Kinin metabolism in the perfused ventilated rat lung. II: Influence of ventilation, perfusion, and perfusate composition variation on bradykinin metabolism in uninjured lung.

Bradykinin metabolism by peptidases of the pulmonary endothelium has been investigated in the previously uninjured, ventilated, and asanguinously perfused rat lung. The influence of short-duration (up to 20 min) abnormal ventilation and perfusion conditions on bradykinin metabolism was assessed. Neither variation of the oxygen concentration (0 to 45%) nor omission of carbon dioxide in the ventilatory gas altered bradykinin metabolism significantly. Tidal volume variation did not alter bradykinin metabolism, and exclusion of one lung from the perfusion circuit reduced the capacity to degrade bradykinin proportionately. Acidification of the perfusion medium to pH 5 did not alter bradykinin metabolism. Acetylsalicylic acid in the perfusate protected the lung from an otherwise irreversible pressure increase associated with high-dose bradykinin perfusion. Endotoxin and hydrogen peroxide in the perfusate did not alter bradykinin metabolism. However, ammonia in the ventilatory gas caused immediate pulmonary edema, diminished lung capacity to metabolize bradykinin and altered the pattern of bradykinin metabolic products. The pulmonary endothelium itself, in the absence of blood, maintains its capacity to metabolize bradykinin under an extraordinary range of conditions.

Quality control of hemoglobin solutions. I. The purity of hemoglobin before modification.

It is not yet clear whether the toxicity of hemoglobin solutions is due to hemoglobin per se, contaminants, or both. To elucidate the problem, it is first necessary to prepare "pure hemoglobin," i.e., hemoglobin completely free of (a) bacterial endotoxins, (b) phospholipids of the red cell membrane, and (c) other proteins and/or peptides. The methods found most sensitive in the determination of such purity are here presented. Consensus on the adoption of these methods and on the meaning of the term "purity" would help to achieve (a) a better understanding of the toxicity problem, (b) a more meaningful exchange of information among investigators and (c) the ultimate development of nontoxic solutions.

Kinin metabolism in the perfused ventilated rat lung. I: Bradykinin metabolism in a system modeling the normal, uninjured lung.

Bradykinin (BK) in an asanguinous salt solution was perfused through intact rat lung. BK concentration varied from 0.0015 to 89 microM. Below 17 microM, the amount degraded was greater than or equal to 90% of the dose. The BK fragment distributions expressed as a percentage of the BK dose degraded were constant. The BK fragments formed and percentage yields were Pro-Pro (2-3) 49%, Arg-Pro-Pro-Gly-Phe (1-5) 32%, Pro-Pro-Gly-Phe-Ser-Pro (2-7) 6%, Arg-Pro-Pro-Gly-Phe-Ser-Pro (1-7) 6%, Arg-Pro-Pro (1-3) 3% and residual BK 4%. Above 17 microM, the amount of BK degraded was not proportional to the dose. Captopril and enalaprilat inhibited BK degradation, and their maximum inhibitions were about 50% and 30%, respectively. The percentage yield of the 1-5 fragment was greatly reduced by both inhibitors, but the percentage yields of the 2-3 and 1-8 fragments were moderately increased. It was concluded that (1) the intact rat lung itself has a very large capacity to degrade BK in the range of 2 mumoles/min/kg body weight; (2) two major and several minor enzyme pathways exist to degrade BK; (3) the relative contributions of these pathways to overall BK degradation remain essentially constant over a bradykinin concentration range from 0.0015 to 17 microM; (4) ACE/kininase-II catalyzed hydrolysis is one of the major pathways but is not the single major route for BK degradation; and (5) the other major BK degradation pathway involves enzymes cleaving the Arg1-Pro2 and Pro3-Gly4 bonds of BK.

Management of massive hemorrhage associated with abdominal trauma.

Control of massive hemorrhage from intra-abdominal organs and major vascular structures may tax the ingenuity of the trauma surgeon. It is emphasized, however, that total blood loss and the amount of transfused blood are far less critical than the duration and severity of shock.

Biocompatibility of hemoglobin solutions. II. The inflammatory reaction of human monocytes and mouse peritoneal macrophages.

This study explored the inflammatory mechanism of toxicity of hemoglobin solutions (Hb-S). Human monocytes and mouse activated peritoneal macrophages were incubated with seven different solutions. The first four consisted of non-cross-linked bovine Hb. Of these, Hb-SI was incompletely purified of stromal phospholipids, Hb-SII was contaminated with environmental bacterial endotoxins, Hb-SIII was pure hemoglobin, and Hb-SIV was pure Hb with the addition of superoxide dismutase (SOD), catalase (CAT), and mannitol (M). The other three solutions were made of pure bovine Hb cross-linked with different agents: Hb-SV, reacted with glutaraldehyde; Hb-SVI reacted with bis-3,5-dibromosalicyl fumarate (DBSF); and Hb-SVII reacted with a ring-opened dialdehyde derivative of 5'(pyro)-phosphate of adenosine (ATP) (o-ATP). The reaction of monocytes and macrophages was studied in terms of (a) O2-derived radicals, as determined by the measurement of H2O2 and lipid peroxides; (b) complement factor C3a desArg; (c) 6-keto-prostaglandin F1 alpha (stable metabolite of prostacyclin); and (d) TxB2 (stable metabolite of thromboxane) released into the culture supernatants. The most significant reactions were obtained with the solutions contaminated with stromal phospholipids or bacterial endotoxins. Pure Hb was less reactive. Further reduction in proinflammatory activity was achieved by the addition of oxygen radical-scavengers (SOD, CAT, and M), or by the cross-linking of Hb with DBSF or o-ATP.

Nephrotoxicity of hemoglobin solutions.

Nephrotoxicity is a problem of hemoglobin solutions (HbS) that still awaits full elucidation and correction. Therefore, a study was conducted using five HbS with different characteristics to replace 1/3 of blood volume in five groups of rabbits. All HbS contained bovine Hb, 6.5 g/dl, dissolved into a balanced electrolyte solution. HbS-I was Hb incompletely purified of stromal phospholipids and environmental bacterial endotoxins, and uncrosslinked; HbS-II was pure Hb non crosslinked; HbS-III was completely purified and crosslinked; HbS-IV was like HbS-III, but with pH 8.4; and HbS-V was like HbS-III, with the addition of mannitol. The effects of blood replacement with these solutions were studied on: (a) PAH clearance (expression of renal plasma flow); (b) endogenous creatinine clearance (expression of glomerular filtration); (c) fractional excretion of sodium and (d) urine/plasma osmolarity (expressions of tubular function). Histological changes were assessed after 24 hours. Significant alterations were observed in decrescent order following the administration of HbS-I, -II and -III, while HbS-IV and -V were well tolerated. These results suggest that the nephrotoxicity of Hb solutions can be prevented by the following steps: (1) complete purification of Hb; (2) complete crosslinking; and (3) protection of the kidney by alkalinization of the urine and/or the addition of mannitol.

Generation of free oxygen radicals and the toxicity of hemoglobin solutions.

The observed toxicity of hemoglobin solutions (HbS) might depend, at least in part, on the tendency of Hb to autoxidation with generation of oxygen free-radicals. Aims of this study were: (1) to quantitate plasma elevations of H2O2 and lipid peroxides after replacement of 1/3 of calculated blood volume in various groups of rabbits with different Hb solutions; (2) to correlate these elevations with parameters of brain, heart, lung, liver and kidney injury or dysfunction; and (3) investigate the protective effect of mannitol as a radical scavenger. One Hb solution contaminated with stromal phospholipids raised H2O2 from 31.2 +/- 1.9 to 166 +/- 20 mumol/ml, lipid peroxides from 1.62 +/- 0.5 to 7.29 +/- 0.3 nmol/ml, CK-BB (brain isoenzyme) from 250 +/- 25 to 470 +/- 50 IU/L, CK-MB (myocardial isoenzyme) from 2.98 +/- 0.03 to 10.73 +/- 1.3 IU/L and SGPT from 38.1 +/- 5 to 167 +/- 45 IU/L, and reduced PaO2 from 87 +/- 10 to 57.5 +/- 2.5 mm Hg and creatinine clearance from 1.5 +/- 0.3 to 0.13 +/- 0.03 mg/min/Kg. These changes were progressively less severe with pure unmodified Hb, pure Hb crosslinked with "o-ATP", and pure crosslinked Hb + mannitol (4 mg/ml). These observations indicate a significant role for oxygen-derived radicals in the toxicity of Hb solutions.

Toxic factors in the red blood cell membrane.

The toxic effects of hemolysed RBCs have been studied for more than 100 years, but the specific factors involved have not been identified. This study focused on phosphatidylethanolamine (PE) and phosphatidylserine (PS), two aminophospholipids that normally reside on the cytoplasmic side of the red cell membrane. An in vitro experiment with murine peritoneal exudate macrophages showed that PE and PS: a) stimulated the production of H2O2, complement factor C3a, prostacyclin, and thromboxane at a dose of 5 micrograms/ml; b) produced cell injury, evidenced by release of lipid peroxides, LDH, and by morphologic changes on phase-contrast and electron microscopy at a dose of 50 micrograms/ml; and c) caused cell death in 50-66% of cells at a dose of 100 micrograms/ml. An in vivo experiment showed that PE and PS injected intravenously into various groups of rabbits: a) caused only transient hypotension at a dose of 0.05 mg/kg body weight; b) caused significant hypotension, cardiac arrhythmias, bronchospasm, activation of intravascular coagulation, complement, platelets, and leukocytes with release of histamine, serotonin, and thromboxane at a dose of 0.10 mg/kg; and c) caused cardiac arrest and death at a dose of 0.30 mg/kg. In contrast, the phospholipids of the outer cell membrane (phosphatidylcholine and phosphatidylinositol) caused minimal toxicity in vitro and none in vivo.

Biocompatibility of hemoglobin solutions. I. Reactions of vascular endothelial cells to pure and impure hemoglobins.

Hemoglobin (Hb) solutions can cause vasoconstriction and activation of intravascular coagulation. Because the endothelium plays a major role in the regulation of vascular tone and hemostasis, a study was conducted of human umbilical vein endothelial cells (EC) incubated with various Hbs. Cell injury was evaluated by electron microscopy and the release of lactic dehydrogenase, H2O2, and procoagulant "tissue factor." Cell reaction was assessed by the measurement of 6-keto-prostaglandin F (PGF)1 alpha (metabolite of prostacyclin) and thromboxane B2 (metabolite of thromboxane A2). Incubation with unmodified bovine hemoglobin for 24 h caused no cell injury and a reaction characterized by 48.4 +/- 8.2% increase in 6-keto-PGF1 alpha production, accompanied by 40.2 +/- 9.4% reduction in thromboxane (Tx)B2 (compared with a control group of EC incubated with saline solution). Incubation with a nonpure Hb solution (Hb plus red blood cell membrane aminophospholipids; a-PLs) caused cell injury with significant release of tissue factor, plus a reaction characterized by 97.5 +/- 12.5% increase in TxB2 production accompanied by 25.3 +/- 3% reduction in 6-keto-PGF1 alpha. A second nonpure Hb [Hb plus bacterial environmental endotoxin (E)] caused cell injury, the release of tissue factor, and increased production of both prostaglandins, with greater release of TxB2 (197 +/- 17%) than of 6-keto-PGF1 alpha (112 +/- 8.3%). These data indicate that the endothelium reacts differently to pure and nonpure hemoglobins. The biocompatibility of Hb solutions, with regard to vasoconstriction and activation of intravascular coagulation, depends on the absence of stromal a-PLs and bacterial E.

Human lung membrane-bound neutral metallo-endopeptidase-catalyzed hydrolysis of bradykinin.

Human lung membrane-bound neutral metallo-endopeptidase (NME; EC 3.4.24.11) has been purified; this enzyme occurred in two forms, NME-I and NME-II. The total NME activity was purified 2,143-fold with the final specific activities for NME-I and NME-II being 750 and 1,124, respectively. The two NME forms were resolved in the final purification step involving ion exchange; in all earlier steps including gel filtration and affinity chromatography (phenyl sepharose) both forms behaved similarly and eluted simultaneously. NME-I and NME-II both had a Mr value of 97,000, and neither form dissociated into subunits. Catalytic actions of NME-I and NME-II upon bradykinin were identical; the Gly4-Phe5 and Pro7-Phe8 bonds of bradykinin were cleaved with the final hydrolytic products for each enzyme being the tetrapeptide, Arg-Pro-Pro-Gly, the tripeptide, Phe-Ser-Pro, and the dipeptide, Phe-Arg. The intermediate products were the heptapeptide, Arg-Pro-Pro-Gly-Phe-Ser-Pro, and the pentapeptide, Phe-Ser-Pro-Phe-Arg. Neither NME-I nor NME-II were inhibited by the angiotensin-converting enzyme inhibitor, captopril. Both enzymes were inhibited by phosphoramidon, dithiothreitol and EDTA. Other peptidase inhibitors and heavy metals were not effective NME inhibitors. Both NME-I and NME-II cleaved angiotensin-I at the Pro7-Phe8 bond, and substance-P at the Glu6-Phe7 bond, with the latter being much slower than the former.

Complement activation and the toxicity of stroma-free hemoglobin solutions in primates.

The toxicity of hemoglobin solutions was studied in the context of their ability to activate serum complement (C). Three bovine polymerized hemoglobin solutions (BPHSs) with different degrees of purity were used for experiments in vitro and in vivo. BPHS-1 contained bacterial endotoxins (E) (5 EU/ml) and stromal phospholipids (PLs) (1.2 mg/dl), BPHS-2 contained only PLs (2.0 mg/dl), while BPHS-3 was completely free of both contaminants. C-activation was studied by the direct measurement of C3a, C4a, and C5a des Arg fragments, using commercially available RIA kits. During 1 hour of incubation with fresh monkey plasma, BPHS-1 and -2 activated both pathways of C, while BPHS-3 caused no activation of any factor. In vivo, Hb solutions were used to replace one-third of blood volume in three groups of six Coebus monkeys each, while fresh homologous plasma was used in a control group of four animals. Impure solutions activated the alternative pathway of C and caused significant reactions of the circulating blood (thrombocytopenia, leukopenia, and disseminated intravascular coagulation) associated with multiorgan dysfunction (cardiac arrhythmias, hypoxemia, reduction of renal clearance of endogenous creatinine, and elevation of liver enzyme SGPT). The pure solution neither activated C nor caused any reaction in the circulating blood. However, it caused a moderate degree of direct tissue injury, evidenced by transient reduction of creatinine clearance and elevation of SGPT. These observations suggest that impure and pure Hb solutions carry separate mechanisms of toxicity. Complement, activated by toxic impurities, plays an active role in the toxicity of impure solutions. C-activation in vitro could be used as a screening test of biocompatibility.

Toxicity of polymerized hemoglobin solutions.

Four solutions of bovine polymerized hemoglobin (BPHS) and rabbit plasma were used to replace one-third of the blood volume in five groups of rabbits. The first three solutions were "impure" because of the presence of stromal phosphatidyl-ethanolamine and phosphatidyl-serine in BPHS-1, environmental endotoxins in BPHS-2, and a large amount of higher molecular weight hemoglobin-glutaraldehyde polymers in BPHS-3. These solutions caused a 33 per cent mortality rate and significant morbidity which was characterized by hemodynamic instability, respiratory and renal insufficiency, elevation of hepatic enzyme levels, thrombocytopenia, leukopenia, disseminated intravascular coagulation (DIC) and activation of the alternate pathway of complement. Histopathologic changes found in the heart, lungs, liver, spleen and kidney were characterized by a combination of ischemic and inflammatory lesions. Fibrin thrombi were visible by immunofluorescence in the microcirculation. In contrast, the fourth solution (BPHS-4) was free of the aforementioned impurities; caused no deaths and minimal morbidity, which was limited to elevated levels of serum glutamic pyruvic transaminase and reduction of creatinine clearance; no DIC or complement activation, and mild histopathologic changes which were exclusively ischemic in nature. The results of this study indicated that the toxicity of polymerized hemoglobin solutions is due principally to the presence of impurities. Pure hemoglobin does exhibit mild toxicity when compared with a control solution which is most likely due to a vasoconstrictor effect of oxyhemoglobin.

A multicatalytic high-molecular-weight neutral endopeptidase from human kidney.

A multicatalytic endopeptidase (ME) with three distinct activities, chymotrypsin-like, cucumisin-like, and trypsin-like, occurred in all rat tissues examined with highest activities in kidney, testes, liver, and spleen; they were assayed with benzyloxycarbonyl-Gly-Gly-Leu-p-nitroanilide (Z-Gly-Gly-Leu-pNA), benzyloxycarbonyl-Leu-Leu-Glu-2-naphthylamide (Z-Leu-Leu-Glu-2NA), and benzyloxycarbonyl-Gly-Gly-Arg-2-naphthylamide (Z-Gly-Gly-Arg-2NA), respectively. All three activities were recovered from a single protein band on a polyacrylamide gel after electrophoresis of purified human kidney ME. The native enzyme had a Mr of 650,000, and it consisted of about 5,135 amino acid residues. After denaturation and electrophoresis on sodium dodecyl sulfate (SDS)-polyacrylamide gels kidney ME dissociated into several low Mr components ranging from 23,000 to 33,000. Kidney ME had a pH optimum of 7.6-8.1 with Z-Gly-Gly-Leu-pNA, 7.3 with Z-Leu-Leu-Glu-2NA, and 9.8 with Z-Gly-Gly-Arg-2NA. SDS enhanced chymotrypsin- and cucumisin-like activities by two to three times whereas trypsin-like activity was not enhanced. The specificity constant (kappa cat/Km) of human kidney ME for Z-Gly-Gly-Leu-pNA was 6.7 X 10(3) M-1 S-1; Z-Gly-Gly-Leu-2NA was not hydrolyzed. The specificity constant for Z-Leu-Leu-Glu-2NA was similar to, and for Z-Gly-Gly-Arg-2NA was one half of that for Z-Gly-Gly-Leu-pNA. ME cleaves only the Phe5-Ser6 bond of bradykinin (BK); however, all three ME activities were inhibited by BK. Strong inhibition of ME by albumin suggests that ME is involved in cleavage of larger polypeptides. Antipain and leupeptin almost completely inactivated the trypsin-like activity whereas they had no significant effect on the other two activities. ME is not a metal-loenzyme nor is the serine residue essential for its activities; however, thiol groups are involved. Na+ and K+ inhibited all ME activities. Trypsin-like activity was more sensitive to divalent cations than the other two.

A high-molecular-mass neutral endopeptidase-24.5 from human lung.

A high-Mr neutral endopeptidase-24.5 (NE) that cleaved bradykinin at the Phe5-Ser6 bond was purified to apparent homogeneity from human lung by (NH4)2SO4 fractionation, ion-exchange chromatography and gel filtration. The final enzyme preparation produced a single enzymically active protein band after electrophoresis on a 5% polyacrylamide gel. Human lung NE had an Mr of 650,000 under non-denaturing conditions, but after denaturation and electrophoresis on an SDS/polyacrylamide gel NE dissociated into several lower-Mr components (Mr 21,000-32,000) and into two minor components (Mr approx. 66,000). The enzyme activity was routinely assayed with the artificial substrate Z-Gly-Gly-Leu-Nan (where Z- and -Nan represent benzyloxycarbonyl- and p-nitroanilide respectively). NE activity was enhanced slightly by reducing agents, greatly diminished by thiol-group inhibitors and unchanged by serine-proteinase inhibitors. Human lung NE was inhibited by the univalent cations Na+ and K+. No metal ions were essential for activity, but the heavy-metal ions Cu2+, Hg2+ and Zn2+ were potent inhibitors. With the substrate Z-Gly-Gly-Leu-Nan a broad pH optimum from pH 7.0 to pH 7.6 was observed, and a Michaelis constant value of 1.0 mM was obtained. When Z-Gly-Gly-Leu-Nap (where -Nap represents 2-naphthylamide) was substituted for the above substrate, no NE-catalysed hydrolysis occurred, but Z-Leu-Leu-Glu-Nap was readily hydrolysed by NE. In addition, NE hydrolysed Z-Gly-Gly-Arg-Nap rapidly, but at pH 9.8 rather than in the neutral range. Although human lung NE was stimulated by SDS, the extent of stimulation was not appreciable as compared with the extent of SDS stimulation of NE from other sources.

One hundred and fourteen fistulas of the gastrointestinal tract treated with total parenteral nutrition.

During a period of five years, all patients with fistulas of the gastrointestinal tract who received total parenteral nutrition as a part of the therapy were evaluated retrospectively. One hundred and eight patients with 114 fistulas of the gastrointestinal tract were studied. There were 58 male and 50 female patients. The mean age of the patients was 58 years old. Fifty-one per cent (58) of all fistulas resulted from surgical complications and 30 per cent from inflammatory disease. The most common origin of the fistula was the small intestine (48 per cent), with the large intestine being the next most common origin (26 per cent). Ninety-eight of the fistulas were treated successfully. The mean time for closure from the time of diagnosis was 30.9 days. Sixty-one per cent (69) of all fistulas closed spontaneously, the use of parenteral nutrition resolved 37 per cent (42) and 24 per cent (27) resolved after surgical control of sepsis. Twenty-five per cent (29) of the fistulas required definitive surgical closure. Sixteen patients died prior to resolution of the fistula, 11 deaths were directly related to septic complications of the fistula and one died as a result of hemorrhage of the fistula tract. The direct fistula-related mortality rate was 10.5 per cent. Whereas the mortality rate for fistulas remained stable, spontaneous closure rates continue to improve. This is attributed to improved "para-surgical" care, appropriate nutritional support and early and aggressive control of sepsis. Ninety to 95 per cent of fistulas that spontaneously resolve will do so within four to five weeks. Inflammatory disease of the intestine and radiation induced fistulas continue to respond poorly to medical management.

Gastrointestinal malignancies in infancy, childhood, and adolescence.

Gastrointestinal and hepatobiliary malignancies in the pediatric age group are rare tumors and are usually treated with frustratingly poor results because of late diagnosis. Current efforts to improve treatment are directed at earlier diagnosis of these tumors in combination with aggressive surgery and adjuvant therapy.