Polycythemic responses to hypoxia: molecular and
genetic mechanisms of chronic mountain sickness
L. C. OU, S. SALCEDA, S. J. SCHUSTER, L. M. DUNNACK,
T. BRINK-JOHNSEN, J. CHEN, AND J. C. LEITER
Departments of Physiology, Pathology, and Medicine, Dartmouth Medical School, Lebanon,
New Hampshire 03756-0001; and Cardeza Foundation for Hematological Research,
Department of Medicine, Thomas Jefferson University, Philadelphia, Pennsylvania 19107
erythropoietin; gene expression; renal oxygenation; renal
work; tissue hypoxia; high altitude
POLYCYTHEMIA is a compensatory mechanism to sustain
O2 delivery during life at high altitude, but excessive
polycythemia is associated with chronic mountain sickness (CMS). There is, on average, an increase in
hematocrit (Hct) after altitude exposure, but there is
considerable individual variation, and only a small
number of high-altitude residents develop CMS. The
origin of the variation among individuals is unknown,
but variation in erythropoietin (EPO) responses to
equivalent hypoxic stress is a tenable hypothesis. EPO,
a glycoprotein growth factor, regulates the rate of red
blood cell production by stimulating the proliferation
1242
and differentiation of erythroid precursor cells (12).
EPO synthesis in adult mammals occurs primarily in
the kidney (19), and hypoxia in the kidney from a
variety of causes (e.g., high-altitude exposure and
anemia) stimulates the synthesis and release of EPO,
which are followed by increased production of red blood
cells. Recent molecular studies in vivo (2, 32–34) and in
vitro (17) showed that hypoxia or cobalt treatment led
to rapid accumulation of EPO mRNA in the kidney or in
cultured hepatoma cell lines, and detection of EPO
mRNA preceded the appearance of EPO in plasma or in
cultured medium. Hence, accelerated EPO gene expression is the first step in the polycythemic responses to
hypoxic or cobalt treatment. The location(s) of the
hypoxic sensing mechanism is controversial. The adequacy of renal tissue oxygenation at the EPOproducing sites is thought to be the immediate signal
regulating EPO production (22), but a more potent
extrarenal sensing mechanism has also been postulated (31). Renal tissue oxygenation depends on the
relationship between renal tissue O2 delivery, a function of arterial O2 content [CaO2, which depends on
arterial PO2 (PaO2) and Hct] and renal blood flow, and
renal tissue O2 consumption (V̇O2), a function of the
excretory activity of the kidney. Therefore, there may
be a link between renal excretory function and the
regulation of EPO production (10, 11, 13). Polycythemia
following EPO release ought to improve renal tissue
oxygenation and reduce EPO synthesis. However,
changes in EPO levels, renal tissue oxygenation, and
Hct are not well correlated in time, and the classic
concept of a negative-feedback control of EPO production by a polycythemic response remains debatable (3,
9, 28).
We have studied two Sprague-Dawley rat strains
with marked differences in the propensity to develop
CMS after chronic exposure to hypoxia: Hilltop (H) rats
develop excessive polycythemia, accentuated hypoxemia, and severe pulmonary hypertension associated
with a high mortality rate; Madison (M) rats develop
only moderate polycythemia and pulmonary hypertension and no significant mortality (30). Excessive polycythemia apparent in H rats after 3 wk of exposure to a
simulated altitude of 5,500 m was associated with
persistent elevation of plasma and renal EPO (28).
Metabolic degradation of EPO did not differ in the two
rat strains, and the persistent elevation of EPO probably resulted from sustained EPO gene expression. In
past experiments, high blood viscosity associated with
excessive polycythemia in the H rats (Hct often .70%)
may have compromised renal oxygenation and thereby
8750-7587/98 $5.00 Copyright r 1998 the American Physiological Society
http://www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Ou, L. C., S. Salceda, S. J. Schuster, L. M. Dunnack, T.
Brink-Johnsen, J. Chen, and J. C. Leiter. Polycythemic
responses to hypoxia: molecular and genetic mechanisms of
chronic mountain sickness. J. Appl. Physiol. 84(4): 1242–
1251, 1998.—We examined erythropoietin (EPO) gene expression and EPO production during hypoxia in two SpragueDawley rat strains with divergent polycythemic responses to
hypoxia. Hilltop (H) rats develop severe polycythemia, severe
hypoxemia, and pulmonary artery hypertension. The H rats
often die from a syndrome indistinguishable from chronic
mountain sickness (CMS) in humans. Madison (M) rats
develop polycythemia and pulmonary artery hypertension
that is modest and suffer no excess mortality. We tested the
hypothesis that these rat strains have different stimulusresponse characteristics governing EPO production. Rats of
each strain were exposed to hypoxia (0.5 atm, 73 Torr inspired
PO2), and renal tissue EPO mRNA and EPO levels, plasma
EPO, ventilation, arterial and renal venous blood gases, and
indexes of renal function were measured at fixed times during
a 30-day hypoxic exposure. During extended hypoxic exposure, H rats had significantly elevated renal EPO mRNA,
renal EPO, and plasma EPO levels compared with M rats.
Ventilatory responses and indexes of renal function were
similar in the strains during the hypoxic exposure. H rats had
greater arterial hypoxemia from the onset of hypoxia and
more severe renal tissue hypoxemia and greater polycythemia after 14 days of hypoxic exposure. When EPO responses
were expressed as functions of renal venous PO2, the two
strains appeared to lie on the same dose-response curves, but
the responses of H rats were shifted along the curve toward
more hypoxic values. We conclude that H rats have significantly greater polycythemia secondary to poorer renal tissue
oxygenation, but the stimulus-response characteristics governing EPO gene expression and EPO production do not seem to
differ between M and H rats. Finally, the regulation of EPO
levels during hypoxia occurs primarily at the transcriptional
level.
POLYCYTHEMIC RESPONSES IN HYPOXIA
enhanced EPO production (13, 28). In addition, there
was a positive correlation between the extent of polycythemia and EPO production, which was not compatible
with a negative-feedback control mechanism regulating EPO production. The purposes of the present study
were to use the H and M rat strains 1) to characterize
EPO gene expression and EPO production during
chronic hypoxia, 2) to examine the relationships among
the levels of EPO gene expression and EPO production
and renal tissue oxygenation, and 3) to examine the
notion of a negative-feedback control mechanism of
EPO production.
MATERIALS AND METHODS
altitude of 5,500 m. In RNase protection assays, total RNA
was probed for EPO mRNA without separation of the poly(A)1
RNA. The details of the RNase protection assay for EPO
mRNA have been described previously (33). Briefly, total
tissue RNA samples prepared by the acid guanidinium thiocyanate-phenol-chloroform method (4) were analyzed using
an RNase protection assay (16). The template for production
of the complementary strand RNA probe was a polymerase
chain reaction-derived rat EPO cDNA fragment cloned into
the plasmid vector pCR1000 (Invitrogen, San Diego, CA). The
labeled RNA probe was generated by in vitro transcription by
T7 RNA polymerase (Life Technologies, Gaithersburg, MD)
using [a-32P]UTP (29.6 TBq/mM; Amersham, Arlington
Heights, IL). Before hybridization, an equal amount of each
RNA sample was visualized by ethidium bromide-stained
agarose gel electrophoresis to verify quantification and integrity of samples. RNA samples were hybridized overnight in 30
µl of hybridization buffer at 45°C with 0.5 3 106 cpm of
labeled probe. RNase digestion was performed at 30°C for 1 h
using RNase T1 (Boehringer-Mannheim, Indianapolis, IN).
The protected fragments were separated on a denaturing 8%
acrylamide-7 M urea gel and analyzed by autoradiography.
The autoradiographs were quantitated by scanning with a
BIO-Imaging analyzer (Fuji Medical System).
EPO was measured by radioimmunoassay using antiserum
against human recombinant EPO (Incstar, Stillwater, MN),
as previously described (28).
Measurements of ventilation, renal function, and renal
tissue oxygenation in fully awake and chronically instrumented rats. Ventilation was measured in a whole body
plethysmograph, as previously described (28). The following
renal variables were measured at each designated time: urine
output, sodium and potassium excretions, plasma sodium
and potassium concentrations, and sodium reabsorption.
Potassium and sodium concentrations were measured with a
flame photometer (model FLM3, Radiometer America, Cleveland, OH). The glomerular filtration rate (GFR) was measured using polyfructosan (15), and renal plasma flow was
measured using p-aminohippurric acid by the method of
Bratton and Marshall as modified by Smith et al. (35). Urine
flow and clearance rates are expressed per 100 g of body
weight. Hct was determined by a micromethod. Arterial and
venous blood gas samples were obtained anaerobically by
withdrawing 300 µl from the rat; only the last 160 µl were
used for analysis. The residue was returned to the rat. pH,
PO2, and PCO2 were measured with microelectrodes at 37°C
(model BMS 3 MK2, Radiometer America). Hypoxic exposure
did not affect the shape of the oxyhemoglobin dissociation
curve in the H or M rats, and the O2 content in arterial and
venous blood was estimated from a rat oxyhemoglobin dissociation curve after correction for the blood pH measured in
vivo (21). Renal blood flow was calculated from renal plasma
flow and the Hct. The renal coefficient of O2 delivery (COD)
was calculated by multiplying renal blood flow by CaO2, and
renal V̇O2 was calculated by multiplying renal blood flow by
the renal arteriovenous O2 content difference.
Surgical preparation. Four to 5 days before an experiment,
catheters were implanted in each animal in the urinary
bladder, femoral artery, and left renal and right external
jugular veins under anesthesia using a combination of ketamine (60 mg/kg body wt im) and pentobarbital sodium (20
mg/kg body wt ip), as previously described (29). After surgery,
each rat was treated with penicillin (100,000 U daily im for 5
days) and returned to the presurgical altitude immediately
after recovery from anesthesia. All rats were allowed free
access to water and laboratory rat chow. The animals were
acclimated to a plastic restraining cage for 2 days before the
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Animals. Male Sprague-Dawley rats weighing 270–320 g
were obtained from Hilltop (H), Scottsdale, PA (altitudesensitive strain) and Madison, WI (M) breeding laboratories.
There were two separate series of experiments: in one series
of experiments, levels of EPO mRNA and EPO were measured; in the other, ventilatory and blood-gas responses, renal
function, and renal oxygenation were measured. In each set of
experiments, five groups of five to six rats from each strain
were exposed to a simulated altitude of 0.5 atm (5,500 m, 73
Torr inspired PO2) for various durations (0, 6, and 24 h and 1,
3, 7, 14, 21, 28, and 30 days). Supplementary experiments
were performed to test the effects of cobalt or combined cobalt
and hypoxia treatment on EPO and EPO mRNA production.
Measurements of EPO mRNA and EPO. Two methods were
used to measure EPO mRNA: Northern and slot-blot analyses
using polyadenylated RNA [poly(A)1 RNA] were used in early
experiments, and a highly sensitive RNase protection assay
was employed later in the study. EPO gene probes for
Northern blot studies were provided by Dr. E. Goldwasser
(University of Chicago). The probes were 1.0- and 1.2-kb Pst I
restriction fragments of the cloned mouse EPO gene. The
1.2-kb fragment contained all of exons 2 and 3 and part of
exon 4. The 1.0-kb fragment contained the remainder of exon
4 and part of exon 5. The probes were labeled with [32P]CTP
(3,000 Ci/mmol; ICN, Costa Mesa, CA) to high specific
activities (2 3 108 dpm/µg) using a nick translation kit (BRL,
Bethesda, MD) according to the manufacturer’s instructions.
Probes were denatured by heating at 100°C for 10 min and
quenched on ice immediately before use. RNAs were isolated
from tissue homogenates by extraction with 4 M guanidine
isothiocyanate and sedimentation of RNA through 5.7 M
cesium chloride (4). Poly(A)1 RNA was isolated by chromatography on oligo dT-cellulose. For electrophoresis, 6–30 µg of
poly(A)1 RNA per lane were loaded on formaldehyde-1.0%
agarose gels. After electrophoresis, the gels were blotted onto
nitrocellulose filters and dried under vacuum at 80°C for 2 h.
Northern and slot blots were hybridized with 32P-labeled
probe at 42°C for 18 h and washed using standard methods.
The blots were exposed to Kodak XAR-5 X-ray film (Eastman
Kodak, Rochester, NY) with an intensifying screen at 270°C.
The EPO mRNA was quantitated by scanning densitometry.
Results from Northern blots were expressed as densitometric
units per microgram of RNA loaded. As an internal control for
RNA loading, Northern and slot blots were probed with a
mouse b-actin probe. Splenic poly(A)1 RNA was used as a
negative control.
Northern blot analysis was not sensitive enough to detect
renal EPO mRNA in rats at sea level or rats exposed to a
simulated altitude of 5,500 m (10.5% O2 or 73 Torr inspired
PO2). For this reason, an RNase protection assay was employed to follow the activity of EPO gene expression during
the development of CMS in rats exposed to a simulated
1243
1244
POLYCYTHEMIC RESPONSES IN HYPOXIA
renal function studies. The animals recovered from surgery
uneventfully, and the majority gained weight by the time of
measurements.
Experimental procedures. During the measurements, each
rat was housed in a restraining cage. A Plexiglas hood was
fitted over the front of the restraining cage so that the desired
gas mixture could be flushed through the hood. Renal hemodynamic values were measured (29), and the exposed end of the
bladder catheter was extended with a short length of polyethylene tubing to allow collection of urine under the restraining
cage. Urine volume was measured by weight. After 30–40
min of equilibration, two to three samples of urine and
arterial and renal venous blood were obtained from each
animal under appropriate PO2 conditions. To avoid possible
adverse effects of repeated blood sampling that might change
the Hct, each rat was studied only once.
RESULTS
Fig. 1. Effect of combined severe hypoxic exposure (7,300 m simulated altitude) and cobalt treatment on plasma erythropoietin (EPO;
A) and renal tissue EPO mRNA (B) in Hilltop (H) and Madison (M)
rats. H rats (lane 1) and M rats (lane 2) were injected with cobalt
chloride (60 mg/kg sc) and 12 h later were exposed to hypobaric
hypoxia (0.4 atm) for 4 h. Control H rats (lane 3) and M rats (lane 4)
were injected with saline and maintained under room air conditions.
At end of exposure, animals were killed, plasma was separated for
EPO assay, and poly(A)1 RNA was isolated from kidney for RNA blot
analysis. Hybridization to radiolabeled mouse b-actin is shown as a
control.
gene in kidneys of H rats during hypoxic exposure. EPO
mRNA was undetectable in the spleen, heart, or liver in
H and M rats under control and hypoxic conditions. In
contrast to the response to 10.5% inspired O2, cobalt
injection (60 mg/kg sc) elicited equivalent accumulation
of EPO mRNA in the kidney 6 h after treatment in both
rat strains (Fig. 3). Thus RNase and Northern blot
assays demonstrate no strain difference in EPO gene
expression after cobalt treatment.
The time courses of changes in circulating and renal
tissue EPO levels during chronic hypoxia in the two rat
strains are portrayed in Fig. 4. There were no strain
differences in plasma or renal tissue EPO levels under
sea-level control conditions. EPO levels rose markedly
at 12 h after exposure to hypoxia in both rat strains.
The mean value of this increase was higher in the H
than in the M rats, but the difference was not statistically significant. The EPO levels declined at 24 h in the
plasma and the kidney in both rat strains, but the
decrease was more profound in the M than in the H
rats. Thereafter, the circulating and renal tissue EPO
levels remained low or continued to decline toward the
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Effect of extreme hypoxia and cobalt on EPO gene
expression and EPO levels in H and M rats. Northern
blot analysis failed to detect EPO mRNA in the kidney
in either rat strain during exposure to a simulated
altitude of 5,500 m, even though poly(A)1 RNA was
used. To test the sensitivity of the Northern blot
analysis, extreme stimuli, severe hypoxia alone (7,300
m simulated altitude or 65 Torr inspired O2 ) or combined severe hypoxia (7,300 m) and cobalt chloride (60
mg/kg sc) treatment, were used. The results are presented in Fig. 1. Under sea-level conditions, no EPO
mRNA was detected in the kidney (Fig. 1B, lanes 3 and
4), and no strain difference in plasma EPO levels was
detected (Fig. 1A, lanes 3 and 4). The EPO mRNA
increased markedly in both rat strains during severe
hypoxia (data not shown) and during combined hypoxia
and cobalt injection (Fig. 1B, lanes 1 and 2), but there
were no strain differences under these conditions. The
combined stimuli elevated the plasma EPO levels from
control values of 59.5 6 12.7 to 2,234.4 6 632.5 mU/ml
in the H rats and from 56.6 6 6.1 to 2,705.7 6 451.3
mU/ml in the M rats. These findings indicate that the
Northern blot analysis could detect EPO mRNA but
was insensitive to the EPO mRNA changes during
exposure to 10.5% inspired O2 (5,500 m). Therefore, a
more sensitive RNase assay was used in subsequent
experiments.
Effect of 10.5% inspired O2 on EPO gene expression
and EPO levels in H and M rats. The results of the EPO
mRNA RNase protection analysis of total RNA from the
kidney are summarized in Fig. 2A, and the quantitative
relationships of renal EPO mRNA as estimated by
phosphor image analysis are shown in Fig. 2B for each
rat strain. At sea level, EPO mRNA levels were equivalent in the two rat strains. After 6 h of hypoxic
exposure, EPO mRNA markedly increased in both rat
strains, but the increase was at least 100% higher in
the H than in the M rats. The EPO mRNA fell precipitously after 24 h of hypoxic exposure but remained
significantly elevated above the control values in the H
rats throughout the entire period of exposure. In contrast, EPO mRNA levels in the M rats fell toward the
sea-level control values after 24 h of hypoxia. The
results demonstrate exaggerated expression of the EPO
POLYCYTHEMIC RESPONSES IN HYPOXIA
1245
sea-level control levels in the M rats, whereas those in
the H rats tended to rise. As a result, the circulating
and renal tissue EPO levels in the chronically hypoxic
H rats were significantly higher than the sea-level
control values and higher than in the hypoxic M rats at
each time studied. The patterns of changes in circulating and renal tissue EPO levels during hypoxic exposure parallel those of EPO mRNA.
Effect of 10.5% inspired O2 on renal function in
conscious H and M rats. We examined a variety of
aspects of renal function in H and M rats under the
hypoxic conditions. These results are summarized in
Table 1. There were no strain differences in urine
output, plasma sodium, sodium excretion, or GFR at
sea level or at any point during hypoxic exposure.
Neither the plasma potassium (range 3.7–4.5 meq/l)
nor potassium excretion (range 0.6–1.1 meq · l21 · min21 )
differed between strains or across times of hypoxic
exposure (data not shown). Plasma sodium was signifi-
Fig. 2. A: autoradiograph of RNase protection assay of total RNA
from yeast tRNA (lane 1), anemic-hypoxic (lane 2) and normal control
rat kidney (lane 3) of another commercially available strain, and
control [sea level (SL)] and acute (6 and 24 h) and chronic (7, 14, 21,
28, and 30 days) hypoxic H and M rat kidneys (100 µg RNA each).
Arrowheads, protected EPO mRNA fragment. Before hybridization,
an equal amount of each RNA sample was visualized by ethidium
bromide staining and agarose gel electrophoresis to verify quantification and integrity of samples. B: relative amount of renal tissue EPO
mRNA in H (r) and M (s) rats during acute and chronic hypoxic
exposure and during posthypoxic recovery. 32P was quantitated as
relative amount of EPO mRNA by a BIO-Imaging analyzer.
cantly less than the sea-level value on days 1 and 14,
when data from H and M rats were pooled (there was a
significant main effect of the duration of altitude exposure). Serum sodium on day 14 was also less than
sea-level control, but this failed to achieve statistical
significance (P 5 0.054). Plasma sodium was significantly less in the M than in the H rats on day 14 only.
The fractional sodium reabsorption was significantly
less than sea-level values on day 1, when results from
H and M rats were pooled at each exposure time.
Sodium excretion did not differ between strains or
among exposure times. However, the largest sodium
excretion occurred on day 1 in both strains, which is
consistent with the lower fractional sodium reabsorption at that time. A reduction in sodium reabsorption
often occurs in the first 24–48 h of acclimatization to
high altitude (18). The GFR was constant at all altitudes and similar in the two strains. The changes in
renal excretory function are statistically significant but
of small magnitude, and the work of the kidney was
maintained during hypoxic exposure.
Effect of 10.5% inspired O2 on systemic and renal
tissue oxygenation in conscious H and M rats. Table 2
summarizes PaO2, renal venous PO2 (PrvO2), CaO2, renal
venous O2 content (CrvO2), renal V̇O2, and COD measured or calculated in H and M rats at sea level and
during hypoxic exposure. There were no strain differences in any of these variables at sea level. PaO2 fell
precipitously at the onset of hypoxic exposure in both
rat strains. PaO2 in the H rats decreased as the hypoxic
exposure persisted, whereas PaO2 in the M rats re-
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Fig. 3. EPO mRNA levels in M and H rats as measured in an RNase
protection assay of total RNA isolated from kidney during room air
(Basal) conditions and after cobalt treatment. Further controls
consisted of examining EPO mRNA levels in normoxic (NK) and
hypoxic (HK) kidney of another rat strain and in yeast tRNA (t).
Except for lane t, in which 80 µg of tRNA were used, 100 µg of RNA
were used in each lane.
1246
POLYCYTHEMIC RESPONSES IN HYPOXIA
Fig. 4. Changes in plasma (A) and renal tissue (B) EPO levels in H
(r) and M (s) rats during acute and chronic exposure to a simulated
altitude of 5,500 m. * P # 0.05 compared with M rats at same time
point.
mained relatively stable and even increased over the
remainder of the hypoxic exposure. As a result, the
mean value of PaO2 was ,41 Torr in the H rats and 48
Torr in the M rats at the end of 30 days. PrvO2 was
similar at sea level in H and M rats and fell at the onset
of hypoxia. PrvO2 values paralleled the arterial values
during hypoxia in both strains. The initial hypoxic
PrvO2 values were similar, but PrvO2 in H rats fell after
the initial hypoxic exposure and remained low throughout the 30-day hypoxic period. In contrast, PrvO2 rose
steadily from the onset of hypoxia toward the sea-level
value in M rats. Although PrvO2 remained significantly
Table 1. Parameters of renal function in rate at sea level and during hypoxia
Urine Output,
µl · min21 · 100 g body wt21
Sea level
Hypoxia
Day 1
Day 3
Day 14
Day 30
H
M
7.4 6 3.0
(5)
9.1 6 5.1
(5)
8.0 6 2.6
(5)
6.0 6 1.4
(5)
8.8 6 4.2
(5)
9.9 6 5.5
(6)
6.2 6 0.7
(5)
7.2 6 2.6
(5)
4.7 6 0.4
(5)
6.3 6 1.9
(5)
Plasma Na, meq/l
H
M
Na Excretion,
µeq · min21 · 100 g body wt21
Na Reabsorption, %
H
M
142.4 6 1.2 142.8 6 0.9
(6)
(4)
0.6 6 0.2
(6)
0.8 6 0.4
(4)
99.4 6 0.2 99.5 6 0.1
(6)
(4)
137.9 6 2.3
(5)
139.4 6 0.9
(5)
140.9 6 2.0
(5)
142.8 6 4.1
(6)
1.2 6 0.4
(5)
0.5 6 0.3
(5)
0.5 6 0.5
(5)
0.5 6 0.4
(6)
0.8 6 0.3
(5)
0.6 6 0.4
(5)
0.4 6 0.1
(5)
0.3 6 0.1
(5)
98.9 6 0.2
(5)
99.4 6 0.3
(5)
99.6 6 0.3
(5)
99.5 6 0.3
(6)
140.3 6 1.2†
(5)
140.1 6 1.8
(5)
136.8 6 1.6*†
(5)
140.9 6 3.3
(5)
H
M
99.2 6 0.3†
(5)
99.4 6 0.1
(5)
99.6 6 0.1
(5)
99.6 6 0.2
(5)
GFR,
µl · min21 · 100 g body wt21
H
M
1,025 6 130
(4)
910 6 130
(4)
958 6 110
(2)
889 6 112
(5)
983 6 11
(2)
807 6 43
(3)
Values are means 6 SD; number in parentheses is number of rats. GFR, glomerular filtration rate. * Significant difference between strains
at a particular time, P # 0.05. †Significantly different from sea-level control when values from H and M rats were pooled (a main effect of altitude and
time), P # 0.05.
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
below sea-level values in both strains, PrvO2 values
were greater in M than in H rats by day 14, and this
persisted to the conclusion of the study. The hemoglobin concentration was similar in the strains at sea
level, rose in both strains, and was significantly greater
than the sea-level value from day 3 until the conclusion
of the study. The increase in hemoglobin was greater in
H rats from day 14 onward. CaO2 fell in both strains at
the onset of hypoxia but rose steadily until day 30,
when CaO2 had returned to the sea-level range. The
restoration of CaO2 to the sea-level range was achieved
by different mechanisms in each strain, as discussed
below. CrvO2 was similar at sea level in the two strains
and fell at the onset of hypoxia. CrvO2 rose in both
strains, but the rise was more rapid in the M rats.
Renal venous CO2 exceeded the sea-level value on day
14 in M rats and on day 30 in H rats. Renal blood flow
was a constant function of Hct in both strains and
increased gradually as the hypoxic exposure progressed and Hct rose (C. D. Thron, J. Chen, J. C. Leiter,
and L. C. Ou, unpublished observations). The renal
COD, the product of renal blood flow and CaO2, fell at
the onset of hypoxic exposure but increased steadily as
Hct rose. The renal COD did not differ between strains,
and in both strains the renal COD was significantly less
than the sea-level value on days 1 and 3 and greater
than the sea-level value by day 30 of hypoxic exposure.
The mechanism whereby equivalent renal O2 delivery
was maintained differed in the two strains: hemoglobin
was greater and PaO2 was lower in the H rats; the
reverse was true in the M rats. Renal V̇O2 was similar
in H and M rats at sea level and at each particular
altitude, but when pooled across all altitude conditions,
renal V̇O2 was significantly lower in the M rats (main
effect of strain).
EPO levels increase during hypoxia (low PaO2 and
low tissue PO2) and carbon monoxide exposure (normal
PaO2 but low tissue PO2). Hence, the hypoxic sensor for
EPO synthesis seems to respond to tissue PO2 (23).
PrvO2 closely approximates renal tissue PO2 (37), and
values of EPO mRNA, renal EPO, and plasma EPO
have been plotted as functions of PrvO2 in Fig. 5. The
EPO and PrvO2 values were obtained from different sets
1247
POLYCYTHEMIC RESPONSES IN HYPOXIA
Table 2. Arterial and renal blood-gas values in rats at sea level and during hypoxia
Hb, g/dl
H
Sea level
M
CaO2 , ml/dl
CrvO2 , ml/dl
Renal V̇O2 ,
µmol · min21 ·
100 g body wt21
Renal COD,
µmol · min21 ·
100 g body wt21
PaO2 , Torr
PrvO2 , Torr
H
M
H
M
H
M
H
M
H
M§
H
M
14.9
6 0.4
(5)
14.4
6 1.1
(5)
80
62
(5)
77
64
(5)
51
61
(5)
48
63
(5)
18.6
6 0.6
(5)
17.4
6 1.2
(5)
10.8
6 0.4
(5)
9.8
6 0.9
(5)
9.0
6 0.8
(7)
7.3
6 1.0
(8)
39.7
6 3.4
(7)
36.3
6 4.3
(8)
15.4
6 1.4
(5)
17.8*
6 0.4
(5)
21.3*†
6 0.7
(5)
25.5*†
6 0.7
(5)
15.7
6 1.9
(5)
17.7*
6 0.3
(5)
19.5*
6 1.0
(5)
21.2*
6 0.4
(5)
45*
60
(5)
42*
61
(5)
38*†
62
(5)
41*†
61
(5)
44*
62
(5)
45*
62
(5)
47*
62
(5)
48*
62
(5)
35*
62
(5)
33*
62
(5)
31*†
62
(5)
34*†
64
(5)
35*
62
(5)
36*
62
(5)
40*
60
(5)
41*
61
(5)
11.0*
6 1.2
(5)
12.0*
6 0.6
(5)
13.1*
6 0.9
(5)
17.0
6 1.0
(5)
10.9*
6 1.4
(5)
12.4*
6 0.9
(5)
14.7*
6 0.8
(5)
18.0
6 0.4
(5)
7.4*
6 1.2
(5)
8.5*
6 0.6
(5)
9.7†
6 0.5
(5)
13.2*
6 0.8
(5)
7.6*
6 1.3
(5)
9.1
6 1.0
(5)
12.0*
6 0.3
(5)
14.9*
6 0.6
(5)
7.4
6 1.9
(4)
8.0
6 1.2
(5)
8.4
6 2.7
(7)
10.6
6 2.1
(9)
6.3
6 1.2
(5)
7.4
6 1.4
(5)
6.5
6 1.7
(7)
8.4
6 2.4
(8)
22.0
6 4.8
(4)
27.4
6 4.4
(5)
36.3
6 5.4
(7)
53.8
6 6.0
(8)
20.2‡
6 3.1
(5)
28.6‡
6 3.6
(5)
35.8
6 8.0
(7)
47.9‡
6 8.1
(8)
Hypoxia
Day 1
Day 3
Day 14
Day 30
of animals studied under identical conditions at equivalent times. In all cases the curves were hyperbolic with
a threshold between 40 and 50 Torr and an asymptote
between 19 and 29 Torr. The hyperbolic equations were
statistically significant, and, even for the worst fit, 58%
of the variation in kidney EPO levels could be attributed to the variation in PrvO2.
Effect of 10.5% inspired O2 on ventilatory response in
conscious H and M rats. To determine the cause of the
accentuated hypoxemia in the H rats during hypoxic
exposure, minute ventilation (V̇E ) at sea level and at a
variety of times during the hypoxic exposure was
measured in H and M rats. The results of measurements of V̇E, PaO2, and arterial PCO2 (PaCO2) are displayed in Fig. 6. There were no differences in V̇E and
PaCO2 in the two rat strains at sea level. V̇E increased
and PaCO2 decreased similarly in response to hypoxic
exposure in the H and M rats, despite the marked
strain differences in PaO2. V̇E remained similar in the H
and M rats as the H rats became more hypoxemic late
in the hypoxic exposure period. However, we have
never detected any differences in hypoxic or hypercapnic ventilatory responses at sea level or at any time
during exposure to simulated high altitude (28). Furthermore, we have seen in previous studies greater
ventilatory responses associated with more severe hypoxemia in the H than in the M rats after chronic
exposure to hypoxia (7) and after monocrotaline treatment (5).
DISCUSSION
In the present study we examined the molecular and
genetic mechanisms of polycythemic responses to hypoxia in H and M rats during chronic hypoxic exposure
Fig. 5. Kidney EPO mRNA (A), renal tissue EPO (B), and plasma EPO levels (C) plotted as functions of renal
venous PO2 (PrvO2) in H (r) and M rats (s); y 5 a/(PrvO2 2 c). Data were pooled from each exposure time. EPO and
PrvO2 values were not obtained from same rats but were obtained from sets of rats in identical treatment protocols.
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Values are means 6 SD; number in parentheses is number of rats. Arterial values are representative of systemic blood; venous values reflect
renal venous blood. Hb, hemoglobin; PaO2 and PrvO2 , arterial and renal venous PO2 , respectively; CaO2 and CrvO2 , arterial and renal venous O2
content, respectively; V̇O2 , O2 consumption; COD, coefficient of O2 delivery (blood flow 3 CaO2). * Significantly different from control value of
that strain, P # 0.05. † Significant difference between strains at that particular time, P # 0.05. ‡ Significantly different from control value
when values from both strains were pooled (a main effect of altitude), P # 0.05. § M rats had a lower renal V̇O2 than H rats when values across
all exposure times were pooled (a main effect of strain), P # 0.05.
1248
POLYCYTHEMIC RESPONSES IN HYPOXIA
and the factors thought to stimulate a polycythemic
response to hypoxia: the balance between renal O2
delivery and renal V̇O2. Because hypoxic exposure
results in development of CMS with excessive polycythemia in the H rats but elicits only moderate polycythemia and no apparent ill effects in the M rats, the
present results provide a spectrum of hypoxic responses within which to analyze the mechanism(s)
underlying the widely different polycythemic responses
in healthy and diseased states at high altitude.
Molecular and genetic mechanisms of different polycythemic responses to hypoxia in H and M rats. The
present study revealed, for the first time, the sequence
of events, starting from enhanced expression of the
EPO gene to elevated plasma and renal tissue EPO
levels to, finally, the polycythemic responses in rats
during chronic exposure to a simulated altitude of
5,500 m. The levels of EPO mRNA that accumulated
with 10.5% inspired O2, as employed in this study, could
not be detected by Northern blot analysis but were
readily detectable by a more sensitive RNase protection
analysis (33). RNase assays are more sensitive than
Northern blot analysis, but a mouse probe was used for
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Fig. 6. Arterial PO2 (PaO2; A), arterial PCO2 (PaCO2; B), and ventilatory
(V̇E; C), responses in conscious H (r) and M (s) rats during exposure to a
simulated altitude of 5,500 m. Values are means 6 SD. bw, Body weight.
*Significant differences (P # 0.05) between strains at each time tested.
the Northern blot analysis in this study, which might
further reduce the sensitivity of the particular Northern blot analysis we performed. The exaggerated EPO
production and excessive polycythemia, which developed in the H rats during hypoxic exposure, originated
from an inordinate and sustained expression of the
EPO gene in the kidney (28, 30). Because no EPO
mRNA was detectable in liver, heart, or spleen by
Northern blot or RNase protection assays under sealevel control and hypoxic conditions in both rat strains,
the kidney must be the primary site of EPO production
in the adult animals exposed to 10% inspired O2. This
conclusion is supported by the negligible increase in
circulating EPO levels in nephrectomized rats of both
strains when exposed to severe hypoxia (28). There
were no strain differences in the rates of EPO degradation in these two rat strains under control and hypoxic
conditions (28). Therefore, the rate of EPO gene expression in the kidney is the major factor determining the
levels of EPO mRNA production and the polycythemic
response to hypoxia in the H and M rats. This conclusion, however, is at variance with the work of Tan et al.
(36). Using a similar RNase protection assay, these
authors observed that EPO mRNA increased significantly in the liver and spleen in rats exposed to
hypoxia. This discrepancy could be due to the difference
in severity of the hypoxic stimulus used in the two
studies: 7% O2 was used in the study of Tan et al.,
whereas 10.5% O2 was employed in the present study.
When the animals were exposed to severe hypoxia
with or without cobalt treatment, circulating EPO and
renal tissue EPO mRNA levels were elevated and
similar in the H and M rats. M rats were capable of
mounting an EPO response; there is no evidence that
deficient EPO-generating capacity in the M rats accounted for the different EPO responses reported here.
The different EPO responses in the two rat strains
appear to be hypoxia specific. Treatment with cobalt,
another erythropoietic stimulus, enhanced EPO gene
expression and EPO production similarly in H and M
rats. Moreover, elevated EPO mRNA and circulating
EPO levels in the chronically hypoxic H rats fell
abruptly to sea-level control values when the animals
were brought down to sea-level conditions. Rather than
strain-specific differences in EPO regulation, more
severe renal tissue hypoxia in the H than in the M rats
during equivalent hypoxic exposure seems to account
for the divergent EPO responses observed in this study
(Fig. 5, Table 2).
Regulation of EPO gene expression and EPO production. The physiological mechanism regulating EPO
gene expression and EPO production remains incompletely understood. Renal tissue oxygenation is undoubtedly the single most important factor (12, 22). The
balance between renal O2 delivery and renal V̇O2 determines the level of renal tissue oxygenation. Renal V̇O2
depends on the work of the kidney, particularly sodium
reabsorption, which is the primary energy-requiring
renal process (22, 38). Inhibition of proximal sodium
reabsorption attenuates the EPO response to hypoxia,
presumably as a result of decreased renal work and
POLYCYTHEMIC RESPONSES IN HYPOXIA
greater, were consistently more hypoxemic than M
rats. We have no evidence that the sensitivity of the O2
sensor for EPO expression differs between strains
acutely or chronically.
More severe renal tissue hypoxia developed in the H
rats during hypoxic exposure as a result of accentuated
arterial hypoxemia. Accentuated hypoxemia in patients with CMS is believed to result from hypoventilation (27, 40), even though hypoventilation has not been
observed consistently in patients with CMS (6). Impaired pulmonary gas exchange is found in patients
with CMS and must contribute to systemic arterial
hypoxemia (39). In this and previous studies (28, 30),
accentuated hypoxemia in the H rats during hypoxic
exposure also resulted from abnormal gas exchange.
Ventilatory responses to a simulated altitude of 5,500
m were similar in the H and M rats; PaCO2 fell in
response to hypoxic exposure to comparable levels in
both rat strains (28), indicative of a comparable alveolar hyperventilation. Severe pulmonary hypertension,
one of the characteristic signs of CMS in humans and
rats, is associated with vascular remodeling (25) and
may impair gas exchange between the alveoli and small
pulmonary arteries and capillaries, worsening arterial
hypoxemia. Suppression of the development of pulmonary hypertension in the H rats during hypoxic exposure was associated with an increased PaO2 and an
attenuated polycythemic response without any change
in ventilation (8). In murine CMS, accentuated hypoxemia follows the development of pathologically elevated pulmonary arterial pressures.
Severe blood hyperviscosity associated with excessive polycythemia did not compromise renal blood flow
or renal O2 delivery during hypoxia in H rats. Contrary
to the belief that polycythemia might reduce renal
blood flow and O2 delivery (13), renal blood flow actually increased during hypoxic exposure as Hct rose and
blood viscosity increased (unpublished observations).
As a result of increased renal blood flow and increased
blood O2 content associated with the polycythemic
response, renal O2 delivery increased beyond the control levels in both rat strains after chronic hypoxia.
Nevertheless, renal tissue hypoxia and EPO gene expression persisted in H rats.
Role of negative-feedback control. The early finding
that a humoral erythropoiesis-stimulating factor (EPO)
increased rapidly on hypoxic exposure and declined
toward control levels led to the notion of a negativefeedback control mechanism regulating erythropoiesis.
According to this concept, hypoxia stimulates EPO
production and erythropoiesis; the resulting polycythemia increases the O2-carrying capacity and O2 content
of blood, and this improves tissue oxygenation and
turns off further production of EPO (14). The idea of
EPO-O2-carrying capacity feedback is inherent in all
EPO dose-response curves published: the Hct is plotted
as the independent variable (14, 23). Nevertheless,
EPO measurements using sensitive radioimmunoassay
for EPO and studies of EPO mRNA levels have made it
clear that a negative-feedback control mechanism does
not account for the regulation of the polycythemic
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
elevated renal tissue oxygenation, but the inhibition
must be sufficient to prevent reabsorption of at least
20% of the filtered load of sodium (10, 11). In the
present study, there were only minor changes in fractional sodium reabsorption and, in general, renal function was preserved without strain differences at sea
level or under hypoxic conditions. Despite stable excretory renal function under control and hypoxic conditions, hypoxic exposure elicited distinctly different
changes in EPO gene expression and EPO production
in the two rat strains. The dissociation between measures of renal function and EPO production under
hypoxic conditions renders unlikely any significant
regulatory role of renal function in the EPO response to
hypoxia.
Role of renal tissue hypoxia. Carbon monoxide and
anemia stimulate EPO gene expression and EPO production in the absence of changes in PaO2; therefore, the
renal hypoxia-sensing process is probably located near
a venous or tissue site (20). The hypoxia-sensing process probably occurs adjacent to the EPO-generating
sites in the kidney in adult mammals (22, 31). Because
PrvO2 approximates the average renal tissue O2 level
(37), PrvO2 is an estimate of the primary hypoxic signal determining EPO production. Renal O2 delivery
and renal V̇O2 did not differ in the two rat strains under
sea-level control or hypoxic conditions (Table 2). Systemic arterial hypoxemia was more severe in the H
than in the M rats, and the severity of hypoxemia
increased in the H rats, but not in the M rats, as the
hypoxic exposure was prolonged (Fig. 6) (7, 28). Not
surprisingly, PrvO2 was lower in the H than in the M
rats after 2 wk of hypoxic exposure (Table 2) (28). There
is no reason to believe that the EPO stimulus-response
characteristics differ between the rat strains during
chronic ($24 h) hypoxia: H rats simply experienced
greater hypoxic stimulation during equivalent altitude
exposure. Furthermore, there can be no effective polycythemic compensation for arterial hypoxemia in the H
rats: the arterial values were below the threshold of the
EPO response (Fig. 5). Thus more severe renal tissue
hypoxia can account for the inordinate EPO gene
expression and EPO production and, thereby, the excessive polycythemia in the H rats during chronic hypoxic
exposure.
The greatest EPO gene expression in each strain
occurred after 6 h of hypoxia. The magnitude of EPO
mRNA and plasma EPO at 6 h exceeded expression at
any other time within each strain, but the PaO2 values
at 6 h were not the most hypoxemic values. We have no
PrvO2 values at 6 h to correlate with the intense EPO
gene expression at 6 h. Therefore, it remains possible
that hypoxia-EPO stimulus-response characteristics
differ between the acute and chronic phases of the
hypoxic exposure. Nevertheless, H rats are more hypoxemic than M rats acutely (PaO2 5 44 6 1 and 46 6 1 Torr
at 6 h in H and M rats, respectively; unpublished
observations) and in the later stages of chronic hypoxia.
Thus the relationship between EPO expression and
hypoxia may differ acutely and chronically, but, in each
time period, H rats, in which EPO expression was
1249
1250
POLYCYTHEMIC RESPONSES IN HYPOXIA
tion or sinus nerve discharge or EPO production, rises
steeply. The mechanism whereby declining O2 levels
are sensed and transformed into ventilatory and EPO
responses is unknown, but it is conceivable that the
renal and carotid sensors share a common mechanism.
For example, both may rely on heme proteins (1, 17,
26).
In summary, the present study examined the genetic
regulatory mechanisms as well as the physiological
significance of the polycythemic response to environmental hypoxia in two rat strains with distinctly
different susceptibility to CMS. The results suggest
that the polycythemia of CMS derives from pulmonary
abnormalities of gas exchange, vascular remodeling,
and more severe arterial hypoxemia in the H rats
rather than different patterns of EPO synthesis and
release to a similar stimulus. EPO levels were appropriate for the levels of renal tissue hypoxia, but PaO2
and PrvO2 were below the EPO response threshold in
the H rats, and no polycythemic response can restore
renal tissue oxygenation in that setting.
This study was supported by National Heart, Lung, and Blood
Institute Grant HL-21159. J. C. Leiter is a recipient of a Clinical
Investigator Award from the American Lung Association.
Address for reprint requests: L. C. Ou, Dept. of Physiology,
Dartmouth Medical School, Borwell Bldg., 1 Medical Center Dr.,
Lebanon, NH 03756-0001.
Received 14 May 1997; accepted in final form 3 December 1997.
REFERENCES
1. Acker, H. PO2 chemoreception in arterial chemoreceptors. Annu.
Rev. Physiol. 51: 835–844, 1989.
2. Beru, N., J. McDonald, C. Lacombe, and E. Goldwasser.
Expression of the erythropoietin gene. Mol. Cell. Biol. 6: 2571–
2575, 1986.
3. Cahan, C., P. L. Hoekje, E. Goldwasser, M. J. Decker, and
K. P. Strohl. Assessing the characteristic between length of
hypoxic exposure and serum erythropoietin levels. Am. J. Physiol.
258 (Regulatory Integrative Comp. Physiol. 27): R1016–R1021,
1990.
4. Chomczynski, P., and N. Sacchi. Single-step method RNA
isolation by acid guanidinium thiocyanate-phenol-chloroform
extraction. Anal. Biochem. 162: 156–159, 1987.
5. Colice, G. L., N. Hill, Y.-J. Lee, H.-K. Du, J. Klinger, J. C.
Leiter, and L.-O. Ou. Exaggerated pulmonary hypertensive
response to monocrotaline in rats susceptible to chronic mountain sickness. J. Appl. Physiol. 83: 25–31, 1997.
6. Cruz, J. C., and S. Recavarren. Chronic mountain sickness: a
pulmonary vascular disease. In: Topics in Environmental Physiology and Medicine: High Altitude Physiology and Medicine,
edited by W. Brendel and R. A. Zink. New York: Springer-Verlag,
1982, p. 271–277.
7. Du, H. K., Y. J. Lee, G. L. Colice, J. C. Leiter, and L. C. Ou.
Pathophysiology of hemodilution in chronic mountain sickness.
J. Appl. Physiol. 80: 574–582, 1996.
8. Du, H. K., J. C. Leiter, and L. C. Ou. Effect of endothelin
antagonist BQ-123 on the development of severe pulmonary
hypertension in CMS in rats (Abstract). FASEB J. 8: A649, 1995.
9. Eckardt, K.-U., J. Dittmer, R. Neumann, C. Bauer, and A.
Kurtz. Decline of erythropoietin formation at continuous hypoxia is not due to feedback inhibition. Am. J. Physiol. 258 (Renal
Fluid Electrolyte Physiol. 27): F1432–F1437, 1990.
10. Eckardt, K.-U., A. Kurtz, and C. Bauer. Regulation of erythropoietin production is related to proximal tubular function. Am. J.
Physiol. 256 (Renal Fluid Electrolyte Physiol. 25): F924–F947,
1989.
11. Eckardt, K.-U., A. Kurtz, H. Scholz, and C. Bauer. Regulation
of erythropoietin formation in vivo. In: Erythropoietin: From
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
response to environmental hypoxia (3, 9). The high
levels of renal EPO mRNA and circulating EPO apparent at the onset of hypoxic exposure declined within the
first 24 h of hypoxia before any detectable polycythemia
or increased O2-carrying capacity developed in this and
other studies (9). The rapid decline in EPO mRNA
during continued hypoxic exposure in the kidney was
also unrelated to the rising circulating EPO levels;
there is no evidence of direct negative feedback in
which EPO might inhibit its own production (9). Furthermore, evidence obtained in the present study does
not support the concept of a negative-feedback mechanism, in which a polycythemic response to hypoxia
suppresses EPO production. Rather, the data in Fig. 5
indicate that feedback control of EPO gene expression
and EPO production operates in terms of PrvO2, and
polycythemia is only one of multiple factors modifying
PrvO2. The pattern of EPO production was biphasic. At
the onset of the hypoxic exposure, there was a sharp
increase in EPO production, but in ,24 h, EPO levels
fell substantially but remained above control values.
EPO levels rose slowly in the late stages of the hypoxic
exposure in the H rats. PrvO2 follows a similar pattern:
an initial sharp drop, a rapid recovery to a stable level
in M and H rats, and a slow decline after 2 wk of
continued hypoxia in H rats (Table 2) (28). The parallel
pattern of changes in PrvO2 and EPO levels during
hypoxia is consistent with feedback control of EPO synthesis and release as a function of PrvO2. Polycythemia
modifies PrvO2 only in M rats after 2 wk of hypoxia. The
early changes in PrvO2 reflect ventilatory acclimatization, acute cardiovascular responses to hypoxia, and
perhaps some benefit from the hemoconcentration after
diuresis in the first days of acclimatization to high
altitude (Table 2) (18). Only the late increase in PrvO2 in
M rats demonstrates the beneficial effect of polycythemia. In contrast, the PrvO2 remained low in the H
strain. This was not due to a lack of O2-carrying
capacity (Hct rose above 70%) or low renal blood flow
(renal COD was equivalent in the H and M rats and
exceeded the sea-level value by day 30 of the hypoxic
exposure) but to a low PaO2. Thus renal hypoxic stimulation persists when arterial hypoxemia is below the
threshold of the EPO response and, therefore, beyond
the capacity of polycythemia to restore renal tissue
oxygenation to suprathreshold levels. The accentuated
hypoxemia of the H rats, which originates in abnormal
pulmonary vascular responses to hypoxia and associated gas exchange abnormalities, creates a situation in
which polycythemia develops but has no significant
effect on the level of EPO stimulation. Consequently,
the signs of CMS develop: profound arterial hypoxemia,
persistent production of EPO, and extraordinary polycythemia.
The EPO-PrvO2 response curves (Fig. 5), whether in
terms of mRNA, renal EPO, or plasma EPO, have a
shape resembling the ventilatory response to hypoxia
and the pattern of carotid body discharge in response to
hypoxia (24): there is little response until a threshold is
reached, and below the threshold the response, ventila-
POLYCYTHEMIC RESPONSES IN HYPOXIA
12.
13.
14.
15.
16.
17.
18.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
two colonies of rats with differing sensitivity to chronic hypoxia.
Am. J. Pathol. 128: 61–66, 1987.
Mills, E., and F. F. Jöbsis. Mitochondrial respiratory chain of
carotid body and chemoreceptor response to changes in oxygen
tension. J. Neurophysiol. 35: 405–428, 1972.
Monge, C. M., and C. C. Monge. High-Altitude Disease.
Springfield, IL: Thomas, 1966.
Ou, L. C., J. Chen, E. Fiore, J. C. Leiter, T. Brinck-Johnsen,
G. F. Birchard, G. Clemons, and R. P. Smith. Ventilatory and
hematopoietic responses to chronic hypoxia in two rat strains. J.
Appl. Physiol. 72: 2354–2363, 1992.
Ou, L. C., J. Silverstein, and B. R. Edwards. Renal function
in rats chronically exposed to high altitude. Am. J. Physiol. 247
(Renal Fluid Electrolyte Physiol. 16): F45–F49, 1984.
Ou, L. C., and R. P. Smith. Probable strain differences of rats in
susceptibilities and cardiopulmonary responses to chronic hypoxia. Respir. Physiol. 53: 367–377, 1983.
Pagel, H., W. Jelkmann, and C. Weiss. O2 supply to the
kidneys and the production of erythropoietin. Respir. Physiol. 77:
111–118, 1989.
Schuster, S. J., E. V. Badiavas, P. Costa-Giomi, R. Weinmann, A. J. Erslev, and J. Caro. Stimulation of erythropoietin
gene transcription during hypoxia and cobalt exposure. Blood 73:
13–16, 1989.
Schuster, S. J., S. T. Koury, M. Bohrer, S. Salceda, and J.
Caro. Cellular sites of extrarenal and renal erythropoietin
production. Br. J. Haematol. 81: 153–159, 1992.
Schuster, S. J., J. H. Wilson, A. J. Erslev, and J. Caro.
Physiologic regulation and tissue localization of renal erythropoietin messenger RNA. Blood 70: 316–318, 1987.
Smith, H. W., N. Finkelstein, L. Aliminosa, B. Crawford,
and M. Graber. The renal clearances of substituted hippuric
acid derivatives and other aromatic acids in dogs and man. J.
Clin. Invest. 24: 388–404, 1945.
Tan, C. C., K.-U. Eckardt, and P. J. Ratcliffe. Organ distribution of erythropoietin messenger RNA in normal and uremic rats.
Kidney Int. 40: 69–76, 1991.
Tenney, S. M. A theoretical analysis of the relationship between
venous blood and mean tissue oxygen pressures. Respir. Physiol.
20: 283–296, 1974.
Valtin, H. Renal Function. Boston, MA: Little, Brown, 1983.
West, J. B. Pulmonary Pathophysiology—The Essentials. Baltimore, MD: Willliams & Wilkins, 1982.
Winslow, R. M., and C. Monge. Hypoxia, Polycythemia, and
Chronic Mountain Sickness. Baltimore, MD: Johns Hopkins
University Press, 1987.
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
19.
Molecular Structure to Clinical Application, edited by C. A.
Baldamus, P. Scigalla, L. Wieczorek, and K. M. Koch. Basel:
Karger, 1989, p. 33–38.
Erslev, A. J. Production of erythrocytes. In: Hematology (2nd
ed.), edited by W. J. Williams, E. Beutler, A. J. Erslev, and R. W.
Rundles. New York: McGraw-Hill, 1977, p. 203–216.
Erslev, A. J., J. Caro, and A. Besarab. Why the kidney?
Nephron 41: 213–216, 1985.
Erslev, A. J., J. Caro, O. Miller, and R. Silver. Plasma
erythropoietin in health and disease. Ann. Clin. Lab. Sci. 10:
250–257, 1980.
Führ, J., J. Kaczmarczyk, and C. D. Krüttgen. Eine einfache
colormetrische Methode zur Inulinbestimmung für Nierenclearanceuntersuchungen bei Stoffwechselgesunden un Diabetikern.
Klin. Wochenschr. 33: 729–730, 1955.
Gilman, M. Ribonuclease protection assay. In: Current Protocols
in Molecular Biology, edited by F. M. Ausubel, R. Brent, R. E.
Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Strahl.
New York: Wiley, 1989, p. 4.7–4.7.8.
Goldberg, M. A., S. P. Dunning, and H. F. Bunn. Regulation of
erythropoietin gene: evidence that the oxygen sensor is a heme
protein. Science 242: 1412–1415, 1988.
Honig, A. Peripheral arterial chemoreceptors and reflex control
of sodium and water homeostasis. Am. J. Physiol. 257 (Regulatory Integrative Comp. Physiol. 26): R1282–R1302, 1989.
Jacobson, L. O., E. Goldwasser, W. Fried, and L. Plzak. Role
of the kidney in erythrocytosis. Nature 179: 633–634, 1957.
Jelkmann, W. Renal erythropoietin: properties and production.
Rev. Physiol. Biochem. Pharmacol. 104: 139–207, 1986.
Jones, S. E., B. G. Malgraith, and H. H. Sculthorpe. Pathological process in disease. II. Blood of the albino rat: approximate
physico-chemical description. Ann. Trop. Med. Parasitol. 44:
168–186, 1950.
Kurtz, A., and K. U. Eckardt. Renal function and oxygen
sensing. In: Erythropoietin: Molecular, Cellular, and Clinical
Biology, edited by A. J. Erslev, J. W. Adamson, J. W. Eschbach,
and C. G. Winearl. Baltimore, MD: Johns Hopkins University
Press, 1991, p. 79–98.
Kurtz, A., K.-U. Eckardt, L. Tannahill, and C. Bauer. Regulation of erythropoietin production. Contrib. Nephrol. 66: 1–16,
1988.
Lahiri, S., and R. G. DeLaney. Relationship between carotid
chemoreceptor activity and ventilation in the cat. Respir. Physiol.
24: 267–286, 1975.
Langleben, D., R. C. Jones, M. J. Aronovitz, N. S. Hill, L.-C.
Ou, and L. M. Reid. Pulmonary artery structural changes in
1251