Nitric Oxide Down-Regulates Endocytosis in Rat Liver Endothelial Cells

JOBNAME: BBRC 222#3 PAGE: 1 SESS: 10 OUTPUT: Thu Jun 6 18:45:42 1996/xypage/worksmart/tsp000/71605g/13

Nitric Oxide Down-Regulates Endocytosis in Rat Liver Endothelial Cells

Iñigo Martinez, Baldur Sveinbjørnsson, and Bård Smedsrød1

Institute of Medical Biology, University of Tromsø, N-9037 Tromsø, Norway

Received April 17, 1996

Hepatic sinusoidal endothelial cells (SEC) were challenged with inducers, blockers or donor of nitric oxide(NO) productionin vitro to test the effect of this reactive nitrogen agent on endocytosis in SEC. NO wasmeasured as its stable form nitrite (NO2

−) in culture media and trace amounts of radioiodinated ligands weretested for endocytosis and binding after 6 and 24 h of incubation. Among several proinflammatory reagentstested, IL-1b induced most strongly NO synthesis after 24 h in culture. Although endocytosis was significantlyenhanced in SEC that had been exposed to IL-1b for 6 h, prolonged exposure (24 h) to this proinflammatorycytokine, which triggered increased production of NO by the cells, yielded a decreased endocytic activity. Thepresence of aminoguanidine, an inhibitor of NO synthase, gave significant up-regulation of endocytosis com-pared with control cells. To fully verify the role of NO as an endocytosis modulator, SEC were preincubated withsodium nitroprusside, an exogenous NO donor. Again, increased concentration of NO2

− in the medium wasassociated with decreased endocytosis of SEC. Binding studies at 4°C revealed that the down-regulation ofendocytosis in SEC after increased exposure to NO was due to a decreased number or affinity of receptors onthe cell surface. © 1996 Academic Press, Inc.

A significant feature of liver sinusoidal endothelial cells (SEC), distinguishing them from en-dothelial cells of other tissues is their very well developed endocytic capacity (1). In fact, thesecells constitute the most important site of elimination of an array of circulating waste macromol-ecules. This relationship offers the possibility to use receptor-mediated endocytosis as a parameterto study the functional state of SEC. We have previously shown that these professional endocytesmay become stimulated and increase their endocytic capacity after exposure to inflammatorymediators like IL-1b and TNF-a (2). These findings correlate with the fact that during the firstphases of inflammation, the catabolism of cells and matrix molecules is elevated. Consequently theamount of waste macromolecules and harmful enzymes in blood increase and the scavengercapacity of SEC would need to be up-regulated.Nitric oxide (NO) is a non-specific signal molecule, secreted locally by most types of liver cells

after stimulation with proinflammatory cytokines or endotoxin (3–6). Elevated levels of NO inresponse to the first wave of inflammatory mediators may exert different effects on the surroundingcells. The effects of NO on the function of Kupffer cells, hepatocytes and liver in general have beenwell studied (7–10). Likewise the pattern of NO production by SEC under different physiologicalfunction of SEC, namely receptor-mediated endocytosis, remains to be elucidated.We here report results from studies on the effect of inducers of NO-synthesis like IL-1b and

NO-synthesis inhibitors like aminoguanidine or dexamethasone, and SNP as exogenous NO donoron the capacity of SEC to perform endocytosis via the collagen-, mannose- and scavenger-receptors. Our results show that SEC respond to increased concentrations of NO by lowering theirendocytic activity. This finding is compatible with the current idea that NO plays a role inresolution of inflammation.

1 To whom correspondence should be addressed. Fax: +47-77 64 54 00.Abbreviations: NO, nitric oxide; iNOS, inducible nitric oxide synthase; TNF-a, tumor necrosis factor-a; IL-1b, inter-

leukin-1b; SEC, sinusoidal endothelial cells; SNP, sodium nitroprusside; Agn, aminoguanidine; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazoleum.

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS222,688–693 (1996)ARTICLE NO. 0805

6880006-291X/96 $18.00Copyright © 1996 by Academic Press, Inc.All rights of reproduction in any form reserved.


MATERIALS AND METHODS

Chemicals.Collagenase, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazoleum (MTT), aminoguanidine (Agn), sodiumnitroprusside (SNP), sulphanilamide, naphtylenediamine dihydrochloride, and bovine serum albumin were purchased fromSigma Chem. Co., (St. Louis, MO, U. S. A.). Human serum albumin and Percoll were from CRTS (Lille, France) andPharmacia Fine Chemicals (Uppsala, Sweden). Human recombinant TNF-a (specific activity: 1 − 2 × 107 U/mg) andrecombinant human IL-1b (specific activity: 1 − 3 × 108 U/mg) were purchased from R&D Systems Europe Ltd (Abingdon,UK). Recombinant murine IFN-g (specific activity: 4.5 × 106 U/mg) was from Genzyme (Cambridge, MA, USA). Allcytokines contained less than 10 EU/mg LPS as determined by Limulus assay. Dexamethasone (water soluble), were fromSigma. LPS from S. enteritidis was purchased from Difco Laboratories, Detroit, MI, USA. Carrier-free Na125I was fromInstitutt for Energiteknikk, (Kjeller, Norway). Cell culture medium RPMI 1640 was purchased from Gibco (Grand Island,NY, USA).Ligands.Mannan was from Sigma. Purified bovine collagen was from Collagen Corporation (Palo Alto, USA). Form-

aldehyde-treated bovine serum albumin was prepared as described (11).Labeling techniques.The three types of ligands were labeled with125I by a direct reaction employing Iodogen (12). The

labeled proteins were purified free of unlabeled125I on a PD-10 column (prepacked Sephadex G-25 from Pharmacia) elutedwith phosphate-buffered saline. The resulting specific radioactivity was about 1 × 106 cpm/mg. Radioactivity was measuredusing a gamma-counter (Auto Gamma Scintillator Spectrometer; Packard Instruments, Warrenville, IL, U. S. A.).Isolation of SEC.Preparation of purified cultures of functionally intact SEC from rat liver has been described elsewhere

(13). Briefly, after collagenase perfusion of the liver, and isopycnic sedimentation of the resulting dispersed cells througha two-step density gradient of Percoll, pure monolayer cultures of SEC were established by selective attachment on asubstrate of fibronectin.Incubations with modifiers of NO-synthesis.90–95% pure monolayer cultures of SEC were established in 24-well dishes

(approx. 5 × 105 cells attached and spread per well). After 8 h of incubation in RPMI renewed every second h, the cells wereincubated with RPMI 1640 with 1% human serum albumin for 6 h or 24 h in thepresence of TNF-a (100 U/ml), IFN-g(50 U/ml), LPS (100 ng/ml), SNP (25mM), dexamethasone (1mg/ml), Agn (200mM) or IL-1b (100 U/ml) separately orin combinations specified in the Results section. Endocytosis and binding experiments, and assays of nitrite and viabilitywere carried out as explained below.Endocytosis studies.SEC cultures were washed and supplied with fresh medium containing 1% human serum albumin

and trace amounts of radiolabeled ligand (30.000 cpm per culture) in a total incubation volume of 250ml per well.Incubations of SEC with radiolabelled collagen or formaldehyde-treated serum albumin were terminated after 1 h bytransferring the media, along with one wash (0.5 ml) of phosphate-buffered saline, to tubes containing 0.75 ml of 20%trichloroacetic acid. This procedure precipitates only undegraded protein of high molecular weight. The extent of degra-dation was determined by measuring the radioactivities in pellet and supernatant after centrifugation. Cell-associated ligandwas quantified by counting radioactivity in washed cells following solubilisation in 1% sodium dodecyl sulphate. Totalendocytosis was obtained by adding cell-associated radioactivity and acid-soluble radioactivity in the medium. Since SECare unable to degrade mannan, this ligand accumulates intracellularly on endocytosis, and analysis of labelled degradationproducts in the medium was not necessary. Statistical significance was determined by Student’s pairedt test and P < 0.05taken as significant.Binding studies.Monolayers of SEC established in 2 cm2 dishes were allowed to bind radioiodinated ligands at 4°C for

2 h. After washing the cultures to remove unbound ligand, cultures were solubilized in 1% sodium dodecyl sulphate. Cellsurface bound ligand was measured as described above.Nitrite assay.NO2

−, the stable end product of NO was measured by a novel colorimetric assay (14). Briefly, 50ml aliquotsof medium were removed from individual wells and treated with an equal volume of Griess reagent (1% sulphanilamide and0.1% napthylenediamine dihydrochloride in 2.5% H3PO4) at room temperature for 10 minutes. The optical density of thesamples was recorded using a TiterTek Multiskan at 540 nm. A standard curve using NaNO2 in culture medium wasemployed as a standard solution for calculating the NO2

− concentration. All values are means ±S.D. of triplicate experi-ments.MTT assay.Viability of the cells in the assays for SNP, Agn, IL-1b and controls was tested by incubating SEC with MTT

for 2 h asdescribed (15).

RESULTS

In order to establish the optimal conditions for induction of NO synthesis by SEC, we performedpreliminary experiments by adding to SEC increasing concentrations of recombinant cytokinesand/or lipopolysaccharide (LPS) and by measuring the level of NO2

− in the culture media after 6and 24 h of incubation (table 1). After 6 h, very low levels of NO2

− could be measured in themedium with hardly any differences in NO2

− production observed. However, 24 h challengeshowed that IL-1b, TNF-a and IFN-g induced NO production, IL-1b being the most effective,

Vol. 222, No. 3, 1996 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

689


whereas LPS alone or in combination with the other three cytokines tested did not result in a moremarked synthesis of NO2

−.Next, we investigated the effect of nitric oxide synthase inhibitors and exogenous donor of NO.

Exposure of the cells to Agn (200mM) or dexamethasone (1mg/ml) resulted in blocking ofinducible NO synthase, giving marked differences in NO2

− levels with untreated cells (table 2). Ofnote, SEC produced considerable amounts of NO2

− after 24 h in culture in the absence of thesereagents. As expected, exposure to the NO donor SNP (25mM) resulted in large levels of NO2

−

after the same incubation period.To ascertain if variation in the NO levels could modulate the endocytic activity in SEC, we

measured the uptake of radioiodinated mannan via the mannose-receptor after treatment of the cellswith the NO-production modifiers. After 6 h, when little or no differences in NO2

− levels wereobserved, cultures treated with IL-1b and dexamethasone showed enhanced endocytosis, whereasAgn and SNP had no effect (Fig. 1A). On the other hand, endocytosis performed after a 24 hexposure was significantly decreased with SNP and increased with Agn. In addition, endocytosisin cultures treated with IL-1b for 24 h was decreased compared with those treated with IL-1b foronly 6 h (Fig. 1B).To determine if the decreased endocytosis observed upon exposure to increased levels of NO

was due to variation in the receptor expression on the cell surface, we performed binding studiesat 4°C with radiolabeled mannan for 2 h (fig. 2). These experiments revealed similar NO-mediatedmodulation patterns with SNP and Agn as those observed at 37°C, indicating that the receptorexpression is down-regulated in the presence of increasing concentrations of NO.The same endocytosis and binding experiments as those performed with mannan were performed

with ligands for the collagen and scavenger receptors of SEC. The effects of NO on the endocytosisand binding via all three receptors were similar (data not shown).

TABLE 1Effect of Cytokines and LPS Alone or in Combination on NO Generation by SEC

Nitrite (mM)

Stimulatory agents 6 h treatment ± S.D. (n4 3) 24 h treatment ± S.D. (n4 3)

Control 7.1 ± 1.0 22.6 ± 1.3IL-1b 10.2 ± 1.4 37.0 ± 1.3TNF-a 7.8 ± 0.6 34.5 ± 2.7IFN-g 8.3 ± 0.8 33.3 ± 1.4LPS 8.8 ± 0.5 28.0 ± 1.9IL-1b + LPS 7.8 ± 0.8 36.1 ± 1.9TNF-a + LPS 8.3 ± 0.9 35.2 ± 1.5IFN-g + LPS 6.9 ± 0.8 33.3 ± 1.1

TABLE 2Effect of Inhibitory Agents of NO Synthesis on Untreated and IL-1b treated SEC and NO

Generation by SNP

Nitrite (mM)

NO-production modifiers 6 h treatment ± S.D. (n4 3) 24 h treatment ± S.D. (n4 3)

Control 7.1 ± 1.0 22.6 ± 1.3Dexamethasone 5.2 ± 0.8 9.2 ± 0.1Agn 5.9 ± 1.4 10.0 ± 0.6IL-1b + Dexamethasone 8.3 ± 0.5 10.7 ± 0.9IL-1b + Agn 9.0 ± 0.9 13.8 ± 0.5SNP 13.1 ± 1.1 43.1 ± 2.2


690


None of the treatments reported here changed viability of SEC as determined by the MTT assays(data not shown).

DISCUSSION

NO is a major secretory product of mammalian cells with critical functions in homeostasis andimmune functions (16). In addition to a low-level constitutive production of NO by SEC, these cellsas well as Kupffer cells and hepatocytes may increase their production of NO in response toproinflammatory agents (3–6). We found that IL-1b was the most effective agent in increasing theNO production by SEC after a 24 hour challenge. Therefore this cytokine was chosen as inducerof NO synthesis in most experiments. IL-1b is a potent biologic signal molecule secreted in the firstphase of inflammation. We have previously reported that exposure of SEC to IL-1b for 6 h triggersenhanced endocytosis of different ligands (2). However, prolonged exposure for 24 h resulted inweaker stimulation of endocytosis, suggesting that increased concentrations of NO in the culturemedium exert opposite effect than proinflammatory cytokines.Establishment of cells in culture and the maintenance in vitro may by itself alter the homeostasis

of the cells. This may result in basal endogenous production of IL-1b and other factors andsubsequent activation of iNOS, resulting in increased levels of NO in the medium. This hypothesiswas tested by measuring production of NO and rate of endocytosis in control vs. Agn treated SECafter cultivation for 24 h. Indeed, blocking of NO synthesis by Agn resulted in enhanced endocyticrate, supporting the idea that NO acts by lowering the endocytic capacity of SEC.Glucocorticoids block the iNOS expression at the transcriptional level in most biological systems

(17,18). As expected, treatment with dexamethasone reduced production of NO to very low levels.Surprisingly, after 6 hours of incubation, when no significant differences in the levels of NO-

FIG. 1. Effect of NO-generating and NO-inhibitory agents on endocytosis of mannan. Uptake for 1 h of 125I-mannan,a specific ligand for mannose receptor, was determined in cultured SEC following a 6 h (A) or 24 h (B)exposure of thecells to Agn (200mM), SNP (25mM), dexamethasone (DX, 1mg/ml), or IL-1b (IL-1, 100 U/ml) alone or together withAgn. Endocytosis of mannan was decreased after 24 h in cells with high levels of NO2

− in the medium and increased inthe presence of low levels of NO2

−. Differences between the two time-points were statistically significant (*p < 0.01 and**p < 0.05). Bars are S.D. (n4 3).


691


production were observed between control cells and dexamethasone treated cells, we noted amarked increase of endocytosis in dexamethasone treated cultures, suggesting that glucocorticoidslike dexamethasone, in addition to blocking the synthesis of NO, exert a direct stimulation ofendoctosis in SEC, almost as fast and potent as proinflammatory mediators do.Endocytosis of several radio-iodinated ligands was studied to determine if modulation of NO

production was related to receptor specificity. The finding that endocytosis via the mannose-,collagen- and scavenger receptors was similarly affected by NO, suggests that NO affects theendocytosis independently of the receptor-type. Binding experiments performed at 4°C to studywhether the receptor expression was affected by NO, showed that the number of receptors on thecell surface is subject to modulation of NO synthesis. We therefore conclude that the NO-dependent decrease of endocytosis in SEC is caused by a decrease in the receptor-number or-affinity, rather than a decreased rate of ligand internalisation.Some authors have postulated that NO exerts autocytotoxic effect on different cell types after in

vitro exposure to NO donors or inducers (19,20). We therefore evaluated cell viability afterincubation with NO-synthesis modifiers by the MTT viability assay. The results of such testsindicated no changes in the viability of the cells. However, our studies with SNP were restrictedto concentrations of 25mM or less, because higher concentrations caused cell death in a dose-dependent manner.At least 50% of the IFN-g-triggered down-regulation of the mannose-receptor expression in

macrophages is due to generation of NO (21). More, NO inhibits leukocyte adhesion to theendothelium, partially by decreasing the expression of adhesion molecules (22). These reports,along with the results from the present study indicate that NO is a signal molecule involved inresolution of inflammation.

ACKNOWLEDGMENTS

This project was supported by the Norwegian Cancer Society (Grant 88 100) and the Norwegian Research Council (Grant312.92/017).

REFERENCES

1. Smedsroød, B., Pertoft, H., Gustafson, S., and Laurent, T. C. (1990)Biochem. J.266,313–327.2. Martinez, I., Sveinbjørnsson, B., Vidal-Vanaclocha, F., Asumendi, A., and Smedsroød, B. (1995)Biochem. Biophys.

Res. Commun.212,235–241.

FIG. 2. Effect of NO-production modifiers on cell-surface receptor expression. Binding experiments for 2-h at 4°C with125I-mannan were performed following 24 h of incubation of SEC with SNP (25mM) or Agn (200mM). Amount of boundligand was significantly enhanced with the iNOS inhibitor Agn and reduced with the exogenous NO donor SNP (*p < 0,01).Bars are S.D. (n4 3).


692


3. Laskin, D. L., Heck, D. E., Gradner, C. R., Feder, L. S., and Laskin, J. D. (1994)J. Leukocyte Biol.56, 751–758.4. Spitzer, J. A. (1994)Hepatology19, 217–228.5. Shshima, H., Oguchi, S., Adachi, M. I., Suzuki, H. S., and Esumi, M. (1992)Biochem. Biophys. Res. Commun.183,

238–244.6. Schmidt, H. H. H. W., Pollock, J. S., Nakare, M., Gorsky, L. D., Förstermann, U., and Murad, F. (1991)Proc. Natl.

Acad. Sci. USA88, 365–369.7. Harbrecht, B. G., and Billiar, T. R. (1995)Shock3, 79–87.8. Milbourne, E. A., and Bygrave, F. L. (1995)Cell. Signal7, 313–318.9. Kurose, I., Kato, S., Ishii, H., Fukumura, D., Miura, S., Suematsu, M., and Tsuchiya, M. (1993)Hepatology18,

380–388.10. Nishida, N., McKuskey, R. S., Mc Donnel, D., and Fox, E. S. (1994)Am J. Physiol.267,G1135–G1141.11. Mego, J. L., Bertini, F., and Mc Queen, D. J. (1967)J. Cell Biol.32, 699–707.12. Fraker, P. J., and Speck, J. C. (1978)Biochem. Biophys. Res. Commun.80, 849–855.13. Smedsrød, B., and Pertoft, H. (1985)J. Leukocyte Biol.38, 213–230.14. Ding, A. H., Nathan, C. F., and Stuehr, D. J. (1991)J. Immunol.141,2407–2412.15. Mosmann, T. (1983)J. Immunol. Methods65, 55–63.16. Nathan, C. (1992)FASEB J.6, 3051–3064.17. Radomski, M. W., Palmer, R. M. J., and Moncada, S. (1990)Proc. Natl. Acad. Sci. USA87, 10043–10047.18. O’Connor, K. J., and Moncada, S. (1991)Biochem. Biophys. Acta1097,227–231.19. Volk, T., Ioannidis, J., Hensel, M., de Groot, H., and Kox, W. J. (1995)Biochem. Biophys. Res. Commun.213,

196–203.20. Estrada, C., Gomez, C., Martin, C., Moncada, S., and Gonzalez, C. (1992)Biochem. Biophys. Res. Commun.186,

475–482.21. Shepherd, V. L., Cowan, H. B., Abdolrasulnia, R., and Vick, S. (1994)Arch. Biochem. Biophys.312,367–374.22. Kubes, P., Suzuki, M., and Granger, D. N. (1991)Proc. Natl. Acad. Sci. USA88, 4651–4655.


693

Documents

Nitric Oxide Down-Regulates Endocytosis in Rat Liver Endothelial Cells