0021 -972X/91/7303-0564$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1991 by The Endocrine Society
Vol. 73, No. 3 Printed in U.S.A.
Insulin Effects on Glucose and Potassium Metabolism in Vivo: Evidence for Selective Insulin Resistance in Humans PINCHAS COHENf, NIR BARZILAI, AMIR LERMAN, HAVA HAREL, PEDRO SZYLMAN, AND EDDY KARNIELI Metabolic Unit, Endocrine Institute, Department of Medicine C, and Department of Nephrology, Rambam Medical Center and Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel
ABSTRACT. The effects of insulin on in vivo glucose use and potassium uptake in healthy humans are well documented. However, the interrelationship between these two processes is not fully defined. In order to characterize it, we have used the euglycemic clamp technique on six normal volunteers, two patients with acanthosis nigricans and insulin resistance (AN), and one patient with idiopathic nonazotemic hyperkalemia (HK). In the basal state, all patients had normal fasting blood sugar, the AN patients had fasting hyperinsulinemia (600% of controls), and the HK patient had an elevated plasma potassium level of 5.1 mmol/L (n = 4.2 ± 0.2 mmol/L). During low dose (1 mU/kg-min), and high dose (10 mU/kg-min) insulin infusions, normals used glucose at a rate of 220 ± 10 and 470 ± 20 mg/ M2 min, respectively. The HK patient had a normal glucose use at both infusion rates, but the AN patients had a 20% decrease of glucose use compared to normals at the two infusion rates. In normal patients, plasma potassium fell by 0.7 and 1.4 mmol/L
at the end of the two infusion periods, respectively. AN patients had a similar fall in potassium, but the HK patient displayed no change in plasma potassium levels during a low dose insulin infusion, and only a 0.6 mmol/L drop during the high dose insulin infusion. These results indicate that: 1) patients with AN are resistant to insulin action on glucose use, 2) AN patients have a normal response to insulin on potassium uptake, 3) HK is a patient with normal response to insulin on glucose use, and 4) this patient is resistant to insulin action on potassium uptake. In conclusion: 1) we have demonstrated the independence of insulin action on glucose and potassium uptake in vivo, 2) we documented the existence of selective insulin resistance in the above patients, 3) we speculate, that in patients with a normal response to insulin on one parameter of its actions, and subnormal response on another parameter, a postreceptor defect rather than a receptor abnormality must exist. (J Clin Endocrinol Metab 73: 564-568,1991)
T
HE hypokalemic (HK) effect of insulin is well recognized and has important physiological (1), and therapeutic (2) implications. Numerous in vitro experiments have demonstrated increased uptake of potassium into cells in the presence of insulin, particularly in skeletal muscle (3). This effect has been attributed to stimulation of the Na-K-ATPase activity by insulin (4), and is accompanied by membrane depolarization (5). In vivo, insulin has been shown to have a role in regulating plasma potassium concentrations both in the basal fasting state (6) and in hyperinsulinemic conditions (7). Use of the euglycemic clamp technique allowed a quantitative documentation of the dose dependence and time course of the HK effect of insulin (7). This effect has been shown to be unaltered by age or beta adrenergic blockade (8), but seems to be diminished in obesity (9)
and Noninsulin Dependent Diabetes Mellitus (10). Additionally it was suggested that abnormal regulation of insulin on potassium homeostasis may be involved in the pathogenesis of hypertension (11), and familial HK periodic paralysis (12). Insulin action on potassium uptake has been suggested to be independent of the presence of glucose (13), but was shown to be related to the effects of insulin on glucose metabolism in normal men (7). Selective insulin resistance has been documented on the effects of insulin on glucose and phosphate metabolism, vs. free fatty acid and potassium metabolism, in experiments performed on normal subjects after a prolonged fast (14). Although this concept has been suggested in additional states (1517), most of the insulin resistant conditions reported to date, display a pan-resistance to insulin actions (9-11). In order to better characterize the relationship between insulin effects on glucose use and potassium uptake, we studied two patients with severe insulin resistance and acanthosis nigricans (AN), and one patient with idiopathic nonazotemic HK, with the euglycemic clamp
Received July 9, 1990. Address correspondence and requests for reprints to: Pinchas Cohen, M.D., Department of Pediatrics, Room S-322, Stanford University Medical Center, Stanford, California 94305. fDr. Cohen is a Juvenile Diabetes Foundation Fellow.
564
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 01:50 For personal use only. No other uses without permission. . All rights reserved.
SELECTIVE INSULIN RESISTANCE technique. Our findings suggest the existence of selective insulin resistance in these patients.
Materials and Methods
565
was used (Velosulin, Nordisk, Denmark). Glucose use was calculated as the mean glucose infusion rate at the last 40 min of each infusion period. Data are presented as mean ± SEM. On line glucose infusion calculations were made with an IBM PC, and further data analysis with the aid of a Macintosh SE.
Subjects
Results
Normal subjects (n = 6), were healthy volunteers without any history of endocrine, metabolic, or kidney disease. All normal subjects had a normal physical examination and normal routine blood chemistry and hematology tests. AN-1 was a 16yr-old male with obesity (205% of ideal body weight), and AN. AN-2 was a 25-yr-old female with obesity (160% of ideal body weight), AN, and mild hirsutism. Both had normal fasting blood glucose as well as serum electrolytes. HK was a 33-yr-old male followed for mild essential hypertension (blood pressure 160/100 mmHg). Repeated laboratory tests revealed potassium levels in the range of 5.1-7.1 mek/L (normal range: 3.5-5.2 mmol/L), the serum HCO3 was 17 mmol/L (normal range: 2428 mmol/L). The rest of the serum electrolytes, BUN, creatinine, and glucose were within normal limits as were the urinalysis, creatinine clearance, and inulin clearance. Plasma renin and aldosterone levels were not elevated but the HK was unresponsive to a trial of Florinef. The patient was diagnosed as having nonazotemic HK and type IV renal tubular acidosis; he was also noted to have elevated levels of digoxin-like immunoreactivity (2500 pg/ml, normal range: 600 ± 150 pg/ml). A full clinical report of this patient has been made by Szylman et al. (17). All subjects were admitted to the Metabolic Unit of Rambam Medical Center (Haifa, Israel), but remained active to approximate their prehospitalization exercise level and maintained their previous diet. The study was approved by the Human Rights Committee of Technion-Israel Institute of Technology, and informed consent was obtained from all subjects. Euglycemic insulin clamp technique Studies were performed after an overnight fast, as previously described (18-20). Subjects were catheterized 2 h before the start of insulin infusion, and basal values of insulin, glucose, and potassium were serially measured. Insulin was administered as primed continuous iv infusions of 1 and 10 mU/kgmin. Blood samples were obtained at 5-min intervals from an arterialized hand vein kept in a 60 C warming box. Euglycemia was maintained at values of 4.72-5.21 mmol/L throughout the experiment. Subjects were clamped for 2 h at each point; potassium values were measured at 10-min intervals and insulin levels at 20-min intervals. Serum was separated immediately and was either assayed or frozen for future analysis. Analytical methods Chemical analysis of insulin was done with a RIA kit by Biodata (Milano, Italy). Glucose was measured every 5 min. with a Glucometer (Ames Co., Clinton, IA) and was crosschecked against a Beckman glucose analyzer 2. Potassium was measured in duplicates by a flame photometer using lithium as an internal standard. Purified monocomponent porcine insulin
Glucose and insulin levels Table 1 summarizes the data on serum glucose and insulin concentrations during the insulin clamp studies in the controls and the three patients. Hyperinsulinemia in the two patients with AN was the only remarkable finding. Glucose use rates Figure 1 displays the mean steady state glucose use rates, at the final 40 min of a 2-h insulin infusion, of 1 mU/kg-min (dark bars) and 10 mU/kg-min {light bars). Values during the low dose infusion were: 220 ± 10 mg/ M2 • min for normals, 210 mg/M2 • min for the HK patient, and 180 mg/M2-min and 165 mg/M2-min for the two patients with AN. Values during the high dose infusion were: 470 ± 20 mg/M2/min for normals (range: 420-540 mg/M2/min), 465 mg/M2/min for the HK patient, and 390 mg/M2 • min and 380 mg/M2 • min for the two patients with AN. Serum potassium levels Figure 2 shows the change in potassium levels from basal values during a 1 mU/kg-min insulin infusion (light bars) and 10 mU/kg-min infusion (dark bars). Basal values of serum potassium were 4.2 ± 0.2 mmol/L in the control group, 5.1 ± 0.1 mmol/L in the HK patient, 4.0 ± 0 . 1 mmol/L, and 4.3 ± 0 . 1 mmol/L in the two patients with AN: Mean serum potassium levels during the last 40 min of the 1 mU/kg • min insulin infusion rate were 3.5 ± 0 . 1 mmol/L in the control group, 5.1 ± 0.1 mmol/L in the HK patient, 3.4 ± 0.1 mmol/L and 3.5 ± 0.1 mmol/L in the two patients with AN. Mean serum potassium levels during the last 30 min of the 10 mU/ kg-min insulin infusion rate were 2.8 ± 0.2 mmol/L in the control group, 4.5 ± 0 . 1 mmol/L in the HK patient, 3.0 ± 0.1 mmol/L and 2.9 ± 0.1 mmol/L in the two patients with AN.
Discussion The role of insulin in regulating potassium homeostasis in vivo has been documented and quantified by several investigators (1, 6-9, 12, 14, 15, 21, 22). Our results in normal subjects undergoing euglycemic clamp studies are
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 01:50 For personal use only. No other uses without permission. . All rights reserved.
COHEN ET AL.
566
JCE & M • 1991 Vol 73 • No 3
TABLE 1. Insulin (I) in pmol/L and glucose (G) in mmol/L values in the normal subjects (n = 6) and the three patients Patient group
KB)
Normals AN1 AN 2 HK
65 ± 7 430 ± 20° 380 ± 14° 84 ±20
KL)
G(B) 4.7 ± 4.9 ± 5.0 ± 4.8 ±
530 ± 980 ± 1120 ± 550 ±
0.2 0.2 0.3 0.3
120 140 125 150
G(L) 4.9 ± 4.9 ± 4.8 ± 5.0 ±
0.1 0.2 0.2 0.2
KH) 7800 ± 9900 ± 9300 ± 8000 ±
1000 1400 700 500
G(H) 5.1 ± 5.0 ± 4.9 ± 5.0 ±
0.1 0.1 0.2 0.2
Glucose values are in mg/dL, insulin values are in /iU/mL. (B) denotes values at the basal state, (L) at the 1 mU/Kg/min insulin infusion period, and (H) at the 10 mU/Kgmin insulin infusion period. " Denotes outside of the 95% confidence limit for normal. 500-1
400-
E
300 -
=
200 -
o
100-
•J
a
5 Normals
AN 1
AN 2
HK
FIG. 1. Glucose use rates during the last 40 min of 2-h insulin infusions of 1 (dark bars) and 10 (light bars) mU/kg-min, in normals (n = 6), two patients with AN (AN 1 and AN 2) and a patient with idiopathic HK.
Normals AN 1 AN 2 HK FIG. 2. Insulin-mediated decrease in plasma potassium. Levels are the mean of five measurements during the final 40 min of the infusion of 1 (light bars) and 10 (dark bars) mU/Kg- min insulin for 2 h to normals (n = 6), patients with AN (AN 1 and AN 2), and a patient with idiopathic HK.
in agreement with those previously reported. We demonstrated a fall of plasma potassium by 0.7 mmol/L during a 1 mU/Kg-min insulin infusion, and 1.4 mmol/ L during a 10 mU/Kg/min. DeFronzo et al. (7) have
found similar reductions of 1.0 and 1.5 mmol/L in plasma potassium during identical insulin infusions. The same investigator in a later study (9) described a fall of 0.8 and 1.2 mmol/L in similar experiments. Minaker et al. (8, 12), performing similar experiments found an equivalent drop in plasma potassium during a low dose insulin infusion but failed to demonstrate an increase of the maximal fall in plasma potassium during high dose insulin infusion. These investigators, however, did find a faster drop in the potassium levels during the higher insulin infusion rates. Other investigators examining only the effect of a low dose insulin infusion, are in agreement with our results (14, 15, 22). Whereas it is impossible to estimate the rate of potassium influx into the tissues in response to insulin in vivo, infusions of 0.5-1.0 mmol potassium chloride during clamp experiments failed to alter plasma potassium levels in normal volunteers (22). AN has been associated with insulin resistance (23) and was previously thought to be associated with defects in the insulin receptor (24). Recently, evidence has been gathering for the possible involvement of postreceptor defects in insulin action in this state (25). Potassium levels have not been reported to be altered in this state, and the effect of insulin on plasma potassium has not been previously evaluated. We have shown here that two patients with AN and severe insulin resistance in regard to glucose use, have a normal response to insulin on potassium uptake. Whereas our findings of defective glucose use in these patients are in agreement with other studies (26, 27), the interpretation of these results has been controversial. The decreased glucose use at submaximal insulin concentrations, (assuming the existence of spare insulin receptors), could be explained by either receptoral or postreceptoral defects in insulin action (28). The decrease in glucose use at high dose insulin infusion is usually thought to be due to a postreceptor defect, but without demonstrating a flattening of the dose response curve of insulin action, this finding may also represent a manifestation of an insulin receptor defect. In conjunction with a normal response to insulin on potassium metabolism, we believe that the impaired glucose use must be a postreceptor defect, since a receptoral defect should affect all of insulin's actions indiscriminately.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 01:50 For personal use only. No other uses without permission. . All rights reserved.
SELECTIVE INSULIN RESISTANCE
rinsulinJ
site of defect in acanthosis .
FIG. 3. A model for insulin action on glucose and potassium metabolism. Details of selective and nonselective insulin resistance are outlined in the text.
567
insulin receptor site of defect in ^> hyperkalemic pt
I primary
Extracellular ^r \ space ^ r glucose \ f metabolism , m
/ 1 1
I effector >*' mechanism ^S^ ^\ \
pathway
\
\S X ^
potassium I transport
a
pathway
other insulin
\
1•
effects In contrast to the patients with AN, our patient with idiopathic nonazotemic HK had a normal response to insulin action on glucose use coupled with a markedly decreased effect of insulin on potassium uptake. Whereas the actual metabolic defect in this patient is not known and may involve more then just the extra renal regulation of potassium by insulin, this patient definitely manifests insulin resistance to one of insulin main actions. Interestingly, Ferranini et al. (21) described the presence of normal potassium uptake, with impaired glucose use in response to insulin, in a group of patients with essential hypertension. Whereas mild hypertension was the presenting problem in this patient, he clearly represent a unique case different than these patients. The presence of an aldosterone level in the high normal range, and lack of response to Florinef in our HK patient, suggests that his defect in potassium uptake may be in a biochemical locus that is under multiple hormonal regulation. In Fig. 3 we have summarized the possible interactions of insulin with its receptor and subsequent postreceptoral effector mechanisms leading to various insulin actions. We postulate, that a defect at the insulin receptor level such as might exist in obesity, will result in a decreased effect of insulin on both glucose and potassium metabolism as described by DeFronzo (9). Defects in insulin effector loci involving such molecules as the glucose transporter, will result in insulin resistance as far as glucose use is concerned (29), but should have a normal response to other insulin actions. AN in our patients is an example to such a case. Finally a patient such as our HK patient may represent a case of a defect in an insulin dependent potassium uptake locus as shown in Fig. 3. In summary, we have shown that as previously described, normal subjects have a reproducible effect of insulin on both potassium and glucose metabolism. In certain cases a selective resistance to insulin action on either glucose or potassium uptake exists. In such cases, a postreceptor defect in insulin action is likely.
References 1. Bia MJ, DeFronzo RA. Extrarenal potassium homeostasis. Am J Physiol. 1981;240:F257-68. 2. Foster DW, McGarry JD. The metabolic derangements and treatment of diabetic ketoacidosis. N Engl J Med. 1983;309:159-69. 3. Zierler KL, Rogus E, Hazlewood CF. Effect of insulin on potassium flux and water and electrolyte content of muscles from normal and from hypophysectomized rats. J Gen Physiol. 1966;49:433-56. 4. Gavryck WA, Moore RD, Thompson RC. Effects of insulin upon membrane bound Na-K-ATPase extracted from frog skeletal muscle. J Physiol. 1975;252:43-58. 5. Zierler KL. Possible mechanisms of insulin action on membrane potential and ion fluxes. Am J Med. 1966;40:735-9. 6. DeFronzo RA, Sherwin RS, Dillingham M, Hendler R, Tamborlane WV, Felig P. Influence of basal insulin and glucagon secretion on potassium and sodium metabolism. J Clin Invest. 1978;61:472-9. 7. DePronzo RA, Felig P, Ferrannini E, Wahren J. Effects of graded doses of insulin on splanchnic and peripheral potassium metabolism in man. Am J Physiol. 1980;238:E421-7. 8. Minaker KL, Rowe JW. Potassium homeostasis during hyperinsulinemia: effects of insulin level, (8-blockade, and age. Am J Physiol. 1982;242:E373-7. 9. DeFronzo RA. Obesity is associated with impaired insulin-mediated potassium uptake. Metabolism. 1988;37:105-8. 10. Rosenstock J, Loizou SA, Brajkovich IE, Mashiter K, Joplin GF. Effect of acute hyperglycaemia on plasma potassium and aldosterone levels in type 2 (non-insulin-dependent) diabetes. Diabetologia. 1982;22:184-7. 11. Halkin H, Modan M, Shefi M, Almog S. Altered erythrocyte and plasma sodium and potassium in hypertension, a facet of hyperinsulinemia. Hypertension. 1988;ll:71-7. 12. Minaker KL, Meneilly GS, Flier JS, Rowe JW. Insulin-mediated hypokalemia and paralysis in familial hypokalemic periodic paralysis. Am J Med. 1988;84:1001-6. 13. Zierler KL. Effect of insulin on potassium efflux from rat muscle in the presence and absence of glucose. Am J Physiol. 1968;198:1066-70. 14. Newman WP, Brodows RG. Insulin action during acute starvation: evidence for selective insulin resistance in normal man. Metabolism. 1983;32:590-6. 15. Geffner ME, Santulli TV, Kaplan SA. Hypertrophic cardiomyopathy in total lipodistrophy: insulin action in the face of insulin resistance? J Pediatr. 1988; 110:161. 16. Zierler KL, Rabinovitz D. Effect of very small concentrations of insulin on forearm metabolism: persistence of its actions on potassium and free fatty acids without its effect on glucose. J Clin Invest. 1964;43:950-62. 17. Szylman P, Wolach B, Winaver J, Pannett R, Cohen P, Shenkman L, Better OS. Nonazotemic hyperkalemia with renal and extra-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 01:50 For personal use only. No other uses without permission. . All rights reserved.
568
18. 19. 20.
21. 22. 23.
COHEN ET AL. renal defects in potassium transport: association with high levels of digoxin-like immunoreactive factor. J Clin Lab Med. 1990;116:315-26. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol. 1979;237:E214-23. Rizza AR, Mandarino LA, Gerich JE. Dose response characteristics for the effects of insulin on production and utilization of glucose in man. Am J Physiol. 1981;240:E630-9. Cohen P, Barzilai N, Barzilai D, Karnieli E. Correlation between insulin clearance and insulin responsiveness: studies in normal, obese, hyperthyroid and Cushing's syndrome patients. Metabolism. 1986;35:744-9. Ferrannini E, Buzzigoli G, Bonadonna R, et al. Insulin resistance in essential hypertension. N Engl J Med. 1987;317:350-7. Burke B, Andrews WJ, Hadden DR. Insulin infusion and serum potassium in normal subjects. Diabetes Res. 1987;6:187-8. Flier JS. Metabolic importance of acanthosis nigricans. Arch Dermatol. 1985;121:193-4.
JCE&M«1991 Vol 73 • No 3
24. Kahn CR, Flier JS, Bar RS, et al. The syndrome of insulin resistance and acanthosis nigricans. Insulin receptor disorders in man. N Engl J Med. 1976;294:739-45. 25. Bar RS, Muggeo M, Roth J, Kahn CR, Havrankova J, ImperutoMcgialy J. Insulin resistance, acanthosis nigricans and normal insulin receptor in a young women: evidence for a postreceptor defect. J Clin Endocrinol Metab. 1978;47:620-5. 26. Grunberger G, Zick Y, Gordon P. Defects in phosphorylation of insulin receptors in cells from insulin resistance patients with normal insulin binding. Science. 1984;223:932-4. 27. Schwewk WF, Rizza RA, Mandarino LJ, Gerich JE, Hayels AB, Haymond MW. Familial insulin resistance and acanthosis nigricans: presence of postreceptor defect. Diabetes 1986;35:33-3. 28. Olefsky JM, Kolterman OG. Mechanism of insulin resistance in obesity and non-insulin diabetes. Am J Med. 1981;70:151-69. 29. Karnieli E, Hissin PI, Simpson IA, Salans LB, Cushman SW. A possible mechanism of insulin resistance in the rat adipose cell in streptozotocin induced diabetes mellitus: depletion of intracellular glucose transport systems. J Clin Invest. 1981;68:811-4.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 01:50 For personal use only. No other uses without permission. . All rights reserved.
Vol. 73, No. 3 Printed in U.S.A.
Insulin Effects on Glucose and Potassium Metabolism in Vivo: Evidence for Selective Insulin Resistance in Humans PINCHAS COHENf, NIR BARZILAI, AMIR LERMAN, HAVA HAREL, PEDRO SZYLMAN, AND EDDY KARNIELI Metabolic Unit, Endocrine Institute, Department of Medicine C, and Department of Nephrology, Rambam Medical Center and Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel
ABSTRACT. The effects of insulin on in vivo glucose use and potassium uptake in healthy humans are well documented. However, the interrelationship between these two processes is not fully defined. In order to characterize it, we have used the euglycemic clamp technique on six normal volunteers, two patients with acanthosis nigricans and insulin resistance (AN), and one patient with idiopathic nonazotemic hyperkalemia (HK). In the basal state, all patients had normal fasting blood sugar, the AN patients had fasting hyperinsulinemia (600% of controls), and the HK patient had an elevated plasma potassium level of 5.1 mmol/L (n = 4.2 ± 0.2 mmol/L). During low dose (1 mU/kg-min), and high dose (10 mU/kg-min) insulin infusions, normals used glucose at a rate of 220 ± 10 and 470 ± 20 mg/ M2 min, respectively. The HK patient had a normal glucose use at both infusion rates, but the AN patients had a 20% decrease of glucose use compared to normals at the two infusion rates. In normal patients, plasma potassium fell by 0.7 and 1.4 mmol/L
at the end of the two infusion periods, respectively. AN patients had a similar fall in potassium, but the HK patient displayed no change in plasma potassium levels during a low dose insulin infusion, and only a 0.6 mmol/L drop during the high dose insulin infusion. These results indicate that: 1) patients with AN are resistant to insulin action on glucose use, 2) AN patients have a normal response to insulin on potassium uptake, 3) HK is a patient with normal response to insulin on glucose use, and 4) this patient is resistant to insulin action on potassium uptake. In conclusion: 1) we have demonstrated the independence of insulin action on glucose and potassium uptake in vivo, 2) we documented the existence of selective insulin resistance in the above patients, 3) we speculate, that in patients with a normal response to insulin on one parameter of its actions, and subnormal response on another parameter, a postreceptor defect rather than a receptor abnormality must exist. (J Clin Endocrinol Metab 73: 564-568,1991)
T
HE hypokalemic (HK) effect of insulin is well recognized and has important physiological (1), and therapeutic (2) implications. Numerous in vitro experiments have demonstrated increased uptake of potassium into cells in the presence of insulin, particularly in skeletal muscle (3). This effect has been attributed to stimulation of the Na-K-ATPase activity by insulin (4), and is accompanied by membrane depolarization (5). In vivo, insulin has been shown to have a role in regulating plasma potassium concentrations both in the basal fasting state (6) and in hyperinsulinemic conditions (7). Use of the euglycemic clamp technique allowed a quantitative documentation of the dose dependence and time course of the HK effect of insulin (7). This effect has been shown to be unaltered by age or beta adrenergic blockade (8), but seems to be diminished in obesity (9)
and Noninsulin Dependent Diabetes Mellitus (10). Additionally it was suggested that abnormal regulation of insulin on potassium homeostasis may be involved in the pathogenesis of hypertension (11), and familial HK periodic paralysis (12). Insulin action on potassium uptake has been suggested to be independent of the presence of glucose (13), but was shown to be related to the effects of insulin on glucose metabolism in normal men (7). Selective insulin resistance has been documented on the effects of insulin on glucose and phosphate metabolism, vs. free fatty acid and potassium metabolism, in experiments performed on normal subjects after a prolonged fast (14). Although this concept has been suggested in additional states (1517), most of the insulin resistant conditions reported to date, display a pan-resistance to insulin actions (9-11). In order to better characterize the relationship between insulin effects on glucose use and potassium uptake, we studied two patients with severe insulin resistance and acanthosis nigricans (AN), and one patient with idiopathic nonazotemic HK, with the euglycemic clamp
Received July 9, 1990. Address correspondence and requests for reprints to: Pinchas Cohen, M.D., Department of Pediatrics, Room S-322, Stanford University Medical Center, Stanford, California 94305. fDr. Cohen is a Juvenile Diabetes Foundation Fellow.
564
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 01:50 For personal use only. No other uses without permission. . All rights reserved.
SELECTIVE INSULIN RESISTANCE technique. Our findings suggest the existence of selective insulin resistance in these patients.
Materials and Methods
565
was used (Velosulin, Nordisk, Denmark). Glucose use was calculated as the mean glucose infusion rate at the last 40 min of each infusion period. Data are presented as mean ± SEM. On line glucose infusion calculations were made with an IBM PC, and further data analysis with the aid of a Macintosh SE.
Subjects
Results
Normal subjects (n = 6), were healthy volunteers without any history of endocrine, metabolic, or kidney disease. All normal subjects had a normal physical examination and normal routine blood chemistry and hematology tests. AN-1 was a 16yr-old male with obesity (205% of ideal body weight), and AN. AN-2 was a 25-yr-old female with obesity (160% of ideal body weight), AN, and mild hirsutism. Both had normal fasting blood glucose as well as serum electrolytes. HK was a 33-yr-old male followed for mild essential hypertension (blood pressure 160/100 mmHg). Repeated laboratory tests revealed potassium levels in the range of 5.1-7.1 mek/L (normal range: 3.5-5.2 mmol/L), the serum HCO3 was 17 mmol/L (normal range: 2428 mmol/L). The rest of the serum electrolytes, BUN, creatinine, and glucose were within normal limits as were the urinalysis, creatinine clearance, and inulin clearance. Plasma renin and aldosterone levels were not elevated but the HK was unresponsive to a trial of Florinef. The patient was diagnosed as having nonazotemic HK and type IV renal tubular acidosis; he was also noted to have elevated levels of digoxin-like immunoreactivity (2500 pg/ml, normal range: 600 ± 150 pg/ml). A full clinical report of this patient has been made by Szylman et al. (17). All subjects were admitted to the Metabolic Unit of Rambam Medical Center (Haifa, Israel), but remained active to approximate their prehospitalization exercise level and maintained their previous diet. The study was approved by the Human Rights Committee of Technion-Israel Institute of Technology, and informed consent was obtained from all subjects. Euglycemic insulin clamp technique Studies were performed after an overnight fast, as previously described (18-20). Subjects were catheterized 2 h before the start of insulin infusion, and basal values of insulin, glucose, and potassium were serially measured. Insulin was administered as primed continuous iv infusions of 1 and 10 mU/kgmin. Blood samples were obtained at 5-min intervals from an arterialized hand vein kept in a 60 C warming box. Euglycemia was maintained at values of 4.72-5.21 mmol/L throughout the experiment. Subjects were clamped for 2 h at each point; potassium values were measured at 10-min intervals and insulin levels at 20-min intervals. Serum was separated immediately and was either assayed or frozen for future analysis. Analytical methods Chemical analysis of insulin was done with a RIA kit by Biodata (Milano, Italy). Glucose was measured every 5 min. with a Glucometer (Ames Co., Clinton, IA) and was crosschecked against a Beckman glucose analyzer 2. Potassium was measured in duplicates by a flame photometer using lithium as an internal standard. Purified monocomponent porcine insulin
Glucose and insulin levels Table 1 summarizes the data on serum glucose and insulin concentrations during the insulin clamp studies in the controls and the three patients. Hyperinsulinemia in the two patients with AN was the only remarkable finding. Glucose use rates Figure 1 displays the mean steady state glucose use rates, at the final 40 min of a 2-h insulin infusion, of 1 mU/kg-min (dark bars) and 10 mU/kg-min {light bars). Values during the low dose infusion were: 220 ± 10 mg/ M2 • min for normals, 210 mg/M2 • min for the HK patient, and 180 mg/M2-min and 165 mg/M2-min for the two patients with AN. Values during the high dose infusion were: 470 ± 20 mg/M2/min for normals (range: 420-540 mg/M2/min), 465 mg/M2/min for the HK patient, and 390 mg/M2 • min and 380 mg/M2 • min for the two patients with AN. Serum potassium levels Figure 2 shows the change in potassium levels from basal values during a 1 mU/kg-min insulin infusion (light bars) and 10 mU/kg-min infusion (dark bars). Basal values of serum potassium were 4.2 ± 0.2 mmol/L in the control group, 5.1 ± 0.1 mmol/L in the HK patient, 4.0 ± 0 . 1 mmol/L, and 4.3 ± 0 . 1 mmol/L in the two patients with AN: Mean serum potassium levels during the last 40 min of the 1 mU/kg • min insulin infusion rate were 3.5 ± 0 . 1 mmol/L in the control group, 5.1 ± 0.1 mmol/L in the HK patient, 3.4 ± 0.1 mmol/L and 3.5 ± 0.1 mmol/L in the two patients with AN. Mean serum potassium levels during the last 30 min of the 10 mU/ kg-min insulin infusion rate were 2.8 ± 0.2 mmol/L in the control group, 4.5 ± 0 . 1 mmol/L in the HK patient, 3.0 ± 0.1 mmol/L and 2.9 ± 0.1 mmol/L in the two patients with AN.
Discussion The role of insulin in regulating potassium homeostasis in vivo has been documented and quantified by several investigators (1, 6-9, 12, 14, 15, 21, 22). Our results in normal subjects undergoing euglycemic clamp studies are
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 01:50 For personal use only. No other uses without permission. . All rights reserved.
COHEN ET AL.
566
JCE & M • 1991 Vol 73 • No 3
TABLE 1. Insulin (I) in pmol/L and glucose (G) in mmol/L values in the normal subjects (n = 6) and the three patients Patient group
KB)
Normals AN1 AN 2 HK
65 ± 7 430 ± 20° 380 ± 14° 84 ±20
KL)
G(B) 4.7 ± 4.9 ± 5.0 ± 4.8 ±
530 ± 980 ± 1120 ± 550 ±
0.2 0.2 0.3 0.3
120 140 125 150
G(L) 4.9 ± 4.9 ± 4.8 ± 5.0 ±
0.1 0.2 0.2 0.2
KH) 7800 ± 9900 ± 9300 ± 8000 ±
1000 1400 700 500
G(H) 5.1 ± 5.0 ± 4.9 ± 5.0 ±
0.1 0.1 0.2 0.2
Glucose values are in mg/dL, insulin values are in /iU/mL. (B) denotes values at the basal state, (L) at the 1 mU/Kg/min insulin infusion period, and (H) at the 10 mU/Kgmin insulin infusion period. " Denotes outside of the 95% confidence limit for normal. 500-1
400-
E
300 -
=
200 -
o
100-
•J
a
5 Normals
AN 1
AN 2
HK
FIG. 1. Glucose use rates during the last 40 min of 2-h insulin infusions of 1 (dark bars) and 10 (light bars) mU/kg-min, in normals (n = 6), two patients with AN (AN 1 and AN 2) and a patient with idiopathic HK.
Normals AN 1 AN 2 HK FIG. 2. Insulin-mediated decrease in plasma potassium. Levels are the mean of five measurements during the final 40 min of the infusion of 1 (light bars) and 10 (dark bars) mU/Kg- min insulin for 2 h to normals (n = 6), patients with AN (AN 1 and AN 2), and a patient with idiopathic HK.
in agreement with those previously reported. We demonstrated a fall of plasma potassium by 0.7 mmol/L during a 1 mU/Kg-min insulin infusion, and 1.4 mmol/ L during a 10 mU/Kg/min. DeFronzo et al. (7) have
found similar reductions of 1.0 and 1.5 mmol/L in plasma potassium during identical insulin infusions. The same investigator in a later study (9) described a fall of 0.8 and 1.2 mmol/L in similar experiments. Minaker et al. (8, 12), performing similar experiments found an equivalent drop in plasma potassium during a low dose insulin infusion but failed to demonstrate an increase of the maximal fall in plasma potassium during high dose insulin infusion. These investigators, however, did find a faster drop in the potassium levels during the higher insulin infusion rates. Other investigators examining only the effect of a low dose insulin infusion, are in agreement with our results (14, 15, 22). Whereas it is impossible to estimate the rate of potassium influx into the tissues in response to insulin in vivo, infusions of 0.5-1.0 mmol potassium chloride during clamp experiments failed to alter plasma potassium levels in normal volunteers (22). AN has been associated with insulin resistance (23) and was previously thought to be associated with defects in the insulin receptor (24). Recently, evidence has been gathering for the possible involvement of postreceptor defects in insulin action in this state (25). Potassium levels have not been reported to be altered in this state, and the effect of insulin on plasma potassium has not been previously evaluated. We have shown here that two patients with AN and severe insulin resistance in regard to glucose use, have a normal response to insulin on potassium uptake. Whereas our findings of defective glucose use in these patients are in agreement with other studies (26, 27), the interpretation of these results has been controversial. The decreased glucose use at submaximal insulin concentrations, (assuming the existence of spare insulin receptors), could be explained by either receptoral or postreceptoral defects in insulin action (28). The decrease in glucose use at high dose insulin infusion is usually thought to be due to a postreceptor defect, but without demonstrating a flattening of the dose response curve of insulin action, this finding may also represent a manifestation of an insulin receptor defect. In conjunction with a normal response to insulin on potassium metabolism, we believe that the impaired glucose use must be a postreceptor defect, since a receptoral defect should affect all of insulin's actions indiscriminately.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 01:50 For personal use only. No other uses without permission. . All rights reserved.
SELECTIVE INSULIN RESISTANCE
rinsulinJ
site of defect in acanthosis .
FIG. 3. A model for insulin action on glucose and potassium metabolism. Details of selective and nonselective insulin resistance are outlined in the text.
567
insulin receptor site of defect in ^> hyperkalemic pt
I primary
Extracellular ^r \ space ^ r glucose \ f metabolism , m
/ 1 1
I effector >*' mechanism ^S^ ^\ \
pathway
\
\S X ^
potassium I transport
a
pathway
other insulin
\
1•
effects In contrast to the patients with AN, our patient with idiopathic nonazotemic HK had a normal response to insulin action on glucose use coupled with a markedly decreased effect of insulin on potassium uptake. Whereas the actual metabolic defect in this patient is not known and may involve more then just the extra renal regulation of potassium by insulin, this patient definitely manifests insulin resistance to one of insulin main actions. Interestingly, Ferranini et al. (21) described the presence of normal potassium uptake, with impaired glucose use in response to insulin, in a group of patients with essential hypertension. Whereas mild hypertension was the presenting problem in this patient, he clearly represent a unique case different than these patients. The presence of an aldosterone level in the high normal range, and lack of response to Florinef in our HK patient, suggests that his defect in potassium uptake may be in a biochemical locus that is under multiple hormonal regulation. In Fig. 3 we have summarized the possible interactions of insulin with its receptor and subsequent postreceptoral effector mechanisms leading to various insulin actions. We postulate, that a defect at the insulin receptor level such as might exist in obesity, will result in a decreased effect of insulin on both glucose and potassium metabolism as described by DeFronzo (9). Defects in insulin effector loci involving such molecules as the glucose transporter, will result in insulin resistance as far as glucose use is concerned (29), but should have a normal response to other insulin actions. AN in our patients is an example to such a case. Finally a patient such as our HK patient may represent a case of a defect in an insulin dependent potassium uptake locus as shown in Fig. 3. In summary, we have shown that as previously described, normal subjects have a reproducible effect of insulin on both potassium and glucose metabolism. In certain cases a selective resistance to insulin action on either glucose or potassium uptake exists. In such cases, a postreceptor defect in insulin action is likely.
References 1. Bia MJ, DeFronzo RA. Extrarenal potassium homeostasis. Am J Physiol. 1981;240:F257-68. 2. Foster DW, McGarry JD. The metabolic derangements and treatment of diabetic ketoacidosis. N Engl J Med. 1983;309:159-69. 3. Zierler KL, Rogus E, Hazlewood CF. Effect of insulin on potassium flux and water and electrolyte content of muscles from normal and from hypophysectomized rats. J Gen Physiol. 1966;49:433-56. 4. Gavryck WA, Moore RD, Thompson RC. Effects of insulin upon membrane bound Na-K-ATPase extracted from frog skeletal muscle. J Physiol. 1975;252:43-58. 5. Zierler KL. Possible mechanisms of insulin action on membrane potential and ion fluxes. Am J Med. 1966;40:735-9. 6. DeFronzo RA, Sherwin RS, Dillingham M, Hendler R, Tamborlane WV, Felig P. Influence of basal insulin and glucagon secretion on potassium and sodium metabolism. J Clin Invest. 1978;61:472-9. 7. DePronzo RA, Felig P, Ferrannini E, Wahren J. Effects of graded doses of insulin on splanchnic and peripheral potassium metabolism in man. Am J Physiol. 1980;238:E421-7. 8. Minaker KL, Rowe JW. Potassium homeostasis during hyperinsulinemia: effects of insulin level, (8-blockade, and age. Am J Physiol. 1982;242:E373-7. 9. DeFronzo RA. Obesity is associated with impaired insulin-mediated potassium uptake. Metabolism. 1988;37:105-8. 10. Rosenstock J, Loizou SA, Brajkovich IE, Mashiter K, Joplin GF. Effect of acute hyperglycaemia on plasma potassium and aldosterone levels in type 2 (non-insulin-dependent) diabetes. Diabetologia. 1982;22:184-7. 11. Halkin H, Modan M, Shefi M, Almog S. Altered erythrocyte and plasma sodium and potassium in hypertension, a facet of hyperinsulinemia. Hypertension. 1988;ll:71-7. 12. Minaker KL, Meneilly GS, Flier JS, Rowe JW. Insulin-mediated hypokalemia and paralysis in familial hypokalemic periodic paralysis. Am J Med. 1988;84:1001-6. 13. Zierler KL. Effect of insulin on potassium efflux from rat muscle in the presence and absence of glucose. Am J Physiol. 1968;198:1066-70. 14. Newman WP, Brodows RG. Insulin action during acute starvation: evidence for selective insulin resistance in normal man. Metabolism. 1983;32:590-6. 15. Geffner ME, Santulli TV, Kaplan SA. Hypertrophic cardiomyopathy in total lipodistrophy: insulin action in the face of insulin resistance? J Pediatr. 1988; 110:161. 16. Zierler KL, Rabinovitz D. Effect of very small concentrations of insulin on forearm metabolism: persistence of its actions on potassium and free fatty acids without its effect on glucose. J Clin Invest. 1964;43:950-62. 17. Szylman P, Wolach B, Winaver J, Pannett R, Cohen P, Shenkman L, Better OS. Nonazotemic hyperkalemia with renal and extra-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 01:50 For personal use only. No other uses without permission. . All rights reserved.
568
18. 19. 20.
21. 22. 23.
COHEN ET AL. renal defects in potassium transport: association with high levels of digoxin-like immunoreactive factor. J Clin Lab Med. 1990;116:315-26. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol. 1979;237:E214-23. Rizza AR, Mandarino LA, Gerich JE. Dose response characteristics for the effects of insulin on production and utilization of glucose in man. Am J Physiol. 1981;240:E630-9. Cohen P, Barzilai N, Barzilai D, Karnieli E. Correlation between insulin clearance and insulin responsiveness: studies in normal, obese, hyperthyroid and Cushing's syndrome patients. Metabolism. 1986;35:744-9. Ferrannini E, Buzzigoli G, Bonadonna R, et al. Insulin resistance in essential hypertension. N Engl J Med. 1987;317:350-7. Burke B, Andrews WJ, Hadden DR. Insulin infusion and serum potassium in normal subjects. Diabetes Res. 1987;6:187-8. Flier JS. Metabolic importance of acanthosis nigricans. Arch Dermatol. 1985;121:193-4.
JCE&M«1991 Vol 73 • No 3
24. Kahn CR, Flier JS, Bar RS, et al. The syndrome of insulin resistance and acanthosis nigricans. Insulin receptor disorders in man. N Engl J Med. 1976;294:739-45. 25. Bar RS, Muggeo M, Roth J, Kahn CR, Havrankova J, ImperutoMcgialy J. Insulin resistance, acanthosis nigricans and normal insulin receptor in a young women: evidence for a postreceptor defect. J Clin Endocrinol Metab. 1978;47:620-5. 26. Grunberger G, Zick Y, Gordon P. Defects in phosphorylation of insulin receptors in cells from insulin resistance patients with normal insulin binding. Science. 1984;223:932-4. 27. Schwewk WF, Rizza RA, Mandarino LJ, Gerich JE, Hayels AB, Haymond MW. Familial insulin resistance and acanthosis nigricans: presence of postreceptor defect. Diabetes 1986;35:33-3. 28. Olefsky JM, Kolterman OG. Mechanism of insulin resistance in obesity and non-insulin diabetes. Am J Med. 1981;70:151-69. 29. Karnieli E, Hissin PI, Simpson IA, Salans LB, Cushman SW. A possible mechanism of insulin resistance in the rat adipose cell in streptozotocin induced diabetes mellitus: depletion of intracellular glucose transport systems. J Clin Invest. 1981;68:811-4.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 01:50 For personal use only. No other uses without permission. . All rights reserved.
