Bacterial Lipopolysaccharide Enhances Chemoattractant - CiteSeerX

4 downloads 0 Views 2MB Size Report
to Fernandez and Hugh [8] as reported previously. [48] . ...... 1965. 12. Guthrie,. L.A. , McPhail,. L.C. , Henson,. P.M. , and Johnston,. RB. Jr. Priming of neutrophils.
Journal

of Leukocyte

Biology

43:547-556

(1988)

Bacterial Lipopolysaccharide Enhances ChemoattractantInduced Elastase Secretion by Human Neutrophils C. Fittschen, Department

R.A.

Sandhaus,

of Pediatrics (CF., Center for Immunology

G.S.

Worthen,

and

P.M.

Henson

P.M.H.) and Medicine (R.A.S., G.S.W.), National and Respiratory Medicine, Denver, Colorado

Jewish

Bacterial lipopolysaccharide (LPS) has previously been shown to enhance a number of chemoattractant-induced responses by human neutrophils. The possible role of elastase, a neutral protease with broad substrate specificity, In neutrophil-medlated vascular injury of a variety of diseases prompted us to examine a) whether or not LPS enhances the direct chemoattractant-lnduced secretion of elastase, b) the quantitative requirements of LPS and chemotactic factors, and c) some structural requirements of LPS for this effect. Our results show that LPS at 10 ng/ml and above, enhanced formyl-methionyl-leucyl-phenylalanine-lnduced neutrophil secretion of elastase, as well as secretion of myeloperoxidase and vitamin B12-blnding protein. This effect was independent of cytochalasins or surface stimulation, and thus may occur during chemotactic factor stimulation in vivo. LPS also enhanced neutrophil secretory responses to the complement fragments C5a, C5a des arg, and, to a lesser degree, to leukotriene B4 and plateletactivating factor. This enhancement effect appeared to require the presence of the lipid A moiety and/or parts of the core polysaccharide but not the 0-antigen portion of the LPS molecule. Our findings identify a possible LPS-dependent mechanism of neutrophil

elastase-mediated Key words:

tissue

chemotactic

injury factors,

in Gram-negative lysosomes,

INTRODUCTION Neutrophil responses to a variety of stimulants that may be encountered in an inflammatory field [28] include secretion of granule contents [3 18,46] and the generation of oxygen radicals [7,38] . While a limited but directed release of these neutrophil products may be required for penetration of tissue barriers such as during vascular egress and during tissue migration [39], massive secretion of granule content and/or oxygen radicals may cause injury of adjacent cells and connective tissue structures [7, 24]. Recent studies from our laboratory have shown that bacterial lipopolysaccharide (LPS) augments both the secretion of lysosomal enzymes [15] and the generation of oxygen radicals [12] by neutrophils stimulated with formly-methionyl-leucyl-phenylalanine (FMLP). Such an enhanced release of toxic products by neutrophils may be important for tissue injury as shown by Smedly et al. [43], who found that neutrophils exposed to small amounts of LPS prior to stimulation with FMLP, but not neutrophils treated with LPS or FMLP alone, damaged cultured vascular endothelial cells. Identification of neutrophil elastase as the injurious principle in the latter study, as well as evidence for protease-mediated damage to endothelial monolayers [13, 14] suggest that neutrophil granule enzymes, and in particular, elastase, ,

© 1988 Alan

R. Lisa,

Inc.

infections. specific

granules

may contribute to neutrophil-mediated injury of vascular endothelium. Early studies suggest that the range of yessel wall components susceptible to proteolytic attack may be even broader. Elastase may be responsible for fragmentation of arterial internal elastic membrane in cardiovascular lesions of experimental animals with serum sickness [26], as well as for the liberation of basement membrane fragments in experimental nephrotoxic nephritis [161 since neutrophil elastase can, in addition to elastin, degrade basement membrane (Type IV) collagen [32]. While the identification of elastase in bronchoalveolar lavage fluid from patients with pulmonary edema due to endotoxemia-associated adult respiratory distress syndrome (ARDS) appear to implicate a contribution of elastase to the pathogenesis of endotoxemia-associated lung disease, several laboratories have shown that toxic oxygen species such as hydrogen peroxide (H2O2) or one of its derivates [40,44,45] may also play a prominent role in this process. Taken together, the activity of neutrophil elastase against connective tissue constituents

Received

August

10, 1987;

accepted

Reprint requests: C. Fiuschen, ogy and Respiratory Medicine,

January

National Denver,

19, 1988.

Jewish Center CO 80206.

for Immunol-

548

Fittschen

et al

as well as its capability to injure endothelial cells [43] suggest that secretion of elastase may be one of the mechanisms by which neutrophils cause injury to vessel walls and other tissues. Neutrophils are known to secrete a large proportion of elastase-containing azurophil granules and of specific granules during phagocytosis [3, 17], and during interaction with opsonin-coated surfaces [19,47] . In contrast, chemoattractants have been reported to induce secretion of low percentages of granule constituents, mostly from specific granules [10,48,51], unless they act in the presence ofcytochalasin B (CB). This combination, however, which elicits extensive release of both azurophil and specific granules [20, 22,52] represents an unphysiologic stimulus as CB is not normally encounterd by neutrophils in vivo. Recent observations of an enhancement of chemoattractant-induced granule secretion by LPS [15] suggest a more natural mechanism by which chemotactic factors might induce secretion of larger amounts of granule components capable of damaging tissue structures. In this process, LPS could play a dual, indirect role by first generating chemoattractant complement fragments [35], and subsequently enhancing the neutrophil secretory response to those fragments and to other inflammatory mediators. In order to further study LPS-mediated enhancement of stimulant-induced neutrophil secretion, we examined quantitative and structural requirements of LPS for the secretion of elastase, comparing the release of this enzyme with that of the azurophil marker myeloperoxidase (MPO) and of the specific granule marker vitamin B12binding protein (Vit B12 BP). We also compared the extent to which LPS enhances the secretagogue effect of the chemotactic peptides FMLP, C5a and C5a des arg, and of the lipids platelet-activating factor (PAF) and leukotriene B4 (LTB4). [24,32]

MATERIALS

AND

METHODS

Reagents All reagents and plasticware employed in studies of neutrophil secretion were tested for the presence of lipopolysaccharide with the Limulus Amebocyte Lysate Kit (Associates of Cape Cod, Woods Hole, MA), which could detect concentrations of 0.01 ng LPS/ml. Using this method, no LPS was detected on sterile plastics and all reagents tested at the concentrations used, contained less than 0. 1 ng/ml LPS. The salts for the preparation of Krebs-Ringer phosphate buffer (pH 7.2) were obtained from Mallinkrodt, (Paris, KY) and had undetectable LPS levels. The buffer, supplemented with LPS-free dextrose (5% dextrose in 0.2% sodium chloride, injectable, Abbott Laboratories, North Chicago, IL) to contain a final concentration of 0.2 % is subsequently called KRP-D. ,

Human serum albumin (Cutter Laboratories, Emergyville, CA), chosen to be free of LPS, was added to give a final concentration of 0.25 % . The complete assay buffer containing dextrose and albumin is subsequently referred to as KRP-DA. The chemoattractant formyl-methionyl-leucyl-phenylalanine (FMLP), Vega Biochemicals, Tucson, AZ) was dissolved in dimethylsulfoxide (DMSO, Fisher Scientific Co. , Fairlawn, NJ) at iO M and kept frozen at 18#{176}C. Before use, this stock was diluted in KRP-DA. Biologically active C5a was prepared from human plasma according to Fernandez and Hugh [8] as reported previously [48] . C5a des arg was derived from C5a by passage through a sephadex column containing carboxypeptidasecoated beads (size 6B) [48]. Both C5a and C5a des arg were stored in phosphate-buffered saline at 5 x 106 M and diluted before use with KRP-DA. Leukotriene B4 (LTB4), a kind gift of Dr. R. Murphy, University of Colorado, was stored in absolute methanol at 18#{176}C; before use, the alcohol was evaporated under a stream of nitrogen and the residue dissolved in KRP-DA. Plateletactivating factor (PAF) ( 1-alkyl-2-acetyl-sn-glycero-3phosphocholine, in which the alkyl chain is mainly C16 and 8) obtained from Avanti Polar Lipids (Birmingham, AL) was dissolved in a chloroform/methanol mixture (10: 1). For preparation of a stock solution, an aliquot was evaporated under a stream of nitrogen; the residue was redissolved in assay buffer at l0’ M and frozen at -

-18#{176}C.

LPS, prepared from Escherichia coli strain K235 by the method of Mclntire et al. [33] was a kind gift of Dr. D. Morrison, University of Kansas, Kansas City. LPS from E. coli strain 0111 :B4, prepared to chromatographical purity according to Westphal, and LPS from Salmonella Minnesota Re 595, prepared by the method of Galanos, were obtained from List Biologicals Inc. (Cambell, CA). The lyophilized LPS was dissolved in LPSfree saline at 1 mg/ml, sonicated on ice using a Sonicator Cell Disrupter (Ultrasonics, Inc. Plainview, NY) with a microtip at an amplitude setting of 2, for two 10 sec periods. Aliquots were kept frozen at 18#{176}C. Before use 100 l of this stock solution was diluted with 900 jtl KRPD without albumin, sonicated again on ice for two 10 sec periods at setting 2 and diluted in KRPDA as required. Human neutrophil elastase was obtained from Elastin Products Co. (Pacific, MO). The specific elastase inhibitor methoxy-succinyl-alanyl-alanyl-prolyl-valine-chloromethyl ketone (AAPVCK) was custom synthesized (Bachem Inc. Torrance, CA). It was dissolved in DMSO and diluted in assay buffer to iO M; the final concentration of DMSO was 0.05 % . A highly purified preparation of alpha1 proteinase inhibitor (alpha1 P1) admixed with 25-40% BSA was obtained from Sigma Chemical -

,

LPS Enhances Co. (St. Louis, MO), and diluted in KRP-D to give 0.3% w/v in the final assay mixture. Insoluble elastin from bovine ligamentum nuchae (Elastin Products Co. Pacific, MO) was tritiated by reduction with 3H NaBH3 as described [2]. Subsequently, the preparation was washed at least ten times to remove unbound label by alternating centrifugation at 3 ,800 x g with suspension in 50 ml in glass-distilled water. Aliquots of approximately 80 mg elastin were stored at -70#{176}C. Before determination of elastase activity, the substrate was prepared as follows: An aliquot was thawed, washed at least five times as described above, suspended in 15 ml of glass-distilled water, and sonicated in an icebath using a Sonicator Cell Disrupter (Ultrasonics, Inc., Plainview, NY) with a microtip at setting 4 for three 10mm periods with intervals of 10 mm to allow the sonicate to cool. Subsequently, the preparation was washed at least another five times as stated above and finally was suspended in 20 ml of isotonic saline. The working solution for quantitative determination of Vitamin B12-binding protein (Vit B12 BP) was prepared by mixing 80 jil cyanocobalamin-57 Co (specific activity 100-300 jCi/rg), as supplied by Amersham Corp. (Arlington Heights, IL) with 1 ml of unlabeled cyanocobalamin (Vit B12) as 67 ng/ml (Sigma Chemical Co. St. Louis, MO) and with 50 ml H2O. ,

,

Neutrophil

Preparation

Neutrophil

Elastase

Secretion

549

centrifuged at 1 ,400 x g and the supernatant of each tube was subdivided into fractions required for each assay. For determination of elastase activity 250 jtl of the supernatant were incubated at 37#{176}C for 1 hr with 200 jtl of the sonicated 3H-elastin suspension of 500 zl Eppendorf tubes (VWR Scientific, Inc. San Francisco, CA). After centrifugation in a Beckman tabletop microfuge 12 (Beckman Instruments, Inc. Fullerton, CA) at setting 12, 200 j.d of the supernatant containing solubilized elastin fragments were mixed with scintillant (Scintiverse II, Fisher Scientific, Fairlawn, NJ) and counted in a scintillation counter (Model LS 1801 Beckman Instruments Inc. Irvine, CA). The enzyme activity measured always was within the linear range of the standard dose-response curve. As the presence of cytoplasmic inhibitors released during homogenization [23] precluded determination of the total neutrophil elastase content, secretion of elastase activity was not expressed as percent release of total cell elastase but as the amount of 3H-elastin solubilized (in counts per mm, CPM in the standard assay). Lactic dehydrogenase (LDH) [19] and myeloperoxidase (MPO) were assayed as previously described [2 1 J. Enzyme activity in the neutrophil supernatant was expressed as a percentage of the enzyme activity in neutrophils (l07/ml) lysed with 10 jl, 10% Triton X-lO0. The secretion of Vitamin B,2-binding protein (Vit B12 BP) was assessed using the method described by Gottlieb [11] The amount of Vit 12 BP secreted was expressed as percent of the total neutrophil content estimated in sonicated neutrophils (107/ml). was

,

,

,

.

Blood was obtained from normal donors after informed consent. Neutrophils were isolated by an LPS-free plasma-Percoll method described previously [15]. The cell preparation thus obtained consisted of 95 % neutrophils that were 99 % viable based on trypan blue exclusion. Neutrophils isolated by this procedure showed less spontaneous shape change, lower baseline secretion of superoxide anion and granule enzymes, and a greater chemotactic response than did cells prepared by FicollHypaque methods [15].

Secretion

Assays

Neutrophils were divided into aliquots as required for each given experiment to allow for triplicates in each treatment group. The cells were centrifuged for 5 mm at 140 x g, resuspended at iO neutrophils/mi in KRP-DA (controls) or in KRP-DA containing LPS and preincubated at 37#{176}C.After 1 hr, stimulants were added at 100 nM concentration as indicated followed by another incubation period of 15 mm at 37#{176}C.In some experiments aliquots of neutrophils were preincubated for 15 mm in KRP-DA containing CB at 5 j.tg/ml followed by FMLP for 15 mm as indicated. Subsequently, the suspension

SDS Electrophoresis of Elastin Fragments Generated by Neutrophil Elastolytic Activity Elastin digests were obtained by incubation for 2 hr of tritiated, insoluble elastin with supernatant from LPSpretreated, FMLP-stimulated neutrophils, or with commercially obtained neutrophil elastase or porcine pancreatic elastase (Elastin Products Co. Pacific, MO), respectively. Sodium dodecyl sulfate polyacrylamide gels (SDS PAGE) were prepared by a modification of Laemli [27] as 4% polyacrylamide stacking gels on a linear 7.518% polyacrylamide gel slab (0.75 mm x 10 cm x 10 cm). The digests were applied and the gels were run at room temperature at 20 mA/gel. After electrophoresis the protein bands were fixed in 50% methanol at room temperature and treated with Enhance#{174}(Dupont, Boston, MA) for 30 mm. The gels were dried and autoradiography was performed using prefogged Kodak x-O mat AR film at -75#{176}C. Stimulant and LPS effects as well as their interactions were examined by analysis of variance. The level of significance was set at 1 %. ,

550

Fittschen

et al Time

RESULTS Elastolysis Neutrophil

by Enzyme(s) Supernate

Released

Into the

Neutrophils preincubated in jig LPS/mI and subsequently stimulated with 100 nM FMLP released considerable amounts of elastolytic activity without a requirement for cytochalasin B (CB). The supernate of neutrophils treated in this manner, as well as commercial elastase purified from human neutrophils and from porcine pancreas, degraded 3H-elastin from bovine ligamentum nuchae. Incubation of these enzyme preparations with 3H-elastin for 2 hr yielded elastin degradation products which by SDS PAGE electrophoresis were estimated to have a MW of 35,000. Undigested elastin did not enter the gel. Elastin degradation was inhibited by the plasma proteinase-inhibitor alpha1 P1 (95 % inhibition) (data not shown), indicating that elastolysis was due to serine proteinase activity, and by the chloromethylketone AAPVCK, a specific serine elastase inhibitor with high affinity for neutrophil elastase (90% inhibition).

Course

of Neutrophil

Secretion

Neutrophil secretion of elastase, of a primary granule constituent myeloperoxidase (MPO) and the secondary granule marker Vit B12 BP as well as release of the cytoplasmic enzyme lactic dehydrogenase (LDH) was followed for 2 hr. During this incubation period, release of LDH, an indicator of cell injury, was less than 3 % in buffer controls and in cells treated with FMLP at 100 nM or LPS at 1 jg/ml alone, and did not exceed 4% in aliquots treated both with LPS and FMLP (Fig. 1). Incubation of neutrophils in buffer, or in LPS alone at 1 ag/mi (from E. coli, strain K235/ml) or 100 nM FMLP alone induced only small crements in MPO release (Fig. 1). But, compared to buffer controls, treatment of neutrophils with LPS or FMLP resulted in a 2-3-fold increase in the secreted elastase activity and in a 5-7-fold increment in release of Vit B12 BP. By contrast, preincubation in lug LPS/m1 followed by FMLP at 100 nM induced increases in the secretion of MPO, elastase, and Vit B12 BP that were approximately 10-fold, 20-fold and 8-fold respectively (Fig. 1). The magnitude of the increase suggested that secretion of elastase and MPO in response to the combined treatment was more than additive. Neutrophil Secretion Stimulated by Various Concentrations of LPS and FMLP

10000

FMLP LDH 10

upo

6-A

.o”---o-

Dubs. 7

I

vit

-----a 5000

8I2

5

3

2000 1000

g g

2 15

a

LPS

10

7 5

3

0

20

40

60

gO

120

MINUTES Fig. 1. Time course of elastase, MPO, and Vit B12 BP secretion by neutrophils incubated in buffer, FMLP, LPS and LPS combined with FMLP. Prior to stimulation with FMLP at 100 nM neutrophils had been preincubated for 1 hr in buffer (controls) or LPS K235 at 1 g/ml. Each point represents the mean ± SE of three experiments performed in triplicate.

Subsequent experiments compared the interactive effects on neutrophil secretion of four concentrations of FMLP, chosen because of their secretory and/or chemotactic activity, with four concentrations of LPS, which previously had been shown to have an enhancing effect on FMLP-induced endothelial cell injury [43]. Incubation of neutrophils with either LPS for 75 mm or buffer for 60 mm followed by FMLP for 15 mm resulted in small increases in secretion of elastase and MPO compared to buffer controls (Figs. 2, 3) whereas preincubation of neutrophils for 1 hr in low concentrations of LPS followed by FMLP for 15 mm caused marked secretion of elastase, MPO and enhanced secretion of Vit B12 BP (Figs. 2-4). Secretion of elastase activity (Fig. 2) appeared more responsive to the combined treatment with LPS and FMLP than did MPO (Fig. 3) and was already measurable at 1 ng LPS/10 nM FMLP. Both elastase and MPO secretion clearly represented more than additive effects of LPS and FMLP alone at concentrations of 100 ng LPS/10 nM FMLP and above (Figs. 2, 3). Secretion of the specific granule marker Vit B12 BP in response to FMLP or LPS alone, as well as to both stimuli combined, was dose dependent (Fig. 4). The pattern of Vit B12 BP secretion differed from that of elastase and MPO secretion by its responsiveness to LPS alone, by its maximal response at 10 nM FMLP combined with 100 or 1,000 ng LPS/m1, and by its additive

LPS Enhances

Neutrophil

Elastase

Secretion

551

40000j‘7.

‘V.

5)Jg CB/mi

l000ng

5jig

LPS/ml

CB/ml

/. 1500

I 01-

/.

-‘1000ng

0.

1 t100

V

10000

0 C 0

A 0 A Cl) A U)

w

0

-#{248}-lOOngLPS/mI

5) 0 5) C,)

5

0

0.

B

2

5000

A U) A

I

0

0

/1.

B

2000

I 0.1

1 FMLP

-10ng ..m.lng ‘0ng

I

LPS/mI LPS/mI

I-I------I-a 0

10 LPS

1

10

100

(nM)

2

ido

1ooII

LPS/ml LPS/ml LPS/ml

0. FMLP

ing Ong

1’O (nM)

LPS/mI

C

100

1000

(ng/ml)

1000

Fig. 3. Myeloperoxidase secretion. Interactive effects of neutrophil exposure to LPS (K235) and stimulation with FMLP on Fig. 2. Elastase secretion. Interactive effects of neutrophil neutrophil MPO secretion (upper panel). Notice the comparaexposure to LPS (K235) and stimulation with FMLP on neutrotively lower response to the combination of LPS and FMLP than to the combination of CB (5 g/ml) and FMLP at 100 nM FMLP. phil elastase secretion (upper panel). Notice the comparatively lower response to the combination of LPS and FMLP than to B indicates MPO release In concurrent buffer controls. The the combination of CB (5 g/ml) and FMLP at 100 nM FMLP. B bottom panel shows the neutrophil response to LPS alone. indicates elastase release In concurrent buffer controls. The Each point represents the mean ± SE of three experiments performed In duplicate. Asterisk denotes significant differbottom panel shows the neutrophil response to LPS alone. Each point represents the mean ± SE of three experiments ences (P 0.01) of combined treatment from both LPS conperformed in triplicate. Asterisk denotes significant differences trols (no FMLP) and FMLP controls (no LPS). (P 0.01) of combined treatment from both LPS controls (no FMLP) and FMLP controls (no LPS). LPS (ngltnl)

nature when LPS and FMLP were used as combined stimuli. Figure 4 also shows that the enhancement of Vit B12 BP secretion by most LPS concentrations followed by FMLP exceeded the enhancement effect of 5 jg CB/ ml followed by 10 nM FMLP. By contrast, CB/FMLP was more active than LPS/FMLP on MPO and elastase release (Fig. 2, 3). Comparison Enhanced

of Different Stimulants on LPSNeutrophil Secretory Activity

In order to study the effect on neutrophil granule secretion of different chemoattractants/stimulants, presumed to be released in an inflammatory field, several concentrations of FMLP, C5a, C5a des arg, LTB4, and PAF were compared for secretagogue activity on neutrophils preincubated for 1 hr in the presence and absence of 100 ng LPS K235/ml. In the absence of LPS the stimulants FMLP, C5a, LTB4 and PAF induced limited release of elastolytic activity and only at the highest (100 nM)

concentration tested (Fig. 5). Desarginated C5a had no effect, consistent with previous observation of its lower biologic activity compared to unmodified C5a [48]. Treatment of neutrophils with 100 ng LPS K 235/mi prior to exposure to peptide stimulants enhanced release of elastolytic activity. This effect was additive at low concentrations (0. 1-1.0 nM for FMLP and C5a, and 0.1-10 nM for CSa des arg) but interactive (more than additive) at higher concentrations. The LPS K235 moderately enhanced the action of the lipid stimulants LTB4 and PAF in an additive manner. In general, stimulant-induced secretion of MPO in the presence and absence of LPS followed patterns similar to that observed for secretion of elastolytic activity (Fig. 6). However, based on the increase over buffer controls, the LPS-induced enhancement of MPO secretion was less extensive than that observed for elastolytic activity. A significant secretion of Vit B12 BP was observed in response to FMLP, C5a, and C5a des arg alone, but not to LTB4 and PAF at the

552

Fittschen

et al

40

30.000

.1O00ng

LPS/mI

-100ng

LPS/ml

10.000

30

-‘*-lOng

LPS/ml

..m-lng

LPS/mI

m-0ng

LPS/ml

9000

20

CS.

C

+

0

LPS

7000

a) 0

a)

CB/ml

10

0. 0

B

Co

0.

0 >

5000

C 0

. 0.1

15

1

10

FMLP

(nM)

A

100

C)

5)

3000

Cl) A

10

U,

1000

LPS

a

5 0

DA

U, B1

w 1

10

LPS

100

I

I

I

.1

1

10

_I

100

I

I

I

.1

1

10

I

100

I

I

I

.1

1

10

100

1000

(ng/ml)

Fig. 4. Vitamin B12-blnding protein secretion. Interactive effects of neutrophil exposure to LPS (K235) and stimulation with FMLP on neutrophil Vit B12 BP secretion (upper panel). Notice the comparatively greater response to the combination of LPS and FMLP than the combination of CB (5 ,g/ml) and FMLP at 100 nM FMLP. B indicates Vit B12 BP release in concurrent buffer controls. The bottom panel shows the neutrophil response to LPS alone. Each point represents the mean ± SE of three experiments performed in duplicate. The asterisk mdicates minimal LPS requirements at a given FMLP concentration for enhanced Vit B12 BP secretion (P0.01).

concentrations used (Fig. 7). In contrast, with LPS followed by stimulant exposure B12 BP secretion to all stimuli.

PAF + LPS

3000

LTB4+ .ZAF

1000

LPSI B! _1

1

Stimulant

10

100

(nM)

Fig. 5. Elastase secretion in response to 100 nM FMLP, C5a, C5a des arg, LTB4 and PAF, respectively, by neutrophils preincubated for 1 hr In buffer or LPS (K235) at 100 ng/ml. B mdicates elastase release in concurrent buffer controls. Each point represents the mean ± SE of three experiments peformed in triplicate. Asterisk denotes significant differences (P0.01) in the response of cells treated with LPS prior to FMLP from those exposed to FMLP alone.

preincubation resulted in Vit

The Enhancement of Neutrophil Secretion by LPS Containing 0-Antigen Polysaccharide and by 0-Antigen-Free LPS While previous studies have shown that most of the biologic activities of LPS are located in the lipid A moiety [3 1 ,36] evidence suggests that certain LPS effects may also be mediated by the polysaccharide side chain which consists of the core polysaccharide and 0-antigen [29,42]. Using two types of LPS which contain a lipid A portion but differ in their side chain (S. minnesota Re 595 lacks most of the core polysaccharide and all of the 0-antigen whereas E. coli 0111 :B4 contains both) [36], we studied the significance of the 0-antigen for the induction of neutrophil secretion. Figure 8 illustrates that the enhancement of FMLP-induced elastase secretion was similar for both types of LPS, indicating that the 0-

antigen was not required for this effect. We also observed no difference for both types of LPS in the augmentation of MPO and Vit B12 BP secretion (data not shown). Importantly, similar to the previous experiment, augmentation of FMLP-induced Vit B12 BP secretion both by Re 595 and by Olll:B4 exceeded that of5 jg CB/m1.

DISCUSSION Neutrophil exposure to a wide variety of particulate and soluble stimuli causes secretion of neutrophil granule constituents [10,18,19,47,50]. Upon stimulation with chemoattractants, neutrophils release predominantly specific granules as reported previously [3, 10,51] and also shown in the present study. A comparison of equimolar concentrations of several chemotactic factors in the ab-

LPS Enhances

Neutrophil

Elastase

.

Secretion

553

A

40

* 4

3

C 0

DA

A C) A Cl)

0 0.

rTI

A U,

1r

10

a

-

LPSI

A A B,

B2

0. _1

I

I

1

10

I

100

I

.1

LTB4+LPS

I

1

PAF

I

I

I

I

I

I

10

100

.1

1

10

100

1

‘.4

I

I

I

_1

1

10

I

100

I .

I

I

1

10

I

100

>

LPS

1.5 20 1 0.5

LPS

r44

15

B, I

I

I

.1

1

10

J

100

I

I

I

I

.1

1

10

100

Stimulant

(nM)

Fig. 6. MPO secretion In response to 100 nM FMLP, C5a, C5a des arg, LTB4 and PAF, respectively, by neutrophils preincubated for 1 hr in buffer or LPS (K235) at 100 ng/ml. B indicates MPO release in concurrent buffer controls. Each point represents the mean ± SE of three experiments performed In duplicate. Asterisk denotes significant differences (PO.O1) in the response of cells treated with LPS prior to FMLP from those exposed to FMLP alone.

sence of cytochalasin B (CB) showed that FMLP and C5a were the most effective inducers of specific granule release, as measured by Vit B12 BP, a selective marker of this granule type. C5a des arg and PAF had intermediate, whereas LTB4 had little activity. Neutrophils stimulated with i0 M FMLP and CSa but not those exposed to i0 M C5a des arg, LTB4 or PAF also secreted small amounts of azurophil granule content, even in the absence of CB. Despite the limited quantity secreted, this observation is nonetheless intriguing because azurophil granules contain neutral proteinases that may play a role in neutrophil extravasation. Previous studies by Russo et al. [39] suggested that neutrophils penetrate basement membranes by proteolytic digestion. In this regard, neutrophil elastase, a proteinase stored in azurophil granules [37] may be important due to its ability to digest basement membrane (Type IV) collagen [32] . Clark et al. [4] have localized elastase cytochemically also in the neutrophil synthetic apparatus

10

LPS

PAF LTB4

5 Bt

.1

1

10

100 -00

Stimulant

(nM)

Fig. 7. Vit B12 BP secretion in response to 100 nM FMLP, C5a, and C5a des arg, LTB4 and PAF, respectively, by neutrophils premncubated for 1 hr In buffer or in LPS (K235) at 100 ng/ml. B indicates Vit B12 BP release in concurrent buffer controls. Each point represents the mean ± SE of three experiments performed In duplicate. Asterisk denotes significant differences (PO.O1) in the response of cells treated with LPS prior to FMLP from those exposed to FMLP alone.

indicating that this enzyme could, in addition, be secreted from this non-granular pool. Recent work by Sandhaus and Henson [41] suggests that elastase may, in fact, be available to neutrophils during diapedesis since neutrophils migrating in a chemotactic gradient through elastincoated filters do solubilize some of that elastin. This observation suggests that migrating neutrophils can cxpress elastolytic activity. Most studies of neutrophil secretion in response to chemoattractants have used pretreatment with cytochalasin B in order to enhance neutrophil response [20,52]. However, the fact that neutrophils normally do not encounter CB in vivo complicates the interpretation of these experiments. Recent studies from our laboratory have shown that pretreatment of neutrophils with bacterial lipopolysaccharide (LPS) enhances granule release [15], an observation of considerable importance because neutrophils encounter LPS in Gram-negative infections. Neutrophils that have been exposed to LPS prior to

554

Fittschen

et al

edly enhanced Vit B12 BP secretion induced by FMLP and, to a lesser extent, that induced by other chemotactic 2600I #{149}-CB+FMLP factors. The overall effect of high concentrations of LPS and stimuli combined, caused release of between 15 and -.-Re595+FMLP 45 % of the total neutrophil Vit B12 BP content. Hence, 12000 ..-01 1 1:B4+FMLP as has been emphasized in other circumstances where 0. 0 neutrophil secretion has been studied [51], the release of 10000 specific granules was generally more extensive than the release of azurophil granules (MPO secretion consis800 tently less than 10%). Cl) 6000 The enhancement effect of LPS on chemotactic factorinduced neutrophil granule secretion resembled in some 4000 w respects that of cytochalasin B, a fungal metabolite that FMLP interferes with neutrophil microfilaments [20, 22,52]. 2000 However, the present studies suggest that LPS acts B ....-Re595 ===I=.!! ‘0111:B4 through a mechanism that differs from CB, since CB I 10 100 1000 proved to more effectively augment the secretion of elasng LPS/ml tase and MPO, while being less effective than LPS in enhancing secretion of Vit B12 BP at the same stimulant Fig. 8. Enhancement of FMLP-induced neutrophil elastase se- concentration. A difference in the mechanism of action cretion by LPS 0111 :B4 and LPS Re595, respectively. Neutroof both agents is further supported by observations showphlls were Incubated for 1 hr in buffer or LPS prior to stimulation ing that LPS induces an irregular shape in neutrophils with 100 nM FMLP. B indicates elastase release In concurrent [15], whereas CB causes them to assume a rounded buffer controls. Note that at the same FMLP concentration LPS shape [22]. Induced less secretion of elastase than did CB (5 g/ml). Each Most of the biological effects of the LPS molecule point represents the mean ± SE of three experiments performed In triplicate. No difference of LPS effects were noted have been associated with its lipid A moiety rather than between Re595 and O111:B4 (PO.O5). with its core or the 0-antigen-polysaccharide portions [3 1 36]. In order to examine a possible contribution of chemoattractants can injure endothelial cells [43] . Since the 0-antigen to the secretory enhancement in neutrophils one of the mechanisms by which neutrophils cause tissue we compared the activity of three types of LPS from injury may involve elastase [43], a broad spectrum prodifferent bacterial strains (E. coli K235 and 0111 :B4, tease that has been implicated in the pathogenesis of a and S. Minnesota Re595) one of which (Re595) lacked variety of inflammatory disorder [ 1 16, 26,30,34] we the 0-antigen and most of the core polysaccharide [31, examined LPS effects on neutrophil elastase secretion 36]. Our results, showing a similar enhancement of neuand found that each of three different LPS species marktrophil secretion for each of the three types of LPS, edly augmented chemoattractant-induced elastase secresuggest that this effect is not mediated by the 0-antigen, tion. but by the lipid A molecule or by components of the core polysaccharide. Neutrophil secretory responses to chemotactic factors The secretion of granule components in response to were markedly enhanced by preincubation in LPS derived from the E. coli strain K235. While in itself an chemotactic stimuli in the presence and absence of LPS ineffective secretagogue for elastase and MPO release, may have physiological and pathological implications. release of elastase in response to LPS K235 augmented the secretion of both enzymes in On one hand, limited response to FMLP, C5a, and CSa des arg in a dosechemotactic factors alone may be instrumental in the dependent manner. LPS appeared to enhance more effecpenetration of tissue barriers by neutrophils, e.g. during tively the secretagogue activity of the peptide stimulants vascular egress or during tissue migration [39], as elasFMLP, CS and CSa des arg, than that of the lipids LTB4 tase is capable of degrading structural components of and PAF. Thus, in the presence of LPS, high concentravessel walls, such as elastin [24], and Types Ill and IV tions of peptide stimuli caused a more than additive collagen [9, 32] . On the other hand, release of larger effect, whereas the effect of lipid stimuli was only amounts of elastase e.g. in endotoxemia [1, 30, 34] could additive. result in vascular injury. Such a possibility is supported While LPS K235 alone failed to induce release of by findings of Smedly et al. [43] of endothelial injury by elastase or MPO, it did prove to be a selective secretneutrophils stimulated with FMLP after preincubation in agogue for specific granules as measured by release of LPS. Vit B12 BP. Lipopolysaccharide K235 in addition markIn addition to the enhancement of neutrophil secretory 2700

C 0

A 0 A

A U) A U) A

,

,

,

,

LPS Enhances activity, LPS also augments the generation of toxic oxygen species [5, 6, 12] which may also contribute to vascular injury [7, 40, 44, 45]. While part of this damage may be due to formation of H202 or one of its metabolites [40, 45], the enhanced release of MPO observed in the present study may also allow the MPO-catalysed formation of hypohalous acid, which may enhance chronic tissue injury [25]. The present study has demonstrated in vitro a remarkable modulating effect by LPS on neutrophil secretion in response to several chemotactic stimuli. The in vivo requirements for similar effects are presently uncertain; however, recent work by Worthen et al. [49] shows that LPS may in vivo be an even more effective neutrophil modulant than in vitro, since pulmonary vascular injury can be elicited by the injection of considerably lower quantities of LPS (i.e. nanograms) than used in the present study, when combined with chemotactic factors. ,

ACKNOWLEDGMENTS We gratefully acknowledge the excellent secretarial assistance by Ms. Lydia Titus and statistical evaluation of the data by John LaBrecque. Supported by NIH Grant No. 24834 work done in the F.L. Bryant, Jr. Research Laboratory for the Mechanisms of Lung Disease, Department of Pediatrics. ,

REFERENCES 1 . Aasen, A.O. , and Ohlsson, K. Release of granulocyte elastase in lethal canine endotoxin shock. Hoppe-Seyler’s Z. Physiol. Chem. 359,683, 1978. 2. Banda, Mi. , and Werb, Z. Mouse macrophage elastase. Purification and characterization as a metalloproteinase. Biochem. J. 193,589, 1981. 3. Bentwood, B.J. and Henson, P.M. The sequential release of granule constituents from human neutrophils. J. Immunol. 124,855, 1980. 4. Clark, J.M., Vaughan, D.Q. , Aiken, B.M.,and Kagan, H.M. Elastase-like enzymes in human neutrophils localized by ultrastructural cytochemistiy. J. Cell Biol. 84, 102, 1980. 5. Dahinden, C. , Galanos, C. , and Fehr, J. Granulocyte activation by endotoxin. I. Correlation between adherence and other granulocyte functions, and role of endotoxin structure on biologic activity. J. Immunol. 130,857, 1983. 6. Dahinden, C. , and Fehr, J. Granulocyte activation by endotoxin. II. Role of granulocyte adherence, aggregation, and effect of cytochalasin B, and comparison with formylated peptide-induced stimulation. J. Immunol. 130,863, 1983. 7. Fantone, J.C. , and Ward, PA. Role of oxygen-derived free radicals and metabolites in leukocyte-dependent inflammatory reactions. Am. J. Pathol. 107,395, 1982. 8. Fernandez, H.N. , and Hugh, T.E. Partial characterization of human C5a anaphylotoxin: I. Chemical description of the carbohydrate and polypeptide portion of human C5a. J. Immunol. 117,1688, 1976. 9. Gadek, J.E., Fells, GA., Wright, D.G., and Crystal, R.G. Human neutrophil elastase functions as a Type III collagen “col-

Neutrophil

Elastase

Secretion

555

lagenase” . Biochem. Biophys. Res. Commun. 95, 1815, 1980. 10. Gallin, J.T. Neutrophil specific granules: A fuse that ignites the inflammatory response. Clin. Res. 32,320, 1984. 11 . Gottlieb, C. , Lau, KS. , Wasserman, L.R. , and Herbert, V. Rapid charcoal assay for intrinsic factor (IF), gastric juice unsaturated B,2-binding capacity, antibody to IF, and serum unsaturated B12-binding capacity. Blood 25,875, 1965. 12. Guthrie, L.A. , McPhail, L.C. , Henson, P.M. , and Johnston, RB. Jr. Priming of neutrophils for enhanced release of oxygen metabolites by bacterial lipopolysaccharide. Evidence for increased activity of the superoxide-producing enzyme. J. Exp. Med. 160,1656, 1984. 13. Harlan, J. M. , Killen, PD. , Harker, L.A. , and Striker, G.E. Neutrophil-mediated endothelial injury in vitro. Mechanisms of cell detachment. J. Clin. Invest. 68, 1394, 1981. 14. Harlan, J.M. , Schwartz, BR. , Reidy, MA. , Schwartz, SM., Ochs, H.D. , and Harker, L.A. Activated neutrophils disrupt endothelial monolayer integrity by an oxygen radical-independent mechanism. Lab. Invest. 52, 141 , 1985. 15. Haslett, C. , Guthrie, L.A. , Kopaniak, MM. , Johnston, RB. Jr., and Henson, P.M. Modulation of multiple neutrophil functions by preparative methods or trace concentrations of bacterial lipopolysaccharide. Am. J. Pathol. 1 19, 101 , 1985. 16. Hawkins, D. , and Cochrane, C.G. Glomerular basement membrane damage in immunological glomerulonephritis. Immunology 14,665, 1968. 17. Hed, J. , and Stendahl, 0. Differences in the ingestion mechanisms of IgG and C3b particles in phagocytosis by neutrophils. Immunology 45,727, 1982. 18. Henson, P.M. Interaction of cells with immune complexes: Adherence, release of constituents, and tissue injury. J. Exp. Med. 134,ll4s, 1971. 19. Henson, P.M. The immunological release of constituents from neutrophil leukocytes. II. Mechanisms of release during phagocytosis,and adherence to nonphagocytosable surfaces. J. Immunol. 107,1547, 1971 20. Henson, P.M. , and Oades, Z.G. Enhancement of immunologically induced granule exocytosis from neutrophils by cytochalasin B. J. Immunol. 110,290, 1973. 21 . Henson, P.M. , Zanolari, B. Schwartzman, NA. , and Hong, SR. Intracellular control of human neutrophil secretion. I. C5ainduced stimulus-specific desensitization and the effects of cytochalasin B. J. Immunol. 121,851, 1978. 22. Hoffstein, S. , Goldstein, I.M. , and Weissmann, G. Role of microtubule assembly in lysosomal enzyme secretion from human polymorphonuclear leukocytes. A reevaluation. J. Cell Biol. 73,242, 23.

24.

25.

26.

27.

28. 29.

1977.

JanoffA., and Blondin, J. Further studies on an esterase inhibitor in human leukocyte cytosol. Lab. Invest. 25,565, 1971. Janoff A. , White, R. , Carp, H. , Harel, S. , Dearing, R. , and Lee, D. Lung injury induced by leukocytic proteases. Am. J. Pathol. 97,111, 1979. Johnson, K.J. , Fantone, J.C. HI, Kaplan, J. , and Ward, PA. In vivo damage of rat lungs by oxygen metabolites. J. Clin. Invest. 67,983, 1981. Kniker, WI. , and Cochrane, C.G. Pathogenic factors in vascular lesions of experimental serum sickness. J. Exp. Med. 122, 83,1965. Laemli, U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227,680, 1970. Larsen, G.L. , and Henson, P.M. Mediators of inflammation. Ann. Rev. Immunol. 1,335, 1983. Lebbar, S. , Cavaillon, J-M. Caroff, M. , Ledar, A. , Borade, H., Sarfati, R. , and Haeffner-Cavaillon, N. Molecular requirement for interleukin 1 induction by lipopolysaccharide-stimulated hu-

556

Fittschen

et aI

man monocytes: involvement of the heptosyl-2-keto 3-deoxyoctulosonate region. Eur. Immunol. 16,87, 1986. 30. Lee, CT. , Fein, AM. , Lippmann, M. , Holtzman, H. , Kimbel, P. , and Weinbaum, G. Elastolytic activity in pulmonary lavage fluid from patients with Adult Respiratory Distress Syndrome. N. Eng. J. Med. 304,192, 1981. 31. Luederitz, 0., Galanos, C., Lehmann, V. , Mayer. H., Rietschel, E.T. , and Weckessec, J. Chemical structure and biological activities of Lipid A’s from various bacterial families. Naturwissenschaften 65,578, 1978. 32. Mainardi, CL. , Dixit, SN. , and Kang, A.H. Degradation of Type IV (basement membrane) collagen by a proteinase isolated from human polymorphonuclear leukocyte granules. J. Biol. Chem. 255,5435, 1980. 33. Mclntire, F.C. , Barlow, G.H. , Sievert, H.W. , Finley, R.A. , and Yoo, AL. Studies on a lipopolysaccharide from E. coil. Heterogeneity and mechanism of reversible inactivation by sodium deoxycholate. Biochemistry 8,4063, 1969. 34. McGuire, W.W. , Spragg, R.C. , Cohen, A.B. . and Cochrane, C.G. Studies on the pathogenesis of the Adult Respiratory Distress Syndrome. J. Clin. Invest. 69,543, 1982. 35. Morrison, D.C. , and Kline, L.F. Activation of the classical and properdin pathways of complement by bacterial lipopolysaccharides (LPS). J. Immunol. I 18,362, 1977. 36. Morrison, D.C. , and Ulevitch. R.J. The effects of bacterial endotoxins on host mediation systems. Am. J. Pathol. 93,526, 1978. 37. Ohlsson, K. , Olsson, I. , and Spitznagel, J.K. Localization of chymotrypsin-like cationic protein, collagenase, and elastase in azurophil granules of human neutrophilic polymorphonuclear leukocytes. Hoppe-Seyler’s Z. Physiol. Chem. 358,361, 1977. 38. Root, R.K. , and Cohen, MS. The microbicidal mechanisms of human neutrophils and eosinophils. Rev. Infect. Dis. 3,565, 1981. 39. Russo, R.G. , Liotta, L.A. , Thorgeirsson, U. , Brundage, R. , and Schiffman, E. Polymorphonuclear leukocyte migration through human amnion membrane. J. Cell Biol. 91,459, 1981. 40. Sacks, T. , Moldow, CF. , Craddock, P.R. , Bowers, T.K. , and Jacob, H.S. Oxygen radicals mediate endothelial cell damage by complement-stimulated granulocytes. An in vitro model of immune vascular damage. J. Clin. Invest. 62, 1161, 1978. 41 . Sandhaus, R.A. , and Henson, P.M. Elastin degradation is re-

quired for directed migration of human neutrophils through an elastin-rich barrier in vitro. In press, 1986. Scibienski, R.J. Immunologic properties of protein-lipopolysaccharide complexes-Ill. Role of carbohydrate in the LPS adjuvant effect. Mol. Immunol. 17, 21, 1980. Smedly, L.A. , Tonnesen, MG. , Sandhaus, R.A. , Haslen, C., Guthrie, L.A. , Johnston, RB. Jr. , Henson, P.M. , and Worthen, G.S. Neutrophil-mediated injury to endothelial cells: Enhancement by endotoxin and essential role of neutrophil elastase. J. Clin. Invest. 77, 1233, 1986. Tate, R.M. , and Repine, i.E. Hydrogen peroxide causes permeability edema and hypertension in isolated salt-perfused rabbit lungs. Chest 5,48s, 1983. Weiss, S.J. , Young, J. , LoBuglio, A.F. , and Slivka, A. Role of hydrogen peroxide in neutrophil-mediated destruction of cultured endothelial cells. J. Clin. Invest. 68,714, 1981. Weissmann, 0. , Korchak, H.M. , Perez, H.D. , Smolen, J.E., Goldstein, J.M. , and Hoffstein, ST. The secretory code of the neutrophil. J. Reticuloendothel. Soc. 26, 687(Suppl), 1979. Weissmann G. , Zurier, RB. , and Hoffstein, S. Leukocytic pro-

42.

43.

44.

45.

46.

47.

teases

48.

49.

Synergistic

50.

51

52.

and

the

immunologic

release

of

lysosomal

enzymes.

Am.

J. Pathol. 68,539, 1972. Webster, R.O. , Hong, SR. , Johnston, RB. Jr. , and Henson, P.M. Biological effects of the human complement fragments C5a and C5a des arg on neutrophil function. Immunopharmacology 2,201, 1980. Worthen, G.S. , Haslett, C. , Rees, A.J. , Gumbay, R.S. Henson, J.E. , and Henson, P.M. Neutrophil-mediated pulmonary injury.

.

effect

of trace

amonts

of lipopolysaccharide

and

neu-

trophil stimuli on vascular permeability and neutrophil sequestration in the lung. Am. Rev. Respir. Dis. 136, 19-28, 1987. Wright, D.G. , Bralove, D.A. , and Gallin, J.I. The differential mobilization of human neutrophil granules. Effects of phorbol myristate acetate and ionophore A23187. Am. J. Pathol. 87,273, 1977. Wright, D.G. , and Gallin, J.T. Secretory responses of human neutrophils: Exocytosis of specific (secondary) granules by human neutrophils during adherence in vitro and during exudation in vivo. J. Immunol. 123,285, 1979. Zurier, RB. , Hoffstein, S. , and Weissmann, G. Cytochalasin B: Effect on lysosomal enzyme release from human leukocytes. Proc. Nat. Acad. Sci. USA. 70,844, 1973.