A method for removing interleukin-1- and tumor necrosis factor-inducing substances from bacterial cultures by ultrafiltration with polysulfone.

The ability of human mononuclear cells (MNC) to produce cytokines is a highly sensitive and biologically relevant test system for the presence of microbial products. The safety of parenteral fluids is presently determined by gelation of the limulus amebocyte lysate (LAL) to endotoxin. In the present study, crude bacterial culture supernatants from Escherichia coli were subjected to ultrafiltration using polysulfone and the ultrafiltrates were tested for their ability to stimulate human MNC. Total interleukin-1 (IL-1) and tumor necrosis factor (TNF) produced by MNC were measured by radioimmunoassay. Endotoxin-like substances in E. coli cultures are rejected by a factor of at least 100,000. Rejection takes place by molecular size exclusion and by absorption. The sensitivity of the LAL and MNC cytokine production were comparable. These studies demonstrate a wide margin of safety for the production of parenteral fluids using ultrafiltration for endotoxin-containing materials.

A method for the detection of erythrocyte-bound interleukin-8 in humans during interleukin-1 immunotherapy.

Interleukin-8 (IL-8) has been recently shown to bind to human erythrocytes with high affinity and is therefore potentially difficult to detect in serum or plasma. IL-8 is transiently elevated in the serum of baboons after the administration of interleukin-1 alpha (IL-1 alpha). The objective of this study was to investigate whether IL-8 can be detected in the plasma or in detergent-lysed erythrocytes from cancer patients undergoing treatment with IL-1 alpha. Using a specific radioimmunoassay (RIA), plasma IL-8 was detected within 1-2 h after the first IL-1 alpha infusion. Thereafter, the levels declined rapidly and after 4-8 h were undetectable. Erythrocyte-bound IL-8 was detectable 1-2 h after the increase in plasma levels. The erythrocyte-bound IL-8 levels were higher than those measured in plasma and remained elevated long after the plasma levels had become undetectable. Erythrocyte membranes accounted for all of the erythrocyte-associated IL-8, as IL-8 was undetectable in the cytosol after erythrocyte lysis. The assay used in these studies detects IL-8 in erythrocyte lysates when it cannot be measured in plasma and may therefore be useful in monitoring IL-8 production in vivo.

Transfer of cytokine-inducing bacterial products across hemodialyzer membranes in the presence of plasma or whole blood.

Cytokine production by peripheral blood mononuclear cells (PBMC) is a sensitive indicator of cytokine-inducing substances which may cross from contaminated dialysate into the blood compartment. The objective of this study was to compare the transfer of cytokine-inducing substances from dialysate contaminated with a culture filtrate from Pseudomonas aeruginosa across dialyzers with low (hemophan) or intermediate ultrafiltration coefficients (modified cellulose triacetate, CTA), under conditions where either 10% plasma or whole blood was circulated in the blood compartment. Eight paired experiments of in vitro dialysis were carried out at 37 degrees C using a countercurrent recirculating loop dialysis circuit with either a new CTA or hemophan dialyzer. 10% plasma in standard tissue culture medium was circulated through the blood compartment and bicarbonate dialysate was circulated in the dialysate compartment. The dialysate was challenged sequentially by log-fold dilutions (10(2), 10(3) or 10(4)) of a Ps. aeruginosa culture filtrate. Samples were drawn from the blood compartment 5 and 15 minutes after each challenge and incubated with suspensions of PBMC in the absence or presence of polymyxin B, in order to block endotoxin. After 24 h at 37 degrees C, total interleukin-1 alpha (IL-1 alpha) was measured by RIA. Although the dialysate contained potent cytokine-inducing substances, there was no significant IL-1 alpha production by PBMC incubated with the plasma mixture from the blood compartment in the majority of experiments with both dialyzers and with each of the three dilutions of the bacterial challenge. Eight experiments were also performed with CTA dialzyers using heparinized whole blood in the blood compartment. Samples of whole blood and dialysate were drawn at baseline, after one hour of dialysis with uncontaminated dialysate and 15 minutes and three hours after dialysis with dialysate contaminated with Ps. aeruginosa filtrate. There was no significant IL-1 alpha production by PBMC isolated from the whole blood 1 h after dialysis with uncontaminated dialysate, and 15 min and 2 h after adding the Ps. aeruginosa filtrate to the dialysate side. In contrast, production of IL-1 alpha by PBMC from the same donors incubated with samples from the dialysate were 263 +/- 50, 1074 +/- 306, 2333 +/- 774 and 2602 +/- 702 pg/2.5 x 10(6) PBMC, respectively at the same four time points. These data suggest that although the Ps. aeruginosa culture filtrate present in the dialysate was a potent inducer of IL-1 alpha, neither dialyzer permitted transfer of cytokine inducing substances from the dialysate into the blood compartment.

TNF-alpha neutralization ameliorates obstruction-induced renal fibrosis and dysfunction.

Upper urinary tract obstruction results in tubulointerstitial fibrosis and a progressive decline in renal function. Although several inflammatory mediators have been implicated in the pathophysiology of renal obstruction, the contribution of TNF-alpha to obstruction-induced fibrosis and renal dysfunction has not been thoroughly evaluated. To study this, male Sprague-Dawley rats were subjected to left unilateral ureteral obstruction vs. sham operation. Rats received either vehicle or a pegylated form of soluble TNF receptor type 1 (PEG-sTNFR1) every 84 h. The kidneys were harvested 1, 3, or 7 days postoperatively, and tissue samples were analyzed for TNF-alpha expression (ELISA), macrophage infiltration (ED-1 staining), transforming growth factor-beta(1) expression (ELISA, RT-PCR), collagen I and IV activity (Western Blot, immunohistochemistry), alpha-smooth muscle actin accumulation (immunohistochemistry, Western blot analysis), and angiotensinogen expression (Western blot). In a separate arm, the glomerular filtration rate (inulin clearance) of rats subjected to unilateral ureteral obstruction in the presence of either vehicle or PEG-sTNFR1 was determined. Renal obstruction induced increased tissue TNF-alpha and transforming growth factor-beta(1) levels, collagen I and IV activity, interstitial volume, alpha-smooth muscle actin accumulation, angiotensinogen expression, and renal dysfunction, whereas treatment with PEG-sTNFR1 significantly reduced each of these markers of renal fibrosis. These results demonstrate that TNF-alpha mediates obstruction-induced renal fibrosis and identify TNF-alpha neutralization as a potential therapeutic option for the amelioration of obstruction-induced renal injury.

Cytokines and biocompatibility.

Issues in biocompatibility have traditionally focused on the interaction of membranes or materials with the activation of complement, red blood cell destruction, clotting dysfunctions, and platelet activation. Other studies are concerned with the replacement fluid formulations and host responses. Recent attention has examined the interaction of the blood monocyte with either complement or membrane surfaces. Activation of monocytes leads to the synthesis of cytokines, namely interleukin 1, tumor necrosis factor, and interleukin 6. These potent polypeptide 'cytokines' affect biological functions in nearly every tissue. Understanding the mechanisms by which gene expression of these cytokines is triggered will help in the better design of membrane material or substitution fluids. These are discussed in the context of increased cytokine synthesis during hemodialysis and continuous ambulatory peritoneal dialysis.

The relationship between biocompatibility and interleukin-1.

Biocompatibility is redefined as the quality of being mutually tolerant with life. In so far as this represents a quality which is as likely to be achieved as is the alchemist's dream of turning lead into gold, a compromise approach is recommended. It is suggested that all extracorporeal or body invasive procedures stimulate the inflammatory defence mechanism of the body by stimulating the monocyte to produce a family of polypeptides currently known collectively as Interleukin-1 (IL-1). So far two dissimilar gene products have been cloned and there are probably more. The IL-1 group of polypeptides possess hormonal functions which orchestrate nearly every instrument of the body's defence system. Inducers of IL-1 are present in dialysate and induce bacterial pyrogen and acetate. In addition bacterial cell wall glycoprotein may be cleaved into muramyl peptides by the release of granulocyte lysozyme at the membrane interface. Muramyl dipeptides have been found in CAPD drain fluid and are more potent inducers of IL-1 than endotoxin. Membrane activation of the fifth component of the complement with the release of C5a will also induce monocytes to produce IL-1. The consequences of repeated stimulation of the acute phase response are undesirable and may include muscle wasting, osteopenia and bone cysts (Shrinking man syndrome), fibrosis of scapulo-humeral joints and the carpal-tunnel syndrome. These latter lesions are often associated with deposition of amyloid fibrils related to beta 2-microglobulin. Efforts to reduce these complications are urgently required.(ABSTRACT TRUNCATED AT 250 WORDS)

Plasma interleukin-1 activity during hemodialysis: the influence of dialysis membranes.

Plasma interleukin-1 (IL-1) activity was measured in 7 stable ESRD patients on regular hemodialysis for no less than 5 months. Predialysis levels were significantly raised compared to 8 normal control subjects. During hemodialysis with four different membranes, plasma IL-1 activity rose with Cuprophan and Hemophan and was unchanged or reduced with Gambrane and Polysulfon. In spite of these differences, body temperature rose in all forms of hemodialysis. Factors responsible for the predialysis elevation included the absence of renal function and/or the repeated stimulus of human blood monocytes by hemodialysis. In view of the uniform increase of body temperature during hemodialysis, the differences in changes of plasma IL-1 activity observed with the various membranes may not be caused by a variable stimulation of monocytes but rather by the presence or absence of the membrane's ability to remove and/or absorb IL-1. Thus, the consequences of monocyte hemodialysis stimulation may be obtained locally, even in the presence of unchanged or reduced plasma IL-1 activity.

Plasma interleukin-1 activity in humans undergoing hemodialysis with regenerated cellulosic membranes.

Because of the similarity between the acute phase response and the post-hemodialysis syndrome, it has been proposed that human blood monocytes adhering to the hemodialysis membrane are exposed to several inducers of interleukin-1 (IL-1). These include endotoxin from the dialysate fluid and C5a activated on the surface of regenerated cellulosic hemodialysis membranes. In order to test this hypothesis, we measured IL-1 plasma activity by employing gel filtration to remove IL-1 inhibitors and subsequent augmentation of lectin-induced murine thymocyte proliferation in each fraction. To evaluate the acute response of end stage renal disease patients during a single hemodialysis with a regenerated cellulosic membrane, plasmas were tested before and after hemodialysis. Control plasmas were obtained from healthy individuals. Plasma IL-1 activity eluted at a molecular weight of predominantly 15 kD. This material was neutralized by anti-human-IL-1 which does not recognize IL-2. IL-1 activity was detected in the pre-dialysis plasma of 6 out of 7 patients and rose in 6 patients following a 5 hour hemodialysis. This increase was associated with a significant rise in body temperature. This study supports the hypothesis that hemodialysis with regenerated cellulose might augment IL-1 production in endstage renal disease patients maintained on longterm hemodialysis. It demonstrates for the first time that dialysis patients have detectable plasma levels of IL-1 compared to normal individuals.

Studies on the ability of hemodialysis membranes to induce, bind, and clear human interleukin-1.

Interleukin-1 (IL-1) is a polypeptide cytokine predominantly produced by monocytes in response to injury or infection. Effects of IL-1 such as fever, acute phase protein synthesis, hypotension, and loss of body mass are complications of hemodialysis. Endotoxin-contaminated dialysate fluid, sodium acetate as a dialysate buffer, and activated complement can induce IL-1 during hemodialysis. In the present study, we investigated the intrinsic property of hemodialysis membranes to stimulate human blood mononuclear cells (MNC) to produce IL-1. Incubation of MNC on sheets of two commonly used hemodialysis membranes, regenerated cellulose (RC) and polyacrylonitrile (PAN) resulted in significant induction of IL-1 in the absence of endotoxin or complement. Production of intracellular and extracellular IL-1 was greater in MNC exposed to PAN compared with RC (p less than 0.01). Similar results were obtained when MNC were exposed to strips of hemodialysis membranes in rotating tubes. However, compared with RC, PAN binds significant amounts of human IL-1. In a model of in vitro hemodialysis, radiolabeled recombinant human IL-1 was added to the blood compartment of a PAN and a RC dialyzer. Forty percent of radiolabeled IL-1 bound to the PAN dialyzer membrane compared with 10% to the membrane of a RC dialyzer. In addition, 22% of radiolabeled IL-1 was found in the dialysate compartment of a PAN dialyzer after 1 hour whereas 1% was found in the dialysate side of a RC dialyzer; these results were confirmed by measuring bioactivity of IL-1. These studies demonstrate the intrinsic property of hemodialysis membranes to stimulate human IL-1 production; in addition they establish that dialysis membranes differ in their ability to bind and clear IL-1.

Transcription, not synthesis, of interleukin-1 and tumor necrosis factor by complement.

Stimulation of interleukin-1 beta (IL-1 beta) and tumor necrosis factor (TNF alpha) production was studied during in vitro hemodialysis (HD) of whole blood using cuprammonium (Cup) or polysulfone (PS) dialyzers. In the absence of LPS, circulation of whole blood for two hours through Cup or PS dialyzers was not sufficient to induce production of IL-1 beta or TNF alpha in peripheral blood mononuclear cells (PBMC) during subsequent 24 hour culture. However, compared to freshly isolated cells, post-HD PBMC were primed to produce more IL-1 beta and TNF alpha when subsequently stimulated with LPS. Despite the lack of spontaneous monokine synthesis after HD, we observed transcription of mRNA coding for IL-1 beta and TNF alpha after two hours of LPS-free HD. When compared to levels of mRNA induced by 5 ng/ml LPS (100%), Cup induced 27 +/- 6% whereas PS did not induce detectable transcription of IL-1 beta. In the case of TNF alpha mRNA, Cup induced 26 +/- 8% and PS 13 +/- 3%. Recombinant C5a induced mRNA for IL-1 beta in PBMC without detectable IL-1 beta protein synthesis. We conclude that transcription of mRNA for IL-1 beta and TNF alpha during HD is primarily caused by complement activation by Cup, but that LPS or other factors are required for translation of IL-1 beta and TNF alpha mRNA transcribed during HD.

Peripheral nerve regeneration is impeded by interleukin-1 receptor antagonist released from a polymeric guidance channel.

Interleukin-1 receptor antagonist (IL-1ra), a true antagonist of the interleukin-1 (IL-1) receptors, is released by activated macrophages and binds specifically to the IL-1 receptors without triggering IL-1 effects. Following peripheral nerve axotomy, activated macrophages release IL-1, which induces the expression of nerve growth factor (NGF) mRNA in Schwann cells. IL-1ra may therefore impede peripheral nerve regeneration by blocking the NGF-mediated effect of IL-1. Peripheral nerve regeneration occurring through polymeric guidance channels releasing IL-1ra was investigated in a 4-mm gap transected mouse sciatic nerve model. Cohorts of five animals were implanted with tubes releasing either bovine serum albumin (BSA), BSA with IL-1ra, or BSA with deactivated IL-1ra (dIL-1ra) for 4 weeks. In vitro release kinetics indicated that after an initial burst, IL-1ra release was linear for the next 3 1/2 weeks. Following implantation of a polymeric guidance channel, a regenerated cable bridged the nerve gap in all animals. The cables were similar in size and were composed of nerve microfascicles containing both unmyelinated and myelinated axons in association with their Schwann cells. Tissue regenerated in tubes releasing BSA-IL-1ra contained, however, significantly fewer myelinated and unmyelinated axons and blood vessels than did tubes releasing BSA alone or BSA-dIL-1ra. We conclude that a naturally occurring antagonist of IL-1 receptors impedes peripheral nerve regeneration, suggesting that macrophages play an essential role in controlling peripheral nerve regeneration through the release of stimulatory and/or inhibitory molecules.

Enhancement of in-vitro human interleukin-1 production by sodium acetate.

Human blood monocytes were incubated in vitro in the presence of various concentrations of sodium acetate or sodium chloride or with medium alone. Intracellular and extracellular levels of interleukin-1 (IL-1) were measured. The production of intracellular IL-1 and the release of extracellular IL-1 were higher in the presence of acetate than in the presence of chloride or in medium alone. The concentrations of acetate used were comparable to those encountered by blood monocytes on the surface of haemodialysis membranes. Since complications of peritoneal dialysis, such as loss of ultrafiltration and progressive fibrosis of the peritoneum, have been associated with the use of sodium acetate as the exchange-fluid buffer, these results suggest that the widespread use of sodium acetate as a buffer during haemodialysis may be contraindicated.

Tumor necrosis factor (TNF)-alpha-induced interleukin-8 in human blood cultures discriminates neutralization by the p55 and p75 TNF soluble receptors.

The dose-dependent increase in mortality in patients with sepsis who are treated with tumor necrosis factor (TNF) p75 soluble receptor Fc conjugate (p75-Fc) remains unexplained. In this study, neutralization of TNF-alpha-induced interleukin (IL)-8 by p75-Fc in whole human blood exhibited a U-shaped inhibition curve, whereas the TNF-soluble p55 receptor, linked to polyethylene glycol (p55-PEG), exhibited a dose-dependent inhibition. Native soluble p75 increased TNF-alpha-induced IL-8, versus a 61% reduction by native p55. Spontaneous IL-8 production was increased by p75-Fc or native p75 but not by p55-PEG or native p55. Unexpectedly, TNF-alpha-stimulated IL-1 receptor antagonist was suppressed by p75-Fc but not by p55-PEG. Studies of binding to TNF trimer revealed that p75-Fc has an affinity 40-fold lower than that of p55-PEG and a faster off rate. Native and p75-Fc pass TNF-alpha to membrane receptors more readily than does native or p55-PEG, which may partly explain the increased mortality in patients with sepsis who are treated with p75-Fc.

Comparison of first use and reuse of cuprophan membranes on interleukin-1 receptor antagonist and interleukin-1 beta production by blood mononuclear cells.

The magnitude of the changes in a variety of blood constituents on exposure to the dialysis membrane has been used as an index of "biocompatibility," and dialyzer reuse has been postulated to improve biocompatibility by attenuating these changes. We studied the hemodialysis-induced changes in the in vitro production of interleukin-1 receptor antagonist (IL-1Ra) and interleukin-1 beta (IL-1 beta) by peripheral blood mononuclear cells (PBMCs), and compared the effect of first use and reuse of cuprophan membranes on these changes. Studies were performed during dialysis with first use and third reuse of the same kidney. The cell content and production of IL-1Ra and IL-1 beta by unstimulated and endotoxin- or IgG-stimulated PBMCs were studied just prior to dialysis, and from the afferent and efferent limbs of the blood circuit 15 minutes after the start of dialysis. Interleukin-1 receptor antagonist and IL-1 beta were measured by specific radioimmunoassay and are expressed as picograms per 2.5 x 10(6) PBMCs. Fifteen minutes after the start of dialysis, the number of PBMCs harvested from 10 mL of blood decreased from 19.8 +/- 4.7 x 10(6) predialysis to 14 +/- 3 x 10(6) (P = 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)

Detection of tumor necrosis factor soluble receptor p55 in blood samples from healthy and endotoxemic humans.

The tumor necrosis factor (TNF) soluble receptor derived from the cell surface p55 TNF receptor (TNFsRp55) is a naturally occurring substance generated during infection and inflammation. TNFsRp55 inhibits biologic effects of TNF. An RIA was developed to quantitate TNFsRp55 in human blood. Recovery of TNFsRp55 from blood anticoagulated with EDTA was optimal compared with recovery from serum or heparinized plasma. TNF did not interfere with the assay. With the RIA based on radiolabeled nonglycosylated Escherichia coli-derived recombinant TNFsRp55, a mean concentration of 198 +/- 15 pg/mL was found in 14 volunteers. When glycosylated CHO cell-derived TNFsRp55 was used, the mean level was 1656 +/- 95 pg/mL. Infusion of endotoxin into volunteers induced TNFsRp55, which peaked at 517 +/- 99 pg/mL for the E. coli-based RIA and 7300 +/- 1810 pg/mL for the CHO cell-based RIA. These findings demonstrate that blood collected in EDTA is optimal for measuring circulating TNFsRp55 and that this soluble receptor is present in health but elevated during endotoxemia.

Gene expression of interleukin-1 beta during hemodialysis.

It is still controversial whether the hemodialysis (HD) procedure is an inflammatory process in vivo. Therefore, we studied the gene expression of interleukin-1 beta (IL-1 beta) as a marker of inflammation in peripheral blood mononuclear cells (PBMC) of patients during HD by Northern blotting and polymerase chain reaction. Compared to PBMC separated pre-HD (1.0 densitometric units), the amount of IL-1 beta mRNA was increased in PBMC leaving the dialyzer (12.2 +/- 2 densitometric units, P < 0.01), but was not increased in PBMC re-entering the dialyzer from the systemic circulation (0.6 +/- 0.1 densitometric units) in all 12 patients studied. The maximal amount of IL-1 beta mRNA in PBMC was seen at five minutes after start of HD. There was a significant correlation between the increase in IL-1 beta mRNA and the increase in activated complement C5a (r = 0.71, P < 0.01). HD using less complement-activating membranes (hemophan, polysulfone, polyamide or polyacrylonitrile) resulted in no detectable IL-1 beta mRNA. Furthermore, a monoclonal antibody against human C5a reduced the increase in IL-1 beta mRNA by 83% (P < 0.05), indicating that C5a plays a major role for induction of IL-1 beta mRNA during HD. This study demonstrates that during HD with regenerated cellulose, gene expression for IL-1 beta takes place in PBMC.

New polyether sulfone dialyzers attenuate passage of cytokine-inducing substances from pseudomonas aeruginosa contaminated dialysate.

The use of bicarbonate dialysate and high-flux and reprocessed dialyzers has raised concerns about the reverse transfer of dialysate contaminants into the blood compartment. This in vitro study was performed to investigate the reverse transfer of soluble Pseudomonas aeruginosa bacterial products across a polyether sulfone (PES), a newly developed synthetic polymer dialyzer. In vitro dialysis was carried out at 37 degreesC in a closed countercurrent recirculating loop dialysis circuit with a new PES dialyzer. An equal mixture of heparinized whole blood (from healthy volunteers) with pyrogen-free tissue culture medium was circulated in the blood compartment, and bicarbonate dialysate was circulated in the dialysate compartment. After 15 min of dialysis, the dialysate was challenged sequentially with 10(-4), 10(-3), and 10(-2) dilutions of a P. aeruginosa culture supernatant. 1-ml samples were drawn from the blood compartment 5 and 15 min after each challenge and incubated upright at 37 degrees C. At the end of 24 h, Triton X-100 was added, in order to extract total interleukin (IL) 6 and IL-8 production by the whole-blood mixture. These cytokines were measured by electrochemiluminescence assays. At dilutions of 10(-4) and 10(-3), the reverse transfer of soluble bacterial products across the dialyzer was negligible. Five and 15 min after contaminating the dialysate with the highest concentration (10(-2) dilution), the increase in IL-6 production was 239 +/- 170% (p = 0.06) and 886 +/- 444% (p = 0.02), respectively. However, comparing the IL-6-inducing potency of the 10(-2) bacterial supernatant dilution to the spontaneous IL-6 production in the blood compartment during dialysis with the same dilution of dialysate contaminant, there was a dramatic reduction in IL-6 production by 94 and 89% at 5 and 15 min, respectively. Similarly, 5 and 15 min after contaminating the dialysate with the 10(-2) dilution, the increase in IL-8 production was 357 +/- 147% (p = 0.07) and 630 +/- 229% (p = 0.04), respectively. However, comparing the IL-8-inducing potency of the 10(-2) bacterial supernatant dilution to the spontaneous IL-8 production in the blood compartment during dialysis with the same dilution of dialysate contaminant, there was a dramatic reduction in IL-8 production by 93 and 92% at 5 and 15 min, respectively. These results demonstrate that PES dialyzers markedly attenuate passage of cytokine-inducing substances from contaminated dialysate, using a method that detects the entire cytokine synthetic output in the blood compartment.

Liposomal delivery of purified inhibitory-kappaBalpha inhibits tumor necrosis factor-alpha-induced human vascular smooth muscle proliferation.

Vessel injury results in the elaboration of various cytokines, including tumor necrosis factor-alpha (TNF-alpha), which may influence vascular smooth muscle cell (VSMC) function and contribute to atherogenesis. We tested the hypothesis that TNF-alpha-induced VSMC proliferation requires activation of the transcription factor nuclear factor-kappaB (NF-kappaB), which could be prevented by delivery of the NF-kappaB inhibitory peptide, IkappaBalpha. TNF-alpha induced concentration-dependent human VSMC proliferation, and neutralizing antibody to interleukin-6 reduced TNF-alpha-induced VSMC proliferation by 65%. In TNF-alpha-stimulated VSMCs, there was a 3-fold increase in NF-kappaB-dependent luciferase reporter activity that was associated with degradation of IkappaBalpha. To determine an essential role for NF-kappaB in TNF-alpha-induced VSMC proliferation, recombinant IkappaBalpha was introduced into VSMCs via liposomal delivery. Under these conditions, TNF-alpha-induced NF-kappaB nuclear translocation and DNA binding were inhibited, NF-kappaB-dependent luciferase activity was reduced by 50%, there was no degradation of native IkappaBalpha detected, interleukin-6 production was reduced by 54%, and VSMC proliferation was decreased by 60%. In conclusion, the mitogenic effect of TNF-alpha on human arterial VSMCs is dependent on NF-kappaB activation and may be prevented by exogenously delivered IkappaBalpha. Furthermore, liposomal delivery of endogenous inhibitory proteins may represent a novel, therapeutically accessible method for selective transcriptional suppression in the response to vascular injury.