- Email: [email protected]
Hepatocyte polyunsaturated fatty acid enrichment increases acute phase protein synthesis
Hepatocyte polyunsaturated fatty acid enrichment increases acute phase protein synthesis Brian Nolan, MD, James Sentementes, BS, and Paul Bankey, MD, PhD, Worcester, Mass, and Dallas, Texas
Background. The production of acute phase proteins by the liver is critical for homeostasis and recovery after injury. Polyunsaturated fatty acids, specifically of the n-3 family, have demonstrated anti-inflammatory properties that promote recovery; however, the effects of these fatty acids on acute phase protein synthesis have not been fully evaluated. Methods. The synthesis of the acute phase proteins α-2 macroglobulin and lipopolysaccharide binding protein was studied in hepatocyte-Kupffer’s cell cocultures from Wistar rats. Cultures were endotoxin stimulated after enrichment with albumin complexed arachidonic acid (20:4n-6) or docosahexaenoic acid (22:6n-3). Protein synthesis was analyzed by [35S]-methionine labeling or Western blotting. Culture interleukin-6 levels were determined. Results. Both polyunsaturated fatty acids increase hepatocyte synthesis of acute phase proteins α2macroglobulin and lipopolysaccharide binding protein compared with controls; however, the response in the docosahexaenoic acid (22:6n-3) treated cultures was significant (P < .05 vs control). Interleukin-6 was also increased in the polyunsaturated fatty acid cultures compared with controls (P < .05 vs control). Cellular phospholipids were significantly enriched with the individual supplemented fatty acids (P < .05 vs bovine serum albumin). Conclusions. Polyunsaturated fatty acids have the capability to increase in vitro acute phase protein synthesis. This may contribute to the observed anti-inflammatory effect of n-3 polyunsaturated fatty acid enrichment. (Surgery 1998;124:471-6.) From the Departments of Surgery and Cell Biology, University of Massachusetts, Worcester, Mass, and University of Texas, Southwestern Medical Center, Dallas, Texas
THE OPTIMAL COMPOSITION of essential fatty acids to provide surgical patients with sepsis or other systemic inflammatory diseases remains controversial.1 Although required to prevent nutrient deficiencies and serving as an alternative caloric source during stress, specific polyunsaturated fatty acids (PUFAs) from the omega-3 (n-3) family have the additional property of modulating inflammation.2-4 The effects on inflammation are multifactorial and include alterations in cytokine production, microvascular perfusion, and eicosanoid generation.5,6 These fatty acids are incorporated into cellular phospholipid pools including the plasma Supported by National Institutes of Health grant R29 GM51059-02 (P.E.B.). Presented at the Fifty-ninth Annual Meeting of the Society of University Surgeons, Milwaukee, Wis, Feb 12-14, 1998. Reprint requests: Paul Bankey, MD, PhD, Department of Surgery, University of Massachusetts Medical Center, 55 Lake Ave N, Worcester, MA 01655-0333. Copyright © 1998 by Mosby, Inc. 0039-6060/98/$5.00 + 0
11/6/91339
membrane and modulate physical properties of the lipid bilayer and receptor-mediated signal transduction.7,8 Sepsis induces systemic inflammation with a high incidence of organ dysfunction. The acute phase response, including the synthesis of proteins with anti-inflammatory and homeostatic functions by hepatocytes, is a critical determinant of host survival.9,10 Interleukin-6 and glucocorticoids are the primary regulatory signals to the hepatocyte to produce acute phase proteins.9 The effects of n-3 PUFAs on hepatocyte acute phase protein synthesis have not been fully evaluated. Enhanced acute phase protein synthesis may have a short-term beneficial anti-inflammatory effect during sepsis. The purpose of this study was to evaluate the effects of PUFA enrichment on hepatocyte production of acute phase proteins using an in vitro model. MATERIAL AND METHODS Animals. Wistar rats were handled according to the guidelines in the “Guide for use of Laboratory Animals,” published by the National Institutes of SURGERY 471
472 Nolan, Sentementes, Bankey
Fig. 1. Autoradiogram representative of culture protein synthesis after individual fatty acid enrichment and LPS stimulation. Synthesis of individual secretory proteins is altered by prior enrichment with PUFAs.
Health, Bethesda, Md (1985). This protocol was approved by the Institutional Review Board for Animal Research of the University of Massachusetts Medical Center. Cell isolation. Hepatocytes are isolated by collagenase perfusion as described previously.11,12 Briefly, the portal vein is cannulated, and then the liver is perfused in situ first with a washout buffer and then switched to 0.05% collagenase (Sigma #C-0130, 200 units/mg; Sigma). The liver is removed, and cells are combed gently in tissue culture medium. Hepatocytes are pelleted and washed, and viability is assessed by trypan blue exclusion. Kupffer’s cells are obtained from separate animals using 0.2% Pronase (Sigma) perfusion through the portal vein in situ. The liver is removed, minced into 5-mm cubes, and then incubated with constant stirring in 0.2% Pronase E (type XIV; Sigma) with DNase added at 20 and 40 minutes (Merck Chemical, Rahway, NJ). Temperature is maintained at 37° C, and pH is maintained at 7.4 to 7.6. The resulting digest is purified by a single centrifugation through 2-step Percoll gradient (Pharmacia). After a 3-hour incubation in RPMI 1640 medium with 5% calf serum (low endotoxin; Gibco) 80% of the adherent population demonstrates phagocytosis of 1.1-mm latex beads indicating Kupffer’s cells.11 Cell culture. Hepatocytes are seeded onto colla-
Surgery August 1998
gen coated plates on day 1 in Williams E medium supplemented with antibiotics, 10% fetal calf serum, HEPES, and insulin. The Kupffer’s cells are isolated the following day and added to cultured hepatocytes at a 2:1 ratio (HC:KC). The cultures are recovered overnight. The medium is removed, and fatty acid containing medium is added for 8 hours. This medium is removed, and the cultures are stimulated with lipopolysaccharide (LPS) (Escherichia coli O111:B4) for 18 hours. Supernatants are collected for analysis, and selected cultures are radiolabeled with [35S]methionine. Measurement of protein synthesis by [35S]-methionine labeling and sodium dodecyl sulfate–polyacrylamide gel electrophoresis or Western blotting. Cells are washed with methionine-free RPMI and then labeled in medium consisting of methionine-free RPMI supplemented with glutamine (15 mmol/L) and 100 µCi/ml [35S]-methionine. After a 6-hour labeling period, the supernatants are collected, pooled, and stored. Label incorporation into secreted protein is determined by precipitation of the sample with cold 20% trichloroacetic acid and scintillation counting. Specific protein secretion is detected on autoradiograms after separation by 8% sodium dodecyl sulfate–polyacrylamide gel electrophoresis.13 The autoradiograms demonstrate a band at 68 kd corresponding to albumin and at 180 kd corresponding to α2-macroglobulin. Unlabeled culture supernatants are evaluated for α2macroglobulin and LPS-binding protein (LBP) by Western blotting. The primary antibody for α2macroglobulin is goat anti-mouse obtained from Accuate, Inc. The primary antibody for LBP is rabbit anti-rat generated by Genosys, Inc, on the basis of a peptide deduced from DNA sequence. Blotted proteins were detected by first incubating the nitrocellulose in a 1:200 dilution of antibody and washing according to the Bio-Rad Immuno-Blot Assay Kit (Bio-Rad, Richmond, Calif). Next, unbound antibody was washed away and the nitrocellulose was incubated with 1:1000 dilution of goat anti-rabbit or anti-goat immunoglobulin G (H+L) conjugated with alkaline phosphatase (Jackson Immunoresearch Laboratories, Inc). Band densities are compared using the Gel-Doc system from Bio-Rad.12,13 Fatty acid–albumin complexes. A 5% solution of bovine serum albumin (BSA) (fatty acid free, low endotoxin) is added to the Na salt of the fatty acid in a concentration of 0.6 mg fatty acid/mL BSA. The fatty acid–BSA solution is allowed to complex overnight at 37° in a sealed tube. This gives a molar ration of 2.24:1 of fatty acid to albumin.14 Analysis of fatty acid enrichment. Total lipids are extracted by using chloroform/methanol, 2:1
Nolan, Sentementes, Bankey 473
Surgery Volume 124, Number 2
Fig. 2. Western blot representative of culture supernatant content of the acute phase α2-macroglobulin after fatty acid enrichment and LPS stimulation. OLA, Oleic acid.
Table. Cell phospholipid fatty acids (% total) after culture treatment Fatty acid 18:1n-9 18:2n-6 20:4n-6 22:6n-3 Total PUFA
BSA
Oleic acid
ArA
3.8 ± 0.5 16.3 ± 0.3 28.1 ± 0.4 5.2 ± 0.2 47.9 ± 0.9
7.7 ± 0.5* 15.9 ± 0.6 26.7 ± 0.7 4.9 ± 0.1 42.1 ± 1.1
3.6 ± 0.5 14.8 ± 0.7 33.9 ± 1.0* 4.0 ± 0.3 55.5 ± 1.7*
DHA 3.5 ± 0.6 14.1 ± 0.5 25.6 ± 0.9 9.8 ± 0.4* 54.3 ± 1.9*
*P < .05 vs BSA treatment.
(vol/vol). The total cellular phospholipids are separated on TLC plates (0.5 mm silica gel) by using the solvent system of petroleum ether/diethyl ether/acetic acid, 80/20/1 (v/v/v). The phospholipids are recovered and fatty acids are esterified by addition of 14% BF3/methanol (wt/vol) before separation by gas chromatography with known methyl ester standards by using a model 428, HewlettPackard system.15 Interleukin-6 assay. Interleukin-6 (IL-6) activity in serum and culture supernatants was measured in the 7TD1 cell proliferation assay.13 The optical density of the samples was compared with a standard curve generated from serial dilutions of murine recombinant IL-6 (Genzyme). Statistical analysis. Means and SDs for each treatment group were calculated. In experiments in which multiple samples were assessed, analysis of variance (F ratio and mean square for error) was used for comparing the data. Tukey’s test was used to perform multiple comparison tests with .05 error rate. Statworks (Cricket Software, Philadelphia, Pa) was used for statistical calculations. RESULTS Effect of fatty acids on cellular phospholipids. Cocultures (HC:KC) incubated for 8 hours in albumin alone (BSA, Table I) had a total PUFA content of 47.9%. In cultures incubated with 10 µg/mL
arachidonic (ArA, 20:4n-6) or docosahexaenoic (DHA, 22:6n-3) acid the total PUFA content increased significantly to 55.5% and 54.3%, respectively. Arachidonic acid increased to 33.9% from 28.1% and DHA to 9.8% from 5.2% (Table). These findings indicate that individual fatty acids are incorporated into cellular phospholipids after only 8 hours in cell culture. Alterations in secretory protein synthesis after fatty acid incubation. There were no significant differences in the rate of protein synthesis as determined by the incorporation of [35S]-methionine in unstimulated cultures treated with the fatty acids (data not shown). Rates of individual protein synthesis were significantly altered between fatty acid treatment groups in LPS-stimulated cultures as shown in the autoradiogram in Fig 1. The PUFA-treated cultures, both ArA and DHA, demonstrated an increase in labeled α2-macroglobulin and decrease of labeled albumin compared with the oleic acid and BSA treated cultures. These results suggest a direct effect of PUFA enrichment on hepatocyte protein synthesis in LPS-stimulated cocultures. Effect of fatty acid treatment on synthesis of the acute phase proteins α-2 macroglobulin and LPSbinding protein. The PUFA-treated cultures also demonstrated increased synthesis of both acute phase proteins α2-macroglobulin and LBP as determined by Western blotting (Fig 2 for α2-macroglob-
474 Nolan, Sentementes, Bankey
Surgery August 1998
Fig. 3. Western blot representative of culture supernatant content of the acute phase protein LPS-binding protein after fatty acid enrichment and LPS stimulation. OIA, Oleic acid.
Fig. 4. Effect of fatty acid treatment on culture acute phase protein levels as determined by Western blotting. Data are expressed as fold increase over non-LPS stimulated cultures. Blots were compared by densitometry and averaged over n = 3 to 4 experiments. P < .05 vs oleic acid (OLA) response.
ulin and Fig 3 for LBP). Densitometry on the Western blots demonstrated a 4.2- ± 1.1-fold increase in α2-macroglobulin in LPS-stimulated oleic acid treated cultures (3.9- ± 0.9-fold increase in BSA only) but that ArA-treated cultures had a 6.3- ± 2.4-fold increase and DHA-treated cultures had a 8.4- ± 2.9-fold increase. LBP synthesis was also increased significantly in the DHA-treated cultures compared with oleic acid (DHA: 6.1 ± 2.1 vs ArA: 4.0 ± 1.7 vs oleic acid 2.7 ± 0.8 fold vs BSA 2.9 ± 0.6). These data are shown graphically in Fig 4. These results indicate a differential acute phase protein response in cultures enriched with DHA and PUFAs compared with nonenriched cultures or cultures enriched with the monounsaturated fatty acid oleic acid. Effect of fatty acid enrichment on culture IL-6 activity. Cocultures enriched with both PUFAs had an increased production of bioactive interleukin-6 in response to stimulation with 100 ng/mL LPS compared with oleic acid enrichment (Fig 5). The increased production of IL-6 may contribute to the
increased synthesis of acute phase proteins in these treatment groups. DISCUSSION Using an in vitro model of acute phase protein synthesis, we have demonstrated that cocultures of hepatocytes and Kupffer’s cells enriched with PUFAs synthesize more α2-macroglobulin and LBP and less albumin than cultures treated with albumin alone or the monounsaturated fatty acid oleic acid. Associated with the increased acute phase protein synthesis is an increase in the production of the regulatory cytokine IL-6. Furthermore, treatment with the n-3 PUFA docosahexaenoic acid, DHA 22:6n-3, resulted in greater although not statistically significant increases in the acute phase proteins compared with the n-6 PUFA arachidonic acid, ArA 20:4n-6. Our experiments showed consistent up-regulation of the acute phase proteins α2-macroglobuin and LBP after enrichment of PUFAs. Other investigators report varied effects of PUFAs on individual acute phase proteins. In one study IL-6–stimulated
Nolan, Sentementes, Bankey 475
Surgery Volume 124, Number 2
human hepatocytes had increased production of prealbumin and decreased production of haptoglobin after treatment with eicosapentaenoic acid, EPA 20:5n-3.16 Both α2-macroglobulin and LBP are IL-6 up-regulated proteins. Further experimentation is required to determine whether the effects of PUFAs are specific to IL-6 up-regulated acute phase proteins or also have effects on IL-1 and tumor necrosis factor (TNF) regulated acute phase proteins.17 Our data would indicate that this effect is not specific to n-3 PUFAs such as DHA but is a generalized response to PUFA enrichment because it also occurs in response to ArA, the n-6 PUFA used in these experiments. Polyunsaturated fatty acid enrichment with both DHA and ArA resulted in increased IL-6 production in the cocultures compared with oleic acid and albumin alone. Other investigators have reported a differential effect of n-3 PUFAs compared with n-6 PUFAs on in vitro macrophage cytokine production.18,19 Previous investigation of Kupffer’s cell cytokine production after a fish oil diet, high n-3 PUFA, compared with a safflower oil diet, high n-6 PUFA, failed to show differential effects on LPS-stimulated TNF and IL-1 production also.6 It is not clear from these studies whether the increased acute phase protein synthesis reflects effects of the PUFAs on the hepatocyte or Kupffer’s cells or both. Our cellular phospholipid data reflect the composition of the cocultures, not individual cell populations. It is possible that our findings reflect only modulation of function in the Kupffer’s cells and that the hepatocyte is responding to an environment with increased levels of cytokines including IL-6. Hepatocyte enrichment with n-3 PUFAs has been demonstrated to alter lipoprotein production and cytokine signaled acute phase proteins.20,21 The effect of DHA and ArA on hepatocyte cultures only is unknown. The mechanism by which PUFAs alter LPS-stimulated acute phase protein production is unclear. Incorporation into cellular membranes in either the Kupffer’s cell or hepatocyte may influence membrane receptor function by altering membrane fluidity or the coupling of receptor and membrane proteins responsible for signal transduction. Fatty acids have previously been reported to have no effect on the number of TNF receptors expressed on hepatocytes but may influence the affinity of TNF for its receptor.22 The hepatocyte has the capacity to metabolize fatty acids by using intracellular eicosanoid pathways, and these metabolites may have a role in acute phase protein synthesis as they do in cytokine synthesis by Kupffer’s cells. Our current understanding of IL-6 signaled acute phase protein synthesis indicates receptor induced dimeriza-
Fig. 5. Effect of enrichment of cultures with individual fatty acids on LPS-stimulated IL-6 bioactivity. Data are expressed as mean ± SEM for 5 to 6 experiments. P < .05 vs BSA alone. OLA, Oleic acid.
tion of gp130 followed by activation of signal transducers and activators of transcription or acute phase response factor nuclear transcription factors. The effect of PUFA enrichment on these responses is unknown.9,23 We have demonstrated by using an in vitro hepatocyte–Kupffer’s cell coculture model that PUFA enrichment before LPS stimulation results in an enhanced acute phase protein synthetic response that is associated with increased levels of regulatory IL-6. Because significant levels of these fatty acids are components of enteral and parenteral nutritonal support regimens for patients with sepsis, this may be a potential pathway to modify the systemic inflammatory response during sepsis. REFERENCES 1. Grimm H, Tibell A, Norrlind B, Blecher C, Wilker S, Schwemmle K. Immunoregulation by parenteral lipids: impact of the n-3 to n-6 fatty acid ratio. JPEN 1994;18:417-21. 2. Cerra F, Bankey P, Holman R, Mazuski J, LiCari J. Omega-3 polyunsaturated fatty acids as modulators of cellular function in the critically ill. Pharmacotherapy 1991;11:71. 3. Blok W, Katan M, van der Meer J. Modulation of inflammation and cytokine production by dietary n-3 fatty acids. J Nutr 1996;126:1515-33. 4. Calder P. N-3 Polyunsaturated fatty acids and immune cell function. Adv Enzyme Regul 1997;37:197-237. 5. Abbate R, Gori A, Martini F, Brunelli T, Filippini M, Francalanci I, et al. n-3 PUFA supplementation, monocyte PCA expression and interleukin-6 production. Prostaglandins Leukot Essent Fatty Acids 1996;54:439-44. 6. Billiar T, Bankey P, Svingen B, Simmons R, Cerra F. Fatty acid intake and kupffer cell function: fish oil alters eicosanoid and monokine production in vitro. Surgery 1988;104:343. 7. Abel S, Gelderblom W, Smuts C, Kruger M. Thresholds and kinetics of fatty acid replacements in different cellular compartments in rat liver as a function of dietary n-6/n-3 fatty acid content. Prostaglandins Leukot Essent Fatty Acids 1997;56:29-39.
476 Nolan, Sentementes, Bankey
8. Clarke S, Jump D. Polyunsaturated fatty acid regulation of hepatic gene transcription. Lipids 1996;31:S7-11. 9 . Moshage H. Cytokines and the hepatic acute phase response. J Pathol 1997;181:257-66. 10. Kushner I. The phenomenon of the acute phase response. Ann NY Acad Sci 1982;389:39-48. 11. West M, Billiar T, Curran R, Hyland B, Simmons R. Evidence that rat kupffer cells stimulate and inhibit hepatocyte protein synthesis in vitro by different mechanisms. Gastroenterology 1989;96:1572-82. 12. Bankey P, Prager M, Geldon D, Taylor S, McIntyre K. Acutephase hepatocytes regulate liver sinusoidal cell mediator production. Arch Surg 1994;129:1166-71-3. 13. Laemmli U. Protein electrophoresis in SDS polyacrylamide gels. Nature 1970;227:680-3. 14. Bankey P, Billiar T, Wang WY, Carlson A, Holman R, Cerra F. Modulation of Kupffer cell membrane phospholipid function by n-3 polyunsaturated fatty acids. J Surg Res 1989;46:439-44. 15. Phinney S, Tanga A, Johnson S, Holman R. Reduced adipose 18:3n-3 with weight loss by very low calorie dieting. Lipids 1990;25:798-806. 16. Wigmore S, Fearon K, Ross J. Modulation of human hepatocyte acute phase protein protein production in vitro by n3 and n-6 polyunsaturated fatty acids. Ann Surg 1997;225:103-11. 17. Wan Y, Freeswick P, Khemlani L, Kispert P, Wang S, Su G, et al. Role of LPS, IL-1, IL-6, TNF, and dexamethasone in regulation of LPS-binding protein expression in normal hepatocytes and hepatocytes from LPS-treated rats. Infect Immun 1995;63:2435-42. 18. Eritsland J, Seljeflot I, Arnesen H, Westvik A, Kierulf P. Effect of long-term, moderate dose supplementation with omega-3 fatty acids on monocyte PCA and release of IL-6 in patients with coronary artery disease. Thromb Res 1995;77:337-46. 19. Endres S, Ghorbani R, Kelly V, Georgilis K, Lonnemann G, van der Meer J, et al. The effect of dietary supplementation with n-3 polyunsaturated fatty acids on the synthesis of interleukin-1 and tumor necrosis factor by mononuclear cells. N Engl J Med 1989;320:265-71. 20. Ernst E, Saradeth T, Achhammer G. n-3 fatty acids and acute-phase proteins. Eur J Clin Invest 1991;21:77-82. 21. Wang H, Chen X, Fisher E. N-3 fatty acids stimulate intracellular degradation of apoprotein B in rat hepatocytes. J Clin Invest 1993;91:1380-9. 22. Kinsella J, Lokesh B, Broughton S, Whelan J. Dietary polyunsaturated fatty acids and eicosanoids: potential effects on the modulation of inflammatory and immune cells—an overview. Nutrition 1990;6:24-62. 23. Clamp A, Gimble R. Fatty acids modulate the affinity of the tumor necrosis factor receptor in cultured rat hepatocytes [abstract].Proc Nutr Soc 1994;53:180A.
DISCUSSION Dr M. Wayne Flye (St Louis, Mo). When Mark Callery was in our laboratory, he showed that there is really a different cytokine profile produced by different macrophage populations, including IL-6. When you mix hepatocytes with the Kupffer’s cell, that is a macrophage population presumably that you are stimulating with
Surgery August 1998 LPS. Have you used other macrophage populations to see whether there is a differential effect? Second, have you used the endotoxin-resistant animals and their hepatocytes and Kupffer’s cells to see whether you get a similar response with the fatty acids? Dr Bankey. The answer is no and no to both of those. I personally have not done experiments looking at other macrophage populations in dealing with the interactions with hepatocytes. A lot of that preliminary work had been done looking at macrophage cell lines and peritoneal macrophages. There were differential effects, but they were not quantitatively significant. When we talk about Kupffer’s cells, I think that we can extrapolate. If we had put alveolar macrophages in there or if we had put peritoneal macrophages, I would expect to see the same sort of results in terms of responses to endotoxin. I have not done any work with endotoxin-resistant animals to see how it affects this culture interaction. There is no doubt that the addition of the hepatocyte has some phenotypic effect on the Kupffer cell’s or the nonparenchymal cells. IL-6 is probably being produced by endothelial cells, and our Kupffer’s cell populations are probably only about 85% to 90% pure. We don’t go through the elutriation step to get 98% purity. Therefore in terms of IL-6, we may have a contribution of the sinusoidal endothelial cells also regulating this response. Dr William G. Cioffi (Providence, RI). You looked at IL-6-dependent acute phase proteins, and there are other acute phase proteins that are driven by other cytokines. Have you looked at IL-1-dependent proteins? Second, you said that your results might suggest that this is an explanation for the anti-inflammatory actions of omega-3 fatty acids, and yet you also showed a significant increase in proinflammatory cytokine production. Could you reconcile that? Dr Bankey. Any speculation into the clinical relevance of this in vitro model is just that, speculation. In the onetwo feeding study by the multicenter trial in which they looked at impact versus Osmolite, they actually looked at C-reactive protein in those patients. There really was no difference in the levels of C-reactive protein. How that all sorts out clinically, I wouldn’t extrapolate too much from our in vitro data. I think we may have a nice culture system, in which we have modified the membrane responsivity of Kupffer’s cells or endothelial cells. We have seen increased IL-6, which results in IL-6-dependent acute phase proteins. Whether that translates into anything that we can take home clinically hasn’t been proved by these experiments. In terms of IL-1-dependent cytokines, these are the only two that we have looked at. A very nice experiment that we are trying to get a resident excited about is to take cultured hepatocytes, enrich them specifically, and then look at IL-6 effects on specific acute phase proteins IL-1 and TNF.
Background. The production of acute phase proteins by the liver is critical for homeostasis and recovery after injury. Polyunsaturated fatty acids, specifically of the n-3 family, have demonstrated anti-inflammatory properties that promote recovery; however, the effects of these fatty acids on acute phase protein synthesis have not been fully evaluated. Methods. The synthesis of the acute phase proteins α-2 macroglobulin and lipopolysaccharide binding protein was studied in hepatocyte-Kupffer’s cell cocultures from Wistar rats. Cultures were endotoxin stimulated after enrichment with albumin complexed arachidonic acid (20:4n-6) or docosahexaenoic acid (22:6n-3). Protein synthesis was analyzed by [35S]-methionine labeling or Western blotting. Culture interleukin-6 levels were determined. Results. Both polyunsaturated fatty acids increase hepatocyte synthesis of acute phase proteins α2macroglobulin and lipopolysaccharide binding protein compared with controls; however, the response in the docosahexaenoic acid (22:6n-3) treated cultures was significant (P < .05 vs control). Interleukin-6 was also increased in the polyunsaturated fatty acid cultures compared with controls (P < .05 vs control). Cellular phospholipids were significantly enriched with the individual supplemented fatty acids (P < .05 vs bovine serum albumin). Conclusions. Polyunsaturated fatty acids have the capability to increase in vitro acute phase protein synthesis. This may contribute to the observed anti-inflammatory effect of n-3 polyunsaturated fatty acid enrichment. (Surgery 1998;124:471-6.) From the Departments of Surgery and Cell Biology, University of Massachusetts, Worcester, Mass, and University of Texas, Southwestern Medical Center, Dallas, Texas
THE OPTIMAL COMPOSITION of essential fatty acids to provide surgical patients with sepsis or other systemic inflammatory diseases remains controversial.1 Although required to prevent nutrient deficiencies and serving as an alternative caloric source during stress, specific polyunsaturated fatty acids (PUFAs) from the omega-3 (n-3) family have the additional property of modulating inflammation.2-4 The effects on inflammation are multifactorial and include alterations in cytokine production, microvascular perfusion, and eicosanoid generation.5,6 These fatty acids are incorporated into cellular phospholipid pools including the plasma Supported by National Institutes of Health grant R29 GM51059-02 (P.E.B.). Presented at the Fifty-ninth Annual Meeting of the Society of University Surgeons, Milwaukee, Wis, Feb 12-14, 1998. Reprint requests: Paul Bankey, MD, PhD, Department of Surgery, University of Massachusetts Medical Center, 55 Lake Ave N, Worcester, MA 01655-0333. Copyright © 1998 by Mosby, Inc. 0039-6060/98/$5.00 + 0
11/6/91339
membrane and modulate physical properties of the lipid bilayer and receptor-mediated signal transduction.7,8 Sepsis induces systemic inflammation with a high incidence of organ dysfunction. The acute phase response, including the synthesis of proteins with anti-inflammatory and homeostatic functions by hepatocytes, is a critical determinant of host survival.9,10 Interleukin-6 and glucocorticoids are the primary regulatory signals to the hepatocyte to produce acute phase proteins.9 The effects of n-3 PUFAs on hepatocyte acute phase protein synthesis have not been fully evaluated. Enhanced acute phase protein synthesis may have a short-term beneficial anti-inflammatory effect during sepsis. The purpose of this study was to evaluate the effects of PUFA enrichment on hepatocyte production of acute phase proteins using an in vitro model. MATERIAL AND METHODS Animals. Wistar rats were handled according to the guidelines in the “Guide for use of Laboratory Animals,” published by the National Institutes of SURGERY 471
472 Nolan, Sentementes, Bankey
Fig. 1. Autoradiogram representative of culture protein synthesis after individual fatty acid enrichment and LPS stimulation. Synthesis of individual secretory proteins is altered by prior enrichment with PUFAs.
Health, Bethesda, Md (1985). This protocol was approved by the Institutional Review Board for Animal Research of the University of Massachusetts Medical Center. Cell isolation. Hepatocytes are isolated by collagenase perfusion as described previously.11,12 Briefly, the portal vein is cannulated, and then the liver is perfused in situ first with a washout buffer and then switched to 0.05% collagenase (Sigma #C-0130, 200 units/mg; Sigma). The liver is removed, and cells are combed gently in tissue culture medium. Hepatocytes are pelleted and washed, and viability is assessed by trypan blue exclusion. Kupffer’s cells are obtained from separate animals using 0.2% Pronase (Sigma) perfusion through the portal vein in situ. The liver is removed, minced into 5-mm cubes, and then incubated with constant stirring in 0.2% Pronase E (type XIV; Sigma) with DNase added at 20 and 40 minutes (Merck Chemical, Rahway, NJ). Temperature is maintained at 37° C, and pH is maintained at 7.4 to 7.6. The resulting digest is purified by a single centrifugation through 2-step Percoll gradient (Pharmacia). After a 3-hour incubation in RPMI 1640 medium with 5% calf serum (low endotoxin; Gibco) 80% of the adherent population demonstrates phagocytosis of 1.1-mm latex beads indicating Kupffer’s cells.11 Cell culture. Hepatocytes are seeded onto colla-
Surgery August 1998
gen coated plates on day 1 in Williams E medium supplemented with antibiotics, 10% fetal calf serum, HEPES, and insulin. The Kupffer’s cells are isolated the following day and added to cultured hepatocytes at a 2:1 ratio (HC:KC). The cultures are recovered overnight. The medium is removed, and fatty acid containing medium is added for 8 hours. This medium is removed, and the cultures are stimulated with lipopolysaccharide (LPS) (Escherichia coli O111:B4) for 18 hours. Supernatants are collected for analysis, and selected cultures are radiolabeled with [35S]methionine. Measurement of protein synthesis by [35S]-methionine labeling and sodium dodecyl sulfate–polyacrylamide gel electrophoresis or Western blotting. Cells are washed with methionine-free RPMI and then labeled in medium consisting of methionine-free RPMI supplemented with glutamine (15 mmol/L) and 100 µCi/ml [35S]-methionine. After a 6-hour labeling period, the supernatants are collected, pooled, and stored. Label incorporation into secreted protein is determined by precipitation of the sample with cold 20% trichloroacetic acid and scintillation counting. Specific protein secretion is detected on autoradiograms after separation by 8% sodium dodecyl sulfate–polyacrylamide gel electrophoresis.13 The autoradiograms demonstrate a band at 68 kd corresponding to albumin and at 180 kd corresponding to α2-macroglobulin. Unlabeled culture supernatants are evaluated for α2macroglobulin and LPS-binding protein (LBP) by Western blotting. The primary antibody for α2macroglobulin is goat anti-mouse obtained from Accuate, Inc. The primary antibody for LBP is rabbit anti-rat generated by Genosys, Inc, on the basis of a peptide deduced from DNA sequence. Blotted proteins were detected by first incubating the nitrocellulose in a 1:200 dilution of antibody and washing according to the Bio-Rad Immuno-Blot Assay Kit (Bio-Rad, Richmond, Calif). Next, unbound antibody was washed away and the nitrocellulose was incubated with 1:1000 dilution of goat anti-rabbit or anti-goat immunoglobulin G (H+L) conjugated with alkaline phosphatase (Jackson Immunoresearch Laboratories, Inc). Band densities are compared using the Gel-Doc system from Bio-Rad.12,13 Fatty acid–albumin complexes. A 5% solution of bovine serum albumin (BSA) (fatty acid free, low endotoxin) is added to the Na salt of the fatty acid in a concentration of 0.6 mg fatty acid/mL BSA. The fatty acid–BSA solution is allowed to complex overnight at 37° in a sealed tube. This gives a molar ration of 2.24:1 of fatty acid to albumin.14 Analysis of fatty acid enrichment. Total lipids are extracted by using chloroform/methanol, 2:1
Nolan, Sentementes, Bankey 473
Surgery Volume 124, Number 2
Fig. 2. Western blot representative of culture supernatant content of the acute phase α2-macroglobulin after fatty acid enrichment and LPS stimulation. OLA, Oleic acid.
Table. Cell phospholipid fatty acids (% total) after culture treatment Fatty acid 18:1n-9 18:2n-6 20:4n-6 22:6n-3 Total PUFA
BSA
Oleic acid
ArA
3.8 ± 0.5 16.3 ± 0.3 28.1 ± 0.4 5.2 ± 0.2 47.9 ± 0.9
7.7 ± 0.5* 15.9 ± 0.6 26.7 ± 0.7 4.9 ± 0.1 42.1 ± 1.1
3.6 ± 0.5 14.8 ± 0.7 33.9 ± 1.0* 4.0 ± 0.3 55.5 ± 1.7*
DHA 3.5 ± 0.6 14.1 ± 0.5 25.6 ± 0.9 9.8 ± 0.4* 54.3 ± 1.9*
*P < .05 vs BSA treatment.
(vol/vol). The total cellular phospholipids are separated on TLC plates (0.5 mm silica gel) by using the solvent system of petroleum ether/diethyl ether/acetic acid, 80/20/1 (v/v/v). The phospholipids are recovered and fatty acids are esterified by addition of 14% BF3/methanol (wt/vol) before separation by gas chromatography with known methyl ester standards by using a model 428, HewlettPackard system.15 Interleukin-6 assay. Interleukin-6 (IL-6) activity in serum and culture supernatants was measured in the 7TD1 cell proliferation assay.13 The optical density of the samples was compared with a standard curve generated from serial dilutions of murine recombinant IL-6 (Genzyme). Statistical analysis. Means and SDs for each treatment group were calculated. In experiments in which multiple samples were assessed, analysis of variance (F ratio and mean square for error) was used for comparing the data. Tukey’s test was used to perform multiple comparison tests with .05 error rate. Statworks (Cricket Software, Philadelphia, Pa) was used for statistical calculations. RESULTS Effect of fatty acids on cellular phospholipids. Cocultures (HC:KC) incubated for 8 hours in albumin alone (BSA, Table I) had a total PUFA content of 47.9%. In cultures incubated with 10 µg/mL
arachidonic (ArA, 20:4n-6) or docosahexaenoic (DHA, 22:6n-3) acid the total PUFA content increased significantly to 55.5% and 54.3%, respectively. Arachidonic acid increased to 33.9% from 28.1% and DHA to 9.8% from 5.2% (Table). These findings indicate that individual fatty acids are incorporated into cellular phospholipids after only 8 hours in cell culture. Alterations in secretory protein synthesis after fatty acid incubation. There were no significant differences in the rate of protein synthesis as determined by the incorporation of [35S]-methionine in unstimulated cultures treated with the fatty acids (data not shown). Rates of individual protein synthesis were significantly altered between fatty acid treatment groups in LPS-stimulated cultures as shown in the autoradiogram in Fig 1. The PUFA-treated cultures, both ArA and DHA, demonstrated an increase in labeled α2-macroglobulin and decrease of labeled albumin compared with the oleic acid and BSA treated cultures. These results suggest a direct effect of PUFA enrichment on hepatocyte protein synthesis in LPS-stimulated cocultures. Effect of fatty acid treatment on synthesis of the acute phase proteins α-2 macroglobulin and LPSbinding protein. The PUFA-treated cultures also demonstrated increased synthesis of both acute phase proteins α2-macroglobulin and LBP as determined by Western blotting (Fig 2 for α2-macroglob-
474 Nolan, Sentementes, Bankey
Surgery August 1998
Fig. 3. Western blot representative of culture supernatant content of the acute phase protein LPS-binding protein after fatty acid enrichment and LPS stimulation. OIA, Oleic acid.
Fig. 4. Effect of fatty acid treatment on culture acute phase protein levels as determined by Western blotting. Data are expressed as fold increase over non-LPS stimulated cultures. Blots were compared by densitometry and averaged over n = 3 to 4 experiments. P < .05 vs oleic acid (OLA) response.
ulin and Fig 3 for LBP). Densitometry on the Western blots demonstrated a 4.2- ± 1.1-fold increase in α2-macroglobulin in LPS-stimulated oleic acid treated cultures (3.9- ± 0.9-fold increase in BSA only) but that ArA-treated cultures had a 6.3- ± 2.4-fold increase and DHA-treated cultures had a 8.4- ± 2.9-fold increase. LBP synthesis was also increased significantly in the DHA-treated cultures compared with oleic acid (DHA: 6.1 ± 2.1 vs ArA: 4.0 ± 1.7 vs oleic acid 2.7 ± 0.8 fold vs BSA 2.9 ± 0.6). These data are shown graphically in Fig 4. These results indicate a differential acute phase protein response in cultures enriched with DHA and PUFAs compared with nonenriched cultures or cultures enriched with the monounsaturated fatty acid oleic acid. Effect of fatty acid enrichment on culture IL-6 activity. Cocultures enriched with both PUFAs had an increased production of bioactive interleukin-6 in response to stimulation with 100 ng/mL LPS compared with oleic acid enrichment (Fig 5). The increased production of IL-6 may contribute to the
increased synthesis of acute phase proteins in these treatment groups. DISCUSSION Using an in vitro model of acute phase protein synthesis, we have demonstrated that cocultures of hepatocytes and Kupffer’s cells enriched with PUFAs synthesize more α2-macroglobulin and LBP and less albumin than cultures treated with albumin alone or the monounsaturated fatty acid oleic acid. Associated with the increased acute phase protein synthesis is an increase in the production of the regulatory cytokine IL-6. Furthermore, treatment with the n-3 PUFA docosahexaenoic acid, DHA 22:6n-3, resulted in greater although not statistically significant increases in the acute phase proteins compared with the n-6 PUFA arachidonic acid, ArA 20:4n-6. Our experiments showed consistent up-regulation of the acute phase proteins α2-macroglobuin and LBP after enrichment of PUFAs. Other investigators report varied effects of PUFAs on individual acute phase proteins. In one study IL-6–stimulated
Nolan, Sentementes, Bankey 475
Surgery Volume 124, Number 2
human hepatocytes had increased production of prealbumin and decreased production of haptoglobin after treatment with eicosapentaenoic acid, EPA 20:5n-3.16 Both α2-macroglobulin and LBP are IL-6 up-regulated proteins. Further experimentation is required to determine whether the effects of PUFAs are specific to IL-6 up-regulated acute phase proteins or also have effects on IL-1 and tumor necrosis factor (TNF) regulated acute phase proteins.17 Our data would indicate that this effect is not specific to n-3 PUFAs such as DHA but is a generalized response to PUFA enrichment because it also occurs in response to ArA, the n-6 PUFA used in these experiments. Polyunsaturated fatty acid enrichment with both DHA and ArA resulted in increased IL-6 production in the cocultures compared with oleic acid and albumin alone. Other investigators have reported a differential effect of n-3 PUFAs compared with n-6 PUFAs on in vitro macrophage cytokine production.18,19 Previous investigation of Kupffer’s cell cytokine production after a fish oil diet, high n-3 PUFA, compared with a safflower oil diet, high n-6 PUFA, failed to show differential effects on LPS-stimulated TNF and IL-1 production also.6 It is not clear from these studies whether the increased acute phase protein synthesis reflects effects of the PUFAs on the hepatocyte or Kupffer’s cells or both. Our cellular phospholipid data reflect the composition of the cocultures, not individual cell populations. It is possible that our findings reflect only modulation of function in the Kupffer’s cells and that the hepatocyte is responding to an environment with increased levels of cytokines including IL-6. Hepatocyte enrichment with n-3 PUFAs has been demonstrated to alter lipoprotein production and cytokine signaled acute phase proteins.20,21 The effect of DHA and ArA on hepatocyte cultures only is unknown. The mechanism by which PUFAs alter LPS-stimulated acute phase protein production is unclear. Incorporation into cellular membranes in either the Kupffer’s cell or hepatocyte may influence membrane receptor function by altering membrane fluidity or the coupling of receptor and membrane proteins responsible for signal transduction. Fatty acids have previously been reported to have no effect on the number of TNF receptors expressed on hepatocytes but may influence the affinity of TNF for its receptor.22 The hepatocyte has the capacity to metabolize fatty acids by using intracellular eicosanoid pathways, and these metabolites may have a role in acute phase protein synthesis as they do in cytokine synthesis by Kupffer’s cells. Our current understanding of IL-6 signaled acute phase protein synthesis indicates receptor induced dimeriza-
Fig. 5. Effect of enrichment of cultures with individual fatty acids on LPS-stimulated IL-6 bioactivity. Data are expressed as mean ± SEM for 5 to 6 experiments. P < .05 vs BSA alone. OLA, Oleic acid.
tion of gp130 followed by activation of signal transducers and activators of transcription or acute phase response factor nuclear transcription factors. The effect of PUFA enrichment on these responses is unknown.9,23 We have demonstrated by using an in vitro hepatocyte–Kupffer’s cell coculture model that PUFA enrichment before LPS stimulation results in an enhanced acute phase protein synthetic response that is associated with increased levels of regulatory IL-6. Because significant levels of these fatty acids are components of enteral and parenteral nutritonal support regimens for patients with sepsis, this may be a potential pathway to modify the systemic inflammatory response during sepsis. REFERENCES 1. Grimm H, Tibell A, Norrlind B, Blecher C, Wilker S, Schwemmle K. Immunoregulation by parenteral lipids: impact of the n-3 to n-6 fatty acid ratio. JPEN 1994;18:417-21. 2. Cerra F, Bankey P, Holman R, Mazuski J, LiCari J. Omega-3 polyunsaturated fatty acids as modulators of cellular function in the critically ill. Pharmacotherapy 1991;11:71. 3. Blok W, Katan M, van der Meer J. Modulation of inflammation and cytokine production by dietary n-3 fatty acids. J Nutr 1996;126:1515-33. 4. Calder P. N-3 Polyunsaturated fatty acids and immune cell function. Adv Enzyme Regul 1997;37:197-237. 5. Abbate R, Gori A, Martini F, Brunelli T, Filippini M, Francalanci I, et al. n-3 PUFA supplementation, monocyte PCA expression and interleukin-6 production. Prostaglandins Leukot Essent Fatty Acids 1996;54:439-44. 6. Billiar T, Bankey P, Svingen B, Simmons R, Cerra F. Fatty acid intake and kupffer cell function: fish oil alters eicosanoid and monokine production in vitro. Surgery 1988;104:343. 7. Abel S, Gelderblom W, Smuts C, Kruger M. Thresholds and kinetics of fatty acid replacements in different cellular compartments in rat liver as a function of dietary n-6/n-3 fatty acid content. Prostaglandins Leukot Essent Fatty Acids 1997;56:29-39.
476 Nolan, Sentementes, Bankey
8. Clarke S, Jump D. Polyunsaturated fatty acid regulation of hepatic gene transcription. Lipids 1996;31:S7-11. 9 . Moshage H. Cytokines and the hepatic acute phase response. J Pathol 1997;181:257-66. 10. Kushner I. The phenomenon of the acute phase response. Ann NY Acad Sci 1982;389:39-48. 11. West M, Billiar T, Curran R, Hyland B, Simmons R. Evidence that rat kupffer cells stimulate and inhibit hepatocyte protein synthesis in vitro by different mechanisms. Gastroenterology 1989;96:1572-82. 12. Bankey P, Prager M, Geldon D, Taylor S, McIntyre K. Acutephase hepatocytes regulate liver sinusoidal cell mediator production. Arch Surg 1994;129:1166-71-3. 13. Laemmli U. Protein electrophoresis in SDS polyacrylamide gels. Nature 1970;227:680-3. 14. Bankey P, Billiar T, Wang WY, Carlson A, Holman R, Cerra F. Modulation of Kupffer cell membrane phospholipid function by n-3 polyunsaturated fatty acids. J Surg Res 1989;46:439-44. 15. Phinney S, Tanga A, Johnson S, Holman R. Reduced adipose 18:3n-3 with weight loss by very low calorie dieting. Lipids 1990;25:798-806. 16. Wigmore S, Fearon K, Ross J. Modulation of human hepatocyte acute phase protein protein production in vitro by n3 and n-6 polyunsaturated fatty acids. Ann Surg 1997;225:103-11. 17. Wan Y, Freeswick P, Khemlani L, Kispert P, Wang S, Su G, et al. Role of LPS, IL-1, IL-6, TNF, and dexamethasone in regulation of LPS-binding protein expression in normal hepatocytes and hepatocytes from LPS-treated rats. Infect Immun 1995;63:2435-42. 18. Eritsland J, Seljeflot I, Arnesen H, Westvik A, Kierulf P. Effect of long-term, moderate dose supplementation with omega-3 fatty acids on monocyte PCA and release of IL-6 in patients with coronary artery disease. Thromb Res 1995;77:337-46. 19. Endres S, Ghorbani R, Kelly V, Georgilis K, Lonnemann G, van der Meer J, et al. The effect of dietary supplementation with n-3 polyunsaturated fatty acids on the synthesis of interleukin-1 and tumor necrosis factor by mononuclear cells. N Engl J Med 1989;320:265-71. 20. Ernst E, Saradeth T, Achhammer G. n-3 fatty acids and acute-phase proteins. Eur J Clin Invest 1991;21:77-82. 21. Wang H, Chen X, Fisher E. N-3 fatty acids stimulate intracellular degradation of apoprotein B in rat hepatocytes. J Clin Invest 1993;91:1380-9. 22. Kinsella J, Lokesh B, Broughton S, Whelan J. Dietary polyunsaturated fatty acids and eicosanoids: potential effects on the modulation of inflammatory and immune cells—an overview. Nutrition 1990;6:24-62. 23. Clamp A, Gimble R. Fatty acids modulate the affinity of the tumor necrosis factor receptor in cultured rat hepatocytes [abstract].Proc Nutr Soc 1994;53:180A.
DISCUSSION Dr M. Wayne Flye (St Louis, Mo). When Mark Callery was in our laboratory, he showed that there is really a different cytokine profile produced by different macrophage populations, including IL-6. When you mix hepatocytes with the Kupffer’s cell, that is a macrophage population presumably that you are stimulating with
Surgery August 1998 LPS. Have you used other macrophage populations to see whether there is a differential effect? Second, have you used the endotoxin-resistant animals and their hepatocytes and Kupffer’s cells to see whether you get a similar response with the fatty acids? Dr Bankey. The answer is no and no to both of those. I personally have not done experiments looking at other macrophage populations in dealing with the interactions with hepatocytes. A lot of that preliminary work had been done looking at macrophage cell lines and peritoneal macrophages. There were differential effects, but they were not quantitatively significant. When we talk about Kupffer’s cells, I think that we can extrapolate. If we had put alveolar macrophages in there or if we had put peritoneal macrophages, I would expect to see the same sort of results in terms of responses to endotoxin. I have not done any work with endotoxin-resistant animals to see how it affects this culture interaction. There is no doubt that the addition of the hepatocyte has some phenotypic effect on the Kupffer cell’s or the nonparenchymal cells. IL-6 is probably being produced by endothelial cells, and our Kupffer’s cell populations are probably only about 85% to 90% pure. We don’t go through the elutriation step to get 98% purity. Therefore in terms of IL-6, we may have a contribution of the sinusoidal endothelial cells also regulating this response. Dr William G. Cioffi (Providence, RI). You looked at IL-6-dependent acute phase proteins, and there are other acute phase proteins that are driven by other cytokines. Have you looked at IL-1-dependent proteins? Second, you said that your results might suggest that this is an explanation for the anti-inflammatory actions of omega-3 fatty acids, and yet you also showed a significant increase in proinflammatory cytokine production. Could you reconcile that? Dr Bankey. Any speculation into the clinical relevance of this in vitro model is just that, speculation. In the onetwo feeding study by the multicenter trial in which they looked at impact versus Osmolite, they actually looked at C-reactive protein in those patients. There really was no difference in the levels of C-reactive protein. How that all sorts out clinically, I wouldn’t extrapolate too much from our in vitro data. I think we may have a nice culture system, in which we have modified the membrane responsivity of Kupffer’s cells or endothelial cells. We have seen increased IL-6, which results in IL-6-dependent acute phase proteins. Whether that translates into anything that we can take home clinically hasn’t been proved by these experiments. In terms of IL-1-dependent cytokines, these are the only two that we have looked at. A very nice experiment that we are trying to get a resident excited about is to take cultured hepatocytes, enrich them specifically, and then look at IL-6 effects on specific acute phase proteins IL-1 and TNF.