Proline in vasoactive peptides: consequences for peptide
hydrolysis in the lung
Marilyn P.
Merker,
Said H.
Audi,
Becky M.
Brantmeier,
Kasem
Nithipatikom,
Robert S.
Goldman,
David L.
Roerig, and
Christopher A.
Dawson
Departments of Anesthesiology, Pharmacology/Toxicology, and
Physiology, Medical College of Wisconsin, Milwaukee 53226;
Department of Biomedical Engineering, Marquette University, Milwaukee
53233; and Veterans Affairs Medical Center, Milwaukee, Wisconsin 53295
 |
ABSTRACT |
To examine the
hypothesis that trans isomers of
bradykinin and
[Gly6]bradykinin are
preferentially hydrolyzed by lung peptidases, we studied the fractional
inactivation of these peptides in the perfused rat lung using a
bioassay after a single-pass bolus injection and high-performance
liquid chromatography after lung recirculation. In the bioassay
studies, when the peptides passed through the lung, 25.6-fold more
bradykinin or 7-fold more
[Gly6]bradykinin was
required to elicit a contraction equivalent to that produced when the
peptides did not pass through the lung. In the recirculation studies,
hydrolysis progress curves with rapid and slow phases were
observed, with a higher fraction of bradykinin than
[Gly6]bradykinin
hydrolyzed in the rapid phase. Cyclophilin increased the hydrolysis
rate during the slow phase for both peptides. Kinetic analysis
indicated that the slowly hydrolyzed peptide fraction, presumably the
cis fraction, was 0.13 for bradykinin
and 0.43 for
[Gly6]bradykinin with
cis-trans
isomerization rate constants of 0.074 and 0.049 s
1, respectively,
consistent with published nuclear magnetic resonance studies.
high-performance liquid chromatography; bradykinin; [6-glycine]bradykinin; lung peptidases; cis-trans
isomerase; cyclophilin
 |
INTRODUCTION |
MANY PROLINE-CONTAINING PEPTIDES exist as relatively
slowly equilibrating mixtures of cis
and trans rotational isomers at X-Pro
imide bonds (11, 33). Some peptidases preferentially hydrolyze trans isomers of
proline-containing peptides in solution (11, 12, 14-16, 23, 25)
and preferentially bind trans isomers of peptidase inhibitors (2, 30, 32). Peptidases on the luminal
endothelial surface of the lung also preferentially hydrolyze trans conformers of certain
proline-containing peptides (5, 23-25). Underlying this phenomenon
is that fact that the
cis-trans isomerization time constants for the relevant imide bonds in these peptides are considerably longer than the pulmonary capillary transit
time, thus sparing a fraction of peptide equivalent to the equilibrium
cis fraction from hydrolysis during
passage through the lung.
To further explore the hypothesis that
cis bonds protect physiologically
relevant proline-containing peptides from hydrolysis in the pulmonary
circulation, we have taken advantage of the fact that two closely
related vasoactive peptides that are substrates for lung peptidases,
bradykinin and
[Gly6]bradykinin, have
different equilibrium
cis-to-trans
ratios at the
X6-Pro7
bond [~0.1 and ~0.4 for bradykinin and
[Gly6]bradykinin,
respectively, by nuclear magnetic resonance (NMR) (17-20)]. In the present study, the fractional hydrolysis of
bradykinin and
[Gly6]bradykinin in
the intact perfused rat lung was compared with both a bioassay
technique and high-performance liquid chromatography (HPLC). A
mathematical model was developed for kinetic interpretation of the
hydrolysis data to estimate
cis-trans
isomerization rate constants and equilibrium
cis-to-trans
ratios of the peptides within the lung. We also examined the impact of
changing the
cis-trans isomerization rates on the hydrolysis kinetics of the peptides in the
lung using the peptidyl prolyl
cis-trans
isomerase cyclophilin, for which both bradykinin and
[Gly6]bradykinin are
substrates (17, 24).
 |
METHODS |
Materials
Bradykinin was purchased from Peninsula Laboratories (Belmont, CA), and
[Gly6]bradykinin was
from Genosys (The Woodlands, TX). HPLC columns were purchased from
MetaChem (Torrance, CA). Fluorescamine, acetonitrile, and all other
chemicals were purchased from Sigma-Aldrich (St. Louis, MO). Purified
human recombinant cyclophilin was prepared under the guidance of Drs.
Wei Li and Robert Handschumacher (Yale University School of Medicine,
New Haven, CT). The cyclophilin was >95% pure as determined by
polyacrylamide gel electrophoresis. The cyclophilin concentration was
determined by titration with tritiated cyclosporin A (22), which was
obtained from Amersham Life Sciences (Arlington Heights, IL). Unlabeled
cyclosporin A was the kind gift of Sandoz Pharmaceutical (Basel,
Switzerland). Stock solutions of bradykinin and
[Gly6]bradykinin (1 mg
peptide/ml H2O) were prepared,
divided into 100-µl aliquots, and lyophilized, and the lyophilized
aliquots were stored at
20°C. The peptide was redissolved in
H2O (100 µl) immediately before
use. For HPLC, all H2O was
filtered in a Quad Style Nanopure water system containing the type I
ORGANICfree cartridge kit (Barnstead) to obtain reagent grade
H2O (18.0 M
· cm resistivity).
Bioassay Studies
For the bioassay tissue, two 22-µm tungsten wires were threaded
through an ~1.5-mm segment of rabbit jugular vein. The wires were
stretched over the jaws of two stainless steel rings, one of which was
anchored and the other attached to a force transducer as previously
described (21, 24). A 750-mg load was applied to the jugular ring, and
the vessel segment was allowed to equilibrate in the perfusion system
in Krebs-Ringer bicarbonate buffer (KRB; pH 7.4;
PO2 100 Torr,
PCO2 40 Torr) containing 5 mM glucose
and 2.5% bovine serum albumin at 37°C for 60 min. The artificial
perfusate was used to minimize any possible contribution of plasma
angiotensin-converting enzyme to peptide metabolism. The tissue
response to bradykinin was then tested, and afterward, the isolated rat
lung was connected to the perfusion circuit so that the jugular vein
was superfused with the venous effluent from the lung.
The isolated rat lungs from eight male Sprague-Dawley rats weighing
between 285 and 355 g were perfused at a flow rate of 0.167 ml/s with
KRB as previously described (23, 24). This produced a pulmonary
arterial pressure of 6.2 ± 1.6 (SD) Torr, with the venous pressure
set at zero. The lungs were ventilated at 30 breaths/min with 6%
CO2 and 15%
O2, with end-inspiratory and
end-expiratory pressures of 5.9 ± 1.3 (SD) and 2.0 ± 0.7 (SD) Torr, respectively. Two injection ports were included in the perfusion circuit such that a 0.1-ml bolus could be introduced either proximal to
the lung into the arterial inflow (lung injection site) or into the
venous outflow (tissue injection site). The arterial injection resulted
in the bolus passing through the lung before it reached the jugular
ring. Injection into the venous outflow resulted in the bolus passing
directly onto the jugular ring without passing through the lung (Fig.
1). To control for any changes in
responsiveness of the tissue or metabolic status of the lung throughout
the course of the studies, injections into the lung and tissue sites
were alternated.
The dose of peptide required to give contractions of equal force was
determined from the log dose-response curves at the lung and tissue
injection sites as previously described (24), and the data were
compared for significance with the Mann-Whitney rank sum test.
HPLC Studies
Lung perfusion. The lungs from 10 male
Wistar rats (235-363 g) were perfused with Krebs bicarbonate
buffer containing 5 mM glucose and 2.5% Ficoll (mol wt ~70,000) (23,
24). The lungs were perfused at a flow rate of 0.50 ml/s, which
produced a pulmonary arterial pressure of 12 ± 3 (SD) Torr, with a
venous pressure of zero. The total volume of the recirculation system
including the reservoir, tubing, and lung was ~8.9 ml. Studies of the
fraction of peptide surviving passage through the lung were initiated
by the addition of bradykinin (2.4, 9.4, or 37.8 nmol) or
[Gly6]bradykinin (0.6, 2.4, or 9.6 nmol) to the reservoir with or without cyclophilin (300 nmol). Perfusate samples (~300 µl) were collected from the venous
outflow tubing at timed intervals. Figure 2
is a diagram of this recirculating lung perfusion system.

View larger version (16K):
[in this window]
[in a new window]
|
Fig. 2.
Diagram of lung recirculation studies in which amount of peptide
emerging in venous effluent is measured by HPLC.
|
|
Immediately after collection of each sample, 200 µl of the sample
were added to 2 ml of ethanol on ice, and the mixture was centrifuged
(5 min at 1,600 g) to remove
precipitated Ficoll. The pellet was discarded, and the supernatant was
dried under N2 at 37°C and
stored at
20°C for HPLC analysis for intact peptide. Standards of bradykinin and
[Gly6]bradykinin were
prepared in perfusate, the Ficoll was precipitated with ethanol, and
the standards were centrifuged and dried under N2 in the same manner as described
above for the samples collected from lung perfusate.
HPLC. To detect the small quantities
of unhydrolyzed bradykinin and
[Gly6]bradykinin
remaining in the recirculating reservoir, it was necessary to develop a
sensitive HPLC assay for these peptides. The HPLC system consisted of a
Hewlett-Packard series 1050 autosampler (model 79852A) fitted with a
100-µl sample loop, a Hewlett-Packard series 1050 dual-piston pump
(model 79852A), a Hewlett-Packard helium degasser (model 79865A), a
MetaChem Inertsil 5-µm octadecylsilane column (30 × 250 mm, 5 µm), and a Perkin-Elmer model 650-10S fluorescence spectrophotometer
containing a 20-µl flow cell as a detector.
The optimal reaction conditions for fluorescamine derivatization of
bradykinin were determined by carrying out the derivatization reaction
at pH 6.05, 7.01, 8.00, and 9.00 and varying the time of reaction from
1 to 5 min. The maximum fluorescence intensity of the peptide
derivative was obtained at pH 8.00 and 9.00 during the first 1-2
min of reaction time, and after 1-2 min, the fluorescence intensity steadily decreased over 5 min. The fluorescence intensity of
the reaction was ~20-25% less in solutions having a pH of 6.05 or 7.01 than in the solutions with the higher pH. On the basis of these
results, for all studies, the derivatization was carried out at pH 8.00 for 1 min before injection onto the HPLC column.
Samples to be analyzed for bradykinin or
[Gly6]bradykinin were
dissolved in 200 µl of 0.1 M borate buffer (pH 8.00). The solution was filtered by centrifugation (1,600 g) through a 0.45-µm syringe filter (Titan) to remove any particulate material. A 60-µl portion of
the filtrate was transferred to a 250-µl capacity glass conical insert that was placed into a 2.0-ml capacity amber autosampler vial.
The derivatization, which was carried out automatically, consisted of
drawing 40 µl of the sample and 10 µl of fluorescamine (0.3 mg/ml
acetonitrile) into the autosampler loop in which the solutions were
mixed at a rate of 0.5 ml/min for 1 min. After 1 additional min to
allow for derivatization, the derivatized sample was automatically
injected onto the Inertsil column that had been equilibrated in 15%
acetonitrile in H2O.
The components of the injected sample were separated at a flow rate of
0.5 ml/min, with a mobil phase consisting of a gradient of 15-45%
acetonitrile in H2O over a period
of 6 min, followed by an isocratic elution with 45% acetonitrile in
H2O for 4 min. The acetonitrile
concentration was then increased to 100% over a 3-min period, and the
HPLC column was washed in 100% acetonitrile for 3 min. The system was
then reequilibrated in 15% acetonitrile in
H2O for 10 min. The
H2O and acetonitrile were sparged
with helium continuously during the chromatography, and the separation was run at room temperature (23°C). The chromatograms were recorded on a Hewlett-Packard model 3396 integrator using the peak area for
quantitation against a standard curve.
Typical standard curves representing the fluorescence intensity of the
fluorescamine derivative as a function of the amount of bradykinin or
[Gly6]bradykinin were
linear over the range of 0.3-48 pmol of peptide injected. The
retention time was 9.61 ± 0.17 (SD) min for 22 standard injections
of bradykinin and 9.62 ± 0.18 (SD) min for 15 standard injections
of [Gly6]bradykinin,
with amounts of peptide-fluorescamine derivative that spanned the
linear range of fluorescence intensity. The amount of unaltered peptide
remaining in each reservoir sample injected was calculated from
standard curves prepared the day the experiment was carried out.
Chromatograms showing the bradykinin present in the venous effluent of
the perfused lung at various times after the peptide was
introduced into the recirculating reservoir are shown in Fig.
3.

View larger version (21K):
[in this window]
[in a new window]
|
Fig. 3.
HPLC chromatograms of fluorescamine-derivatized bradykinin prepared
from lung venous effluent at indicated times after addition of
bradykinin to reservoir. Standard, representative chromatogram of a
sample collected from perfusion system without the lung (reservoir and
tubing only and equivalent to 0 time). Arrows, elution time of
bradykinin peak.
|
|
 |
RESULTS |
To examine the hypothesis that the equilibrium
cis-to-trans
ratio of the
X6-Pro7
bond of bradykinin and
[Gly6]bradykinin is a
factor determining the extent of peptide hydrolysis in a single pass
through the pulmonary capillary bed, the effects of passage through the
lungs on contraction of the rabbit jugular vein were examined for both
peptides. Figure 4 shows representative jugular vein contraction dose-response curves for bradykinin and [Gly6]bradykinin
injected into the tissue or lung injection sites. The rightward shift
of the dose-response curves obtained after lung site injections
compared with tissue site injections reveals peptide hydrolysis in the
lungs. The shift was greater for bradykinin than for
[Gly6]bradykinin,
consistent with a larger fractional hydrolysis of bradykinin. On
average, for all such studies carried out with bradykinin, passage
through the lung increased the dose required by 25.6 ± 6.4 (SE)-fold to produce a contraction equivalent to one in which
bradykinin did not pass through the lung
(P < 0.001). With
[Gly6]bradykinin,
passage through the lung increased the dose required by 7.0 ± 1.4 (SE)-fold to produce a contraction equivalent to one in which
[Gly6]bradykinin did
not pass through the lung (P < 0.001), which was significantly less than that for bradykinin
(P < 0.003).

View larger version (17K):
[in this window]
[in a new window]
|
Fig. 4.
Contractions of rabbit jugular vein ring induced by bolus injections of
bradykinin (A) or
[Gly6]bradykinin
(B) introduced into tissue and lung
injection sites. In this example, passage through lung increased amount
of peptide required to produce a contraction equivalent to one in
which peptide did not pass through lung by ~12.5-fold for bradykinin
and 3.6-fold for
[Gly6]bradykinin.
|
|
To obtain more quantitative information regarding the contribution of
cis-trans
isomerization to the hydrolysis kinetics of the two peptides in the
pulmonary circulation, the amount of bradykinin or
[Gly6]bradykinin
emerging in the venous effluent of the lung perfused from a
recirculating reservoir containing the peptides was measured by HPLC
(Figs. 5 and
6). For both peptides, there were two
phases to the hydrolysis reaction, an initial rapid phase and a
subsequent slower phase. The rapid phase is reflected by the fact that
by the first sampling time (10 s), there was a substantially lower concentration of peptide in the venous effluent than was originally in
the reservoir. The peptide surviving the initial rapid phase was
subsequently much more slowly hydrolyzed over a period of ~40 s as
shown for bradykinin in Fig. 5. The fraction of
[Gly6]bradykinin
surviving passage through the lung at the earliest time point was
higher than that of bradykinin, but like bradykinin, the peptide
surviving the rapid phase subsequently decreased much more slowly until
there was no detectable
[Gly6]bradykinin
remaining in the venous effluent after 80 s (Fig. 6). In three control
experiments for each peptide, with no lung in the perfusion circuit,
the rate of peptide degradation over the time course of the experiments
was not significant, indicating that the disappearance of intact
peptide from the lung venous effluent samples required that the peptide
pass through the lung. Over the range of peptide doses studied, no
systematic effect of dose on the fraction of peptide remaining with
respect to time was detected, indicating that lung hydrolysis was first
order under the conditions of the study. The addition of cyclophilin to
the reservoir containing either peptide increased both the fraction of
peptide hydrolyzed in the lung in the rapid phase of the reaction and
the rate of hydrolysis of the slower phase (Figs. 5 and 6).

View larger version (11K):
[in this window]
[in a new window]
|
Fig. 5.
Hydrolysis of bradykinin in lung perfused in series with recirculating
reservoir containing bradykinin or bradykinin plus cyclophilin. Solid
lines, Eqs. 5,
6,
10,
11, and
16 fit to the data, with resulting
estimate for initial fraction of peptide in
cis (unhydrolyzed) form ( ) of 0.11 and reaction rate constant
(k1) of 0.14 s 1 for bradykinin and 2.43 s 1 for bradykinin plus
cyclophilin.
|
|

View larger version (14K):
[in this window]
[in a new window]
|
Fig. 6.
Hydrolysis of
[Gly6]bradykinin in
lung perfused in series with recirculating reservoir containing
[Gly6]bradykinin or
[Gly6]bradykinin plus
cyclophilin. Solid lines, Eqs. 5,
6,
10,
11, and
16 fit to the data, with resulting
estimate for of 0.31 and
k1 of 0.051 s 1 for
[Gly6]bradykinin and
1.81 s 1 for
[Gly6]bradykinin plus
cyclophilin.
|
|
To further evaluate the hypothesis that the
trans isomers of bradykinin and
[Gly6]bradykinin were
preferentially hydrolyzed in the lung and to estimate the fraction of
cis peptide at equilibrium and
cis-trans isomerization rate constant for each peptide, we carried out the following kinetic analysis, the underlying hypotheses of which are
presented in Fig. 7. The total
concentration (b) of unaltered peptide at any time (
) was defined as
the sum of the concentrations of the
cis (c) and the
trans (t) forms
|
(1)
|
To interpret the hydrolysis-progress curves [b(
)], we
hypothesized the following kinetic processes, which included the
kinetics of
cis-trans
isomerization and the enzymatic hydrolysis of the trans isomer(s) of the peptides
|
(2)
|
|
(3)
|
where
e is the concentration of bradykinin and
[Gly6]bradykinin
peptidases in the lung, k1 and
k
1 are the cis-trans and
trans-cis isomerization rate constants, respectively, and k2 is the rate constant of the formation of the
trans peptide isomer-enzyme complex (te), which either
dissociates to form e and t with the rate constant
k
2 or proceeds to form the hydrolysis product
(p) with the rate constant k4.

View larger version (14K):
[in this window]
[in a new window]
|
Fig. 7.
Schematic diagram of hypothesized reactions occurring in lung,
reservoir, and tubing during peptide recirculation studies.
Cis-trans
isomerization occurs throughout recirculation system. F, perfusate flow
into lung; Fin, flow into
reservoir; c and t, concentration of
cis and
trans forms, respectively, of
peptides; subscripts R, T, and L, reservoir, tubing, and lungs,
respectively; e, concentration of bradykinin or
[Gly6]bradykinin
peptidase; p, hydrolysis product; k1,
k 1, k2,
k 2, and k4, reaction rate
constants (see text for definitions). Diagram also indicates venous
outflow site at which perfusate is sampled for determination of peptide
concentration.
|
|
Initially, before exposure to the peptidases in the lung, the
cis and
trans forms are in equilibrium, and we
defined the initial fraction of peptide in the
cis (c) form (
) as
|
(4)
|
where
the dissociation constant
(Kd) = k
1/k1.
Under the assumption that the reservoir is well mixed, the
concentrations of the cis and
trans forms of peptide in the
reservoir are described by the following species balance equations
|
(5)
|
and
|
(6)
|
with
the following initial conditions
|
(7)
|
where
cR and
tR are the concentrations
(nmol/ml) of the cis and
trans isomers, respectively, in the
reservoir; VR(
) is the volume
(ml) of the reservoir at time
; and
Fin is the flow (ml/s) into the
reservoir. During sampling
|
(8)
|
otherwise
|
(9)
|
where
F is the perfusate flow (ml/s) into the lung.
The concentrations of the cis and
trans forms of the peptide in the
tubing connecting the lung and the reservoir at distance x from the inlet to the tubing at time
are described by the following species balance equations
|
(10)
|
and
|
(11)
|
with
the following initial conditions
|
(12)
|
and
boundary conditions
|
(13)
|
where
cT(x,
)
and
tT(x,
)
are the concentrations of the cis and
trans forms, respectively, of the
peptide within the tubing at distance
x from the inlet to the tubing at time
and WT is the
average flow velocity (cm/s) in the tubing.
The spatial and temporal variations in the concentrations of the
cis and
trans forms of the peptide in the
lungs at distance x from the inlet and
time
are described by the following species balance equations
|
(14)
|
and
|
(15)
|
where
Vmax = ek4, the maximum rate of
peptide hydrolysis (nmol/s); Km = (k
2 + k4)/k2, the concentration of
peptide resulting in Vmax/2;
WL is the average flow
velocity in the lungs; and
cL(x,
)
and
tL(x,
)
are the vascular concentrations of the cis and
trans forms, respectively, of the
peptide within the lungs at distance x
at time
. Under the assumption that
Vmax/Km
F, Eqs. 14 and 15 reduce to
|
(16)
|
with
the following initial condition
|
(17)
|
and
boundary condition
|
(18)
|
where
x = z is the tubing outlet.
The only model parameters are
and
k1 because
k
1 = k1
/(1
). Furthermore, because the only impact of cyclophilin is to
increase k1, only
k1 was allowed to
change when the model equations were fit to the data obtained when
cyclophilin was present, and
was kept constant. Thus in the four
studies with bradykinin and the three with
[Gly6]bradykinin that
included experiments both with and without cyclophilin, optimization
for the three parameters,
and the two
k1 values (without and with cyclophilin), was carried out with a modified Levenberg-Marquardt algorithm to find the model parameter values for
which the numerical solution of Eqs.
5, 6,
10,
11, and
16 produce the best fit to the data
sets from experiments with and without cyclophilin simultaneously. In
the three studies with bradykinin and the three with
[Gly6]bradykinin in
which only the peptides in the absence of cyclophilin were studied,
optimization was carried out for only two parameters,
and
k1 without cyclophilin.
Examples of the model solutions calculated with the parameters
estimated from Eqs. 5, 6, 10,
11, and 16 are shown in Figs. 5 and 6. Also
in Figs. 5 and 6 are the data representing the amount of bradykinin or
[Gly6] bradykinin, respectively, remaining in
the venous effluent expressed as a percentage of the amount expected if
none had been metabolized. The mean estimates for the model parameters
obtained from all of the studies are shown in Table
1.
For both peptides, the agreement between the model solution and the
data is consistent with the hypothesis that a fraction of each was
preferentially hydrolyzed in the lung and that, during the period of
recirculation through the lungs, the rate of peptide hydrolysis was
dependent primarily on the rate of
cis-trans
isomerization, which was increased by cyclophilin.
 |
DISCUSSION |
The results are consistent with the hypothesis that
cis conformers of at least one of the
X-Pro bonds in bradykinin and
[Gly6]bradykinin are
refractory to enzymatic hydrolysis in the lung. In addition, the
cis-trans
isomerization time constants for both peptides are longer than the
pulmonary capillary transit time. Three types of evidence support this
hypothesis. The first is that for both peptides, in both the
single-pass bioassay and recirculation-HPLC studies, there was a
fraction of peptide that was not hydrolyzed during transit through the
lungs. In both kinds of experiments, a greater fraction of bradykinin
than [Gly6]bradykinin
was in the more rapidly hydrolyzable form, an observation consistent
with the fact that the equilibrium
cis-to-trans
ratio of the
Gly6-Pro7
bond of
[Gly6]bradykinin
(~0.38-0.4) has been found to be greater than that of the
Ser6-Pro7
bond of bradykinin (~0.1-0.13) as measured by NMR (17-20).
Finally, in the recirculation-HPLC study, the hydrolysis rates of the
more slowly hydrolyzed fractions of both peptides were increased by adding the peptidyl prolyl
cis-trans
isomerase cyclophilin.
In the bioassay studies, when the peptide passed through the lungs
before reaching the jugular vein rings, it took ~25.6 times as much
bradykinin and 7 times as much
[Gly6]bradykinin to
produce the same jugular vein contraction as was produced when the
peptide did not pass through the lungs, consistent with a substantially
larger fraction of
[Gly6]bradykinin than
of bradykinin surviving passage through the lung. Additionally, the
bioassay results reveal that
[Gly6]bradykinin had
~40% of the activity of bradykinin as an agonist for contraction of
the jugular vein, which is consistent with the findings of London et
al. (20) and Stewart (34) that
[Gly6]bradykinin had
from ~20 to 70% of the activity of bradykinin in bioassays for
contraction of rat uterus muscle or guinea pig ileum.
In the lung recirculation-HPLC studies of bradykinin and
[Gly6]bradykinin
hydrolysis kinetics, the ability of the model to fit the data is also
consistent with the hypothesis that the
trans isomers are the preferred
peptidase substrates and that the effect of cyclophilin is to increase
the rate of
cis-trans
isomerization, reflected by the increase in
k1. There were
two phases to the hydrolysis reactions, rapid and slow. The fraction of
peptide hydrolyzed in the rapid phase was higher for bradykinin than
for [Gly6]bradykinin,
which is reflected in the estimated values of the equilibrium
cis fractions (
) for bradykinin and
[Gly6]bradykinin (0.13 and 0.43, respectively) obtained from the model fits to the data. These
values are close to NMR measurements of the
cis fractions of these peptides at the
Ser6-Pro7
(~0.1-0.13) and
Gly6-Pro7
(0.38-0.40) bonds, respectively (17-20). The increase in the rate of
cis-trans
isomerization caused by cyclophilin allowed for a sufficient amount of
cis isomer to be converted to
trans within the capillary transit
time such that a fraction of peptide in excess of the equilibrium
trans fraction could be hydrolyzed
within the lung.
The results of the bioassay and HPLC studies can be quantitatively
compared as follows. Assuming a rat lung capillary mean transit time of
~3 s for a flow rate of 0.17 ml/s [assuming that the pulmonary
capillary blood volume-to-lung weight ratio is about the same in a rat
as in a rabbit (1)], taking into account the
cis-trans
isomerization rate constants obtained from the recirculation-HPLC studies, and all else being equal, an estimated equilibrium
cis fraction for bradykinin and
[Gly6]bradykinin from
the bioassay studies would be 7.5 ± 1.8 (SE) and 24.0 ± 5.6%
(SE), respectively, based on the shift in the dose-response curves
between the lung and tissue injection sites. The reason for the
quantitative differences between such estimates for the
equilibrium cis peptide fraction
obtained from the bioassay and HPLC-recirculation studies is not clear,
but factors that might contribute to an underestimation of the
cis peptide fraction from the bioassay
results include differences in dispersion of the bolus reaching the
bioassay tissue between the two injection sites; the possibility that
bradykinin might release other vasoactive agents from the lung that may
affect the bioassay tissue response (35); and, finally, the possibility
that bradykinin receptors in the jugular vein are preferentially
activated by one isomer or the other, a concept that has been
previously suggested (18, 20). Although such estimates of the
equilibrium cis contents of the
peptides from the bioassay study are lower than those calculated from
the HPLC-recirculation data, the relative differences are similar;
i.e., in the bioassay and HPLC experiments, the unhydrolyzable fraction
was 3.1- and 3.2-fold higher in the two kinds of experiments, respectively, for
[Gly6]bradykinin than
for bradykinin.
These relative differences in the equilibrium
cis fraction estimated by the bioassay
and HPLC studies are consistent with NMR measurements. Bradykinin
contains three proline residues, and in the studies of London and
colleagues (18, 19), the cis content
of each of the imide bonds was estimated to be up to 10% at
equilibrium with NMR spectroscopy, and it also has been suggested that
all of the observed cis proline
resonances observed in bradykinin were probably associated with
Pro7 (20). Other studies of the
Ser6-Pro7
bond demonstrated
cis-to-trans
ratios of 0.15 and 0.13 for the bradykinin fragment Ser-Pro-Phe-Arg and
[p-fluoro-Phe8]bradykinin,
respectively (17, 19). Carbon-13 NMR spectroscopic studies of
[90%-1,2-13C2-Gly6]bradykinin
and of the
[p-fluoro-Phe8]
analog of
[Gly6]bradykinin
indicate
cis-to-trans
ratios of 0.38 and 0.4, respectively (17, 20). Thus, with the data
obtained from the
[p-fluoro-Phe8]
analogs, in which the two peptides were directly compared,
[Gly6]bradykinin has a
3.1 higher equilibrium cis content
than the bradykinin, very close to the relative difference between the estimates obtained by both the bioassay and HPLC studies. Also consistent with our observations, the
cis-trans
isomerization rate constants for the
Ser6-Pro7
and
Gly6-Pro7
bonds in
[p-fluoro-Phe8]bradykinin
and the
[p-fluoro-Phe8]
analog of
[Gly6]bradykinin were
found to be 0.048 and 0.021 s
1, respectively, under the
NMR conditions of the study (17), which compares reasonably well with
0.074 and 0.049 s
1,
respectively, under the conditions of the present study.
Angiotensin-converting enzyme has been shown to account for the
majority of pulmonary hydrolysis of bradykinin (9, 10, 28). The results
of the present study are consistent with this finding because the only
difference between bradykinin and
[Gly6]bradykinin is
the substitution of Gly for Ser at the sixth position in the peptide,
and the differences in the kinetics of peptide hydrolysis can be
accounted for by the magnitude of the difference in the
cis-to-trans
ratio at the
X6-Pro7
bond. Therefore, it seems likely that a preference of
angiotensin-converting enzyme for a
trans isomer at the
X6-Pro7
bond is responsible for the fact that a fraction of peptide equivalent to the cis fraction at that bond
escapes inactivation in the lung. However, aminopeptidase P also
hydrolyzes bradykinin in the pulmonary circulation (27), and there is
evidence that it also has a preference for
trans isomers (15, 31). Thus the
impact of any potential cis-trans
isomerization at the
Arg1-Pro2
or
Pro2-Pro3
bond may also be involved in determining the extent of peptide inactivation in the lung.
To our knowledge, there is no evidence that the concentration of
cyclophilins in plasma is sufficient to affect the rate of cis-trans
isomerization of imide bonds involved in the hydrolysis of peptides in
the lung (7, 8, 29). Thus, under physiological conditions, our results
predict that cis peptides in the
venous blood would escape hydrolysis in the lung. Although the
relevance of this finding to the bioactivity of peptides is as yet
unknown, it has been suggested that different subtypes of bradykinin
receptors may preferentially bind bradykinin conformers containing
either cis or
trans isomers (18, 20). In fact, the
50-70% lower activity of
[Gly6]bradykinin
compared with that of bradykinin in some bioassay preparations was
interpreted as a possible indication that the cis peptide either did not bind to the
receptor or had a lower affinity than the
trans peptide (20). Because one-half
of the bradykinin entering the systemic circulation could still be in the cis conformation 9 s after leaving
the lung [about two times the resting arterial mean transit time
(26)], there could be a concentration gradient of differing
bradykinin conformers throughout the circulation. This phenomenon might
play a role in regulating the proportion of bradykinin conformers with
varying biological activity available to different receptor subtypes in
the vasculature.
Comments on HPLC Method
Analytic information regarding bradykinin inactivation in the lung has
been obtained with several different techniques including radio- and
chemiluminoimmunoassay (4, 6) and HPLC of radioactive bradykinin
peptides (27). However, many of these assays have the drawback that
they make use of radioactive bradykinin analogs and/or
antibodies to bradykinin for which there are no available [Gly6]bradykinin
counterparts. Therefore, this study required that we develop a
sensitive analytic methodology that could be used to directly compare
the kinetics of hydrolysis of these peptides under first-order
conditions. We chose
4-phenylspiro[furan-2(3H),1'-phtalan]-3,3'-dione (fluorescamine) derivatization of peptides in the lung perfusate followed by HPLC detection of intact peptides. Advantages of this technique include that the derivatization reaction takes place rapidly
and requires a single step that can be carried out at room temperature
and that the product is highly fluorescent (3, 36) and thus sensitive.
Most importantly, bradykinin and
[Gly6]bradykinin were
equally amenable to derivatization and could be separated from other
fluorescent components with the same HPLC methodology. In addition,
neither the fluorescamine nor its rapidly produced hydrolysis products
are fluorescent (36). The problem of instability known to occur with
fluorescamine derivatives was overcome by using automated precolumn
derivatization of individual samples immediately before injection onto
the HPLC column. Although Boppana et al. (3) previously explored the
utility of automated precolumn fluorescamine derivatization with a
variety of peptides, their studies of bradykinin did not include
studies in biological fluids or physiological solutions. Thus, to our
knowledge, detection of bradykinin and
[Gly6]bradykinin by
precolumn derivatization with fluorescamine followed by HPLC represents
a new method for the quantitation of these peptides in physiological
solutions such as lung perfusate.
 |
APPENDIX |
The motivation for the model hypothesis that
Vmax/Km
is very large relative to F, that is, that all of the
trans isomer is hydrolyzed within the
transit time of the pulmonary capillary bed, came from the results of a
previous study by Merker and Dawson (24) that revealed
that when cyclophilin was included in a bolus of bradykinin injected
into the arterial inflow of the lung, bradykinin was not detectable in
the venous effluent. Under the assumption that cyclophilin did not
affect bradykinin peptidase activity, the conclusion was that the
cis content rather than the hydrolysis rate constant
(Vmax/Km)
determined the extent of hydrolysis. The results of the present study
are consistent with this conclusion to the extent that the model fits
the data. However, the model also provides an additional means of
evaluating the hypothesis as illustrated in Fig.
8. The value of
Vmax/Km
was successively decreased until the model fit to the data
deteriorated. This deterioration is revealed by the sharp rise in the
coefficient of variation between the model fit and the data (Fig. 8,
). This occurred at a
Vmax/Km
of ~2 ml/s for both bradykinin and
[Gly6]bradykinin, thus
indicating a lower bound on
Vmax/Km,
consistent with the data of ~2 ml/s. Setting
Vmax/Km
at this lower bound and fitting the model to the data had little effect
on the estimated values of
or
k1 without
cyclophilin. However, when cyclophilin was present, the higher value of
k1 resulted in a
high correlation between
Vmax/Km
and k1 with
cyclophilin. In this case, the lower bound on
Vmax/Km
resulted in a correspondingly higher estimate of
k1 with
cyclophilin. Thus the values of
k1 with
cyclophilin in Table 1, obtained under the assumption that
Vmax/Km
F, actually represent lower bounds on
k1. When this
assumption is relaxed, and the
Vmax/Km
values are set at their lower bounds (~ 2 ml/s), the estimated values
for k1 with
cyclophilin represent upper bounds. The estimated upper bounds on
k1 in the
presence of cyclophilin were, on average, ~90 and 37 s
1 for bradykinin and
[Gly6]bradykinin,
respectively. To put the lower bounds on
Vmax/Km in perspective, if
Vmax/Km
were actually close to the lower bound of ~2 ml/s, the fraction of
the trans form of either bradykinin or
[Gly6]bradykinin
surviving hydrolysis on passage through the lungs would still be
<2%. Thus the overall conclusion from this evaluation is that the
estimated values of
and
k1 without
cyclophilin are virtually independent of the values of
Vmax/Km
that are consistent with the data, whereas the values of
k1 with
cyclophilin estimated under the assumption that
Vmax/Km
F and with no additional experimental data could be considered the
most conservative estimate for the cyclophilin-stimulated
cis-to-trans
conversion rate constant.

View larger version (18K):
[in this window]
[in a new window]
|
Fig. 8.
Impact of decreasing values of hydrolysis rate constant [ratio of
maximum rate of peptide hydrolysis to concentration of peptide
resulting in
Vmax/2
(Vmax/Km)]
on estimates of model parameters (A) and
k1 without
(k1;
B) and with cyclophilin
(k1CyP;
C). , Coefficient of variation
(CV) between model fit and data; , parameter values obtained for
bradykinin; , parameter values obtained for
[Gly6]bradykinin.
|
|
 |
ACKNOWLEDGEMENTS |
This work was supported by National Heart, Lung, and Blood
Institute Grant HL-52108; the Whitaker Foundation; and the Department of Veterans Affairs.
 |
FOOTNOTES |
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Address for reprint requests: M. P. Merker, VA Medical Center, Research
Service 151, Milwaukee, WI 53295.
Received 9 July 1998; accepted in final form 13 November 1998.
 |
REFERENCES |
1.
Audi, S. H.,
J. H. Linehan,
G. S. Krenz,
C. A. Dawson,
S. B. Ahlf,
and
D. L. Roerig.
Estimation of the pulmonary capillary transport function in isolated rabbit lungs.
J. Appl. Physiol.
78:
1004-1014,
1995[Abstract/Free Full Text].
2.
Audi, S. H.,
D. P. Schuster,
M. P. Merker,
R. D. Bongard,
J. H. Linehan,
and
C. A. Dawson.
Pulmonary angiotensin converting enzyme ligand binding kinetics (Abstract).
FASEB J.
10:
A99,
1996.
3.
Boppana, V. K.,
C. Miller-Stein,
J. F. Politowski,
and
G. Rhodes.
High-performance liquid chromatographic determination of peptides in biological fluids by automated pre-column fluorescence derivatization with fluorescamine.
J. Chromatogr.
548:
319-327,
1991[Medline].
4.
Campbell, D. J.,
A. Kladis,
and
A. M. Duncan.
Bradykinin peptides in kidney, blood, and other tissues of the rat.
Hypertension
21:
155-165,
1993[Abstract].
5.
Dawson, C. A.,
R. D. Bongard,
D. A. Rickaby,
J. H. Linehan,
and
D. L. Roerig.
Effect of transit time on metabolism of a pulmonary endothelial enzyme substrate.
Am. J. Physiol.
257 (Heart Circ. Physiol. 26):
H853-H865,
1989[Abstract/Free Full Text].
6.
Decarie, A.,
G. Drapeau,
J. Closset,
R. Couture,
and
A. Adam.
Development of digoxigenin-labeled peptide: application to chemiluminoenzyme immunoassay of bradykinin in inflammed tissues.
Peptides
15:
511-518,
1994[Medline].
7.
Dupuis, J.,
C. A. Goresky,
J. W. Ryan,
J. L. Rouleau,
and
G. G. Bach.
Pulmonary angiotensin-converting enzyme substrate hydrolysis during exercise.
J. Appl. Physiol.
72:
1868-1886,
1992[Abstract/Free Full Text].
8.
Endrich, M. M.,
and
H. Gehring.
The V3 loop of human immunodeficiency virus type-1 envelope protein is a high-affinity ligand for immunophilins present in human blood.
Eur. J. Biochem.
252:
441-446,
1998[Abstract].
9.
Erdos, E. G.,
and
R. A. Skidgel.
The angiotensin I-converting enzyme.
Lab. Invest.
56:
345-348,
1987[Medline].
10.
Ferreira, S. H.,
and
J. R. Vane.
The disappearance of bradykinin and eledoisin in the circulation and vascular beds of the cat.
Br. J. Pharmacol. Chemother.
30:
417-424,
1967[Medline].
11.
Fischer, G.
Peptidyl-prolyl cis/trans isomerases and their effectors.
Angew. Chem. Int. Ed. Engl.
33:
1415-1436,
1994.
12.
King, G. K.,
C. R. Middlehurst,
and
P. W. Kuchel.
Direct NMR evidence that prolidase is specific for the trans isomer of imidopeptide substrates.
Biochemistry
25:
1054-1062,
1986[Medline].
13.
Landaw, E. M.,
and
J. J. I. DiStefano.
Multiexponential, multicompartmental, and noncompartmental modeling. II. Data analysis and statistical considerations.
Am. J. Physiol.
246 (Regulatory Integrative Comp. Physiol. 15):
R665-R677,
1984[Medline].
14.
Lin, L. N.,
and
J. F. Brandts.
Evidence suggesting that some proteolytic enzymes may cleave only the trans form of the peptide bond.
Biochemistry
18:
43-47,
1979[Medline].
15.
Lin, L. N.,
and
J. F. Brandts.
Role of cis-trans isomerism of the peptide bond in protease specificity. Kinetic studies on small proline-containing peptides and on polyproline.
Biochemistry
18:
5037-5042,
1979[Medline].
16.
Lin, L. N.,
and
J. F. Brandts.
Evidence showing that a proline-specific endopeptidase has an absolute requirement for a trans peptide bond immediately preceding the active bond.
Biochemistry
22:
4480-4485,
1983[Medline].
17.
London, R. E.,
D. G. Davis,
R. J. Varvek,
J. M. Stewart,
and
R. E. Handschumacher.
Bradykinin and its Gly6 analog are substrates of cyclophilin: a fluorine-19 magnetization transfer study.
Biochemistry
29:
10298-10302,
1990[Medline].
18.
London, R. E.,
J. M. Stewart,
and
J. R. Cann.
Probing the role of proline in peptide hormones. NMR studies of bradykinin and related peptides.
Biochem. Pharmacol.
40:
41-48,
1990[Medline].
19.
London, R. E.,
J. M. Stewart,
J. R. Cann,
and
N. A. Matwioff.
13C and 1H nuclear magnetic resonance studies of bradykinin and selected peptide fragments.
Biochemistry
17:
2270-2277,
1978[Medline].
20.
London, R. E.,
J. M. Stewart,
R. Williams,
J. R. Cann,
and
N. A. Matwiyoff.
Carbon-13 NMR spectroscopy of [20%-1,2-13C2-Gly6]-bradykinin. Role of serine in reducing structural heterogeneity.
J. Am. Chem. Soc.
101:
2455-2462,
1979.
21.
Madden, J. A.,
C. A. Dawson,
and
D. R. Harder.
Hypoxia-induced activation in small isolated pulmonary arteries of the cat.
J. Appl. Physiol.
59:
113-118,
1985[Abstract/Free Full Text].
22.
Merker, M.,
and
R. E. Handschumacher.
Uptake and nature of the intracellular binding of cyclosporin A in a murine thymoma cell line, BW5147.
J. Immunol.
132:
1-6,
1984[Free Full Text].
23.
Merker, M. P.,
I. M. Armitage,
S. H. Audi,
L. T. Kakalis,
J. H. Linehan,
J. R. Maehl,
D. L. Roerig,
and
C. A. Dawson.
Impact of angiotensin-converting enzyme substrate conformation on fractional hydrolysis in the lung.
Am. J. Physiol.
270 (Lung Cell. Mol. Physiol. 14):
L251-L259,
1996[Abstract/Free Full Text].
24.
Merker, M. P.,
and
C. A. Dawson.
Cyclophilin facilitated bradykinin inactivation in the perfused rat lung.
Biochem. Pharmacol.
50:
2085-2091,
1995[Medline].
25.
Merker, M. P.,
C. A. Dawson,
R. Bongard,
D. Roerig,
S. Haworth,
and
J. Linehan.
Angiotensin-converting enzyme preferentially hydrolyzes the trans isomer of a proline-containing substrate.
J. Appl. Physiol.
75:
1519-1524,
1993[Abstract].
26.
Milnor, W. R.
Hemodynamics. Baltimore, MD: Williams and Wilkins, 1982.
27.
Prechel, M. M.,
A. T. Orawski,
L. L. Maggiora,
and
W. H. Simmons.
Effect of a new aminopeptidase P inhibitor, apstatin, on bradykinin degredation in the rat lung.
J. Pharmacol. Exp. Ther.
275:
1136-1142,
1995[Abstract].
28.
Regoli, D.,
and
J. Barabe.
Pharmacology of bradykinin and related kinins.
Pharmacol. Rev.
32:
1-46,
1980[Medline].
29.
Ryffel, B.
Cyclosporin binding proteins. Identification, distribution, function and relation to FK binding proteins.
Biochem. Pharmacol.
46:
1-12,
1993[Medline].
30.
Schuster, D. P.,
T. J. McCarthy,
M. J. Welch,
S. Holmberg,
P. Sandiford,
and
J. Markham.
In vivo measurements of pulmonary angiotensin-converting enzyme kinetics. II. Implementation and application.
J. Appl. Physiol.
78:
1169-1178,
1995[Abstract/Free Full Text].
31.
Simmons, W. H.,
and
A. T. Orawski.
Membrane-bound aminopeptidase P from bovine lung.
J. Biol. Chem.
267:
4897-4903,
1992[Abstract/Free Full Text].
32.
Skoglof, A.,
I. Nilsson,
S. Gustafsson,
J. Deinum,
and
P. Gothe.
Cis-trans isomerization of an angiotensin converting enzyme inhibitor. An enzyme kinetic and nuclear magnetic resonance study.
Biochim. Biophys. Acta
1041:
22-30,
1990[Medline].
33.
Stein, R. L.
Mechanism of enzymatic and nonenzymatic prolyl cis-trans isomerization.
In: Advances in Protein Chemistry. Accessory Folding Proteins, edited by G. Lorimer. San Diego, CA: Academic, 1993, p. 1-24.
34.
Stewart, J. M.
Chemistry and biological activity of peptides related to bradykinin.
In: Bradykinin, Kallidin and Kallikrein, edited by E. G. Erdos. Berlin, Germany: Springer-Verlag, 1979, p. 227-265.
35.
Terragno, N. A.,
and
A. Terragno.
Release of vasoactive substances by kinins.
In: Bradykinin, Kallidin and Kallikrein, edited by E. G. Erdos. Berlin, Germany: Springer-Verlag, 1979, p. 401-421.
36.
Udenfriend, S.,
S. Stein,
P. Bohlen,
W. Dairman,
W. Leimgruber,
and
M. Weigele.
Fluorescamine: a reagent for assay of amino acids, peptides, proteins, and primary amines in the picomole range.
Science
178:
871-872,
1972[Medline].
Am J Physiol Lung Cell Mol Physiol 276(2):L341-L350