Role of Adenlne Nucleotides, Adenosine,
and Inorganic Phosphate En the Regulation
of Skeletal Muscle Blood Flow
By James G. Dobson, Jr., Rafael Rubio, and Robert M . Berne
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
ABSTRACT
Experiments were performed on isolated frog sartorius muscle and in situ
dog skeletal muscle to determine whether adenine nucleotides and their
degradation products are released during contraction in concentrations capable
of producing arteriolar dilation. ATP was not detectable ( < 1 O - 8 M ) in the
bathing solution of the resting or contracting frog sartorius muscle. Inorganic
phosphate (P,) in the muscle bath increased from 9 x lO"6** to 28 X K H ^
with 30 minutes of contraction (2 Hz) or with rest. With the dog hindlimb
preparation, ATP, ADP, and AMP were not detectable ( < 5 X 1 0 ~ 8 M ) in
the venous blood collected after 5 minutes of ischemic contraction whereas P t
was present at a concentration of 3.7 X ICMM. Arterial blood levels required to
elicit detectable vasodilation for ATP, ADP, AMP, and P, were 28.7 X 1 ( H M ,
27.1 X K H M , 31.4 X K H M and 7.2 X I O ^ M , respectively. The adenosine
concentration in dog muscle increased from 0.7 to 1.5 nmole/g with ischemic
contraction, and hypoxanthine and inosine increased from 4.5 to 8.5 nmole/g
and 2.0 to 5.5 nmole/g, respectively. The adenosine concentration in venous
plasma collected from the hindlimb immediately after termination of the
ischemic contraction period was 2.2 X 10~ 7 M as compared to 0.4 X 10~ 7 M in
control venous and arterial blood samples. Hypoxanthine and inosine
concentrations in venous blood increased 22- and 270-fold, respectively,
following ischemic contraction. The calculated interstitial fluid adenosine
concentration was twice the arterial concentration of adenosine required to
elicit maximal arteriolar dilation. These findings suggest that adenosine may
play a role in the metabolic regulation of skeletal muscle blood flow, whereas
ATP, ADP, AMP, and P, may not.
KEY WORDS
autoregulation of blood flow inosine
hypoxanthine
skeletal muscle ischemia
skeletal muscle contraction
peripheral resistance
reactive hyperemia
functional hyperemia
purine derivatives in muscle
dog
frog
vascular smooth muscle tone
• For a number of years, the adenine
nucleotides have been periodically suggested
as mediators of the vasodilation associated
with an increase in metabolic activity or
From the Department of Physiology, University of
Virginia School of Medicine, Charlottesville, Virginia
22901.
This investigation was supported by G""ant HE
10384 from the National Heart and Lung Institute.
Dr. Dobson's present address is Division of Pharmacology, Department of Medicine, School of Medicine,
University of California, La Jolla, California 92037.
Because the present editor is one of the authors of
this paper, we asked Dr. Julius H. Comroe, Jr., to act
as editor. He sent the manuscript out for reviews and
evaluation and made the final decision on its acceptability.
Received July 19, 1971. Accepted for publication
August 23, 1971.
Circulation Rtsurch, Vol. XXIX, Ouobtr 1971
reduced oxygen supply to skeletal muscle,
despite the lack of experimental verification
for this hypothesis The principal reasons for
entertaining this idea have been the potent
vasodilator property of this group of compounds, their presence in high concentration
in muscle, and their key role in the regulation
of energy metabolism.
Recently it has been reported that adenosine triphosphate (ATP) is released from
isolated contracting frog sartorius muscle (1)
and exercising human forearm muscle (2). In
addition, inorganic phosphate (Pi), a vasoactive product of nucleotide hydrolysis, has also
been postulated as a regulatory agent in blood
375
376
DOBSON, RUBIO, BERNE
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
flow to skeletal muscle, since levels of Pi have
been shown to increase in venous plasma from
contracting cat gastrocnemius muscle (3).
With respect to the heart, it has been
proposed that adenosine functions as a
mediator in the regulation of myocardial
blood flow (4), and the possibility exists that
adenosine may also serve a similar role in the
metabolic regulation of skeletal muscle blood
flow. Adenosine decreases skeletal muscle
vascular resistance when administered intraarterially (5, 6), and accumulates in intact
contracting and in isolated, ischemic, contracting rat muscle (7). Furthermore, venous
blood from heart and contracting skeletal
muscle produces dilation of resistance vessels
when infused into a forelimb artery but
produces constriction when infused into a
renal artery (5). The only endogenous compounds known to elicit these responses are
adenosine and adenylic acid (AMP).
In view of the current interest in nucleotides and their derivatives as possible mediators of metabolic vasodilation and the availability of improved methodology for detection
and quantification of adenine compounds, the
present study was undertaken to determine
the role of ATP, adenosine diphosphate
(ADP), AMP, adenosine, and P, in the
regulation of skeletal muscle blood flow.
Methods
FROG SARTORIUS NERVE-MUSCLE PREPARATION
Sartorius nerve-muscle preparations were carefully dissected from pithed summer or winter
frogs (Rana pipiens). Before dissection, the
hindlimbs were perfused via the dorsal aorta with
10 ml of amphibian Ringer's solution to flush the
blood from the muscle vasculature. The solution
had a composition (in ITIM) of 111.60 NaCl, 1.88
KC1, 1.08 CaCl2 • 2H 2 O, 2.38 NaHCOs> 0.09
NaH 2 PO 4 and 11.11 glucose. This solution was
bubbled with air and had a pH of 7.4. The
muscles were mounted at their resting length on
glass rods and immersed in 300 ml of Ringer's
solution for a 30-minute equilibration period at
25°C. After equilibration, the preparations were
placed in a chamber containing 3.0 ml of fresh,
aerated Ringer's solution. The sciatic nerve was
suspended in air just above the solution on a
platinum wire electrode, and the nerve was
stimulated at 2—5 Hz for 30 minutes. Other
preparations were allowed to rest for 30 minutes.
The bathing solutions were dien removed and
immediately analyzed for ATP and P|.
DOG HINDLIMB PREPARATION
Twelve mongrel dogs weighing 18—22 kg were
anesthetized with pentobarbital sodium (30
mg/kg, iv). The right leg was skinned to the
ankle, and the circulation below this point was
occluded by binding the ankle widi a tight
ligature. A stimulating electrode was placed on
the peripheral end of the sciatic nerve at die level
of the quadratus femoris muscle, and the leg was
wrapped with a plastic sheet to prevent dehydration during the course of the experiment. The
abdominal cavity was opened and the right
common iliac vein and right external iliac artery
were prepared for cannulation. After intravenous
administration of 20,000 units of heparin, both
vessels were severed and silicone rubber Tcannulas were inserted between die cut ends of
each. The side branches of the T-cannulas were
used for the collection of blood samples. Arterial
blood flow was continuously recorded with a 3.5
mm cannulating electromagnetic flow probe
(Biotronex). Adenosine, P,, and the sodium salts
of ATP, ADP, and AMP were freshly prepared in
0.9$ NaCl and were infused into the arterial
cannula at a constant rate between 1.4 and 8.8
ml/min. Systemic arterial pressure was continually recorded from the left common carotid artery
on a Honeywell Visicorder. Hindlimb blood flow
achieved a steady, low, control value 30—40
minutes after insertion of the cannulas. Control
arterial and venous blood samples were simultaneously collected after the initial equilibration
period. The hindlimb muscle mass was then made
ischemic for 5 minutes by occlusion of the arterial
cannula and was induced to contract by electrical
stimulation of the sciatic nerve with 5-10 v at
20-30 Hz during the ischemic period (Fig. 1).
Experimental arterial and venous blood samples
(40 ml) were simultaneously collected over a
period of 8—40 seconds immediately upon
termination of ischemic contraction. In some
experiments, sartorius muscle samples were
obtained by instantaneous freezing in situ by
compression of the muscle into "wafers" (thickness 1-2 mm) with a pair of clamps precooled in
liquid nitrogen. Six percent (w/v) dextran (mol
wt = 60,000-90,000) in mammalian saline containing (in HIM) 118.41 NaCl, 4.69 KC1, 2.52
CaCl2 • 2H2O, 25.00 NaHCO 8 , 1.18 MgSO4 •
7H 2 O and 1.18 KH2PO4 was infused intravenously
over a 30-minute period to replace the blood withdrawn (approximately 160 ml). Following a second equilibration period, the control and experimental ischemic contractions were repeated and
additional blood and muscle samples were obtained.
CircmUtion Rtstrrcb,
Vol. XXIX, October 1971
377
ADENOSINE AND MUSCLE BLOOD FLOW
300
V«*WV^
ISO
£
o
t
V
150
a
o
o
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
50
10
z
10
1J
30
35
MINUTES
FIGURE 1
A typical pressure-flow recording from a dog hindlimb skeletal muscle preparation. Between
0 and 5 minutes the hindlimb muscle was stimulated to contract at 20—30 Hz while the
arterial cannula was occluded. The brief decline in the flow tracing on release of the occlusion
(5 minutes) was due to blood sampling procedures (see text). Note that arterial pressure (AP)
was essentially constant during the course of the experiment.
Five minutes of ischemic muscle contraction
constitutes a severe challenge. However, this
procedure is not completely unphysiological, since
25-30 minutes after stimulation, hindlimb blood
flow returned to control levels and the vascular
bed showed normal reactive hyperemic responses
to occlusions of the arterial blood supply for 5—10
seconds. This restoration of resistance vessel tone
suggests that the experimental procedure had no
lasting deleterious effect on the muscle vasculature, as illustrated in Figure 1.
ANALYTICAL PROCEDURES
Blood samples were collected in polyethylene
centrifuge tubes immersed in ice and containing
an equal volume of cold mammalian saline. Each
sample was rapidly transferred to a refrigerated
centrifuge and the bulk of die cellular elements
was gently separated from the plasma by a force
of 800-1,000 g for 10 minutes at 0°C. The dilute
plasma was decanted and ccntrifugcd a second
time at 25,000 g for 10 minutes to remove the
balance of the formed elements. It was then
decanted and small aliquots (1-3 ml) were
assayed immediately for ATP and Fh while the
remainder of the dilute plasma was subjected to
ultrafiltration as previously described (8). ComCircttUtum Ruurcb, Vol. XXIX, Octobtr 1971
plete ultrafiltration required 20-24 hours at 0°C
and removed proteins with molecular weight
greater than approximately 30,000. The ultrafiltrates were stored at —20°C.
Tissue samples were stored (1-5 days) under
liquid nitrogen prior to initiation of analytical
procedures. The frozen muscle "wafers" were
finely powdered in a hollow stainless steel
cylinder fitted with a stainless steel piston, both
precooled in liquid nitrogen. The pulverized
muscles were then weighed and transferred to
glass homogenizer tubes packed in dry ice. Based
on tissue weight, 10 volumes of ice-cold 0.5N
perchloric acid were rapidly added, and the
mixture immediately homogenized for 90 seconds
at 0°C with a blade homogenizer (Polytron) at a
speed of 4,000-8,000 rpm. The homogenate was
centrifuged at 7,000 g at 0°C for 10 minutes. The
resulting supernatant fluid was filtered through
fiber glass to remove insoluble lipids and then
neutralized at 0°C to pll 7 with KOH. The
neutralized extract was centrifuged, the precipitate (KC1O4) washed with 2-5 ml of distilled
H 2 O, and the supernatant fluids combined for
storage at -20°C.
Activated charcoal was added to the ultrafiltrates of dilute plasma (2-5 mg/ml of original
378
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
blood sample) and to the neutralized acid
extracts of muscle (50-100 mg/g wet weight) for
the adsorption of purine derivatives. Each
charcoal suspension was shaken for 3 hours at
25°C, separated by centrifugation (1,500 g for 10
minutes) and the supernatant fluid discarded.
The charcoal was washed twice with 10 ml of
water and centrifuged and the wash discarded.
The purine derivatives were eluted from the
charcoal by shaking vigorously for 90 minutes
with 30 ml of 10$ (v/v) pyridine in 50$ (v/v)
ethanol. The suspension was layered on an acidwashed celite column (3.0 g of celite in a tube
2.25 cm in diameter) and the eluate slowly
filtered. Upon completion of filtration, an additional 15 ml of the elution fluid was placed on the
column. Eluates were combined and dried with a
stream of filtered air at 50°C, and the remaining
solids were redissolved in 0.6 ml of H 2 O and
either assayed for ADP and AMP or chromatographed for isolation of adenosine, inosine, and
hypoxanthine.
The chromatographic procedures have been
described previously (8). In brief, adenosine,
inosine, and hypoxanthine were separated on
thin-layer (0.5 mm) ion-exchange cellulose
(Selectacel no. 69 phosphate) plates. The
ascending chromatograms were developed with
15$ (v/v) aqueous ethanol in closed tanks at 4°C
for 4 hours. Ultraviolet fluorescent spots, which
represented the isolated compounds, were eluted
from the cellulose by washing four times with 5
ml of 50$ (v/v) ethanol adjusted to pH 10 with
NH 4 OH. Each cellulose suspension was stirred for
20 minutes and the supernatant fluid collected by
centrifugation at 1,500 g for 10 minutes.
Supernatant fluids were combined and air dried,
and the residue redissolved in 1.5 ml H 2 O. The
redissolved precipitate was centrifuged and used
for enzymatic quantification of adenosine, inosine,
and hypoxanthine, The percent recoveries for
known quantities of adenosine, hypoxanthine, and
inosine added to muscle extracts and plasma were
81.2 ± 2 . 1 , 85.6 ± 1 . 8 , 84.2 ± 2 . 2 and 71 ± 1.9,
74.2 ± 2.5, 73.5 ± 2.4, respectively.
The two nucleosides and hypoxanthine were
measured enzymatically by conversion to uric
acid, after the spectropnotometric enzymatic
methods of Kalckar (9, 10). Assays were
performed using 0.2-0.5 ml of the chromatographically separated compounds in a 1.0-ml
cuvette. The appropriate buffer and enzymes
were added and absorbance changes recorded
with a Hitachi dual-wavelength spectrophotometer (Perkin-Elmer model 356), which permits
measurement of 10~10 moles of adenosine,
hypoxanthine, or inosine.
ATP was measured fluorometrically by the
firefly method (11) in an Aminco-Bowman
DOBSON, RUBIO, BERNE
spectrophotofluorometer at room temperature.
The sartorius muscle bathing solution and 1-ml
aliquots of the diluted dog plasma were rapidly
injected into a reaction cuvette previously placed
in the spectrophotometer. The cuvette contained
1.0 ml H 2 O and 0.5 ml of a reconstituted extract
of firefly lanterns (Sigma, FLE-50). The peak
intensity of the initial light flash emitted at 550
nm was recorded on an X-Y recorder. With this
assay, ATP concentrations as low as 1 X 10~®M
and 5 X IO-^M could be detected in the bathing
solutions and plasma samples, respectively. All
assays were performed in duplicate. The amount
of ATP present in each sample was determined
by comparison with flash intensities obtained from
standard ATP solutions. Known amounts of ATP
were added to the samples, and die assay
repeated to determine if the sample itself
inhibited the enzymatic reactions, thereby decreasing the initial intensity of the light flash.
Plasma levels of ADP and AMP were independently measured by selective enzymatic reduction
of NAD+ after the method of Adam (12).
Concentrations of ADP and AMP in water or
saline as low as 1 X I O ^ M could be detected and
such concentrations produced absorbance changes
of 0.0006 and 0.0012, respectively. In plasma
samples, the sensitivity was reduced, but charcoal
adsorption of ADP and AMP from the dilute
plasma increased the sensitivity of the assay and
permitted detection of ADP and AMP in plasma
at concentrations as low as 4.8 and 4.6 X IO-^M,
respectively. The percent recoveries of known
amounts of ADP and AMP using this procedure
were 74.3 ± 1.5 and 78.4±1.3, respectively.
Inorganic phosphate levels were determined on
0.5-1.5 aliquots of muscle bathing solutions and
diluted plasma samples by selective and quantitative precipitation of Pj with a suspension of
Ca(OH) 2 in Cad^, according to the method of
Seraydarian et al. (13). The precipitates were
dissolved
in
2.0
ml of
O.25N
HC1, and
P,
determined colorimetrically with the acid molybdate method of Fiske and Subbarow (14).
Results
FROG SARTORIUS NERVE-MUSCLE PREPARATION
ATP was not detectable in the bathing
solution after a 30-minute exposure to either
resting or contracting muscle. If the nucleotide was present in the bathing solution,
it was present in concentrations less than
1.0 X 10~*M, the lower limit of sensitivity of
our assay system. The concentrations of Pt in
the bathing solutions from both resting and
contracting muscle showed increases of apCircuUiion Rtstarcb, Vol. XXIX, October 1971
379
ADENOSINE AND MUSCLE BLOOD FLOW
TABLE 1
Freshly prepared ATP, ADP, and AMP
and P, (5 X 10" 3 M) in normal
mammalian saline were infused intra-arterially
at different rates for a period of 1 minute. The
resulting arterial plasma concentration was
calculated for each experiment on the basis of
the hindlimb blood flow at the onset of
perfusion and the hematocrit ratio (range
0.43-0.45). The plasma levels of the adenine
nucleotides and Pi necessary to elicit a
detectable vasodilation are shown in Table 2.
A detectable vasodilation in the hindlimb
preparation is defined in this study as a 1035
increase in blood flow.
The concentrations of ATP, ADP, and AMP
in arterial plasma required to elicit a detectable vasodilation ranged from 27.1 to
31.4 X 10^M. These concentrations are approximately sixfold greater than the lowest
measurable plasma level for each nucleotide.
The arterial plasma concentration of Pi
required to produce detectable vasodilation
was 7.2 X ICHM, a level that was twice as
great as that found in the plasma samples.
Adenosine in Muscle.—The adenosine concentration in control dog sartorius muscle was
0.7 nmole/g (wet weight) and increased to
1.5 nmole/g in ischemic contracting muscle
(Fig. 2). The concentration of hypoxanthine
and inosine in control sartorius muscle was 4.5
and 2.0 nmole/g, respectively. With ischemic
muscle contraction the hypoxanthine and
inosine levels increased to 8.5 and 5.5
nmole/g, respectively (Fig. 3).
Adenosine in Plasma.—Adenosine concentrations in control arterial and venous plasma,
as well as in arterial plasma obtained after a
Concentration of Pt in Frog Sartorius Bathing Solution after the Thirty-Minute Experimental Period
Increase In Pi
concentration
above that of
bathing solution*
X 10-hi
Period
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Resting
Contracting
Resting
Contracting
Resting
Contracting
Contracting
Contracting
Contracting
Contracting
Contracting
Contracting
(IO-SM),
20
21
19
18
19
17
19
20
19
18
20
19
(2 Hz)
(2 Hz)
(2
(2
(2
(2
(2
(5
(5
Hz)
Hz)
Hz)
Hz)
Hz)
Hz)
Hz)
ATP was not detectable in any sample. The lowest
measurable concentration in this study was 1.0 X
IO-SM. *P, concentration of freshly prepared bathing
solution was 9 X
proximately twice the initial concentration of
fresh bathing solution (Table 1).
DOG HINDLIMB PREPARATION
Nucleotides and Inorganic Phosphate.—
ATP, ADP, and AMP were not detected in
control arterial and venous plasma or in
arterial and venous plasma from blood
collected immediately on termination of the
period of ischemic contraction. Arterial and
venous plasma levels of P| following ischemic
contraction, as well as during control periods,
were 3.7 X 1(HM. Table 2 shows the concentrations of ATP, ADP, AMP, and Pi in arterial
and venous plasma of dog hindlimb skeletal
muscle.
TABU 2
Concentration of ATP, ADP, AMP and Pi m Arterial and Venous Plasma of Skeletal Muscle of
the Dog Hindlimb
Compound
Wo.
Expbi
ATP
ADP
AMP
P.
7
6
6
5
Determined arterial and renous
plasma levels at end of ischemlc
contraction
<5.0
<4.8
<4.6
3.7
X 10-'M*
X 10-«M*
X 10-»M*
X 10-*M
Arttnial plpamp Invela
neceasar; to elicit a
detectable vasodllationt
28.7
27.1
31.4
7.2
*Loweat detectable plasma levels of these nucleotides. fValues represent mean
CircuUiion Reswcb, Vol. XXIX, October 1971
±
±
±
±
4.2
5.4
5.1
1.8
X 10-»M
X lO"8!!
X 10" 8 M
X 10-*M
1 8B.
380
DOBSON, RUBIO, BERNE
rates that produced an arterial plasma concentration of 2.8 ± 0.6 X 1(HM (approximately
the level found in venous blood after ischemic
hindlimb contraction) elicited a 25 ±5% increase in hindlimb blood flow. However,
increasing the arterial plasma adenosine concentration to 33.2 ± 2.6 X 10- 7 M, a level equal
to approximately one-half the sum of the
concentrations of adenosine, hypoxanthine,
and inosine found in venous blood after
ischemic contraction, resulted in a 300 ± 20%
increase in hindlimb blood flow. This degree
of hyperemia is similar in magnitude to that
observed following ischemic muscle contraction of the hindlimb in the present experiments.
1.6
g>
1.4
%
1-2
"OJ
1.0
_
o
E
0.8
0.6
rli
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
0.4
LU
Discusiion
Q
FROG SARTORIUS MUSCLE
0.2
CONTROL
ISCHEMIC
CONTRACTING
FIGURE 2
Adenosine concentrations in sartorius muscle from
the dog hindlimb skeletal muscle preparation. Control
muscles (N = 8,) were obtained from resting, normally
perfused preparations. Ischemic contracting muscles
(N = 8) were obtained from preparations stimulated to
contract (20—30 Hz) for 5 minutes while arterial inflow
was occluded. Vertical bars represent 1 SEM.
period of ischemic hindlimb contraction,
ranged from 0.3 to 0.5 X 10~7M (Fig. 4).
However, venous plasma levels of adenosine
increased to 2.2 X 10~7M following ischemic
contraction (Fig. 4). The control arterial and
venous, as well as experimental arterial,
plasma levels of hypoxanthine, ranged from
1.3 to 2.1 X 10~7M, whereas the inosine levels
were lower and ranged from trace amounts to
0.1 X 10~7M (Fig. 5). After ischemic contraction, venous plasma levels of hypoxanthine
increased to 38.0 X 10~7M and inosine increased to 2 7 . 0 X 1 0 - 7 M (Fig. 5). Neither of
these degradation products of adenosine
possess any significant vasoactivity.
Adenosine Infusion.—Intra-arterial infusion
of adenosine into the hindlimb preparation at
The results of the experiments on the frog
sartorius nerve-muscle preparations indicate
that ATP is probably not released from
contracting skeletal muscle in vitro in amounts
required to significantly alter blood flow.
These findings are not in agreement with
those of Boyd and Forrester (1), who
reported ATP concentrations of 1.97—19.7 X
IO^M in bathing solutions (2 ml) from
contracting (2 Hz for 30 minutes) frog
sartorius muscle. Although the sensitivity of
the firefly assay system employed in the
•
HYPOXANTHINE
CO
o
CONTROL
ISCHEMIC
CONTRACTING
FIGURE 3
Hypoxanthine and inosine concentrations in sartorius
muscle from dog hindlimb skeletal muscle preparation.
See legend of Figure 2.
Circtdmtion Rtstarcb, Vol. XXIX, October
1971
381
ADENOSINE AND MUSCLE BLOOD FLOW
z
g
_
2.4 _
at
Z
_l_
2.0 -
a
z
0 5
1.6 -
| 2
1.2
o
Z
UJ
a
<
<
2
0.8
-
—
—
04
tn
a.
0
IIArterial
II
Ariarial
>
V»noui
i
Vanaut
1
HYPEREMIA
CONTROL
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
FIGURE 4
Plasma adenosine concentrations in blood collected
from 12 dog hindlimb skeletal muscle preparations.
Control arterial and venous levels represent plasma
concentrations from well-perfused, resting hindlimb
muscle; hyperemic levels represent arterial and venous
plasma concentrations immediately on termination of
5 minutes of ischemic contraction. Vertical bars represent 1 SEM for 24 determinations (2 per preparation).
present studies allowed measurement of ATP
concentrations as low as 1.0 X K H M , ATP was
never detected in sartorius muscle bathing
solutions from either resting or contracting
muscle. Our results are also at variance with
those of Abood et al. (15), who reported an
increase in the release of ATP, ADP, AMP,
creatine phosphate, and Pi with in vitro
contraction (0.33 Hz for 20 minutes) of frog
sartorius muscle.
The reason for the differences between our
results and those of Boyd and Forrester (1)
and Abood et al. (15) is not readily apparent.
It is possible that injury to frog sartorius
muscle cells during dissection can release
sufficient amounts of cell cytoplasm, which
contains 2 to 3 X 10" 8 M ATP (16), to elevate
the concentration of ATP in the extracellular
fluid and bathing solution. Furthermore, a
contracting, as opposed to a resting, muscle
could liberate ATP and other phosphates from
damaged cells by compression by adjacent,
intact contracting cells. In our studies, the
44.0
40.0
V 36.0
O
"" 32.0 -
1
HYPOXANTHINE
^M
INOSINE
X
1
28.0
2
2.4
LU
£
T
2.0
o
u
1.6
T
<
S
'-2
0.4
Arterial
Venous
CONTROL
Arttria!
V*noui
HYPEREMIA
FIGURE 5
Plasma hypoxanthine and inosine concentrations in blood collected from eight dog hindlimb
skeletal muscle preparations. See legend of Figure 4.
Circulation Rtittrcb, Vol. XXIX, Octobtr 1971
382
muscles were repeatedly rinsed with frog
Ringer's solution before the start of the
experiments to wash all cytoplasm from cells
that may have been damaged during dissection. In some experiments in which this
precaution was not taken, small concentrations of ATP were detectable.
DOG HINDLIMB PREPARATION
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Five minutes of ischemic muscle contraction
in the dog hindlimb skeletal muscle preparation constitutes a strong stimulus but was used
because such a maneuver should reveal
whether any of the adenine nucleotides or P!
are released from the muscle into the venous
blood. Since ATP, ADP, and AMP were not
detected in venous plasma from either resting
or ischemic contracting dog hindlimb skeletal
muscle, and since plasma concentrations of
these compounds necessary to elicit a detectable vasodilation were approximately six times
greater than the lowest measurable plasma
concentrations, we conclude that the adenine
nucleotides are not directly involved in
skeletal muscle blood flow regulation.
The plasma concentrations of the adenine
nucleotides required to elicit a detectable
vasodilation in our dog hindlimb preparation
were similar to those necessary to induce a
noticeable decrease in vascular resistance in
the intact dog forelimb (17). Forrester and
Lind (2) reported ATP concentrations in
venous plasma from resting human forearm
muscle 1.3—4.6 times greater than that required
to elicit a detectable vasodilation in the
hindlimb preparation. With sustained contraction of forearm muscle, the venous blood ATP
concentrations were 300 times greater. These
large plasma ATP concentrations from the
human forearm are not reconcilable with our
results or with the results of others (18, 19)
and may reflect contamination from damaged
cellular elements in the blood. Moreover,
Forrester and Lind (2) suggested that platelets probably contributed to their plasma ATP
concentrations. Although not reported in our
studies, ATP was occasionally found in plasma
after the first centrifugation (800-1,000 g) but
not after the second centrifugation, suggesting
that the plasma obtained after the initial low-
DOBSON, RUBIO, BERNE
speed centrifugation still contained formed
elements (probably erythrocytes and platelets)
that could have been responsible for the ATP
observed. In two samples, in which hemolysis
was apparent, ATP was also detected.
Control and experimental arterial and venous plasma levels of Pi of the dog hindlimb
skeletal muscle preparation were less than
3.7 X 10-*M. An arterial plasma concentration
of twice this value was necessary to elicit a
small arteriolar dilation. However, Pi was
reported by Hilton and Chir (3) to be
released from isolated, blood-perfused, contracting white-fibered skeletal muscle in the
cat, in proportion to the frequency of
contraction. Our hindlimb preparation contained both white- and red-fibered muscle,
which may account for the difference between
our results and those of Hilton and Chir (3).
Barcroft et al. (20) reported that phosphate
concentrations increased by 20% in blood from
contracting human forearm muscle in which
blood flow increased tenfold. However, when
blood phosphate concentrations were increased as much as fourfold above control by
infusion of a mixture of NaH2PO4 and
Na2HPO4, flow did not change. Although
Hilton and Chir (3) suggested that P, may be
involved in active hyperemia, organic phosphate in addition to Pi may have contributed
to the response, since they used the Fiske and
Subbarow (14) method for Pi determination.
This method used without prior selective
isolation of P(, as performed in the present
studies according to the methods of Seraydarian et al. (13), does not eliminate contamination with organic phosphate. Although our
results indicate that the adenine nucleotides
probably do not contribute to the increase in
plasma phosphate levels reported by these
investigators (3), it is possible that some
unidentified organic phosphate(s) with unknown vasoactive property may be responsible for the apparent increase in plasma Pi
levels.
Resting skeletal muscle of the dog contains
measurable amounts of adenosine that increase when oxygen demand exceeds oxygen
supply. These findings are in agreement with
Circulation Ri**rch, Vol. XXIX, October 1971
ADENOSINE AND MUSCLE BLOOD FLOW
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
previous studies on rat skeletal muscle (7).
The adenosine present in the muscle is
presumably restricted to the extracellular
fluid, as has been postulated for the myocardium (4), due to the ubiquitous nature of
adenosine deaminase. For example, at extracellular adenosine concentrations far in excess
of those existing in anoxic muscle, adenosine
was not detectable within red cell ghosts because of the high activity of adenosine deaminase (21). These ghost cells are capable of
incorporating extracellular adenosine into intracellular adenine nucleotides or deaminating
it to inosine, but free adenosine was found in
the ghost cells only at an extracellular
adenosine concentration of 1 mM. If indeed
adenosine is confined to the extracellular space
in skeletal muscle, and the extracellular fluid
volume constitutes 20% of the skeletal muscle
volume (22), the concentration of adenosine
in the interstitial fluid after ischemic contraction would be 7.5 nmole/ml. This calculated
interstitial fluid adenosine concentration (75
X ±0~ 7 M) is greater than twice the arterial
plasma adenosine concentration (approximately 33 X K H M ) required by intra-arterial
infusion to produce a degree of hyperemia
equal to that observed following ischemic
muscle contraction. This suggestion is based
on the assumption that since both interstitial
fluid and plasma are in communication with
the vascular smooth muscle of the resistance
vessels, equal concentrations of adenosine in
each of these compartments would have
comparable effects in attenuating vascular
smooth muscle tone.
Adenosine in venous plasma from the dog
hindlimb increased about fivefold with ischemic contraction, whereas the levels of
hypoxanthine and inosine increased 22- and
270-fold, respectively, above control levels.
Studies on the rate of adenosine degradation
in blood (8) indicate that about half of the
adenosine is deaminated in the time required
for the blood to pass from the capillaries to
the collecting tube. Adenosine present in
interstitial fluid is also converted to inosine
and hypoxanthine in its passage through the
CircuUtton Rutarcb, Vol. XXIX, October 1971
383
vessel wall, since endothelial cells and pericytes contain adenosine deaminase and nucleoside phosphorylase (23) (enzymes that
catalyze the conversion of adenosine to inosine
and hypoxanthine, respectively). Hence, the
concentration of adenosine at the level of the
resistance vessels is undoubtedly much greater
than that present in the blood collected from
the iliac vein, a conclusion supported by the
greater ratio of hypoxanthine and inosine to
adenosine in venous plasma as compared to
this ratio in the muscle.
Therefore, it appears that the concentration
of adenosine reached in the interstitial fluid
during ischemic contraction of the muscle is
more than sufficient to account for the
decrease in vascular resistance during the
period of reactive hyperemia. These observations represent the first demonstration of
adenosine in skeletal muscle effluents and
suggest that adenosine may play a role in the
local regulation of skeletal muscle blood flow
in a manner similar to that proposed for
cardiac muscle (4). Blood flow regulation by
adenosine in skeletal muscle would predict
that with an increase in oxygen consumption
or a reduction in the arterial oxygen supply,
cellular AMP levels would increase at the
expense of ATP and the 5'-nucleotidase which
in skeletal muscle is present in vessel walls
and in muscle tissue adjacent to blood vessels
(24), would dephosphorylate AMP to adenosine at a faster rate. Consequently, the
interstitial fluid concentration of adenosine
would increase, which in turn would cause
skeletal muscle resistance vessels to dilate and
muscle blood flow to increase to a level that
would maintain muscle oxygen balance. Other
vasoactive factors such as H + , K+ reduced P(>>,
or increased osmolarity probably participate
in the reactive hyperemia following ischemic
contraction, but the extent of their participation, as well as that of adenosine remains to be
elucidated. However, it appears unlikely that
adenine nucleotides and inorganic phosphate
are involved in the local regulation of skeletal
muscle blood flow in a manner similar to that
postulated for adenosine.
384
DOBSON, RUBIO, BERNE
sine 5'-monophosphate. In Methods of Enzymatic Analysis, edited by H. Bergmeyer. New
York, Academic Press, 1965, pp 573-577.
References
1. BOYD,
LA.,
AND FORRESTER,
T.:
adenosine triphosphate from
muscle in vitro. J Physiol
1968.
2.
Release
of
frog skeletal
199:115-135,
HILTON, S.M., AND CHIR, B.: New candidate for
mediator of functional vasodilation in skeletal
muscle. Circ Res 28 (suppl I ) : 1-70-72,
1971.
4.
RUBIO,
R.,
AND BERNE,
R.M.:
Release
of
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
adenosine by the normal myocardium in dogs
and its relationship to the regulation of
coronary resistance. Circ Res 25:407-415,
1969.
5.
SCOTT, J.B., DAUGHERTY, R.M., JR., DABNEY, J.M.,
AND HADDY, F.J.: Role of chemical factors in
regulation of flow through kidney, hindlimb,
and heart. Am J Physiol 208:813-824, 1965.
6.
HASHIMOTO, K., AND KUMAKURA, S.: Pharmaco-
14.
CURNISH, R.R.: Adenosine and adenine nucleotides as possible mediators of cardiac and
skeletal muscle blood flow regulation. Circ Res
28 (suppl. I ) : 1-115-119, 1971.
8.
9.
10.
11.
12.
RUBIO,
R.,
BERNE,
R.M.,
AND KATORI,
15.
C.H.,
AND SUBBAROW,
Y.:
Colorimetric
ABOOD, L.G., KOKETSU, K., AND MIYAMOTO, S.:
Outflux of various phosphates during membrane depolarization of excitable tissues. Am J
Physiol 202:469-474, 1962.
16. CARLSON, F.D., AND SIGER, A.: Creatine phos-
phoryltransfer reaction in iodoacetate-poisoned
muscle. J Gen Physiol 43:301^313, 1959.
17. FROHLICH, E.D.: Local effect of adenosine
mono-, di-, and triphosphate on vessel resistance. Am J Physiol 204:28-30, 1963.
18.
BRASHEAR, R.E., Ross, J.C., AND SMITH,
C.N.:
Rest and exercise plasma nucleotides. Resp
Physiol 5:401-406, 1968.
19.
KONTOS, H.A., GOLDIN, D., RICHARDSON,
D.W.,
AND PATTERSON, J.L., JR.: Effect of dipyridamole on vasodilator responses to ischemia,
hypoxia, ATP, and AMP. Am J Physiol
214:107-114, 1968.
20.
BARCROFT,
H., FOLEY,
T.H.,
AND MCSWINEY,
R.R.: Experiments on the liberation of phosphate from the muscles of the human forearm
during vigorous exercise and on the action of
sodium phosphate on forearm muscle blood
vessels. J Physiol 213:411-420, 1971.
M.:
Release of adenosine in reactive hyperemia of
the dog heart. Am J Physiol 216:56-62,
1969.
KALCKAR, H.M.: Differential spectrophotometry
of purine compounds by means of specific
enzymes: I. Determination of hydroxypurine
compounds. J Biol Chem 167:429-443, 1947.
KALCKAR, H.M.: Differential spectrophotometry
of purine compounds by means of specific
enzymes: II. Determination of adenine compounds. J Biol Chem 167:445-459, 1947.
STREHT.FR, B.L.: Adenosine 5'-triphosphate and
creatine phosphate: Determination with luciferase. In Methods of Enzymatic Analysis, edited by H. Bergmeyer. New York, Academic
Press, 1965, pp 559-568.
ADAM, H.: Adenosine 5'-diphosphate and adeno-
FISKE,
determination of phosphorus. J Biol Chem
66:375-400, 1925.
logical features of the coronary, renal, mesenteric and femoral arteries. Jap J Physiol
15:540-551, 1965.
7. BERNE, R.M., RUHIO, R., DOHSON, J.G., JR , AND
SERAYDARIAN,
K.,
MOMMAERTS,
W.F.H.M.,
WALLNER, A., AND GUILLORY, R.J.: Estimation
of the true inorganic phosphate content of fiog
sartorius muscle. J Biol Chem 236:2071-2075,
1961.
FORRESTER, T., AND LLND, A.R.: Identification of
adenosine triphosphate in human plasma and
the concentration in the venous effluent of
forearm muscle before, during and after
sustained contractions. J Physiol 204:347-364,
1969.
3.
13.
21.
SCHRADER, J., AND BERNE, R.M.: Metabolism of
the coronary vasodilator adenosine by human
red cell ghosts (abstr.). Fed Proc 30:488,
1971.
22.
ZlERLER, K.L., ROGUS, E . , AND HAZELWOOD, C . F . :
Effect of insulin on potassium flux and water
and electrolyte content of muscles from normal
and from hypophysectomized rats. J Gen
Physiol 49:433-456, 1966.
23.
RUBIO, R., AND WIEDMEIER, V.T.:
Localization
of nucleoside phosphorylase and its role in the
determination of myocardial purine distribution
(abstr.). Fed Proc 30:488, 1971.
24.
DOBSON, J.G., JR-, RUBIO, R., AND BERNE, R.M.:
Adenine nucleotide degradation in heart and
skeletal muscle (abstr.). Physiologist 13:181,
1970.
Circulation Rtstarcb, Vol. XXIX, October 1971
Role of Adenine Nucleotides, Adenosine, and Inorganic Phosphate in the Regulation of
Skeletal Muscle Blood Flow
James G. Dobson, Jr., Rafael Rubio, Jr. and Robert M. Berne, Jr.
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Circ Res. 1971;29:375-366
doi: 10.1161/01.RES.29.4.375
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1971 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/29/4/375
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/