Am J Physiol Regulatory Integrative Comp Physiol
280: R1308–R1314, 2001.
Sites and ionic mechanisms of hypoxic vasoconstriction
in frog skin
GARY M. MALVIN AND BENJIMEN R. WALKER
Lovelace Respiratory Research Institute, Albuquerque 87108; and Department of Cell Biology
and Physiology, University of New Mexico School of Medicine, Albuquerque, New Mexico 87131
Received 24 July 2000; accepted in final form 11 December 2000
hypoxic pulmonary vasoconstriction; amphibia; cutaneous
respiration; cutaneous blood flow
O2 availability. HVC facilitates O2 uptake efficiency
and blood oxygenation by diverting blood away from
poorly oxygenated gas-exchange regions to better oxygenated ones.
In contrast to other respiratory organs, the mechanisms mediating HVC in mammalian lungs have been
well studied. An accepted hypothesis is that alveolar
hypoxia alters the redox potential in the vascular
smooth muscle cell. This decreases membrane K⫹ conductance, causing depolarization. Depolarization increases the open probability of L-type Ca2⫹ channels,
facilitating Ca2⫹ entry into the cell, which leads to
vascular smooth muscle contraction and vasoconstriction (19). Although both pulmonary arteries and veins
are vasoactive, HVC occurs primarily in the arterial
pulmonary circulation. This appears to be due to the
presence of L-type Ca2⫹ channels in arterial vascular
smooth muscle but not in the pulmonary venous circulation (18).
Amphibian skin is another important respiratory
organ that displays HVC (10, 11). In contrast to the
pulmonary microcirculation, the cutaneous microcirculation can be readily studied in vivo with little or no
surgical intervention (8). The purpose of this study was
to test the hypothesis that the mechanism mediating
HVC in amphibian skin is similar to that in mammalian lung.
METHODS
Animals
Albino Xenopus laevis (Xenopus I, Ann Arbor, MI) were
used because their unpigmented skin allows excellent visualization of the cutaneous microcirculation (10). They were
kept in aquaria at ⬇22°C at least 2 wk before experimental
use and fed crickets twice per week.
Experimental Setup
(e.g., vertebrate lungs, fish
gills, and amphibian skin) regulate local blood flow to
match local O2 availability at the respiratory surface
(3, 7, 11, 14). Hypoxic vasoconstriction (HVC), a blood
flow reduction in response to low respiratory surface
PO2, is the primary mechanism matching blood flow to
All experiments were performed at room temperature
(⬇22°C). Frogs were anesthetized with methohexital Na⫹
(Brevital-Na⫹, Eli Lilly; 40 mg/kg im). This barbiturate was
used because it rapidly produces anesthesia lasting ⬇1–2 h,
and barbiturates have no effect on pulmonary hypoxic vasoconstriction in mammals (1). After anesthesia was induced,
Address for reprint requests and other correspondence: B. R.
Walker, Dept. of Cell Biology and Physiology, Univ. of New Mexico
School of Medicine, 915 Camino de Salud, Albuquerque, NM 87131
(E-mail: [email protected]).
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
MANY RESPIRATORY ORGANS
R1308
0363-6119/01 $5.00 Copyright © 2001 the American Physiological Society
http://www.ajpregu.org
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.5 on June 15, 2017
Malvin, Gary M., and Benjimen R. Walker. Sites and
ionic mechanisms of hypoxic vasoconstriction in frog skin.
Am J Physiol Regulatory Integrative Comp Physiol 280:
R1308–R1314, 2001.— We tested the hypothesis that the
cellular mechanisms mediating hypoxic vasoconstriction
(HVC) in frog skin, an important vertebrate respiratory organ, are similar to those mediating HVC in the pulmonary
vasculature of mammals. An accepted hypothesis in the lung
is that alveolar hypoxia alters the redox potential in vascular
smooth muscle cells of arterial vessels. This decreases membrane K⫹ conductance, causing depolarization. Depolarization increases the open probability of L-type Ca2⫹ channels,
facilitating Ca2⫹ entry into the cell, which leads to vascular
smooth muscle contraction and vasoconstriction. We studied
the cutaneous microcirculation of the frog (Xenopus laevis)
web by enclosing the web in a transparent chamber that was
ventilated with different gas mixtures. Arteriolar and venular diameters were measured by video microscopy. Drugs
were applied topically or intravascularly. A dose-dependent
constriction to hypoxia occurred in arterioles but not venules,
although both vessel types constricted to similar degrees to
the thromboxane mimetic U-46619. The magnitude of HVC
was not associated with arteriolar size. Constriction of arterioles with 4-amino pyridine, a K⫹-channel antagonist, was
blocked by the L-type Ca2⫹-channel blocker nifedipine. Nifedipine also antagonized HVC and hypercapnic vasoconstriction. Bay K 8664, a drug that increases the open probability of L-type Ca2⫹ channels, augmented HVC. These data
support our hypothesis that the cellular mechanisms mediating HVC are similar in frog skin and mammalian lungs.
This similarity between amphibain and mammalian tissues
suggests that the mechanisms of HVC may have arisen
relatively early in vertebrate evolution. In addition, because
of its structural simplicity and easy accessibility, frog skin
may be a useful tissue for studying this general phenomenon
in vivo.
HYPOXIC VASOCONSTRICTION IN FROG SKIN
Microcirculatory Measurements
Vessels were chosen on the basis of visual clarity. Good
visual clarity was necessary to accurately measure vessel
diameters. This selection criterium may have biased our
results if there was any association between visual clarity
and responsiveness of the vessels. Vessel diameters were
determined from the videotapes. The external diameters of
arterioles and venules were measured from printed video
images by a naive technician. Measurements were made
every 30 s.
Drugs
All drugs were dissolved in DMSO (Sigma Chemical), then
diluted 1:1,000 in distilled water for topical application or in
frog Ringer solution for intravascular administration. The
Ringer solution contained (in mM): 76.1 NaCl, 30.7 NaHCO3,
1.4 NaH2PO4, 2.5 KCl, 3.1 MgSO4, 4.0 CaCl2, and 5.6 glucose. PCO2 of the Ringer solution was ⬇12 Torr. DMSO
diluted 1:1,000 had no effect on vascular dimensions. The
drugs were: nifedipine (Research Biochemicals International), 4-AP (Sigma Chemical), U-46619 (Cayman Chemicals),
and S(⫺)-Bay K 8644 (Research Biochemicals International).
The drug doses chosen for testing were based on doses used
in previous lung experiments (18).
Experimental Protocols
1. Site of hypoxic vasoconstriction. The effect of N2 was
tested on arterioles and venules by exposing the web to 10
min of air, then 10 min of N2, followed by 10 min of air. One
arterial and one venous vessel were tested per frog. Responding vessels, the arterioles, constricted during the N2 exposure
and relaxed to initial normoxic levels during the second air
period. The percent change in arteriolar diameter after 10
min of N2 exposure was plotted against the vessel diameter 1
min before N2 was administered to the web. In addition,
arteriolar data from 44 additional frogs used in other protocols that were treated with 10 min of air followed by 10 min
of N2 were plotted. A correlation between the N2 response
and initial vessel diameter would indicate a size dependence
of hypoxic responsiveness.
2. General vasoactivity of arterioles and venules. Only
arterioles responded to N2, suggesting that either hypoxia
selectively affects the arterioles or that frog web venules do
not have the ability to constrict. To assess these possibilities,
we applied U-46619, a thromboxane mimetic that constricts
both arterial and venous vessels in the lung (18). If U-46619
constricted both arterioles and venules in the web, as it does
in the lung, we could conclude that hypoxia specifically affects arterioles in the web. Three doses of U-46619 (0.01, 0.1,
and 1.0 ␮M) and of vehicle were applied topically (120 ␮l) to
the web under normoxic conditions. Ten minutes after application of U-46619 or vehicle, vessel diameters were measured. Different animals were used for each dose tested.
Adjacent arterioles and venules were measured in each animal.
3. Dose-response relationship of arterioles to O2. We then
tested whether the response to O2 was dose dependent. The
web chamber was initially ventilated with air (21% O2) for 10
min. Then the web was exposed to four different O2 concentrations for 10 min each in a random order. The vessel
diameters at the end of 10-min periods were analyzed and
compared with the diameters measured at the end of the
initial normoxic period. The O2 concentrations tested after
the initial air exposure were 0, 8, 21, and 100%.
4. Effect of time on HVC. The effect of time on HVC was
tested by exposing the web to consecutive 10-min periods of
air and N2. The response to N2 was tested three times in
succession. Values obtained at the end of the air periods were
compared with the values obtained at the end of the subsequent hypoxic period. In another experiment, we characterized the response of arteriolar diameter to 40 min of continuous N2 to assess the duration of HVC in frog skin.
5. Effects of 4-AP and 4-AP ⫹ nifedipine on HVC. 4-AP
blocks voltage-sensitive K⫹ channels, and nifedipine inhibits
the activity of L-type Ca2⫹ channels. Because topical application of 4-AP caused slight muscle twitching in the foot,
drugs in this experiment were applied intravascularly, as
described above. During air exposure, arteriolar diameter
was measured for 10 min, then vehicle (frog Ringer solution
with 0.1% DMSO) was infused (0.1 ml over 1 min). Ten
minutes later, either 4-AP (10 ␮M), 4-AP (10 ␮M) ⫹ nifedipine (10 ␮M), or vehicle was infused in exactly the same
manner. Vessel diameters were measured every 30 s. Diameter values obtained just before infusion were compared with
the largest change in diameter observed after infusion. Different animals were used for the three different treatments.
Vasoconstriction by 4-AP and reversal of 4-AP-induced constriction by nifedipine would support our hypothesis that a
reduction in membrane potassium conductance leads to cell
depolarization, subsequent increased membrane conductance of Ca2⫹ through L-type Ca2⫹ channels, and vascular
constriction.
6. Effects of nifedipine on HVC. The effects of nifedipine on
HVC were tested by exposing the web to 10-min periods of
air, N2, then air. Nifedipine (120 ␮l) was then applied to the
skin surface. Ten minutes later, the web was exposed to N2
for 10 min. Each concentration of nifedipine (10 ␮M and 100
␮M) was tested in a separate group of animals. Values 1 min
before nifedipine or vehicle application and values at the end
of the 10-min hypoxic period were used in the statistical
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.5 on June 15, 2017
the frog was placed supine on a wire mesh, and the foot was
positioned in a round plastic chamber (1.4 cm high; 8.5 cm
radius) constructed from a Petri dish. The foot web was
spread loosely by pinning the distal ends of the toes on each
side of the foot to modeling clay within the chamber. Moist
gauze was placed between the most posterior part of the leg.
This created a seal that allowed the web to be exposed to
different chamber gases while the rest of the aminal was
exposed to the atmosphere. The chamber had gas inflow and
outflow ports and a hole in the top for topical administration
of drug solutions. The hole could be covered by a glass
coverslip. The gas ports allowed the chamber to be ventilated
with different gas mixtures. Chamber ventilation was 1 liter/
min causing a change in chamber gas composition of ⬎99% in
1 min when the inflow gas was changed. The web chamber
was placed beneath a microscope (Nikon Optiphot) and transilluminated. Magnified images of the web microcirculation
were recorded by video microscopy (video camera, Dage-MTI,
model CCD-72; video monitor, Hitachi, VM-1220U; and video
recorder, Panasonic, AG-1950). Magnification from skin to
monitor was ⫻670 with a ⫻20 objective.
In experiments testing the effects of 4-amino pyridine
(4-AP) on vessel diameter, its vehicle (frog Ringer solution
with 0.1% DMSO) and a combination of 4-AP and nifedipine
were directly infused into the web circulation. Intravascular
infusion was necessary because topical application of 4-AP
caused slight muscle twitching in the foot, making microvascular measurements impossible. In these experiments, a
small arterial vessel in the distal gastrocnemius muscle was
cannulated occlusively with tapered PE-10 tubing. Infusate
then traveled into the main artery perfusing the foot web.
R1309
R1310
HYPOXIC VASOCONSTRICTION IN FROG SKIN
Fig. 1. Effect of 100% N2 on arteriolar and venular diameter. Values
are means ⫾ SE. The number of animals tested is n. *Significantly
different from 0 (P ⬍ 0.001; paired t-test). There was no significant
effect of N2 on venular diameter (P ⫽ 0.095; paired t-test).
Fig. 2. Relationship between normoxic arteriolar diameter and %reduction in arteriolar diameter after 10 min of N2 exposure. Each
point represents a datum from 1 animal. The line is the least-squares
regression line. There was no significant correlation between the 2
variables (r ⫽ 0.185; P ⫽ 0.153; n ⫽ 60).
RESULTS
1. Site of HVC. During the initial air periods, the
arterioles and venules had mean (⫾SD) diameters of
44 ⫾ 12 and 51 ⫾ 13 ␮m, respectively. Exposure of the
foot web to 100% N2 constricted arterioles by 42 ⫾ 5%
(mean ⫾ SE; P ⬍ 0.001) but had no significant effect on
venules (P ⫽ 0.095; Fig. 1). Among arterioles, there
was no correlation between initial vessel size and percent change in vessel diameter during N2 exposure (r ⫽
0.185; P ⫽ 0.153; Fig. 2).
2. General vasoactivity of arterioles and venules.
U-46619 constricted both arterioles (P ⬍ 0.05) and
venules (P ⬍ 0.05) in a dose-dependent manner (Fig.
3). There was no significant difference in the percent
change in vessel diameter between arterioles and
venules (P ⬎ 0.17).
Fig. 3. Effect of U-46619 on the diameters of arterioles and venules.
Values are means ⫾ SE. *Significantly different from vehicle values
(P ⬍ 0.05; 2-way ANOVA and Dunnett’s test). There was no significant difference in the %change in vessel diameter between arterioles
and venules (P ⬎ 0.05; 2-way ANOVA).
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.5 on June 15, 2017
analysis. In preliminary experiments, longer normoxic nifedipine periods had no significant effect on the response to the
subsequent hypoxic period.
7. Effects of Bay K 8644 on arterioles and venules during
normoxia and hypoxia. Bay K 8644 increases the open probability of L-type Ca2⫹ channels. Bay K 8644 (50 ␮M; 120 ␮l)
or vehicle (120 ␮l) was applied topically to the web under
normoxic conditions. Ten minutes after Bay K 8644 or vehicle
application, the web was exposed to air, 8% O2, and 0% O2.
The order of gas exposures was randomized. Diameters of
both an arteriole and venule were measured simultaneously
in each animal tested with Bay K 8664 and vehicle. Values 1
min before Bay K 8664 or vehicle application and at the end
of each succeeding 10-min period were analyzed. Enhancement of HVC by Bay K 8644 would support the involvement
of L-type Ca2⫹ channels in HVC. Venules were studied to test
the possibility that venule unresponsiveness to hypoxia is
due to L-type Ca2⫹ channels that have an unusually low open
probability at hypoxic membrane potentials or that hypoxia
has a relatively small effect on smooth muscle membrane
potential. Either of these possibilities would be supported if
venular HVC occurred in the presence of Bay K 8644.
8. Effects of nifedipine on hypercapnia- and hypercapnia
plus hypoxia-induced arteriolar constriction. Hypercapnia at
the gas-exchange surface reduces blood flow both in mammalian lungs (2) and in frog skin (10). To determine whether
hypercapnic vasoconstriction may involve a mechanism similar to that for HVC, the effects of nifedipine on local hypercapnia (5% CO2–95% air) and on a combination of hypercapnia and hypoxia (5% CO2–95% N2) were investigated as
described in the protocol testing the effects of nifedipine on
HVC. Attenuation of the hypercapnic and hypercapnic plus
hypoxic responses by nifedipine would indicate that L-type
Ca2⫹ channels are involved in HVC as well as in hypercapnic
vasoconstriction.
Statistical analysis. Because vessels of different sizes were
studied, diameter changes were expressed as a percent
change from control-period diameters. To assess the significance of these values, an arcsin square-root transformation
was performed to distribute the data normally before statistical testing. Paired t-tests, unpaired t-tests, and 1- and
2-way repeated-measures ANOVA were used where appropriate. Multiple comparisons of sample means were performed with either a Dunnett’s test or Neuman-Keuls test.
The specific test used is indicated in RESULTS or figure legends. Statistical significance was set at a level of P ⫽ 0.05.
HYPOXIC VASOCONSTRICTION IN FROG SKIN
3. Dose-response relationship of arterioles to O2. Hypoxic vasoconstriction of web arterioles was dose dependent (Fig. 4). Hyperoxia had no significant effect on
arteriolar diameter (P ⬎ 0.54).
4. Effect of time on HVC. Time had no significant
effect on HVC during three consecutive 10-min periods
of air and N2 exposure (P ⬎ 0.357; Fig. 5). HVC is not
a brief, transient response because time had little effect on HVC during a continuous 40-min N2 exposure
(Fig. 6). The data in Fig. 6 also suggest that the response to N2 is biphasic (larger during the first several
minutes than at later times). We assessed that possibility by comparing arteriolar diameters after 2 min
(the time when constriction appeared greatest) and
40 min. Although there was a significant difference
(P ⫽ 0.048; ANOVA), no significant difference existed between the 4- and 40-min values (P ⫽ 0.14;
ANOVA).
5. Effects of 4-AP and 4-AP plus nifedipine on HVC.
4-AP infused into the web circulation constricted web
arterioles (Fig. 7). Coinfusion of nifedipine abolished
the response to 4-AP (P ⫽ 0.50; Fig. 7). The vehicle had
no effect on vessel diameters (P ⬎ 0.26; paired t-test).
In two experiments, intravascular infusion of nifedi-
Fig. 5. Effect of repeated exposures to N2 on arteriolar diameter.
Each bar represents the mean ⫾ SE of arteriolar diameter after 10
min exposure to N2. Ten minutes of air exposure separated the
hypoxic periods. Time had no significant effect on hypoxic vasoconstriction (HVC) during 3 consecutive periods of air and N2 exposure
(P ⫽ 0.357; repeated-measures ANOVA).
Fig. 6. Effect of 40 min of 100% N2 on arteriolar diameter. N2 was
initiated at time 0. Values are means ⫾ SE.
pine alone had no apparent effect on vessel diameter
(data not shown).
6. Effects of nifedipine on HVC. Nifedipine significantly attenuated the arteriolar response to 100% N2
(P ⬍ 0.012; Fig. 8). The magnitude of HVC inhibition
by 10 ␮M nifedipine was not significantly different
from inhibition by 100 ␮M nifedipine (P ⫽ 0.18).
7. Effects of Bay K 8644 on arterioles and venules
during normoxia and hypoxia. Bay K 8644 (50 ␮M) had
no significant effect on arteriolar diameter during air
(P ⬎ 0.50; data not shown) or 8% O2 exposure (P ⬎ 0.8)
but potentiated the constriction to 100% N2 (P ⬍ 0.001;
Fig. 9). Bay K 8644 had no significant effect on venular
diameters during air, 8% O2, or 100% N2 exposure (P ⬎
0.33; Fig. 10).
8. Effects of nifedipine on hypercapnia- and hypercapnia plus hypoxia-induced arteriolar constriction.
Exposure of the foot web to N2, 5% CO2/95% air, or to
5% CO2/95% N2 constricted arterioles (P ⬍ 0.05). There
was no significant difference in the degree of constriction among the three gas treatments (P ⬎ 0.05; Fig.
11). Nifedipine significantly attenuated the arteriolar
Fig. 7. Effects of 4-aminopyridine (4-AP), 4-AP ⫹ nifedipine (Nif),
and vehicle (V) on arteriolar diameter. Values are means ⫾ SE.
*Significant difference between V and 4-AP (P ⫽ 0.022; paired t-test).
There was no significant difference between V and 4-AP ⫹ Nif (P ⫽
0.50; paired t-test).
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.5 on June 15, 2017
Fig. 4. Effect of O2 concentration ([O2]) on arteriolar diameter.
Values are means ⫾ SE. *Significantly different from the normoxic
value (P ⬍ 0.05; repeated-measures ANOVA and Dunnett’s test).
R1311
R1312
HYPOXIC VASOCONSTRICTION IN FROG SKIN
responses to the different gas mixtures (P ⬍ 0.05).
During nifedipine treatment, there was no significant
effect of gas treatment (P ⬎ 0.05).
DISCUSSION
Hypoxic vasoconstriction occurs in a variety of species and respiratory organs. The present study illustrates that this response in amphibian skin exhibits
numerous similarities to the more investigated response in the mammalian lung. For example, the sites
of vasoconstriction in the skin are restricted to the
arterial circulation, which is similar to hypoxic pulmonary vasoconstriction (5). Unlike the rabbit lung (5),
however, we observed no difference in hypoxic responsiveness between arterioles of different sizes, although
the range in vessel size studied was only between ⬃20
and 75 ␮m initial diameter. The lack of venous responsiveness to hypoxia is not due to a general inability of
Fig. 9. Effects of 50 ␮M Bay K 8644 on arterioles during exposure to
8% and 0% O2. Values are means ⫾ SE. *Bay K 8644 values are
significantly different from corresponding vehicle values (P ⫽ 0.001;
paired t-test).
Fig. 10. Effects of 50 ␮M Bay K 8644 on venules during exposure to
8% and 0% O2. Values are means ⫾ SE. Bay K 8644 had no
significant effect on venular diameters (P ⬎ 0.33; paired t-tests).
veins to constrict, because arterioles and venules both
responded similarly to the vasoconstrictor thromboxane mimetic U-46619. Similar observations have been
made in isolated, perfused rat lungs (18). In addition,
hypoxic vasoconstriction in this preparation appears to
be PO2 dependent, because N2 elicited greater vasoconstriction than did 8% O2. Thus the segmental profile of
hypoxic vasoconstriction and its PO2 dependence are
similar between amphibian skin and mammalian
lungs.
In addition to similarities in the sites of the response
to hypoxia, data from the frog web strongly suggest
that a common mechanism of hypoxic vasoconstriction
exists between this bed and the mammalian lung. In
the mammalian pulmonary circulation, it has been
proposed that hypoxic vasoconstriction is a contractile
response of vascular smooth muscle cells to increased
Fig. 11. Effects of Nif on hypoxia-, hypercapnia-, and hypercapnia ⫹
hypoxia-induced arteriolar constriction. Values are means ⫾ SE.
*Nif significantly changed the response to gas treatments (P ⬍ 0.05;
2-way repeated-measures ANOVA). There was no significant difference between gas treatments before or after Nif (P ⬎ 0.05; 2-way
repeated-measures ANOVA).
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.5 on June 15, 2017
Fig. 8. Effects of Nif and V on HVC. Values are means ⫾ SE. Both
doses of Nif significantly attenuated the arteriolar response to 100%
N2 (P ⬍ 0.024; paired t-test). The %reduction in diameter by 10 mM
Nif was not significantly different from the %reduction in diameter
by 100 mM Nif (P ⫽ 0.36; t-test). *Difference from N2 ⫹ V.
HYPOXIC VASOCONSTRICTION IN FROG SKIN
lung involves potassium channels and alterations in
membrane potential. Interestingly, hypercapnia-induced arteriolar constriction has also been observed in
the amphibian lung (6); however, the mechanism of
this response has not been studied.
In contrast to what is found in the pulmonary circulation of mammals (3), hypercapnia did not potentiate
the hypoxic response of frog skin arterioles. The reason
for this difference is not known. However, it is possible
that both 100% N2 and 5% CO2 alone produce a maximal vasoconstriction in the skin so that the combination of these gases cannot reduce vessel diameter further.
We speculate that the mechanism mediating HVC
arose early in vertebrate evolution because the hypoxic
vascular responses in amphibian skin and mammalian
lung appear to share mechanistic similarities. Additional studies on other species would be needed to
assess further this possibility.
The response of the skin to environmental PO2 will
act to match cutaneous blood flow to O2 availability.
This is probably beneficial for many amphibians because there are large PO2 variations in bodies of fresh
water. Surface water PO2 can vary between near 0 Torr
at night to several hundred Torr during the day,
whereas lake bottoms are often anoxic for periods of
weeks to months. Reducing cutaneous blood flow in
severe aquatic hypoxia or anoxia will minimize O2 loss
from the blood to the environment. Increasing skin
perfusion in hyperoxic water will augment cutaneous
O2 uptake. Local responses to PO2 will also distribute
blood flow within the skin to match regional PO2. Local
variations in skin surface PO2 can exist when animals
are partially submerged or when the ventral surface
contacts the ground.
We acknowledge the excellent technical assistance of Michelle
Duran, Lisa Roberts, and Yvonne Villalobos.
This research was supported by National Heart, Lung, and Blood
Institute Grants HL-38942 (G. M. Malvin), HL-07758 (G. M. Malvin)
and HL-58124 (B. R. Walker).
REFERENCES
1. Eisenkraft JB. Effects of anesthetics on the pulmonary circulation. Br J Anaesth 65: 63–78, 1990.
2. Gao Y, Tassiopoulos AK, McGraw DJ, Hauser MC, Camporesi EM, and Hakim TS. Segmental pulmonary vascular responses to changes in pH in rat lungs: role of nitric oxide. Acta
Anaesthesiol Scand 43: 64–70, 1999.
3. Grover RF, Wagner WW Jr, McMurtry IF, and Reeves JT.
Pulmonary circulation. In: Handbook of Physiology, Section 2:
The Cardiovascular System, edited by Shepherd JT and Abboud
F. Bethesda, MD: Am Physiol Soc, 1983, p. 103–136.
4. Hasunuma K, Rodman DM, and McMurtry IF. Effects of K⫹
channel blockers on vascular tone in the perfused rat lung. Am
Rev Respir Dis 144: 884–887, 1991.
5. Kato M and Staub NC. Response of small pulmonary arteries to
unilobar hypoxia and hypercapnia. Circ Res 19: 426–444, 1966.
6. Koyama T and Horimoto M. Pulmonary microcirculatory response to localized hypercapnia. J Appl Physiol 53: 1556–1564,
1982.
7. Koyama T and Horimoto M. Blood flow reductions in local
pulmonary microvessels during acute hypoxia imposed on a
small fraction of the lung. Respir Physiol 52: 181–189, 1983.
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.5 on June 15, 2017
calcium influx secondary to membrane depolarization.
This response has been observed in isolated pulmonary
vascular myocytes (9, 15) and is associated with decreased voltage-dependent potassium currents in these
cells (15). Thus it has been postulated that hypoxia
reduces potassium efflux, eliciting depolarization and
calcium entry through voltage-sensitive L-type calcium
channels. Data from the present study strongly suggest that a similar mechanism exists in the frog cutaneous vasculature. For example, hypoxic vasoconstriction was greatly blunted following topical application
of the L-type calcium-channel blocker nifedipine. Similar attenuation of hypoxic pulmonary vasoconstriction
has been observed with L-type calcium-channel blockers in the mammalian lung (13). Also, consistent with
earlier experiments on the pulmonary circulation (4,
15), we observed that administration of the voltagesensitive potassium-channel blocker 4-AP produced
vasoconstriction similar to that seen with hypoxia and
that this response was also inhibited by nifedipine. Our
observation that the L-type channel activator BAY K
8644 greatly potentiates hypoxic vasoconstriction further supports a role for this pathway of calcium influx
in the response. The lack of effect of BAY K 8644 on
resting tone suggests that the vascular smooth muscle
cells are relatively hyperpolarized under normoxic conditions, because this dihydropyridine acts to shift the
voltage-activation curve to the left for calcium influx
(16). Once again, the enhancement of hypoxic vasoconstriction in frog skin by BAY K 8664 parallels findings
in the mammalian lung (12, 17). Although nifedipine
consistently attenuated hypoxic vasoconstriction,
blockade of the response was not complete even with
higher doses of the antagonist. It is unclear whether
this residual response represents limitations of the
delivery of the antagonist by topical application or
whether additional parallel vasoconstrictor pathways
contribute to the hypoxic response similar to those
suggested in studies of lungs from other species. Thus
the characteristics of the vascular responses to hypoxia
in amphibian skin and the mammalian lung are strikingly similar and suggest a shared conserved mechanism in these anatomically distinct respiratory organs.
In addition to the response to hypoxia, we examined
the response to hypercapnia in our preparation. Consistent with our earlier work (10), we observed cutaneous vasoconstriction in response to 5% CO2 and determined that this response is also sensitive to nifedipine.
These data suggest that hypercapnia and hypoxia may
share the same signaling pathway to elicit vasoconstriction. This hypothesis is further supported by the
observation that the response to the two stimuli were
not additive and were inhibited to a similar degree by
L-type Ca⫹2-channel blockade. The mechanism of hypercapnic vasoconstriction in mammalian lungs is not
well studied; however, a recent study by Gao et al. (2)
showed that the response occurs primarily in small
arteries of the isolated rat lung and that it is prevented
by preconstriction of the lung with KCl. These data
suggest that, similar to the response to hypoxia, the
mechanism of hypercapnic vasoconstriction in the rat
R1313
R1314
HYPOXIC VASOCONSTRICTION IN FROG SKIN
8. Krogh A. Studies on the physiology of capillaries. II. The reactions to local stimuli of the blood vessels in the skin and web of
the frog. J Physiol (Lond) 55: 412–422, 1921.
9. Madden JA, Dawson CA, and Harder DR. Hypoxia-induced
activation in small isolated pulmonary arteries from the cat.
J Appl Physiol 59: 113–118, 1985.
10. Malvin GM. Microcirculatory effects of hypoxic and hypercapnic
vasoconstriction in frog skin. Am J Physiol Regulatory Integrative Comp Physiol 264: R435–R439, 1993.
11. Malvin GM and Hlastala MP. Effects of environmental O2 on
cutaneous blood flow and diffusing capacity in amphibians. Respir Physiol 76: 229–242, 1989.
12. McMurtry IF. Bay K 8644 potentiates and A23187 inhibits
hypoxic vasoconstriction in rat lungs. Am J Physiol Heart Circ
Physiol 249: H741–H746.
13. McMurtry IF, Davidson AB, Reeves JT, and Grover RF.
Inhibition of hypoxic pulmonary vasoconstriction by calcium
channel antagonists in isolated rat lungs. Circ Res 38: 99–104,
1976.
14. Pettersson K and Johansen K. Hypoxic vasoconstriction and
the effects of adrenaline on gas exchange efficiency in fish gills.
J Exp Biol 97: 263–272, 1982.
15. Post JM, Hume JR, Archer SL, and Weir EK. Direct role for
potassium channel inhibition in hypoxic pulmonary vasoconstriction. Am J Physiol Cell Physiol 262: C882–C890, 1992.
16. Sanguinetti MC, Krafte DS, and Kass RS. Voltage-dependent modulation of Ca channel current in heart cells by Bay
K8644. J Gen Physiol 88: 369–392, 1986.
17. Tolins M, Weir EK, Chesler E, Nelson DP, and From AHL.
Pulmonary vascular tone is increased by a voltage-dependent
calcium channel potentiater. J Appl Physiol 60: 942–948, 1986.
18. Walker BR. Evidence for uneven distribution of L-type calcium
channels in rat pulmonary circulation. Am J Physiol Heart Circ
Physiol 269: H2051–H2056, 1995.
19. Weir EK and Archer SL. The mechanism of acute hypoxic
pulmonary vasoconstriction: the tale of two channels. FASEB J
9: 183–189, 1995.
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.5 on June 15, 2017