0021-972X/83/5606-1294$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright© 1983 by The Endocrine Society

Vol. 56, No. 6 Printed in U.S.A.

Hepatic Removal of Insulin in Normal Man: Dose Response to Endogenous Insulin Secretion* R. PHILIP EATON, RICHARD C. ALLEN, AND DAVID S. SCHADE University of New Mexico School of Medicine, Departments of Medicine and Mathematics, Albuquerque, New Mexico 87131

ABSTRACT. The hepatic extraction of insulin in normal man was evaluated by kinetic analysis of peripheral insulin behavior in the plasma following stimulation of endogenous insulin secretion. Prehepatic insulin production was determined by deconvolution of plasma connecting peptide behavior (C-peptide), and hepatic extraction of the secreted insulin determined with a three-compartment model for hepatic, vascular, and extravascular plasma spaces. Three dosages of oral glucose (10, 25, and 100 g) administered to normal volunteers resulted in 1.8 ± 0.5, 2.7 ± 1.1, and 7.2 ± 1.6 U endogenous insulin secretion, respectively. Total hepatic exposure to insulin exceeded the endogenous secretion due to recycling to the liver from the systemic

T

HE LIVER occupies a central role in the production and disposal of glucose (1) and is regulated in part by small changes in peripheral plasma insulin concentration. The interpretation of peripheral insulin levels relative to hepatic glucose regulation is complicated by the fact that approximately half of endogenously secreted insulin is removed from the portal blood by the liver before entrance into the systemic circulation (2). Moreover, it has been suggested (3) that this hepatocytebound/degraded insulin may regulate glucose metabolism within the liver. Previous studies in the perfused isolated rat liver have suggested that hepatic insulin removal is proportional to arterial hormone levels within physiological ranges, and becomes saturated at higher perfusate levels of insulin (4-6). In vivo studies in man in response to both endogenous insulin secretin (7,8) and exogenous portal insulin infusion (9, 10) have demonstrated a reduction in the percentage of hepatic insulin removal with increasing insulin exposure, though the quantitative relationships between these events have not been defined in a fashion comparable to liver perfusion studies. Received August 13,1982. Address requests for reprints to: R. Philip Eaton, M.D., Department of Medicine, The University of New Mexico, Albuquerque, New Mexico 87131. * This work was supported by Grant RR-997-06 from the NIH Clinical Research Program, Grant IP50-HD-11327 from HEW NICHD, Grant IR01-AM-25132 from NIH and a grant from the KROC Foundation.

circulation. Decreasing insulin extraction by the liver (67—5342%) in the presence of increasing insulin exposure (2.6-4.413.2 U) was observed during the dose-response to glucose. The rates of hepatic insulin extraction observed with arginine (58 ± 9% with 3.2 U), and a normal meal (50 ± 9% with 7.6 U) were intermediate between the extremes seen with the 10- and 100-g glucose challenge. These results quantitate hepatic exposure of insulin in man during differing stimuli of endogenous insulin secretion, and demonstrate reduced fractional hepatic extraction with increasing insulin exposure. (J Clin Endocrinol Metab 56: 1294,1983)

Efforts to quantitate hepatic insulin extraction in vivo in man have been complicated by the difficulty in measuring portal and hepatic vein insulin which is dependent upon invasive technology for blood sampling in a high flow vascular system (11). The availability of the noninvasive determination of prehepatic insulin production (portal vein insulin) utilizing peripheral connecting-peptide (C-peptide) behavior (12), and of posthepatic insulin behavior (peripheral vein insulin) based upon the investigation of exogenous insulin infusion by compartmental analysis (13-15), make possible an alternative approach to the question of hepatic insulin extraction. In the present investigation, we have examined hepatic removal of insulin following oral glucose stimulation in a dose-response fashion in normal man, and compared these results with those obtained during iv arginine stimulation and the ingestion of a mixed caloric meal. These studies quantitate insulin removal by the liver in relation to the total endogenous insulin exposure in man, and demonstrate reduced hepatic extraction of insulin during increased insulin exposure.

Subjects and Methods Patient selection (Table 1) Seven healthy adult male subjects volunteered for investigation of hepatic removal of insulin, with demographic characteristics given in Table 1. All had negative family histories

1294

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 January 2017. at 06:28 For personal use only. No other uses without permission. . All rights reserved.

HEPATIC REMOVAL OF INSULIN

1295

TABLE 1. Clinical data

Pt.

Age (yr)

Wt (kg)

Height (cm)

Ponderal index"

Glycosylated hemoglobin (mol/100 ml)

1 2 3 4 5 6 7

24 25 20 26 27 24 22

65.4 68.9 65.3 68.9 77.4 64.1 78.4

180 180 170 175 180 178 188

22.4 22.8 23.6 23.4 23.7 22.5 22.8

4.5 4.8 4.7 4.9 5.4 5.1 4.3

96 84 91 95 87 88 96

78 68 108 74 101 85 81

Mean ± SD

24 ± 2

69.8 ± 5.8

179 ± 5

23.0 ± 0.5

4.8 ± 0.4

91 ± 5

85 ±14

0 6

Fasting plasma 2-h postprandial plasma glucose* glucose* (mg/dl) (mg/dl)

Ponderal index, 10 [(wt) VsJ/height. During 100 g GTT.

of diabetes mellitus, a normal red blood cell glycosylated hemoglobin (16), and a normal oral glucose tolerance test (100 g) (17). The mean age was 24 ± 2 yr, and the mean ponderal index was 23 ± 0.5, indicating a nonobese body mass (18). All studies were performed at the General Clinical Research Center, where subjects received a diet consisting of 43-48% carbohydrates, 12-17% protein, and 35-40% fat. Human Research Review Committee approval and informed patient consent for the studies were obtained. All subjects were investigated during 1) an iv arginine infusion, 2) a normal meal ingestion, and 3) an oral glucose tolerance series with three different dosages of glucose. Experimental protocol Studies were initiated at 0600 h after an overnight fast with random activity permitted within the Clinical Research Center. After a stabilization period of 30 min, baseline venous blood samples were obtained at 10-min intervals for 30 min before the infusion of arginine or the ingestion of the meal or glucose. All subjects received a 30-min continuous infusion of L-arginine (10%) of 300 ml, with blood sampling at -30, -20, -10, 0, 5, 10, 20, 30, 40, 50, 60, 75, and 90 min. On alternate days, all subjects received a 412 calorie meal composed of eggs, bread, orange juice, jelly, and margarine containing 18 g protein, 50 g carbohydrate, and 15 g fat. After the baseline samples, blood was obtained at 10, 20, 30, 40, 60, 90,120, 150,180, 210, and 240 min. Investigation of the hepatic removal of insulin in response to progressively increased insulin exposure was performed during oral 10-, 25-, and 100-g glucose tolerance tests in all seven subjects. After the baseline studies and after glucose ingestion, blood samples were otained at 10, 20, 30, 45, 60, 90, 120, 150, 180, 210, and 240 min. Blood was immediately placed into heparinized tubes and the plasma separated by centrifugation. Double antibody RIA of plasma C-peptide was performed with antiserum directed against synthetic human C-peptide as previously reported (12). The intraassay coefficient of variation was 6%, and the detection limit varied between 0.05-0.10 ng/ml C-peptide in individual assays. Plasma C-peptide was assayed as nanograms per ml and converted to molar secretion based upon a mol wt of 3300. Molar secretion of C-peptide was converted to the secretion of

insulin on the basis of a mol wt of 6000 for insulin and the relationship of 1 /*U insulin for each 40 pg insulin. Plasma insulin was assayed by RIA and plasma glucose by glucoseoxidase. In determining the hormone production for each patient, the calculated plasma volume (41.3 ml/kg BW) (19) was used as the dilutional space for C-peptide. Kinetic analysis Endogenous insulin behavior throughout each tolerance test was determined using the three compartmental model previously defined from insulin kinetic analysis (13-15). As shown in Fig. 1, this model describes the hepatic plasma space [compartment-1 (cmpt)], the extrahepatic plasma space (cmpt-2), and the extravascular dilutional space (cmpt-3) as reported by Sherwin et al. (13). Whereas these investigators examined the kinetic events after tracer-insulin injection into cmpt-2, we have evaluated these events in response to endogenous insulin secretion into cmpt-1. Determination of this pancreatic prehepatic endogenous insulin secretion was performed by deconvolution as previously described in detail from our laboratory

Where X O ,,+A 2 .| =2.l7min" Pancreatic Insulin Production

0.0437

0,0216

(Variable) max=2,l7

0268

max=2.l7 (Variable)

0.07

FIG. 1. The three-cmpt model (hepatic, vascular, and extravascular pools) utilized in the kinetic analysis of the insulin. The figure provides the fractional turnover constants (min"1) derived from the literature (13-15) and identifies the variable parameters determined from these studies (Xo,i and A2,i).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 January 2017. at 06:28 For personal use only. No other uses without permission. . All rights reserved.

EATON, ALLEN, AND SCHADE

1296

utilizing plasma C-peptide concentration (12). In this procedure, a least squares spline fitting of the C-peptide data by interactive computer analysis permits evaluation of the precursor production of insulin using a two-cmpt model for C-peptide removal. The derivation of the dilutional compartments comes from published data concerning kinetic analysis of exogenous infusions of native insulin. The initial volume of distribution of insulin has been reported to be 45 ml/kg (13), and the plasma flow to the liver as 840 ml/min by Sherwin et al. (13). Their studies define the extravascular dilutional space as 95 ml/kg, and the rate constants for plasma flow between this compartment, the plasma, and the liver as shown in Fig. 1. We have extended the model to define the plasma space of the liver as 4.95 ml/kg in man based upon the data of Greenway and Stark (20). The model defines two pathways of insulin removal and/or degradation, one hepatic pathway (X0>i) and one nonhepatic pathway (70,2) (13). By different methodology Fugleberg et al. (21) have also proposed the existence of two independent processes for insulin removal from the plasma in man. Similar experiments in dogs suggest that the liver is one site and the kidneys, muscle, and adipose tissue a second site of insulin removal and degradation (22, 23). Renal insulin removal has been studied in man with direct arterio-venous sampling by Chamberlain and Stimmler (24). Based upon a renal plasma flow of 650 ml min"1, and an arterio-venous insulin extraction ratio of 0.29 ± 0.12, a minimum fractional insulin removal of 0.06 ± 0.03 min"1 can be defined. Similar arterio-venous sampling studies across forearm muscle and adipose tissue have variously reported the fractional removal of insulin to range from 0.005 min"1 to 0.01 min"1 (25-27). Our model thus describes net insulin fractional loss from the vascular compartment exclusive of hepatic removal as 0.07 min"1 representing renal, muscle, and adipose tissue removal. These removal functions are presumed to be constant over the course of each study, representing first order kinetic removal functions of insulin concentration presented to the organs (22). Evaluation of hepatic extraction of insulin To evaluate hepatic extraction of insulin, the model generated solution in cmpt-2 was fit to the experimental peripheral plasma insulin data by least squares criteria. The only model parameters which were variable in the fitting process were the fractional loss from the liver compartment (X0,i), and the fractional return to the plasma compartment (X2,i). They were constrained by the relation: X0>i + X2)1 = 2.17 min"1 where the plasma flow out of the liver (2.17 min"1 X Vi ml) equals the plasma flow into the liver (0.268 min"1 X V2 ml). A 2-min delay in the transit of insulin through the liver (28), and a 16-min delay in the transit through the extravascular space provide a significant improvement in the shape of the insulin solution. Calculations were performed using the KABIS computer program (29) on an IBM-3032 computer. A quantitative evaluation of the model-determined insulin solution relative to the plasma insulin data is given by the normalized residual error (29).

JCE & M • 1983 Vol56«No6

Results Response to arginine infusion (Table 2) (Fig. 2) Plasma C-peptide concentration rose in a biphasic fashion as shown in Fig. 2, left, from basal levels of 1.6 ± 0.3 ng/ml to a maximum of 4.3 ±0.1 ng/ml. Prehaptic insulin production rate increased from basal levels of 3.7 ± 0.8 /iU/ml • min, to a maximum of 32 ± 5 /LtU/ml • min during the initial 5 min of the arginine challenge. The secondary surge achieved insulin production of 16.2 ± 1.0 jiU/ml-min at approximately 20 min after the initiation of the arginine infusion (Fig. 2, middle). As shown in Fig. 2, right, basal insulin concentration averaged 7 ± 2 /iU/ml before arginine infusion, and rose to a maximum of 30 ± 5 /uU/ml during the infusion in a biphasic fashion. In this panel, the shaded area defines the SEM of the individual peripheral plasma insulin data from all seven studies, with the mean solution in cmpt2 of the model plotted at each time of sampling. The fit of the model solution to the experimental data has been determined with a mean normalized residual error of 17 ± 6% in the seven studies. (Table 2). The rate of fractional hepatic extraction of insulin averaged 1.26 ± 0.19 min"1, which represents 58% of the total hepatic insulin flux (portal-venous + hepatic arterial) irreversibly removed by the liver during arginine stimulation of endogenous insulin secretion. Cumulative insulin secretion from the pancreas averaged 2.3 ± 0.3 U in response to arginine, with a total hepatic flux of 3.2 ± TABLE 2. Kinetic parameters of insulin metabolism after arginine stimulation and meal ingestion

Patient

Prehepatic insulin secretion (U)'

Hepatic extraction rate (Xo,i) (min"1

Fractional hepatic extraction (%)

± SEM))

Arginine tolerance test 1 2 3 4 5 6 7 Mean ±SD

Meal ingestion 1 2 3 4 5 6 7 Mean ±SD

hepatic insulin flux (U)°

Normalized residual error of data solution (%)b

2.1 1.9 2.6 2.2 2.7 2.2 2.5 2.3 ±0.3

1.24 ± 0.03 1.55 ± 0.03 1.13 ± 0.06 1.02 ± 0.06 1.44 ± 0.05 1.31 ± 0.07 1.11 ± 0.03 1.26 ± 0.05 ±0.19 ±0.02

57 72 52 47 66 60 51 58 ±9

3.0 2.8 3.4 3.4 3.6 2.8 3.8 3.2 ±0.4

10 15 20 17 21 26 10 17 ±6

3.9 3.3 4.8 3.7 6.9 3.9 4.4 4.4 ±1.2

1.26 ± 0.04 1.02 ± 0.06 0.77 ± 0.03 0.96 ± 0.04 1.32 ± 0.06 1.09 ± 0.04 1.25 ± 0.06 1.09 ± 0.05 ±0.20 ±0.01

58 47 35 44 61 50 57 50 ±9

6.0 5.7 10.7 6.5 10.6 7.0 6.4 7.6 ±2.1

18 21 11 14 24 14 25 18 ±5

1

Units of flux during the time required to return to the basal level. ' Normalized residual error expressed as a percent of the plasma insulin data.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 January 2017. at 06:28 For personal use only. No other uses without permission. . All rights reserved.

1297

HEPATIC REMOVAL OF INSULIN FlG. 2. Model analysis of insulin behavior in seven normal subjects for an iv Arginine Tolerance Test. The plasma Cpeptide concentration (mean ± SEM) is shown in the top left. The pancreatic insulin production, as derived by deconvolution of the plasma C-peptide data, is shown in the middle. The mean plasma insulin concentration determined in cmpt-2 of the three-cmpt model is shown in the right, with the SEM of the data depicted by the shaded area. ARG. INF., Arginine infusion.

' Derived Pone reahc Insulin Production in eait

I 20.

v\ IW

|

I ~

60

Response to meal ingestion (Table 2; Fig. 3) As shown in Fig. 3A, plasma C-peptide concentration rose from basal levels of 1.6 ± 0.5 ng/ml to a maximum of 5.4 ± 0.8 ng/ml after the meal ingestion. Prehepatic insulin production rate increased from basal levels of 3.9 ± 0.5 /uU/ml-min, to a maximum of 17.5 ± 1.7 /iU/mlmin. As shown in B, basal insulin concentration averaged 7 ± 2 /zU/ml before ingestion of the 412 calorie meal, .

'

•PC

0.4 U due to the recycling of insulin to the liver via the hepatic artery and portal vein. Of the total hepatic insulin flux, 1.9 ± 0.3 U was quantitatively removed.

20

I

SEM)

Derived Poncreolie Insulin Production (meant SEM)

and rose to a maximum of 50 ± 14 /xU/ml during the absorption period. In Fig. 3C, the shaded area defines the SEM of the individual peripheral plasma insulin data from all seven studies, with the mean solution in empt2 of the model plotted at each time of sampling. The fit of the model solution to the data resulted in a mean normalized residual error of 18 ± 5% in the seven studies (Table 2). Cumulative insulin secretion from the pancreas averaged 4.4 ± 1.2 U with a total hepatic insulin flux of 7.6 ± 2.1 U or 2Va times that seen with the arginine infusion. The rate of fractional hepatic extraction of insulin averaged 1.09 ± 0.20 min"1, with a quantitative insulin removal of 3.8 ± 1.2 U. Thus, 50% of the total hepatic insulin exposure (portal-venous + hepatic arterial) was irreversibly removed by the liver during this mixed meal induced stimulation of endogenous insulin secretion. The derived hepatic plasma insulin concentration (cmpt-1) after the meal ingestion is graphically shown in Fig. 3D. The basal insulin concentration of 28 ± 3 /*U/ ml and the maximum concentration achieved of 116 ± 9 /xU/ml represent elevations several orders of magnitude above that of the peripheral plasma given for comparison inC. Dose-response to oral glucose ingestion (Table 3) (Fig. 4)

FIG. 3. Model analysis of insulin behavior in seven normal subjects during the ingestion of a normal mixed meal of 412 calories. The plasma C-peptide concentration (mean ± SEM) is shown in A. The pancreatic insulin production as derived by deconvolution of the plasma C-peptide data is shown in B. The mean plasma insulin concentration determined in cmpt-2 of the model is shown in C, with the SEM of the peripheral plasma data depicted by the shaded area. The derived hepatic plasma concentration of insulin determined in cmpt-1 of the model is shown in D as the mean ± SEM.

The initial fasting plasma C-peptide concentration in the seven subjects averaged 1.65 ± 0.2 ng/ml, and rose to a maximum of 3.7 ± 0.7, 4.2 ± 0.3, and 6.9 ± 0.7 ng/ ml, respectively, during the 10-, 25-, and 100-g glucose challenge. The solution for prehepatic pancreatic insulin production rate demonstrated a rise from basal production of 3.5 ± 0.3 AiU/ml-min to a maximum of 14.9,16.7, and 26.0 /iU/ml-min post ingestion of 10, 25, and 100 g glucose, respectively. Cumulative insulin secretion during the 300 min of observation above basal was 1.8 ± 0.5, 2.7 ± 1.1, and 7.2 ± 1.6 U, respectively, spanning the range observed with both arginine infusion and meal ingestion (Table 2). The 21 sets of the insulin dose-response to glucose were resolved utilizing the 3-cmpt model for insulin

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 January 2017. at 06:28 For personal use only. No other uses without permission. . All rights reserved.

EATON, ALLEN, AND SCHADE

1298

TABLE 3. Kinetic parameters of insulin metabolism during the dose response to oral glucose ingestion

Patient

Pancreatic insulin secretion (U)°

Hepatic extraction rate (Kx) 1 (min"

Fractional hepatic extraction (%)

±SEM))

Total hepatic insulin flux

Normalized residual error of data solution

/TT\a IU)

(%)

10 g oral GTT*

±0.5

1.59 ± 0.06 1.52 ± 0.07 1.51 ± 0.06 1.30 ± 0.05 1.49 ± 0.02 1.60 ± 0.04 1.20 ± 0.04 1.46 0.05 ±0.15 ±0.01

1.8 2.2 2.5 2.9 5.1 1.8 2.9 2.7 ±1.1

0.92 ± 0.04 1.46 ± 0.08 0.93 ± 0.09 1.05 ± 0.04 1.39 ± 0.03 1.26 ± 0.07 1.11 ± 0.06 1.16" 0.06 ±0.22 ±0.02

5.6 6.0

0.88 ± 0.05 1.02 ± 0.05 0.82 ± 0.04 0.84 ± 0.04 1.24 ± 0.06 0.93 ± 0.02 0.70 ± 0.03 0.92c-d 0.04 ±0.17 ±0.01

1 2 3 4 5

1.6 1.7 1.3 1.5 2.7

6

2.2

7

1.8 1.8

Mean ±SD

73 70 69 60 69 74 55 67

±7

2.0 2.3 1.8 2.1 3.6 3.1 3.1 2.6 ±0.7

27 28 23 16 11 18 14 20

±6

25 g oral GTT

1 2 3 4 5 6 7 Mean ±SD

41

67 43 48 64 58 51 53C ±10

3.4 3.1 4.5 5.2 7.3 2.5 4.9 4.4C ±1.6

13

33 17 13 11 24 22 19 ±8

100 g oral GTT 1 2 3 4

5 6 7 Mean ±SD

7.2

6.1 10.2 7.6 7.7 ±1.6

41 47 38 39

57 43 32 42^ ±8

10.3 10.0 14.2 12.0 15.5 13.8 16.6 13.2^ ±2.5

17 17 13 13 22

14 10 15 ±4

° Unite of flux during the time required to return to the basal level. 6 GTT, glucose tolerance test. c Significance of difference from results of 10 g oral G T T is P < 0.01 by paired t statistics. d Significance of difference from results of 25 g oral G T T is P < 0.01 by paired t statistics.

kinetic behavior (Fig. 1), and the mean solution to the plasma insulin data obtained within cmpt-2 shown in Fig. 3. A uniform distribution of the residual solution was pbtained throughout the entire study, with a mean normalized residual error of 20%, 19%, and 15% for the 3 sets of studies (Table 3). The rate of fractional hepatic extraction of insulin as determined from compartmental kinetic analysis, decreased progressively with an average of 1.46 ± 0.15 min"1, 1.16 ± 0.22 min"1, and 0.92 ± 0.17 min"1 in response to 10, 25, and 100 g oral glucose, respectively (P < 0.01). This phenomenon occurred in parallel with a progressive increase in the total hepatic flux of insulin which averaged 2.6 ± 0.7, 4.4 ± 1.6, and 13.2 ± 2.5 U, respectively, with increasing glucose administration. The minute flux of insulin through the liver (2.17 min"1 X Ci nV/ral) in comparison to that removed by the liver (X0,i min"1 x Ci /uU/ml) is shown in Fig. 4, bottom. With increasing insulin exposure, the total insulin extraction progressively increased from 1.7 ± 0.5 U

JCE & M • 1983 Vol56«No6

of the 2.6 U of total flux (10-g glucose ingestion), to 2.4 ± 1.1 U of the 4.4 U of total flux (25 g glucose), and to 5.4 ± 1.6 U of the 13.2 U of total flux (100 g glucose). However, when expressed as fractional hepatic insulin extraction of available insulin, a progressive decrease from 67% during the lowest insulin exposure, 53% at the intermediate exposure, and 42% with the maximum insulin exposure occurred. Discussion Our data obtained by compartmental analysis of the kinetic behavior of endogenous insulin secretion describe hepatic insulin extraction to be approximately 50 ± 9% during the ingestion of a standard mixed calorie meal. A similar degree of hepatic extraction of insulin was observed in man in 1959 by Madison et al. (30) by hepatic vein monitoring of portal I125-insulin infusion and confirmed with hepatic arterio-venous difference by Samols and Ryder (31). This value is similar to the 45% retention and degradation of insulin reported in the perfused rat liver (32). Moreover, from considerations of the in vitro fractional degradation velocity of insulin by isolated hepatocytes of 0.03 ± 0.011 min"1 (33), and the rate of dissociation of intact insulin from hepatocytes of 0.0385 ± 0.0075 min"1 (33), a net retention-degradation of 43.8% of insulin has been estimated for individual liver cells (32). Our in vivo data utilizing endogenous insulin and C-peptide thus extend these observations based upon nonhuman and tracer-labeled insulin studies in cells and the perfused liver, and upon tracer studies and hepatic A-V sampling in man. The present studies demonstrate decreasing insulin extraction by the liver (67-53-42%) in the presence of increasing insulin exposure (2.6-4.4-13.2 U) observed during the dose-response investigation of glucose ingestion. Similar qualitative changes in response to differing dosages of insulin achieved by sequential peripheral (34) as well as portal insulin infusion in man have been reported (7, 9, 10). Moreover, in the perfused rat liver fractional extraction of insulin has been similarly shown to decrease with increasing insulin exposure (5). The mechanism(s) underlying this phenomena are unresolved. Studies by Terris and Steiner (32) in the perfused rat liver suggest that uptake of insulin reflects the binding of the hormone to receptor sites on the plasma membrane of hepatocytes followed by enzymatic degradation of the insulin. However, examination of the time sequence of binding and degradation, demonstrates that the amount of insulin degraded increases more slowly than the total amount of insulin bound (32, 33). As reviewed by Terris and Steiner (32), this could represent (1) uptake by nonparenchymal cells with reduced degradation efficiency, or (2) reduced affinity of insulin

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 January 2017. at 06:28 For personal use only. No other uses without permission. . All rights reserved.

HEPATIC REMOVAL OF INSULIN

1299

Peripherol Plosmg Insulin Behovior

100 grom GTT 7.211.6 Units Insulin Secreted

FIG. 4. Plasma insulin concentration (SEM in shaded area) and the model solution in compt-2 of the dose-response to Oral Glucose Ingestion at 10, 25, and 100 g is shown in the top. The total minute flux of insulin flowing through the liver (2.17 min"1 x Ci jtU/ml) superimposed upon the minute flux of insulin extracted by the liver (Xo.imin"1 x Ci ml) is shown in the bottom.

67.5

135

203

270

0

67.5

135

203

270 0

67.5

135

203

270

Hepatic Plosmo Insulin Flux