Extracellular ATP decreases trophoblast invasion, spiral artery remodeling and immune cells in the mesometrial triangle in pregnant rats.

INTRODUCTION: Preeclampsia is characterized by deficient trophoblast invasion and spiral artery remodeling, a process governed by inflammatory cells. High levels of the danger signal extracellular adenosine triphosphate (ATP) have been found in women with preeclampsia and infusion of ATP in pregnant rats induced preeclampsia-like symptoms such as albuminuria and placental ischemia. We hypothesized that ATP inhibits trophoblast invasion and spiral artery remodeling and affects macrophages and natural killer (NK) cells present in the rat mesometrial triangle. METHODS: Pregnant rats were infused with ATP or saline (control) on day 14 of pregnancy. Rats were sacrificed on day 15, 17 or 20 of pregnancy and placentas with mesometrial triangle were collected. Sections were stained for trophoblast cells, α-smooth muscle actin (spiral artery remodeling), NK cells and various macrophage populations. Expression of various cytokines in the mesometrial triangle was analyzed using real-time RT-PCR. RESULTS: ATP infusion decreased interstitial trophoblast invasion on day 17 and spiral artery remodeling on day 17 and 20, increased activated tartrate resistant acid phosphatase (TRAP)-positive macrophages on day 15, decreased NK cells on day 17 and 20, and decreased inducible nitric oxide synthase (iNOS)-positive and CD206-positive macrophages and TNF-α and IL-33 expression at the end of pregnancy (day 20). DISCUSSION: Interstitial trophoblast invasion and spiral artery remodeling in the rat mesometrial triangle were decreased by infusion of ATP. These ATP-induced modifications were preceded by an increase in activated TRAP-positive macrophages and coincided with NK cell numbers, suggesting that they are involved. CONCLUSION: Trophoblast invasion and spiral artery remodeling may be inhibited by ATP-induced activated macrophages and decreased NK cells in the mesometrial triangle in rat pregnancy.

OS053. Trophoblast invasion in the mesometrial triangle of atp infused pregnant rats.

INTRODUCTION: Poor placentation (disturbed and decreased trophoblast invasion) is a hallmark of preeclampsia (PE), which is a major complication of pregnancy. Unfortunately, the cause and mechanism of disturbed trophoblast invasion is still unknown. OBJECTIVES: The pro-inflammatory agent ATP has been shown to induce PE-like signs, after a single infusion in pregnant rats. These PE-like characteristics include proteinuria and decreased fetal weight. Since purinergic ATP receptors are expressed on trophoblast cells, we aimed to study the effect of ATP infusion on trophoblast invasion in pregnant rats in this pilot study. METHODS: Pregnant rats received a single ATP (n=4) or saline (control,ni=5) infusion via a permanent jugular vein cannula on day 14 of pregnancy. At the time of maximal trophoblast invasion (day 17 of pregnancy) rats were sacrificed and placentas with mesometrial triangle were collected, fixed in zinc-buffer and embedded in paraffin. 4 µm sections were stained with monoclonal α-cytokeratin antibodies. In the mesometrial triangle, the maternal part of the rat placenta, the percentage of surface area of trophoblast invasion was evaluated using computerized image analysis. Also, the depth and width of invasion were analyzed by subdividing the mesometrial triangle in three concentric depth levels of equal width. In addition, trophoblast invaded versus non-invaded spiral arteries in the mesometrial triangle were quantified. RESULTS: In the mesometrial triangle, no changes in percentage of surface area of trophoblast invasion and percentage of invaded spiral arteries were observed after ATP infusion. However, the pattern of trophoblast invasion appeared to be disturbed in ATP infused rats, with a decreased depth of invasion and an increased width of invasion, resulting in a trend towards a decreased depth/width ratio of trophoblast invasion in ATP infused rats. CONCLUSION: In this (pilot) study we showed an altered trophoblast invasion pattern in the mesometrial triangle of the placenta, although no significant differences in the total surface area of trophoblast invasion were seen in experimental versus control pregnant animals. e mechanism by which ATP induces this altered trophoblast invasion pattern and its potential contribution to the pathophysiology of this experimental PE in the pregnant rat awaits further investigation.

Nitric oxide inhibition enhances platelet aggregation in experimental anti-Thy-1 nephritis.

In the present paper we studied the role of nitric oxide radicals (NO) on platelet aggregation, fibrinogen deposition, superoxide formation, peroxynitrite formation, hemodynamics, and leukocyte migration in the Thy-1 model of glomerulonephritis. To first study the baseline kinetics of these parameters, groups of anti-Thy-1-treated rats were sacrificed at 1 h, 4 h, 24 h, 3 days, 7 days, and 14 days and compared to controls. Urinary protein excretion was significantly elevated in Thy-1 nephritis at 3 and 7 days. Glomerular macrophages, PMNs, and superoxide anion-positive cells were significantly increased in Thy-1 nephritis. Nitrotyrosine immunoreactivity was absent during the entire study period. Glomerular platelet aggregation was significantly increased in anti-Thy-1 injected rats at 1 h, 4 h, 24 h, and 3 days. Glomerular fibrinogen deposition was significantly elevated at all time points. To elucidate the role of NO in this process, additional groups of anti-Thy-1-injected rats were treated with the NOS inhibitor l-NAME and studied at 24 h. Urinary protein excretion was significantly higher in l-NAME treated Thy-1 rats compared to nontreated Thy-1 rats. Plasma and urine nitrite/nitrate levels were significantly lower in l-NAME-treated Thy-1 rats compared to nontreated Thy-1 rats. Compared to nontreated Thy-1 rats, there were no differences in intraglomerular leukocyte accumulation after treatment with l-NAME. In contrast, we observed a marked increase in platelet aggregation following l-NAME treatment. From these data we conclude that the inflammatory infiltrate in Thy-1 nephritis develops independent of NO radical production, whereas NO radicals prevent the accumulation of platelet aggregates.

Cytomegalovirus infection increases the expression and activity of ecto-ATPase (CD39) and ecto-5'nucleotidase (CD73) on endothelial cells.

We describe enhanced expression and enzymatic activity of ecto-ATPase and ecto-5'nucleotidase on CMV infected endothelial cells as compared to uninfected cells. These ectoenzymes play a major role in modulation of platelet activation and aggregation. Furthermore, adenosine has a modulatory effect upon inflammation. Addition of ATP, ADP or AMP to cultures of CMV infected or uninfected endothelial cells revealed increased turnover of AMP in CMV infected endothelial cells. In addition, the superoxide production by stimulated polymorphonuclear cells was inhibited in the presence of CMV infected endothelial cells as compared to uninfected cells, probably due to the enhanced activity of ecto-5'nucleotidase and associated to production of adenosine.

Induction of experimental proteinuria in vivo following infusion of human plasma hemopexin.

BACKGROUND: The human plasma constituent hemopexin (Hx), following incubation with renal tissue, is able to induce glomerular alterations in vitro that are similar to those seen in minimal change disease (MCD). Whether this acute phase reactant is also able to induce proteinuria and minimal change-like alterations in vivo is questioned. METHODS: In the first set of experiments, Hx (4.0 mg in 5.0 mL saline) or equal amounts of control fraction, that is, heat-inactivated Hx (HI-Hx), were infused into conscious rats (N = 6) that had been surgically equipped with a cannula inserted into the suprarenal artery (SRA), enabling direct contact of the infusate and the renal microvasculature. Each animal received HI-Hx at day 1 for 15 minutes (flow rate 20.0 mL/h), subsequently followed by saline for seven hours (Flow rate 5.0 mL/h), after which the cannula was disconnected. At day 2, identical infusions in the same rat were carried out, using native Hx. Urine samples collected every 30 minutes during the experiments were monitored for protein content using standard methods. In the second set of experiments, unilateral perfusion was done ex vivo in anesthetized rats with Hx (N = 5) or HI-Hx (N = 3; 1.5 mg/mL; 4.0 mL during 6 min). After reconnection of the circulation, urine samples of both kidneys were collected every 30 minutes during five hours via ureter cannulation. Urinary protein (expressed as the difference in excretion between perfused and nonperfused kidney) was calculated in mg/24 h. In additional experiments, rats were sacrificed two hours after perfusion of Hx or heat-inactivated (control) Hx (first set of experiments) or after five hours (second set of experiments), and kidneys were processed for immunohistochemical and ultrastructural examination. RESULTS: The results of experiment 1 show a significant increase of proteinuria after Hx infusion versus HI-Hx (means +/- SD, 41.91 +/- 16.01 mg/24 h vs. control, 21.22 +/- 5.69 mg/24 h; P

Pregnancy aggravates proteinuria in subclinical glomerulonephritis in the rat.

Because subclinical renal disease may be aggravated during pregnancy--as reflected in the occurrence of proteinuria, for example--we investigated whether a subclinical glomerulonephritis (SG) in the non-pregnant rat (passive Heymann nephritis), a condition without proteinuria, is aggravated when the animals become pregnant and, if so, whether this is associated with a glomerular inflammatory reaction. SG was induced in non-pregnant rats 8 days before pregnancy. On day -1, part of the group of rats became pregnant. Three experiments were performed. In experiment 1, 4-hour urine albumin excretions and blood pressure (tail cuff) were measured. In experiment 2, the glomerular filtration rate (GFR) was measured with the chromium 51-labeled ethylenediaminetetraacetic acid method, while in experiment 3, parameters characteristic of a glomerular inflammation were studied. Experiment 1 revealed that non-pregnant rats with SG did not exhibit proteinuria. However, after the rats became pregnant, a significant proteinuria occurred, without an increase in systolic blood pressure. Experiment 2 revealed that the GFR did not increase in pregnant rats with SG, while experiment 3 showed that only pregnant animals exhibited a significant glomerular inflammation; this glomerular inflammation was characterized by intraglomerular influx of polymorphonuclear cells and monocytes. The results suggest that an SG in the rat may flare up during pregnancy. This exacerbation is characterized by proteinuria and coincides with a glomerular inflammatory reaction. It is suggested that proteinuria and the glomerular inflammatory reaction are causally related and promoted by the pregnant condition.

Is 100KF an isoform of hemopexin? Immunochemical characterization of the vasoactive plasma factor 100KF.

The human vasoactive plasma factor 100KF has been proposed to play a role in minimal change disease in relapse. Since preliminary data suggested similarity between 100KF and the human plasma glycoprotein hemopexin (Hx), this study was conducted to compare 100KF with purified Hx for sequence homology, immunostaining properties in Western and dot-blot assays, ability to affect glomerular ecto-ATPase and glomerular polyanions in vitro, as well as their glomerular permeability increasing effect following alternate perfusion into the rat kidney ex vivo. 100KF was purified from normal pooled plasma according to standard chromatographic techniques, and from the same batch Hx was prepared using affinity chromatography. A second batch of Hx was prepared directly from human serum according to a standard protocol. (For comparison, additional Hx samples obtained from other centers were also included in the study.) The results show: (1) 100% homology of 100KF with plasma Hx after internal sequence analysis; (2) positive staining of the eluate with both monoclonal and polyclonal anti-Hx IgG as well as anti-100KF IgG in dot-blot assays, and similar bands on Western blotting using the same antibodies; (3) affection of glomerular polyanions and glomerular ecto-ATPase after incubation of kidney tissue with either 100KF or Hx (1.5 respectively 1.0 mg/ml; 1.0 h, 37 degrees C), as detected by computerized histochemical quantification; and (4) significant enhancement of urinary protein leakage after Hx perfusion followed by diluted rat serum into the rat kidney ex vivo (Hx: 210.65+/-49.79 microg protein leakage per min versus heat-inactivated Hx control: 112.2+/-49.18 microg per min [both n = 6]). From these data and from the observation that both Hx and 100KF activity can be inhibited by serine protease inhibitors but not by broad spectrum collagenase inhibitors, it is concluded that Hx may be closely related or identical to the active moiety of 100KF.

Effect of estradiol and progesterone on the low-dose endotoxin-induced glomerular inflammatory response of the female rat.

PROBLEM: Is the endotoxin-induced glomerular inflammatory response of the female rat under ovarian control? METHOD OF STUDY: Ovariectomized rats (OVX), with or without progesterone (OVX-P) or estradiol (OVX-E) treatment, as well as rats in the follicular or luteal phase of the ovulatory cycle were infused with endotoxin or saline and sacrificed 3 days later. Cryostat kidney sections were immunohistologically stained for the presence of neutrophils and monocytes (MO) and the expression of adhesion molecules. RESULTS: After endotoxin, the glomerular number of neutrophils and the number of MAC-1 positive cells were increased in luteal-phase and in OVX-P rats, and the number of glomerular Mø was increased in luteal-phase, OVX, OVX-E, and OVX-P rats. Endotoxin increased ICAM-1 expression in all groups of rats, except in follicular-phase rats. The glomerular number of LFA-1- and VLA-4 positive cells following endotoxin were only increased in OVX rats. CONCLUSIONS: It is concluded that endotoxin-induced monocyte infiltration and ICAM-1 expression are inhibited by a factor produced during the follicular phase, probably by developing follicles. Infiltration of neutrophils and expression of MAC-1, LFA-1, VLA-4 seem to be under control of progesterone or estradiol.

Pregnancy enhances the sensitivity of glomerular ecto-adenosine triphosphate-diphosphohydrolase to products of activated polymorphonuclear leukocytes.

To test the hypothesis that pregnancy enhances the sensitivity of glomerular ecto-adenosine triphosphate-diphosphohydrolase to products of activated polymorphonuclear leukocytes, cryostat-cut kidney sections of pregnant and cycling rats were exposed to activated polymorphonuclear leukocytes and subsequently stained for ecto-adenosine triphosphate-diphosphohydrolase activity. The results show that the levels of ecto-adenosine triphosphate-diphosphohydrolase activity of pregnant rats showed a significantly greater decrease after incubation with activated polymorphonuclear leukocytes than did those of cycling rats.

Oxygen-dependent injury by a human plasma factor associated with minimal change disease.

The mechanism by which a human plasma factor associated with proteinuria is able to cause experimental glomerular albumin leakage is unknown. This factor (called 100KF) is able to induce glomerular alterations in the rat kidney, similar to those seen in minimal change disease, including loss of glomerular sialoglycoproteins and decreased expression of glomerular ecto-ATPase. It was previously shown that 100KF is able to stimulate release of reactive oxygen species in inflammatory cells in vitro. This prompted us to test whether 100KF-induced injury is oxygen dependent. The expression of glomerular sialoglycoproteins and ecto-ATPase was evaluated by standard histochemistry and computerized image analysis and expressed in arbitrary units. Rat kidney sections were incubated with or without 100KF under normal or oxygen-poor, i.e., nitrogen, conditions, or with supplementation of superoxide dismutase (SOD, 100 U/ml). The effect of 100KF on glomerular ecto-ATPase was oxygen dependent (32.98+/-2.14 under air vs. 65.20+/-5.53 under nitrogen, P< or =0.01), in contrast to the 100KF-induced loss of glomerular sialoglycoproteins that was not significantly altered under nitrogen (62.67+/-10.08 under air vs. 61.74+/-26.05 under nitrogen). Supplementation of SOD to 100KF solution under normal incubation conditions also suggested oxygen-dependent impairment of glomerular ecto-ATPase. Alternate perfusion ex vivo of the rat kidney with 100KF followed by diluted plasma showed that enhanced leakage of plasma proteins could be inhibited with SOD, indicating oxygen dependency of this 100KF-induced enhanced permeability (60.25+/-19.32 microg urinary albumin/ml after 100KF perfusion vs. 25.23+/-12.05 microg/ml after 100KF plus SOD, P< or =0.01). We conclude that the action of 100KF upon specific glomerular matrix molecules is oxygen dependent, as is the albumin leakage induced by 100KF in the present ex vivo model.

Superoxide-mediated glomerulopathy in the endotoxin-treated pregnant rat.

In the present study the role of superoxide in the glomerular damage in the low-dose endotoxin-infused pregnant rats was investigated. On day 14 of pregnancy, 12 rats were infused for 1 h with 1.0 microgram/kg bw endotoxin via a permanent jugular vein cannula. Of these rats, 6 were treated with SOD both prior to endotoxin infusion (7,000 U/kg) and 30 min (7,000 U/kg) and 4 h (14,000 U/kg) after the start of the infusion (SOD rats). The other 6 rats received no SOD treatment (endotoxin rats). Control pregnant rats were infused for 1 h with saline (saline rats; n = 6). Urinary albumin was measured on days 15 and 19 of pregnancy. On day 21, rats were sacrificed and kidney specimens were snap-frozen. Cryostat kidney sections were stained for fibrinogen, ecto-ATP diphosphohydrolase (e-ATPase) activity, polymorphonuclear cells, monocytes and various adhesion molecules on the endothelium and the leukocytes. SOD treatment appeared to significantly prevent the increased urinary albumin excretion and the decrease of glomerular e-ATPase activity which were observed in endotoxin-treated rats. This effect of SOD treatment after endotoxin infusion was associated with a significant inhibition of glomerular monocyte influx and a significant inhibition of adhesion molecule expression (glomerular ICAM-1 and VCAM-1 and leukocyte LFA-1 and VLA-4). The present data suggest that in the endotoxin-infused pregnant rat, production of superoxide in the first few hours after the infusion plays a role in the induction of glomerular damage, leading to albuminuria and diminished e-ATPase expression during the following days.

Aspirin treatment of the low-dose-endotoxin-treated pregnant rat: pathophysiologic and immunohistologic aspects.

In the present study, we evaluated the effect of low-dose aspirin (acetylsalicylic acid (ASA); 1.0 mg/kg daily) on blood pressure, albumin excretion, glomerular fibrinogen deposits, and glomerular (basement) membrane-bound adenosine diphosphatase (ecto-ADPase) activity, as well as on glomerular inflammation in pregnant rats infused with low-dose endotoxin (1.0 mg/kg). Rats (day 14 of pregnancy) were infused with endotoxin (ET rats) or saline (control rats) and received ASA in their drinking water. These rats were compared with non-ASA-treated rats. Blood pressure and albumin excretion were measured from day 15 to day 21, and glomerular fibrinogen and ecto-ADPase activity were measured at day 21. Glomerular inflammation was evaluated at various times after the start of the infusion. The results show that treatment with ASA had a significant beneficial effect on hypertension and inflammation induced by endotoxin in pregnant rats, whereas it reduced albumin excretion and glomerular fibrinogen deposits in some of the rats.

Dephosphorylation of endotoxin by alkaline phosphatase in vivo.

Natural substrates for alkaline phosphatase (AP) are at present not identified despite extensive investigations. Difficulties in imagining a possible physiological function involve its extremely high pH optimum for the usual exogenous substrates and its localization as an ecto-enzyme. As endotoxin is a substance that contains phosphate groups and is usually present in the extracellular space, we studied whether AP is able to dephosphorylate this bacterial product at physiological pH levels. We tested this in intestinal cryostat sections using histochemical methods with endotoxin from Escherichia coli and Salmonella minnesota R595 as substrate. Results show that dephosphorylation of both preparations occurs at pH 7.5 by AP activity. As phosphate residues in the lipid A moiety determine the toxicity of the molecule, we examined the effect of the AP inhibitor levamisole in vivo using a septicemia model in the rat. The results show that inhibition of endogenous AP by levamisole significantly reduces survival of rats intraperitoneally injected with E. coli bacteria, whereas this drug does not influence survival of rats receiving a sublethal dose of the gram-positive bacteria Staphylococcus aureus. In view of the endotoxin-dephosphorylating properties of AP demonstrated in vitro, we propose a crucial role for this enzyme in host defense. The effects of levamisole during gram-negative bacterial infections and the localization of AP as an ecto-enzyme in most organs as well as the induction of enzyme activity during inflammatory reactions and cholestasis is in accordance with such a protective role.

Plasma endothelin-1 and tumor necrosis factor-alpha concentrations in pregnant and cyclic rats after low-dose endotoxin infusion.

Plasma endothelin-1 and tumor necrosis factor-alpha were determined in pregnant and cyclic rats after infusion of either endotoxin (1.0 microgram/kg of body weight) or saline solution. After endotoxin, but not after saline solution, administration there was a transient endothelin-1 response in pregnant rats but not cyclic rats. In both reproductive conditions there was an equally high transient tumor necrosis factor-alpha response after endotoxin.

Reproductive condition and the low-dose endotoxin-induced inflammatory response in rats. Glomerular influx of inflammatory cells and expression of adhesion molecules.

These experiments were designed to study the increased sensitivity of pregnant rats to endotoxin. Pregnant (Pr), cyclic (C), and progesterone (P)-treated pseudopregnant rats with or without a decidualized uterus (PSP and DEC rats, respectively) received infusions of an ultra-low dose of endotoxin (1.0 microg/kg BW) and were killed 3 days later. Pr, PSP, and DEC rats were infused on Day 14, C rats on diestrus. Endotoxin-infused rats were compared with saline-infused rats in the same reproductive conditions. The inflammatory reaction of the glomeruli of the kidneys was studied by immunohistochemical methods using 4-microm cryostat sections stained with specific monoclonal antibodies against neutrophils (polymorphonuclear cells, PMNs) and monocytes (MOs), and against the adhesion molecules ICAM-1 and VCAM-1 on the endothelium, and LFA-1, MAC-1, and VLA-4 on the leukocytes. Endotoxin infusion increased glomerular PMN and MO number in Pr, PSP, and DEC rats, all of which have elevated P levels, but not in C rats, which do not. The endotoxin-induced expression of adhesion molecules, associated with this influx of inflammatory cells, varied with the reproductive condition. In C rats there was no increased adhesion molecule expression after endotoxin treatment, in Pr rats there was increased expression of both the combinations ICAM-1/LFA-1 and VCAM-1/VLA-4. DEC rats did not express either of these combinations (although there was expression of ICAM-1); PSP rats expressed the combination ICAM-1/MAC-1. Adhesion molecule expression thus seems to be regulated by ovarian (e.g., P) and placental factors (e.g., of trophoblastic and decidual origin). Because the different combinations of adhesion molecules in the various reproductive conditions after exposure to endotoxin led to more or less the same leukocyte influx under these conditions, the increased sensitivity to endotoxin of pregnant individuals cannot be reduced to differences in leukocyte influx into the glomeruli.

Modulation of glomerular ECTO-ADPase expression by oestradiol. A histochemical study.

The effect of 17-beta-oestradiol (OE2) upon the activity of the glomerular anti-thrombotic ecto-enzyme ADPase was studied in cyclic and ovariectomized (OVX) Wistar rats. On day 0 (i.e. at the time of ovariectomy or 11 days after ovariectomy) rats received OE2-releasing Silastic implants or empty implants and were sacrificed on day 3, 10 or 21. Cryostat kidney sections were histochemically stained for ecto-ADPase activity using enzyme-histochemistry and glomerular reaction product was quantitatively evaluated by computerized image analysis. Both the histological distribution of reaction product in each glomerulus, as reflected by the relative glomerular area covered with reaction product, as well as enzyme activity, as reflected by staining intensity of the reaction product, were scored. The results show significantly decreased histological distribution after OVX; OVX, however, did not change enzyme activity. It further appeared that OE2 (partly) prevented the decrease of histological distribution in OVX rats, while the enzyme activity was significantly increased by exogenous OE2. In cyclic rats, OE2 did not change histological distribution, although OE2 significantly increased enzyme activity in these rats. It is concluded that glomerular ecto-ADPase expression in the rat kidney is influenced by one or more ovarian factor(s), a very likely candidate being oestradiol. These results may thus point to a dual action of OE2 upon haemostasis: In addition to the known enhancement of procoagulatory plasma factors by OE2, also anti-aggregatory effects may be stimulated by OE2 as reflected by upregulation of vessel wall associated ecto-ADPase activity.

A physiologic function for alkaline phosphatase: endotoxin detoxification.

Alkaline phosphatase (AP), a common enzyme present in many species including humans, has been studied extensively. Although the enzyme is routinely applied as a marker for liver function, its biologic relevance is poorly understood. The reason for this is obvious: the pH optimum of AP in vitro, as measured with the usual test substrates (+/-10.5), greatly exceeds the physiologic pH range as it occurs in biologic tissues. We now hypothesize that this relatively high pH optimum in vitro is related to dissociation of acidic groups in the protein preparation, which leads to the formation of negatively charged groups in the vicinity of the active site of the enzyme. These negatively charged groups may promote the activity of AP. We examined the possibility that endotoxin is a natural substrate for this enzyme because this phosphorylated substance is able to supply multiple negatively charged residues in the microenvironment of the enzyme at a physiologic pH level. Phosphate groups in the endotoxin molecule are known to be essential for the biologic activities of this bacterial product. The present study demonstrates that in intestinal and renal tissue specimens in vitro, AP is endowed with endotoxin dephosphorylating activity at pH levels closer to the physiologic range. This is also illustrated by our experiments in vivo showing that the toxicity of endotoxin is significantly reduced after exposure to AP preparations, as tested by inducing a local intradermal inflammatory reaction in rats. Collectively, our data suggest that the ubiquitous enzyme AP may accomplish protection against endotoxin, an equally ubiquitous product of Gram-negative bacteria that may cause lethal complications after an infection with these micro organisms.