ON THE ACTION OF PHLORHIZIN ON THE KIDNEY. By E. B. MAYRS. (From the Department of Pharmacology, Edinburgh.) GLUCOSE is the most important diffusible substance in the blood which is completely held back by the normal kidney. On Cushny's theory(l) of kidney action, the glomerulus and capsule are completely permeable to glucose, but the glucose does not pass into the urine because it is all absorbed by the tubule cells. Strong evidence for this view is afforded by Wearn's demonstration (2) that in the frog injection of glucose is followed by the appearance of a reducing substance in the glomerular filtrate, though none may be present in the urine. A similar conclusion is arrived at by Clark(3) from perfusion experiments on the frog's kidney. Further, in the albuminuria occurring in renal disease, proteins can sometimes be shown to have passed through B ow ma n's capsule (4), and yet no glucose appears in the urine. It is difficult to see how the highly diffusible glucose can in such cases fail to reach the urine unless it is absorbed by the renal tubules. The diffusibility of sugar would also lead us to expect that injury to the kidney might cause the appearance of glucose in the urine, but the only condition in which this occurs from renal changes is under poisoning with phlorhizin, in which there is reason to believe that the tubule cells are the site of action; if this consists in hindered absorption by them, the concentration of glucose by the kidney after administering phlorhizin might approach that of a "no-threshold" substance (1), but could not exceed it. It seems desirable then to determine how far this occurs, and I have made some experiments on rabbits. Sulphate was chosen as the standard "no-threshold" body, in preference to employing the normal urea of the blood; because in the rabbit sulphate (which is practically non-toxic) is more completely "no-threshold" in character than urea (5); and intravenous injection of sodium sulphate has the additional advantage of ensuring the diuresis necessary for short experiments. A similar method of investigation could be used for human subjects, and for that purpose urea, creatinine, or phosphate could be selected as the standard for comparison, and no injection would be necessary. The rabbits were given varying amounts of phlorhizin subcutaneously, in warm saline or dilute alcohol. The quantity injected was always fairly
462 E. B. MA YRS. large, in order to produce the maximum effect; while in some cases several doses were given, the treatment extending over 18 to 48 hours. One rabbit had injections of phlorhizin for 14 days, and a typical ketosis resulted, with acetone and diacetic acid in the urine; the blood sugar was exceptionally high. The last dose was always given a short time (I to 2 hours) before the ansesthetic (urethane). The experiments were controlled by observations of the excretion of glucose by the normal kidney which often occurs to a slight extent in aneesthesia, and also by observing the efficiency of the kidney in dealing with injected glucose. In spite of the rapid loss of sugar caused by phlorhizin the level of blood sugar was usually not low, but the pronounced hyperglyeamia of narcosis was prevented. This is illustrated by the following results, in which the plasma concentrations of glucose are given. Only the third rabbit got phlorhizin. Normal plasma I hour after 0-2 grnm. 1-1k hours after phlorhizin per kilo anesthetic 1 *11% -.30% 2 *13% -30%O 3 *13%.09% *12% It was not the purpose of this investigation to determine the maximum urine concentration of glucose which could be produced by the drug, since this is governed by the rate of excretion of fluid; but to observe the relative extents to which glucose and sulphate were concentrated in their simultaneous passage through the kidney. This rendered it necessary to estimate glucose and sulphate in the plasma and urine, in order to compare their urine/plasma ratios, and also made it desirable to curtail the period of observation as much as possible (3-7 minutes), in case the plasma concentrations of the two substances should be varying at different rates. The actual percentages of glucose in the urine were always fairly low because of the diuresis. Sulphate was estimated by the benzidine method, enough being injected to justify the application of this procedure to incinerated plasma; and glucose by the method of Folin and Wu (6) which was suitable for urine as well as plasma, since the fluid excreted could not in any of the experiments have contained enough creatinine or other reducing substances to interfere with the accuracy of the results, and could be diluted to about the same glucose concentration as the standards. The figures obtained are given in the accompanying table. The last column shows the extent to which sulphate is concentrated in relation to glucose, and the two preceding columns (from which this relation is calculated) show the concentration ratios of these substances, that is, the number of times their concentrations are increased in their
PHLORHIZIN AND KIDNEY ACTION. 463 Last Intra- Conc. Total dose venous Vol. of phlor of injection. of 0/ 0/0 0/0 0/0 Urine Urine S04 by Wt. of hizin phlor- c.c. of urine. Na2904 glucose Na2SO4 glucose Ns2SO4 glucose/ kidney, rabbit, given. hizin. 100/0 c.c. per in in in in Plasma Plasma glueose No. kilos. grms. grms. Na2804 minute plasma plasma urine urine Na2SO4 glucose =1 1 2.3 nil - 23'0 1.50 '452 *29 2'558 *01(?) 5-66 '03 2 1X7 nil - 10.0 2-20 *347 '44 1'476 *41 4'25 93-3 2.4 nil - 23.7* 3'67 *369 '55 1-023 1-04 2'77 1'89 1'47 4 2-0 nil - 20.5t 1'77 '304.50 1'374 1'43 4'52 2'86 1'58 5 2'1 *210 *210 21'0 2'83 *285 *13 1-386 *24 4'86 1'85 2'63 6 2'3 '460 *460 11iS.50 *244 *17 2'047 1'10 8-43 6'47 1'30 7 2-0 *820 *820 20-0 2'27 *367 a1 1'214 *34 3'31 227 1-46 8 1'9 1'122 '748 18-7 *86 *403 *20 1'210.61 3'00 3'05 *98 9 1-4 *852 *568 14-2 *48 *224 *17 1-095 *58 4'89 3-41 1'43 10 2'2 1'800 '900 22'5 2'47 *631 *10 1'833 *24 2'90 2'40 1'21 11 2-0 1'600 1-200 40 0 '33 1-374 *29 2'915 '44 2'12 1'52 141 12 2-1 1'300 1300 21'0 1'10 *583 *22 2'340 '65 4'01 2-95 1'36 13 1-6 1-400 '200 250 '90 '724 '45 2-547 '83 3'52 1-84 1'91 * Containing 2'37 grms. of glucose. t Containing 2'05 grns. of glucose. passage from plasma to urine. Many of the rabbits showed a fairly high percentage of glucose in the plasma, due no doubt to the effect of the anesthetic, and this is naturally most noticeable in the first two experiments when no phlorhizin was given. The last rabbit in the series was one to which small doses (0.1 grm.) of the drug had been given for about a fortnight, and its high blood sugar is somewhat remarkable. A secondary defensive mechanism may have been brought into action. As an average of the results, sulphate seems to be concentrated about 1'37 times as well as glucose even after the maximum phlorhizin effect on the kidney has been obtained; the limits of variation being shown in Exp. 8 ('98) and Exp. 13 (1.91). It seems clear that in Exp. 5 when only 0.1 grm. per kilo was given the maximum action was not produced. Increasing the dose above 0X2 grm. per kilo does not appear to cause any corresponding increase in the effect of the drug. It is evident, therefore, that if concentration by the kidney is brought about by absorption of water from glomerular filtrate, some glucose must be absorbed also even under phlorhizin; for otherwise the urine/plasma ratio of glucose could not be less than that of sulphate. Hence, phlorhizin does not, as a rule, completely prevent the re-absorption of glucose. If the sulphate ion is entirely unabsorbed, then the greatest theoretical action that phlorhizin could exert would equalise the concentration ratios of glucose and sulphate, and it is interesting to observe that this effect has actually been produced in Exp. 8. Cushny's hypothesis demands that glucose shall never be concentrated to a greater degree than sulphate, which is a "no-threshold" substance; and in no experiment of this series has such a result been obtained. In Exps. 3 and 4, an intravenous injection of 1 grm. of glucose per kilo
464 E. B. MA YRS. has produced sufficient hyperglycsamia (about 0.5%) to make the kidney excrete glucose almost as effectively as if it were under the influence of phlorhizin; though with less diuresis and more time for re-absorption this might not have occurred. The narcotic hyperglyeemia of the first two experiments has not had this effect. It may be simply a question of degree, or perhaps part of the normal blood sugar exists in a form different from that of injected glucose(7), which may act to some extent as a foreign body. Previous investigations have shown that urea is less efficiently concentrated than sulphate by the rabbit's kidney(5). When a mixture of these substances is excreted the urine/plasma ratio of sulphate is higher than that of urea. It was found to be from 1.50 to 2*75 times as high in different experiments, and the average was 2*01 times. It is therefore necessary for the stability of the absorption theory to suppose that some urea is re-absorbed from the tubules by their lining epithelium, and this point should be considered in deciding the value of renal efficiency tests based on the concentration of urea by the kidney. Since the response of the kidney to phlorhizin has also been used as a test for renal efficiency, it is of interest to see whether there is any theoretical advantage in employing this method. In pathological conditions of the kidney urea is often held back until it reaches a high level in the blood, and in these cases its concentration in the urine is low. The efficiency of excretion may be judged by comparing the percentages of urea in the plasma and urine, and observing the extent of their departure from the normal. On the absorption hypothesis urea retention is probably not due to inability of urea to reach the tubules, since the larger molecule of albumen can pass through the glomerular membrane, but to inability of the damaged tubule cells to avoid re-absorbing some of the urea in an attempt to concentrate it by removing water against its osmotic resistance. Phlorhizin has been found to produce less glycosuria in cases of nephritis(8), and has been used as an efficiency test on the ground that its effectiveness in causing excretion of blood sugar depends on the integrity of the renal epithelium. This is analogous to the retention of urea. After phlorhizin glucose should behave in much the same way as urea, and if the latter can diffuse back from the tubules into the blood the same assumption may be applied equally to glucose. In order to demonstrate the superiority of the phlorhizin test it would be necessary, in the first place, to show that when acting on the normal kidney phlorhizin can render glucose more completely " no-threshold" in character than urea ordinarily is; and, secondly, to show that a dose
PHLORHIZIN AND KIDNEY ACTION. 465 large enough to produce a fairly constant effect in health would not be injurious to the subject of the test. Evidence has been obtained that a rabbit fully under the influence of phlorhizin can excrete glucose slightly better than urea. The concentration ratio of urea was determined in Exp. 12, and was found to be 2-45. The corresponding glucose concentration ratio was 2-95. This difference is not of much importance, and as the normal human kidney is probably better able to concentrate urea than that of the rabbit, one cannot conclude that there is any material advantage in using the phlorhizin test. As regards the question of dosage it can only be said that 0 1 grm. per kilo subcutaneously was not found sufficient to produce the maximum effect in the rabbit, and that this dose is relatively very much larger than has been given to human subjects undergoing a renal efficiency test. Hence, unless the drug is given in excessive amounts, one cannot be quite sure that the kidney is in all cases under its influence to the same degree; and if too small a dose were employed, even the healthiest kidney could not be expected to concentrate glucose very much. SUMMARY. A procedure has been described for measuring the effect of phlorhizin on the kidney of animals. It is suggested that a similar method might be used for determining the efficiency of the human kidney in excreting any constituent of the urine; provided that one of the "no-threshold" substances is known to be excreted in a normal manner. The method described avoids uncertainty as to whether the concentrations observed have been influenced by diuresis, since both the standard and the substance being examined are affected to the same extent. The nature of the renal mechanism for the retention of glucose has been discussed, and it has been shown that phlorhizin does not usually enable the kidney to excrete glucose quite as effectively as " no-threshold" substances. The use of phlorhizin for renal efficiency tests has been criticised, and the dose usually given has been shown to be quite inadequate. All the results obtained are in accord with the absorption theory of the excretion of urine. The expenses of this research have been defrayed for the most part by a grant from the Moray Fund of Edinburgh University.
466 B. B. MA YRS. 0 REFERENCES. (1) Cushny. The Secretion of Urine. 1917. (2) Wearn. Amer. Journ. Physiol. 59. Proc. 490. 1922. (3) Clark. This Journ. 56. p. 201. 1922. (4) Posner. Virchow's Arch. f. Path. Anat. 79. p. 311. 1880. (5) Mayrs. This Joun. 56. p. 58. 1922. (6) Folin and Wu. Journ. Biolog. Chem. 41. p. 367. 1920. (7) Winter and Smith. This Journ. 57. p. 100. 1922. (8) Weber. Arch. f. exp. Path. u. Pharm. 54. p. 1. 1906. CAMBRIDGE: PRINTED BY W. LEWIS AT THE UNVERSITY PRESS