucose (chocolate) as the first food taken in the morning. The figures do not include volume of excretion, but the qualitative results seem definite (Table VIII).
TABLE VIII.
Ingestion of commercial glucose by normal subjects results in an in- creased excretion of non-fermentable reducing substances, but not of glu- cose unless tolerance has been exceeded.
a. Fasting. b. 10 gm. of glucose.
a. Fasting. b. 30 gm. of glucose..
a. Fasting. 0.036 b. 20 gm. of glucose.. 0.053
a. Fasting. ,. b. 85 gm. of glucose..
0.045 O.l!O
-. Total.
per cent
0.019 0.028
per cent
0.020 0.027
0.033 0.036 0.085 0.098
0.034 0.051
0.047 0.090
NOW fermentable. Fermentable.
per cent
0 025
a, collected 15 minutes before taking glucose. b, collected approximately 13 to 2 hours after taking glucose.
In only one case did glucose at any time appear in the urine, and this was after the ingestion of a quantity of glucose which may well have exceeded the tolerance of that individual. The marked increase in the concentration of non-fermentable reducing substances after the ingestion of the chocolate is too uniform to be an accident of secretion and is probably due to the non-glucose carbohydrates present in the material ingested.
In six normal individuals, following an ordinary mixed meal including bread, meat, milk, potatoes, and fruit in varying quanti- ties and combinations, there was a marked increase in the excretion of non-glucose reducing substances, both in concentration and total output. But in no case was fermentable sugar present either before or after the ingestion of food (Table IX).
The threshold tolerance for glucose, indicated by the foregoing experiments, is not peculiar to the fasting state. The ingestion of commercial glucose and of a miscellaneous meal by non-fasting individuals approximately 5 hours after the preceding meal, gives results qualitatively similar to those obtained in fasting subjects.
SUMMARY AND CONCLUSIONS.
By incubating urine at 37-38“ with appropriate quantities of yeast, any glucose present up to 0.4 per cent may be quantitatively fermented within 40 minutes. The presence of glucose in con- centrations as low as 0.010 per cent may thus be determined, provided certain substances in urine which obscure the results of fermentation are first removed. Lloyd’s alkaloidal reagent, suggested by Folin for use in determination of urine sugar, has been found to be useful also in this regard.
Using this method, it is found that glucose is not normally excreted in the urine. What has heretofore been considered to be glucose, fermentable by yeast, is in reality a group of sub- stances which only gradually decompose under the conditions of fermentation, possibly due to bacterial decomposition. Such decomposition is a matter of days, while the fermentation of glucose is completed in less than an hour. It is these substances which are partially removed by the preliminary shaking of urine with Lloyd’s reagent.
Until a certain critical level of blood sugar has been reached, the kidney interposes an absolute barrier against the excretion of glucose. The ingestion of moderate quantities of glucose by the normal individual fails to exceed this threshold value, and is accordingly not followed by the appearance of even traces of glucose in the urine. It is only when such large quantities as 100 gm. of glucose are taken that this critica, level of blood sugar is exceeded in a certain proportion of normal individuals, and then only does fermentable sugar appear in the urine. Once
begun, however, the excretion of glucose continues even after the blood sugar has receded to below this level.
It is quite doubtful whether the normal individual, on an average diet, ever shows true glucose in the urine. The increase in urine sugar following food (and glucose) intake, discovered by Benedict .and termed by him glycuresis, represents, not glucose, but non- fermentable substances. In the case of food it is probably true, as has been suggested, that t,hese are either foreign non-assimilable carbohydrates originally present in the food, or decomposition products formed in the process of preparing food or of its digestion. The increase in their excretion following the ingestion of glucose is somewhat difficult to understand, and may be due to any of several factors suggested: variations in the rate of secretion, a trace of impurities in the glucose used, or a mobilization of these substances in the blood coincident with the rise in blood sugar.
I wish to express my thanks to Dr. W. A. Perlzweig, whose kind advice and cooperation at every turn have made this investigation possible.
2. Benedict, S. R., J. Bid. Chem., 1925, lxiv, 207. 3. Benedict, S. R., and Osterberg, E., J. Biol. Chem., 1918, xxxiv, 195. 4. Benedict, S. R., and Osterberg, E., J. BioZ. Chem., 1923, Iv, 769. 5. Benedict, S. R., and Osterberg, E., J. Biol. Chem., 1921, xlviii, 51. 6. Benedict, S. R., Osterberg, E., and Neuwirth, I., J. Biol. Chem., 1918,
xxxiv, 217. 7. Blatherwick, M. R., Bell, M., Hill, E., and Long, M. L., J. Biol. Chem.,
10. Folin, O., and Berglund, II., J. BioZ. Chem., 1922, li, 209. 11. Folin, O., and Berglund, H., J. BioZ. Chem., 1922, Ii, 213. 12. Folin, O., and Wu, H., J. BioZ. Chem., 1920, xli, 367. 13. Greenwald, I., Gross, J., and Samet, a., J. BioZ. Chem., 192425, lxii, 401. 14. Hiller, A., Linder, G. C., and Van Slyke, D. D., J. BioZ. Chem., 1925,
lxiv, 625. 15. H&t, H. F., J. Mctabol. Research, 1923, iv, 317. 16. Lund, G. S., and Wolf, C. G. L., Biochem. J., 1926, xx, 259. 17. MacLean, H., Biochem. J., 1907, ii, 431.
18. Myers, V. 6., Proc. Sot. Exp. Biol. and Med., 1914, xvi, 13. 19. Neuwirth, I., J. Biol. Chem., 1922, Ii, 11. 20. Oppler, B., Z. physiol. Chem., 1911, lxxv, 71. 21. Patterson, J., Biochem. J., 1926, 43. xx, 22. Pavy, F. W., and Siau, R. L., J. Physiol., 19oo-01, i, 26, 282. 23. Shaffer, P. A., and Hartmann, A. F., J. Biol. Chem., 192G21, XIV, 365. 24. Sumner, J. B., J. Biol. Chem., 1921, xlvii, 5.
