Gadolinium chloride inhibits lipopolysaccharide-induced mortality and in vivo prostaglandin E2 release By splenic macrophages.

The monocytic phagocytic system, consisting primarily of tissue macrophages of the liver and spleen, produces prostaglandin E(2) (PGE(2)), a modulator of the septic response. Macrophages are known to internalize gadolinium chloride (GD), a lanthanide metal, which inhibits phagocytic function. Thus we studied the effect of in vivo GD on lipopolysacchride (LPS)-induced mortality and on LPS-stimulated PGE(2) release by cultured splenic macrophages. GD (7 mg/kg intravenously) given on the two days prior to LPS challenge (30 mg/kg intravenously) completely prevented the uniform mortality in rats. This protective effect was transient since rechallenge with LPS 10 days later was uniformly lethal. Previous work in this laboratory has established a critical role of arginine concentration on macrophage behavior in vitro. Therefore, to establish culture conditions reflective of the milieu within the portal venous system, alanine and arginine levels were measured in the portal and hepatic veins of normal and endotoxemic (LPS, 10 mg/kg intraperitoneally) rats. In contrast to alanine levels, which were not altered by endotoxemia, there was a reduction of arginine concentrations from a range of 50 to 250 micromol/L in normal rats to a range of 10 to 50 micromol/L after LPS challenge. Consequently subsequent in vitro assays of splenic macrophage secretory behavior were performed in concentrations of 1200 micromol/L arginine (in standard RPMI-1640), as well as in concentrations reflective of physiologic arginine levels (10 and 100 micromol/L in modified RPMI-1640). Rat splenic macrophages harvested after two consecutive days of either in vivo saline or GD injection (7 mg/kg intravenously) were stimulated with LPS (0.025 to 2.5 microg/ml). At 72 hours of culture, the release of PGE(2) by splenic macrophages from GD-treated rats was significantly (P <0.0001) reduced at all LPS concentrations. Increased PGE(2) production was not present when the splenic macrophages were cultured in the supraphysiologic arginine (1200 micromol/L) concentration. The results demonstrate the relevance of physiologic arginine concentrations in cell culture studies and suggest that the protection conferred by GD against septic mortality may be related to downregulation of the release of immunosuppressive PGE(2) by the monocytic phagocytic system.

Gadolinium blocks rat Kupffer cell calcium channels: relevance to calcium-dependent prostaglandin E2 synthesis and septic mortality.

Hepatic Kupffer cells (KC), the major tissue macrophage population, produce the septic response mediators, tumor necrosis factor alpha (TNF-alpha) and prostaglandin E2 (PGE2), and have been shown to internalize gadolinium chloride (GD), a rare earth metal of the lanthanide series. Because GD pretreatment of rats has been shown to inhibit the mortality of sepsis, we studied the secretory response to lipopolysaccharide (LPS) by KC isolated from rats injected with either saline or GD (7 mg/kg, intravenously) on the 2 days before KC isolation. Using culture conditions modified to reflect the intrasinusoidal milieu of arginine (RPMI-1640 media with 10 or 100 micromol/L arginine), KC from GD-treated rats responded to LPS (0. 0025 microg/mL) with significantly (P <.01) reduced PGE2 release. In contrast, TNF-alpha release by treated KC was significantly (P <.05) enhanced, consistent with the loss of PGE2 autocoid inhibition of TNF-alpha. Calcium flux is an early signaling event in eicosanoid synthesis, and GD is known to block calcium channels. Therefore, KC were loaded with fura-2-AM to study the effect of GD on KC calcium flux. GD prevented ionomycin and platelet-activating factor (PAF)-mediated [Ca++]i increase and calcium-dependent PGE2 synthesis, while GD did not affect PGE2 synthesis when protein kinase C (PKC) was directly activated with tetradecanoylphorbolacetate (TPA). The inhibition of calcium flux and calcium-dependent PGE2 synthesis in the major cell of the monocytic phagocytic system by GD may explain the previously reported ability of this lanthanide to prevent the mortality of endotoxemia.

Gadolinium chloride inhibits Kupffer cell nitric oxide synthase (iNOS) induction.

Kupffer cells (KC) are the phagocytic macrophages of the liver. The rare earth metal, gadolinium (GdCl3), is a lanthanide, which, after phagocytosis by the KC, has been found to alter various aspects of KC physiology. In this study, we describe for the first time that the in vivo administration of GdCl3 to rats decreases the release of NO by isolated rat KC in response to lipopolysaccharide. Western blot analysis shows decreased expression of both inducible nitric oxide synthase as well as total cellular calmodulin after GdCl3 treatment. Possible mechanisms for this phenomenon are suggested.

Outcome of Kupffer cell antigen presentation to a cloned murine Th1 lymphocyte depends on the inducibility of nitric oxide synthase by IFN-gamma.

The interaction between lymphocytes and the resident hepatic macrophage, the Kupffer cell (KC), is relevant to the phenomenon of immune tolerance to Ags entering the liver. Tolerance to Ag administered via the portal vein can be prevented by the rare earth lanthanide metal, gadolinium (Gd). Therefore, we studied the ability of OVA-responsive, H-2d-restricted Th1 clones to proliferate in response to KCs from DBA/2J (H-2d) mice that had been injected with either saline (control) or a Gd solution. Whereas control KCs functioned as effective APCs, KCs from Gd-injected mice (GdKC) were incapable of sustaining the proliferative response of the Th1 clone to the 16 mer of OVA (323-339). This lack of proliferation was determined not to be caused by impaired Ag processing, but rather was the result of IFN-gamma-stimulated nitric oxide (NO) release by the APC: 1) In vitro addition of the nitric oxide synthase (NOS) inhibitor NG-methyl-L-arginine (NMMA) restored the ability of the Gd-treated KC to stimulate clone proliferation. 2) Additional of anti-IFN-gamma, but not anti-IL-2 or anti-IL-4, prevented the induction of NOS in the Gd-exposed KC and was associated with clone proliferation. 3) IFN-gamma levels from clone-GdKC-OVA cocultures closely paralleled the nitrite released by GdKCs. 4) Only the addition of rIFN-gamma, and not IL-2 or IL-4, to cultures of purified GdKCs resulted in the release of nitrite. The results of the study suggest an autocrine loop initiated by the interaction of the clone's TCR with the class II MHC molecule presenting processed OVA on the surface of KC. This interaction stimulates the Th1 lymphocyte to release IFN-gamma, which in turn induces NO release by KCs isolated from Gd-injected mice. This release of NO blocks Th1 proliferation. Such a feedback loop may have particular relevance to Ag-specific tolerance, which is not only induced by the administration of Ag into the portal vein, but is also prevented by Gd pretreatment of the recipient animal.

Immunological tolerance to a defined myelin basic protein antigen administered intrathymically.

To explore the mechanisms responsible for the development of tolerance to allografts after intrathymic (IT) injection of alloantigen, the well-defined model of experimental autoimmune encephalomyelitis (EAE), which mimics the human autoimmune disease multiple sclerosis, was used. This inflammatory neurologic syndrome is initiated by myelin basic protein (MBP)-reactive CD4+ T lymphocytes restricted to self-MHC class II molecules. Naive adult, EAE-susceptible Lewis (RT1(1) rats were treated IT, i.v., or i.p. with a single dose (100 micrograms) of guinea pig-myelin basic protein (GP-MBP 1-176) in PBS plus 1 ml rabbit anti-rat lymphocyte serum i.p. Twenty-one days later, all rats were challenged by intradermal hind footpad injections of 50 micrograms GP-MBP in PBS emulsified in CFA. Only IT, but not i.p. or i.v., administration of GP-MBP plus anti-lymphocyte serum conferred marked resistance to a subsequent systemic challenge of GP-MBP, as demonstrated by the prevention of weight loss and paralysis characteristic of EAE. The IT administration dramatically decreased the size and number of histologic perivascular infiltrates observed per visual field in spinal cord of the tolerant animals and decreased GP-MBP-specific T lymphocyte in vitro proliferation (p < 0.01), whereas proliferation to a nonspecific mitogen (Con A) was not altered. With the addition of rIL-2, the decreased Ag-specific proliferative responses of IT-treated animals increased to control levels. Adoptive transfer of 100 x 10(6) splenocytes from tolerant hosts i.v. to naive syngeneic Lewis rats challenge with 100 micrograms GP-MBP in CFA had no effect on clinical or histologic EAE. Exposure of MBP to maturing thymocytes results in functionally immunounresponsive lymphocytes and prevention of autoimmune EAE.

The Kupffer cell in endotoxin tolerance: mechanisms of protection against lethal endotoxemia.

Kupffer cells (KC) of the hepatic sinusoid respond to endotoxemia by producing mediators which promote or inhibit systemic inflammatory responses. Sublethal lipopolysaccharide (LPS) pretreatment confers tolerance to the lethality of a subsequent LPS exposure. However, the precise role of the KC in endotoxin tolerance (ET) remains unclear. This study evaluated the effect of ET induction upon the rat KC production of the mediators tumor necrosis factor-alpha (TNF-alpha), prostaglandin E2 (PGE2), and interleukin-6 (IL-6), and upon the in vivo phagocytic capacity of the KCs. 3 days prior to KC isolation, age-matched rats received either 5 mg/kg LPS (ET) or normal saline (nontolerant, NT), which protected 100% of the ET rats against an LPS dose 3 days later which was lethal in 72% of NT rats. On an in vitro LPS rechallenge, ET KC produced significantly lower amounts of TNF than NT KC (p < .01). In contrast, the ET KC produced significantly more PGE2 (p < .05) and IL-6 (p < .001) than the NT KC. The percentage of KC phagocytosing fluorescent latex spheres in vivo was increased 7-fold in the ET rats. Thus, ET induction, which protects rats against subsequent lethal endotoxemia, selectively alters KC mediator production and phagocytic capacity. These findings strongly implicate the KC in the mediation of early endotoxin tolerance.

Autoregulation by eicosanoids of human Kupffer cell secretory products. A study of interleukin-1, interleukin-6, tumor necrosis factor-alpha, transforming growth factor-beta, and nitric oxide.

OBJECTIVE: Methods employed previously to analyze the secretory behavior of rodent Kupffer cells (KC) were used to examine the human KC's secretory response to lipopolysaccharide (LPS). SUMMARY BACKGROUND DATA: As the resident hepatic macrophage, the KC resides at the interface between the portal and systemic circulations. Consequently, this cell may play an integral role in the immune response to antigens and bacteria in the sinusoid. Study of cytokine production by the KC has relied predominantly on the rat as the source of these cells. Whether human KCs respond similarly to rat KCs after LPS stimulation has been a matter of speculation. METHODS: Kupffer cells obtained from seven human livers were tested under conditions identical to those used to study rat KCs. Kupffer cells rested for 12 hours after isolation were stimulated with LPS (2.5 micrograms/mL). Arginine concentration in the culture medium varied from 0.01 to 1.2 mM. To examine the role of eicosanoids, parallel culture wells received indomethacin (10 microM). Culture supernatants were assayed for interleukin-1 (IL-1), interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-alpha), transforming growth factor-beta (TGF-beta), prostaglandin E2 (PGE2), and nitric oxide. RESULTS: Similar to the rat KC, LPS-stimulated human KCs released IL-1, IL-6, TNF-alpha, TGF-beta, and PGE2. However, unlike rat KCs, nitric oxide could not be detected, regardless of whether the human KCs were exposed to LPS, interferon-gamma (INF-gamma), or LPS + IFN-gamma. Similar to rat KCs, indomethacin prevented PGE2 release while significantly upregulating TNF-alpha, IL-1, and IL-6, but not TGF-beta, consistent with an autoregulatory control of eicosanoids over proinflammatory cytokines. As has been shown in the rat, physiologic levels of L-arginine (0.01 mM) significantly enhanced LPS-induced PGE2 secretion relative to the response in medium containing standard L-arginine concentration (1.2 mM); however, unlike the rat KC, the human's cytokine response to LPS was not downregulated by this enhanced PGE2 release. CONCLUSIONS: Although many functional features are shared by rat and human KCs, significant differences do exist. Such discrepancies reinforce the need to proceed with caution when generalizing from the results obtained in other species to human physiology.

Lanthanide "blockade" of antigen-presenting cells suppresses lymphocyte proliferation by inducing nitric oxide synthesis.

Introduction of antigen into the portal venous circulation can induce a state of antigen-specific tolerance. Reversal of this phenomenon by the administration of gadolinium chloride (GD) before portal venous inoculation with antigen has been thought to result from the inhibition of macrophage phagocytosis by GD, thereby permitting passage of antigen through the hepatic sinusoids into the systemic circulation where sensitization can occur. We recently demonstrated that i.v. GD also greatly enhances the ability of the hepatic macrophage, or Kupffer cell (KC), to suppress lymphocyte proliferation in vitro. In this study, lymphocytes from ovalbumin (OVA)-immunized Lewis rats were cocultured with OVA plus KC or adherent splenocytes from syngeneic rats injected i.v. with GD or saline. Whereas small numbers of either hepatic or splenic untreated macrophages enhanced lymphocyte proliferation, both macrophage populations treated with GD were markedly inhibitory. Neither the presence of lymph node macrophages nor of normal KC was able to prevent the effect of GD-treated KC indicating a mechanism of active suppression rather than defective macrophage antigen uptake, processing, and presentation. While cyclooxygenase inhibition had no effect, the addition of inhibitors of nitric oxide synthase, i.e., aminoguanidine and NG-methyl-L-arginine, fully reversed the suppression of proliferation by GD-treated macrophages. Nitrite, but not PGE2, levels correlated with the degree of inhibition by the GD-treated macrophages. The suppression of lymphocyte proliferation and elevated supernatant nitrite levels associated with the presence of GD-treated macrophages were reversed when the macrophages were preincubated with EDTA indicating that GD's effect was directly on macrophages and due to its persistence in cultured cells. The findings of this study suggest that GD's in vivo prevention of portal venous tolerance may be due to enhanced nitric oxide synthesis whose effect on lymphocytes could be to inhibit adhesion or replication in the hepatic sinusoid.

Lymphocyte suppression by Kupffer cells prevents portal venous tolerance induction: a study of macrophage function after intravenous gadolinium.

Ag administration into the portal vein can induce specific tolerance to that Ag, known as portal venous tolerance. Because intrahepatic mechanisms of tolerance induction are still largely undefined, we studied the in vitro response of OVA-sensitized Lewis rat lymphocytes to OVA presented by normal syngeneic rat Kupffer cells (KC) or KC that had been treated in vivo with gadolinium chloride (GD), a rare earth metal, which prevents the induction of portal venous tolerance. KC (2.5 x 10(4)) were able to present OVA to 5 x 10(5) OVA-sensitized APC-depleted lymphocytes as effectively as could lymph node APC. However, the use of GD-treated KC was associated with a significantly (P < 0.001) impaired response of OVA-sensitized APC-depleted lymphocytes to OVA. Although GD nearly abrogated in vivo phagocytosis of fluorescent latex beads by both KC and adherent splenocytes, expression of the class II MHC molecule (Ia) by KC was only slightly reduced by GD treatment. Unresponsiveness of OVA-sensitized lymphocytes to OVA was not related to enhanced PGE2 release by GD-treated KC, as determined both by PGE2 levels in culture supernatants and by cyclooxygenase inhibition. However, the marked ability of GD-treated KC to inhibit the response to OVA by primed lymph node populations containing lymphocytes and APCs supports an active suppressive mechanism. Prevention of the induction of portal venous tolerance by GD, the lack of in vitro KC Ag presentation by GD-treated KC, and active immunosuppression by GD-treated KC support a model of tolerance induction within the liver wherein Ag presentation and lymphocyte proliferation are necessary for the development of tolerance.

Beneficial effect of syngeneic pancreatic islet transplantation on liver atrophy and hepatocellular function after portacaval shunt.

BACKGROUND: Hepatic insufficiency, which continues to be a source of morbidity after portacaval shunt (PCS), can be prevented by syngeneic pancreatic islet transplantation into the portal vein before PCS. This study investigated the ability of syngeneic pancreatic islet transplantation after PCS to prevent hepatic atrophy and rescue hepatocellular function. METHODS: Approximately 1200 to 1400 syngeneic rat pancreatic islets were transplanted through a heparinized catheter into the left lobes of the liver 3, 7, and 21 days after end-to-side PCS. Normal rats received no treatment, and PCS control rats received PCS only, without islet transplantation. Hepatocellular function (caffeine clearance) and hepatic blood flow (indocyanine green clearance) were analyzed at 42 and 49 days after PCS. On day 51 after PCS, the left and right lobes of the liver were divided, weighed, and sectioned for histologic studies. RESULTS: Caffeine clearance in the animals at 3 days (p less than 0.05) and at 7 days (p less than 0.05) after end-to-side PCS was significantly improved versus control PCS animals, indicating that hepatocellular function could be rescued after creation of a PCS. Indocyanine green clearance of all groups with PCS was significantly (p less than 0.001) decreased versus normal animals, showing that hepatic blood flow was uniformly decreased by PCS in all groups. The weight of the transplanted left lobes was significantly greater than the untransplanted right lobes of the groups at 3 days (p less than 0.01) and at 7 days (p less than 0.05) after end-to-side PCS compared with control animals, indicating that liver atrophy was prevented in the islet-transplanted lobes but not in those lobes without a transplant. CONCLUSIONS: Islet transplantation early after PCS can prevent liver atrophy and significantly improve hepatocellular function.

Diurnal rhythm of alpha 2-noradrenergic receptors in the paraventricular nucleus and other brain areas: relation to circulating corticosterone and feeding behavior.

The paraventricular nucleus alpha 2-noradrenergic system and the glucocorticoid hormone, corticosterone, are known to modulate feeding behavior and exhibit a circadian pattern which may be related to the natural periodicity of feeding in the rat. The results of the present study indicate that the binding of [3H]p-aminoclonidine to alpha 2-noradrenergic receptors specifically in the paraventricular nucleus varies concomitantly with plasma corticosterone levels, as well as spontaneous feeding. A monophasic peak of paraventricular noradrenergic receptor binding is detected at the onset of the dark period, when corticosterone levels are highest and feeding is initiated. On the other hand, the supraoptic nucleus exhibits the reverse diurnal pattern, i.e., a significant decline of [3H]p-aminoclonidine binding at the onset of the dark period. Other hypothalamic and extra-hypothalamic areas fail to show significant changes in alpha 2-noradrenergic receptors as a function of the diurnal cycle. This study supports other evidence indicating a close interaction between circulating corticosterone and alpha 2-noradrenergic receptors in specific hypothalamic areas. It also reveals a potential importance for this interaction in control of the natural feeding rhythm.

Interaction between corticosterone and alpha-2-noradrenergic system of the paraventricular nucleus in relation to feeding behavior.

Food consumption elicited by injection of norepinephrine (NE) into the paraventricular nucleus (PVN) of satiated rats has been shown to be dependent on the glucocorticoid corticosterone. To determine the specific nature of this dependence of NE-induced feeding on corticosterone, the efficacy of PVN-injected NE and its interaction with peripherally administered corticosterone was examined in adrenalectomized rats. NE, at doses ranging from 10 to 160 nmol, failed to elicit a reliable feeding response in these adrenalectomized animals. This loss of noradrenergic responsiveness developed at least as early as 4 h after adrenalectomy and continued until the end of the test sequence 6-8 weeks post-surgery. Single subcutaneous injections of corticosterone, administered to adrenalectomized rats, significantly restored the NE feeding response when injected 15-120 min prior to NE PVN injection, but not when administered 5 min before. Corticosterone was effective at doses of 0.5-4.0 mg/kg. Tests with other steroid hormones, namely, the synthetic glucocorticoid dexamethasone, the mineralocorticoid deoxycorticosterone, and the gonadal hormones testosterone and progesterone, were generally ineffective in restoring the sensitivity of the PVN to noradrenergic stimulation. Radioimmunoassay of circulating corticosterone in adrenalectomized rats, as a function of dose and time after injection of corticosterone, indicated that physiological levels of the hormone, at least 3 micrograms%, are required for a patent behavioral effect. These findings demonstrate a specific, sensitive, and rapid dependence of the PVN alpha 2-noradrenergic eating response on circulating corticosterone.

Hypophysectomy disturbs the noradrenergic feeding system of the paraventricular nucleus.

Injection of norepinephrine (NE) into the hypothalamic paraventricular nucleus (PVN) of satiated rats is known to stimulate eating behavior. In addition, drinking behavior is potentiated just prior to the onset of eating, followed by a strong inhibition of water intake. To understand the relationship between these PVN noradrenergic phenomena and endocrine processes associated with the PVN, chronically hypophysectomized animals were tested for their behavioral responsiveness to PVN NE injection. Pituitary ablation was found to abolish the NE-elicited eating response and the NE drinking suppressive effect. However, hypophysectomy had no impact on the NE-elicited preprandial drinking response, nor did it affect drinking produced by carbachol, angiotensin, and histamine, or the feeding and drinking responses induced by insulin. These results demonstrate that hypophysectomy disturbs PVN noradrenergic mechanisms in a behaviorally and pharmacologically specific specific manner.

Noradrenergic feeding elicited via the paraventricular nucleus is dependent upon circulating corticosterone.

Feeding behavior elicited by injection of norepinephrine (NE) into the paraventricular nucleus (PVN) of satiated rats has been shown to be abolished by hypophysectomy. To determine the specific nature of this dependence of NE's action on pituitary hormones, the efficacy of PVN-injected NE was examined in rats subjected to hypophysectomy, as well as to adrenalectomy, thyroidectomy, and gonadectomy, and also in operated rats receiving hormone replacement therapy. The feeding response induced by NE was almost completely abolished in adrenalectomized as well as hypophysectomized animals, but remained unimpaired after thyroidectomy and gonadectomy. The NE response was significantly restored in hypophysectomized rats by daily subcutaneous injections of corticosterone, but not by thyroxine, testosterone, insulin, or the mineralocorticoid deoxycorticosterone. In adrenalectomized rats, subcutaneous corticosterone implants as well as daily corticosterone injections (as opposed to deoxycorticosterone injections) effectively restored the NE eating effect. Radioimmunoassay of plasma corticosterone indicated that the level of hormone was positively correlated with the strength of the animals' response to the NE injection. These findings demonstrate that the loss of response to NE subsequent to hypophysectomy is due to a disruption of the hypothalamo-pituitary-adrenal axis. The glucocorticoids of the adrenal gland appear to be the essential humoral factor interacting permissively with PVN-injected NE to elicit feeding.