N-Acetyl-Pro-Gly-Pro
(acPGP) A biomarker and therapeutic
target for neutrophilic inflammatory diseases
A novel peptide
CXCR ligand derived from extracellular matrix degradation
during airway inflammation
We describe the
tripeptide neutrophil chemoattractant N-acetyl Pro-Gly-Pro
(PGP), derived from the breakdown of extracellular matrix
(ECM), which shares sequence and structural homology
with an important domain on alpha chemokines. PGP caused
chemotaxis and production of superoxide through CXC
receptors, and administration of peptide caused recruitment
of neutrophils (PMNs) into lungs of control, but not
CXCR2-deficient mice. PGP was generated in mouse lung
after exposure to lipopolysaccharide, and in vivo
and in vitro blockade of PGP with monoclonal
antibody suppressed PMN responses as much as chemokine-specific
monoclonal antibody. Extended PGP treatment caused alveolar
enlargement and right ventricular hypertrophy in mice.
PGP was detectable in substantial concentrations in
a majority of bronchoalveolar lavage samples from individuals
with chronic obstructive pulmonary disease, but not
control individuals. Thus, PGP's activity links degradation
of ECM with neutrophil recruitment in airway inflammation,
and PGP may be a biomarker and therapeutic target for
neutrophilic inflammatory diseases.
A peptide fragment derived from extracellular matrix
collagen attracts inflammatory cells by mimicking the
activity of a chemokine—potentially contributing
to lung injury.
Send in the neutrophils.
Early in an inflammatory response, neutrophils react
to CXC chemokines, such as IL-8 in people, by following
the chemotatic gradient and emigrating into the interstitial
space. This chemotatic response depends upon the ability
of IL-8 to engage either of its two receptors, CXCR1
and CXCR2. Proteolytic cleavage of collagen, presumably
by matrix metalloproteinases (MMPs) and subsequent acetylation,
generates an acetylated tripeptide, acPGP, which mimics
key sequences of IL-8, stimulates CXCR1 and CXCR2 and
prolongs influx of neutrophils. Peter M Henson & R William
Vandivier. Nature Medicine,12, 280 - 281 (2006)
(a) The PGP peptide, but not the PGG
control peptide causes an increase in airway PMNs after
intratracheal administration (250 g
in 50 l)
to 6–8-week-old C57 BL/6J mice; BAL and differential
cell counts were performed 5 h after treatment (*P
< 0.05 compared to PBS group). (b)
PGP is a chemoattractant of human PMNs in an in
vitro transwell chemotaxis assay, whereas PGG had
no chemotactic activity; 2
105 PMNs were placed on one side of a 3 m
filter with peptides present on the other side, and
migrated cells were photographed and quantified after
1 h at 37 °C. (c) Dose-response curves
for human PMNs (open squares), HL-60 cells (filled squares),
bone marrow–derived mouse PMNs (open triangles)
and mouse peritoneal PMNs (filled triangles). Weathington N.M. , et al. Nature
Medicine12, 317 - 323 (2006)
Structural homology between PGP and the neutrophil chemokines.
(a) Alignment of a collagen fragment
isolated from arthritic inflammatory tissue and possessing
the PGP sequence19 with similar sequences contained
in neutrophil-attracting ELR+ CXC chemokine family members.
Chemokines shown are rat, mouse and human CXCL1, CXCL2
and CXCL3 and human CXCL8 and representative ELR- CXCLs
and CCLs lacking activity on PMN CXCR1 and CXCR2. ELR
sequences are green, the GP-containing sequences from
collagen and the chemokines are red, and structural
cysteines are blue. The nuclear magnetic resonance imaging
solution structure (b) of PGP (referring
to the N-Acetyl-PGP molecule22) is similar to the SGP
motif found in the structure of human IL-8 (ref. 21)
and related chemokines (c). This motif
is required for chemokine binding and neutrophil activation20,
and is shown (d; yellow) to be solution
accessible in the IL-8 structure and close in space
to the important ELR motif (green) present in the neutrophil
specific CXC chemokines. For reference, the structural
cysteines are shown in red. Weathington N.M. , et al. Nature
Medicine12, 317 - 323 (2006)
Extended exposure to PGP causes alveolar enlargement
and right ventricular hypertrophy. C57Bl/6
mice were intranasally treated with PGP (250 g
in PBS) or PBS twice weekly for 12 weeks, and evaluated
2 weeks after the last treatment. H&E-stained lung
sections from PBS- (a) and PGP-treated
mice (b) were evaluated for alveolar
enlargement. Scale bar, 100 m.
(c) A 21% increase in Lm in the PGP-treated
group (*P = 0.021). (d) Right
ventricular (RV) mass is proportionally greater than
the rest of the lower heart (that is, the left ventricle
(LV) and the septum (S)) for PGP-treated mice than for
PBS-treated controls (22%, *P = 0.025). Weathington N.M. , et al. Nature
Medicine12, 317 - 323 (2006)
Kinetics of neutrophil influx, CXC chemokine
and PGP production in the mouse airway after exposure
to aerosolized LPS (a).
Monoclonal antibodies against KC and MIP-2 or PGP
inhibit in vitro PMN chemotaxis caused by
bronchoalveolar lavage fluids from LPS-exposed mice
(b), and decrease the in vivo
neutrophil burden in lungs of LPS-exposed BALB/c mice
(c). (a) Mice were
exposed to 100 g/ml
aerosolized LPS for 1 h at time 0. BAL samples (n
= 3 per time point) were collected and analyzed for
neutrophil number (green squares, outer left axis;
numbers are percentage of PMNs of total cells); KC
and MIP-2 chemokine concentrations by ELISA (blue
diamonds and black triangles, respectively, inner
left axis); and PGP (red circles, right axis) by ESI-LC-MS/MS.
(b) Twenty-four hours after LPS exposure,
BAL samples were collected, and supernatants were
incubated for 1 h with monoclonal antibodies against
KC (50 g/ml),
MIP-2 (50 g/ml),
PGP (9A4, 5 g/ml)
or isotype control antibody (50 g/ml)
before PMN chemotaxis. Scale shows activity of BAL
samples from LPS-exposed mice versus LPS-naive mice,
showing mean
s.e.m. from six mice. (c) Mice exposed
to aerosolized LPS were intratracheally administered
PBS, isotype control monoclonal antibody (50 g
total), monoclonal antibody against MIP-2 and KC (25
g
each) or monoclonal antibody 9A4 (30 g)
in 35 l
PBS 1 and 10 h after LPS. PMNs were counted from 24
h BAL samples (n = 3 per group); isotype
controls IgG2a (third bar) and IgG1 (fourth bar) are
controls for MIP2- and KC-specific antibody and 9A4
antibody, respectively. *P < 0.05, **
P < 0.01 compared to positive controls
by ANOVA for b and c.
Weathington N.M. , et al. Nature
Medicine12, 317 - 323 (2006)
BAL fluid samples from individuals with COPD
contain elevated levels of PGP. BAL samples
from humans were assayed for PGP using ESI-LC-MS/MS.
Incidence of PGP is greater among individuals with COPD
(3 of 5) than controls (2 of 18, P = 0.014),
and detected levels of PGP are higher in individuals
with COPD (363) than controls (22; P = 0.015).
Weathington N.M. , et al. Nature
Medicine12, 317 - 323 (2006)
PMN chemotaxis to PGP is dependent on the CXC chemokine
receptors. (a,b)
PMNs were incubated with 1 g/ml
monoclonal antibody against CXCR1, CXCR2, both or 2
g/ml
IgG2a isotype control for 1 h at 22 °C before assaying
chemotaxis to 50 ng/ml IL-8 (a) or
100 g/ml
PGP (b). (c) The dose-response
curve for the inhibition of PGP-mediated PMN chemotaxis
with monoclonal antibodies against both CXCR1 and CXCR2.
(d) RBL cells stably transfected to
express CXCR1 or CXCR2 migrate to PGP (g/ml
as indicated, P < 0.05 for CXCR1 and CXCR2
at 10 and 100) and IL-8 (50 ng/ml), but not PGG, whereas
vector-transfected controls are unresponsive to either
ligand. (e) Cxcr2-/- mice
and BALB/c control mice (n = 5 per group) were
intratracheally administered 250 g
PGP in 50 l
PBS or PBS alone. BAL and cell counts were performed
5 h later; because Cxcr2-/- mice are smaller
than controls, PMN counts are shown indexed to the values
for PBS-treated mice from each group, rather than absolute
PMN numbers, which are (left to right) 5.3 and 34.9
for wild-type mice and 2.1 and 2.4 for knockout mice
(PMN
104 per ml BAL fluid). *P < 0.05, **P
< 0.01 compared to controls by t-tests in
a,d and e;
ANOVA in b. Weathington N.M. , et al. Nature
Medicine12, 317 - 323 (2006)
g
in 50 
105 PMNs were placed on one side of a 3 
s.e.m. from six mice. (c) Mice exposed
to aerosolized LPS were intratracheally administered
PBS, isotype control monoclonal antibody (50 
