flow chamber, Telstar CAM 7001. Insulin-binding and ..... centrations in liver. Biochem. Soc. Trans. 14,. I83-I93. KRuPP M. & LANE M.D. (I98I) Regulation of.
Br. J. Exp. Path. (I989) 70, I99-205
Lipopolysaccharide-induced insulin resistance in monolayers of cultured hepatocytes M. Teresa Portoles, Raffaella Pagani, M. Jesus Ainaga, Ines Diaz-Laviada and Angel M. Municio Department of Biochemistry
&
Molecular Biology I, Faculty of Chemistry, Universidad Complutense, 28040-Madrid, Spain
Received for publication 29 April I988 November I988 Accepted for publication io
Summary. In order to clarify the endotoxin effect on the hepatic removal of insulin, the influence of lipopolysaccharide (LPS) from E. coli oiii:B4 on the insulin binding and endocytosis in cultured hepatocytes from adult male rats has been investigated. LPS decreases both processes in a time and temperature-dependent manner, showing a major effect at short time and low temperature, according to the characteristics of LPS binding and uptake.
Keywords: endotoxic shock, lipopolysaccharide (E. coli), hepatocyte cultures, insulin binding, insulin resistance, insulin internalization
It is well known that endotoxaemia exerts a LPS with cell membranes. As we have premajor impact on metabolic requirements, viously shown, LPS from Escherichia coli and that some metabolic-endocrine de- O I1: B4 binds to the hepatocyte membrane rangements may represent the primary by means of hydrophobic and electrostatic defects in this type of shock. non-specific interactions (Pagani et al. During severe endotoxic shock, liver funcI98I). This fact has been related to an tion is impaired with a subsequent depresincrease of membrane microviscosity (Porsion of hepatic gluconeogenesis (Filkins & toles et al. I 9 87). Cornell 1974; Groves et al. I974; Cerra et al. Following the initial binding of [1251]insuI979). Thus, the supply of available glucose lin to the surface of isolated hepatocytes, the is often less than that required for normal hormone is progressively internalized by the metabolism. On the other hand, abnormal cell in a time and temperature-dependent insulin levels have been reported (Clowes et manner (Olefsky & Kao I982; Poole et al. I982). The uptake may serve as a means of al. I974; Hinshaw et al. I975) as well as ineffectiveness of insulin for the uptake of down-regulating the insulin-receptor conglucose. Increased insulin resistance in some centration (Krupp & Lane I98I; Draznin et al. I984). tissue has also been described (Clowes et al. In an attempt to clarify the molecular I978). Primary steps of endotoxicosis are some- mechanisms involved in the hormonal altehow mediated by the direct interaction of rations induced by the endotoxin at cellular Correspondence: Dr M.T. Portoles, Departamento de Bioquiniica y Biologia Molecular I, Facultad de Ciencias Quimicas, Universidad Complutense, 28040-Madrid, Spain. '99
M.T. Portols et al. level, the influence ofLPS on molecular bases phere for 24 h until the formation of a of insulin action has been studied using continuous monolayer. Cell culture was performed under sterile conditions in a laminar hepatocytes in monolayer cultures. flow chamber, Telstar CAM 7001. Materials and methods Insulin-binding and internalization assays Materials Hepatocyte monolayers, on glass circular Lipopolysaccharide from E. coli oiii:B4, micro-coverslips, were washed with Wilobtained according to Westphal's method liam's E medium (500 p1) and preincubated (Westphal et al. I952), was supplied by Difco in either the presence or the absence of LPS (I00 /2g/3 x IO' cells/500 MI William's E (Detroit, MI, USA). Unlabelled insulin was medium) during 30 min at 25TC, under C02/ obtained from Novo Espania SA (Spain). Monoiodinated ['25I]insulin (30 pCi/Pg) was 02 (5%:95%) atmosphere. The medium was then aspirated and the cells were washed purchased from Amersham International (UK). William's E medium was obtained from with William's E medium (500 !s) and incubated with a constant concentration of Flow Laboratories (Irvine, Ayrshire, UK), [1251]insulin (400 pg/3 x io5 cells) in either penicillin and streptomycin from Antibiotithe presence or the absence of increasing cos SA (Spain) and foetal calf serum from Gibco Europe (Paisley, Renfrewshire, UK). concentrations of unlabelled hormone according to the experiment. Incubations Bovine serum albumin and collagenase (type I) were from Sigma Chemical Co (St Louis, were carried out in William's E medium with MI, USA). Other chemical reagents were albumin (I%) for varying times and at different temperatures (4, 2 5 and 3 70C) purchased from Merck (Darmstadt, FRG). under C02/02 (5%:95%) atmosphere. The assay was stopped by aspirating incubation Animals medium and washing monolayers with WilMale Wistar rats (I50-200 g bodyweight), liam's E medium (500 M1) containing albufasted over 20 h with water ad libitum, were min (I%). used in all the experiments. The internalization of [125I]labelled insulin was determined by the acidification technique (Olefsky & Kao I982). Cell-associated Isolation and culture of hepatocytes radioactivity was extracted with acid by a Hepatocytes were isolated by the perfusion modification of the method of Haigler et al. technique, using collagenase in Krebs- (I980) to remove material bound to the cell Ringer bicarbonate solution (KRB-medium), surface. After stopping incubation and according to the method of Berry and Friend adding William's E medium (500 .1) with (I969) as described previously (Pagani et al. albumin (i%), monolayers were incubated with 2 ml of barbital sodium acetate buffer I98 I; Portoles et al. I98 7). Yield of viable hepatocytes was (pH 3 o), containing 28 mM Na acetate, 20 (I00 ± 20) X Io6 per rat liver. Cell viability, mM Na barbital and I I 7 mM NaCl, at 40C for determined by the trypan blue exclusion test, 6 min. was 85 to 95%. The acid-extractable radioactivity meaThe isolated hepatocytes, cultured in Wilsures the surface-bound insulin, and the liam's E medium and in the presence of I0% non-acid-extractable radioactivity correfoetal calf serum, were grown at 370C in sponds to the internalized hormone. Internatissue culture plates (24 flat-bottomed wells) lized plus surface-bound insulin gives the (3 X Io5 cells/well) with circular glass micro- total specific associated hormone. In these experiments, non-specific binding coverslips, under C02/02 (5%:95%) atmos200
20I LPS-induced insulin resistance is defined as the amount of [125I]insulin Table i. Effect of LPS on the insulin specific remaining in the indicated fractions (total binding at different temperatures. Cultured hepacell-associated radioactivity, non-acid-ex- tocytes (3 x IO' cells/500 pl William's E medium) preincubated with or without LPS (ioo jg/ tractable radioactivity and acid-extractable were 3 x I0o cells) during 30 min at 2 50C; medium was radioactivity), in the presence of a great aspirated and cells were then incubated with the excess (50 pg/3 x Io0 cells/5oo MI medium) labelled hormone (400 pg/3 x Iol cells) at differof unlabelled insulin. ent temperatures. Non-specific binding was estiMicrocoverslips were transferred to plastic mated as indicated under Materials and methods. tubes for counting in a Beckman counter Values are given as percentages ofthe control cells (Gamma 5500) to estimate the cell-asso- at every temperature assayed ciated radioactivity.
Statistics All assay points were performed in triplicate. Data given in figures and tables represent the mean values of three determinations ± standard deviations (s.d.). Results In order to study the influence of lipopolysaccharide from E. coli O I I I: B4 on the binding of insulin to cultured hepatocytes, monolayers were preincubated with endotoxin previously to the addition of the hormone. The non-specific binding was estimated as the amount of [125']insulin bound to cells in the presence of a great excess of unlabelled insulin. Subtraction of non-specifically bound hormone from the total radioactive uptake generated specific binding curves. As shown in Fig. i, preincubation with lipopolysaccharide decreases the insulin specific binding at every hormone dose assayed. This decrease was dependent on the endotoxin dose (unpublished results), and the relatively high amount of LPS (pharmacological dose) used in the assays is justified to obtain clear results within the time scale of the experimental models. Table i shows the effect of preincubation of hepatocyte cultures with LPS on the [1251]insulin specific binding at different temperatures and after go min incubation with the hormone. Values are given as percentages of the insulin binding to the control cells at every temperature assayed. The LPSinduced decrease on the insulin binding is
t(OC)
4
25
37
B(%)
4I±7
70±I4
I00±9
100 m 80
i5 60
: 40
C\\
20 N
_\
1 2 103
104 105 Insulin (pg/b06 cells)
106
Fig. I. Effect of LPS on insulin uptake by cultured hepatocytes. Cultured hepatocytes (3 x Io' cells/ 500 yUl William's E medium) were preincubated with (0) or without (0) LPS (IOO ug/3 x IO' cells) during 30 min at 250C; medium was aspirated and cells were then incubated with a constant concentration of ['251]insulin (400 pg/ 3 x Io' cells) and increasing concentrations of unlabelled hormone (o, 102, 03, I04, I05, 106, IO7 and io8 pg/To6 cells) for go min at 25°C. Values are percentages of specific binding of [1251]insulin referred to the binding control of the labelled hormone to the cells in the absence of unlabelled native insulin.
M.T. Portoles et al. more evident at 4, lower at 25, and disap- bound hormone) and internalized insulin to pears at 3 7C. the total insulin association. As incubation In order to know the LPS effect on the proceeds, total insulin association is less insulin internalization and its time and tem- affected by endotoxin, whereas the LPS effect perature dependence, various studies were on the surface-bound insulin disappears and carried out using the acid extraction tech- even becomes higher than control values. On nique. the other hand, a significant decrease of Figure 2 shows the total specific insulin insulin internalization produced by LPS is association, the surface-bound insulin, and observed at all incubation times. its internalized state in cultured hepatocytes The temperature dependence of this effect preincubated with LPS and incubated with of LPS on the internalization of insulin was ['251]insulin during different times at 37TC. investigated with hepatocyte cultures preinAt short incubation times (30 min) LPS cubated with endotoxin and incubated with clearly decreases the contribution of both the hormone. Figure 3 shows that the desurface-bound insulin (acid extractable crease of insulin internalization induced by LPS is more evident at low temperatures (40C) than at 3 70C. 202
tit)*
'" t',;,'.',i',t'.
X
J
, , 4 t. - .-J 4i
> > ! i^Jg~',;.9 . i;'4*'F"l{Ji t'; .'. 3
