Neutrophil elastase increases secretory leukocyte protease

inhibitor transcript levels in airway epithelial cells

JOAN M. ABBINANTE-NISSEN, LEIGH G. SIMPSON, AND GEORGE D. LEIKAUF
Pulmonary Cell Biology L&oratory, Departments of Environmental Health, Physiology/Biophysics,
and Medicine, University of Cincinnati Medical Center, Cincinnati, Ohio 45267-0056

Abbinante-Nissen, Joan M., Leigh G. Simpson, and Gene regulation of an antiprotease by a proteolytic
George D. Leikauf. Neutrophil elastaseincreasessecretory enzyme has been examined previously, wherein neutro-
leukocyte proteaseinhibitor transcript levels in airway epithe- phi1 elastase produced a dose- and time-dependent in-
lial cells.Am. J. Physiol. 265 (Lung Cell. Mol. Physiol. 9): L286- crease in transcript levels and rate of synthesis of one of
LX& 1993.-Airway inflammation is often associatedwith the its own inhibitors, al-AT (23). Because neutrophils re-
infiltration of activated neutrophils and subsequentprotease
lease significant quantities of protease in the airway,
release.Although aiding in the digestion and phagocytosis of
foreign proteins and microorganisms,neutrophil proteasescan SLPI transcript levels also may be responsive to local
indiscriminately damagehealthy lung tissue.In the conducting changes in neutrophil protease concentrations. To ex-
airway, proteases,particularly neutrophil elastase,are counter- amine whether neutrophil proteases can increase SLPI
balancedby severalantiproteases,including secretory leukocyte synthesis, SLPI transcript levels were measured in hu-
proteaseinhibitor (SLPI). SLPI can be produced locally by a man airway epithelial cells treated with various neutro-
number of cells including the airway epithelial cell. To examine phi1 and nonneutrophil proteases.
the effects of neutrophil granule components on SLPI tran-
script levels, airway epithelial cells were treated (up to 96 h) METHODS
with elastase,other proteases,or enzymesisolatedfrom human
sputum. We found that neutrophil elastaseincreased SLPI Experimental design. Airway epithelial cells at confluency
transcript levels in primary and transformed human airway were washedthree times with Dulbecco’sphosphate-buffered
epithelial cells in a time- and dose-dependentmanner. Other saline (PBS) and placed in MCDB-153 containing 110 PM cal-
neutrophil products, suchascathepsin G, myeloperoxidase,and cium and antibiotics for 36 h, then cells were incubated in
lysozyme, had little or no effect on SLPI transcript levels.How- control or enzyme-containing medium. Two transformed air-
ever, two nonneutrophil proteases, trypsin and pancreatic way cell lines, SHTEo- and BEAS-2B, were comparedwith
elastase,also increasedSLPI transcript levels at higher doses primary humanairway epithelial cellsderived from three donors
than that required of neutrophil elastase.Two inflammatory for SLPI transcript levelsin the presenceand absenceof human
cytokines, tumor necrosisfactor-a and interleukin-8, produced neutrophil elastase (derived from human sputum). The
little or no effect on SLPI transcript levels. This study demon- SHTEo- cell line respondedsimilarly to human primary cells
strates one way in which SLPI is regulated,via a proteasethat and was therefore used in subsequentexperiments. To deter-
it inhibits, neutrophil elastase. mine time to maximal accumulation of SLPI transcript, cells
were continuously exposedto 2 U/ml elastaseand total RNA
antileukoprotease;antiprotease; asthma; bronchiectasis; bron- harvested from 8 to 96 h. In the remaining experiments, RNA
chitis; inflammation; serineprotease was harvested 48 h after the initiation of treatment. The dura-
tion of exposureto elastasenecessaryto alter SLPI transcript
AIRWAY INFLAMMATION, present in asthma, bronchitis, levelswasdeterminedby incubating cellswith elastase(1,2, and
and bronchiectasis, is characterized by the presence of 10 U/ml) for 1, 8, and 48 h.
activated neutrophils in the epithelium. Neutrophil To examinethe specificity of the responseto elastase,several
products normally function to digest microorganisms additional tests were conducted with SHTEo- cells. To deter-
and necrotic tissue (27). The neutrophil’s armamentar- mine whether the effect of elastaseon SLPI transcript levels
ium is comprised of 40 components, localized to the could be due to proteolytic activity, cells were treated with
plasma membrane or to intracellular granules. Azuro- elastasein the presenceand absenceof a protease inhibitor,
philic (intracellular) granule components include neu- either rSLP1 or al-AT. Each inhibitor wasmixed and incubated
tral serine proteases, elastase and cathepsin G, as well as with elastasefor 15 min at room temperature in either a three-
fold molar excessor equimolar to elastasebefore addition to
other enzymes such as myeloperoxidase and lysozyme SHTEo- cells. To compare responsesto several neutrophil
(27). The concentrations of these enzymes can be ele- azurophilic granule components, SLPI transcript levels were
vated in diseased airways, with elastase, for example, measuredin cells treated with either elastase(0.1-30 U/ml),
ranging from 3 to 30 PM (18, 30). Elastase is counter- cathepsinG (0.1-30 U/ml), lysozyme (0.1-30 pg/ml), or myelop-
balanced by two major antiproteases, secretory leuko- eroxidase (0.1-30 pg/ml), each isolated from purulent human
cyte protease inhibitor (SLPI) and ai-antitrypsin (ai- sputum. This responsewas compared with that produced by
AT) (6). SLPI is a single nonglycosylated polypeptide other nonneutrophil proteases:bovine pancreatic trypsin (O.l-
chain of 107 amino acids (-11,700 Da) (9, 20, 29, 31) 100 U/ml) and porcine pancreatic elastase(0.1-100 U/ml).
with a boomerang-shaped tertiary structure comprised SLPI transcript levels were alsomeasuredfollowing treatment
of two domains, with the antiprotease active site for with tumor necrosisfactor-a! (TNF-(x) (0.1-100 rig/ml) or inter-
leukin-8 (IL-B) (0.01-100 rig/ml) for 72 h or lipopolysaccharide
elastase and trypsin residing in the COOH-terminal do- (10 or 100 pg/ml) derived from either Escherichiu coli or
main (4, 8). SLPI is produced locally in the airway ep- Pseudomonasaeruginosafor 48 h and comparedwith the re-
ithelium (3, 20, 32), unlike al-AT, which is predomi- sponseobservedwith elastase.To examinewhether the effect of
nantly produced by the liver and to a lesser extent by neutrophil elastasecould be attributed to possiblelipopolysac-
peripheral blood monocytes and alveolar macrophages charide contamination, polymyxin B (50 pg/ml) was preincu-
(15, 22). bated with elastasebefore treatment of cultured SHTEo- cells.
L286 1040-0605/93 $2.00 Copyright 0 1993 the American Physiological Society

Downloaded from journals.physiology.org/journal/ajplung (105.111.135.139) on August 7, 2020.

 ELASTASE INCREASES SLPI TRANSCRIPT LEVELS L287
Cell culture. Human airway epithelial cellswere grown by the sulfonic acid (MOPS), 8.3 mM sodium acetate, and 0.52 mM
method describedpreviously (13, 14). Briefly, normal human EDTA, size separated in a 2% agarosegel containing 3.7%
tracheobronchial tissue was obtained at autopsy (within 20 h formaldehyde, 20 mM MOPS, 16mM sodiumacetate and 1 mM
postmortem) or following surgery, from three donors. Tracheal EDTA, and membranetransferred via capillary action. For dot-
tissuewasobtained from donor 1 (37 yr) who smoked2-3 packs blot analysis, 1 pg of total RNA were denatured (60°C for 15
per day (ppd)/20 yr. Bronchus tissuewasobtained from donor 2 min) in 100 ~1 7.4% formaldehyde-7x standard saline citrate
(ageunknown), following a lung resection. Segmentalbronchus (SSC) solution (1X SSC = 150 mM NaCl and 15 mM sodium
tissuewasobtained from donor 3 (49 yr), who smoked2-3 ppd citrate), snapcooled(0°C) for 2 min, and loadedonto the nylon
for 30 yr and wasdiagnosedwith chronic obstructive pulmonary membrane.Each cDNA was [a-32P]dCTP labeled by random
disease.The tissue was dissectedinto l-cm2 piecesand main- primer extension (5) (average specific activity = 9.5 X lo8
tained in CMRL 1066 medium supplementedwith 1% fetal counts min-l pg cDNA-’ and 2.64 X lo6 counts.min-l *ml
l l

bovine serum, 0.8 PM insulin, 0.1 FM hydrocortisone, 0.3 PM hybridization buffer-l) and the blots hybridized according to
P-retinol acetate, 1 mM glutamine, and antibiotics (1.0 pg/ml the method of Church and Gilbert (2). Each blot wasexposedto
amphotericin B, 50 pg/ml gentamicin, 10 U/ml penicillin, and Kodak XAR-5 X-ray film for autoradiography.
10 pg/ml streptomycin) (50% 02-45%N2-5% Co,). Data analysis. Autoradiographs were analyzed with a scan-
After lo-14 days, eachpiece was further dissectedinto 0.25- ning densitometer(modelGS 300) and the total areaunder each
cm2 explants and transferred to coated dishes. Dishes were peak integrated by the associateddata system (model GS 350,
coated by incubation (2 h, 36.5”C) with 3 ml MCDB-153 me- Hoefer Scientific Instruments, San Francisco, CA) or the Jan-
dium containing collagen (Vitrogen 100, 30 pg/ml), fibronectin de1Video Analysis Software (JAVA 1.4,Jandel Scientific, Corte
(10 cLglml),and bovine serumalbumin (100 pg/ml). Excessso- Madera, CA). The Student’s t test of independentsampleswas
lution wasremoved and the dish washedoncewith PBS before usedto determinesignificance(P < 0.05) of the SLPI transcript
use.Cellsweregrown in MCDB-153 mediumcontaining 110 pM levels following normalization to @actin, and all data are pre-
calcium and supplementedwith 0.9 PM insulin, 0.2 PM hydro- sented as the mean t SE SLPI-to-&tin ratio. Blot analysis
cortisone, 0.8 nM epidermalgrowth factor, 0.1 PM transferrin, conductedwith 24 control and 24 elastase-treatedRNA samples
0.5 PM each phosphoethanolamine,and ethanolamine, 10 nM from SHTEo- cells showed no increasein p-actin transcript
3,3’,5-triiodo-L-thyronine, 30 hg/ml bovine pituitary extract, levels following elastasetreatment.
and antibiotics (3.5% CO, at 36.5”C). Once the epithelial cells Materials. SLPI cDNA probe and recombinant human secre-
had radiated to >5 mm, they were removed using 0.025% tory leukocyte proteaseinhibitor (rSLP1) were provided by R.
trypsin, 0.02% ethylene glycol-bis(P-aminoethyl ether)- C. Thompson (Synergen, Boulder CO). CMRL-1066, Dulbec-
N,N,N’,N’-tetraacetic acid (EGTA), and 1% polyvinylpyrroli- co’s PBS, amphotericin B (Fungizone), and gentamicin were
done in 25 mM N-2-hydroxyethylpiperazine-N’-2-ethane- from GIBCO/BRL, Grand Island, NY. DMEM, MCDB-153,
sulfonic acid (HEPES) -buffered saline solution. Cells were glutamine, penicillin/streptomycin, insulin, hydrocortisone,
seeded(10,000ce11s/cm2) on coated60-mm dishesand grown for P-retinol acetate, bovine serumalbumin (fraction V/fatty acid
5-13 days in supplementedMCDB-153. free), bovine pancreatic trypsin (Type XI, 6700 U/mg protein,
Transformed cell lines were obtained from two sourcesand mol wt 23,345), cu,-antitrypsin (humanplasma),lipopolysaccha-
maintained in our laboratory. Human tracheal epithelial cells ride, E. coli, serotype055:B5, P. aeruginosa, serotype 10 (Habs),
transformed with replication-origin defective simian virus 40, polymyxin B sulfate, Antifoam A, and Kodak XAR-5 film were
SHTEo-, were obtained from D. C. Gruenert (7), and human from Sigma Chemical, St. Louis, MO. Fibronectin, transferrin,
bronchial epithelial cells transformed with an adenovirus-12/ and epidermalgrowth factor were from Collaborative Research,
simian virus 40 hybrid virus, BEAS-2B, were obtained from C. Bedford, MA. Collagen(Vitrogen 100)wasfrom CollagenCorp.,
C. Harris and J. Lechner (25). The SHTEo- cells were grown Palo Alto, CA. Fetal bovine serum(Defined) wasobtained from
(passage126-150) on collagen-fibronectin-coatedculture flasks Biocell, Ranch0 Dominguez, CA. Human neutrophil elastase
containing Dulbecco’smodified Eagle’smedium (DMEM) sup- (875 U/mg protein, mol wt 27,000), cathepsin G (4,115 U/mg
plemented with 5% fetal bovine serum and antibiotics. The protein, mol wt 25,000),porcine pancreatic elastase(135U/mg
BEAS2B cells were grown (passage55-66) on coated culture protein, mol w-t25,500),myeloperoxidase(31 U/mg protein, mol
flasks containing supplementedMCDB-153 medium as above. wt 120,000),and lysozyme (80,000U/mg protein, mol wt 27,000)
Both cell lines were seededlO,OOO-50,000cells/cm2on coated were from Elastin Products, Owensville, MO. Interleukin-8 (re-
35-mm dishes. combinant, monocyte) and tumor necrosisfactor-a (recombi-
Isolation and measurement of steady-state mRNA. After ex- nant) were from Promega, Madison, WI. [a-32P]dCTP (3,000
posure, each culture was washed once with PBS. Cells lifted Ci/mmol) was from Du Pont-New England Nuclear, Boston,
from the culture dish during protease treatment were centri- MA. Randomprimer extension kit (hexamers)wasfrom Boeh-
fuged (150g for 10 min) and returned to the dish. Cellsremain- ringer Mannheim, Indianapolis, IN. Magnagraph nylon mem-
ing on the dish and nonadherent cells were lysed using 4 M brane was from Micron Separations,Westboro, MA.
guanidinium thiocyanate, 25 mM sodium citrate (pH 7), 0.5%
Antifoam A, and 100 mM 2-mercaptoethanol.Total RNA was RESULTS
isolatedby the method of Chomczynski and Sacchi (1) and the
RNA resuspended in 10 mM tris(hydroxymethyl)ami- Primary human and SHTEo- cells respond to elustase.
nomethane hydrochloride (pH 7.5), 1 mM EDTA, and 0.5% Primary human airway epithelial cells and SHTEo- cells
sodiumdodecyl sulfate. Total RNA yield wascalculatedby mea- possess basal levels of SLPI transcript as demonstrated
suringthe absorbanceat 260and 280nm (assumingthat an A260 by dot blot [Fig. 1: control (C)] and Northern analysis
of 1 = 40 pg RNA)(model DU46, Beckman Spectrophotometer, with constitutive levels being greater in the primary cells
Fullerton, CA). than in the SHTEo- cells (Fig. 2: larws 1 and 3). Elastase
The levels of steady-state SLPI and P-actin transcripts were
evaluated by Northern and dot-blot analysisusing the Magna- treatment (2 U/ml for 48 h) produced a similar increase in
graph nylon membraneaccording to the manufacturer’sproto- SLPI transcript in both the primary human and SHTEo-
col. For Northern analysis, 1.5 pg total RNA were denatured cells [Fig. 1: elastase (E) and Fig. 2: lanes 2 and 41. An-
(65°C for 15 min) in a buffer solution containing 8.4% formal- other transformed cell line, BEAS-BB, also possessed
dehyde,51.6% formamide, 10.4mM 3-(N-morpholino)propane- basal levels, but was relatively unresponsive to elastase

Downloaded from journals.physiology.org/journal/ajplung (105.111.135.139) on August 7, 2020.

L288 ELASTASE INCREASES SLPI TRANSCRIPT LEVELS

lowing 8 h continuous elastase (2 U/ml) treatment with

25
r maximal transcript accumulation observed following 72 h
of exposure (Fig. 3). SLPI transcript levels following 96 h
elastase treatment were less than that observed at 72 h,
but remained elevated above control.
Elustase exposure length required for an increase in
SLPI transcript. SLPI transcript accumulation, mea-
sured 48 h after initiation of treatment, was increased
following as little as 1 h elastase (2 and 10 U/ml) treat-
ment (Fig. 4). Lower elastase concentrations (1 and 2
U/ml) produced increases that were greater following 8
and 48 h treatment. A higher elastase concentration (10
U/ml) produced a greater response at 8 h, but at 48 h
SLPI transcript levels were decreased. Because 8 h treat-
C E C E
ment with all doses increased transcript levels, in the
remaining dose-response and specificity experiments en-
Donor 1 Donor 2 Donor 3 zyme treatments were 8 h with subsequent RNA harvest
Fig. 1. Secretory leukocyte protease inhibitor @LPI) transcript levels in 48 h after initiation of treatment.
primary human airway epithelial cells following neutrophil elastase
treatment (2 U/ml for 48 h). SLPI transcript levels in primary human Specificity and dose dependency of the response. To
epithelial cells, from 3 donors, treated with either medium alone (C) or determine whether the elastase-induced increase in SLPI
medium containing elastase (E). Total cellular RNA was analyzed by transcript levels is due to proteolytic activity, elastase
dot-blot analysis with subsequent autoradiographic densitometric anal- was used in the presence of an antiprotease, rSLP1 or
ysis. Values are means +. SE SLPI-to-actin ratio of 3-4 trials/treatment (pi-AT. Either antiprotease inhibited the elastase-in-
from each donor.
duced increase in SLPI transcript levels at the molar
ratios tested (Fig. 5), suggesting that the elastase-induced
A Primary B 9HTEo’ increase in SLPI transcript levels is due to proteolytic
activity. Also, rSLP1 or (wi-AT added alone had no effect
on SLPI transcript levels.
C E C E The effect of elastase on SLPI transcript levels was
compared with that of other serine proteases. SLPI tran-
script levels increased following rl U/ml elastase treat-
ment with the maximal accumulation following treat-
ment with 10 U/ml elastase. Treatment with 30 U/ml
elastase resulted in levels less than maximal (Fig. 6.4).
Cathepsin G, a se&e protease also released from the
azurophilic granule, did not alter SLPI transcript levels
at any dose tested (Fig. 6A). However, the optimum pH
for cathepsin G is 8.0 (Elastin Products), and these ex-
periments were conducted at pH 7.4. SLPI transcript

Fig. 2. SLPI transcript levels in primary and transformed (SHTEo-)

human airway epithelial cells following neutrophil elastase treatment (2
U/ml for 48 h). Shown are Northern blot analyses of 1.5 rg total cellular
RNA/lane from cells treated with medium alone (C) (A and B, 1st lane)
or medium containing elastase (E) (A and B, 2nd lane), hybridized with
cY-s*P-labeled SLPI (bottom) and @-actin (top) cDNA probes. Autorad-
iograph in A was exposed with 2 amplifying screens for 3 h and that in
B with 2 screens for 5 h. Constitutive SLPI levels in primary cells were
greater than that in SHTEo- cells.

treatment. Because SHTEo- and primary human epithe-

lial cells responded similarly to elastase treatment, alter- Fig. 3. Time course of SLPI transcript accumulation. SLPI transcript
ations in SLPI transcript levels were further studied us- levels were measured in SHTEo- cells treated with medium alone (0) or
cells treated with 2 U/ml elastase (a) from 8 to 96 h. Values are means
ing SHTEo- cells. f SE SLPI-to-actin ratio for 3 trials done on 2 separate occasions (n =
Increased SLPI transcript level is time dependent. In- 6 samples). Elastase treatment significantly increased SLPI transcript
creases in SLPI steady-state mRNA were detectable fol- levels from 8 to 72 h (P < 0.05).

Downloaded from journals.physiology.org/journal/ajplung (105.111.135.139) on August 7, 2020.

 ELASTASE INCREASES SLPI TRANSCRIPT LEVELS L289
granule components, myeloperoxidase and lysozyme,
were examined for their effect on SLPI transcript levels.
Myeloperoxidase modestly increased transcript levels at
doses of 10 and 30 pg/ml, whereas lysozyme did not alter
SLPI transcript levels at any dose tested (Fig. 6C). In
these experiments, myeloperoxidase was tested in the ab-
sence of its substrate, Hz02 (27).
Because elastase was obtained from human sputum
that could also contain cytokines or bacteria, we also
measured the effects of TNF-ar, IL-8, and lipopolysaccha-
ride (LPS) on SLPI transcript levels. SHTEo- cells were
treated for 72 h with either TNF-a! or IL-8 and total RNA
immediately harvested. TNF-a! had a modest effect at
lo-100 rig/ml, whereas IL-8 had no effect on SLPI tran-
script levels at the doses tested (mean t SE, SLPI-to-
0 8 16 24 32 40 48
actin ratio: 72 h control = 0.62 t 0.16; 30 rig/ml TNF-a!
Hours = 1.03 t 0.60; 30 rig/ml IL-8 = 0.59 t 0.16; and 2 U/ml
Fig. 4. Effect of duration of elastase treatment on SLPI transcript lev- elastase = 10.45 t 3.30).
els. SHTEo- cells were treated with 1 U/ml (o), 2 U/ml (a), or 10 U/ml Treatment of SHTEo- cells with 10 pg/ml (Fig. 7) or
(A) neutrophil elastase for 1,8, and 48 h and harvested for total RNA 48
h after initiation of treatment. Values are means k SE SLPI-to-actin 100 pg/ml (SLPIto-actin ratio = 0.29 t 0.07) E. cob LPS
ratio for 3 trials done on 3 separate occasions (n = 9). Transcript levels for 48 h had little or no effect on SLPI transcript levels.
were significantly elevated over control (SLPI-to-actin =O.lO k 0.01) LPS derived from P. aerugimsa also had little or no effect
for all doses and times except 1 h of 1 U/ml elastase treatment (P < on SLPI transcript levels at 10 and 100 pg/ml (mean k
0.05).
SE SLPI-to-actin ratio of control = 0.46 t 0.02,lO pg/ml
3 LPS = 0.61 t 0.05, 100 pg/ml LPS = 0.59 t 0.08). In
addition, treatment of cells with elastase (2 U/ml) plus
polymyxin B (50 pg/ml), an inhibitor of the lipid A por-
tion of LPS (16), produces a response equal to that of
elastase alone (Fig. 7).

DISCUSSION

Airway inflammation is observed in diseases such as

asthma, bronchitis, and bronchiectasis (lo), and follow-
ing exposure to environmental agents, such as ozone (26).
Inflammation is often accompanied by increased levels of
protease (12,18,30), which can compromise the capacity
of the antiproteolytic screen and result in indiscriminate
0 tissue damage. To minimize tissue destruction, antipro-
-3:l 1:l -3:l 1:l
C E tease regulation may be sensitive to local levels of pro-
SLPl:E alAT:E tease. This study investigated a possible regulatory feed-
Fig. 5. Effect of antiproteases on elastase-induced increase in SLPI back mechanism by which a protease modulates the
transcript levels. SHTEo- cells were treated with 2 U/ml elastase in expression of one of its own inhibitors. We demonstrated
presence (hatched bars) and absence (solid bars) of an antiprotease that SLPI transcript levels in cultured human airway
(rSLP1 or al-AT) for 48 h. Control (open bar) was treated with medium
alone for 48 h. Concentration of each inhibitor was either a 3-fold molar epithelial cells increase in a dose- and time-dependent
excess (3: 1) or equimolar (1:l) to elastase. Values are means k SE manner with neutrophil elastase treatment. Pretreatment
SLPI-to-actin ratio. of cells with rSLP1 or CQ-AT inhibited the elastase effect,
suggesting that proteolytic activity may be responsible
levels were also measured following treatment with two for the alterations in SLPI transcript levels.
serine, nonneutrophil proteases, trypsin and pancreatic Elastase is a major serine protease secreted by neutro-
elastase, to determine whether other serine proteases, like phils during airway injury (27), and its involvement in
neutrophil elastase, could increase SLPI transcript levels. inflammatory processes has been previously examined.
Trypsin also was examined because SLPI possesses An early response to airway insult is accelerated glyco-
trypsin inhibitory activity (4, 8). Treatment of SHTEo- protein secretion, that functions to clear antigens or mi-
cells with trypsin increased SLPI transcript levels at croorganisms from the airway. Neutrophil elastase rap-
doses > 3 U/ml (Fig. 6B). We found that porcine pan- idly stimulates glycoprotein secretion from hamster
creatic elastase increased transcript levels at doses as low tracheal ring organ cultures (19) and bovine submucosal
as 1 U/ml (Fig. 6B). The pancreatic elastase, however, gland serous cells (28) within 1 h of initiation of treat-
has a high optimum pH (8.8, Elastin Products), and these ment. In addition, elastase may indirectly increase the
experiments were conducted at pH 7.4. recruitment of neutrophils by increasing transcript levels
Inasmuch as elastase is only one of the neutrophil prod- of a neutrophil chemotactic factor, IL-8, produced by a
ucts released during inflammation, two other azurophilic variety of cell types including airway epithelial cells (17).

Downloaded from journals.physiology.org/journal/ajplung (105.111.135.139) on August 7, 2020.

 ELASTASE INCREASES SLPI TRANSCRIPT LEVELS
-
- B
Fig. 6. Dose-response relationships of dif-
ferent proteases/enzymes on SLPI tran-
script levels in human airway epithelial
cells. SHTEo- cells were treated for 8 h
with neutrophil and nonneutrophil en-
zymes and total RNA harvested 48 h after
PPE 4 initiation of treatment: A, elastase (0) and
cathepsin G (a); B, porcine pancreatic
/ T elastase (A) and trypsin (A); C, myeloper-
oxidase (0) and lysozyme (m). Values are
means k SE SLPI-to-actin ratio for 3 tri-
als done on 3 separate occasions (n = 9
samples) (A and B) or 3 trials done on 2
occasions (n = 6 samples) (C). Control
values were 0.12 k 0.01, 0.22 k 0.03, and
0.21 * 0.02 for A, B, and C, respectively.
0.1 1 10 1000.1 1 10 1000.1 1 10 100 All tests were conducted at pH 7.4.
U/ml U/ml
Enzyme Concentration

observed at elastase concentrations as low as 0.1 nM (28).

IL-8 transcript levels were increased at concentrations
from 0.1 to 50 nM elastase (18). We found an increase in
SLPI transcript levels at doses as low as 40 nM (1 U/ml),
with maximal transcript accumulation following 400 nM
(10 U/ml) neutrophil elastase. Thus SLPI synthesis may
be initiated when local concentrations of elastase become
excessive.
We also examined the effect of other neutrophil and
nonneutrophil serine proteases on SLPI transcript levels.
Cathepsin G had no effect on SLPI transcript levels;
however, it was not used at its optimum pH (8.0). How-
ever, Sommerhoff et al. (28) demonstrated that elastase
E P E LPS E
and cathepsin G could stimulate secretion from bovine
+ submucosal gland serous cells at the pH 7.4. The effec-
P Lp+S tiveness of cathepsin G in stimulating glycoprotein secre-
Fig. 7. Effect of lipopolysaccharide (LPS) on SLPI transcript levels. tion may be due to species differences or differences in
SHTEo- cells were treated with medium alone (C), 2 U/ml elastase (E), cell type. We found that pancreatic elastase, a serine,
50 pg/ml polymyxin B (P), elastase with polymyxin B (E + P), 10 clglml
E. coli LPS (LPS), or elastase with LPS (E + LPS) for 48 h and nonneutrophil protease, increases SLPI transcript levels,
subsequently harvested for total cellular RNA. Values are means k SE but to a lesser degree than neutrophil elastase. Again, this
SLPI-to-actin ratio. may have been the consequence of not using the optimum
pH (8.8) for pancreatic elastase, whereas neutrophil
IL-8 transcript levels maximally accumulate within 3-6 h elastase was used nearer its optimum pH (7.5) (Elastin
following elastase treatment in human bronchial epithe- Products). Trypsin, another serine, nonneutrophil pro-
lial cells (18). Neutrophils assist in the phagocytosis and tease, was also tested because SLPI possesses antitrypsin
digestion of foreign antigens, microorganisms, and ne- activity. Trypsin increased SLPI transcript levels similar
crotic tissue (27). In contrast, local antiprotease synthesis to pancreatic elastase, but was effective at higher doses
and release can downregulate inflammatory responses than that required of neutrophil elastase. The diminished
and thus may have a slower time course of upregulation. sensitivity to trypsin and pancreatic elastase, relative to
In this study we found that SLPI transcript levels max- neutrophil elastase, may reflect tissue specificity, as these
imally accumulate 72 h following the initiation of elastase enzymes are less likely to be present in the airway.
treatment. Similarly, increases in al-AT transcript levels We also examined the effect of two nonprotease neu-
also exhibit a slow response to elastase (23). Thus the trophil enzymes, myeloperoxidase and lysozyme, which
sequence of responses of airway epithelial cells to elastase can be released concurrently with elastase. Both had little
treatment is as follows: increased airway secretions or no effect on SLPI transcript levels; however, myelo-
within 1 h, increased chemoattractant transcript accu- peroxidase was tested in the absence of its substrate
mulation within 6 h, followed by increased antiprotease w202) (27).
transcript levels within 72 h. Because of their ability to either attract and/or activate
The above sequence of inflammatory events is further neutrophils and thus increase local elastase concentra-
supported by the differences in sensitivity to elastase. tions, two cytokines, TNF-cu and IL-8, were also exam-
Increased glycoprotein secretion from serous cells was ined for their ability to increase SLPI transcript levels.

Downloaded from journals.physiology.org/journal/ajplung (105.111.135.139) on August 7, 2020.

 ELASTASE INCREASES SLPI TRANSCRIPT LEVELS L291
SLPI transcript levels were modestly increased following and Cancer Research Institute, San Francisco, CA) and BEAS-2B cells
high doses of TNF-(w, but were unaffected by IL-8. by Dr. Curtis Harris, (National Cancer Institute, Bethesda, MD).
Support was provided by the Center for Indoor Air Research grant
Because LPS, a potential contaminant of the protease/ (90-010); the Health Effects Institute, agreement 90-13, an organiza-
enzyme preparations, stimulates the expression of in- tion jointly funded by the United States Environmental Protection
flammatory mediators in airway epithelial cells (24)) we Agency (Assistance Agreement X-812059) and automotive manufactur-
asked whether LPS could increase SLPI transcript levels. ers; the National Institute Of Environmental Health Sciences Grant
P30-ES06096; the Electrical Power Research Institute Grant RP2155-
SHTEo- cells were treated with LPS derived from either 03; and Procter and Gamble UAARP-21314.
E. coli or P. aeruginosa, but they had little or no effect on J. M. Abbinante-Nissen is a recipient of a University of Cincinnati
SLPI transcript levels. Polymyxin B, a LPS inhibitor Graduate Assistantship and this work was conducted in partial fulfill-
(16), also did not diminish the elastase-induced increase ment of the requirements for the PhD degree at the University of
in SLPI transcript levels. Together, these results suggest Cincinnati.
The contents of this article do not necessarily reflect the views or
that the elastase effect we observed was unlikely to be a policies of the Health Effects Institute, Environmental Protection
result of LPS in the preparation. Agency, or automotive manufacturers.
As previously mentioned, neutrophil elastase also reg- Address for reprint requests: G. D. Leikauf, Pulmonary Cell Biology
ulates the expression of another of its inhibitors, al-AT. Laboratory ML 056, University of Cincinnati Medical Center, Cincin-
nati, OH 45267-0056.
Elastase (2 PM for 24 h) increased transcription of al-AT
in peripheral blood monocytes three to fourfold and in Received 9 October 1992; accepted in final form 4 May 1993.
bronchoalveolar macrophages up to eightfold (23). Other
REFERENCES
serine proteases tested, including cathepsin G and
trypsin, were ineffective at increasing al-AT transcript 1. Chomczynski, P., and N. Sacchi. Single-step method of RNA
isolation by acid guanidinium thiocyanate-phenol-chloroform ex-
levels, whereas porcine pancreatic elastase was as effec- traction. Anal. Biochem. 162: 156-159, 1987.
tive as neutrophil elastase (23). The sensitivities of SLPI 2. Church, G. M., and W. Gilbert. Genomic sequencing. Proc.
and al-AT to neutrophil elastase are similar, both requir- N&l. Acud. Sci. USA 81: 1991-1995, 1984.
ing higher elastase concentrations than other responses 3. De Water, R., L. N. A. Willems, G. N. P. Van Muijen, C. 0.
such as glycoprotein secretion and chemotactic factor Franken, J. A. M. Fransen, J. H. Dijkman, and 3. A.
Kramps. Ultrastructural localization of bronchial antileukopro-
production. Together the time course and dose-response tease in central and peripheral human airways by a gold-labeling
data for SLPI and al-AT give support to the concept that technique using monoclonal antibodies. Am. Rev. Respir. Dis. 133:
antiproteases function in the eventual downregulation of 882-889, 1986.
the inflammatory process. The mechanism by which 4. Eisenberg, S. P., K. K. Hale, P. Heimdal, and R. C.
elastase stimulates SLPI transcription was not examined Thompson. Location of the protease-inhibitory region of secre-
tory leukocyte protease inhibitor. J. Bzid. Chem. 265: 7976-7981,
in this study. However, Perlmutter et al. (11, 21, 23) 1990.
reported previously that a stable al-AT-neutrophil 5. Feinberg, A. P., and B. Vogelstein. A technique for radiola-
elastase complex increases al-AT transcript levels via a beling DNA restriction endonuclease fragments to high specific
serpin-enzyme complex receptor . Our data suggest that activity. Ad. Biochem. 132: 6-13, 1983.
the formation of a protease-antiprotease complex may 6. Gauthier, F., U. Fryksmark, K. Ohlsson, and J. G. Bieth.
Kinetics of the inhibition of leukocyte elastase by the bronchial
not be involved in elastase-induced increases in SLPI inhibitor. B&him. Biophys. Acta 700: 178-183, 1982.
transcript levels, because both rSLP1 and al-AT inhib- 7. Gruenert, D. C., C. B. Basbaum, M. J. Welsh, M. Li, W. E.
ited the stimulation of SLPI by elastase, as reflected in Finkbeiner, and J. A. Nadel. Characterization of human tra-
the marked decrease in SLPI transcript levels with cheal epithelial cells transformed by an origin-defective simian
virus 40. Proc. Natl. Acad. Sci. USA 85: 5951-5955.
cotreatment compared with elastase alone. Had a rSLPI- 8. Grutter, M. G., G. Fendrich, R. Huber, and W. Bode. The
or cul-AT- elastase complex been required for activation, 2.5 A X-ray crystal structure of the acid-stable proteinase inhib-
then cotreatment would have resulted in increases of itor from human mucous secretions analyzed in its complex with
SLPI transcript levels. bovine a-chymotrypsin. EMBO J. 7: 345-351, 1988.
The time course and sensitivity of inflammatory events 9. Heinzal, R., H. Appelhans, H-G Gassen, U. Seemuller, M.
Arnhold, H. Fritz, F. Lottspeich, K. Wiedenmann, and W.
in response to neutrophil products, particularly neutro- Machleidt. The neutrophil elastase-cathepsin G inhibitor of hu-
phi1 elastase, suggest an overall scheme of events that man mucus tissues and secretions (antileukoprotease, HUSI- I):
could occur following airway injury. Increases in airway complete primary structure as revealed by protein and DNA
secretions and neutrophil infiltration occur early to aid in sequencing. In: Pulmonary Emphysema and Proteolysis, edited
by J. C. Taylor and C. Mittman. Orlando, FL: Academic, 1986,
the phagocytosis and clearance of antigens or microor- p. 297-306.
ganisms. As shown, neutrophil elastase is capable of reg- 10. Hubbard, R. C., M. L. Brantly, and R. G. Crystal. Proteases.
ulating these early inflammatory events, resulting in the In: The Lung, edited by R. G. Crystal and J. B. West. New York:
amplification of these processes. Antiprotease regulation, Raven, 1991, p. 1763-1773.
although less sensitive to neutrophil products and occur- 11. Joslin, G., R. J. Fallon, J. Bullock, S. P. Adams, and D. H.
ring later in the process, works to combat excesses in Perlmutter. The SEC receptor recognizes a pentapeptide ne-
odomain of cxl-antitrypsin-protease complexes. J. Biol. Chem. 266:
proteolytic enzymes that can cause indiscriminate tissue 11282-11288, 1991.
damage and thus may overall downregulate the inflam- 12. Koren, H. S., R. B. Devlin, D. E. Graham, R. Mann, M. P.
matory process. McGee, D. H. Horstman, W. J. Kozumbo, S. Becker, D. E.
House, W. F. McDonnell, and P. A. Bromberg. Ozone-in-
The authors thank Dr. Paul Biddinger (University of Cincinnati, duced inflammation in the lower airways of human subjects. Am.
Dept. of Pathology and Lab Medicine) and the Cooperative Human Rev. Respir. Dis. 139: 407-415, 1989.
Tissue Network for assistance in tissue procurement. SLPI cDNA and 13. Lechner, J. F., and M. A. LaVeck. A serum-free method for
rSLP1 were gifts from Dr. Robert Thompson, Synergen, Boulder, CO. culturing normal human bronchial epithelial cells at clonal den-
SHTEo- cells were donated by Dr. Dieter Gruenert, (Cardiovascular sity. J. Tissue Culture Methods 9: 43-48, 1985.

Downloaded from journals.physiology.org/journal/ajplung (105.111.135.139) on August 7, 2020.

L292 ELASTASE INCREASES SLPI TRANSCRIPT LEVELS

14. Leikauf, G. D., H.-E. Claesson, C. A. Doupnik, S. regulates the synthesis of its inhibitor, al-proteinase inhibitor,
Hybbinette, and R. C. Grafstrijm. Cysteinyl leukotrienes en- and exaggerates the defect in homozygous Pi22 cyl PI deficiency.
hance growth of human airway epithelial cells. Am. J. Physiol. 259 J. CZin. Invest. 81: 1774- 1780, 1988.
(Lung Cell. Mol. Phyd. 3): L255-L261, 1990. 24. Piquet, P. F., M. A. Collart, G. E. Graw, Y. Kapanci, and
15. Mornex, J.-F., A. Chytil-Weir, Y. Martinet, M. Courtney, P. Vassalli. Tumor necrosis factor/cachectin plays a key role in
J.- P. LeCocq, and R. G. Crystal. Expression of the alpha-l- bleomycin-induced pneumopathy and fibrosis. J. Exp. Med. 170:
antitrypsin gene in mononuclear phagocytes of normal and alpha- 655-663, 1989.
l-antitrypsin-deficient individuals. J. Clin. Inuest. 77: 1952-1961, 25. Reddel, R. R., Y. Ke, B. I. Gerwin, M. G. McMenamin, J.
1986. F. Lechner, R. T. Su, D. E. Brash, J.-B. Park, J. S. Rhim,
16. Morrison, D. C., and D. M. Jacobs. Binding of polymyxin B to and C. C. Harris. Transformation of human bronchial epithelial
the lipid A portion of bacterial lipopolysaccharides. Immunochem- cells by infection with SV40 or adenovirus-12 SV40 hybrid virus,
istry 13: 813-818, 1976. or transfection via strontium phosphate coprecipitation with a
17. Nakumura, H., K. Yoshimura, H. A. Jaffe, and R. G. plasmid containing SV40 early region genes. Cancer Res. 48: 1904-
Crystal. Interleukin-8 gene expression in human bronchial epi- 1909, 1988.
thelial cells. J. BioZ. Chem. 266: 19611-19617, 1991. 26. Seltzer, J., B. G. Bigby, M. Stulbarg, M. J. Holtzman, J. A.
18. Nakumura, H., K. Yoshimura, N. G. McElvaney, and R. G. Nadel, I. F. Ueki, G. D. Leikauf, E. J. Goetzl, and H. A.
Crystal. Neutrophil elastase in respiratory epithelial lining fluid Boushey. O,-induced change in bronchial reactivity to metha-
of individuals with cystic fibrosis induces interleukin-8 gene ex- choline and airway inflammation in humans. J. AppZ. Physiol. 60:
pression in a human bronchial epithelial cell line. J. Clin. Inuest. 1321-1326, 1986.
89: 1478-1484, 1992. 27. Sibille, Y., and H. Y. Reynolds. Macrophages and polymor-
19. Niles, R. M., T. G. Christensen, R. Breuer, P. J. Stone, and phonuclear neutrophils in lung defense and injury. Am. Reu.
G. L. Snider. Serine proteases stimulate mucous glycoprotein Respir. Dis. 141: 471-501, 1990.
release from hamster tracheal ring organ culture. J. Lab. CZin. 28. Sommerhoff, C. P., J. A. Nadel, C. B. Basbaum, and G. H.
Med. 108: 489-497, 1986. Caughey. Neutrophil elastase and cathepsin G stimulate secre-
20. Ohlsson, K., M. Rosengren, G. Stetler, M. Brewer, K. K. tion from cultured bovine airway gland serous cells. J. CZin. Inuest.
Hale, and R. C. Thompson. Structure, genomic organization, 85: 682-689, 1990.
and tissue distribution of human secretory leukocyte-protease in- 29. Stetler, G., M. T. Brewer, and R. C. Thompson. Isolation
hibitor (SLPI): a potent inhibitor of neutrophil elastase. In: PuZ- and sequence of a human gene encoding a potent inhibitor of
monaly Emphysema and Proteolysis, edited by J. C. Taylor and C. leukocyte proteases. Nucleic Acids Res. 14: 7883-7896, 1986.
Mittman. Orlando, FL: Academic, 1986, p. 307-324. 30. Stockley, R. A., and D. Burnett. Alphal-antitrypsin and leu-
21. Perlmutter, D. H., G. I. Glover, M. Rivetna, C. S. kocyte elastase in infected and noninfected sputum. Am. Rev.
Schasteen, and R. J. Fallon. Identification of a serpin-enzyme Respir. Dis. 120: 1081-1086, 1979.
complex receptor on human hepatoma cells and human mono- 31. Thompson, R. C., and K. Ohlsson. Isolation, properties, and
cytes. Proc. NatZ. Acad. Sci. USA 87: 3753-3757, 1990. complete amino acid sequence of human secretory leukocyte pro-
22. Perlmutter, D. H., F. Sessions Cole, P. Kilbridge, T. H. tease inhibitor, a potent inhibitor of leukocyte elastase. Proc. NatZ.
Rossing, and H. R.Colten. Expression of the al-proteinase in- Acad. Sci. USA 83: 6692-6696, 1986.
hibitor gene in human monocytes and macrophages. Proc. NatZ. 32. Willems, L. N. A., J. A. Kramps, T. Stijnen, P. J. Sterk, J.
Acad. Sci. USA 82: 795-799, 1985. J. Weening, and J. H. Dijkman. Antileukoprotease-containing
23. Perlmutter, D. H., J. Travis, and P. I. Punsal. Elastase bronchiolar cells. Am. Rev. Respir. Dis. 139: 1244-1250, 1989.

Downloaded from journals.physiology.org/journal/ajplung (105.111.135.139) on August 7, 2020.