New selective peptidyl di(chlorophenyl) phosphonate esters for visualizing and blocking neutrophil proteinase 3 in human diseases

Article abstract of DOI:10.1074/jbc.M114.591339

The function of neutrophil protease 3 (PR3) is poorly under-stood despite of its role in autoimmune vasculitides and its possible involvement in cell apoptosis. This makes it different from its structural homologue neutrophil elastase (HNE). Endogenous inhibitors of human neutrophil serine proteases preferentially inhibit HNE and to a lesser extent, PR3. We constructed a single-residue mutant PR3 (I217R) to investigate the S4 subsite preferences of PR3 and HNE and used the best peptide substrate sequences to develop selective phosphonate inhibitors with the structure Ac-peptidyl^P(0-C₆H₄-4-Cl)₂. The combination of a prolyl residue at P4 and an aspartyl residue at P2 was totally selective for PR3. We then synthesized N-terminally biotinylated peptidyl phosphonates to identify the PR3 in complex biological samples. These inhibitors resisted proteolytic degradation and rapidly inactivated PR3 in biological fluids such as inflammatory lung secretions and the urine of patients with bladder cancer. One of these inhibitors revealed intracellular PR3 in permeabilized neutrophils and on the surface of activated cells. They hardly inhibited PR3 bound to the surface of stimulated neutrophils despite their low molecular mass, suggesting that the conformation and reactivity of membrane-bound PR3 is altered. This finding is relevant for autoantibody binding and the subsequent activation of neutrophils in granulomatosis with polyangiitis (formerly Wegener disease). These are the first inhibitors that can be used as probes to monitor, detect, and control PR3 activity in a variety of inflammatory diseases.

Full text of DOI:10.1074/jbc.M114.591339

Enzymology:
New Selective Peptidyl Di(chlorophenyl)
Phosphonate Esters for Visualizing and
Blocking Neutrophil Proteinase 3 in
Human Diseases
Carla Guarino, Monika Legowska, Christophe
Epinette, Christine Kellenberger, Sandrine
Dallet-Choisy, Marcin Sienczyk, Guillaume
Gabant, Martine Cadene, Jérôme Zoidakis,
Antonia Vlahou, Magdalena Wysocka,
Sylvain Marchand-Adam, Dieter E. Jenne,
Adam Lesner, Francis Gauthier and Brice
Korkmaz
J. Biol. Chem. 2014, 289:31777-31791.
doi: 10.1074/jbc.M114.591339 originally published online October 6, 2014
Access the most updated version of this article at doi: 10.1074/jbc.M114.591339
Find articles, minireviews, Reflections and Classics on similar topics on the JBC Affinity Sites.
Alerts:
•
•
When this article is cited
When a correction for this article is posted
Click here to choose from all of JBC's e-mail alerts
Supplemental material:
http://www.jbc.org/content/suppl/2014/10/06/M114.591339.DC1.html
This article cites 51 references, 21 of which can be accessed free at
http://www.jbc.org/content/289/46/31777.full.html#ref-list-1

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 289, NO. 46, pp. 31777–31791, November 14, 2014
2014 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A.
©
New Selective Peptidyl Di(chlorophenyl) Phosphonate Esters
for Visualizing and Blocking Neutrophil Proteinase 3 in
□
S
*
Human Diseases
Received for publication, June 27, 2014, and in revised form, September 23, 2014 Published, JBC Papers in Press, October 6, 2014, DOI 10.1074/jbc.M114.591339
§§
‡
§
‡
¶
‡
Carla Guarino , Monika Legowska , Christophe Epinette , Christine Kellenberger , Sandrine Dallet-Choisy ,
ʈ
‡‡
‡‡
Marcin Sie n´ czyk , Guillaume Gabant**, Martine Cadene**, J e´ r oˆ me Zoidakis , Antonia Vlahou ,
Magdalena Wysocka , Sylvain Marchand-Adam , Dieter E. Jenne , Adam Lesner , Francis Gauthier ,
§
‡
§§
§
‡
‡
1
and Brice Korkmaz
‡
From the INSERM U-1100/EA-6305 Centre d’Etude des Pathologies Respiratoires and Universit e´ Fran c¸ ois Rabelais, 37032 Tours,
France, Faculty of Chemistry, University of Gdansk, 80-952, Gdansk, Poland, Architecture et Fonction des Macromol e´ cules
Biologiques, CNRS-Unit e´ Mixte de Recherche (UMR),13288 Marseille, France, Wroclaw University of Technology, Faculty of
§
¶
ʈ
Chemistry, Division of Medicinal Chemistry and Microbiology, 50-370 Wroclaw, Poland, **Centre de Biophysique Mol e´ culaire,
‡
‡
UPR4301 CNRS, 45071 Orl e´ ans, France, Biotechnology Division, Biomedical Research Foundation, Academy of Athens, 11527
§
§
Athens, Greece, and Comprehensive Pneumology Center, Institute of Lung Biology and Disease, German Center for Lung
Research (DZL), 81377 Munich and Max Planck Institute of Neurobiology, 82152 Planegg-Martinsried, Germany
Background: Proteinase 3 activity is poorly controlled by physiological inhibitors, and its biological function is not well
understood.
Results: We have designed irreversible phosphonate inhibitors based on structural differences between proteinase 3 and
elastase.
Conclusion: They selectively inhibit proteinase 3 in biological fluids and can act as activity-based probes.
Significance: These inhibitors will help clarify proteinase 3 function.
The function of neutrophil protease 3 (PR3) is poorly under- finding is relevant for autoantibody binding and the subsequent
stood despite of its role in autoimmune vasculitides and its pos- activation of neutrophils in granulomatosis with polyangiitis (for-
sible involvement in cell apoptosis. This makes it different from merly Wegener disease). These are the first inhibitors that can be
its structural homologue neutrophil elastase (HNE). Endoge- used as probes to monitor, detect, and control PR3 activity in a
nous inhibitors of human neutrophil serine proteases preferen- variety of inflammatory diseases.
tially inhibit HNE and to a lesser extent, PR3. We constructed a
single-residue mutant PR3 (I217R) to investigate the S4 subsite
preferences of PR3 and HNE and used the best peptide substrate
There are ϳ100 active serine proteases with a trypsin/chy-
sequences to develop selective phosphonate inhibitors with the motrypsin-like fold that are encoded by distinct human genes,
P
structure Ac-peptidyl (O-C H -4-Cl) . The combination of a but the physiological substrates and in vivo function of most of
6
4
2
prolyl residue at P4 and an aspartyl residue at P2 was totally them are still poorly characterized. Although they are potential
selective for PR3. We then synthesized N-terminally biotiny- therapeutic targets in a large number of diseases, only a few
lated peptidyl phosphonates to identify the PR3 in complex bio- inhibitors, primarily those that interfere with the coagulation
logical samples. These inhibitors resisted proteolytic degrada- cascade (factor Xa, thrombin inhibitors), have been approved
2
tion and rapidly inactivated PR3 in biological fluids such as for clinical use (for review see Ref. 1). Human proteinase 3 (PR3)
inflammatory lung secretions and the urine of patients with (EC 3.4.21.76) is a neutrophilic serine protease that shares many
bladder cancer. One of these inhibitors revealed intracellular PR3 structural and functional characteristics with human neutrophil
in permeabilized neutrophils and on the surface of activated cells. elastase (HNE) (EC 3.4.21.37) (2, 3). Large amounts of both
They hardly inhibited PR3 bound to the surface of stimulated neu- proteases are stored intracellularly in so-called primary gran-
trophils despite their low molecular mass, suggesting that the con- ules and contribute to the breakdown of extracellular matrix
formation and reactivity of membrane-bound PR3 is altered. This components in infectious and inflammatory diseases, especially
those of the lung (4). PR3 has also been identified as the princi-
pal autoantigen in one clinical subtype of systemic autoimmune
*
This work was supported by the Minist e` re de l’Enseignement Sup e´ rieur et
de la Recherche, the R e´ gion Centre, and the Fonds Europ e´ en de D e´ vel-
vasculitis, granulomatosis with polyangiitis (GPA) (formerly
oppement R e´ gional (project INFINHI) and Polish Ministry of Science and Wegener disease) (5–7). The PR3 in activated neutrophils with
Higher Education (project IP2012 0596 72; to M. W.). This work was also
supported by European Union Seventh Framework Programme (FP7/
destabilized lysosomal membranes can induce apoptosis and
2
007–2013; to D. E. J.) under Grant 261382 and by European Union Com-
2
m
mission (FP7/ TransBioBC; to J. Z. and A. V.) under Grant 601933.
This article contains supplemental Video 1.
Supported by the Alexandre von Humboldt Foundation. To whom corre-
spondence should be addressed. Tel.: 33-2-47-36-62-53; Fax: 33-2-47-36-
The abbreviations used are: PR3, proteinase 3; PR3 , membrane-bound PR3;
□S
HNE, human neutrophil elastase; GPA, granulomatosis with polyangiitis; pNA,
para-nitroaniline; ABZ, ortho-aminobenzoic acid; EDDnp, N-(2.4-dinitrophe-
nyl)ethylenediamine; Bt, biotin; CF, cystic fibrosis; Fmoc, (fluoromethyloxycar-
bonyl chloride); DMF, dimethylformamide; RT, room temperature.
1
6
0-46; E-mail: brice.korkmaz@inserm.fr.
NOVEMBER 14, 2014•VOLUME 289•NUMBER 46
JOURNAL OF BIOLOGICAL CHEMISTRY 31777

Selective PR3 Activity-based Probes
hence accelerate their death in inflamed tissues (8). Unlike
Synthesis of Peptidyl Phosphonate Inhibitors—The first step
HNE, PR3 is also present in highly mobile secretory vesicles and in the synthesis of the phosphonic analog of alanine was the
is translocated to the outer plasma membrane under certain preparation of tri(4-chlorophenyl)phosphite from 4-chloro-
conditions of priming (9). Furthermore, very small amounts of phenol and phosphorus trichloride (21). Briefly, phosphorus
PR3 are constitutively exposed on the outer surface of circulat- trichloride (10 mmol) was added to 4-chlorophenol (30 mmol)
ing neutrophils (10). This genetically determined constitutive dissolved in acetonitrile (50 ml), and the mixture was refluxed
distribution is a unique feature of human PR3 that may explain for 6 h. The volatile elements were removed in a vacuum, and
its function of autoantibody target in vasculitides (11). Natu- the resulting crude phosphite, a yellow oil, was used directly in
rally occurring inhibitors of PR3 in the extracellular compart- an amidoalkylation reaction. It was mixed with benzyl carba-
ment and blood plasma target HNE preferentially, which makes mate (12 mmol) and acetaldehyde (12 mmol) and refluxed in
investigating and understanding its biological function partic- acetic acid for 3 h (Oleksyszyn and Powers (22)). The resulting
P
carboxybenzyl-Ala (O-C H -4-Cl) was a white solid, 56%;
ularly complex (12).
6
4
2
yield, purity after 56%; high resolution mass spectrometry cal-
Peptidyldiphenyl phosphonate inhibitors are irreversible
transition state inhibitors that form a tetrahedral adduct with
the serine 195 residue (chymotrypsin numbering) of the cata-
lytic triad (13, 14). They selectively inhibit serine proteases, are
chemically stable in several buffers and in the plasma under
acidic and neutral conditions, and are effective at low concen-
trations (15). They can also be used as activity-based probes for
labeling serine proteases at the cell surface (16) and even within
the cell when synthesized in a membrane-permeable form (17).
These inhibitors, therefore, seem to be most appropriate for
dissecting the intracellular and extracellular biological roles of
enzymatically active PR3 whether free or membrane-bound.
We and others have shown that the substrate binding site of
PR3 extends on both side of the catalytic site and that the Asp
residues at P2 and P2Ј (nomenclature of Schechter and Berger
ϩ
culated for (C H Cl NO P)H , 480.0534; found, 480.0533;
22 20
2
5
1
H NMR (CDCl , ppm), ␦ 7.43-6.97 (m, 14H), 5.22-5.08 (m,
3
3
1
2
H), 4.74-4.37 (m, 1H), 1.56 (dd, J ϭ 18.2, 7.4 Hz, 3H); P NMR
(
CDCl , ppm): ␦ 19.56 (s). The carboxybenzyl protecting group
3
was then removed by incubation with 33% hydrobromic acid in
acetic acid (2 h), and the volatile components were removed
under reduced pressure. The product was crystallized from
P
methanol/diethyl ether to give HBr*H N-Ala (OPh) as a white
2
2
solid, 97%; high resolution mass spectrometry calculated for
ϩ
1
(
(
(
C H Cl NO P)H , 346.0167; found, 346.0172; H NMR
14 14
2
3
DMSO-d , ppm): ␦ 8.85 (s, 3H), 7.57-7.44 (m, 4H), 7.36-7.16
6
3
1
m, 4H), 4.45-4.24 (m, 1H), 1.55 (dd, J ϭ 18.3, 7.2 Hz, 3H);
P
NMR (CDCl , ppm): ␦ 16.49 (s).
3
The peptides were synthesized manually by the solid-phase
method using Fmoc (fluoromethyloxycarbonyl chloride) chemis-
try. The following amino acid derivatives were used: Fmoc-Pro,
Fmoc-Tyr(tBu), Fmoc-Asp(OtBu). The protected derivative
of the C-terminal amino acid residue, Fmoc-Asp(OtBu), was
(
18)) are essential to obtain selectivity toward PR3 (19, 20). Our
goal was to produce an inhibitor that was selective for PR3 and
had a sequence that binds only to the nonprime subsites of the
protease. Having an Asp at P2 is not sufficient to ensure a selec-
tive interaction with PR3; we therefore used the difference
between the structures of the S4 subsites of PR3 and HNE to
determine whether the cooperation between the S4 and the S2
subsites could provide inhibitors selective for PR3.
Ϫ
attached to the 2-chlorotritylchloride resin (substitution of Cl
1
.46 meq/g) (Calbiochem) in the presence of an equimolar
amount of diisopropylethylamine (based on the amino acid)
under anhydrous conditions in dichloromethylene solution.
Peptide chains were elongated in consecutive cycles of depro-
tection and coupling. They were deprotected with 20% piperi-
dine in dimethylformamide (DMF)/n-methylpyrrolidone (1:1,
v/v) plus 1% Triton X-100, whereas chain elongation used
standard differential interference contrast/hydroxybenzotria-
zole chemistry; 3 eq of protected amino acid derivatives were
used. The same method was used to couple biotin and fluores-
cent compounds (used as N-protected Fmoc derivatives). A
We designed a tetrapeptide to be the peptide moiety of a PR3-
selective, irreversible, easy-to-handle chlorodiphenyl phospho-
nate inhibitor. This compound has proved to be a successful probe
for detecting PR3 activity in biological samples or for visualizing
and monitoring PR3 both inside cells and at the cell surface.
EXPERIMENTAL PROCEDURES
Production of proI217R—The proI217R mutant was produced
in Sf9 insect cells using the pMT/BiP/proPR3/His vector as a
matrix and two complementary primers (5Ј-ccaaggaatagactc-
1
0-fold molar excess of N-acetylimidazole in DMF was used for
P
N-terminal acetyl derivates. Bt-[PEG] -Pro-Tyr-Asp-Ala (O-
6
6
C H -4-Cl) was synthesized by coupling the ␣-(9-fluorenyl-
6
4
2
cttcgtgaggtggggatgtgcc-3Ј and 5Ј-ggcacatccccacctcacgaaggagt- methyloxycarbonyl)amino-␻-carboxyl poly(ethylene glycol) Fmoc-
ctattccttgg-3Ј). The mutation was introduced using the Quik- polyethylene glycol-66 (PEG ; M 3000 Da) to the amino group of
66
r
Change Lightning Site-Directed Mutagenesis kit (Stratagene, Pro. The N-terminal biotin group was introduced stirring a 5-fold
La Jolla, CA), and its presence was checked by sequencing molar (1.8 mmol) excess of biotin and 1,3-diisopropylcarbidiimide
(
MWG Biotech). We established a stable cell line using in 2 ml of anhydrous DMSO for 6 h at 30 °C. The synthesized
antibiotic selection, and the cells were cultured in Schneider peptides were cleaved from the resin with trifluoroacetic acid/
medium supplemented with 10% fetal bovine serum. We used hexane/acetic acid (1:6:1, v/v/v).
CuSO to induce synthesis of the protein, which was purified by
Fully protected peptides were dissolved in DMF, and their
4
affinity chromatography on a chelating Sepharose fast flow carboxyl groups were activated with 1,3-diisopropylcarbidi-
P
resin column (Amersham Biosciences), eluted with a gradient of imide to which were added HBrϫH N-Ala (OPh) in DMF and
2
2
imidazole. The proI217R was activated as described previously for diisopropylethylamine. The mixture was stirred for 6 h, and the
recombinant wtPR3 and the K99L (19).
DMF was evaporated off under reduced pressure. The resulting
3
1778 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 289•NUMBER 46•NOVEMBER 14, 2014

Selective PR3 Activity-based Probes
compound was suspended in trifluoroacetic acid/phenol/tri- with bladder cancer were adjusted to give a final PR3 concen-
isopropylsilane/H O (88:5:2:5, v/v/v/v) for 2 h to remove side- tration of 10–100 nM (4–40 ng). These samples were incubated
2
chain protecting groups.
with compound 5 (150 nM final) for 20 min at 37 °C in 50 mM
The crude peptides were purified by high performance liquid HEPES, 0.75 M NaCl, and 0.05% Nonidet P-40 (pH 7.4). The reac-
chromatography (HPLC) on a Beckman Gold System (Beck- tion was stopped by adding 1 volume of 2ϫ SDS reducing buffer
man) with an RP Kromasil-100, C8, 5 ␮m column (8 ϫ 250 mm) and heating at 90 °C for 5 min. The components of the mixture
(
Knauer, Berlin, Germany). The solvent systems were 0.1% tri- were separated by SDS-PAGE, 12% SDS-polyacrylamide gel elec-
fluoroacetic acid (A) and 80% acetonitrile in A (B). Either iso- trophoresis under denaturing conditions (4–40 ng of protein per
cratic conditions or a linear gradient were applied (flow rate 3.0
ml/min, monitored at 226 nm). The purity of the synthesized pep-
tides was checked on an RP Kromasil 100, C8, 5 ␮m column (4.6 ϫ
lane). They were transferred to a nitrocellulose (Hybond)-ECL
(
Enhanced Chemiluminescence) membrane at 4 °C.
Extravidin-peroxidase Detection—Free sites on the mem-
brane were blocked with 3% bovine serum albumin (BSA) in
PBS (phosphate-buffered saline) or with 0.1% Tween in PBS for
2
50 mm) (Knauer). The peptides were eluted with a linear gradient
of the above solvent system (10–90% B) for 30 min and a flow rate
1
ml/min, monitored at 226 nm. Mass spectrometric analysis was
9
0 min at room temperature (RT). Membranes were then given
done on a MALDI MS (a Biflex III MALDI-TOF spectrometer,
Bruker Daltonics, Ettlingen, Germany) using a CCA matrix.
Enzyme Studies—Free wtPR3, I217R, K99L, and HNE (Bio-
centrum, Krakow, Poland) were titrated with ␣-1-proteinase
two quick washes with PBS, Tween 0.1% and incubated for 2 h
at RT with extravidin horseradish peroxidase (HRP) (Sigma)
(
diluted 1/4000 in 3% BSA in PBS, Tween 0.1%). The extravidin-
HRP-treated membrane was washed (3 ϫ 10 min) with PBS,
Tween 1% and then incubated with HRP substrate for 1 min. Reac-
tive bands were identified by chemiluminescence (ECL kit).
Immunodetection—Free sites on the membranes were blocked
by incubation with 5% nonfat dried milk in PBS, 0.1% Tween for 90
min at RT. They were washed twice with PBS, Tween 0.1% and
incubated overnight with a rabbit primary anti-PR3 antibody (1:700)
followed by a goat anti-rabbit IgG secondary antibody (1:7000).
These membranes were then washed and processed as above.
ELISA—The wells of Nunc Maxisorp plates were coated with
purified wtPR3 and HNE by incubating them with enzyme (0.05
inhibitor (23). The specificity constants k /K for fluorescence
cat
m
resonance energy transfer (FRET) substrates and para-nitroani-
lide (pNA) substrates synthesized by Genecust (Dudelange Lux-
embourg) were determined under first-order conditions (23). All
assays were performed at 37 °C in 50 mM HEPES, 0.75 M NaCl, and
0
.05% Nonidet P-40 (pH 7.4).
We measured the inactivation of proteases by substrate ana-
log inhibitors in the presence of the substrate by competition
for the enzyme binding site by the method of Tian and Tsou
(
24). Product formation in the presence of an irreversible inhib-
itor approaches an asymptote in this system, as described by:
log([P∞] Ϫ [P]) ϭ log[P∞] Ϫ 0.43A[Y]t, where [P∞] is the con-
centration of product formed at time approaching infinity, [P]
is the concentration of product at time t, [Y] is the inhibitor
concentration, and A is the apparent inhibition rate constant in
␮
M in 100 ␮l PBS) at 37 °C for 1 h. The plates were washed 4
P
times with 200 ␮l of PBS, Tween 0.1%. Bt-[PEG] -PYDA (O-
6
6
C H -4-Cl) (compound 9) (2.5 ␮M final) in 100 ␮l of HEPES
6
4
2
buffer was added to each well, and the plates were incubated at
3
7 °C for 180 min. The wells were washed 4 times with 200 ␮l of
the presence of the substrate. A is given by: A ϭ k /(1 ϩ
ϩo
PBS, Tween 0.1% and then saturated by adding 2% BSA in PBS
(200 ␮l/well) and incubated for 1 h at 37 °C. The plates were
washed again (4 ϫ 200 ␮l PBS, Tween 0.1%) and incubated with
extravidin-HRP (1/2000) or streptavidin-HRP (1/15000) in
Ϫ1
K
[S]), where k is the rate constant for association of the
ϩo
Ϫ1
inhibitor with the enzyme, K is the inverted Michaelis con-
stant, and [S] is the substrate concentration. The apparent inhi-
bition rate constant A is the slope of a plot of log([P∞] Ϫ [P])
0.1% Tween, 2% BSA in PBS for 2 h at RT. The plates were washed
against t, to give the second-order rate constant of inhibition k
.
ϩo
again and incubated with the chromogenic substrate ortho-phe-
nylenediamine (Sigma). The reaction was stopped by adding
H SO (2 M), and the absorbance at 492 nm was measured.
The rates of inhibition of PR3 and HNE were measured in 50 mM
HEPES, 0.75 M NaCl, and 0.05% Nonidet P-40 (pH ϭ 7.4).
2
4
Identification of Cleavage Sites by Mass Spectrometry—HPLC-
purified fragments of FRET substrates were diluted 20-fold in ace-
tonitrile:water:formic acid (45:54.8:0.2) and analyzed by direct
infusion in a 4-GHz MaXis ultra high resolution quadrupole-
TOF mass spectrometer (Bruker Daltonics) equipped with an
electrospray ion source. Spectra were acquired in positive ion
MS mode over a 300–3000 m/z range with a nebulizer gas pres-
sure of 0.3 bars. The drying gas flow was 4 liter/min, and the
temperature was 180 °C. The acquisition rate was 1 Hz corre-
sponding to spectra summations of 5494. External calibration was
performed with the ESI-L Low Concentration Tuning Mix (Agi-
Flow Cytometry and Confocal Microscopy—Neutrophils were
isolated from heparinized peripheral blood samples from
healthy volunteers who gave their informed consent (25). Puri-
fied quiescent neutrophils were suspended in PBS and fixed
with 3.7% (v/v) paraformaldehyde (20 min at RT). They were
washed twice with PBS and permeabilized with 0.5% (v/v) Tri-
ton X-100 (10 min at RT). Permeabilized cells were washed
twice in PBS and then incubated for 20 min at RT with 10 ␮M
P
Bt-[PEG] -PYDA (O-C H -4-Cl) (compound 9). These cells
66
6
4
2
were then washed twice in PBS to remove unbound inhibitors, and
lent Technologies). Electrospray ionization-high resolution mass free binding sites were blocked by incubation with 10% (w/v) BSA
spectrometry spectra were processed using DataAnalysis 3.1 soft- in PBS for 1 h. They were then incubated with Alexa Fluorꢀ 488
ware (Bruker Daltonics), and cleavage peptides were assigned streptavidin (Invitrogen, diluted 1/1000 in 3% (w/v) PBS) for 1 h at
using Paws Version 8.5.0.3 (ProteoMetrics, New York, NY).
RT. Flow cytometry or confocal microscopy were performed
Detection of PR3 in Biological Fluids—The volumes of cystic immediately after this last wash. For flow cytometry, cells were
fibrosis (CF) sputa and 10ϫ concentrated urines from patients suspended in 200 ␮l of PBS and examined in a flow cytometer
NOVEMBER 14, 2014•VOLUME 289•NUMBER 46
JOURNAL OF BIOLOGICAL CHEMISTRY 31779

Selective PR3 Activity-based Probes
FIGURE 1. Hydrolysis of the polyvalent FRET substrate ABZ-QDMAVVQSVPQ-EDDnp (substrate 3) by wtPR3, the PR3 variant I217R, and HNE. A, HPLC
C18 analysis of substrate fragments after extensive hydrolysis. Fragments were detected at 220 nm (blue) and 360 nm (green) as described (19). The insets show
the relative positions of the key residues at positions 99 and 217 (chymotrypsin numbering) in the crystal structures of PR3 (PDB code 1FUJ) (26) and HNE (PDB
code 1PPF) (49) and the molecular model of the PR3 variant I217R. The active site serine and Ile-217 are shown in yellow and orange, respectively. B, identifi-
cation of the cleavage sites by mass spectrometry (top) and summary of the identified cleavage sites (bottom).
equipped with a 488-nm argon laser. For microscopy, 200 ␮l of philic in PR3 than in HNE mainly because the Leu at position 99
suspended cells were centrifuged (500 g, 5 min) on a glass slide, is replaced by a Lys in the vicinity of both the S2 and S4 pockets
treated with DAPI (4Ј,6-diamidino-2-phenylindole), and exam- (26). Similarly, there is a charged arginyl residue at position 217
ined under a Zeiss Axio Imager fluorescence microscope.
in HNE in the vicinity of the S4, whereas it is an Ile in PR3. We
investigated the importance of the S4 subsite in PR3 by produc-
ing a new PR3 variant with an Arg at position 217 (I217R) (Fig.
RESULTS
Design of the Selective PR3-recognizing Peptide Sequence— 1A). The activities of the wild type and the two PR3 mutants
The charge distribution of the solvent-accessible surfaces in the (I217R, K99L (19)) were then assayed with peptide-based FRET
vicinity of the substrate binding sites of PR3 and HNE differ protease substrates derived from the PR3/NE substrate ABZ-
significantly (20, 26). The S2 pocket is smaller and more hydro- QPMAV2VQSVPQ-EDDnp (substrate 1) (19) bearing either a
3
1780 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 289•NUMBER 46•NOVEMBER 14, 2014

Selective PR3 Activity-based Probes
TABLE 1
Kinetics of synthetic substrate cleavage by proteinase 3 and elastase
The arrows indicate the cleavage sites: 2PR3, 1I217R, 8HNE. n.h., no hydrolysis.
*
Values are the means Ϯ S.D. of three experiments.
Values were from Jégot et al., (19).
#
TABLE 2
Rates of inhibition of wild type proteinase 3 and elastase by peptide phosphonates
n.s., not significant.
*
Values are the means Ϯ S.D. of three experiments.
positively charged residue ABZ-QKMAV2VQSVPQ-EDDnp ing the Pro at P4 by Asp and the Ala at P2 by Asp. The resulting
substrate 2) or a negatively charged residue QDMAV- substrate (substrate 5) was not hydrolyzed by wtPR3 but was
VQSVPQ-EDDnp (substrate 3) at P4. The wtPR3, the PR3 cleaved by I217R. As expected K99L did not cleave substrates
(
2
mutants, and HNE all cleaved substrate 1 at the V-V bond with with an Asp at either P2 or P4 (Table 1). HNE still cleaved
similar specificity constants (Table 1). Although replacing P4 substrate 5, but cleavage was one bond downstream from the
Pro by a positively charged residue in substrate 1 (substrate 2) V-V bond (Fig. 1, A and B), which is not a drawback when
had no significant influence on the rate of hydrolysis, an Asp at constructing inhibitors with no prime residues. We deduced
P4 (substrate 3) strongly impaired cleavage by wtPR3 and K99L from these data that the sequence Asp-X-Asp-Ala (DXDA)
but not by I217R, which cleaved at the V-V bond with a signif- should be optimal for discriminating between PR3 and HNE in
icantly larger k /K (Table 1 and Fig. 1, A and B). Although terms of selectivity, but that PR3 would have a greater affinity
cat
m
replacing the Ala at P2 by Asp in substrate 1 (substrate 4) for Pro-X-Asp-Ala (PXDA).
resulted in a large increase in the specificity constant with
We previously observed that a wide range of amino acid resi-
wtPR3 (19), I217R still cleaved at the same V-V bond with a dues could be accommodated at the PR3 P3 subsite, whereas Tyr
similar specificity constant (Table 1). These observations was more appropriate for PR3 (12). We, therefore, constructed
strongly suggest that the S4 pocket greatly influences the sub- pNA substrates having the peptide sequences Ac-DYDA- and Ac-
strate specificity of PR3. This was also demonstrated by replac- PYDA- from P4 to P1. The specificity constant for Ac-PYDA-
NOVEMBER 14, 2014•VOLUME 289•NUMBER 46
JOURNAL OF BIOLOGICAL CHEMISTRY 31781

Selective PR3 Activity-based Probes
pNA (substrate 6) was higher than for Ac-DYDA-pNA (sub-
strate 7), but substrate 7 was also hydrolyzed at a significant
rate (Table 1). These two sequences were selected to con-
struct peptidyl di(chlorophenyl) phosphonate esters as
inhibitors of PR3.
P
Synthesis and Inhibitory Properties of Ac-peptidyl (O-C H -4-
2
6
4
Cl) Phosphonate Inhibitors and Their Biotinylated Derivatives—
The phosphonate inhibitors and their biotinylated derivatives
(
Table 2) synthesized, purified by HPLC, and checked by mass
spectrometry as described above were assayed for their capac-
ities to inhibit the human neutrophil serine proteases PR3 and
HNE by measuring residual activity after incubation with
increasing amounts of each compound. We used the FRET sub-
strates developed for each protease (ABZ-VADnVADYQ-
EDDnp (27) (where nV stands for norvaline) and ABZ-APQQ-
IMDDQ-EDDnp (28)). PR3 (1 nM final) was almost completely
P
inhibited by incubation for 20 min with 0.5 ␮M Ac-PYDA (O-
C H -4-Cl) (compound 1), whereas HNE was not significantly
6
4
2
P
inhibited. A far higher concentration of Ac-DYDA (O-C H -
6
4
4
-Cl) (compound 2) was necessary to completely inhibit PR3,
2
in keeping with the kinetic data reported above, whereas HNE
was not inhibited. We confirmed these observations by meas-
uring the second order rate constants k /[I] in competition
obs
experiments between the PR3 substrate and each inhibitor (24).
FIGURE 2. Inhibition of wtPR3 and its molecular variants I217R and K99L
P
by increasing the amounts of Ac-PYFA (O-C H -4-Cl) (compound 3) (A) The second order rate constant k /[I] for compound 1 was ϳ80
6
4
2
obs
P
and Ac-PYDA (O-C H -4-Cl) (compound 1) (B) showing the relationship
6
4
2
times greater than that for compound 2 (Table 2). Replacing the
between the S4 and the S2 subsites of PR3.
charged Asp residue at P2 in compound 1 with an uncharged
P
FIGURE 3. Selective inhibition of PR3 by Bt-PYDA (O-C H -4-Cl) (compound 5). A, chemical structure of the biotinylated phosphonate inhibitor Bt-PY-
6
4
2
P
DA (O-C H -4-Cl) (compound5). B, residualactivitiesofPR3andHNEincubatedwithcompound5. PR3(1nM)andHNE(1nM)wereincubatedfor20minat37 °C
6
4
2
with various concentrations of compound 5. Inhibition was monitored by measuring the residual activities of PR3 on ABZ-VADnVADYQ-EDDnp (20 ␮M) and
HNE on ABZ-APQQIMDDQ-EDDnp (20 ␮M). C, Western blot analysis of the irreversible PR3-compound 5 complex. PR3 (10 and 50 nM) and HNE (10 and 50 nM)
were incubated for 20 min at 37 °C with compound 5 (50 nM final), and the mixtures were analyzed by Western blotting (4 ng of PR3 and 20 ng of HNE) and
revealed using extravidin peroxidase.
3
1782 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 289•NUMBER 46•NOVEMBER 14, 2014

Selective PR3 Activity-based Probes
؉
FIGURE 4. Chemical structures of H N-PYDA (O-C H -4-Cl) (compound 8), Ac-PYDA (O-C H -4-Cl) (compound 1), HX-PYDA (O-C H -4-Cl) (com-
P
P
P
2
6
4
2
6
4
2
6
4
2
P
pound 6), and Ac-AHX-PYDA (O-C H -4-Cl) (compound 7). AHX, 6-amino hexanoic acid.
6
4
2
hydrophobic Phe residue (compound 3) reduced the k /[I] phosphonate inhibitors and the active site of PR3. A five-carbon
obs
8
0-fold (Table 2). Whereas compound 3 inhibited recombinant aliphatic chain improved the specificity constant, as much as
K99L, I217R, and wtPR3 to similar extents, compound 1 inhibited biotin, which means that a hydrophobic moiety is enough to
wtPR3 far better than the two PR3 variants, illustrating the close stabilize the inhibitor within the PR3 active site (Table 2). Intro-
relationship between the S2 and S4 subsites (Fig. 2).
ducing PEG as a spacer in compound 5 (compound 9) resulted in
66
We replaced the N-terminal acetyl group by biotin in com- an inhibitor that was not fully selective for PR3 (Table 2) but still
pounds 1 and 2 to generate molecular probes for identifying selectively visualized the PR3-bound biotinylated inhibitor
PR3 activity in liquid biological samples, cells, or tissues (com- despite the fact that biotin was not recognized by extravidin or
pounds 4 and 5). This resulted in significantly greater k /[I] streptavidin in the HNE-compound 9 complex under non-de-
obs
values (Table 2). Thus incubating PR3 (1 nM final) for 20 min naturing/non-reducing conditions (Fig. 5). We next examined
P
with a small molar excess of Bt-PYDA (O-C H -4-Cl) (com- the capacities of these phosphonate inhibitors to inhibit PR3
6
4
2
pound 5) completely inhibited the enzyme, whereas HNE was when it was in a complex biological environment and resisted
not inhibited when the inhibitor concentration was even 100- degradation by other proteases.
P
fold higher (Fig. 3, A and B). The biotinylated inhibitor also
Selective Inhibition of PR3 by Bt-PYDA (O-C H -4-Cl) (com-
6
4
2
enabledustoidentifyPR3byWesternblottingunderreducingand pound 5) in a Human Neutrophil Lysate and in Biological
denaturing conditions but not HNE (Fig. 3C). Non-denaturing Samples—PR3 and its glycosylated forms present in a lysate of
conditions prevented biotin being recognized by extravidin, sug- purified human neutrophils were totally inhibited by Bt-
P
gesting that biotin helps the inhibitor interact with the binding PYDA (O-C H -4-Cl) when assayed under the same experi-
6
4
2
site or stabilizes it. We constructed additional inhibitors with mental conditions as those used for purified proteases (Fig. 6, A
the same peptide sequence but bearing an aliphatic chain, a and B). Furthermore, the inhibitor suffered no loss of activity
positive charge at their N terminus (compounds 6–9), or a when it was incubated with the neutrophil lysate for 20 min
hydrophobic residue (compounds 10 and 11) (Table 2 and Fig. before HPLC analysis, indicating that the inhibitor resisted deg-
4
) to examine the way biotin influences the interaction between radation in an environment replete with proteases (Fig. 6C).
NOVEMBER 14, 2014•VOLUME 289•NUMBER 46
JOURNAL OF BIOLOGICAL CHEMISTRY 31783

Selective PR3 Activity-based Probes
3
1784 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 289•NUMBER 46•NOVEMBER 14, 2014

Selective PR3 Activity-based Probes
bladder cancer (29), but there was no indication as to whether
PR3 was proteolytically active or whether phosphonate inhibi-
tors resist degradation in this heterogeneous environment.
Urine samples from control subjects and patients with a blad-
der cancer (n ϭ 30) were concentrated 20-fold (controls) or
1
0-fold, PR3 activity was measured, and compound 5 was added
before analysis by Western blotting. No activity was detected in
the urines of controls in these experimental conditions. The
active PR3 concentration in the cancer patient samples was
estimated to be 0.1–1 nM based on the rate of substrate hydrol-
ysis. Incubation for 20 min with compound 5 (50 nM final) com-
pletely inhibited the PR3 activity in the concentrated urine
samples (Fig. 7D). Urinary PR3 activity was selectively labeled
and compared well with the data from anti-PR3 antibodies (Fig.
7
, E and F). Thus, the concentration of proteolytically active
PR3 in the urine of patients with bladder cancer is elevated. It
probably originates from cancer-associated neutrophils, as sup-
ported by our observation that other neutrophil-derived mark-
ers such as HNE and myeloperoxidase are also present in the
urine of these patients (not shown).
Characterization of PR3 in Purified Neutrophils—Because
the biotin in compound 9 still reacts with extravidin/streptavi-
FIGURE 6. Selective inhibition of PR3 in a neutrophil lysate. A, PR3 (ϳ1 nM)
P
and HNE (ϳ1 nM) were incubated for 20 min at 37 °C with 10 nM Bt-PYDA (O- din after the inhibitor has interacted with its target protease, we
C H -4-Cl) (compound 5 (CP5)), and the inhibition was monitored by mea-
6
4
2
looked at the intracellular distribution of PR3 in purified per-
meabilized neutrophils. The whole cytoplasm of permeabilized
suring the hydrolysis of 20 ␮M ABZ-VADnVADYQ-EDDnp (PR3) or 20 ␮M
ABZ-APQQIMDDQ-EDDnp (HNE). B, detection of the irreversible complex of
compound 5 with purified PR3 and PR3 in neutrophil lysate using extravidin- neutrophils was intensely labeled but not the perinuclear envi-
peroxidase: samples containing ϳ50 nM PR3 were incubated for 20 min at
ronment, as previously observed for HNE (30) (Fig. 8 and
3
7 °C with compound 5 (50 nM final). The resulting products were separated
by electrophoresis and transferred to a nitrocellulose membrane. C, stability supplemental Video 1). There was no PR3 labeling at the cell
of compound 5 (50 nM) (solid line) after incubation for 20 min at 37 °C with a
neutrophil lysate (dotted line) and analysis by HPLC at 220 nm. The arrow
indicates compound 5.
surface which confirms that the small variable fraction of PR3
constitutively exposed on the surface of quiescent neutrophils
is inactive (31). Primed, activated neutrophils bear an “induced
We then analyzed the supernatants of 10 sputa from CF pool of active PR3” on their membranes, the binding of which
patients for their PR3 activity before and after incubation with clearly differs from that of HNE and affects the kinetic proper-
m
compound 5 (150 nM final). All the biological samples cleaved ties of the membrane-bound PR3 (PR3 ) (32). This results in
the selective PR3 substrate in the absence of inhibitor, and increased resistance to inhibition by ␣-1-proteinase inhibitor,
cleavage was always at the PR3-specific bond. The estimated its main natural inhibitor in plasma (31). The resistance has
PR3 concentration of these samples was in the 0.1–1 ␮M range been attributed to steric hindrance of interactions with high
based on the rate of substrate hydrolysis. Substrate hydrolysis molecular mass (50 kDa) serpin (33, 34). We examined this
was totally inhibited after incubation with compound 5 for 20 hypothesis using the cell-impermeable low M compound 1 to
r
min, whereas HNE activity remained unchanged (Fig. 7A). inhibit the PR3 on the surface of activated neutrophils purified
Compound 5 was also used to selectively label PR3 activity in from control subjects and patients with GPA. The phosphonate
CF sputum, and the findings compared with those obtained inhibitor behaved much like ␣-1-proteinase inhibitor as,
with anti-PR3 antibodies (Fig. 7, B and C). The inhibitor was not despite its very small size (0.7 kDa), a large molar excess of
m
broken down by crude supernatants, and the rate of PR3 inac- inhibitor was needed to fully inhibit PR3 (Fig. 9A). We used
tivation was comparable to that of purified PR3. Thus, PR3 flow cytometry and compound 9 to show that the PR3-inhibitor
inhibition was not impaired by other soluble compounds in CF complexes remained at the surface of activated neutrophils
sputa. The same results were obtained with crude bronchoal- (Fig. 9B). Because the membrane binding site on the PR3 sur-
veolar fluid samples (n ϭ 5) from patients with idiopathic pul- face involves a hydrophobic patch located near the S4 subsite
m
monary fibrosis (not shown).
(34, 35), we postulated that PR3 was in a resting state confor-
We also used the urine of patients with bladder cancer as a mation that prevented small molecule inhibitors from rapidly
m
model for detecting proteolytically active PR3 by compound 5. interacting with PR3 (Fig. 9C). We checked this by comparing
PR3 has been shown to be an antigenic marker of aggressive the K values of the interactions between a peptide PR3 sub-
m
FIGURE 5. Selective recognition of the PR3-compound 9 complex by extravidin/streptavidin under non-denaturing/non-reducing conditions. A,
P
chemical structures of Bt-[PEG]66-PYDA (O-C H -4-Cl) (compound 9). B, visualization by ELISA of the PR3-compound 9 (CP5) complex as revealed with strepta-
6
4
2
vidin-HRP (1/2000). C, Western blot analysis of the PR3-compound 9 complex under non-denaturing/non-reducing conditions. PR3 and HNE (50 and 100 nM)
were incubated for 20 or 180 min at 37 °C with compound 9 (150 nM), and the mixtures were analyzed by Western blotting using extravidin-peroxidase. D,
selectiveinteractionofbiotinylatedcompound9withPR3inaneutrophillysateasshownbyWesternblottingundernon-denaturing/non-reducingconditions
and extravidin-peroxidase staining.
NOVEMBER 14, 2014•VOLUME 289•NUMBER 46
JOURNAL OF BIOLOGICAL CHEMISTRY 31785

Selective PR3 Activity-based Probes
FIGURE 7. Selective inhibition of PR3 in biological samples. A, inhibition of PR3 in a representative CF sputum supernatant. The volume of CF sputum was
adjusted to give final concentrations of PR3 and HNE of ϳ1 nM. Diluted samples were incubated with compound 5 (Cp5, 50 nM final) for 20 min at 37 °C. The
residual activities of PR3 and HNE were measured as described in Fig. 3. Purified PR3 (1 nM) and HNE (1 nM) were used as controls. B, selective labeling of PR3
activityinCFsputum. CFvolumeswereadjustedtocontain4or20ngPR3(10and50nM final)andincubatedfor20minat37 °Cwithcompound5(150nM final),
and the products were revealed with extravidin peroxidase. C, Western blot analysis of CF sputa with anti-PR3 antibodies showing active/inactive/degraded
PR3andadditionalbandsduetoantibodiesinteractingwithunrelatedCFsputumproteins. D, inhibitionofPR3inconcentrated(10ϫ)urinefromthreepatients
with bladder cancer. Proteolytically active PR3 in unconcentrated urine was ϳ0.1–1 nM based on the rate of substrate hydrolysis. E, selective labeling of PR3
activity in concentrated (10ϫ) urine (U) samples. F, Western blot analysis of concentrated urine samples (10ϫ) using anti-PR3 antibodies.
strate, structurally similar to the phosphonate inhibitor, mea- as proteases with almost indistinguishable substrate specifici-
sured with the soluble and the membrane-bound proteases. ties. But evidence is slowly accumulating that each protease has
The K (determined by Hanes plot of the hydrolysis of PR3 a unique and non-redundant biological function that is mir-
m
m
substrate) for PR3 was ϳ6 times higher than that for soluble rored by subtle differences in their overall structure. These
m
PR3 indicating that the substrate binding region of PR3 is not affect their subcellular distributions and their substrate speci-
freely accessible to substrates or inhibitors, unlike that of the ficities (3, 36). For example, human PR3 has a hydrophobic
soluble, active form. By contrast, the K of an HNE peptide patch on its surface that is responsible for its exposure on the
m
substrate (ABZ-APQQIMDDQ-EDDnp) was the same when it surface of quiescent neutrophils, whereas HNE does not. This
was measured in a suspension of activated neutrophils and with probably explains why PR3 is a disease-specific major autoan-
purified HNE (not shown).
tigen in GPA (35). The substrate binding specificities of PR3
and HNE extend to both sides of the cleaved peptide bond, so
that subsites from S4 to S2Ј influence the rate of substrate cleav-
DISCUSSION
PR3 and HNE are structurally very similar, and both are pro- age and sensitivity to inhibitors (20, 37).
duced, stored and secreted at the same time by the same leuko-
Their stability, their lack of toxicity, and their selectivity
cytes. Both cleave protein substrates preferentially after small make peptide derivatives of aminoalkyl phosphonate diphenyl
hydrophobic residues like Val/Ala and have long been regarded esters convenient tools for in vitro and in vivo studies on the
3
1786 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 289•NUMBER 46•NOVEMBER 14, 2014

Selective PR3 Activity-based Probes
NOVEMBER 14, 2014•VOLUME 289•NUMBER 46
JOURNAL OF BIOLOGICAL CHEMISTRY 31787

Selective PR3 Activity-based Probes
m
m
FIGURE 9. Inhibition and labeling of PR3 . A, residual activity of soluble (gray bars) and PR3 from purified neutrophils of a healthy individual (black bars) and
a GPA patient (white bars) in the presence of increased concentrations of compound 1. Purified neutrophils were activated with A23187 (1 ␮M) for 15 min at
m
3
7 °C (25). Purified PR3 (1 nM) and PR3 (ϳ1 nM based on the rate of substrate hydrolysis by ϳ1 million cells) were incubated with a range of concentrations of
m
compound 1 for 20 min at RT in PBS, and residual activities were measured fluorometrically. A large molar excess of compound 1 was required to inhibit PR3
on activated neutrophils. B, labeling of cell surface PR3 . Shown is a representative flow cytometry analysis of A23187-activated neutrophils before and after
m
incubationwithcompound9. Activatedneutrophilswereincubatedfor20minatRTwith50 ␮M compound9andvisualizedwithstreptavidin-AlexaFluorꢁ488.
The gray (mean fluorescence intensity ϭ 562) peak corresponds to unspecific binding of streptavidin-Alexa Fluorꢁ 488, and the black (mean fluorescence
m
intensity ϭ 1439) peak corresponds to labeled active PR3 . The 2.5 Ϯ 0.4 (mean value of 4 independent experiments) -fold increase in mean fluorescence
m
intensity shows that induced PR3 interacts with compound 9 and that the complex lies on the cell surface. C, proposed explanation for the low activity of
m
m
induced PR3 and its stable exposure on activated neutrophils. Although constitutive PR3 is proteolytically inactive, induced PR3 , which may be co-localized
withNB1(CD177)inlipiddraftsontheneutrophilsurface(51), isinarestingstateconformationthatimpairstheproductivebindingofsubstratesandinhibitors
m
(
left). An excess of substrate or inhibitor favors the conversion of latent PR3 to its fully active form (right).
function and activity of serine proteases during the course of a atom, thus making it more susceptible to the nucleophilic
variety of diseases (14). The phosphonate moiety of these irre- attack of the active site serine residue (38). The selectivity of
versible inhibitors interacts with the S1Ј site and forms a cova- phosphonate inhibitors depends only on the non-prime speci-
lent bond with Ser-195. We have introduced an electron-with- ficity of the target protease, whereas the C-terminal phospho-
drawing chlorine atom in the para position of the phenyl ester nate group occupies the S1Ј subsite of the protease. Selective
ring to increase the electrophilic potential of the phosphorous PR3 substrates have exploited both the prime and non-prime
FIGURE 8. Selective labeling of intracellular PR3 in fixed, permeabilized quiescent neutrophils. A, residual activities of PR3 and HNE in permeabilized cells
incubated for 90 min at RT with the polyvalent serine protease inhibitor Ala-Ala-Pro-Val-CH Cl (AAPV-CH Cl, 2 mM) for 20 min at RT with compound 9 (Cp9, 10
2
2
␮M), or with the elastase inhibitor EPI-hNE-4 (1 ␮M). Inhibition was monitored using 100,000–150,000 quiescent cells. The residual activity of PR3 was measured
using ABZ-VADnVADYQ-EDDnp (20 ␮M final) and that of HNE with ABZ-APQQIMDDQ-EDDnp (20 ␮M final). B, side (SS)- and forward (FS)-scatter of fixed,
permeabilized purified neutrophils showing the homogeneity of the neutrophil population. C, flow cytometry analysis of permeabilized neutrophils that had
been incubated with compound 9. The fluorescence of cells increased dramatically after incubation with compound 9 (green peak, mean fluorescence
intensity ϭ 67073), but it increased much less when the cells had been first treated with the polyvalent serine protease inhibitor Ala-Ala-Pro-Val-CH Cl (blue
2
peak, mean fluorescence intensity ϭ 1900). D, fluorescence remained unchanged in cells preincubated with the elastase inhibitor EPI-hNE-4 (50) (magenta
peak). The black peak shows the fluorescence of permeabilized neutrophils in the presence of streptavidin-Alexa Fluorꢁ 488. E, confocal microscopy of
permeabilized neutrophils treated first with EPI-hNE-4 and then incubated with compound 9 (right). Confocal microscope image stacks are as revealed with
DAPI and streptavidin-Alexa Fluorꢁ 488.
3
1788 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 289•NUMBER 46•NOVEMBER 14, 2014

Selective PR3 Activity-based Probes
specificity pockets (27). We wanted to produce PR3 inhibitors vated neutrophils. This particular binding depends on a hydro-
that consisted only of non-prime residues and discriminated phobic patch located beyond the S4 subsite of the protease
between PR3 and HNE. We, therefore, investigated the non- (Phe-165, Phe-166, Phe-215, Ile-217, Trp-218, Phe-224) (26).
prime subsite specificity of PR3 using two PR3 mutants that The other neutrophilic serine proteases, HNE and cathepsin G,
mimic the S4 and S2 subsites of HNE, which are the most crit- lack this hydrophobic patch (35, 44) as does rodent PR3. This
ical non-prime sites for distinguishing between PR3 and HNE difference between humans and rodents seriously complicates
(
26).
the use of rodents for studies on PR3 in vivo (45). We previously
We used several FRET substrates that differed at the P4 and reported a weak PR3 activity in suspensions of activated human
m
P2 positions to compare wtPR3, the two mutants of PR3 and neutrophils (31). We also observed that induced PR3 was
HNE, followed by peptidyl-pNA substrates that have no prime poorly inhibited by its main natural plasma inhibitor ␣-1-pro-
side residues. These studies revealed a close cooperation teinase inhibitor and attributed this to steric hindrance at the
between the S4 and the S2 subsites, which we exploited in our cell surface. We now find that low molecular weight phospho-
design of two short tetrapeptide sequences that are selectively nate inhibitors behave in the same way as ␣-1-proteinase inhib-
recognized by PR3. DYDA and PYDA were chosen as the P4-P1 itor and that a large molar excess of inhibitor is need to fully
sequences to synthesize peptide derivatives of aminoalkyl phos- inhibit PR3 in a suspension of activated neutrophils. This phe-
phonate diphenyl esters. The two tetrapeptide derivatives were nomenon cannot be readily accounted for by steric hindrance.
selective for PR3, but the second order inhibition constant k
/
The PR3 hydrophobic patch should not impair access of low M
obs
r
[
I] for the prolyl-containing inhibitor was 100 times larger, in phosphonate inhibitors to the PR3 active site cleft. The weak
m
keeping with the lesser ability of the PR3 S4 subsite to accom- PR3 activity and its resistance to inhibition might be due to its
modate a negatively charged residue. Unexpectedly, biotinyla- conformation being distorted; we recently used the monoclonal
P
tion of the N terminus of Ac-PYDA (O-C H -4-Cl) improved Ab MCPR3–7 to show that PR3 is in equilibrium between a
6
4
2
the k /[I] ϳ25 times. This was probably because biotin highly favored active conformation and an inactive conforma-
obs
improved the binding of the inhibitor to the protease. This idea tion (46). This antibody inhibits PR3 activity by an allosteric
is supported by the finding that biotin interacts poorly with mechanism and competes with the neutrophil membrane for
extravidin/streptavidin once the inhibitor is bound to PR3. We the binding of PR3. Interaction of the hydrophobic patch of
m
also showed that attaching a short aliphatic sequence like PR3 with NB1 (CD177) or lipid rafts on the neutrophil surface
hexanoic acid to the P4 prolyl residue of inhibitors improved might well employ the same mechanism to affect the substrate
ϩ
inhibitor binding, whereas introducing a NH Pro at P4 greatly and inhibitor binding subsites. The conversion of the resting or
2
impaired the interaction. This apparently contradicts previous latent conformation to the fully active form would be induced
ϩ
results showing that an NH -terminal residue poorly influ- by an excess of substrate or inhibitor. This is consistent with
3
enced the cleavage of FRET substrates bearing a prime exten- our observation that the K for cleavage of a peptide substrate
m
sion and no Asp at P2 (20). Stabilizing the short tetrapeptide by PR3 in a suspension of neutrophils is about six times greater
m
sequence, as was done here depends greatly on the Asp at P2. than that of PR3 in solution. PR3 , therefore, seems to function
This stabilization is most certainly compromised by the pres- much like other serine proteases, granzyme K, FIXa, DF, degP,
ϩ
ence of NH Pro in the environment of P2 ϭ Asp. This further and ␣-tryptase that have a resting state conformation due to
2
emphasizes the cooperation between the S4 and S2 subsites in one or more distorted loops surrounding the active site cleft
PR3.
that impair the productive binding of substrates and inhibitors
We also verified that phosphonate inhibitors are still active (47).
m
and resist degradation in a milieu replete with proteases, such
The relative resistance of induced PR3 to inhibition and its
as neutrophil lysates, bronchoalveolar fluids, and sputa from apparently reduced enzymatic activity make it different from
patients with inflammatory lung diseases or the urine of
patients with bladder cancer. A spacer like polyethylene glycol
placed between the biotin moiety and the inhibitor P4 residue
preserved inhibitory properties and allowed the biotin moiety
to be recognized by extravidin/streptavidin. Thus the biotiny-
lated inhibitor can be used as a probe for visualizing PR3 activity
in situ in permeabilized neutrophils by confocal microscopy.
Unlike HNE, whose distribution is mainly perinuclear (30, 39),
PR3 activity is present in punctate areas throughout the cyto-
plasm of neutrophils. This agrees with the results obtained by
immunocytochemistry showing that the PR3 autoantigen, the
target of cANCA (anti-neutrophil cytoplasmic antibodies), is
distributed throughout the cytoplasm, whereas the HNE anti-
gen (40, 41), the target of some pANCA (42, 43), is concentrated
in the perinuclear region.
HNE, which remains fully active at the neutrophil surface and is
easily removed from the surface by ␣-1-proteinase inhibitor
m
(
31, 48). This peculiar behavior of PR3 results in the pro-
longed exposure of a latent form of the protease at the cell
surface during inflammatory episodes. This exposure of a latent
activatable form plus the permanent, genetically determined,
exposure of inactive PR3 at the cell surface could well explain
why this protease is the major autoantigen mediating the bind-
ing of anti-PR3 antibodies and the subsequent activation of
blood neutrophils in GPA.
Acknowledgments—We acknowledge the Association Vaincre La
Mucoviscidose (AVLM). We thank Lise Vanderlynden (INSERM
U-1100), Julien Burlaud-Gaillard (D e´ partement des Microscopies,
Universit e´ Francois Rabelais), and Heike Reimann (Comprehensive
Pneumology Center, Institute of Lung Biology and Disease) for tech-
nical assistance. The English text was edited by Dr. Owen Parkes.
The most striking feature of PR3 is its constitutive expression
as a proteolytically inactive form at the neutrophil surface and
the peculiar binding of its induced form to the surface of acti-
NOVEMBER 14, 2014•VOLUME 289•NUMBER 46
JOURNAL OF BIOLOGICAL CHEMISTRY 31789

Selective PR3 Activity-based Probes
REFERENCES
derivatives: the influence of the ester group on their biological activity.
Bioorg. Med. Chem. Lett. 16, 2886–2890
1
. Turk, B. (2006) Targeting proteases: successes, failures, and future pros-
pects. Nat. Rev. Drug Discov. 5, 785–799
2
2. Oleksyszyn, J., and Powers, J. C. (1989) Irreversible inhibition of serine
proteases by peptidyl derivatives of ␣-aminoalkylphosphonate diphenyl
esters. Biochem. Biophys. Res. Commun. 161, 143–149
2
. Korkmaz, B., Moreau, T., and Gauthier, F. (2008) Neutrophil elastase,
proteinase 3, and cathepsin G: physicochemical properties, activity, and
physiopathological functions. Biochimie 90, 227–242
2
3. Korkmaz, B., Attucci, S., Hazouard, E., Ferrandiere, M., Jourdan, M. L.,
Brillard-Bourdet, M., Juliano, L., and Gauthier, F. (2002) Discriminating
between the activities of human neutrophil elastase and proteinase 3 using
serpin-derived fluorogenic substrates. J. Biol. Chem. 277, 39074–39081
4. Tian, W. X., and Tsou, C. L. (1982) Determination of the rate constant of
enzyme modification by measuring the substrate reaction in the presence
of the modifier. Biochemistry 21, 1028–1032
3
4
. Korkmaz, B., Horwitz, M. S., Jenne, D. E., and Gauthier, F. (2010) Neutro-
phil elastase, proteinase 3, and cathepsin G as therapeutic targets in hu-
man diseases. Pharmacol. Rev. 62, 726–759
2
2
. Owen, C. A. (2008) Roles for proteinases in the pathogenesis of chronic
obstructive pulmonary disease. Int. J. Chron. Obstruct. Pulmon. Dis. 3,
2
53–268
5. Korkmaz, B., Attucci, S., Juliano, M. A., Kalupov, T., Jourdan, M. L., Ju-
liano, L., and Gauthier, F. (2008) Measuring elastase, proteinase 3, and
cathepsin G activities at the surface of human neutrophils with fluores-
cence resonance energy transfer substrates. Nat. Protoc. 3, 991–1000
6. Fujinaga, M., Chernaia, M. M., Halenbeck, R., Koths, K., and James, M. N.
5
6
7
. Jenne, D. E., Tschopp, J., Ludemann, J., Utecht, B., and Gross, W. L. (1990)
Wegener’s autoantigen decoded. Nature 346, 520
. Kallenberg, C. G. (2008) Pathogenesis of PR3-ANCA associated vasculitis.
J. Autoimmun. 30, 29–36
2
2
2
. Csernok, E., and Gross, W. L. (2013) Current understanding of the patho-
genesis of granulomatosis with polyangiitis (Wegener’s). Expert Rev. Clin.
Immunol. 9, 641–648
(
1996) The crystal structure of PR3, a neutrophil serine proteinase antigen
of Wegener’s granulomatosis antibodies. J. Mol. Biol. 261, 267–278
7. Korkmaz, B., Attucci, S., Moreau, T., Godat, E., Juliano, L., and Gauthier, F.
8
. Gabillet, J., Millet, A., Pederzoli-Ribeil, M., Tacnet-Delorme, P., Guillevin,
L., Mouthon, L., Frachet, P., and Witko-Sarsat, V. (2012) Proteinase 3, the
autoantigen in granulomatosis with polyangiitis, associates with calreticu-
lin on apoptotic neutrophils, impairs macrophage phagocytosis, and pro-
motes inflammation. J. Immunol. 189, 2574–2583
(
2004) Design and use of highly specific substrates of neutrophil elastase
and proteinase 3. Am. J. Respir. Cell Mol. Biol. 30, 801–807
8. Derache, C., Epinette, C., Roussel, A., Gabant, G., Cadene, M., Korkmaz,
B., Gauthier, F., and Kellenberger, C. (2012) Crystal structure of greglin, a
novel non-classical Kazal inhibitor, in complex with subtilisin. FEBS J.
9
. Witko-Sarsat, V., Cramer, E. M., Hieblot, C., Guichard, J., Nusbaum, P.,
Lopez, S., Lesavre, P., and Halbwachs-Mecarelli, L. (1999) Presence of
proteinase 3 in secretory vesicles: evidence of a novel, highly mobilizable
intracellular pool distinct from azurophil granules. Blood 94, 2487–2496
2
79, 4466–4478
2
9. Zoidakis, J., Makridakis, M., Zerefos, P. G., Bitsika, V., Esteban, S., Frantzi,
M., Stravodimos, K., Anagnou, N. P., Roubelakis, M. G., Sanchez-Carbayo,
M., and Vlahou, A. (2012) Profilin 1 is a potential biomarker for bladder
cancer aggressiveness. Mol. Cell Proteomics 10.1074/mcp.M111.
1
0. Halbwachs-Mecarelli, L., Bessou, G., Lesavre, P., Lopez, S., and Witko-
Sarsat, V. (1995) Bimodal distribution of proteinase 3 (PR3) surface ex-
pression reflects a constitutive heterogeneity in the polymorphonuclear
neutrophil pool. FEBS Lett. 374, 29–33
0
09449 1–15
3
3
3
0. Kasperkiewicz, P., Poreba, M., Snipas, S. J., Parker, H., Winterbourn, C. C.,
Salvesen, G. S., and Drag, M. (2014) Design of ultrasensitive probes for
human neutrophil elastase through hybrid combinatorial substrate library
profiling. Proc. Natl. Acad. Sci. U.S.A. 111, 2518–2523
1
1
1
1. Schreiber, A., Busjahn, A., Luft, F. C., and Kettritz, R. (2003) Membrane
expression of proteinase 3 is genetically determined. J. Am. Soc. Nephrol.
14, 68–75
1. Korkmaz, B., Jaillet, J., Jourdan, M. L., Gauthier, A., Gauthier, F., and
Attucci, S. (2009) Catalytic activity and inhibition of wegener antigen pro-
teinase 3 on the cell surface of human polymorphonuclear neutrophils.
J. Biol. Chem. 284, 19896–19902
2. Korkmaz, B., Kellenberger, C., Viaud-Massuard, M. C., and Gauthier, F.
(
2013) Selective inhibitors of human neutrophil proteinase 3. Curr.
Pharm. Des. 19, 966–976
3. Sie n´ czyk, M., and Oleksyszyn, J. (2009) Irreversible inhibition of serine
proteases: design and in vivo activity of diaryl ␣-aminophosphonate de-
rivatives. Curr. Med. Chem. 16, 1673–1687
2. Korkmaz, B., Lesner, A., Letast, S., Mahdi, Y. K., Jourdan, M. L., Dallet-
Choisy, S., Marchand-Adam, S., Kellenberger, C., Viaud-Massuard, M. C.,
Jenne, D. E., and Gauthier, F. (2013) Neutrophil proteinase 3 and dipepti-
dyl peptidase I (cathepsin C) as pharmacological targets in granulomatosis
with polyangiitis (Wegener granulomatosis). Semin. Immunopathol. 35,
1
4. Grzywa, R., and Sie n´ czyk, M. (2013) Phosphonic esters and their applica-
tion of protease control. Curr. Pharm. Des. 19, 1154–1178
5. Boduszek, B., Brown, A. D., and Powers, J. C. (1994) ␣-Aminoalkylphos-
phonate di(chlorophenyl) esters as inhibitors of serine proteases. J. En-
zyme Inhib. 8, 147–158
1
4
11–421
3
3. Hajjar, E., Mihajlovic, M., Witko-Sarsat, V., Lazaridis, T., and Reuter, N.
(
2008) Computational prediction of the binding site of proteinase 3 to the
1
6. Brown, C. M., Ray, M., Eroy-Reveles, A. A., Egea, P., Tajon, C., and Craik,
C. S. (2011) Peptide length and leaving-group sterics influence potency of
peptide phosphonate protease inhibitors. Chem. Biol. 18, 48–57
7. Woodard, S. L., Jackson, D. S., Abuelyaman, A. S., Powers, J. C., Winkler,
U., and Hudig, D. (1994) Chymase-directed serine protease inhibitor that
reacts with a single 30-kDa granzyme and blocks NK-mediated cytotox-
icity. J. Immunol. 153, 5016–5025
plasma membrane. Proteins. 71, 1655–1669
3
3
4. Broemstrup, T., and Reuter, N. (2010) How does proteinase 3 interact with
lipid bilayers? Phys. Chem. Chem. Phys. 12, 7487–7496
1
5. Korkmaz, B., Kuhl, A., Bayat, B., Santoso, S., and Jenne, D. E. (2008) A
hydrophobic patch on proteinase 3, the target of autoantibodies in We-
gener granulomatosis, mediates membrane binding via NB1 receptors.
J. Biol. Chem. 283, 35976–35982
1
8. Schechter, I., and Berger, A. (1967) On the size of the active site in pro-
teases. I. Papain. Biochem. Biophys. Res. Commun. 27, 157–162
9. J e´ got, G., Derache, C., Castella, S., Lahouassa, H., Pitois, E., Jourdan, M. L.,
36. Hajjar, E., Broemstrup, T., Kantari, C., Witko-Sarsat, V., and Reuter, N.
(2010) Structures of human proteinase 3 and neutrophil elastase: so sim-
ilar yet so different. FEBS J. 277, 2238–2254
1
Remold-O’Donnell, E., Kellenberger, C., Gauthier, F., and Korkmaz, B. 37. Hajjar, E., Korkmaz, B., and Reuter, N. (2007) Differences in the substrate
(
2011) A substrate-based approach to convert SerpinB1 into a specific
binding sites of murine and human proteinase 3 and neutrophil elastase.
FEBS Lett. 581, 5685–5690
inhibitor of proteinase 3, the Wegener’s granulomatosis autoantigen.
FASEB J. 25, 3019–3031
38. Sie n´ czyk, M., Podgórski, D., Bła¿ejewska, A., Kulbacka, J., Saczko, J., and
Oleksyszyn, J. (2011) Phosphonic pseudopeptides as human neutrophil
elastase inhibitors: a combinatorial approach. Bioorg. Med. Chem. 19,
1277–1284
2
0. Korkmaz, B., Hajjar, E., Kalupov, T., Reuter, N., Brillard-Bourdet, M.,
Moreau, T., Juliano, L., and Gauthier, F. (2007) Influence of charge distri-
bution at the active site surface on the substrate specificity of human
neutrophil protease 3 and elastase: a kinetic and molecular modeling anal-
ysis. J. Biol. Chem. 282, 1989–1997
39. Clark, J. M., Vaughan, D. W., Aiken, B. M., and Kagan, H. M. (1980)
Elastase-like enzymes in human neutrophils localized by ultrastructural
cytochemistry. J. Cell Biol. 84, 102–119
2
1. Sie n´ czyk, M., and Oleksyszyn, J. (2006) Inhibition of trypsin and urokinase
by Cbz-amino(4-guanidinophenyl)methanephosphonate aromatic ester 40. Benson, K. F., Li, F. Q., Person, R. E., Albani, D., Duan, Z., Wechsler, J.,
3
1790 JOURNAL OF BIOLOGICAL CHEMISTRY
VOLUME 289•NUMBER 46•NOVEMBER 14, 2014

Selective PR3 Activity-based Probes
Meade-White, K., Williams, K., Acland, G. M., Niemeyer, G., Lothrop, 46. Hinkofer, L. C., Seidel, S. A., Korkmaz, B., Silva, F., Hummel, A. M., Braun,
C. D., and Horwitz, M. (2003) Mutations associated with neutropenia in
dogs and humans disrupt intracellular transport of neutrophil elastase.
Nat. Genet. 35, 90–96
D., Jenne, D. E., and Specks, U. (2013) A monoclonal antibody (MCPR3–7)
interfering with the activity of proteinase 3 by an allosteric mechanism.
J. Biol. Chem. 288, 26635–26648
4
1. K o¨ llner, I., Sodeik, B., Schreek, S., Heyn, H., von Neuhoff, N., Germe-
shausen, M., Zeidler, C., Kr u¨ ger, M., Schlegelberger, B., Welte, K., and
Beger, C. (2006) Mutations in neutrophil elastase causing congenital neu-
tropenia lead to cytoplasmic protein accumulation and induction of the
unfolded protein response. Blood 108, 493–500
4
7. Hink-Schauer, C., Est e´ banez-Perpi n˜ a´ , E., Wilharm, E., Fuentes-Prior, P.,
Klinkert, W., Bode, W., and Jenne, D. E. (2002) The 2.2-A crystal structure
of human pro-granzyme K reveals a rigid zymogen with unusual features.
J. Biol. Chem. 277, 50923–50933
4
8. Korkmaz, B., Attucci, S., Jourdan, M. L., Juliano, L., and Gauthier, F. (2005)
Inhibition of neutrophil elastase by ␣1-protease inhibitor at the surface of
human polymorphonuclear neutrophils. J. Immunol. 175, 3329–3338
9. Bode, W., Wei, A. Z., Huber, R., Meyer, E., Travis, J., and Neumann, S.
4
4
4
2. Csernok, E., L u¨ demann, J., Gross, W. L., and Bainton, D. F. (1990) Ultra-
structural localization of proteinase 3, the target antigen of anti-cytoplas-
mic antibodies circulating in Wegener’s granulomatosis. Am. J. Pathol.
4
137, 1113–1120
(
(
1986) X-ray crystal structure of the complex of human leukocyte elastase
PMN elastase) and the third domain of the turkey ovomucoid inhibitor.
3. Tervaert, J. W., Mulder, L., Stegeman, C., Elema, J., Huitema, M., The, H.,
and Kallenberg, C. (1993) Occurrence of autoantibodies to human leuco-
cyte elastase in Wegener’s granulomatosis and other inflammatory disor-
ders. Ann. Rheum. Dis. 52, 115–120
EMBO J. 5, 2453–2458
5
0. Attucci, S., Gauthier, A., Korkmaz, B., Del e´ pine, P., Martino, M. F.,
Saudubray, F., Diot, P., and Gauthier, F. (2006) EPI-hNE4, a proteolysis-
resistant inhibitor of human neutrophil elastase and potential anti-inflam-
matory drug for treating cystic fibrosis. J. Pharmacol. Exp. Ther. 318,
4. Kantari, C., Millet, A., Gabillet, J., Hajjar, E., Broemstrup, T., Pluta, P.,
Reuter, N., and Witko-Sarsat, V. (2011) Molecular analysis of the mem-
brane insertion domain of proteinase 3, the Wegener’s autoantigen, in
RBL cells: implication for its pathogenic activity. J. Leukoc. Biol. 90,
8
03–809
9
41–950
51. von Vietinghoff, S., Tunnemann, G., Eulenberg, C., Wellner, M., Cristina
Cardoso, M., Luft, F. C., and Kettritz, R. (2007) NB1 mediates surface
expression of the ANCA antigen proteinase 3 on human neutrophils.
Blood 109, 4487–4493
4
5. Korkmaz, B., Jenne, D. E., and Gauthier, F. (2013) Relevance of the mouse
model as a therapeutic approach for neutrophil proteinase 3-associated
human diseases. Int. Immunopharmacol. 17, 1198–1205
NOVEMBER 14, 2014•VOLUME 289•NUMBER 46
JOURNAL OF BIOLOGICAL CHEMISTRY 31791