Iron chelates bind nitric oxide and decrease mortality in an experimental model of septic shock

The hydroxamic acid siderophore ferrioxamine B [FeIII(HDFB)+] and the iron complex of diethylenetri-aminepentaacetic acid [FeIII(DTPA)2-] protected mice against death by septic shock induced by Corynebacterium parvum + lipopolysaccharide. Although FeIII(DTPA)2- was somewhat more effective than FeIII(HDFB)+, the iron-free ligand H4DFB+ was significantly more effective than DTPA. The hydroxamic acid chelator has a much higher iron affinity than the amine carboxylate, allowing for more efficient formation of the FeIII(HDFB)+ complex upon administration of the iron-free ligand. Electrochemical studies show that FeIII(DTPA)2- binds NO stoichiometrically upon reduction to iron(II) at biologically relevant potentials to form a stable NO adduct. In contrast, FeIII(HDFB)+ is a stable and efficient electrocatalyst for the reduction of NO to N2O at biologically relevant potentials. These results suggest that the mechanism of protection against death by septic shock involves NO scavenging and that particularly effective drugs that operate a low dosages may be designed based on the principle of redox catalysis. These complexes constitute a new family of drugs that rely on the special ability of transition metals to activate small molecules. In addition, the wealth of information available on siderophore chemistry and biology provides an intellectual platform for further development.

Efficacy of treatment with the iron (III) complex of diethylenetriamine pentaacetic acid in mice and primates inoculated with live lethal dose 100 Escherichia coli

The iron (III) complex of diethylenetriamine pentaacetic acid (DTPA iron [III]) protected mice and baboons from the lethal effects of an infusion with live LD100 Escherichia coli. In mice, optimal results were obtained when DTPA iron (III) was administered two or more hours after infection. Prevention of death occurred in spite of the fact that the adverse effects of TNF-alpha were well underway in the mouse model. The half-life of DTPA iron (III) was 51 +/- 9 min in normal baboons; primary clearance was consistent with glomerular filtration. In septic baboons, survival was observed after administration of two doses of DTPA iron (III) at 2.125 mg/kg, the first one given before, or as late as 2 h after, severe hypotension. Administration of DTPA iron (III) did not alter mean systemic arterial pressure, but did protect baboons in the presence of high levels of TNF-alpha and free radical overproduction. Furthermore, exaggerated production of nitric oxide was attenuated. The mechanism of protection with DTPA iron (III) is not obvious. Because of its ability to interact in vitro with free radicals, its poor cell permeability, and its short half-life, we postulate that DTPA iron (III) and/or its reduced form may have protected the mice and baboons by sequestration and subsequent elimination of free radicals (including nitric oxide) from their systems.

Xenogeneic ICAM-1 gene transfer suppresses tumorigenicity and generates protective antitumor immunity

Tumorigenicity in Fischer rats was not significantly reduced when the rat ICAM-1 gene was overexpressed in the rat tumor cell lines, JM-1 and SST-2. When these rat tumor cell lines were genetically modified with a gene encoding human ICAM-1, tumorigenicity was dramatically reduced. Expression of xenogeneic ICAM-1 did not alter the growth rate, expression of the major histocompatibility complex, nor morphological appearance of the cells. However, it did facilitate a tumor-specific immunological recognition and rejection of the genetically modified tumor cells. This effect resulted in a tumor-specific, long-term protective immunity directed against genetically unmodified tumor cells. Most importantly, administration of tumor cells genetically modified with genes encoding xenogeneic ICAM-1 can facilitate an immunological response to genetically unaltered pre-existing tumors. Transferring splenocytes from animals 'vaccinated' with the xenogeneic ICAM-1 gene altered tumor cells was able to transfer the antitumor response into recipient animals. Furthermore, transfer of CD8+ T lymphocytes produced the same result. These results suggested that tumors specific CD8+ T lymphocytes were activated by the xenogeneic altered tumor cells. THis activation generated the long-term, tumor-specific immunity.

Targets of nitric oxide in a mouse model of liver inflammation by Corynebacterium parvum

Treatment of mice with Corynebacterium parvum induces chronic inflammation. This treatment followed by an injection of lipopolysaccharide (LPS) produces hepatic necrosis and death. We examined liver tissue by using electron paramagnetic resonance (EPR) spectroscopy and found that, in addition to the previously reported nonheme nitrosyl complexes, heme nitrosyl complexes were also formed. Hemoglobin nitrosyl complexes measured in the whole blood of mice treated with C. parvum were not increased after additional LPS treatment. However, this treatment significantly increased the heme nitrosyl complexes in the liver, whereas the nonheme nitrosyl complex concentration was unaffected. EPR signals from whole blood and liver tissues from mice treated with C. parvum and C. parvum + LPS were inhibited by prolonged treatment with NG-monomethyl-L-arginine (L-NMA). Nitric oxide (.NO) is known to bind to cytochrome P450 heme, and we consistently found a suppression of EPR signals attributable to ferric low-spin cytochrome P450/P420 peaks in the livers of mice treated with C. parvum and C. parvum + LPS. By performing analyses of EPR spectra obtained from hepatocytes exposed to .NO, we were able to unambiguously identify EPR signals attributable to cytochrome P420 and nonheme nitrosyl complexes in the livers of both treatments. Deconvolution of the composite in vivo EPR spectra indicated that hemoglobin nitrosyl complexes contributed weakly in the C. parvum livers, but threefold more in the C. parvum + LPS livers, suggesting that hemorrhage may have occurred. Experiments with L-NMA treatment revealed that this additional .NO production did not correlate with hepatic necrosis and onset of death. Immunoprecipitation of liver cytosols from C. parvum- and (C. parvum + LPS)-treated mice using an antibody against mouse inducible nitric oxide synthase showed that this enzyme was indeed present in the cytosolic fractions and was absent in those from control livers. Our novel detection of cytochrome P420 nitrosyl complex in vivo may be linked to any role of hepatic P450's functions during liver inflammation.

Ca2+ ionophore-induced cyclic adenosine-3',5'-monophosphate elevation in human neutrophils. A calmodulin-dependent potentiation of adenylate cyclase response to endogenously produced adenosine: comparison to chemotactic agents

The cyclic adenosine-3',5'-monophosphate (cAMP) elevation caused by exposure of human neutrophils to the Ca2+ ionophore A23187 was prevented when endogenously produced adenosine was either removed by preincubation with adenosine deaminase or blocked from binding to the adenosine receptor by antagonists [theophylline or (E)-4-(1,2,3,6-tetrahydro-1,3-dimethyl-2,6-dioxo-9H-purin-8-yl)cinnamic acid]. In the absence of endogenous adenosine, A23187 potentiated the neutrophil cAMP response to 2-chloroadenosine, prostaglandin E1, and isoproterenol. When neutrophil suspensions were preincubated with concentrations of Ro 20-1724, which appeared to maximally inhibit cAMP phosphodiesterase, A23187 was still able to substantially elevate cAMP levels, suggesting that A23187 increases cAMP by amplifying adenylate cyclase responsiveness to the agonist rather than by inhibiting cAMP phosphodiesterase. The ability of A23187 to augment the cAMP elevation caused by 2-chloroadenosine was persistent over a 10-min period. The neutrophil cAMP elevations caused by chemoattractants leukotriene B4, C5a, and N-formyl-L-methionyl-L-leucyl-L-phenylalanine (FMLP) were all prevented when endogenously produced adenosine was eliminated from the cell suspensions by the addition of adenosine deaminase. The A23187-induced cAMP elevation was inhibited completely by the calmodulin inhibitors chlorpromazine, trifluoperazine and N-(6-aminohexyl)-5-chloro-1-naphthalenesulfonamide, whereas cAMP levels induced by FMLP, leukotriene B4 and C5a were less affected. It appears that A23187 raises cAMP in human neutrophils by a calmodulin-dependent potentiation of adenylate cyclase responsiveness to endogenously produced adenosine while the chemoattractant-induced cAMP elevations (FMLP), leukotriene B4, and C5a), although possibly Ca2+ dependent, are less sensitive to calmodulin inhibitors and may involve additional biochemical events.

3-deazaadenosine inhibits leukocyte adhesion and ICAM-1 biosynthesis in tumor necrosis factor-stimulated human endothelial cells

Previous reports demonstrate that cultured human umbilical vein endothelial cells (HEC) treated with TNF and other inflammatory mediators show an increased capacity to adhere human neutrophils. This increase is associated with the up-regulation of intercellular adhesion molecule 1 (ICAM-1) and other adhesion molecules on the HEC surface. We have found that 200 microM 3-deazaadenosine (c3Ado) prevented this TNF-induced increase in HEC adhesiveness. This effect resulted from interactions of c3Ado with HEC and not with polymorphonuclear neutrophils. Transport of c3Ado into the HEC was required for its activity, as evidenced by antagonism with the nucleoside transport inhibitor, nitrobenzylthioinosine. Treatment of HEC with c3Ado led to the intracellular buildup of S-adenosylhomocysteine and to the metabolic formation of S-3-deazaadenosylhomocysteine and 3-deazaadenosine 5'-triphosphate, events that appeared not to contribute to c3Ado activity. Exogenous L-homocysteine potentiated c3Ado activity, and this potentiation was prevented by the S-adenosylhomocysteine hydrolase inhibitor, periodate-oxidized adenosine. By using the mAb RR1/1, we have determined that c3Ado also inhibited the TNF-induced expression of ICAM-1 on the surface of the HEC, as well as cytosol-associated ICAM-1. Northern blot and in vitro translation analyses of poly(A+) RNA from c3Ado-treated HEC revealed that this nucleoside analog selectively decreased steady-state levels of ICAM-1 mRNA. The capacity of c3Ado to selectively inhibit HEC adhesiveness, ICAM-1 production, and steady-state levels of ICAM-1 mRNA may contribute to the drug's activity as an anti-inflammatory agent.

3-deazaadenosine inhibits leukocyte adhesion and ICAM-1 biosynthesis in tumor necrosis factor-stimulated human endothelial cells

Previous reports demonstrate that cultured human umbilical vein endothelial cells (HEC) treated with TNF and other inflammatory mediators show an increased capacity to adhere human neutrophils. This increase is associated with the up-regulation of intercellular adhesion molecule 1 (ICAM-1) and other adhesion molecules on the HEC surface. We have found that 200 microM 3-deazaadenosine (c3Ado) prevented this TNF-induced increase in HEC adhesiveness. This effect resulted from interactions of c3Ado with HEC and not with polymorphonuclear neutrophils. Transport of c3Ado into the HEC was required for its activity, as evidenced by antagonism with the nucleoside transport inhibitor, nitrobenzylthioinosine. Treatment of HEC with c3Ado led to the intracellular buildup of S-adenosylhomocysteine and to the metabolic formation of S-3-deazaadenosylhomocysteine and 3-deazaadenosine 5'-triphosphate, events that appeared not to contribute to c3Ado activity. Exogenous L-homocysteine potentiated c3Ado activity, and this potentiation was prevented by the S-adenosylhomocysteine hydrolase inhibitor, periodate-oxidized adenosine. By using the mAb RR1/1, we have determined that c3Ado also inhibited the TNF-induced expression of ICAM-1 on the surface of the HEC, as well as cytosol-associated ICAM-1. Northern blot and in vitro translation analyses of poly(A+) RNA from c3Ado-treated HEC revealed that this nucleoside analog selectively decreased steady-state levels of ICAM-1 mRNA. The capacity of c3Ado to selectively inhibit HEC adhesiveness, ICAM-1 production, and steady-state levels of ICAM-1 mRNA may contribute to the drug's activity as an anti-inflammatory agent.

Chemotactic peptide induces cAMP elevation in human neutrophils by amplification of the adenylate cyclase response to endogenously produced adenosine

The transient increase in human neutrophil cAMP levels induced by the chemoattractant N-formyl-L-methionyl-L-leucyl-L-phenylalanine (FMLP) is shown to be caused by amplification of adenylate cyclase response to endogenously produced adenosine. The FMLP-stimulated increase in neutrophil cAMP was potentiated markedly by a nonmethylxanthine cAMP phosphodiesterase inhibitor (Ro 20-1724). By inhibiting the degradation of newly formed cAMP, Ro 20-1724 rendered the FMLP-induced cAMP elevation persistent rather than transient. The role of endogenously produced adenosine in this phenomenon is demonstrated by the ability of either adenosine deaminase or theophylline, an adenosine receptor antagonist, to prevent FMLP-stimulated cAMP elevation. The general nature of the FMLP-potentiated cAMP response is indicated by the finding that FMLP-treated neutrophils, in the presence of exogenously supplied adenosine deaminase, exhibited augmented cAMP generation in response to three different types of receptor agonists: 2-chloroadenosine, prostaglandin E1, and L-isoproterenol. Moreover, like the neutrophil cAMP increase caused by FMLP alone, the ability of FMLP to augment cAMP response to 2-chloroadenosine in adenosine deaminase-treated cells was short-lived and declined after 1.0 min of exposure to FMLP. Preincubation of neutrophil suspensions with the adenylate cyclase inhibitor SQ 22,536 completely prevented FMLP-induced cAMP generation. Furthermore, when neutrophil suspensions were preincubated with concentrations of Ro 20-1724, which apparently maximally inhibit cAMP phosphodiesterase, a 30-s incubation with FMLP still resulted in substantially elevated cAMP levels. It therefore appears that FMLP raises cAMP by activating adenylate cyclase rather than inhibiting cAMP phosphodiesterase.

Separable image warping with spatial lookup tables

Image warping refers to the 2-D resampling of a source image onto a target image. In the general case, this requires costly 2-D filtering operations. Simplifications are possible when the warp can be expressed as a cascade of orthogonal 1-D transformations. In these cases, separable transformations have been introduced to realize large performance gains. The central ideas in this area were formulated in the 2-pass algorithm by Catmull and Smith. Although that method applies over an important class of transformations, there are intrinsic problems which limit its usefulness.The goal of this work is to extend the 2-pass approach to handle arbitrary spatial mapping functions. We address the difficulties intrinsic to 2-pass scanline algorithms: bottlenecking, foldovers, and the lack of closed-form inverse solutions. These problems are shown to be resolved in a general, efficient, separable technique, with graceful degradation for transformations of increasing complexity.

Inhibition of neutrophil adherence to endothelial cells by 3-deazaadenosine

Treatment of cultured human umbilical vein endothelial cells (HUVE) with tumor necrosis factor alpha (TNF-alpha) increases their capacity to adhere human neutrophils. We have found that 3-deazaadenosine (c3Ado), when added in conjunction with TNF-alpha, inhibited this increase in neutrophil adherence. This activity of c3Ado was potentiated by the addition of L-homocysteine thiolactone (Hcy). The ability of c3Ado to inhibit neutrophil adherence to HUVE may contribute to the anti-inflammatory activity of this nucleoside analogue.

Effects of adenosine on neutrophil polarization induced by N-formyl-methionyl-leucyl-phenylalanine, sodium propionate and colchicine

Polarization of human neutrophils (a characteristic bipolar shape change) can be induced by the chemotactic peptide N-formyl-methionyl-leucyl-phenylalanine (FMLP); sodium propionate, which causes a rapid acidification of the cytosol; or colchicine, which disrupts microtubules. We have previously reported that adenosine, endogenously produced in human neutrophil suspensions, inhibits FMLP-induced polarization. We report here that endogenously produced adenosine also inhibits sodium propionate-induced polarization but has no effect on colchicine-induced polarization. These results suggest that neutrophil polarization may be a multistep process inducible by compounds that trigger different biochemical events.

3-Deazaadenosine 5'-triphosphate: a novel metabolite of 3-deazaadenosine in mouse leukocytes

Evidence has been obtained for the metabolic formation of small amounts (1-2% of the ATP pool) of 3-deazaadenosine 5'-triphosphate (c3ATP) from 3-deazaadenosine (c3Ado) in mouse cytolytic lymphocytes and mouse resident peritoneal macrophages. With intact leukocytes, pharmacological evidence was obtained that adenosine kinase was not the enzyme chiefly responsible for the phosphorylation of c3Ado. Moreover, in the presence of MgCl2, NaCl and IMP, purified rat liver 5'-nucleotidase catalyzed the phosphorylation of c3Ado to 3-deazaadenosine 5'-monophosphate (c3AMP). Two lines of evidence suggest that the metabolic formation of c3ATP is not involved in the inhibition of leukocyte function caused by c3Ado. First, the inhibitory action of c3Ado on antibody-dependent phagocytosis and lymphocyte-mediated cytolysis was reversed markedly upon removal of the drug from the medium. However, the intracellular content of c3ATP remained constant in lymphocytes and macrophages after removal of c3Ado. Second, in macrophages and in lymphocytes, similar intracellular amounts of c3ATP were formed from both c3Ado and 3-deazaadenine under conditions in which the former was biologically active and the latter was essentially inactive. Thus, it appears unlikely that the novel c3ATP metabolite is of relevance for the mechanism of action of c3Ado in mouse leukocytes.

Taxol inhibits N-formyl-methionyl-leucyl-phenylalanine (FMLP)-induced human neutrophil polarization and H2O2 production while decreasing [3H]FMLP binding

We have studied the effect of taxol on two N-formyl-methionyl-leucyl-phenylalanine (FMLP)-induced neutrophil functions and the possible mechanism by which it inhibits these functions. Taxol inhibited FMLP-induced human neutrophil polarization (a characteristic change in neutrophil shape in response to a chemotactic stimulus) and H2O2 generation. Taxol also decreased the specific binding of [3H]FMLP to human neutrophils at 4 degrees C. The decreased binding of FMLP to its receptor may be responsible for the inhibition by taxol of FMLP-induced polarization and H2O2 generation.

Imidazo[4,5-c]pyridines (3-deazapurines) and their nucleosides as immunosuppressive and antiinflammatory agents

A variety of imidazo[4,5-c]pyridines (3-deazapurines) were synthesized. With use of these aglycons as pentosyl acceptors, the corresponding ribonucleosides and 2'-deoxyribonucleosides were prepared by an enzymatic method involving transfer of the pentosyl moiety from appropriate pyrimidine nucleosides. With most of the imidazo[4,5-c]pyridines, the products obtained from the enzyme-catalyzed reactions were pentosylated exclusively in the 1-position. However, some 3-pentosylation occurred with aglycons that had H or N3 in the 4-position. In addition to the 2'-deoxy congener of the ribonucleoside of 4-amino-1H-imidazo[4,5-c]pyridine, the 5'-deoxy and 2',5'-dideoxy congeners were synthesized. All of the aglycons and their nucleosides were tested for toxicity to mammalian cells in culture. None were markedly cytotoxic. These compounds were also evaluated for their ability to inhibit lymphocyte-mediated cytolysis in vitro. 3-Deazaadenosine (23) and its 2'-deoxy congener (38) were the most potent inhibitors (ED50 = 20 microM). In addition to these two in vitro tests, in vivo inhibition of the inflammatory response in the rat carregeenan pleurisy model was determined. 3-Deazaadenosine (23) was the most potent compound (ED50 = 3 mg/kg) in this in vivo test.

Effects of adenosine deaminase inhibitors on lymphocyte-mediated cytolysis

Four compounds that inhibit adenosine deaminase, erythro-9-(2-hydroxy-3-nonyl)adenine, 2'-deoxycoformycin, coformycin, and 9-(1-hydroxy-2-octyl)adenine have been studied in an in vitro lymphocyte-mediated cytolysis assay. At low concentration (congruent to 10 microM) these agents enhance the activities of a number of inhibitory purine nucleosides, including adenosine and 2'-deoxyadenosine. The LMC-inhibitory activity of Ado but not dAdo is further enhanced by 5-iodotubercidin, uridine, 4-(3-butoxy-4-methoxybenzyl)-2-imidazolidinone, or L-homocysteine and is antagonized by theophylline. The inhibition of LMC by Ado and dAdo is increased by nitrobenzyl-thioinosine. Lymphocyte-mediated cytolysis was inhibited by EHNA or HOA alone (IC50 congruent to 150 microM), but not by dCF and CF (even at 400 microM). Inhibition of LMC by EHNA, HOA, Ado, or dAdo could not be attributed to changes in nucleoside 5'-triphosphate or S-adenosylhomocysteine levels. Inhibition of LMC by Ado appears to be related to increases in lymphocyte cAMP levels, while the mechanism of action of dAdo remains obscure. Lymphocyte-mediated cytolysis may be inhibited by EHNA and HOA through modulation of cAMP metabolism.

3-Deazaadenosine-induced disorganization of macrophage microfilaments

3-Deazaadenosine (c3Ado) has been reported to inhibit a number of cellular functions. These biological effects of c3Ado have generally been attributed to its ability to act as inhibitor and substrate of S-adenosylhomocysteine hydrolase. In this report, it is revealed by fluorescence microscopy that c3Ado caused disorganization of the microfilament system of mouse macrophages at concentrations (greater than or equal to 5 microM) similar to those that inhibited antibody-dependent phagocytosis and zymosan-stimulated H2O2 production by these cells. Inhibition of phagocytosis and perturbation of microfilaments by c3Ado were completely abrogated by washing the macrophages free of this agent and allowing the cells a 30-min recovery period. Furthermore, these effects of c3Ado on phagocytosis and microfilaments appeared to be independent of the increase in S-adenosylhomocysteine and S-3-deazaadenosylhomocysteine that occurred in these macrophages. First, periodate-oxidized adenosine and 3-deaza(+/-)aristeromycin, two other inhibitors of S-adenosylhomocysteine hydrolase that caused greater increases in macrophage S-adenosylhomocysteine than did c3Ado, had no effect on either phagocytosis or microfilaments. Second, pretreatment of macrophages with periodate-oxidized adenosine (to inhibit S-adenosylhomocysteine hydrolase) prevented the subsequent metabolism of c3Ado to S-3-deazaadenosylhomocysteine but did not diminish the effects of c3Ado on phagocytosis or microfilaments. These results demonstrate that c3Ado can perturb the microfilament system of cells and provide an alternative mechanism for the biological effects of c3Ado.

Effects of calmodulin antagonists on immune mouse lymphocytes

The nature of the Ca2+ requirement of lymphocyte-mediated cytolysis (LMC) has been explored pharmacologically with a number of putative calmodulin antagonists. N-(6-Aminohexyl)-5-chloro-1-naphthalenesulfonamide (W-7), N-(6-aminohexyl)-1-naphthalenesulfonamide (W-5), trifluoperazine, and chlorpromazine were found to inhibit LMC (IC50 values = 8.9, approximately 50, 7.4, and 9.4 microM, respectively) at concentrations which were not detectably toxic to either the effector or the target cell. Pimozide inhibited LMC by 50% at 15 microM but caused a substantial decrease in lymphocyte ATP content and viability at this concentration. 1-[Bis(p-chlorophenyl)methyl]-3-[2,4-dichloro-beta-(2,4-dichlorobenzy loxy) phenethyl]imidazolium chloride (R 24 571, calmidazolium), which has been reported to be the most potent antagonist of isolated calmodulin, caused a marked decrease in lymphocyte ATP content and viability at concentrations greater than 4 microM and inhibited LMC only slightly at similar concentrations. Trifluoperazine sulfoxide and chlorpromazine sulfoxide were not inhibitory to LMC at less than or equal to 20 microM. LMC was inhibited in a sustained manner when cytolytic lymphocytes, but not target cells, were pretreated separately with W-7 or chlorpromazine at 37 degrees and were then washed free of exogenous drug prior to the start of the LMC assay. The above cellular effects of the calmodulin antagonists were reduced in magnitude when the serum concentration in the culture medium was increased (from 5% to 20%). The inhibition of LMC by micromolar concentrations of W-7, trifluoperazine, and chlorpromazine, as well as the relative inactivities of W-5 versus W-7 and of the sulfoxide derivatives of trifluoperazine and chlorpromazine, are consistent with calmodulin's being a lymphocyte receptor whose occupancy by Ca2+ is required for the performance of this cytolytic function. However, this conclusion must be tempered by the finding that even W-7, trifluoperazine, and chlorpromazine can exert nonspecific effects on the energy metabolism and viability of the cytolytic lymphocytes at concentrations of drug severalfold higher than those required to inhibit LMC.

Antagonism by taxol of effects of microtubule-disrupting agents on lymphocyte cAMP metabolism and cell function

Several microtubule-disrupting agents (colchicine, demecolcine, vinblastine, vincristine, podophyllotoxin, and nocodazole) have been shown to inhibit lymphocyte-mediated cytolysis. These agents also enhanced the prostaglandin E1-induced rise in cAMP levels in these cytolytic lymphocytes. Taxol, a natural product alkaloid that has been shown to enhance microtubule polymerization and to stabilize microtubules, antagonized both of these effects of the microtubule-disrupting agents in the cytolytic lymphocytes. Taxol also antagonized the enhancement of cAMP increases by colchicine in lymphocytes stimulated by 2-chloroadenosine, isoproterenol, and cholera toxin. The enhancement of the prostaglandin E1-induced cAMP response caused by treatment of the lymphocytes with either cytochalasin B or 3-deazaadenosine in the presence or absence of L-homocysteine was not antagonized by taxol. Taxol, colchicine, or the combination of these two agents did not affect ATP levels in cytolytic lymphocytes. These results support a modulatory role for microtubules in both the cytolytic process and the production of cAMP in these lymphocytes.

3-Deazaadenosine. S-adenosylhomocysteine hydrolase-independent mechanism of action in mouse lymphocytes

3-Deazaadenosine (c3Ado) inhibits the ability of specifically sensitized mouse lymphocytes to lyse tumor cells, and this effect of c3Ado on immune function is accompanied by a buildup of both S-adenosylhomocysteine (AdoHcy) and S-3-deazaadenosylhomocysteine (c3AdoHcy) within the lymphocytes (Zimmerman, T. P., Wolberg, G., and Duncan, G. S. (1978) Proc. Natl. Acad. Sci. U.S.A. 75, 6220-6224). Several types of evidence have now been obtained which indicate that neither of these latter two biochemical events contributes to this biological activity of c3Ado. First, periodate-oxidized adenosine (Adox) and 3-deaza(+/-)aristeromycin, two compounds which are not inhibitory to this lymphocyte function, were both more effective than c3Ado in elevating lymphocyte AdoHcy levels and in inhibiting protein carboxymethylation within these cells. Second, pretreatment of lymphocytes with Adox (to inhibit AdoHcy hydrolase) prevented the metabolic formation of c3AdoHcy during subsequent incubation of the cells with c3Ado without diminishment of the biological activity of c3Ado. Third, the lymphocyte buildup of c3AdoHcy exhibited a concentration dependence upon c3Ado which was biphasic and greatly dissimilar to that for the inhibition of lymphocyte function by c3Ado. These results are not compatible with the widely held view that c3Ado, as a single agent, affects various cellular functions as a consequence of elevations in intracellular AdoHcy and/or c3AdoHcy and suggest that c3Ado inhibits this particular lymphocyte function by an unknown mechanism which is independent of the interaction of c3Ado with AdoHcy hydrolase. During this study, it was found that pretreatment of lymphocytes with Adox prevented the potentiation by homocysteine of this immunoinhibitory activity of c3Ado and reduced the biological effect of the c3Ado/homocysteine combination to that observed with c3Ado alone. This result indicates that the ability of homocysteine to potentiate this biological activity of c3Ado requires the metabolic formation of c3AdoHcy catalyzed by AdoHcy hydrolase.

In vivo and in vitro antitumor activity expressed by cells of concomitantly immune mice

Growth of a primary tumor is often accompanied by the development of resistance to subsequent challenge implants of the same tumor, i.e., concomitant immunity. Using the P815 mastocytoma tumor, the kinetics of concomitant immunity was found to be governed by duration of exposure to the tumor and tumor mass. By implanting small "challenges" prior to the immunizing tumor, resistance to the growth of existing tumor foci was demonstrated. Winn-type assays revealed that antitumor activity was present in cell populations from the peritoneal exudate and lymph node draining the tumor. Peritoneal exudate cells, when infused systemically, were also able to confer protection against P815 mastocytoma challenge, suggesting their role as mediators of concomitant immunity. The 51Cr release technique indicated that cytolytic activity in lymph node cells, peritoneal exudate cells, and the spleen was present over a time course parallel to the kinetics of in vivo challenge. The peritoneal resident cell population was only slightly active; thus, effectors accumulated in the inflammatory exudate. Removal of specific subsets of cells from effector populations with antibody to surface markers and complement produced similar effects on both Winn and cytolytic assays. Anti-Thy 1.2 ablated measurable activity. It was substantially but not completely reduced by anti-Lyt 1.1 and only to a small degree by anti-Lyt 2.1.

Inhibition of lymphocyte function by 9-deazaadenosine

9-Deazaadenosine (c9Ado), a novel C-nucleoside, has been found to inhibit lymphocyte-mediated cytolysis (LMC) in a time-dependent manner. c9Ado inhibited LMC by 50% at concentrations of 10 and 0.07 microM after drug-pretreatment periods of 3 and 22 hr, respectively, although a 1-hr pretreatment of cytolytic lymphocytes with 100 microM c9Ado had no effect upon this lymphocyte function. c9Ado was metabolized rapidly and extensively to 9-deazaadenosine 5'-triphosphate (c9ATP) both by mouse cytolytic lymphocytes and by human erythrocytes. Adenosine kinase purified from rabbit liver phosphorylated c9Ado with a Km of 200 microM and a Vmax of 8% that for adenosine. The metabolic buildup of c9ATP in lymphocytes was accompanied by a large, time-dependent decrease in cellular ATP and by smaller percentage decreases in CTP, UTP and GTP. Among other biochemical effects examined, c9Ado was found to cause a decrease in lymphocyte cAMP content and appeared to be neither an inhibitor nor a substrate for S-adenosylhomocysteine hydrolase. Consistent with this latter result, L-homocysteine thiolactone had no effect on the inhibition of LMC by c9Ado. Neither the inhibition of LMC by c9Ado nor the metabolic formation of c9ATP in lymphocytes was affected by erythro-9-(2-hydroxy-3-nonyl)adenine (EHNA), indicating that c9Ado is not a substrate for adenosine deaminase. 5-Iodotubercidin, a non-competitive inhibitor (Kis = 9 nM, Ku = 20 nM) of adenosine kinase, prevented the above effects of c9Ado on lymphocyte function, c9ATP formation, and ATP levels. Either complete preservation (with coformycin) or partial replenishment (with adenosine plus EHNA) of ATP levels in c9Ado-treated lymphocytes resulted in partial restoration of cytolytic function to cells containing large amounts of c9ATP. These results suggest that c9Ado is inhibitory to LMC both because it causes a decrease in the absolute concentration of ATP within the cytolytic lymphocytes and because it permits the establishment within these cells of an unfavorable c9ATP:ATP ratio which impedes the utilization of ATP in a reaction essential to the execution of this lymphocyte function.

Effect of acyclovir on various murine in vivo and in vitro immunologic assay systems

In two in vitro tests, lymphocyte-mediated cytotoxicity and neutrophil chemotaxis, acyclovir showed no inhibitory effects at concentrations as high as 600 microM. The compound inhibited rosette formation with nonimmune mouse lymphocytes in vitro by approximately 50 percent at 15.8 microM. The significance of this inhibition is unclear. In four in vivo tests in mice which measured humoral and cell-mediated immunity (complement-dependent cellular cytotoxicity, complement-independent cellular cytotoxicity, delayed hypersensitivity and graft versus host reaction) acyclovir showed no inhibitory effects at single doses up to 200 mg/kg given on day 2 after antigenic stimulation. Four daily doses of acyclovir at 50 mg/kg per day had no effect on the numbers of hemolytic IgM antibody-forming cells in the spleen when assayed on day 4. At the higher dosage of 100 mg/kg per day for four days, there was a slight reduction in the numbers of these cells. There was no significant decrease in hemagglutinin or hemolysin antibody titers after four daily doses of acyclovir up to 200 mg/kg.

Inhibition of lymphocyte-mediated cytolysis and cyclic AMP phosphodiesterase by erythro-9-(2-hydroxy-3-nonyl)adenine

The adenosine deaminase (ADA) inhibitor erythro-9-(2-hydroxy-3-nonyl)adenine (EHNA), at low concentrations (less than 10 microM), enhances the inhibitory activity of adenosine against lymphocyte-mediated cytolysis (LMC) without itself being inhibitory. At higher concentrations, EHNA alone is inhibitory to LMC with an IC50 of 160 microM. This inhibition is reversible upon washout, appears to affect an early stage of the lytic process, and does not appear to involve changes in basal levels of cyclic AMP (cAMP), ribonucleoside 5'-triphosphate pool sizes, S-adenosylhomocysteine levels, or protein carboxymethylation. EHNA does enhance the cAMP response of cytolytic lymphocytes (CL) to activators of adenylate cyclase such as prostaglandin E1. EHNA inhibits lymphocyte high-affinity cAMP phosphodiesterase at immunosuppressive levels, exhibiting hyperbolic mixed-type inhibition (Ki = 83 microM, alpha = 0.47, beta = 0.18). Whereas inhibition of intralymphocytic ADA is complete at low concentrations (less than 25 microM) of EHNA, inhibition of LMC and intralymphocytic cAMP phosphodiesterase increases linearly with EHNA concentration to at least 200 microM. The presence of 200 microM EHNA during the centrifugation of mixtures of CL and EL4 leukemia target cells leads to increased CL cAMP levels. 2'-Deoxycoformycin, a more potent ADA inhibitor than EHNA, is not inhibitory to LMC and shows none of these cAMP-related effects. These results suggest that CL-target cell contact stimulates adenylate cyclase in the CL and that EHNA inhibits LMC due to its enhancement of this target cell-stimulated elevation of cAMP.

Immune reaction of tumor-bearing mice to Propionibacterium acnes and the antitumor effect of the bacteria

The relationship between immunological reaction to Propionibacterium acnes (PA) and the antitumor effect of the injected bacterium was investigated. The aim was to determine whether the strength of the immune reaction to the bacterium can be used to predict its antitumor effectiveness. C3Hf/Sed mice received SC injections (right thigh) of viable cells of a methylcholanthrene-induced fibrosarcoma. When the tumor grew to 5 mm, the hosts received 350 micrograms PA IV as the antitumor treatment. Cellular immunity (footpad test) to PA was assayed in one group of these mice 14 days later, and in the other anti-PA agglutinins were determined 28 days later. The PA injection cured 22 of 58 mice in the first, and 20 of 46 mice in the second group. Footpad reaction and agglutinin titers to PA in cured mice were not statistically different from those in mice eventually killed by the tumor. Therefore, the strength of the immune reaction to PA in tumor-bearing mice could not be used to predict the antitumor effectiveness of the bacterium.

Immunosuppressive effects of the S-adenosylhomocysteine hydrolase inhibitor, 3-deazaadenosine

Immunosuppressive effects of 3-deazaadenosine (3-DAA), an inhibitor of S-adenosylhomocysteine hydrolase, were tested in vivo in immune assays against sheep red blood cells (SRBC), involving serum titrations for hemagglutinins and hemolysins, cellular cytotoxicity tests and the direct plaque-forming cell assay. At daily doses up to 100 mg/kg, the compound was suppressive when injected before antigen and the effect appeared to be dose-dependent (ED50 = 52.6 +/- 4.9 mg/kg). When doses of 25 mg/kg of 3-DAA were given before antigen, co-injections of 250 mg/kg of L-homocysteine (L-HC) potentiated the suppressive effect, although L-HC alone was inactive. Daily administration of 100 mg/kg of 3-DAA or 250 mg/kg of L-HC alone was not suppressive when given after the antigen; however, in combination they were able to induce suppression. The possible biochemical mechanisms of the suppression, particularly those involving the inhibition of S-adenosylmethionine-dependent methylation reactions, are discussed.

In vitro cytotoxicity expressed by cells active against established tumors in vivo

Although antitumor activity by host cells has been documented in vivo and in vitro, the cellular relationships between these two classes of studies are not clear. Cells capable of causing the regression of solid tumors are generated in lymph nodes draining sites of immunization with Corynebacterium parvum:irradiated P815 mastocytoma admixtures. These cells are active in a 51Cr release assay at a low effector:target ratio producing a characteristic low level of specific 51Cr release which required 24 hr for optimal development. The activity is immunologically specific for the immunizing tumor and is mediated by nonadherent, rapidly dividing (vinblastine-sensitive) cells. They are absent in thymectomized animals and susceptible to alpha-Thy 1.2 antibody and complement. They are present in peritoneal exudates, consistent with the systemic resistance demonstrable in the animal model. The properties and development kinetics of effector cells measured by 51Cr release correlate closely with those of cells showing in vivo activity, supporting the identity of the two populations.

Modulation of cyclic AMP metabolism by S-adenosylhomocysteine and S-3-deazaadenosylhomocysteine in mouse lymphocytes

Mouse lymphocytes incubated with micromolar concentrations of adenosine or 3-deazaadenosine, in medium supplemented with L-homocysteine, rapidly accumulated supramillimolar concentrations of S-adenosylhomocysteine (AdoHcy) or S-3-deazaadenosylhomocysteine (c3AdoHcy), respectively. Lymphocytes thus preloaded with high levels of AdoHcy or c3AdoHcy exhibited markedly enhanced (5- to 40-fold) cyclic AMP responses to prostaglandin E1, adenosine, 2-chloroadenosine, isoproterenol, and cholera toxin. This enhancement of cyclic AMP response by intracellular AdoHcy or c3AdoHcy was attributable both to amplification of the activity of adenylate cyclase [ATP pyrophosphate-lyase (cyclizing), EC 4.6.1.1] and to inhibition of cyclic AMP phosphodiesterase (3',5'-cyclic-nucleotide 5'-nucleotidohydrolase, EC 3.1.4.17). Basal and prostaglandin E1- and isoproterenol-stimulated activities of adenylate cyclase, assayed in lymphocyte homogenates, were increased 1.3- to 2.0-fold after treatment of the cells with homocysteine plus either adenosine or 3-deazaadenosine. AdoHcy and c3AdoHcy were found to be competitive inhibitors (with Ki values of 1.7 and 4.8 mM, respectively) of the high-affinity cyclic AMP phosphodiesterase present in lymphocyte homogenates. It is evident, therefore, that increased cellular levels of AdoHcy or c3AdoHcy can affect cellular physiology via perturbation of cyclic AMP metabolism as well as via inhibition of S-adenosylmethionine-dependent methylation reactions.

Adenosine analogues as substrates and inhibitors of S-adenosylhomocysteine hydrolase in intact lymphocytes

A number of adenosine analogues have been examined for their ability to interact with S-adenosyl-L-homocysteine (SAH) hydrolase in intact mouse lymphocytes. In the presence of erythro-9-(2-hydroxy-3-nonyl)adenine, 3-deazaadenosine, 8-azaadenosine, formycin A, 2-aminoadenosine, 2-fluoroadenosine, N6-methyladenosine, N6-hydroxyadenosine, purine ribonucleoside and inosine were each metabilized to radioactive analogues of SAH when cells were labeled with either L-[2-3H]methionine or L-[35S]homocysteine. Tubercidin was shown to undergo metabolism both to S-[3H]tubercidinyl-L-methionine and to S-[3H]tubercidinyl-L-homocysteine in cells labeled with [2-3H]methionine. 9-beta-D-Arabinofuranosyladenine and 2'-deoxyadenosine caused marked elevations of [3H]SAH in cells preloaded with [2-3H]methionine but were not themselves metabolized detectably to SAH analogues. Adenine and 5'-deoxyadenosine also caused substantial elevations of [3H]SAH under these same conditions. Some of the adenosine analogues shown to be metabolized to SAH analogues also caused an elevation of SAH in the lymphocytes. These results indicate the potential of adenosine analogues to interfere with cellular methylation reactions due either to their inhibition of SAH hydrolase or to their metabolism, via this enzyme, to SAH analogues.

Inhibition of lymphocyte-mediated cytolysis by 3-deazaadenosine: evidence for a methylation reaction essential to cytolysis

3-Deazaadenosine (deazaAdo) inhibits lymphocyte-mediated cytolysis in vitro at micromolar concentrations and is potentiated markedly in this activity by L-homocysteine thiolactone. DeazaAdo alone causes a rapid, dose-dependent buildup of S-[(3)H]adenosylhomocysteine (AdoHcy) and S-[(3)H]adenosylmethionine in cytolytic lymphocytes labeled with L-[2-(3)H]methionine; smaller amounts of S-3-[(3)H]deazaadenosylhomocysteine (deazaAdoHcy) are also formed in these cells. The simultaneous addition of deazaAdo and L-homocysteine thiolactone to the lymphocytes results in a massive intracellular accumulation of deazaAdoHcy. Both the inhibition of lymphocyte-mediated cytolysis and the cellular accumulation of [(3)H]AdoHcy caused by deazaAdo alone are reversed rapidly by removal of drug from the medium. However, the inhibition of cytolysis and the large cellular buildup of deazaAdoHcy resulting from treatment of the lymphocytes with deazaAdo plus L-homocysteine thiolactone are dissipated more slowly under these same conditions. Unlike adenosine, deazaAdo is not potentiated in its inhibition of lymphocyte-mediated cytolysis by Ro 20-1724 [4-(3-butoxy-4-methoxybenzyl)-2-imidazolidinone], an inhibitor of cyclic AMP phosphodiesterase, and has little or no effect upon the level of lymphocytic cyclic AMP. DeazaAdo is not metabolized detectably to 5'-nucleotides in the lymphocytes and does not cause a decrease in the pool sizes of CTP, UTP, ATP, or GTP. Both AdoHcy and deazaAdoHcy have been reported to be powerful inhibitors of a variety of S-adenosylmethionine-utilizing methyltransferases. The present results, therefore, indicate that the effect of deazaAdo upon lymphocyte-mediated cytolysis is due ultimately to the inhibition of an unidentified but crucial methyltransferase within the cytolytic lymphocytes and provide an insight into the biochemical processes involved in lymphocyte-mediated cytolysis.

Antibody to Corynebacterium parvum in normal human and animal sera

Using a microtiter bacterial agglutination test, we have estimated antibodies to Corynebacterium parvum in "normal" human and "normal" and immune animal sera. Widely differing levels of C. parvum antibodies were found in the normal human sera. The median titer for all 310 human sera was 1:128, whereas that for the 1- to 17-year and 18- to 50-year subgroups was 1:64 and 1:512, respectively. Antibody titers in the various animal species were generally much lower.

2-Fluoroadenosine 3':5'-monophosphate. A metabolite of 2-fluoroadenosine in mouse cytotoxic lymphocytes

2-Fluoroadenosine (F-Ado) is a potent inhibitor of lymphocyte-mediated cytolysis studied in vitro. The inhibition of cytolysis by F-Ado was potentiated markedly by an inhibiotr (Ro 20-1724) of adenosine 3':5'-monophosphate (cAMP) phosphodiesterase and, unlike the inhibition caused by adenosine, was irreversible when the cytotoxic lymphocytes were incubated with F-Ado and were then washed free of exogenous nucleoside. Incubation of cytotoxic lymphocytes with F-Ado resulted in the rapid, dose-dependent formation of 2-fluoroadenosine 5'-triphosphate (F-ATP); the build-up of F-ATP within these cells was accompanied by a reciprocal depletion of ATP. Once formed intracellularly, the F-ATP was not diminished during a subsequent 30-min incubation of the cells in F-Ado-free medium. 2-Fluoroadenosine 3':5'-monophosphate (F-cAMP), a novel compound, was synthesized chemically. This cAMP analogue was found to be highly cross-reactive in a radioimmunoassay specific for cAMP and to be equipotent to cAMP in its ability to activate a crude preparation of protein kinase derived from rat brain. A column chromatographic procedure was devised whereby F-cAMP and cAMP could be purified simultaneously from tissue extracts. Treatment of cytotoxic lymphocytes with F-Ado resulted in the formation of presumptive F-cAMP in amounts greater than that of cAMP, as determined by the concentration of F-Ado added to the medium and was not observed when the lymphocytes were incubated with either adenosine or 2-chloroadenosine, two agents which caused large increases in cAMP. The simultaneous presence of Ro 20-1724 enhances greatly the formation of F-cAMP from F-Ado without affecting the pool size of F-ATP. Removal of exogenous F-Ado from cells previously incubated with this drug and subsequent incubation of these cells in drug-free medium did not result in a substantial reduction in intracellular F-Ado (via prior incubation with F-Ado); 2'-deoxyadenosine was also effective in this capacity, while 9-beta-D-arabinofulanosyladenine was without effect. The level of cAMP was elevated transiently, in a dose-dependent manner, by F-Ado, and returned to control value after removal of exogenous F-Ado from the cells. Ro 20-1724 enhanced greatly this transient elevation of cAMP caused by F-Ado.

Adenosine inhibition of lymphocyte-mediated cytolysis: possible role of cyclic adenosine monophosphate

The in vitro destruction of tumor cells by specifically sensitized mouse lymphocytes was inhibited by adenosine; this inhibition was markedly potentiated by the presence of an inhibitor of adenosine deaminase. The inhibition of cytolysis by adenosine was accompanied by a rapid elevation in lymphocytic adenosine 3',5'-monophosphate (cyclic AMP) concentrations. Both the inhibition of cytolysis and the elevation of cyclic AMP were reversed by prolonged incubation of the lymphocytes in the presence of adenosine or, more rapidly, by removal of the adenosine. Low concentrations of adenosine also caused an elevation of cyclic AMP in human lymphocytes, and this effect of adenosine may contribute to the lack of immune response associated with adenosine deaminase deficiency.

Mouse virulence of K(L) antigen-containing strains of Escherichia coli

The mouse virulence of two K antigen-containing (L variety) strains of Escherichia coli (serotype O2:K1) isolated from human septicemia, and of their variants which lacked K antigen, was studied. The strains containing envelope antigen (K+) were highly virulent when injected intracerebrally or when suspended in mucin and injected intraperitoneally. After intraperitoneal injection of E-107 K+ (but not K-), there was a marked initial growth in the peritoneal cavity followed by bacteremia and infection of all the organs examined. In the mucin-enhanced lethal infection, this growth continued until death of the animal; in the nonlethal infection, growth ceased and the count dropped quickly after approximately 5 hr. Host defenses were depressed greatly by intraperitoneally, but not intravenously, administered mucin. Bacteria were most virulent when injected intraperitoneally. In vitro phagocytosis of the K+ bacteria required opsonins not needed for phagocytosis of the smooth K- variants. Opsonins were found in immunized rabbit and normal mouse sera. Immune rabbit sera contained antibodies with anti-K specificity which were opsonic in vitro and highly protective in vivo when administered passively. There appears to be a lesser anti-O opsonic and protective activity involving one of the strains (E-107 K+), and colonial morphology, agglutination, and absorption tests indicated a low amount of K antigen on this organism. No anti-O opsonic or protective activity could be shown involving the other strain (E-102 K+). When standard serological typing procedures were used, these two strains appeared to be identical serologically, but they differed greatly in sensitivity to immune rabbit serum in phagocytosis experiments in vitro.