Biol. Chem. Hoppe-Seyler Vol. 373, pp. 691-698, August 1992
The Enzymatic and Release Characteristics of Sheep Neutrophil Elastase: A Comparison with Human Neutrophil Elastase Wolfgang Georg JUNGER*, Seth HALLSTROM b , Forrest Chen Liua, Heinz REDL b , Günther SCHLAG*5 a h
University of California San Diego Medical Center, Department of Surgery, San Diego, California, USA Ludwig-Boltzmann-Institut für Experimentelle und KlinischeTraumatologie, Wien, Österreich
(Received 8 January / 28 April 1992)
Summary: Sheep are often used to study tissue damage following shock after traumatic injury and in the course of other diseases. The processes involved are thought to be caused at least in part by elastase released from polymorphonuclear leukocytes (PMNs). Since little is known about elastase and its role as a mediator of tissue damage in sheep, we studied the biochemical properties and release characteristics to sheep leukocyte elastase (SLE) in comparison of those of human leukocyte elastase (HLE). Both enzymes showed similar molecular masses, amino-acid compositions, N-terminal amino-acid sequences, and abilities to digest elastin substrates. Differences, however, were found in kinetic parameters measured with the elastase-specific substrate N-methoxysuccinyl(L-alanyl)2-L-prolyl-L-valine-4-nitroanilide (MeoSucAAPV-pNa). The Michaelis constant (Km) of ovine elastase was nearly 10 times higher (1.82mM) than the
Km of HLE (0.21mM). Values of SLE calculated for kcal were 70% and for kcat/Km 8% of corresponding values determined for HLE. In addition, significant differences between sheep and human PMNs were found in in vitro stimulation experiments. In contrast to human PMNs, sheep neutrophils released no active elastase, and only 50 to 70% of the H2O2 produced by human PMNs. This failure to release active elastase could not be explained by a lower elastase content of sheep PMNs, as there were no significant differences found between the elastase contents of sheep and human PMNs. We conclude that elastase liberated by stimulated sheep PMNs is inactivated by a concomitantly released proteinase inhibitor also located within the sheep PMNs. These results together with our earlier findings indicate that sheep elastase, in contrast to human elastase, may play a minor role in mediating tissue damage and organ failure.
Neutrophile Elastase vom Schaf: Ein Vergleich mit humaner neutrophiler Elastase bezüglich Enzymcharakteristik und Freisetzung Zusammenfassung: Das Schaf dient oft als Modell zur Untersuchung von Gewebsschädigung im Zuge von traumatischem Schock und anderen Krankheitsbildern. Elastase ist eine von polymorphkernigen neutrophilen Granulozyten (PMN) freigesetzte Proteinase und wird als ein maßgeblicher Mediator die-
ser gewebsschädigenden Prozesse erachtet. Da wenig über Schafleukozytenelastase (SLE) und ihren Beitrag zur Gewebsschädigung im Schaf bekannt ist, untersuchten wir die biochemischen Eigenschaften und den Freisetzungsmechanismus dieses Enzyms und verglichen diese Daten mit denen der humanen
Enzymes: Human leukocyte elastase (lysosomal elastase from human leukocytes) EC 3.4.21.37; Sheep leukocyte elastase (lysosomal elastase from sheep leukocytes) EC 3.4.21.-. Abbreviations: BBSS, Hank's balanced salt solution; HLE, human leukocyte elastase; MeoSuc-AAPV-pNa, N-methoxysuccinyl-(L-alanyl)2-L-prolyl-L-valyl-4-nitroanilide; PMNs, polymorphonuclear leukocytes; PMA, phorbol-12-myristate-13-acetate; PITC, phenylisothiocyanate; PTH, phenylthiohydantoin; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; SLE, sheep leukocyte elastase.
Brought to you by | University of Arizona
Copyright © by Walter de Gruyter & Co · Berlin · New York Authenticated
Download Date | 5/26/15 10:38 PM
692
W.G. Junger, S. Hallström, F.C. Liu, H. Redl and G. Schlag
leukozytären Elastase (HLE). Beide Enzyme zeigten ähnliche Molmassen, Aminosäurenzusammensetzungen, W-terminale Aminosäurensequenzen und Fähigkeiten, Elastin zu spalten. Wir fanden jedoch Unterschiede der enzymkinetischen Parameter mit dem oft verwendeten elastasespezifischen Substrat vY-methoxysuccinyl-(L-alanyl) 2 -L-prolyl-L-valine-4-nitroanilid (MeoSuc-AAPV-pNa). Die Michaelis Konstante C^m) von Schafelastase war nahezu lOfach höher (1.82mM) als der A:m-Wert von HLE (0.21mM). Die Werte von SLE für £cat betrugen nur 70% und für kcj Km lediglich 8% der für HLE ermittelten Daten. Abgesehen davon konnten signifikante Unterschiede in der Elastasefreisetzung von in vitro stimulierten Schafgranulozyten und humanen Granulozyten beobachtet werden. Im Gegensatz zu humanen PMN setzten Schafgranulozyten keine aktive Elastase frei.
Vol. 373 (1992)
Außerdem produzierten Schaf-PMN nur 50 bis 70% der von humanen Zellen freigesetzten Menge an H2O2. Die Tatsache, daß Schaf-PMN keine aktive Elastase freisetzten, konnte nicht auf einen niedrigeren Elastasegehalt in Schafgranulozyten zurückgeführt werden, da sich Schaf- und humane Granulozyten nur unbedeutend in ihrem Elastasegehalt unterschieden. Wir schließen aus diesen Untersuchungen, daß aus stimulierten Schafgranulozyten freigesetzte Elastase durch einen gleichzeitig aus diesen Zellen liberierten Proteinaseinhibitor inaktiviert wird. Gemeinsam mit unseren früheren Untersuchungen weisen diese Ergebnisse darauf hin, daß Schafelastase im Gegensatz zur humanen Elastase eine untergeordnete Rolle als Mediator von Gewebsund Organschädigung spielt.
Key terms: human and sheep neutrophil elastase; biochemical characterization; PMN activation; in vitro elastase release.
The serine proteinase, elastase, is one of the most abundant lysosomal proteinases contained in polymorphonuclear leukocytes (PMNs). Excessive release of elastase after PMN stimulation is thought to contribute substantially to processes resulting in proteolytic degradation of plasma proteins and structural elements of the microvascular system in the course of several diseases'1'2'. PMN margination and activation is a common observation after severe injury leading to traumatic shock. Irreversible multi-organ failure syndrome (MOFS) and adult respiratory distress syndrome (ARDS) are thought to be primarily caused by the proteolytic action of elastase, especially as plasma proteinase inhibitor activity is reduced by concomitantly produced oxidants'1'41. Several animal models using sheep have been established to evaluate the correlation of microvascular tissue damage in the lung with PMN margination and activation in various diseases'5'101. Sheep are widely used as animal models of acute and chronic lung injury, primarily because measurements of vascular permeability changes are available by a lung lymph node preparation technique introduced by Staub and co-workers'111. Despite the common use of sheep in experiments designed to assess the role of lysosomal proteinases in tissue destruction and to explore possible pharmacologic approaches to prevent proteolytic tissue injury with specific proteinase inhibitors, little is known about the properties of neutrophil elastase in sheep and its resemblance with the human enzyme. As the sheep is used extensively in our group as a model of septic shock, we became interested in ovine elastase
and its possible role as a mediator of tissue damage in comparison with the human enzyme. In a previous paper we described the isolation and a preliminary biochemical characterization of an elastase-like neutral proteinase from sheep PMNs as well as of a serine proteinase inhibitor located in the cytosol of sheep PMNs'12l In order to complete the characterization of sheep elastase and to evaluate its potential relevance in provoking tissue damage, we gathered additional data on the biochemical properties of this enzyme and the mechanism of its in vitro release by sheep PMNs. In addition we compared these biochemical and release characteristics of sheep elastase to human elastase, as we believe that these data might be of interest to experimenters using sheep as an animal model in studying the role of neutrophil elastase in certain diseases.
Materials and Methods Elastases Sheep elastase and the cytosolic serine proteinase inhibitor we isolated from ovine neutrophils as previously described1121. Purifii
Characterization of sheep elastase Amino acid composition The amino-acid composition was determined according to the method of Bergman et al.l'3]. Purified elastase was hydrolysed with amino-acid-free OM HC1 (Pierce Chemical Company, Rockford, IL,
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
Vol. 373 (1992)
Sheep and Human Elastase: Enzyme and Release Characteristics
USA) at 110 °C for 24 h.The hydrolysis products were derivatized with phenylisothiocyanate (PITC; Pierce), lyophilized, and dissolved in HPLC starting buffer. A gradient profile was applied using 0.03M NaOH, adjusted to pH 6.6 with IM H3PO4, and 60% acetonitrile in water (Merck, Darmstadt, Germany) at a flow rate of lm//min. The Cjg-Spherisorb S3 ODS2 column (3-μ,ηι spherical particles; Phase Separations, Queensferry, Clwyd, U.K.) with dimensions of 125 x 4.6 mm was thermostated at 37 °C in a Precitherm PFVwaterbath (Boehringer Mannheim GmbH, Mannheim, Germany); detection wavelength was 254 nm.The HPLC system consisted of a Wisp 710 Β automated sample applicator, two Model 510 pumps, a Data Modul Integrator, a Waters Automated Gradient Controller, and a Model 481 LC Spectrophotometer LambdaMax (Waters Associates, Milford, MA, USA). N-Terminal amino-acid sequence Elastase was subjected to SDS-PAGE gel electrophoresis as described'121, and electroblotted to a PVDFmembrane (Imrnobilon-P transfer membrane; Millipore Corporation, Bedford, MA, USA) according to the method described by Matsudaira'14'. The membrane was stained with Coomassie Brilliant Blue R-250 and stored at - 70 °C for amino-acid sequence analysis.Three bands representing the isoforms (see Fig. 1) were cut out of the membrane and applied to the reaction chamber of the sequencer. Automated Edman degradations were performed on an Applied Biosystems Model 470 A Gas Phase Sequencer employing the O3RPTH program as supplied by the manufacturer[15] (Applied Biosystems, Foster City, CA, USA). Phenylthiohydantoin (PTH) amino acid derivatives were identified using an Applied Biosystems Model 120 on-line HPLC system. Data reduction was accomplished using a Perkin-Elmer Model 7500 computer with Chrom 3 software (Perkin-Elmer, Norwalk, CT, USA). Enzymatic activity Enzymatic activity was measured in an assay buffer consisting of 50mMTris/HCl with'lOOmM NaCl, and 0.05% Triton X-100 (Alcylphenyl polyethylenglycole; Fluka AG, Buchs, Switzerland), pH 7.4. The synthetic substrate N-methoxysuccinyl-(L-alanyl)2-L-prolyl-L-valine 4-nitroanilide (MeoSuc-AAPV-pNa) (Protagen AG, L ufelfingen, Switzerland) was dissolved in DMSO (Dimethylsulfoxide; Fluka AG) and used at a final assay concentration of ImM. Buffer containing MeoSuc-AAPV-pNa (900 μ/) was mixed with 100 μ/ of sample and measured for 5 min at 405 nm and 37 °C.The molar extinction coefficient of the synthetic substrate was 10200M"1 cm'1^. Enzymatic activity was expressed in international units (1U = 16.67 nkat, the activity necessary to convert 1 μ,ητιοί MeoSuc-AAPV-pNa/min). For all kinetic experiments a Beckman DU-70 Spectrophotometer (Beckman Instruments, Palo Alto, CA, USA) was used. Michaelis constant Km The Michaelis constants Km (with MeoSuc-AAPV-pNa) of SLE and HLE were determined according to Lineweaver and Burk'17'. MeoSuc-AAPV-pNa was added to yield final assay concentrations ranging from 0.28 to ImM for HLE, and 0.3 to 2mM for SLE. The final assay concentration of DMSO was kept constant at 1% (v/v) by preparing substrate dilutions in DMSO. All measurements were performed in triplicate. The Km values were determined from straight lines obtained by linear regression with correlation coefficients r > 0.99. Molar concentrations ofelastase solutions Molar concentrations of elastase solutions were determined by equivalent titration according to Braun et al.[181. Increasing amounts of elastase solutions of unknown concentrations were incubated with constant, known concentrations of the serine proteinase inhibitor Eglin c (courtesy of Dr. Schnebli, Ciba-Geigy,
693
Basel, Switzerland). Elastase and inhibitor solutions were mixed and incubated in the assay buffer for 10 min at 37 °C. After addition of MeoSuc-AAPV-pNa at a final assay concentration of ImM, the mixtures were incubated for 10 min, the reactions were stopped by adding 100 μΐ glacial acetic acid, and released 4-nitroaniline was measured at 405 nm. Residual enzymatic activities were plotted versus the volume of added elastase solution and the equivalent molar concentration determined graphically as described1181. Degradation of elastin substrates The capability of ovine and human neutrophil elastase to degrade elastin was compared. [3H]elastin (3.95 /u.Ci/mg) and human leukocyte elastase (HLE) were kindly provided by Dr. Schnebli (CibaGeigy, Basel, Switzerland). Orcein-elastin (bovine neck ligament) was purchased from Sigma. Equal amounts of elastin derivatives were washed twice in assay buffer (used in enzymatic measurements) to remove unbound tritium or orcein dye and suspended in the same buffer. Aliquots of the suspensions were mixed with lOOpmol of human or ovine neutrophil elastase. Normal saline was used as a blank. The mixtures were incubated for 2 h at 37 °C with constant agitation. After centrifugation, liberated orcein dye was measured in the supernatants at a wavelength of 578 nm. Liberation of tritium was measured after mixing 10 μΐ of the supernatant with 4.5 ml of Ready Value scintillation liquid (Beckman Instruments) with a BF5000/300 beta-counter (Berthold, Wildbad, Germany) for 4 min.
In vitro stimulation experiments PMN isolation Human and ovine PMNs were isolated from buffy coats of healthy individuals or animals, respectively, by counter flow centrifugation employing a Beckman J2-21M/E with a JE6B-rotor (Beckman Instruments) as previously described'12'. This method yielded PMN preparations with purities > 98%, as examined by differential cell counting, and cell viabilities > 95%, as determined by trypan-blue dye exclusion. Analysis of released elastase activity Equal concentrations of sheep or human PMNs (1.0 x 106 cells/m/) were stimulated by incubation with either 2 mg/m/ opsonized zymosan or Ο.Ιμ,Μ phorbol-12-myristate-13-acetate (PMA; Sigma Chemical Company, St. Louis, MO, USA) in Hanks' balanced salt solution (HBSS). Zymosan was opsonized with autologous plasma according to the method described by Lieners et al.'19' immediately before use. Opsonized zymosan was washed 5 times with HBSS to remove serum arproteinase inhibitor. The cell suspensions were incubated at 37 °C for 1 h, centrifuged, and analysed for released elastase activity as described above. To determine if activated sheep PMNs released an excess of their cytosolic proteinase inhibitor, 100 μΐ of supernatants were incubated with 10 μΐ purified SLE at 37 °C in the buffer used for kinetic measurements. After 10 min of pre-incubation, the enzymatic activity of SLE was determined as described above. Determination ofH2O2 HaOa production of activated PMNs was determined fluorometrically according to the method described by Ruch et al.[20]. In brief: 0.5 ml of a 0.4mM HVA homovanillic acid (HVA; Serva, Heidelberg, Germany) in HBSS, containing 4 U/m/ horseradish peroxidase (HRP; Boehringer Mannheim) was mixed with 0.1 m/of an appropriate dilution of stimulating agent. The mixtures were brought to a volume of 1.9 ml with HBSS and pre-incubated for 5 min at 37 °C.Then 100 μΐ of cell suspensions were added to obtain final cell concentrations of 1.0 x 106 cells/assay and incubated for an additional 30 min at 37 °C under steady shaking in a water bath. Liberation of H2O2 was terminated by addition of 250 μΐ of an ice-
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
694
W.G. Junger, S. Hallstr m, F.C. Liu, H. Redl and G. Schlag
Vol. 373 (1992)
cold solution containing O.lM glycine and 15mM EDTA, pH 12 (adjusted with NaOH).The assay mixtures were centrifuged and concentrations of the reaction product 2,2'-dihydroxy-3,3'-dimethoxydiphenyl-5,5'-diacetic acid in the supernatants measured with a F-4000 Fluorescence Spectrophotometer (Hitachi Ltd., Tokyo, Japan) at 25 °C using an excitation wavelength of 312 nm and an emission wavelength of 420 nm. Degradation of [3H]elastin by stimulated PMNs Sheep and human PMNs (1.5 x 106 cells/assay), respectively, were incubated with l mg [ 3 H]elastin in HBSS at 37 °C. Final assay volume was 150 μ,/. The cells were stimulated with PMAat a final concentration of Ο.Ιμ,Μ. Aliquots of the reaction mixture (50 μ,/) were taken at different time intervals and centrifuged for 2 min in a microcentrifuge. Released tritium was measured in 45 μΐ of the supernatants as described above. Elastase content of sheep PMNs Granule-rich preparations of known amounts of purified ovine PMNs were washed and extracted according to the method described'121. Total enzymatic activities of the extracts were determined as indicated above.The original elastase amount contained in the PMNs was estimated, based on the total extracted activity and the purified enzyme's specific enzymatic activity, which had been determined with 2841.5 U/g using the substrate MeoSucAAPV-pNa. Other methods Protein assays Protein concentrations were determined by the bicinchoninic acid protein assay (BCA) reagent (Pierce Chemical Company) with human serum albumin as standard, or with an Antek 771 Pyroreactor nitrogen analyser (Antek, Houston, TX, USA) using urea as standard. Association rate constant The association rate constant (kon) of the reaction of sheep elastase with the cytosolic inhibitor of sheep PMNs was determined according to the method described by Vincent and Lazdunski'21'.
Results Ovine elastase was isolated from purified sheep granulocytes resulting in a chromatographically homogeneous preparation as described earlieril2]. SLE in this preparation had a specific enzymatic activity of 2840 U/g protein. The total activity increased after the first Chromatographie purification step probably due to the removal of a proteinase inhibitor. SDS-PAGE electrophoresis of the purified SLE preparation yielded three protein bands with molecular masses of 22, 24, and 25 kDa, which were used for N-terminal amino-acid sequence analysis (see Fig. 1). The amino-acid composition determined for SLE was compared to data published for human neutrophil elastase as shown in Table l.The amino-acid distribution in HLE had been determined by analysis of hydrolysed protein' 22 ^, and by calculations based on sequencing data performed by Sinha et al.1231. SLE
Fig. 1. Sheep elastase immobilized on PVDF membrane for determination of the N-terminal amino-acid sequence. All three bands (arrows) representing the elastase isoforms were used for sequencing.
Table 1. Amino-acid composition of sheep neutrophil elastase in comparison to data for human leukocyte elastase published by Travis et al. [22] and Sinha et al.[23]. (The values shown are mol/100 mol amino acids). Data for SLE and HLE by Travis et al.'22' were obtained by hydrolysis (Asn, Gin, and Trp not determined), values for HLE by Sinha et al.'231 were derived from sequencing data. Amino acid
Asp + Asn Glu + Gin Ser Gly Thr His Ala Pro Arg Tyr Val Met Cys He Leu Phe Lys Trp
SLE
[mol-%]
HLE Travis et al. [mol-%]
HLE Sinha et al. [mol-%]
8.15 9.06 9.83 13.32 3.94 2.71 7.72 5.73 9.66 3.00 6.95 0.56 2.87 3.45 7.39 2.59 2.65 n.d.
10.62 7.96 5.75 12.39 3.10 1.77 10.62 4.42 9.73 0.88 11.06 0.88 2.65 4.87 8.85 3.98 0.44 n.d.
9.63 6.88 5.50 11.01 2.75 2.29 10.09 4.13 8.72 0.92 12.39 1.38 3.67 5.50 9.63 4.13 0.00 1.38
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
Vol. 373 (1992)
695
Sheep and Human Elastase: Enzyme and Release Characteristics
contains higher amounts of serine and lysine, and smaller quantities of alanine and valine than HLE. The lower valine content of SLE could be due to incomplete hydrolysis of valyl bonds during the 24-h hydrolysis period. Besides these exceptions, similar amino-acid distribution patterns of human and sheep elastase were seen. The first residues of the N-terminal amino-acid sequence of SLE were determined and compared to the sequence published for HLE by Sinha et alJ23l Table 2 displays the N-terminal sequences of SLE and HLE. With the exception of the amino acids at positions 6 and 15, the sequences of these two elastases are identical. Throughout the sequence analysis no evidence of microheterogeneity could be found , which indicates that the SLE isoforms may share the same peptide moiety and differ only in the carbohydrate residue. (All three bands were used for the determination. See Fig. l.)The same has been reported for Despite the similarities in the structure of sheep and human elastase, marked differences were found in their kinetic properties with the widely used elastasespecific substrate MeoSuc-AAPV-pNa. The Michaelis constant Km of sheep elastase with MeoSucAAPV-pNa was 1.82 x 10~3M, which is about 10 times higher than the Km value determined for human elastase (0.21 x 10"3M). The turnover number fccat was calculated with 16 s'1 for SLE and 23 s~l for HLE. The kcJKm values of SLE and HLE were determined as 8800 and 111900M"1 s"1, respectively. The capability of SLE to degrade 3H- and orceinlabeled elastin, however, was similar to that of HLE (Table 3). After 2 h of incubation at 37 °C, SLE had
Table 2. Comparison of N-terminal amino-acid sequence of SLE with the sequence of HLE determined by Sinha et al.'23'. Only positions 6 and 15 do not correspond. Position
SLE
HLE
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
He Val Gly Gly Arg Ala Ala Arg Pro His Ala Trp Pro Phe He
He Val Gly Gly Arg Arg Ala Arg Pro His Ala Trp Pro Phe Met
TableS. Equal molar concentrations (100 pmol) of SLE and HLE were mixed with [3H]elastin (4.48 mg) and orcein-elastin (2.20 mg) in 0.2 ml of buffer (50mM Tris/HCl, lOOmM NaCl and 0.05% Triton X-100,pH7.4). Liberated orcein dye and tritium were measured after 2 h incubation at 37 °C (n = 4 assays; values: mean ± SD). Leukocyte elastase
Orcein elastin dye liberation (A578)
KT5 x 3H-labeled elastin tritium liberation [cpm]
SLE HLE
0.299 ± 0.045 0.136 ±0.052
4.060 ± 0.062 3.426 ±0.076
liberated slightly more tritium than HLE and digested about two times more orcein-elastin than the same molar amount of HLE. Although these assays can only provide semi-quantitative data, they clearly indicate that SLE has a capability to degrade the natural substrate elastin, which is slightly more efficient than that of HLE. The in vitro stimulation experiments performed with sheep and human PMNs, however, showed a significant difference in elastase release. As summarized in Table 4, neither of the stimuli used to activate sheep PMNs resulted in a release of measurable elastase activity. Human PMNs liberated relatively low amounts of active elastase when stimulated with opsonized zymosan. Activation with PMA, however, caused the liberation of 23.4 ± 4.9 mU elastase per 106 cells. Sheep PMNs were found to produce about 50% less H2O2 than human cells when stimulated with PMA, and 70% less with zymosan activation.
Table 4. Liberation of elastase and H2O2 and by sheep and human PMNs in response to activation with 2 mg/m/ opsonized zymosan (zymo) or Ο.Ιμ,Μ phorbol-12-myristat-13-acetate (PMA). Values given are mean ± SD of 6 determinations (n = 6). Liberation product
Stimuli
Sheep PMNs Human PMNs
Elastase [mU/106PMN]
zymo PMA
0 0
1.5 ± 0.6 23.4 ± 4.9
H202 [nmol/106PMN]
zymo PMA
3.1±0.8 20.0 + 4.2
10.6 ± 6.8 42.6 ±18.2
We examined the elastase content of sheep PMNs in order to determine if a lower elastase content in the ovine neutrophil may be partly responsible for the fact that activation of sheep PMNs does not lead to liberation of measurable elastase activity. The content of elastase in sheep granulocytes was estimated to be Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
696
Vol. 373 (1992)
W.G. Junger, S. Hallstr m, EC. Liu, H. Redl and G. Schlag
1.3 ± 0.7 pg elastase per PMN, based on activity measurements. This value may be up to twice as high if losses of enzymatic activity during the isolation procedure and the contamination of the extract with the cytosolic inhibitor in sheep PMNs are taken into account. The anticipated elastase content of sheep PMNs, however, would still be well within the range of 1-4 pg/PMN described for human neutroThe absence of active elastase in the supernatants of in vitro stimulated sheep PMNs indicates that SLE may be released as a complex with the cytosolic proteinase inhibitor in sheep PMNs. To test if the cytosolic proteinase inhibitor is released by stimulated sheep PMNs into the supernatants, sheep PMNs at a concentration of l x 106 cells/m/ were stimulated with 0.2μΜ PMA for l h at 37 °C. The supernatants were incubated with purified SLE and the residual activity was measured. The addition of purified SLE generated elastase activity, suggesting that the cytosolic inhibitor may be destroyed by the added SLE, resulting in liberation of active elastase from the elastase-inhibitor complex. Based on the increase in total activity in the assay mixture, we estimated the elastase activity which would be released by stimulated sheep PMNs if the enzyme were not neutralized by the cytosolic inhibitor with 32.4 mU/106 PMNs (Table 5).
1
2 3 4 Incubation time [h]
5
6
Fig. 2. Degradation of [3H]elastin by in vitro stimulated human and sheep PMNs. [3H]elastin was incubated with human and sheep PMNs, respectively. The cells were stimulated as described in the text. Sheep cells were capable of degrading this substrate, but at a significantly lower rate than human PMNs.
Discussion
The biochemical data on sheep neutrophil elastase presented here and in a previous paper^ demonstrate clearly its similarity with the human enzyme. Besides the molecular masses, and the isoelectric points of the three isoforms, very similar amino-acid compositions (Table 1) and N-terminal amino-acid sequences (Table 2) of human and ovine elastase were found. Furthermore we could show that ovine and human neutrophils do not differ significantly in their Although no active elastase was found in the superelastase contents. Ohlsson[26] found 3-4 pg/PMN, natants of in vitro stimulated sheep PMNs, [3H]elasSinha et al.[24] reported 1.0 pg/cell, and Campbell[25] tin was degraded when directly incubated with stimucalculated 1.5 pg/cell for the elastase amount conlated sheep PMNs. This suggests that SLE can be tained in human neutrophils. With an estimated 1.3 ± partly dissociated from the complex formed with the 0.7 pg elastase per cell, the elastase content in sheep cytosolic proteinase inhibitor when a high affinity PMNs lies well within the range found for human substrate such as elastin is present. Degradation of shown to contain high [3H]elastin by in vitro stimulated sheep PMNs, how- neutrophils. Sheep PMNs were [12] levels of an elastase inhibitor ; if loss of activity ever, was significantly less effective than that by during enzyme isolation is taken into account (see rehuman PMNs (Fig. 2). sults) the estimated value for the elastase content could be even higher.
fable 5. Elastase activity in supernatants of stimulated sheep PMNs was generated by incubation of the supernatants with purified SLE. The elastolytic activity of the supernatant increased with prolonged incubation with purified SLE. The elastase activity released by 106 PMN was calculated based on the activity liberated by incubation of PMN supernatants with purified SLE as described in the text. Incubation time [min]
Theoretically released elastase activity [mU/106PMN]
0 5 10
0.0 20.5 32.4
Differences, however, were observed in the kinetic properties of SLE and HLE with the chromogenic elastase substrate MeoSuc-AAPV-pNa. The Michaelis constant Km of SLE was nearly 10 times higher than that of HLE, indicating a lower affinity of SLE for this synthetic substrate. The turnover number of SLE was 30% lower than that of HLE, and the kcJKm value of HLE was nearly 13 times higher than that of SLE. The ability of SLE to degrade elastin, one of its natural substrates which are destroyed in the process of tissue damage, however, was even somewhat better than that of HLE (Table 3). Besides the differences in the interaction with the synthetic elastase substrate, more important differences
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
Vol. 373 (1992)
Sheep and Human Elastase: Enzyme and Release Characteristics
were found in the elastase release mechanisms of human and sheep PMNs (Table 4). Unlike human neutrophils, sheep PMNs did not liberate any active elastase when stimulated with either PMA or opsonized zymosan. This cannot be ascribed to differences in the elastase amounts present in human and sheep PMNs, as discussed above. Although H2O2 production by sheep PMNs was about 50% lower than that of human neutrophils, a higher threshold of ovine versus human cells in response to the stimuli used is not likely to be solely responsible for the significant differences seen in elastase liberation. In an earlier publication we showed that sheep PMNs contain considerable amounts of an effective serine proteinase inhibitor in their cytosol[12]. The association rate constant kon of the inhibitor with SLE was determined as 2.3 x K^ivrV1, which is close to the value published for the analogue inhibitor/elastase pair in porcine neutrophils[27]. The presence of this inhibitor together with the low solubility, as well as the strong tendency of SLE to adhere to surfaces at low ionic strength may also have led to an underestimation of the elastase content in sheep PMNs by others[24]. As described earlier, we used a IM NaCl solution to extract elastolytic activity after separation of lysosomal granule preparations from the PMN cytosolic fraction, precautions necessary to avoid excessive loss of active enzyme during the isolation process[12]. Interestingly, we have not been able to demonstrate the release of an active cytosolic inhibitor proportion following stimulation of sheep PMNs with PMA, the latter inducing an overwhelming production of highly reactive oxygen species (Table 4). Obviously, the sheep cytosolic inhibitor seems to be as sensitive to oxidative inactivation as the corresponding cytosolic inhibitor of human PMNs[28] , rendering the molecule unable to inhibit exogenously added elastase activity. The simultaneous release of the inhibitor and enzyme, however, may allow immediate complex formation, thus protecting at least part of the inhibitor from oxidative denaturation. This conclusion can be indirectly drawn from our experiment with exogenous elastase yielding an increase of enzyme activity probably due to the proteolytic degradation of the complex. Therefore, in analogy to data published by Geiger et al J29] concerning pig PMN granulocytes, the cytosolic inhibitor seems to be released from ovine PMNs together with elastase, thus preventing extracellular enzyme activity. By this means the inhibitor does protect not only the PMN from autolysis, but also the vascular system from the proteolytic action of liberated elastase. Sinha et al.[24J isolated and characterized ar
697
proteinase inhibitor («i-PI) from sheep plasma and showed that sheep c^-PI is much less effective in neutralizing elastase than human a\-PI.Thus, sheep «ι-PI seems to be of minor importance as a protecting agent in the circulation. This finding supports our conclusion that sheep elastase is not released in an active form due to neutralization by the cytosolic inhibitor. A very stable enzyme-inhibitor complex similar to the human elastase-α^-ΡΙ complex would be essential for protecting the sheep from elastase-induced tissue damage. We observed, however, that elastin partly dissociated the elastase-inhibitor complex formed with the cytosolic inhibitor, resulting in slow degradation of this substrate. Yet, whether the cytosolic inhibitor can be displaced more effectively by other substrates with higher affinity in vivo resulting in elastase-induced tissue damage awaits further investigation. In summary, clear evidence is provided by our data that the isolated enzyme represents the ovine counterpart of human leukocyte elastase and that these two elastases have many similar biochemical properties. The role of sheep elastase as a mediator of tissue damage following PMN activation in traumatic shock and other diseases is, however, still questionable, since ovine PMNs seem capable of strictly controlling extracellular elastase activity, most likely with the assistance of their effective cytosolic inhibitor.
The authors are grateful to Dr. David Hoy t for his generous support, Dr. Hans Peter Schnebli for his valuable help, Lucy Schwab for carefully correcting the manuscript, and Matt Williamson for determination of the N-terminal amino-acid sequence. This study was supported in part by a grant of the Nationalbankfonds.
References 1 2 3 4 5 6 7 8 9
Nuytinck, J.K.S., Goris, R.J.A., Redl, H., Schlag, G. & van Munster, P.J.J. (im] Arch. Surg. 121,886-890. Jochum, M., Machleidt, W. & Fritz, H. (1991) in Molecular Aspects of Inflammation (Sies, H., Flohe, L. & Zimmer, G., eds.) pp. 73-92, Springer-Verlag, Berlin. Janoff, A. (1985) Am. Rev. Respir. Dis. 132, 417-433. Cochrane, G.G., Spragg, R.G., Revak, S.D. & Schraufstatter, I. (1983) Chest 83, 67S-70S. Brigham, K.L., Bowers, R. & Haynes, J. (1979) Circ. Res. 45,292-297. Heflin, A.C.J. & Brigham, K.L. (1981) /. Clin. Invest. 68, 1253-1260. Meyrick, B. & Brigham, K.L. (1983) Lab. Invest. 48,458470. Belenko, M., Chanana, A.D., Joel, D.D., Fagerhol, M.K. & Janoff, A. (1986) Am. Rev. Respir. Dis. 133, 866874. Traber, D.L., Schlag, G., Redl, H., Strohmaier, W. & Traber, L.D. (1987) Circ. Shock22,185-193.
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
698
W.G. Junger, S. Hallström, F.C. Liu, H. Redl and G. Schlag
Redl, H., Schlag, G., Vogl, C., Schiesser, A., Paul, E., Thurnher, M., Junger, W., Traber, L.D. & Traber, D.L. (1988) Biol. Chem. Hoppe-Seyler369, Suppl. 153-156. Staub, C.N., Bland, R.D., Brigham, K.L., Demling, R., Erdmann, A.J. & Woolverton, C.W. (1975) /. Surg. Res. 19, 315-320. Junger, W., Hallström, S., Redl, H. & Schlag, G. (1988) Biol. Chem. Hoppe-Seyler369, Suppl. 63-68. Bergman, T, Carlquist, M. & Jörnvall, H. (1986) in Advanced Methods in Protein Microsequence Analysis (Wittmann-Liebold, B., Salnikow, J. & Erdmann, V.A., eds.) pp. 45-55, Springer-Verlag, Berlin-Heidelberg. Matsudaira, P. (1990) in Methods in Enzymology, vol. 182 (Deutscher, M.P., ed.) pp. 602-613, Academic Press, San Diego, Hewick, R.M., Hunkapiller, M.W., Hood, L.E. & Dreyer, W.J. (1981)7. Biol. Chem. 256,7990-7997. Fiedler, F., Geiger, R., Leysath, G. & Hirschauer, C. (1978) Biol. Chem. Hoppe-Seyler359,1667-1673. Lineweaver, H. & Burk, D. (1934) J. Am. Chem. Soc. 56, 658-666. Braun, N.J., Bodmer, J.L., Virca, G.D., Metz-Virca, G., Maschler, R., Bieth, J.G. & Schnebli, H.P. (1987) Biol. Chem. Hoppe-Seyler368, 299-308. Lieners, C., Redl, H., Schlag, G. & Hammerschmidt, D.E. (1989) Inflammation 13, 621-630.
20 21 22
23 24 25 26
27 28 29
Vol. 373 (1992)
Ruch, W, Cooper, P.H. & Baggiolini, M. (1983) 7. Im· munol. Methods 63, 347-351. Vincent, J.-P. & Lazdunski, M. (1972) Biochemistry 11, 2967-2977. Travis, J., Baugh, R., Giles, P.J., Johnson, D., Bowen, J. & Reilly, C.F. (1978) in Neutral Proteases of Human Polymorphonuclear Leukocytes (Havemann, K. & Janoff, A., eds.) pp. 118-128, Urban & Schwarzenberg, Baltimore-Munich. Sinha, S., Watorek, W, Karr, S., Giles, J., Bode, W. & Travis, J. (1987) Proc. Natl. Acad. Sei. USA 84,2228-2232. Sinha, U, Sinha, S. & Janoff, A. (1988) Am. Rev. Respir. Dis. D7, 558-563. Campbell, E.J. (1987) Am. Rev. Respir. Dis. 135, 984. Ohlsson, K. (1978) in Neutral Proteases of Human Polymorphonuclear Leukocytes (Havemann, K. & Janoff, A., eds.) p. 45, Urban & Schwarzenberg, BaltimoreMunich. Trefz, G. & Geiger, R. (1987) Biol. Chem. Hoppe-Seyler 368,1343-1353. Thomas, R.M., Nauseef, W.M., Iyer, S.S., Peterson, M.W, Stone, P.J. & Clark, R.A. (1991) 7. Leuk. Biol. 50, 568-579. Geiger, R., Sokal, S., Siebeck, M., Hoffmann, H. &Trefz, G. (1988)7. Gin. Chem. Clin. Biochem. 26, 605-609.
Dr. Wolfgang G. Junger*, Dr. Forrest C. Liu; University of California San Diego Medical Center, Department of Surgery 8236, Division of Trauma, 225 Dickinson Street, San Diego, California 92103, USA; Dr. Seth Hallström, Dr. Heinz Redl, Prof. Dr. Günther Schlag; Ludwig-Boltzmann-Institut für Experimentelle und Klinische Traumatologie, Donaueschingenstr. 13, A-1200Wien, Österreich. * To whom correspondence should be sent.
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
The Enzymatic and Release Characteristics of Sheep Neutrophil Elastase: A Comparison with Human Neutrophil Elastase Wolfgang Georg JUNGER*, Seth HALLSTROM b , Forrest Chen Liua, Heinz REDL b , Günther SCHLAG*5 a h
University of California San Diego Medical Center, Department of Surgery, San Diego, California, USA Ludwig-Boltzmann-Institut für Experimentelle und KlinischeTraumatologie, Wien, Österreich
(Received 8 January / 28 April 1992)
Summary: Sheep are often used to study tissue damage following shock after traumatic injury and in the course of other diseases. The processes involved are thought to be caused at least in part by elastase released from polymorphonuclear leukocytes (PMNs). Since little is known about elastase and its role as a mediator of tissue damage in sheep, we studied the biochemical properties and release characteristics to sheep leukocyte elastase (SLE) in comparison of those of human leukocyte elastase (HLE). Both enzymes showed similar molecular masses, amino-acid compositions, N-terminal amino-acid sequences, and abilities to digest elastin substrates. Differences, however, were found in kinetic parameters measured with the elastase-specific substrate N-methoxysuccinyl(L-alanyl)2-L-prolyl-L-valine-4-nitroanilide (MeoSucAAPV-pNa). The Michaelis constant (Km) of ovine elastase was nearly 10 times higher (1.82mM) than the
Km of HLE (0.21mM). Values of SLE calculated for kcal were 70% and for kcat/Km 8% of corresponding values determined for HLE. In addition, significant differences between sheep and human PMNs were found in in vitro stimulation experiments. In contrast to human PMNs, sheep neutrophils released no active elastase, and only 50 to 70% of the H2O2 produced by human PMNs. This failure to release active elastase could not be explained by a lower elastase content of sheep PMNs, as there were no significant differences found between the elastase contents of sheep and human PMNs. We conclude that elastase liberated by stimulated sheep PMNs is inactivated by a concomitantly released proteinase inhibitor also located within the sheep PMNs. These results together with our earlier findings indicate that sheep elastase, in contrast to human elastase, may play a minor role in mediating tissue damage and organ failure.
Neutrophile Elastase vom Schaf: Ein Vergleich mit humaner neutrophiler Elastase bezüglich Enzymcharakteristik und Freisetzung Zusammenfassung: Das Schaf dient oft als Modell zur Untersuchung von Gewebsschädigung im Zuge von traumatischem Schock und anderen Krankheitsbildern. Elastase ist eine von polymorphkernigen neutrophilen Granulozyten (PMN) freigesetzte Proteinase und wird als ein maßgeblicher Mediator die-
ser gewebsschädigenden Prozesse erachtet. Da wenig über Schafleukozytenelastase (SLE) und ihren Beitrag zur Gewebsschädigung im Schaf bekannt ist, untersuchten wir die biochemischen Eigenschaften und den Freisetzungsmechanismus dieses Enzyms und verglichen diese Daten mit denen der humanen
Enzymes: Human leukocyte elastase (lysosomal elastase from human leukocytes) EC 3.4.21.37; Sheep leukocyte elastase (lysosomal elastase from sheep leukocytes) EC 3.4.21.-. Abbreviations: BBSS, Hank's balanced salt solution; HLE, human leukocyte elastase; MeoSuc-AAPV-pNa, N-methoxysuccinyl-(L-alanyl)2-L-prolyl-L-valyl-4-nitroanilide; PMNs, polymorphonuclear leukocytes; PMA, phorbol-12-myristate-13-acetate; PITC, phenylisothiocyanate; PTH, phenylthiohydantoin; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; SLE, sheep leukocyte elastase.
Brought to you by | University of Arizona
Copyright © by Walter de Gruyter & Co · Berlin · New York Authenticated
Download Date | 5/26/15 10:38 PM
692
W.G. Junger, S. Hallström, F.C. Liu, H. Redl and G. Schlag
leukozytären Elastase (HLE). Beide Enzyme zeigten ähnliche Molmassen, Aminosäurenzusammensetzungen, W-terminale Aminosäurensequenzen und Fähigkeiten, Elastin zu spalten. Wir fanden jedoch Unterschiede der enzymkinetischen Parameter mit dem oft verwendeten elastasespezifischen Substrat vY-methoxysuccinyl-(L-alanyl) 2 -L-prolyl-L-valine-4-nitroanilid (MeoSuc-AAPV-pNa). Die Michaelis Konstante C^m) von Schafelastase war nahezu lOfach höher (1.82mM) als der A:m-Wert von HLE (0.21mM). Die Werte von SLE für £cat betrugen nur 70% und für kcj Km lediglich 8% der für HLE ermittelten Daten. Abgesehen davon konnten signifikante Unterschiede in der Elastasefreisetzung von in vitro stimulierten Schafgranulozyten und humanen Granulozyten beobachtet werden. Im Gegensatz zu humanen PMN setzten Schafgranulozyten keine aktive Elastase frei.
Vol. 373 (1992)
Außerdem produzierten Schaf-PMN nur 50 bis 70% der von humanen Zellen freigesetzten Menge an H2O2. Die Tatsache, daß Schaf-PMN keine aktive Elastase freisetzten, konnte nicht auf einen niedrigeren Elastasegehalt in Schafgranulozyten zurückgeführt werden, da sich Schaf- und humane Granulozyten nur unbedeutend in ihrem Elastasegehalt unterschieden. Wir schließen aus diesen Untersuchungen, daß aus stimulierten Schafgranulozyten freigesetzte Elastase durch einen gleichzeitig aus diesen Zellen liberierten Proteinaseinhibitor inaktiviert wird. Gemeinsam mit unseren früheren Untersuchungen weisen diese Ergebnisse darauf hin, daß Schafelastase im Gegensatz zur humanen Elastase eine untergeordnete Rolle als Mediator von Gewebsund Organschädigung spielt.
Key terms: human and sheep neutrophil elastase; biochemical characterization; PMN activation; in vitro elastase release.
The serine proteinase, elastase, is one of the most abundant lysosomal proteinases contained in polymorphonuclear leukocytes (PMNs). Excessive release of elastase after PMN stimulation is thought to contribute substantially to processes resulting in proteolytic degradation of plasma proteins and structural elements of the microvascular system in the course of several diseases'1'2'. PMN margination and activation is a common observation after severe injury leading to traumatic shock. Irreversible multi-organ failure syndrome (MOFS) and adult respiratory distress syndrome (ARDS) are thought to be primarily caused by the proteolytic action of elastase, especially as plasma proteinase inhibitor activity is reduced by concomitantly produced oxidants'1'41. Several animal models using sheep have been established to evaluate the correlation of microvascular tissue damage in the lung with PMN margination and activation in various diseases'5'101. Sheep are widely used as animal models of acute and chronic lung injury, primarily because measurements of vascular permeability changes are available by a lung lymph node preparation technique introduced by Staub and co-workers'111. Despite the common use of sheep in experiments designed to assess the role of lysosomal proteinases in tissue destruction and to explore possible pharmacologic approaches to prevent proteolytic tissue injury with specific proteinase inhibitors, little is known about the properties of neutrophil elastase in sheep and its resemblance with the human enzyme. As the sheep is used extensively in our group as a model of septic shock, we became interested in ovine elastase
and its possible role as a mediator of tissue damage in comparison with the human enzyme. In a previous paper we described the isolation and a preliminary biochemical characterization of an elastase-like neutral proteinase from sheep PMNs as well as of a serine proteinase inhibitor located in the cytosol of sheep PMNs'12l In order to complete the characterization of sheep elastase and to evaluate its potential relevance in provoking tissue damage, we gathered additional data on the biochemical properties of this enzyme and the mechanism of its in vitro release by sheep PMNs. In addition we compared these biochemical and release characteristics of sheep elastase to human elastase, as we believe that these data might be of interest to experimenters using sheep as an animal model in studying the role of neutrophil elastase in certain diseases.
Materials and Methods Elastases Sheep elastase and the cytosolic serine proteinase inhibitor we isolated from ovine neutrophils as previously described1121. Purifii
Characterization of sheep elastase Amino acid composition The amino-acid composition was determined according to the method of Bergman et al.l'3]. Purified elastase was hydrolysed with amino-acid-free OM HC1 (Pierce Chemical Company, Rockford, IL,
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
Vol. 373 (1992)
Sheep and Human Elastase: Enzyme and Release Characteristics
USA) at 110 °C for 24 h.The hydrolysis products were derivatized with phenylisothiocyanate (PITC; Pierce), lyophilized, and dissolved in HPLC starting buffer. A gradient profile was applied using 0.03M NaOH, adjusted to pH 6.6 with IM H3PO4, and 60% acetonitrile in water (Merck, Darmstadt, Germany) at a flow rate of lm//min. The Cjg-Spherisorb S3 ODS2 column (3-μ,ηι spherical particles; Phase Separations, Queensferry, Clwyd, U.K.) with dimensions of 125 x 4.6 mm was thermostated at 37 °C in a Precitherm PFVwaterbath (Boehringer Mannheim GmbH, Mannheim, Germany); detection wavelength was 254 nm.The HPLC system consisted of a Wisp 710 Β automated sample applicator, two Model 510 pumps, a Data Modul Integrator, a Waters Automated Gradient Controller, and a Model 481 LC Spectrophotometer LambdaMax (Waters Associates, Milford, MA, USA). N-Terminal amino-acid sequence Elastase was subjected to SDS-PAGE gel electrophoresis as described'121, and electroblotted to a PVDFmembrane (Imrnobilon-P transfer membrane; Millipore Corporation, Bedford, MA, USA) according to the method described by Matsudaira'14'. The membrane was stained with Coomassie Brilliant Blue R-250 and stored at - 70 °C for amino-acid sequence analysis.Three bands representing the isoforms (see Fig. 1) were cut out of the membrane and applied to the reaction chamber of the sequencer. Automated Edman degradations were performed on an Applied Biosystems Model 470 A Gas Phase Sequencer employing the O3RPTH program as supplied by the manufacturer[15] (Applied Biosystems, Foster City, CA, USA). Phenylthiohydantoin (PTH) amino acid derivatives were identified using an Applied Biosystems Model 120 on-line HPLC system. Data reduction was accomplished using a Perkin-Elmer Model 7500 computer with Chrom 3 software (Perkin-Elmer, Norwalk, CT, USA). Enzymatic activity Enzymatic activity was measured in an assay buffer consisting of 50mMTris/HCl with'lOOmM NaCl, and 0.05% Triton X-100 (Alcylphenyl polyethylenglycole; Fluka AG, Buchs, Switzerland), pH 7.4. The synthetic substrate N-methoxysuccinyl-(L-alanyl)2-L-prolyl-L-valine 4-nitroanilide (MeoSuc-AAPV-pNa) (Protagen AG, L ufelfingen, Switzerland) was dissolved in DMSO (Dimethylsulfoxide; Fluka AG) and used at a final assay concentration of ImM. Buffer containing MeoSuc-AAPV-pNa (900 μ/) was mixed with 100 μ/ of sample and measured for 5 min at 405 nm and 37 °C.The molar extinction coefficient of the synthetic substrate was 10200M"1 cm'1^. Enzymatic activity was expressed in international units (1U = 16.67 nkat, the activity necessary to convert 1 μ,ητιοί MeoSuc-AAPV-pNa/min). For all kinetic experiments a Beckman DU-70 Spectrophotometer (Beckman Instruments, Palo Alto, CA, USA) was used. Michaelis constant Km The Michaelis constants Km (with MeoSuc-AAPV-pNa) of SLE and HLE were determined according to Lineweaver and Burk'17'. MeoSuc-AAPV-pNa was added to yield final assay concentrations ranging from 0.28 to ImM for HLE, and 0.3 to 2mM for SLE. The final assay concentration of DMSO was kept constant at 1% (v/v) by preparing substrate dilutions in DMSO. All measurements were performed in triplicate. The Km values were determined from straight lines obtained by linear regression with correlation coefficients r > 0.99. Molar concentrations ofelastase solutions Molar concentrations of elastase solutions were determined by equivalent titration according to Braun et al.[181. Increasing amounts of elastase solutions of unknown concentrations were incubated with constant, known concentrations of the serine proteinase inhibitor Eglin c (courtesy of Dr. Schnebli, Ciba-Geigy,
693
Basel, Switzerland). Elastase and inhibitor solutions were mixed and incubated in the assay buffer for 10 min at 37 °C. After addition of MeoSuc-AAPV-pNa at a final assay concentration of ImM, the mixtures were incubated for 10 min, the reactions were stopped by adding 100 μΐ glacial acetic acid, and released 4-nitroaniline was measured at 405 nm. Residual enzymatic activities were plotted versus the volume of added elastase solution and the equivalent molar concentration determined graphically as described1181. Degradation of elastin substrates The capability of ovine and human neutrophil elastase to degrade elastin was compared. [3H]elastin (3.95 /u.Ci/mg) and human leukocyte elastase (HLE) were kindly provided by Dr. Schnebli (CibaGeigy, Basel, Switzerland). Orcein-elastin (bovine neck ligament) was purchased from Sigma. Equal amounts of elastin derivatives were washed twice in assay buffer (used in enzymatic measurements) to remove unbound tritium or orcein dye and suspended in the same buffer. Aliquots of the suspensions were mixed with lOOpmol of human or ovine neutrophil elastase. Normal saline was used as a blank. The mixtures were incubated for 2 h at 37 °C with constant agitation. After centrifugation, liberated orcein dye was measured in the supernatants at a wavelength of 578 nm. Liberation of tritium was measured after mixing 10 μΐ of the supernatant with 4.5 ml of Ready Value scintillation liquid (Beckman Instruments) with a BF5000/300 beta-counter (Berthold, Wildbad, Germany) for 4 min.
In vitro stimulation experiments PMN isolation Human and ovine PMNs were isolated from buffy coats of healthy individuals or animals, respectively, by counter flow centrifugation employing a Beckman J2-21M/E with a JE6B-rotor (Beckman Instruments) as previously described'12'. This method yielded PMN preparations with purities > 98%, as examined by differential cell counting, and cell viabilities > 95%, as determined by trypan-blue dye exclusion. Analysis of released elastase activity Equal concentrations of sheep or human PMNs (1.0 x 106 cells/m/) were stimulated by incubation with either 2 mg/m/ opsonized zymosan or Ο.Ιμ,Μ phorbol-12-myristate-13-acetate (PMA; Sigma Chemical Company, St. Louis, MO, USA) in Hanks' balanced salt solution (HBSS). Zymosan was opsonized with autologous plasma according to the method described by Lieners et al.'19' immediately before use. Opsonized zymosan was washed 5 times with HBSS to remove serum arproteinase inhibitor. The cell suspensions were incubated at 37 °C for 1 h, centrifuged, and analysed for released elastase activity as described above. To determine if activated sheep PMNs released an excess of their cytosolic proteinase inhibitor, 100 μΐ of supernatants were incubated with 10 μΐ purified SLE at 37 °C in the buffer used for kinetic measurements. After 10 min of pre-incubation, the enzymatic activity of SLE was determined as described above. Determination ofH2O2 HaOa production of activated PMNs was determined fluorometrically according to the method described by Ruch et al.[20]. In brief: 0.5 ml of a 0.4mM HVA homovanillic acid (HVA; Serva, Heidelberg, Germany) in HBSS, containing 4 U/m/ horseradish peroxidase (HRP; Boehringer Mannheim) was mixed with 0.1 m/of an appropriate dilution of stimulating agent. The mixtures were brought to a volume of 1.9 ml with HBSS and pre-incubated for 5 min at 37 °C.Then 100 μΐ of cell suspensions were added to obtain final cell concentrations of 1.0 x 106 cells/assay and incubated for an additional 30 min at 37 °C under steady shaking in a water bath. Liberation of H2O2 was terminated by addition of 250 μΐ of an ice-
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
694
W.G. Junger, S. Hallstr m, F.C. Liu, H. Redl and G. Schlag
Vol. 373 (1992)
cold solution containing O.lM glycine and 15mM EDTA, pH 12 (adjusted with NaOH).The assay mixtures were centrifuged and concentrations of the reaction product 2,2'-dihydroxy-3,3'-dimethoxydiphenyl-5,5'-diacetic acid in the supernatants measured with a F-4000 Fluorescence Spectrophotometer (Hitachi Ltd., Tokyo, Japan) at 25 °C using an excitation wavelength of 312 nm and an emission wavelength of 420 nm. Degradation of [3H]elastin by stimulated PMNs Sheep and human PMNs (1.5 x 106 cells/assay), respectively, were incubated with l mg [ 3 H]elastin in HBSS at 37 °C. Final assay volume was 150 μ,/. The cells were stimulated with PMAat a final concentration of Ο.Ιμ,Μ. Aliquots of the reaction mixture (50 μ,/) were taken at different time intervals and centrifuged for 2 min in a microcentrifuge. Released tritium was measured in 45 μΐ of the supernatants as described above. Elastase content of sheep PMNs Granule-rich preparations of known amounts of purified ovine PMNs were washed and extracted according to the method described'121. Total enzymatic activities of the extracts were determined as indicated above.The original elastase amount contained in the PMNs was estimated, based on the total extracted activity and the purified enzyme's specific enzymatic activity, which had been determined with 2841.5 U/g using the substrate MeoSucAAPV-pNa. Other methods Protein assays Protein concentrations were determined by the bicinchoninic acid protein assay (BCA) reagent (Pierce Chemical Company) with human serum albumin as standard, or with an Antek 771 Pyroreactor nitrogen analyser (Antek, Houston, TX, USA) using urea as standard. Association rate constant The association rate constant (kon) of the reaction of sheep elastase with the cytosolic inhibitor of sheep PMNs was determined according to the method described by Vincent and Lazdunski'21'.
Results Ovine elastase was isolated from purified sheep granulocytes resulting in a chromatographically homogeneous preparation as described earlieril2]. SLE in this preparation had a specific enzymatic activity of 2840 U/g protein. The total activity increased after the first Chromatographie purification step probably due to the removal of a proteinase inhibitor. SDS-PAGE electrophoresis of the purified SLE preparation yielded three protein bands with molecular masses of 22, 24, and 25 kDa, which were used for N-terminal amino-acid sequence analysis (see Fig. 1). The amino-acid composition determined for SLE was compared to data published for human neutrophil elastase as shown in Table l.The amino-acid distribution in HLE had been determined by analysis of hydrolysed protein' 22 ^, and by calculations based on sequencing data performed by Sinha et al.1231. SLE
Fig. 1. Sheep elastase immobilized on PVDF membrane for determination of the N-terminal amino-acid sequence. All three bands (arrows) representing the elastase isoforms were used for sequencing.
Table 1. Amino-acid composition of sheep neutrophil elastase in comparison to data for human leukocyte elastase published by Travis et al. [22] and Sinha et al.[23]. (The values shown are mol/100 mol amino acids). Data for SLE and HLE by Travis et al.'22' were obtained by hydrolysis (Asn, Gin, and Trp not determined), values for HLE by Sinha et al.'231 were derived from sequencing data. Amino acid
Asp + Asn Glu + Gin Ser Gly Thr His Ala Pro Arg Tyr Val Met Cys He Leu Phe Lys Trp
SLE
[mol-%]
HLE Travis et al. [mol-%]
HLE Sinha et al. [mol-%]
8.15 9.06 9.83 13.32 3.94 2.71 7.72 5.73 9.66 3.00 6.95 0.56 2.87 3.45 7.39 2.59 2.65 n.d.
10.62 7.96 5.75 12.39 3.10 1.77 10.62 4.42 9.73 0.88 11.06 0.88 2.65 4.87 8.85 3.98 0.44 n.d.
9.63 6.88 5.50 11.01 2.75 2.29 10.09 4.13 8.72 0.92 12.39 1.38 3.67 5.50 9.63 4.13 0.00 1.38
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
Vol. 373 (1992)
695
Sheep and Human Elastase: Enzyme and Release Characteristics
contains higher amounts of serine and lysine, and smaller quantities of alanine and valine than HLE. The lower valine content of SLE could be due to incomplete hydrolysis of valyl bonds during the 24-h hydrolysis period. Besides these exceptions, similar amino-acid distribution patterns of human and sheep elastase were seen. The first residues of the N-terminal amino-acid sequence of SLE were determined and compared to the sequence published for HLE by Sinha et alJ23l Table 2 displays the N-terminal sequences of SLE and HLE. With the exception of the amino acids at positions 6 and 15, the sequences of these two elastases are identical. Throughout the sequence analysis no evidence of microheterogeneity could be found , which indicates that the SLE isoforms may share the same peptide moiety and differ only in the carbohydrate residue. (All three bands were used for the determination. See Fig. l.)The same has been reported for Despite the similarities in the structure of sheep and human elastase, marked differences were found in their kinetic properties with the widely used elastasespecific substrate MeoSuc-AAPV-pNa. The Michaelis constant Km of sheep elastase with MeoSucAAPV-pNa was 1.82 x 10~3M, which is about 10 times higher than the Km value determined for human elastase (0.21 x 10"3M). The turnover number fccat was calculated with 16 s'1 for SLE and 23 s~l for HLE. The kcJKm values of SLE and HLE were determined as 8800 and 111900M"1 s"1, respectively. The capability of SLE to degrade 3H- and orceinlabeled elastin, however, was similar to that of HLE (Table 3). After 2 h of incubation at 37 °C, SLE had
Table 2. Comparison of N-terminal amino-acid sequence of SLE with the sequence of HLE determined by Sinha et al.'23'. Only positions 6 and 15 do not correspond. Position
SLE
HLE
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
He Val Gly Gly Arg Ala Ala Arg Pro His Ala Trp Pro Phe He
He Val Gly Gly Arg Arg Ala Arg Pro His Ala Trp Pro Phe Met
TableS. Equal molar concentrations (100 pmol) of SLE and HLE were mixed with [3H]elastin (4.48 mg) and orcein-elastin (2.20 mg) in 0.2 ml of buffer (50mM Tris/HCl, lOOmM NaCl and 0.05% Triton X-100,pH7.4). Liberated orcein dye and tritium were measured after 2 h incubation at 37 °C (n = 4 assays; values: mean ± SD). Leukocyte elastase
Orcein elastin dye liberation (A578)
KT5 x 3H-labeled elastin tritium liberation [cpm]
SLE HLE
0.299 ± 0.045 0.136 ±0.052
4.060 ± 0.062 3.426 ±0.076
liberated slightly more tritium than HLE and digested about two times more orcein-elastin than the same molar amount of HLE. Although these assays can only provide semi-quantitative data, they clearly indicate that SLE has a capability to degrade the natural substrate elastin, which is slightly more efficient than that of HLE. The in vitro stimulation experiments performed with sheep and human PMNs, however, showed a significant difference in elastase release. As summarized in Table 4, neither of the stimuli used to activate sheep PMNs resulted in a release of measurable elastase activity. Human PMNs liberated relatively low amounts of active elastase when stimulated with opsonized zymosan. Activation with PMA, however, caused the liberation of 23.4 ± 4.9 mU elastase per 106 cells. Sheep PMNs were found to produce about 50% less H2O2 than human cells when stimulated with PMA, and 70% less with zymosan activation.
Table 4. Liberation of elastase and H2O2 and by sheep and human PMNs in response to activation with 2 mg/m/ opsonized zymosan (zymo) or Ο.Ιμ,Μ phorbol-12-myristat-13-acetate (PMA). Values given are mean ± SD of 6 determinations (n = 6). Liberation product
Stimuli
Sheep PMNs Human PMNs
Elastase [mU/106PMN]
zymo PMA
0 0
1.5 ± 0.6 23.4 ± 4.9
H202 [nmol/106PMN]
zymo PMA
3.1±0.8 20.0 + 4.2
10.6 ± 6.8 42.6 ±18.2
We examined the elastase content of sheep PMNs in order to determine if a lower elastase content in the ovine neutrophil may be partly responsible for the fact that activation of sheep PMNs does not lead to liberation of measurable elastase activity. The content of elastase in sheep granulocytes was estimated to be Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
696
Vol. 373 (1992)
W.G. Junger, S. Hallstr m, EC. Liu, H. Redl and G. Schlag
1.3 ± 0.7 pg elastase per PMN, based on activity measurements. This value may be up to twice as high if losses of enzymatic activity during the isolation procedure and the contamination of the extract with the cytosolic inhibitor in sheep PMNs are taken into account. The anticipated elastase content of sheep PMNs, however, would still be well within the range of 1-4 pg/PMN described for human neutroThe absence of active elastase in the supernatants of in vitro stimulated sheep PMNs indicates that SLE may be released as a complex with the cytosolic proteinase inhibitor in sheep PMNs. To test if the cytosolic proteinase inhibitor is released by stimulated sheep PMNs into the supernatants, sheep PMNs at a concentration of l x 106 cells/m/ were stimulated with 0.2μΜ PMA for l h at 37 °C. The supernatants were incubated with purified SLE and the residual activity was measured. The addition of purified SLE generated elastase activity, suggesting that the cytosolic inhibitor may be destroyed by the added SLE, resulting in liberation of active elastase from the elastase-inhibitor complex. Based on the increase in total activity in the assay mixture, we estimated the elastase activity which would be released by stimulated sheep PMNs if the enzyme were not neutralized by the cytosolic inhibitor with 32.4 mU/106 PMNs (Table 5).
1
2 3 4 Incubation time [h]
5
6
Fig. 2. Degradation of [3H]elastin by in vitro stimulated human and sheep PMNs. [3H]elastin was incubated with human and sheep PMNs, respectively. The cells were stimulated as described in the text. Sheep cells were capable of degrading this substrate, but at a significantly lower rate than human PMNs.
Discussion
The biochemical data on sheep neutrophil elastase presented here and in a previous paper^ demonstrate clearly its similarity with the human enzyme. Besides the molecular masses, and the isoelectric points of the three isoforms, very similar amino-acid compositions (Table 1) and N-terminal amino-acid sequences (Table 2) of human and ovine elastase were found. Furthermore we could show that ovine and human neutrophils do not differ significantly in their Although no active elastase was found in the superelastase contents. Ohlsson[26] found 3-4 pg/PMN, natants of in vitro stimulated sheep PMNs, [3H]elasSinha et al.[24] reported 1.0 pg/cell, and Campbell[25] tin was degraded when directly incubated with stimucalculated 1.5 pg/cell for the elastase amount conlated sheep PMNs. This suggests that SLE can be tained in human neutrophils. With an estimated 1.3 ± partly dissociated from the complex formed with the 0.7 pg elastase per cell, the elastase content in sheep cytosolic proteinase inhibitor when a high affinity PMNs lies well within the range found for human substrate such as elastin is present. Degradation of shown to contain high [3H]elastin by in vitro stimulated sheep PMNs, how- neutrophils. Sheep PMNs were [12] levels of an elastase inhibitor ; if loss of activity ever, was significantly less effective than that by during enzyme isolation is taken into account (see rehuman PMNs (Fig. 2). sults) the estimated value for the elastase content could be even higher.
fable 5. Elastase activity in supernatants of stimulated sheep PMNs was generated by incubation of the supernatants with purified SLE. The elastolytic activity of the supernatant increased with prolonged incubation with purified SLE. The elastase activity released by 106 PMN was calculated based on the activity liberated by incubation of PMN supernatants with purified SLE as described in the text. Incubation time [min]
Theoretically released elastase activity [mU/106PMN]
0 5 10
0.0 20.5 32.4
Differences, however, were observed in the kinetic properties of SLE and HLE with the chromogenic elastase substrate MeoSuc-AAPV-pNa. The Michaelis constant Km of SLE was nearly 10 times higher than that of HLE, indicating a lower affinity of SLE for this synthetic substrate. The turnover number of SLE was 30% lower than that of HLE, and the kcJKm value of HLE was nearly 13 times higher than that of SLE. The ability of SLE to degrade elastin, one of its natural substrates which are destroyed in the process of tissue damage, however, was even somewhat better than that of HLE (Table 3). Besides the differences in the interaction with the synthetic elastase substrate, more important differences
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
Vol. 373 (1992)
Sheep and Human Elastase: Enzyme and Release Characteristics
were found in the elastase release mechanisms of human and sheep PMNs (Table 4). Unlike human neutrophils, sheep PMNs did not liberate any active elastase when stimulated with either PMA or opsonized zymosan. This cannot be ascribed to differences in the elastase amounts present in human and sheep PMNs, as discussed above. Although H2O2 production by sheep PMNs was about 50% lower than that of human neutrophils, a higher threshold of ovine versus human cells in response to the stimuli used is not likely to be solely responsible for the significant differences seen in elastase liberation. In an earlier publication we showed that sheep PMNs contain considerable amounts of an effective serine proteinase inhibitor in their cytosol[12]. The association rate constant kon of the inhibitor with SLE was determined as 2.3 x K^ivrV1, which is close to the value published for the analogue inhibitor/elastase pair in porcine neutrophils[27]. The presence of this inhibitor together with the low solubility, as well as the strong tendency of SLE to adhere to surfaces at low ionic strength may also have led to an underestimation of the elastase content in sheep PMNs by others[24]. As described earlier, we used a IM NaCl solution to extract elastolytic activity after separation of lysosomal granule preparations from the PMN cytosolic fraction, precautions necessary to avoid excessive loss of active enzyme during the isolation process[12]. Interestingly, we have not been able to demonstrate the release of an active cytosolic inhibitor proportion following stimulation of sheep PMNs with PMA, the latter inducing an overwhelming production of highly reactive oxygen species (Table 4). Obviously, the sheep cytosolic inhibitor seems to be as sensitive to oxidative inactivation as the corresponding cytosolic inhibitor of human PMNs[28] , rendering the molecule unable to inhibit exogenously added elastase activity. The simultaneous release of the inhibitor and enzyme, however, may allow immediate complex formation, thus protecting at least part of the inhibitor from oxidative denaturation. This conclusion can be indirectly drawn from our experiment with exogenous elastase yielding an increase of enzyme activity probably due to the proteolytic degradation of the complex. Therefore, in analogy to data published by Geiger et al J29] concerning pig PMN granulocytes, the cytosolic inhibitor seems to be released from ovine PMNs together with elastase, thus preventing extracellular enzyme activity. By this means the inhibitor does protect not only the PMN from autolysis, but also the vascular system from the proteolytic action of liberated elastase. Sinha et al.[24J isolated and characterized ar
697
proteinase inhibitor («i-PI) from sheep plasma and showed that sheep c^-PI is much less effective in neutralizing elastase than human a\-PI.Thus, sheep «ι-PI seems to be of minor importance as a protecting agent in the circulation. This finding supports our conclusion that sheep elastase is not released in an active form due to neutralization by the cytosolic inhibitor. A very stable enzyme-inhibitor complex similar to the human elastase-α^-ΡΙ complex would be essential for protecting the sheep from elastase-induced tissue damage. We observed, however, that elastin partly dissociated the elastase-inhibitor complex formed with the cytosolic inhibitor, resulting in slow degradation of this substrate. Yet, whether the cytosolic inhibitor can be displaced more effectively by other substrates with higher affinity in vivo resulting in elastase-induced tissue damage awaits further investigation. In summary, clear evidence is provided by our data that the isolated enzyme represents the ovine counterpart of human leukocyte elastase and that these two elastases have many similar biochemical properties. The role of sheep elastase as a mediator of tissue damage following PMN activation in traumatic shock and other diseases is, however, still questionable, since ovine PMNs seem capable of strictly controlling extracellular elastase activity, most likely with the assistance of their effective cytosolic inhibitor.
The authors are grateful to Dr. David Hoy t for his generous support, Dr. Hans Peter Schnebli for his valuable help, Lucy Schwab for carefully correcting the manuscript, and Matt Williamson for determination of the N-terminal amino-acid sequence. This study was supported in part by a grant of the Nationalbankfonds.
References 1 2 3 4 5 6 7 8 9
Nuytinck, J.K.S., Goris, R.J.A., Redl, H., Schlag, G. & van Munster, P.J.J. (im] Arch. Surg. 121,886-890. Jochum, M., Machleidt, W. & Fritz, H. (1991) in Molecular Aspects of Inflammation (Sies, H., Flohe, L. & Zimmer, G., eds.) pp. 73-92, Springer-Verlag, Berlin. Janoff, A. (1985) Am. Rev. Respir. Dis. 132, 417-433. Cochrane, G.G., Spragg, R.G., Revak, S.D. & Schraufstatter, I. (1983) Chest 83, 67S-70S. Brigham, K.L., Bowers, R. & Haynes, J. (1979) Circ. Res. 45,292-297. Heflin, A.C.J. & Brigham, K.L. (1981) /. Clin. Invest. 68, 1253-1260. Meyrick, B. & Brigham, K.L. (1983) Lab. Invest. 48,458470. Belenko, M., Chanana, A.D., Joel, D.D., Fagerhol, M.K. & Janoff, A. (1986) Am. Rev. Respir. Dis. 133, 866874. Traber, D.L., Schlag, G., Redl, H., Strohmaier, W. & Traber, L.D. (1987) Circ. Shock22,185-193.
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
698
W.G. Junger, S. Hallström, F.C. Liu, H. Redl and G. Schlag
Redl, H., Schlag, G., Vogl, C., Schiesser, A., Paul, E., Thurnher, M., Junger, W., Traber, L.D. & Traber, D.L. (1988) Biol. Chem. Hoppe-Seyler369, Suppl. 153-156. Staub, C.N., Bland, R.D., Brigham, K.L., Demling, R., Erdmann, A.J. & Woolverton, C.W. (1975) /. Surg. Res. 19, 315-320. Junger, W., Hallström, S., Redl, H. & Schlag, G. (1988) Biol. Chem. Hoppe-Seyler369, Suppl. 63-68. Bergman, T, Carlquist, M. & Jörnvall, H. (1986) in Advanced Methods in Protein Microsequence Analysis (Wittmann-Liebold, B., Salnikow, J. & Erdmann, V.A., eds.) pp. 45-55, Springer-Verlag, Berlin-Heidelberg. Matsudaira, P. (1990) in Methods in Enzymology, vol. 182 (Deutscher, M.P., ed.) pp. 602-613, Academic Press, San Diego, Hewick, R.M., Hunkapiller, M.W., Hood, L.E. & Dreyer, W.J. (1981)7. Biol. Chem. 256,7990-7997. Fiedler, F., Geiger, R., Leysath, G. & Hirschauer, C. (1978) Biol. Chem. Hoppe-Seyler359,1667-1673. Lineweaver, H. & Burk, D. (1934) J. Am. Chem. Soc. 56, 658-666. Braun, N.J., Bodmer, J.L., Virca, G.D., Metz-Virca, G., Maschler, R., Bieth, J.G. & Schnebli, H.P. (1987) Biol. Chem. Hoppe-Seyler368, 299-308. Lieners, C., Redl, H., Schlag, G. & Hammerschmidt, D.E. (1989) Inflammation 13, 621-630.
20 21 22
23 24 25 26
27 28 29
Vol. 373 (1992)
Ruch, W, Cooper, P.H. & Baggiolini, M. (1983) 7. Im· munol. Methods 63, 347-351. Vincent, J.-P. & Lazdunski, M. (1972) Biochemistry 11, 2967-2977. Travis, J., Baugh, R., Giles, P.J., Johnson, D., Bowen, J. & Reilly, C.F. (1978) in Neutral Proteases of Human Polymorphonuclear Leukocytes (Havemann, K. & Janoff, A., eds.) pp. 118-128, Urban & Schwarzenberg, Baltimore-Munich. Sinha, S., Watorek, W, Karr, S., Giles, J., Bode, W. & Travis, J. (1987) Proc. Natl. Acad. Sei. USA 84,2228-2232. Sinha, U, Sinha, S. & Janoff, A. (1988) Am. Rev. Respir. Dis. D7, 558-563. Campbell, E.J. (1987) Am. Rev. Respir. Dis. 135, 984. Ohlsson, K. (1978) in Neutral Proteases of Human Polymorphonuclear Leukocytes (Havemann, K. & Janoff, A., eds.) p. 45, Urban & Schwarzenberg, BaltimoreMunich. Trefz, G. & Geiger, R. (1987) Biol. Chem. Hoppe-Seyler 368,1343-1353. Thomas, R.M., Nauseef, W.M., Iyer, S.S., Peterson, M.W, Stone, P.J. & Clark, R.A. (1991) 7. Leuk. Biol. 50, 568-579. Geiger, R., Sokal, S., Siebeck, M., Hoffmann, H. &Trefz, G. (1988)7. Gin. Chem. Clin. Biochem. 26, 605-609.
Dr. Wolfgang G. Junger*, Dr. Forrest C. Liu; University of California San Diego Medical Center, Department of Surgery 8236, Division of Trauma, 225 Dickinson Street, San Diego, California 92103, USA; Dr. Seth Hallström, Dr. Heinz Redl, Prof. Dr. Günther Schlag; Ludwig-Boltzmann-Institut für Experimentelle und Klinische Traumatologie, Donaueschingenstr. 13, A-1200Wien, Österreich. * To whom correspondence should be sent.
Brought to you by | University of Arizona Authenticated Download Date | 5/26/15 10:38 PM
