315
Basic Polyamino Acids Rich in Arginine,
Lysine, or Ornithine Cause Both
Enhancement of and Refractoriness to
Formation of Endothelium-Derived Nitric
Oxide in Pulmonary Artery and Vein
Louis J. Ignarro, Michele E. Gold, Georgette M. Buga, Russell E. Byrns,
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Keith S. Wood, Gautam Chaudhuri, and Gerard Frank
The objective of this study was to elucidate the mechanism by which polyamino acids containing
L-arginine, L-lysine or L-ornithine cause endotheh'um-dependent relaxation of bovine intrapulmonary artery and vein. Basic but not acidic or neutral polypeptides ranging in average molecular
weights from 17 to 225 kDa elicited time- and concentration-dependent relaxation and cyclic GMP
accumulation in precontracted rings of artery and vein by endotheh'um-dependent mechanisms.
Vascular responses were markedly inhibited by oxyhemoglobin, methylene blue, or potassium.
The basic polyamino acids stimulated the formation and/or release of an endothelium-derived
relaxing factor (EDRF) identified as nitric oxide (NO) in perfused segments of both artery and
vein as assessed by bioassay. The polyamino adds and A23187 released a similar endotheliumderived NO (EDNO) from artery and vein, as assessed by the similar half-life (3-5 seconds),
antagonism by superoxide anion or oxyhemoglobin, enhancement by superoxide dismutase, and
lack of influence by indomethacin. The bask polyamino adds elicited potent relaxant responses
with ECJO values ranging from 3x10"' to 2xlO" 7 M, and a direct correlation was obtained
between molecular weight and relaxation potency irrespective of the basic amino acid incorporated. Prolonged contact of arterial or venous rings with basic polyamino acids resulted in the
rapid development of marked refractoriness to relaxation and cyclic GMP formation on addition
of polyamino add. Moreover, refractoriness developed to the vascular responses of other
endothelium-dependent vasodilators but not to glyceryl trinitrate or isoproterenol. The mechanism of refractory responses was attributed to interference with EDNO formation and release as
assessed by bioassay and chemical assay. The hypothesis is forwarded that the basic polyamino
adds serve as partial substrates for the enzyme system that catalyzes the conversion of L-arginine
to NO. Prolonged contact between substrate and enzyme results in enzyme desensitization and the
development of refractoriness or a form of tolerance to vasodilators whose action is mediated by
EDNO. (Circulation Research 1989;64:315-329)
R
ecent studies have indicated that small vasoactive polypeptides containing basic amino
acids such as L-arginine or L-lysine cause
relaxation of vascular smooth muscle by diverse
From the Department of Pharmacology (L.J.I., M.E.G.,
G.M.B., R.E.B., K.S.W.), and the Department of Obstetrics and
Gynecology (G.C.), University of California Los Angeles, School
of Medicine, Los Angeles, and the Department of Medicine
(G.F.), Wadsworth VA Medical Center, Los Angeles.
Supported in part by National Institutes of Health Grant HL35014 and a grant from the Laubisch Fund for Cardiovascular
Research.
Address for reprints: Dr. Louis J. Ignarro, Department of Pharmacology, UCLA School of Medicine, Los Angeles, CA 90024.
Received February 22, 1988; accepted July 25, 1988.
mechanisms of action. Moreover, the mechanism of
action of a given polypeptide may be different in
one type of blood vessel or species than in another.
Endothelium-dependent arterial relaxation has been
demonstrated for numerous polypeptides including
bradykinin, vasoactive intestinal polypeptide, substance P, thrombin, cholecystokinin, bombesin, neurotensin, melittin, and others.'-' 2 Bradykinin was
demonstrated also to elicit endothelium-dependent
relaxation of venous smooth muscle." 12 Where
studied, such polypeptides have been shown to
stimulate endothelium-dependent cyclic GMP accumulation in vascular smooth muscle. 61112 The direct
formation and/or release of endothelium-derived
316
Circulation Research
Vol 64, No 2, February 1989
relaxing factor (EDRF) from intact blood vessels or
cultured endothelial cells has been shown to occur
in response to bradykinin and substance p.'*-'*
Some of these polypeptides relax vascular smooth
muscle by mechanisms involving stimulation of
prostacyclin formation.1217-19 In addition, some of
the above polypeptides, as well as parathyroid
hormone and several atriopeptins that contain basic
amino acids, were found to elicit endotheliumindependent arterial relaxation. 4719 - 22
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Thus, there appears to be no unifying mechanism
of action for such vasodilator polypeptides. In a
more extensive study we reported that bradykinin
and vasoactive intestinal polypeptide each elicit
endothelium-dependent relaxation of bovine intrapulmonary artery by two independent mechanisms
involving both cyclic GMP formation triggered by
EDRF and cyclic AMP formation triggered by prostacyclin or a related prostaglandin.12 On the other
hand, bradykinin relaxed veins by endotheliumdependent mechanisms involving only EDRF release
and cyclic GMP accumulation, whereas vasoactive
intestinal polypeptide was inactive.12 A recent brief
report indicated that two relatively large polyamino
acids, which can also be termed polypeptides, composed of the basic amino acids L-arginine or L-lysine
caused endothelium-dependent relaxation of rat aorta
that was inhibited by methylene blue, whereas
acidic polyamino acids, polyamines, and amino
acids were inactive.23 The objective of the present
study was to elucidate the mechanism of vascular
smooth muscle relaxation elicited by such basic
polyamino acids and to compare this with the
vascular actions of bradykinin and endotheliumindependent relaxants. Polyamino acids ranging in
molecular weight from 17 to 225 kDa and composed
of L-arginine, L-lysine, L-ornithine, or D-lysine were
tested on unrubbed and rubbed rings of bovine
intrapulmonary artery and vein in the absence and
presence of inhibitors such as oxyhemoglobin, methylene blue, and indomethacin. Both vascular cyclic
GMP and cyclic AMP levels were monitored, and
the capacity of the basic polyamino acids to stimulate the formation and/or release of endotheliumderived nitric oxide (EDNO) from intact artery and
vein was assessed by bioassay and chemical assay.
Materials and Methods
Preparation of Rings of Bovine Intrapulmonary
Artery and Vein
Bovine lungs were obtained from freshly slaughtered cows under 5 years of age as described
previously.24 Intrapulmonary arterial and venous
branches were rapidly isolated, gently cleaned of
parenchyma, fat, and adhering connective tissue,
and placed in cold preoxygenated Krebs-bicarbonate
solution. Segments of the second arterial branch
and underlying second venous branch extending
into the larger lobe were isolated. Outside diameters were 4-6 mm (artery) and 6-8 mm (vein).
Vessel segments were sliced into rings (4 mm wide)
with a specially designed microtome.23 Rings prepared in this manner possessed an intact or functional endothelium as assessed by 80-100% relaxation responses to 10"7-10"6 M acetylcholine
(arteries) or 10~8 M bradykinin (veins). These vascular rings are referred to in the text as unrubbed.
Endothelial cells were largely removed from vascular rings by everting the rings (intimal surface
outside), gently rubbing the intimal surface with
moistened filter paper for 30-40 seconds, and again
everting the rings (intimal surface inside). This
procedure of endothelium denudation was very
effective and did not result in damage to the medial
layer due to stretching or tearing that could occur
through the luminal insertion of a wooden stick or
cotton swab. These endothelium-denuded rings
sharply contracted in response to 10~6 M acetylcholine (arteries) or 10"8 M bradykinin (veins), and
relaxed to 10~'-10~7 M NO (arteries and veins).
Mounting Rings and Recording of Muscle Tension
Arterial and venous rings were mounted by means
of fine nichrome wires in jacketed, 25 ml capacity,
drop-away chambers containing Krebs-bicarbonate
solution (37° C) gassed with 95% O2-5% CO2 as
described previously.24-25 Changes in isometric force
were measured, and length-tension relations were
determined initially for unrubbed and endotheliumdenuded rings of artery and vein exactly as described
previously for isolated rings of bovine pulmonary
artery and vein.24 Arterial and venous rings were
routinely depolarized by addition of 120 mM KC1
following 2 hours of equilibration at optimal tension, and were subsequently washed and allowed to
equilibrate for 45 minutes before initiating any given
protocol. This procedure increases and stabilizes
any subsequent submaximal precontractile responses
to U46619, phenylephrine, and other contractile
agents, presumably by loading the smooth muscle
cells with calcium. This procedure has been
employed routinely for bovine pulmonary vessels in
this laboratory.2425
Bioassay Cascade Superfusion Technique
A modification2627 of the procedure developed by
Vane28 was employed. Briefly, isolated segments (57 cm) of endothelium-intact, bovine intrapulmonary
artery or vein were cleaned of adhering connective
tissue and any small branches were ligated with
titanium hemostatic clips (Pilling Co, Fort Washington, Pennsylvania) to prevent leakage during perfusion of the vessel. Vessel segments were fitted at
either end with polyethylene tubing, were submerged in chambers containing oxygenated Krebsbicarbonate solution at 37° C, and were perfused (3.5
ml7min) with oxygenated Krebs-bicarbonate solution
at 37° C. The perfusate was allowed to superfuse
three isolated, helically cut, precontracted strips of
endothelium-denuded artery or vein arranged in a
cascade at 37° C. Vascular strips were contracted by
Ignarro et al
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
10 fxM phenylephrine (artery) or 0.01 /xM U46619
(vein) delivered by superfusion from a separate line,
and responses were monitored exactly as described
previously.26 Arterial and venous segments as well
as vascular strips were pretreated with 10 fiM indomethacin as described26 to prevent the formation of
prostaglandins, especially prostacyclin, whose formation could be provoked by various chemicals and
manipulations.14 The time delay between each vascular strip in the cascade was 2.5 seconds, as was the
delay between the perfused vessel and the first
vascular strip, to observe a decrement in the relaxant
responses to EDRF during the cascade superfusion.
At the start and end of each experiment, glyceryl
trinitrate, a stable endothelium-independent vasodilator, was superfused (0.1 fjM) over the vascular
strips to be certain that the magnitudes of relaxation
did not deviate appreciably among the three strips
during the experiment.14
Determination of Cyclic Nucleotide Levels
Cyclic GMP and cyclic AMP determinations were
made in arterial and venous rings that had been
equilibrated under tension and depolarized with
KC1. Tone was monitored until the time of freezeclamping. The use of drop-away bath chambers,
freeze-clamping of rings, preparation and extraction
of tissues for cyclic nucleotide determinations, and
radioimmunoassay procedures were described
previously.252930 Cyclic GMP and cyclic AMP levels were determined in aliquots from the same ring
extract. None of the test agents added to bath
chambers interfered directly with antigen-antibody
binding in the radioimmunoassay procedures. Recoveries of standard amounts of added cyclic nucleotides were determined periodically and the values
Ach
ARTERY
VEIN
ARTERY
KCl
30 m M
ARC 115 kDa
a
Polyamino Acid Relaxation and Cydic GMP Formation
317
ranged from 92% to 104%. Therefore, no corrections for sample recoveries were made.
Chemical Assay of Nitric Oxide
The concentration of NO in the superfusion media
collected after perfusion of intact bovine pulmonary
artery with and without A23187 or basic polyamino
acids was determined by spectrophotometric procedures based on the diazotization of sulfanilic acid by
NO at acidic pH, exactly as described previously.27
Scanning Electron Microscopy
Isolated arterial rings were prepared for scanning
electron microscopy using standard techniques described previously.31 In brief, following myographic
protocols, arterial rings were immediately fixed in
4% glutaraldehyde. Arterial rings were subsequently
postfixed in osmium tetroxide and then coated with a
gold-palladium film. Samples were examined and
photographed in an ETEC Auto-Scan Electron Microscope (Haywood, California).
Chemicals and Solutions
Acetylcholine chloride, bradykinin triacetate,
A23187, phenylephrine hydrochloride, methylene
blue, methemoglobin, indomethacin, sodium dithionite, dithiothreitol, pyrogallol, superoxide dismutase (bovine liver), all polyamino acids, amino acids,
bradykinin analogues, neurotensin, and other peptides were obtained from Sigma Chemical Co, St.
Louis, Missouri. Glyceryl trinitrate (nitroglycerin;
10% wt/wt triturated mixture in lactose) was a gift
from ICI Americas, Inc (Wilmington, Delaware).
U46619 ([15S]-hydroxy-lla, 9a [epoxymethano]
prosta 5Z, 13E dienoic acid) was provided by The
Upjohn Co (Kalamazoo, Michigan) and was dissolved in absolute ethanol at a concentration of 5 mg/
FIGURE 1. Concentration-dependent relaxant
effects of basic polyamino acids in artery and
vein precontracted by U46619 but not potassium. Unrubbed arterial and venous rings were
mounted under 4 g of tension and were precontracted submaximally (60-80% of maximum) by
addition of 10-* M U46619 (U4) or 30 mM KCl
as indicated. Acetylcholine (Ach) and bradykinin (BKN) were added cumulatively at the bath
concentrations indicated (expressed as exponents to the base power 10). Poly-L-arginine 115
kDa (ARG), poly-L-ornithine 50 kDa (ORN), and
poly-L-lysine 41 kDa (L YS) were added cumulatively at molar bath concentrations of /.6x
10-\a), 5xlO-9(b), 1.6xl(r'(c), 5*10-'(d), and
/.6x 10'7(e) as indicated. Only one concentrationresponse relation was obtained for a given vascular ring. Each of the individual tracings illustrated is representative from a total of four to six
separate experiments.
318
Circulation Research
Vol 64, No 2, February 1989
TABLE 1. Potencies of Basic Polyamino Adds as Relaxants of
Bovine Intrapulmonary Artery and Vein
EC* (M)
Polyamino Acid
Poly-L-lysine 225 kDa
Poly-L-ornithine 120 kDa
Poly-L-lysine 115 kDa
Poly-L-arginine 115 kDa
Poly-L-omithine 50 kDa
Poly-L-lysine 41 kDa
Poly-L-arginine 41 kDa
Poly-L-lysine 17 kDa
Artery
Vein
3.0±0.2xl0"'
6.0±0.5xl0-'
6.2±0.5xl0-'
6.8±0.5xl0" 9
2.2+0.3x10"'
2.4±0.4xl0" 8
2.8±0.3xlO-»
2.2±O.lxlO"7
3.3±O.3xlO"'
5.6±0.4xl0"'
7.1 ±0.6x10-'
6.2±0.2xl0"'
2.6+0.4x10"'
1.9±0.3xl0"'
2.4±0.4xl0"»
2.1±0.2xl0" 7
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Unrubbed vascular rings were mounted under 4 g of tension
and were precontracted submaximally (65-75% of maximum) by
addition of U46619 (10"9-10"8 M). After peak contractile responses
developed, the indicated polyamino acids were added at cumulatively increasing concentrations at threefold increments until
100% relaxation was observed. Percent relaxation signifies the
percent decrease in submaximal precontractile tone elicited by
U46619. Only one concentration-response relation was established for each ring due to the subsequent development of
tolerance. Values represent the mean±SEM from eight rings
isolated from four animals (two rings per animal). Values corresponding to a comparable molecular weight, irrespective of the
polyamino acid, are significantly different (p<0.05) from values
for the other molecular weights within the indicated column
(artery or vein).
ml. Dilutions were prepared in cold distilled water
just before use. Solutions of hygroscopic acetylcholine chloride were prepared, aliquoted, and stored
frozen as described previously.25 Bradykinin and its
analogues, other small peptides, pyrogallol and superoxide dismutase are unstable and were prepared
fresh in distilled water just before use. Polyamino
acids were prepared in distilled water and stored
frozen. Oxyhemoglobin was prepared as described
previously.32 Krebs-bicarbonate solution consisted
of (mM) NaCl 118, KC1 4.7, CaCl2 1.5, NaHCO3 25,
MgSO4 1.1, KH2PO4 1.2, and glucose 5.6. Depolarizing KC1 solution had a composition similar to
Krebs-bicarbonate solution except the NaCl was
replaced with an equimolar concentration of KC1.
Calculations and Statistical Analysis
Relaxation was measured as the decrease in tension below the elevated tension elicited by precontracting arterial or venous smooth muscle with
U46619. Values in the tables and in Figures 4 and 5
are expressed as the mean±SEM and represent
unpaired data. Comparisons were made using the
Student's / test for unpaired values. The level of
statistically significant difference was p<0.05.
Results
Endothelium-Dependent Arterial and Venous
Relaxant Effects of Basic Polyamino Acids
High molecular weight basic polyamino acids
consisting of L-arginine, L-lysine, or L-ornithine
produced potent, marked, and concentrationdependent relaxant responses in U46619- or
phenylephrine- but not potassium-precontracted
unrubbed rings of bovine intrapulmonary artery and
vein (Figure 1). Relaxation elicited by the basic
polyamino acids was generally slow in onset and
development when compared with the relaxant
responses to either acetylcholine in artery or bradykinin in vein. Prior endothelial denudation of the
vascular rings abolished all relaxant responses. Table
1 indicates the eight basic polyamino acids studied
together with their approximate average molecular
weights and EC50 values as arterial and venous
smooth muscle relaxants. Compounds are listed in
order of decreasing potencies as arterial relaxants.
There was a direct correlation between molecular
weight and relaxation potency, different polyamino
acids of comparable molecular weights showed similar potencies, and potencies as arterial relaxants
were similar to potencies as venous relaxants for
each polyamino acid tested. Moreover, the relaxant
potencies of poly-D-lysine 40 and 225 kDa were
similar to the potencies of the corresponding polyL-amino acids, and the responses were endotheliumdependent (not shown). In contrast to the basic
compounds, acidic polyamino acids such as polyL-aspartic acid 40 kDa, poly-L-glutamic acid 75
kDa, and poly-gamma-benzyl-L-glutamic acid 115
kDa were completely inactive in arterial and venous
rings at concentrations of up to 10~6 M. Similarly,
neutral polyamino acids of comparable molecular
weights (poly-L-glycine, poly-L-leucine, or polyL-tyrosine) were completely inactive, as were the
basic monoamino acids themselves such as Larginine, L-lysine, and L-ornithine, and polyamines
such as spermine, spermidine, and putrescine.
Effects of Inhibitors on Vascular Relaxant
Responses to Basic Polyamino Acids
The above data suggest that the EDRF first
described by Furchgott and Zawadzki33 mediates
the relaxant effects of the basic polyamino acids.
Oxyhemoglobin, which binds and inactivates EDRF
and NO, and methylene blue, which inhibits the
stimulatory action of EDRF and NO on soluble
guanylate cyclase, each abolished or markedly inhibited the arterial and venous relaxant effects of basic
polyamino acids. Data for four such polypeptides
are illustrated in Figure 2. As observed previously
in this and other laboratories,2432-34 oxyhemoglobin
and methylene blue rendered arterial and venous
rings supersensitive to the contractile effects of
agonists. Therefore, lower concentrations of U46619
were used to contract rings in the presence of either
inhibitor. The following agents were tested by preincubating arterial or venous rings for 15-30 minutes before initiating relaxations and were found to
be without appreciable effect in altering relaxant
responses: 10~5 M indomethacin, 10"* M atropine,
10~6 M propranolol, 10~5 M chlorpheniramine, 10~5
M quinacrine (mepacrine), 10"5 M nordihydroguaiaretic acid, and 10"4 M SKF-525A.
Ignarro et al Polyamino Acid Relaxation and Cyclic GMP Formation
ARTERY
ARG IIS kDa
319
LYS 115 kDa
b
-8
VEIN
ORN 50 kDa
ORN 50 kDa
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
FIGURE 2. Inhibition by oxyhemoglobin and methylene blue of arterial and venous smooth muscle relaxation elicited by
basic polyamino acids. Unrubbed arterial and venous rings were mounted under 4 g of tension and were precontracted
submaximally (70-85% of maximum) by addition of 10'' M or I0~9 M U46619 (U4) as indicated. Poty-L-arginine 115 kDa
(ARG), poly-L-tysine 115 kDa (LYS), poly-L-lysine 225 kDa (LYS), andpoly-L-ornithine 50 kDa (ORN) were added at the
indicated bath concentrations (b, 5xlO'9 M; c, 7.6x70"* M; d, 5x10'" M). Following washing and 45 minutes of
equilibration, either 1 yM oxyhemoglobin (HbO2) or 10 fiM methylene blue (MB) was added 10 minutes before
precontracting rings with U46619 as indicated. The basic polyamino acids were again tested as described above. Each
of the individual tracings illustrated is representative from a total offour separate experiments.
Bioassay Cascade Superfusion
The release of EDNO from intact segments of
perfused artery and vein by the basic polyamino
acids is illustrated in Figure 3. Superfusion of
glyceryl trinitrate over the three endotheliumdenuded vascular strips caused relaxant responses
that were of comparable magnitudes in all three
strips. In contrast, perfusates from intact artery or
vein in the absence or presence of A23187 evoked
decremental relaxant responses down the cascade
of strips, thus reflecting the very short half-life of
arterial and venous EDNO (3-5 seconds as estimated from the tracings). Poly-L-arginine 41 kDa
(artery and vein) and poly-L-lysine 41 kDa (vein),
when perfused through intact vessels, relaxed bioassay strips similarly to perfused A23187 (Figure 3).
The higher molecular weight basic polyamino acids
were more potent than lower molecular weight
analogues (not shown). In the presence of perfused
superoxide dismutase, the relaxant effects and duration of relaxation as well as the chemical stability of
EDNO were all enhanced in comparison to control
responses. On the other hand, superfused pyrogallol, which generates superoxide anion in oxygenrich aqueous media,35 nearly abolished the EDNOmediated relaxant responses. Only the data obtained
in artery with poly-L-arginine 41 kDa are illustrated
in Figure 3. Superfused oxyhemoglobin abolished
the venous relaxant responses of EDNO released by
A23187, poly-L-arginine 41 kDa, and poly-L-lysine 41
kDa (Figure 3). Oxyhemoglobin similarly inhibited
arterial and venous EDNO released by several different basic polyamino acids (not shown).
Stimulation of Arterial and Venous Cyclic GMP
Accumulation Elicited by Basic Polyamino Acids
In view of the observations that the basic polyamino acids caused slowly developing relaxant
responses, whereas other endothelium-dependent
relaxants such as acetylcholine, bradykinin, and
A23187 caused rapidly developing relaxant responses,
a time-course analysis of cyclic GMP accumulation
and relaxation in response to a relatively high concentration of poly-L-arginine 115 kDa was conducted in
artery and vein (Figure 4). Peak cyclic GMP accumulation was observed at about 180 seconds after addition of the polyamino acid, and a good correlation was
observed between cyclic GMP accumulation and relaxation in both artery and vein. A similar time-course
pattern was observed for poly-L-arginine 41 kDa,
poly-L-lysine 41 and 225 kDa, and poly-L-ornithine 50
kDa in artery and vein (not shown). The onset of
cyclic GMP accumulation consistently preceded the
onset of relaxation in both artery and vein. Relaxation
continued for 2-3 minutes beyond the time of peak
cyclic GMP accumulation. Thus, a contact time of 180
seconds was selected for subsequent experiments.
Figure 5 illustrates that poly-L-arginine 115 kDa (artery)
and poly-L-lysine 225 kDa (vein) produced
concentration-dependent cyclic GMP accumulation
and relaxation, and a good correlation was observed
between cyclic GMP accumulation and relaxation in
320
Circulation Research
Vol 64, No 2, February 1989
Artery
1
SOD
,
1"
GTN
Perf
A23
' T^i
' ~\ ~
-j
ARG
1
-^
A23
1
ARG
,
r
" 1"
100 " Artery
l
75
A23 ARG
I
II
Poly-L-Arg 115 kDo
3x10-8 M
-r
I »
g
I
T
y^
JL
M^
50
x^—
-
240
- 180
- 120
-
60
n
4
ft
Cl l>
o
0
%
L ^
15mln i
j
Vein
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
1
GTN
Perf
—1
—\s~
I
1~
3
-.
sj
•*|
PYR
—
^
15mln
A23
ARG
1 1
LYS
1
HbO2
f
A23 ARG LYS
IM
I
120
180
240
Tim* (*«c)
~^—
f
3. Release of EDNO from perfused artery and
vein by basic polyamino acids. Endothelium-denuded
strips of bovine intrapulmonary artery or vein were
precontracted sub maximally (50-70% of maximum) by
superfusion with U46619 (10'^-lQ-' M). The numbers 1,
2, and 3 signify the cascade arrangement of the strips.
Glyceryl trinitrate (GTN, 10~7 M) was superfused over the
strips for 1 minute. P'erfsignifies perfusate delivered from
the unrubbed segment of artery or vein. Breaks in the
tracings represent periods of tissue equilibration. A23187
(A23, 3xlO~7 M), poly-L-arginine 41 kDa (ARG, 3x10-'
M), and poly-L-lysine 41 kDa (LYS, 1.6x10-' M) were
perfused through artery and/or vein for 3 minutes as
indicated. Superoxide dismutase (SOD, 100 units/ml) was
perfused through the arterial segment during concomitant perfusion with A23 or ARG as indicated. Pyrogallol
(PYR, 10~5 M) was superfused over the strips during
concomitant perfusion of artery with A23 or ARG. Oxyhemoglobin (HbO2, 10'6 M) was superfused over the
strips during concomitant perfusion of vein with A23,
ARG, or L YS. Each panel of tracings is representative
from a total offour separate experiments.
FIGURE
both artery and vein. None of the basic polyamino
acids stimulated the formation of cyclic GMP in
endothelium-denuded artery or vein (not shown). Table
2 illustrates that oxyhemoglobin abolished and methylene blue markedly inhibited both arterial cyclic
GMP accumulation and relaxation elicited by polyL-arginine 115 kDa, poly-L-lysine 41 kDa, and polyL-lysine 225 kDa. Similar observations were made in
vein with poly-L-arginine 115 kDa and poly-L-lysine
225 kDa (not shown).
4. Time-dependent relaxation and cyclic GMP
accumulation in artery and vein elicited by polyL-arginine 115 kDa. Unrubbed arterial and venous rings
were mounted under 4 g of tension, and were precontracted submaximally (65-80% of maximum) by addition
ofU46619 (10~9-10~s M). After peak contractile responses
developed, 3x10'' M poly-L-arginine 115 kDa (PolyL-Arg) was added. Rings were quick-frozen at the indicated times after addition of the polyamino acid. Relaxation was monitored up until the time of tissue freezing.
Percent relaxation signifies percent decrease in submaximal precontractile tone elicited by U46619. Values represent the mean±SEM from 12-16 rings isolated from
three to four animals (four rings per animal).
FIGURE
Refractoriness to Endothelium-Dependent Arterial
and Venous Relaxation Produced by Basic
Polyamino Acids
Throughout the conduct of this study we recognized that relatively high concentrations of the
basic polyamino acids which produced 100% relaxant responses were much less effective in causing
relaxation in the same vascular preparations during
second exposure to the relaxant. No such diminution in relaxant responses to acetylcholine, bradykinin, or A23187 was ever observed. Figure 6
illustrates that a second exposure to low concentrations of poly-L-ornithine 50 kDa produced relaxant
responses that were similar in magnitude to the
initial relaxant responses and contractile responses
to U46619 were unaltered. Relaxant responses to a
second exposure of higher concentrations of polyL-ornithine 50 kDa, however, were markedly diminished but glyceryl trinitrate still produced maximal
relaxation (Figure 6). Concentration-response curves
to glyceryl trinitrate or NO were unaltered in the
Ignarro et al
100
- Artsry
Poly-L-Arg 115 kDa
180 we
Polyamino Acid Relaxation and Cydic GMP Formation
TABLE 2. Inhibition by Oxyhemogk>bln and Methylene Blue of
Arterial Relaxation and Cyclic GMP Accumulation Elicited by
Basic Polyamioo Adds
300
75
225
Test agents
50
None
Poly-L-arginine 115 kDa,
5xlO"'M
150
JL
n
• 25
75
o
S
8
0
10"8
3*10-8
10' 7
o I
•o
100
5
I
-
Vein
300
Poly-L-Lyi 225 kDo
180 M K
2.
225 f|j
150
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
T
50
25
0
0
|
JL
75
I
3x10" 9
75
10"8
3x10-8
321
0
Concentration (M)
FIGURE 5. Concentration-dependent relaxation and
cyclic GMP accumulation in artery and vein elicited by
poly-L-arginine 115 kDa and poly-L-lysine 225 kDa.
Unrubbed arterial and venous rings were mounted under
4 g of tension, and were precontracted submaximally
(65-80% of maximum) by addition of U46619 (W^-IO''
M). After peak contractile responses developed, polyL-arginine 115 kDa (Poly-L-Arg) or poly-L-lysine 225 kDa
(Poty-L-Lys) was added to artery or vein, respectively, at
the indicated bath concentrations. Rings were quickfrozen at 180 seconds after addition of polyamino acid, or
at the time of peak contraction to U46619 in the absence
of added test agent (zero concentration). Relaxation was
monitored up until the time of tissue freezing. Percent
relaxation signifies percent decrease in submaximal precontractile tone elicited by U46619. Values represent the
mean±SEM using 12-16 rings isolated from three to four
animals (four rings per animal).
presence of basic polyamino acids (not shown).
Moreover, rings that were pretreated with polyamino acids were supersensitive to the contractile
effects of U46619. Therefore, 10-fold lower concentrations of U46619 were used to precontract the
refractory rings.
Arterial rings were pretreated with high concentrations of poly-L-arginine 41 kDa and subsequently
challenged with acetylcholine after tissue washing,
equilibration and precontraction. The relaxant
responses to acetylcholine were abolished, whereas
maximal relaxation was still obtained with glyceryl
trinitrate (Figure 6). In arterial rings cut from the
same vessel used to prepare rings for the preceding
experiment, acetylcholine failed to produce any
impairment of relaxation elicited by itself (Figure 6)
or by poly-L-arginine 41 kDa (not shown). Similarly,
Cyclic GMP
(pmol/g tissue)
Percent
relaxation
34±3
212+31
22±4*
81±11*
72±9
0'
16±3*
Poly-L-lysine 41 kDa, 5x 10"8 M
+1 /xM oxyhemoglobin
+10 JAM methylene blue
168±24
18+3*
44+6*
58±7
0*
8±2*
Poly-L-lysine 225 kDa, 5x10"' M
+1 ^iM oxyhemoglobin
+10 fiM methylene blue
232±40
46±5*
74 ±9*
80+11
4±1*
17+3*
+1 /JM oxyhemoglobin
+ 10 ^iM methylene blue
Unrubbed arterial rings were mounted under 4 g of tension and
were precontracted submaximally (65-75% of maximum) by
addition of U46619 (10"'-10"' M). Oxyhemoglobin and methylene blue were added 10 minutes before initiation of precontraction. After peak contractile responses developed, the indicated
polyamino acids were added (5x10"' M). Rings were quickfrozen 180 seconds after addition of polyamino acid. Control
rings (indicated as None) were quick-frozen at the time of peak
contractile responses to U46619. Percent relaxation signifies
percent decrease in submaximal precontractile tone elicited by
U46619. Values represent the mean±SEM from 18 rings isolated
from three animals (six rings per animal).
*Significantly different (p<0.01) from corresponding values
obtained in the absence of inhibitor.
high concentrations of poly-L-lysine 225 kDa abolished relaxant responses to A23187, and rendered
the rings supersensitive to the contractile effects of
U46619, whereas maximal relaxation was obtained
with glyceryl trinitrate (Figure 6). A23187 did not
produce any impairment of relaxation to itself or to
poly-L-lysine 225 kDa. Arterial rings pretreated
with poly-L-lysine 225 kDa that showed refractoriness to acetylcholine did show partial relaxation in
response to bradykinin (not shown in Figure 6 but is
demonstrated by bioassay in Figure 7). Indomethacin abolished this partial relaxant response to
bradykinin.
Experiments were conducted to determine whether
refractoriness develops to cyclic GMP accumulation
as well. Table 3 illustrates that pretreatment with
poly-L-lysine 225 kDa nearly abolished cyclic GMP
accumulation and relaxation in response to either the
polyamino acid or acetylcholine, whereas no inhibition developed to the actions of glyceryl trinitrate.
Similarly, refractoriness developed to the actions of
poly-L-lysine 225 kDa and bradykinin but not glyceryl trinitrate in vein (Table 3). Resting levels of
cyclic GMP in rings pretreated with the polyamino
acid did not differ from control rings.
The bioassay cascade superfusion technique was
used to ascertain whether poly-L-lysine 225 kDa
(3x 1(T7 M) could impair the release of EDNO from
perfused arterial segments elicited by 10~8 M poly-
322
Circulation Research
ORN SO kDa
ARTERY
Vol 64, No 2, February 1989
ORN SO kDa
c d •
ARTERY
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
ARTERY
U4-«
L-lysine 225 kDa and 10~6 M A23187. After obtaining control responses to perfused A23187 and the
polyamino acid, and to superfused 10~7 M glyceryl
trinitrate, a 30-fold higher concentration of polyamino acid was perfused through the artery (but not
superfused over the bioassay strips) for 30 minutes.
The arterial perfusate was then allowed to superfuse the strips, and A23187 and the polyamino acid
were retested. No evidence of EDNO release from
arterial segments pretreated with high concentrations of poly-L-lysine 225 kDa was seen, whereas
the relaxant responses to superfused glyceryl trinitrate were comparable to control responses (Figure
7). Release of EDNO from perfused intact artery
was quantified by chemical assay as described.27
Table 4 illustrates that arterial perfusates with or
without A23187 or poly-L-lysine 225 kDa evoke the
formation and/or release of EDNO that is associated with concomitant relaxation of the superfused
arterial strips. After pretreatment of intact artery
with the basic polyamino acid for 30 minutes,
however, both relaxation and formation/release of
EDNO in response to arterial perfusion are markedly impaired.
Perfusion of normal artery with A23187 or bradykinin in the absence of indomethacin caused the
release of an EDRF that appeared to be much more
FIGURE 6. Development of refractoriness to
endothelium-dependent arterial relaxation
caused by high concentrations of basic polyamino acids. Unrubbed arterial rings were
mounted under 4 g of tension, and precontracted submaximally (60-80% of maximum)
by addition ofU46619 (U4, KT'-IO'' M). PolyL-ornithine 50 kDa (ORN), poly-L-arginine 41
kDa (ARG), and poly-L-lysine 225 kDa (LYS)
were added cumulatively at bath concentrations of 1.6x10-' M (c), 5xl(T' M (d), and/or
1.6X10T1 M (e) as indicated. Acetylcholine
(Ach), A23187, and glyceryl trinitrate (GTN)
were added cumulatively at the indicated bath
concentrations (expressed as exponents to the
base power 10). Horizontal arrows signify t's~
sue washing and 45 minutes of equilibration in
between the two sets of responses. The upper
left pair of tracings represent responses to
lower concentrations of ORN 50 kDa that do
not cause refractoriness or enhanced contractile responses to U4. The upper right pair of
tracings represent responses to higher concentrations of ORN 50 kDa that induce refractoriness and enhance contractile responses to U4,
as do the center left and lower left pairs of
tracings. The center right and lower right pairs
of tracings represent responses to Ach and
A23187, respectively, which do not cause refractoriness. Each of the individual sets of
tracings illustrated is representative from a
total of four to six separate experiments.
stable chemically than that released in the presence
of indomethacin (Figure 7). This was particularly
true for bradykinin and is attributed to the generation of a prostanoid relaxing factor such as prostacyclin. Following treatment of the perfused artery
with poly-L-lysine 225 kDa as described above, the
release of EDRFs by A23187 in the presence of
oxyhemoglobin was nearly abolished, whereas
release by bradykinin was only slightly inhibited.
Such relaxant responses to perfused bradykinin
were nearly abolished, however, in the presence of
indomethacin.
Scanning electron microscopy was employed to
determine whether the basic polyamino acids caused
morphologic damage to vascular endothelial cells.
The photomicrographs illustrated in Figures 8 and 9
reveal that exposure of arterial rings to high concentrations of poly-L-ornithine 50 kDa, polyL-arginine 41 kDa, or poly-L-lysine 225 kDa for 15
minutes did not cause overt endothelial cell damage. Rubbing the intimal surface as described in the
methods section completely denuded the surface of
endothelial cells (Figure 8). Mounting the arterial
rings in organ baths followed by manipulations such
as washing and contracting the tissues alters the
morphological appearance of the endothelial cells
when compared with that of freshly sliced rings
Ignarro et al
Artery
Tol
GTN Perf A23
1
Polyamino Add Relaxation and Cyclic GMP Formation
LYS
T
A23
1
LY 5
"
GTN
\
2
3
si
—1
t
1
* | 15min
Artery
Tol
GTN A23
Ind
BKN~|~A23 BKN GTN
BKN ~T~A23
1
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
2
\
3
—i
~~\
—
—L
~1
s| 1/
•*| 15min
FIGURE 7. Development of refractoriness to formation
and/or release of EDNO from perfused artery caused by
poly-L-lysine 225 kDa. Endothelium-denuded strips of
bovine intrapulmonary artery were precontracted submaximally (50-70% of maximum) by superfusion with
U466J9 (10-9-10-* M). The numbers 1, 2, and3 signify the
cascade arrangement of the strips. Glyceryl trinitrate
(GTN, 10~7 M) was superfused over the strips for 1
minute. Perf signifies perfusate delivered from the
unrubbed segment of artery. Breaks in the tracings represent periods of tissue equilibration. A23187 (A23,3 x 10~7
M), bradykinin (BKN, 3x10-' M), and poly-L-lysine 225
kDa (LYS, L6xlO'' M) were perfused through the artery
for 3 minutes as indicated. Tol signifies tolerance (impaired
EDNO formation/release) induced by perfusion of artery
with 1.6xlO'7 M poly-L-lysine 225 kDa for 30 minutes,
without allowing the perfusate to superfuse the strips. In
the lower panel, 10'7 M oxyhemoglobin was added to the
superfusion medium after induction of tolerance for the
duration of the experiment. Ind signifies indomethacin
(5xlO~5 M), which was present in both the perfusion and
superfusion media for the duration of the experiment.
Each panel of tracings is representative from a total of
three separate experiments.
(Figure 8). Tissue manipulation or prolonged exposures to bathing solutions caused retraction ofendothelial cell borders, perforations, and detachments
of cell edges without causing denudation or loss of
responsiveness to endothelium-dependent relaxants
(Figure 9A). Exposure to the basic polyamino acids
as indicated did not exaggerate the above morphologic changes (Figure 9B-D).
323
Vascular Actions of Bradykinin Analogues and
Other Peptides
Bradykinin, Lys-bradykinin, and Met-Lysbradykinin relaxed arterial rings with comparable
potencies (Figure 10). However, changing the Lproline to D-phenylalanine at position 7 or Larginine to L-lysine at position 1 abolished relaxant
activity (Figure 10). Similar observations were made
with venous rings. Neurotensin and several analogues of neurotensin and bradykinin, each of which
contains arginine and/or lysine residues, were completely inactive at concentrations up to 3xl0~ 7 M
(not shown). The r>Phe analogue of bradykinin was
found to antagonize the relaxant effect of bradykinin but not that of poly-L-arginine 115 kDa (Figure
10). The above observations indicate that the endothelial cell receptors involved with the relaxant
action of bradykinin are different from those involved
with the basic polyamino acids.
Discussion
The present observations indicate that high molecular weight polyamino acids composed of the basic
amino acids L-arginine, L-lysine, or L-ornithine cause
relaxation of isolated rings of bovine intrapulmonary artery and vein by endothelium-dependent
mechanisms that are attributed to the formation and
release of one type of EDRF. This particular EDRF
is EDNO as demonstrated recently in bovine pulmonary artery and vein by this laboratory16-2732 and
in cultured porcine aortic endothelial cells by Palmer
et al.36 The average molecular weights of the basic
polyamino acids tested ranged from 17 to 225 kDa,
and the EC*, values ranged from 3 x 10~9 to 2x 10"7 M
in artery and vein. No significant relaxant responses
were elicited by acidic or neutral polyamino acids of
comparable molecular weights, polyamines, or basic
amino acids. However, poly-D-lysine was just as
potent a relaxant as poly-L-lysine. There was a close
direct correlation between molecular weight and relaxation potency among the basic polyamino acids, and
this relation was similar in artery and vein. The most
potent compound tested was poly-L-lysine 225 kDa,
which had an EC*, of 3x10"' M and a minimal
effective concentration of about 10"'° M. The present
findings confirm and extend those of a recent report,23
which indicated that poly-L-arginine and polyL-lysine, each 70 kDa, caused endothelium-dependent
relaxation of rat aorta.
The general mechanism of endothelium-dependent
vascular smooth muscle relaxation elicited by the
basic polyamino acids is attributed to enhanced formation and release of EDNO. This view is supported
by the following experimental evidence. The basic
polyamino acids produced endothelium-dependent,
time-dependent, and concentration-dependent increases in arterial and venous cyclic GMP levels,
and cyclic GMP accumulation correlated well with
concomitant relaxant responses in the same vascular rings. Both cyclic GMP accumulation and relax-
324
Circulation Research
Vol 64, No 2, February 1989
TABLE 3. Impairment by High Concentrations of Poly-L-Lysint 225 kDa of Subsequent Relaxation and Cydk GMP Accumulation in Artery
and Vein
No pre treatment
Test agents
Artery
None
Poly-L-lysine 225 kDa
(5xlO-»M)
Acetylcholine (10"* M)
Glyceryl trinitrate (10"* M)
Vein
None
Poly-L-lysine 225 kDa
(5xl0-«M)
Bradykinin (10"' M)
Glyceryl trinitrate (KT6 M)
Pre treatment with poly-L-lysine 225
kDa
Percent
relaxation
Cyclic GMP
Cyclic
GMP
Percent
relaxation
34±3
221±34
77±9
36±5
60±9*
8+2*
332±46
298 ±37
82±11
85 ±14
62±8*
286+33
0*
87±12
33±5
254±37
79±14
38±4
41±6*
0*
284±41
263 ±36
71±12
78±10
54±8*
233±37
0'
8I±14
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Unrubbed arterial and venous rings were mounted under 4 g of tension, and were precontracted submaximally (65-75% of maximum)
by addition of U46619 (10"'-10"' M). The indicated test agent was added after peak contractile responses developed. Rings were
quick-frozen either at 180 seconds after addition of poly-L-lysine 225 kDa or at 60 seconds after addition of acetylcholine, bradykinin or
glyceryl trinitrate as indicated. Control rings (indicated as None) were quick-frozen at the time of peak contractile responses to U46619.
Pretreatment signifies rings that were initially relaxed by exposure to 5x 10"7 M poly-L-lysine 225 kDa for 20 minutes followed by tissue
washing and 45 minutes of equilibration. Percent relaxation signifies percent decrease in submaximal precontractile tone elicited by
U46619. Values represent the mean±SEM from 12 rings isolated from three animals (four rings per animal).
'Significantly different (p<0.01) from corresponding values obtained in the absence of tolerance.
ation were abolished or markedly inhibited by oxyhemoglobin or methylene blue. Hemoglobin and
related hemoproteins have been shown to inhibit
the biological actions of at least one EDRF 3437 and
NO.30.M.39 Methylene blue inhibits the actions of
this EDRF and NO by preventing their activation of
soluble guanylate cyclase, thereby attenuating cyclic
TABLE 4. Influence of Poly-L-Lysine 225 kDa on Formation/
Release of EDNO and on Arterial Relaxation
NO
Test conditions
ml)
Control
Before pretreatment
Arterial perfusate
0
+0.3MMA23187
+5 nM poly-L-lysine 225 kDa
After pretreatment
Arterial perfusate
+0.3/iM A23187
+5 nM poly-L-lysine 225 kDa
0.04±0.01
Percent
relaxation
0.42+0.05
12±2
73±9
60±8
0
0.05±0.01
0.06±0.02
0
12±1
10±2
0.51±0.07
Endothelium-denuded strips of bovine intrapulmonary artery
were precontracted submaximally (60-75% of maximum) by
superfusion with U46619 (10"'-10"8 M). One strip per experiment was superfused.27 Control signifies absence of arterial
perfusion in that strips were superfused only with Krebsbicarbonate solution. Pretreatment signifies perfusion of intact
artery with 1.6xlO"7 M poly-L-lysine 225 kDa for 30 minutes,
without allowing the perfusate to superfuse the strip. Arterial
perfusate signifies superfusion of strip with perfusates delivered
through intact artery in the absence or presence of A23187 or
poly-L-lysine 225 kDa, as indicated. The superfusion medium
from each strip was collected (25 ml), starting at the time of
superfusion with arterial perfusate (except Control), into a flask
containing 2.5 ml of 4 M HC1. Aliquots were assayed for NO as
described.27 Percent relaxation signifies percent decrease in
submaximal precontractile tone elicited by U46619. Values represent the mean±SEM from four separate experiments.
GMP accumulation in vascular smooth muscle.25-38-41
Indomethacin failed to modify the vascular actions of
the basic polyamino acids, thereby indicating that
cyclooxygenase products of arachidonic acid such as
prostacyclin are not involved. Moreover, no metabolite of arachidonic acid appears to be involved as
neither quinacrine nor nordihydroguaiaretic acid inhibited relaxant responses to the polyamino acids. The
bioassay technique revealed that both the basic
polyamino acids and the calcium ionophore A23187
caused the release of EDNO from perfused artery
and vein. An approximate half-life of 3-5 seconds
was observed for EDNO, and the apparent stability
of EDNO was increased by superoxide dismutase
and markedly decreased by superoxide anion (generated by pyrogallol). These observations are consistent with those reported for EDNO released from
cultured aortic endothelial cells by bradykinin or
A23187 as assessed by bioassay. 364243 Arterial rings
that were precontracted by depolarizing potassium
solution were markedly less sensitive than were
U46619-precontracted rings to the relaxant effects of
acetylcholine and basic polyamino acids. It is well
appreciated, although not well understood, that
EDRF-elicited vascular relaxant responses are
greatly diminished in potassium-depolarized vascular
preparations, whereas endothelium-independent
vasodilators markedly relax such preparations.12'25-2744
Finally, the release of EDNO from artery and vein
caused by basic polyamino acids was verified by
chemical assay, as was previously done for EDNO
released by A23187."
The precise mechanism by which the basic polyamino acids stimulate the formation and/or release
of EDNO is presently unknown. Possible mecha-
Ignarro et al
Polyamino Add Relaxation and CydJc GMP Formation
325
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
FIGURE 8. Scanning electron microscopy of arterial rings before and after endothelial denudation. Intimal surfaces were
photographed at X6O0 magnification by scanning electron microscopy. Panel A: Arterial ring was fixed immediately after
removal from the arterial vessel. Panel B: Arterial ring was prepared, rubbed along its intimal surface, and then fixed.
nisms include interaction with 1) extracellular receptors or ion channels that are coupled to EDNO
formation and/or release, 2) intracellular processes
that govern the release of preformed or stored
EDNO, and 3) intracellular enzyme systems associated with EDNO formation. The first possibility is
unlikely because the data are not consistent with an
interaction between polyamino acids and selective
extracellular receptors. The basic polyamino acids
elicit much slower rates of vascular smooth muscle
relaxation than do acetylcholine, bradykinin, or
A23187, as observed with isolated vascular rings
and by bioassay. Clearly, the data indicate that the
extracellular receptors with which bradykinin and
its active analogues interact are not involved in the
actions of the basic polyamino acids. Moreover, the
finding that the polyamino acids impair or cause
refractoriness to the relaxant effects of acetylcholine, bradykinin, and A23187 argue against a selective interaction between basic polyamino acids and
extracellular receptors. Finally, the findings that
basic polyamino acids rich in arginine, lysine, or
ornithine are equally active and that the L- and
D-isomers of polylysine are equipotent as relaxants
further support the view that the basic polyamino
acids do not interact with selective cell surface
receptors. The second possibility appears to be
unlikely at present because NO is highly unstable,
possessing a half-life of only 3-5 seconds1627-3236
and is lipophilic, so that storage of preformed free
NO may not occur to any appreciable extent.
The most likely mechanism of action of the basic
polyamino acids is that they penetrate endothelial
cells by diffusion or special transport systems and
facilitate the formation of EDNO. These basic polyamino acids could serve as precursors for the synthesis of NO. More specifically, the basic amino
groups could be cleaved and oxidized to NO or a
more stable nitroso-derivative by enzyme systems
that apparently catalyze the conversion of Larginine to NO in cultured porcine aortic endothelial cells as reported recently.45 The component
pathways of this enzyme system are presently
unknown but may possess broad substrate specificity for polypeptides rich in various basic amino
acids. L-Arginine may serve as an endogenous
precursor to EDNO/ 5 but exogenous peptides or
polyamino acids rich in arginine or other basic
amino acids could behave as alternate or partial
enzyme substrates. Consistent with this proposed
mechanism of action are the observations that 1)
vascular smooth muscle relaxation occurs in
response only to basic and not to the corresponding
acidic or neutral polyamino acids, 2) relaxant potencies increase with increasing molecular weights
(greater number of moles of basic amino functions
per mole of basic polyamino acid), 3) the rates of
relaxation in response to these substances are much
slower than those to agents that interact with membrane receptors or ion channels, and this may be
attributed to a slower rate of conversion of the basic
polyamino acids to NO; and 4) prolonged interaction
between polyamino acid and vascular preparations
326
Circulation Research
Vol 64, No 2, February 1989
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
FIGURE 9. Scanning electron microscopy of arterial rings after treatment with basic polyamino acids. Intimal surfaces
were photographed at x600 magnification in Panels A, B, and C, and at xl,500 magnification in Panel D by scanning
electron microscopy. Panel A: Unrubbed arterial ring was mounted in a bath chamber under tension, depolarized with
KC1, contracted with 70~* M U46619, rinsed with Krebs solution and then fixed. Panel B: Unrubbed arterial ring was
mounted and precontracted with U46619 ad described above, challenged with 10~7 M poly-L-ornithine 50 kDa for 15
minutes, rinsed and then fixed. Panel C: Unrubbed arterial ring was handled as in Panel B except that 10'7 M
poty-L-tysine 225 kDa was tested. Panel D: Same as Panel C except at higher magnification.
Ignarro et al
Polyamino Add Relaxation and Cydic GMP Formation
ARTERY
BKN
-"-in
lyi-fiKN
d'-io
"-9
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
FIGURE 10. Effects of bradykinin analogues on arterial
smooth muscle relaxation. Unrubbed arterial rings were
mounted under 4 g of tension, and precontracted submaximally (60-80% of maximum) by 10'9 M or W* M
U46619 (U4) as indicated. Bradykinin (BKN), Lysbradykinin (Lys-BKN), Met-Lys-bradykinin
(MetLys-BKN), [Lys1]-bradykinin ([Lys']-BKN), [D-Phe7]bradykinin ([D-Phe7]-BKN), and poly-L-arginine 115 kDa
(ARG 115 kDa) were added cumulatively at the bath
concentrations indicated (expressed as exponents to the
base power 10). The bath concentrations for ARG 115
kDa are a, 1.6xl0r9 M; b, 5 x / 0 - ' M; c, 1.6x10-' M.
Acetylcholine (Ach) was added at 1O~6 M where indicated. Horizontal arrows signify that 10'7 M [DPhe7]-BKN was added 10 minutes before precontracting
rings by addition of U46619. Each of the individual
tracings illustrated is representative from a total of three
to four separate experiments.
causes refractoriness or a form of tolerance characterized by impairment of EDNO formation in
response to any vasodilator whose action is mediated by EDNO.
The refractoriness to vascular smooth muscle
relaxation that develops to higher concentrations of
the basic polyamino acids or after prolonged tissue
contact is attributed to a desensitization of one or
more enzymes that may be involved in the conversion of the basic amino groups to NO. If arginine is
indeed the natural substrate for the enzyme that
initiates the reaction sequence leading to the formation of NO,45 then the basic polyamino acids could
conceivably act as alternate or partial substrates
whereby they compete with arginine for enzyme
binding sites and initially serve as substrates. As
these binding sites become saturated with polyamino
327
acid, however, enzyme desensitization occurs and
the formation of EDNO is impaired. This hypothesis
is supported by the observations that 1) refractoriness or tolerance occurs to the endotheliumdependent relaxant effects not only of all basic
polyamino acids but also to acetylcholine, bradykinin, and A23187 but does not occur to endotheliumindependent relaxants, and 2) impairment of NO
formation/release from perfused polyamino acidpretreated artery and vein, as assessed by bioassay
and chemical assay, occurs in response to various
endothelium-dependent vasodilators. Meaningful time
course experiments to determine whether vascular
preparations can recover from the refractoriness to
endothelium-dependent relaxants were not conducted
because the tissues begin to deteriorate after 7-8
hours in the tissue baths. Prolonged contact of the
basic polyamino acids with the endothelium-denuded
vascular strips in the bioassay cascade does not alter
relaxation in response to other relaxants such as
glyceryl trinitrate delivered by superfusion. Similarly,
isolated arterial and venous rings rendered tolerant by
pretreatment with basic polyamino acids undergo
complete relaxation in response to glyceryl trinitrate.
The apparent tolerance and cross-tolerance
observed with the basic polyamino acids cannot be
attributed to 1) desensitization at the level of soluble guanylate cyclase or action of cyclic GMP on
calcium-binding proteins because relaxant responses
to glyceryl trinitrate are no different from control
responses, 2) damage to the vascular endothelium
as assessed by scanning electron microscopy and
because bradykinin can still evoke the formation/
release of a prostanoid relaxing factor, 3) desensitization of selective extracellular endothelial receptors because refractoriness to all tested endotheliumdependent relaxants occurs and because the basic
polyamino acids do not interact with the receptors
that interact with bradykinin, Lys-bradykinin, and
Met-Lys-bradykinin; and 4) direct inactivation of
EDNO because relaxant responses to authentic NO
are no different from control. Thus, the mechanism
of refractoriness induced by the basic polyamino
acids may occur as a result of impairment of NO
formation in vascular endothelial cells.
The small arginine-containing polypeptide, bradykinin, and two of its close structural analogues elicit
endothelium-dependent relaxation by mechanisms
that are unrelated to those for the arginine-rich
polyamino acids. Bradykinin and vasoactive intestinal polypeptide, which contains both arginine and
lysine, interact with selective extracellular receptors that are coupled to the biosynthesis and/or
release of both EDNO, which stimulates cyclic
GMP formation, and prostacyclin, which stimulates
cyclic AMP formation.12 Atriopeptins, which contain both arginine and lysine, interact with vascular
smooth muscle receptors linked to particulate guanylate cyclase and cause endothelium-independent
relaxation.22 The actions of the arginine- or lysinerich polyamino acids tested in this study are attrib-
328
Circulation Research
Vol 64, No 2, February 1989
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
uted to endothelium-dependent mechanisms that do
not involve interactions with receptors for the
smaller arginine- or lysine-containing polypeptides.
Competitive antagonists of bradykinin do not alter
relaxant responses to basic polyamino acids, and
the latter do not elevate vascular levels of cyclic
AMP. Thus, there is no common or unifying mechanism of vasodilator action for polypeptides containing arginine and/or lysine.
In conclusion, basic polyamino acids rich in arginine, lysine, or ornithine elicit endothelium- and
cyclic GMP-dependent relaxation of artery and vein
by a novel mechanism. The hypothesis offered is
that basic polyamino acids act as partial or alternate
substrates for the enzyme system(s) that convert
L-arginine to NO, 45 thereby causing increased formation of EDNO. The refractoriness to other
endothelium-dependent vasodilators that develops
in vascular preparations after prolonged contact
between endothelium and basic polyamino acids is
attributed to enzyme desensitization that results in
decreased EDNO formation. The proposed mechanisms of action and refractoriness for the basic
polyamino acids are mutually compatible. It appears
that, although both EDNO and prostacyclin are
released concomitantly by diverse stimuli, 12144647
the two events can be dissociated not only by the
use of indomethacin but also by basic polyamino
acids. The basic polyamino acids are useful pharmacological probes for future studies on the biosynthesis and physiological actions of EDNO.
Acknowledgments
The authors wish to thank Karen Lee Roberts
and Ljuba Abascal for their expert editorial assistance, and Diane Rome Peebles for preparing the
illustrations.
References
1. Altura BM, Chand N: Bradykinin-induced relaxation of
renal and pulmonary arteries is dependent upon intact endothelial cells. Br J Pharmacol 1981 ;74:10-11
2. Chand N, Altura BM: Acetylcholine and bradykinin relax
intrapulmonary arteries by acting on endothelial cells: Role
in lung vascular disease. Science 1981 ;213:1376—1378
3. Chand N, Altura BM: Endothelial cells and relaxation of
vascular smooth muscle cells: Possible relevance to lipoxygenases and their significance in vascular diseases. Microcirculation 1981;1:297-317
4. Cherry PD, Furchgott RF, Zawadzki JV, Jothianandan D:
Role of endothelial cells in relaxation of isolated arteries by
bradykinin. Proc Nail Acad Sci USA 1982;72:2l06-2110
5. De Mey JG, Claeys M, Vanhoutte PM: Endotheliumdependent inhibitory effects of acetylcholine, adcnosine
triphosphate, thrombin and arachidonic acid in the canine
femoral artery. J Pharmacol Exp Ther 1982^22:166-173
6. Rapoport RM, Murad F: Agonist-induced endotheliumdependent relaxation in rat thoracic aorta may be mediated
through cGMP. Circ Res 1983^52:352-357
7. D'Orleans-Juste P, Dion S, Mizrahi J, Regoli D: Effects of
peptides and non-peptides on isolated arterial smooth muscles: Role of endothelium. Eur J Pharmacol 1985;114:9-21
8. Katusic ZS, Shepard JT, Vanhoutte PM: Vasopressin causes
endothelium-dependent relaxations of the canine basilar
artery. Circ Res 1984^5:575-579
9. Davies JM, Williams KI: Endothelial-dependent relaxant
effects of vaso-active intestinal polypeptide and arachidonic
acid in rat aortic strips. Proslaglandins 1984;27:195-202
10. Forstermann U, Neufang B: Endothelium-dependent vasodilation by melittin: Are lipoxygenase products involved?
Am J Physiol 1985;249:H14-H19
11. Gruetter CA, Lemke SM: Bradykinin-induced endotheliumdependent relaxation of bovine intrapulmonary artery and
vein. Eur J Pharmacol 1986;122:363-367
12. Ignarro LJ, Byms RE, Buga GM, Wood KS: Mechanisms of
endothelium-dependent vascular smooth muscle relaxation
elicited by bradykinin and VIP. Am J Physiol 1987;
253:H1074-H1082
13. Cocks TM, Angus JA, Campbell JH, Campbell GR: Release
and properties of endothelium-dcrived relaxing factor (EDRF)from endothelial cells in culture. J Cell Physiol 1985;
123:310-320
14. Gryglewski RJ, Moncada S, Palmer RMJ: Bioassay of prostacyclin and endothelium-derived relaxing factor (EDRF)
from porcine aortic endothelial cells. Br J Pharmacol 1986;
87:685-694
15. Forstermann U, Goppelt-Strube M, Frolich JC, Busse R:
Inhibitors of acyl-Coenzyme A: Lysolecithin acetyltransferase activate the production of endothelium-derived vascular relaxing factor. J Pharmacol Exp Ther 1986,238:352-359
16. Ignarro LJ, Byrns RE, Buga GM, Wood KS, Chaudhuri G:
Pharmacological evidence that endothelium-derived relaxing
factor is nitric oxide: Use of pyrogallol and superoxide
dismutase to study endothelium-dependent and nitric oxideelicited vascular smooth muscle relaxation. J Pharmacol
Exp Ther 1988,244:181-190
17. Ganz P, Sandrock AW, Landis SC, Leopold J, Gimbrone
MA, Alexander RW: Vasoactive intestinal peptide: Vasodilatation and cyclic AMP generation. Am J Physiol 1986;
250:H755-H760
18. Itoh T, Sasaguri T, Makita Y, Kanmura Y, Kuriyama H:
Mechanisms of vasodilation induced by vasoactive intestinal
polypeptide in rabbit mesenteric artery. Am J Physiol 1985;
249:H231-H240
19. Schoeffter P, Stoclet J-C: Effect of vasoactive intestinal
polypeptide (VIP) on cyclic AMP level and relaxation in rat
isolated aorta. Eur J Pharmacol 1985;109:275-279
20. Duckies SP, Said SI: Vasoactive intestinal peptide as a
neurotransmitter in the cerebral circulation. Eur J Pharmacol 1982;78:371-374
21. Lee TJF, Saito A, Berezin I: Vasoactive intestinal polypeptide-like substance: The potential transmitter for cerebral
vasodilation. Science 1984,224:898-901
22. Winquist RJ, Faison EP, Waldman SA, Schwartz K, Murad
F, Rapoport RM: Atrial natriuretic factor elicits an endothelium independent relaxation and activates paniculate guanylate cyclase in vascular smooth muscle. Proc Nail Acad
Sci USA 1984;81:7661-7664
23. Thomas G, Mostaghim R, Ramwell PW: Endothelium dependent vascular relaxation by arginine containing polypeptides. Biochem Biophys Res Commun 1986; 141:446-451
24. Ignarro LJ, Byms RE, Wood KS: Endothelium-dependent
modulation of cyclic GMP levels and intrinsic smooth muscle tone in isolated bovine intrapulmonary artery and vein.
Circ Res 1987;60:82-92
25. Ignarro LJ, Burke TM, Wood KS, Wolin MS, Kadowitz PJ:
Association between cyclic GMP accumulation and
acetylcholine-elicited relaxation of bovine intrapulmonary
artery. / Pharmacol Exp Ther 1984^28:682-690
26. Ignarro LJ, Buga GM, Chaudhuri G: EDRF generation and
release from perfused bovine pulmonary artery and vein.
Eur J Pharmacol 1988;149:79-88
27. Ignarro LJ, Buga GM, Wood KS, Byms RE, Chaudhuri G:
Endothelium-derived relaxing factor produced and released
from artery and vein is nitric oxide. Proc Nail Acad Sci USA
1987;84:9265-9269
28. Vane JR: The use of isolated organs for detecting active
substances in the circulating blood. Br J Pharmacol Chemother 1964^23:360-373
Ignarro el al
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
29. Gruetter CA, Gruetter DY, Lyon JE, Kadowitz PJ, Ignarro
LJ: Relationship between cyclic GMP formation and relaxation of coronary arterial smooth muscle by glyceryl trinitrate, nitroprusside, nitrite and nitric oxide: Effects of methylene blue and methemoglobin. J Pharmacol Exp Ther 1981;
219:181-186
30. Ignarro LJ, Lippton H, Edwards JC, Baricos WH, Hyman
AL, Kadowitz PJ, Gruetter CA: Mechanism of vascular
smooth muscle relaxation by organic nitrates, nitrites, nitroprusside and nitric oxide: Evidence for the involvement of
S-nitrosothiols as active intermediates. J Pharmacol Exp
Ther 1981;218:739-749
31. Frank GW, Bevan JA: Electrical stimulation causes
endothelium-dependent relaxation in lung vessels. Am J
Physiol 1983;244:H793-H798
32. Ignarro LJ, Byrns RE, Buga GM, Wood KS: Endotheliumderived relaxing factor from pulmonary artery and vein
possesses pharmacological and chemical properties that are
identical to those for nitric oxide radical. Ore Res 1987;
61:866-879
33. Furchgott RF, Zawadzki JV: The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by
acetylcholine. Nature 1980;288:373-376
34. Martin W, Villani GM, Jothianandan D, Furchgott RF:
Selective blockade of endothelium-dependent and glyceryl
trinitrate-induced relaxation by hemoglobin and by methylene blue in the rabbit aorta. / Pharmacol Exp Ther 1985;
232:708-716
35. Marklund S, Marklund G: Involvement of the superoxide
anion radical in the autoxidation of pyrogallol and a convenient assay for superoxide dismutase. Eur J Biochem 1974;
47:469-474
36. Palmer RMJ, Ferrige AG, Moncada S: Nitric oxide release
accounts for the biological activity of endothelium-derived
relaxing factor. Nature 1987;327:524-526
37. Martin W, Villani GM, Jothianandan D, Furchgott RF:
Blockade of endothelium-dependent and glyceryl trinitrateinduced relaxation of rabbit aorta by certain ferrous hemoproteins. J Pharmacol Exp Ther 1985233:679-685
38. Gruetter CA, Barry BK, McNamara DB, Gruetter DY,
Kadowitz PJ, Ignarro LJ: Relaxation of bovine coronary
Polyamino Acid Relaxation and Cyclic GMP Formation
39.
40.
41.
42.
43.
44.
45.
46.
47.
329
artery and activation of coronary arterial guanylate cyclase
by nitric oxide, nitroprusside and a carcinogenic nitrosoamine. J Cyclic Nucleotide Res 1979;5:211-224
Gruetter CA, Barry BK, McNamara DB, Kadowitz PJ,
Ignarro LJ: Coronary arterial relaxation and guanylate cyclase
activation by cigarette smoke, N'-nitrosonornicotine and
nitric oxide. J Pharmacol Exp Ther 198O;214:9-I5
Gruetter CA, Kadowitz PJ, Ignarro LJ: Methylene blue
inhibits coronary arterial relaxation and guanylate cyclase
activation by nitroglycerin, sodium nitrite and amyl nitrite.
Can J Physiol Pharmacol 1981^9:150-156
Ignarro LJ, Harbison RG, Wood KS, Kadowitz PJ: Activation of purified soluble guanylate cyclase by endotheliumderived relaxing factor from intrapulmonary artery and vein:
Stimulation by acetylcholine, bradykinin and arachidonic
acid. J Pharmacol Exp Ther 1986^37:893-900
Gryglewski RJ, Palmer RMJ, Moncada S: Superoxide anion
is involved in the breakdown of endothelium-derived vascular relaxing factor. Nature 1986,320:454-456
Moncada S, Palmer RMJ, Gryglewski RJ: Mechanism of
action of some inhibitors of endothelium-derived relaxing
factor. Proc Natl Acad Sci USA 1986;83:9164-9168
Furchgott RF: The role of endothelium in the responses of
vascular smooth muscle to drugs. Ann Rev Pharmacol 1984;
24:175-197
Palmer RMJ, Ashton DS, Moncada S: Vascular endothelial
cells synthesize nitric oxide from L-arginine. Nature 1988;
333:664-666
Ignarro LJ, Harbison RG, Wood KS, Wolin MS, McNamara
DB, Hyman AL, Kadowitz PJ: Differences in responsiveness of intrapulmonary artery and vein to arachidonic acid:
Mechanism of arterial relaxation involves cyclic GMP and
cyclic AMP. J Pharmacol Exp Ther 1985;233:560-569
de Nucci G, Gryglewski RJ, WarnerTD, Vane JR: Receptormediated release of endothelium-derived relaxing factor and
prostacyclin from bovine aortic endothelial cells is coupled.
Proc Natl Acad Sci USA 1988;85:2334-2338
KEY WORDS • bovine pulmonary artery and vein • vascular
smooth muscle • cyclic GMP • nitric oxide • arginine •
lysine • tolerance • polyamino acid • endotheliumdependent vasodilation
Basic polyamino acids rich in arginine, lysine, or ornithine cause both enhancement of and
refractoriness to formation of endothelium-derived nitric oxide in pulmonary artery and
vein.
L J Ignarro, M E Gold, G M Buga, R E Byrns, K S Wood, G Chaudhuri and G Frank
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Circ Res. 1989;64:315-329
doi: 10.1161/01.RES.64.2.315
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1989 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/64/2/315
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/