Vol. 87, Issue 5, 1901-1908, November 1999
Role of the spleen in the exaggerated polycythemic response to
hypoxia in chronic mountain sickness in rats
H. Y.
Kam1,
L. C.
Ou1,
C. D.
Thron2,
R. P.
Smith2, and
J. C.
Leiter1,3
Departments of 1 Physiology,
2 Pharmacology and Toxicology, and
3 Medicine, Dartmouth Medical
School, Lebanon, New Hampshire 03756-001
 |
ABSTRACT |
In a rat model of chronic mountain sickness, the
excessive polycythemic response to hypoxic exposure is associated with
profound splenic erythropoiesis. We studied the uptake and distribution of radioactive iron and red blood cell (RBC) morphology in intact and
splenectomized rats over a 30-day hypoxic exposure.
Retention of 59Fe in the plasma
was correlated with 59Fe uptake by
both spleen and marrow and the appearance of
59Fe-labeled RBCs in the
blood. 59Fe uptake in
both the spleen and the marrow paralleled the production of nucleated
RBCs. Splenic 59Fe uptake was
~10% of the total marrow uptake under normoxic conditions but
increased to 60% of the total marrow uptake during hypoxic exposure.
Peak splenic 59Fe uptake and
splenomegaly occurred at the most intense phase of erythropoiesis and
coincided with the rapid appearance of
59Fe-labeled RBCs in the blood.
The bone marrow remains the most important erythropoietic organ under
both resting and stimulated states, but inordinate splenic
erythropoiesis in this rat strain accounts in large measure for the
excessive polycythemia during the development of chronic mountain
sickness in chronic hypoxia.
interaction of splenic and medullary erythropoiesis; high altitude; splenomegaly; splenectomy; leukocytosis; thrombocytosis; thrombocytopenia
| |
INTRODUCTION |
WE HAVE CHARACTERIZED a rat model of human chronic
mountain sickness (CMS). As in the human disease, the rats manifest
excessive polycythemia and hypervolemia (16). In our investigations, we studied two strains of Sprague-Dawley rats: the Hilltop strain is
hypoxia sensitive and develops CMS, whereas the Madison strain is
relatively resistant to hypoxia and does not develop CMS. Splenic erythropoiesis contributes significantly to the overall polycythemic response to hypoxia in the Hilltop strain (17, 18). In contrast, the
polycythemic response to the identical hypoxic conditions is only
moderate in the Madison strain, and splenic erythropoiesis plays a
minimal role (17).
Splenic erythropoiesis has been well documented in rodents,
particularly in mice (5). The prominence of splenic erythropoiesis in
adult rodents during erythropoietic stimulation was attributed to the
limited marrow space, which might restrict the production of red blood
cells (RBCs) (3). Thus splenectomy reduced the polycythemic response
after bleeding or led to death when medullary erythropoiesis was
chemically ablated (23). On the other hand, a normal and complete
polycythemic response to chronic hypoxia was found in splenectomized
mice (13) and rats (24), suggesting that splenic erythropoiesis serves
only an accessory function. Hence, there is an intricate relationship
between splenic and medullary erythropoiesis that remains to be
investigated. The goal of the present study was to determine the
relative contributions of splenic and medullary erythropoiesis in the
development of the excessive polycythemic response to chronic hypoxia
in the CMS-susceptible, Hilltop rat strain. The temporal patterns of splenic and medullary erythropoiesis were followed in intact and splenectomized animals over a period of 30 days of hypoxic exposure by
means of radioactive-iron uptake and morphological analyses of the
marrow and spleen. The effect of splenectomy and hypoxia on other
formed elements of the blood during hypoxic exposure was also examined.
| |
MATERIALS AND METHODS |
Radioactive-iron uptake studies.
Adult male Sprague-Dawley rats weighing 270-290 g were purchased
from the Hilltop Breeding Laboratories (Scottsdale, PA). Six separate
groups of rats were exposed to a simulated altitude of 18,000 ft (5,500 m or an inspired PO2 = 73 Torr) for 0, 1, 3, 5, 10, or 30 days in a large environmental chamber (24). Each
group of 10 animals consisted of 2 subgroups with equal numbers of
intact and splenectomized rats. Splenectomy was performed while the
rats were under ether anesthesia. After surgery, the rats were treated
with penicillin (100,000 U daily im) for 5 days, although exposure to
high altitude began after a 2- to 3-day postsurgical recovery period.
In the intact group, one or two animals received sham operations for
each time subgroup. No significant effect of the sham operation was
observed in any of the measured variables, and the data from animals
with and without sham operations were pooled to form the intact subgroup.
After altitude exposure, the animals were given 15 µCi
59Fe (29 µCi/mg) via the jugular
vein under ether anesthesia and returned to room air [sea-level
(SL) controls] or to the hypoxic environment (experimental
groups). Sixty minutes after isotope injection, the animals were killed
while under ether anesthesia by exsanguination via the vena cava. The
half-life of 59Fe incorporation is
40 min in rats (22). A sampling time of 60 min allowed adequate
59Fe incorporation and was
consistent with previous studies. Radioactivity was assayed in tissue
samples (~1 g) from spleen, liver, femur, whole blood, and plasma
(Beckman Gamma Counter, Irvine, CA). Radioactivity was expressed as
counts per minute (cpm) per unit volume in the case of whole blood,
plasma, and RBCs. Radioactive counts were consistently >10,000
cpm/sample, and background activity was low. For the organ-uptake data,
radioactivity was expressed per whole spleen or femur per 100 g body
weight. The marrow distribution of
59Fe radioactivity has been
studied in rats (9), and the ratio of femur
59Fe uptake to total marrow
59Fe uptake remained constant
under a variety of conditions, although the effect of hypoxia was not
studied. Each femur represented 9.6% of the total
59Fe uptake; therefore, the total
marrow radioactivity in each animal was calculated as the product of
the radioactivity of 1 femur multiplied by 10.6 (9). The total marrow
radioactivity in each animal was corrected for differences in body
weight and expressed as total medullary radioactivity per 100 g body
weight. Normalization by body weight was used to correct for
treatment-related body weight differences and to facilitate comparison
with past work. Identical results were obtained with and without body
weight normalization.
Tissue radioactivity that was due to the residual blood
59Fe was determined by using
51Cr as described previously (20),
and the net tissue uptake of 59Fe
was calculated after subtraction of radioactivity attributable to
residual blood in tissue.
Morphological studies.
At the end of the exposure period, animals were killed under ether
anesthesia by exsanguination via the vena cava. Blood samples were
collected in EDTA-treated tubes, and complete cell and platelet counts
were made in a Coulter s-plus IV (Coulter Electronics, Hialeah, FL).
Blood smears were prepared for differential cell counts. Reticulocytes
were counted by using new methylene blue.
The spleen was removed intact, weighed, fixed in 10% Formalin (4%
formaldehyde gas dissolved in water), embedded in paraffin, sectioned
at 6 mm, and stained with hematoxylin and eosin. Slides containing
these sections were labeled with a code number to blind the evaluator
to the treatment. Hematopoietic activity in sections of the spleen was
quantified by measuring the relative numbers of nucleated RBCs (NRBCs).
Slides of splenic sections were ranked according to their
erythropoietic activity in side-by-side comparisons, and "ties"
were not included. The rank orders were used in two statistical tests:
Wilcoxon's one-way comparison of all possible pairs of treatments and
Wilcoxon's one-way comparison of several treatments with a control
(26).
The femur was cleaned of surrounding tissue, and bone mass was
determined before and after demedullation. The weight of the total
femoral marrow was calculated from the difference in bone mass before
and after demedullation. For demedullation, the femur was split along
its main axis, and ~20-µg samples of bone marrow were taken; the
remainder was removed by swabbing the bone cavity clean with lens
paper. The samples of bone marrow were drawn into a 100-µl disposable
pipette. The mass of this marrow was determined by weighing the pipette
before and after drawing in the marrow sample. The bone marrow was
expelled from the pipette onto the center of a small piece of 100-mesh
stainless steel sheet set in a petri dish containing 1 ml of bovine
serum. The sample was quantitatively transferred onto the sheet by
repeatedly drawing serum in and out of the pipette. The stainless steel
mesh sheet was folded up with the marrow sample inside, and a cell
suspension was made by quickly applying pressure to the screen. This
technique of preparing cell suspensions minimized damage to the cells.
Bovine serum was used instead of homologous serum because of the
limited amount of serum available in the high-altitude rats. Comparison of the cell suspensions obtained from both the homologous serum and
bovine serum revealed no major differences in preservation of the
morphological integrity of the cells. The percentage of damaged cells
ranged from 5 to 15% in both cases. NRBC counts of the SL, control
rats obtained by this procedure were similar to those reported by
Fruhman and Gordon (6) and Donohue et al. (4). Total cell counts and
NRBC counts were made in a hematocytometer (Improved Neubauer,
Philadelphia, PA). Marrow smears were made from the other femur. Marrow
smears were examined for numbers of NRBCs and total RBCs and for other
cell types. The absolute numbers of each cell type in one femur were
calculated from the percentage of various cell types in the smears and
the total cell numbers in the whole femur and were expressed per 100 g
body weight.
Statistics.
Comparisons between measurements in the intact and splenectomized
animals at various durations of high-altitude exposure was performed by
using a two-way analysis of variance. Comparisons were reported as
statistically significant when the level of probability for rejecting
the null hypothesis was 
0.05. Data were reported as means ± SD
and P values as 0.05 or 0.01.
| |
RESULTS |
Body weight was significantly less (P < 0.001) in the splenectomized animals compared with intact animals
when considered across all study times. However, the profile of body
weight as a function of time (an initial drop in body weight at the
onset of hypoxic exposure with a nadir on day
3 followed by steady growth and recovery of body
weight) was similar in both treatment groups. There was no significant
interaction between treatment group and altitude-exposure time.
Kinetics of 59Fe uptake during hypoxia.
The temporal patterns of the changes in
59Fe activity of whole blood,
plasma and packed RBCs 1 h after
59Fe injection are shown in Fig.
1. After a latency of 24 h, the radioactivity of the whole blood increased slightly, but significantly, at 72 and 120 h of hypoxia and then returned to normal control values
(Fig. 1A). Retention of injected
59Fe in the plasma dropped
precipitously after 72 h of hypoxia and remained low throughout the
entire period of exposure (Fig. 1B). The 59Fe retained in the plasma in
intact animals was consistently lower than in splenectomized animals,
indicating that removal of the injected
59Fe was more rapid in the intact
group. The 59Fe uptake of the
packed RBCs measured 1 h after injection increased greatly after 72 h
of hypoxia, then declined somewhat, but remained significantly elevated
above the SL control values (Fig.
1C). More rapid removal of injected
59Fe from the plasma was
associated with a more rapid increase in the RBC
59Fe in intact animals. In
contrast, 59Fe radioactivity
persisted longer in the plasma of the splenectomized rats, and RBC
59Fe rose more slowly and to a
lesser extent.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 1.
Changes in 59Fe distribution in
whole blood (A), plasma
(B), and packed red blood cells
(RBCs) (C) in intact ( ) and
splenectomized (Splenx;  ) rats during exposure to a simulated
altitude of 5,500 m. Values are means ± SD
from at least 5 animals. HA, high altitude; SL, sea level; cpm,
counts/min. * P 0.05 compared
with SL controls of either intact or splenectomized rats.
 P 0.05 in
comparisons between intact and splenectomized rats at the same
time point.
|
|
The net total tissue 59Fe uptake
for the marrow, spleen, and liver are summarized in Fig.
2. The total marrow
59Fe uptake 1 h after injection in
splenectomized animals (Fig. 2A)
exceeded total marrow 59Fe uptake
in intact, control animals at SL. Subsequently, there was a dip in
59Fe uptake at 24 h, after which
marrow activity increased ~40% over the SL control values.
Furthermore, 59Fe uptake in
splenectomized animals remained significantly higher than in the SL
controls even after 30 days of hypoxia. In the intact animals (Fig.
2A), marrow
59Fe uptake, measured 1 h after
injection, increased immediately, reached a maximum uptake (50%
increase over the SL control level) by the third day, and then
gradually decreased toward the control values. Except for the first 24 h of hypoxia, the mean marrow 59Fe
uptake values were consistently higher in the splenectomized than in
the intact animals. Most strikingly, hypoxic exposure greatly
stimulated splenic 59Fe uptake in
the intact animals (Fig. 2B). The
59Fe uptake increased ~2-fold
after 24 h of hypoxia, reached a value 10-fold greater than the SL
control level at the end of 72 h, and declined slowly to ~5-fold the
control value after 30 days. Under SL conditions, splenic
59Fe uptake was ~10% (49 × 103 vs. 545 × 103 cpm/100 g body wt) of the
total marrow uptake, but it increased to ~60% (500 × 103 vs. 750 × 103 cpm/100 g body wt) of the
total marrow uptake after hypoxic exposure. There was a hypoxia-related
decrease in 59Fe uptake, measured
1 h after injection, in the liver that was greater in the intact rats
compared with the splenectomized animals (Fig.
2C).
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 2.
59Fe uptake in various organs in
intact () and splenectomized () rats during exposure to a
simulated altitude of 5,500 m. A:
marrow. B: spleen.
C: liver. Values are means ± SD
from at least 5 animals. BW, body wt.
* P 0.05 compared with SL
controls.
|
|
We assume that the marrow and the spleen represent the primary
erythropoietic organs, and 59Fe
uptake by these organs relates directly to erythropoietic activity. Hence, the total marrow 59Fe
uptake represents the total erythropoietic activity in the splenectomized animals, and the combined
59Fe uptake of the spleen and
marrow represents the total erythropoietic activity in the intact
animals. The changes in the total erythropoietic activity with time of
hypoxic exposure in intact and splenectomized animals are shown in Fig.
3. The temporal patterns of total
erythropoietic activity were similar to those of the packed-RBC
59Fe activity on the one hand
(Fig. 1C), but mirror images of the 59Fe retained in the plasma on the
other hand (Fig. 1B). The uptake and
disposition of exogenous 59Fe in
plasma are best revealed in Fig. 4. The
concentration of injected 59Fe
retained in the plasma was inversely related to the total
erythropoietic organ 59Fe uptake
(Fig. 4A), but the appearance of the
59Fe-labeled RBCs in the blood was
linearly related to the total erythropoietic organ
59Fe uptake (Fig.
4B).
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 3.
Changes in total 59Fe uptake of
erythropoietic organs in intact (marrow and spleen) and in
splenectomized (marrow only) rats during exposure to a simulated
altitude of 5,500 m. Values are means ± SD from at least
5 animals. * P 0.05 compared
with SL controls within each treatment group.
P 0.05 in comparisons
between intact and splenectomized rats at same time point.
|
|
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 4.
Correlation between 59Fe retained
in plasma and total erythropoietic organ
59Fe uptake
(A) and relationship between
circulating packed-RBC 59Fe and
total 59Fe uptake of
erythropoietic organs (B) in intact
() and splenectomized () rats. Correlation coefficient
(r2) between
59Fe retained in plasma and total
erythropoietic 59Fe uptake was
0.834 (P < 0.01), and
r2 between
packed-RBC 59Fe and erythropoietic
organ 59Fe uptake was 0.891 (P < 0.01). Mean values of total
59Fe uptake of erythropoietic
organs of intact (marrow and spleen) and splenectomized (marrow only)
animals in Fig. 3 and mean values of the packed-RBC
59Fe in Fig. 1 were used to
generate this plot, and curves were fitted by linear regression.
|
|
In the intact animals (Fig. 5), splenic
erythropoietic activity was linearly correlated with the packed-RBC
59Fe activity. The comparable
relationship for the medullary component, however, was not linear; a
second-order polynomial fit the points significantly better than did a
straight line. The initial component of the marrow relationship
paralleled the splenic relationship; marrow activity increased as
the packed-RBC 59Fe activity
increased, but, at the most intense phase of erythropoiesis, the
increased 59Fe content of packed
RBCs was not reflected by increased
59Fe uptake by the marrow. The
splenic erythropoietic activity correlated more closely with the
release of young RBCs into the circulation at this time. The
two-component marrow response occurred, however, only in intact
animals. Medullary erythropoietic activity in splenectomized animals
was linearly correlated with the packed-RBC
59Fe activity. This observation
suggests that spleen was the major erythropoietic organ providing new
RBCs during the early phase of intense hypoxia-stimulated
erythropoiesis.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 5.
Relationship between organ-specific
59Fe uptake of spleen and bone
marrow and production of young RBCs in intact ( and  ) and
splenectomized () animals. Mean values of Fig. 1 (packed-RBC
59Fe) and Fig. 2 (marrow and
splenic 59Fe uptake) were used to
generate this plot. Lines for splenectomized animals and splenic
59Fe uptake were fitted by linear
regression (r2 = 0.910 and 0.982, respectively). Marrow response in intact animals was
fitted by using a second-order polynomial
(r2 = 0.822).
|
|
Morphological changes in the spleen and marrow during hypoxia.
Figure 6 depicts the morphological changes
in the spleen as a function of the time of hypoxic exposure. The spleen
became significantly hypertrophied after 72 h of hypoxia and remained so for the rest of the exposure period except at the 10th day of
hypoxia, when the splenic weight did not differ significantly from the
SL control value. Increased numbers of NRBCs in splenic sections
demonstrated that splenic erythrocytic activity was also significantly
increased after 72 h of hypoxia. This increase was sustained throughout
the subsequent course of the hypoxic exposure. The histological
evaluation was based on the NRBC population density, and the twofold
increase in splenic volume further augmented the total NRBC population.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 6.
Changes in total splenic mass (A)
and relative number of nucleated RBCs (NRBC;
B) of spleen during exposure to a
simulated altitude of 5,500 m. In A,
values are means ± SD from at least 5 animals, and
* P 0.05 compared with SL
control values. Effect of hypoxic exposure on NRBC density is plotted
in B. There were 30 samples (6 study
times, each with 5 rats), and NRBC density was ranked from 1 (no NRBC)
to 30 (maximal NRBC density) within this population of 30 rats. Ranked
score for each animal (ordinate) is plotted as a function of duration
of altitude exposure (abscissa).
|
|
Figure 7 summarizes the morphological
changes in the marrow. On exposure to hypoxia, the NRBC population of
both the intact and splenectomized rats increased rapidly and reached
maximal and constant levels after 72 h of hypoxia (Fig.
7A). The increase in NRBC
populations in both intact and splenectomized rats was more than twice
the SL control values. Although the mean increase in the NRBCs was
consistently higher in the splenectomized than in the intact animals,
the differences were not statistically significant. Total bone marrow
cellularity in intact and splenectomized animals followed strikingly
different patterns (Fig. 7B). In the intact rats, the total cell counts decreased after 24 h of hypoxia but
then increased linearly up to twice the control value. In contrast, the
total cell counts in splenectomized animals increased immediately and
reached a peak after 5 days of hypoxia. Thereafter, they declined
slightly. Thus the total cell counts in the splenectomized animals were
higher in the early phase and lower in the later phase of hypoxic
exposure compared with intact animals. The mature RBC count in the
marrow of intact animals dipped initially (Fig. 7C) but increased steadily to a
level significantly above both the SL controls and marrow RBC counts in
splenectomized animals at 30 days of exposure. There were no
significant changes in the mature RBC counts in splenectomized animals
over the period of hypoxic exposure.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 7.
Changes in NRBC number (A), total
cellularity (B), and mature RBCs
(C) in femur in intact () and
splenectomized () rats during exposure to hypoxia. Values are means ± SD from at least 5 animals.
* P 0.05 compared with SL
controls within each treatment group.
P 0.05 in comparisons
between intact and splenectomized rats at same time point.
|
|
Changes in circulating RBCs during hypoxia.
The whole blood hemoglobin concentration (Fig.
8A) and
the hematocrit (Fig. 8B) increased
over the time of hypoxic exposure in both intact and splenectomized
rats as expected. In most cases, the mean values for these variables
were significantly higher in the intact than in the splenectomized
animals. The circulating reticulocyte numbers (Fig.
8C), after a 24-h latency, increased rapidly and reached peak values in both groups of animals after 120 h
of hypoxia. The reticulocyte count rapidly declined in the intact
animals, but the decline was delayed in the splenectomized animals, and
reticulocyte counts were significantly higher in splenectomized
compared with intact rats after 10 days of hypoxia.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 8.
Changes in hemoglobin concentration
(A), hematocrit (Hct;
B), and reticulocyte number
(C) in blood in intact () and
splenectomized () rats during exposure to hypoxia. Values are means ± SD from at least 5 animals.
* P 0.05 compared with SL
controls within each treatment group.
P 0.05 in comparisons
between intact and splenectomized rats at same time point.
|
|
The mean corpuscular volume increased by ~15% after 72 h of hypoxia
and remained elevated throughout the entire period of exposure in
intact and splenectomized animals. The
mean corpuscular hemoglobin content increased along with the increase in cell size so that the mean corpuscular hemoglobin concentration remained unchanged during the hypoxic exposure. The temporal pattern of
changes in mean corpuscular volume and mean corpuscular hemoglobin content were similar in the intact and splenectomized rats (data not
shown). Electrophoretic analysis of the hemoglobin species in
circulating RBC showed no differences between the intact and splenectomized animals.
Changes in other blood cells during hypoxia.
As shown in Fig.
9A,
hypoxic exposure resulted in differing early responses in marrow white
blood cell (WBC) counts between the intact and splenectomized animals.
WBC counts decreased in intact rats, but increased in the
splenectomized animals. The transient decrease in marrow WBC counts in
the intact animals was primarily due to a fall in lymphocytes (data not
shown). The increase in marrow WBC in the splenectomized animals,
however, was due to increased lymphocytes and polymorphonucleocytes
(data not shown). Both circulating WBC (Fig.
9B) and platelet counts (Fig.
9C) increased during the first few
days of hypoxic exposure. Thereafter, WBC counts declined toward, and
platelet counts fell below, the SL control values. Splenectomy
significantly increased both the circulating WBC and platelet counts
under SL normoxic conditions and augmented the hypoxia-induced
transient leukocytosis and thrombocytosis.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 9.
Changes in marrow white blood cell (WBC) population
(A), circulating WBC population
(B), and platelet
(C) counts in intact () and
splenectomized () animals during exposure to a simulated altitude of
5,500 m. Values are means ± SD from at least 5 animals.
* P 0.05 compared with SL
controls within each treatment group.
P 0.05 in comparisons
between intact and splenectomized rats at same time point.
|
|
| |
DISCUSSION |
Flow of exogenous 59Fe in rats during
hypoxic exposure.
The present study demonstrates that
59Fe was rapidly used by the
erythropoietic organs for the synthesis of hemoglobin and returned to
the circulation in the form of newly released young RBCs. The evidence
for this is that, first, the rapid disappearance of plasma 59Fe was closely associated with a
rapid increase in 59Fe uptake by
the erythropoietic organs (marrow and spleen; Figs. 1, 2, and 4). In
fact, the retention of plasma 59Fe
was linearly and inversely related to
59Fe uptake by the erythropoietic
organs (Fig. 4). Second, the appearance of newly formed RBCs in the
circulation correlated well with either the total
59Fe uptake of the erythropoietic
organs (bone marrow and spleen) in the intact animals or with the
marrow 59Fe uptake in the
splenectomized animals (Fig. 5).
Relative roles of splenic and medullary erythropoiesis in the
polycythemia response to hypoxia.
Under normoxic conditions the total splenic
59Fe uptake represents only 10%
of the total marrow uptake (49 × 103 vs. 550 × 103 cpm/100 g body wt). Hypoxic
exposure increased the splenic uptake by >10-fold, but increased the
medullary uptake by <50%. The maximum increase in
59Fe uptake occurred after 3 days
of hypoxia in both the spleen and marrow. Because the basal uptake in
the marrow was much higher than in the spleen, the absolute
59Fe uptake remained significantly
higher in the marrow than in the spleen during the entire period of
hypoxic exposure. Consistent with these ferrokinetic findings, the
studies of splenic and marrow morphology showed that nucleated RBC
numbers increase markedly at the onset of hypoxic exposure, reached a
peak after 3 days of hypoxia, and remained elevated in both the spleen
and the bone marrow throughout the remainder of the 30-day hypoxic
exposure. Thus the present study demonstrated that hypoxic exposure
rapidly and simultaneously stimulated both splenic and medullary
erythropoiesis. As expected, the polycythemic response to hypoxia was
consistently greater in the intact compared with the splenectomized
animals (Fig. 8). As mentioned earlier, Hilltop rats develop CMS, which is characterized by excessive polycythemia and profound splenic erythropoiesis. In contrast, Madison rats are resistant to CMS and
develop only moderate polycythemia and minor splenic erythropoiesis (17). These previous observations suggest that the inordinate splenic
erythropoiesis seen in the Hilltop rats during chronic hypoxia may
account for the excessive polycythemic response seen in this rat
strain, and the finding in the present work that splenectomy significantly reduced the polycythemic response to chronic hypoxia appears to support this conclusion. The inordinate splenic
erythropoiesis in the Hilltop strain may originate from enhanced gene
expression and production of erythropoietin (EPO) during chronic
hypoxic exposure that was not seen in Madison rats (19). In the absence of the spleen, which is a strong competitor for EPO, circulating EPO
levels may be higher in splenectomized animals compared with intact
animals. Indeed, the medullary erythropoiesis, evident by the medullary
59Fe uptake and NRBC population,
increased in splenectomized animals. However, the polycythemic response
to hypoxia was lower in the splenectomized animals. Madison rats do
develop mild polycythemia and splenic erythrocytosis. But the total
splenic erythropoietic activity was fivefold greater in Hilltop rats
compared with Madison rats. Inordinate levels of circulating EPO alone
cannot account for the excessive polycythemia seen in the Hilltop rat
strain; the spleen is necessary as an additional erythropoietic site
that facilitates full expression of the polycythemic response to
hypoxia. This conclusion is consistent with the notion of a limited
marrow space in rodents (3).
Interaction between splenic and medullary erythropoiesis in
chronic hypoxia.
In the murine model of CMS, both splenic and medullary erythropoiesis
were stimulated concomitantly during hypoxic exposure in these rats (as
shown by the radioactive-iron uptake and the increased population of
NRBCs). There is less evidence of splenic erythropoiesis in adult
humans with CMS. Nonetheless, the relative contribution of the spleen
and marrow to the increased circulating RBC count remains an issue,
particularly the mechanism(s) of coordination between these two sites
of erythropoiesis. During early hypoxia, the rapid release of young
RBCs into the circulation correlated more with the increase in
59Fe uptake in the spleen than
with 59Fe uptake in the marrow
(Figs. 1, 2, and 5). The spleen may be the primary immediate source of
new RBCs on exposure to hypoxia. A similar correlation between the
rapid appearance of new RBCs in the circulation and medullary
59Fe uptake was established in the
splenectomized animals (note the surge in
59Fe uptake in the marrow; Fig.
2). Augmented 59Fe uptake by the
marrow in splenectomized animals could result simply from the absence
of competition for the available
59Fe or from an inhibitory
influence of the spleen over medullary erythropoiesis. Furthermore,
medullary 59Fe uptake increased
steadily during hypoxic exposure in the intact animals, but fell below
the control level on the first day of hypoxia in the splenectomized
animals (Figs. 2 and 3). The conspicuous fall in medullary
59Fe uptake in splenectomized
animals was associated with a parallel fall in the number of the
labeled RBCs in the circulation (Fig. 1). This observation seems to
indicate a depletion of some precursor cells that normally take up
iron, become mature RBCs, and are released into the circulation in
response to erythropoietic stimulation, as was seen in the intact
animals. The apparent depletion of this particular stage of the
precursor cells in splenectomized animals may be the consequence of
accelerated maturation of these precursors in the marrow in the absence
of the spleen. Finally, the number of mature RBCs in the marrow, after
a slight fall, accumulated during the time of hypoxic exposure in the
intact animals, whereas counts of mature RBCs in the splenectomized
animals remained constant (Fig. 7). Because hypoxic exposure increased
the NRBC population comparably in both the intact and splenectomized
animals, the gradual accumulation of mature RBCs with time of exposure
in the intact animals probably reflects decreased release of mature
RBCs in the presence of the spleen. The cause of the accelerated
maturation of the precursor cells and the increased release of mature
RBCs from the marrow in the absence of the spleen is not known.
However, EPO is known to accelerate both the maturation and release of the RBCs from the marrow (8), and one of the characteristic responses
of Hilltop rats to hypoxia is inordinate expression of the EPO gene,
leading to exaggerated production and sustained high titers of EPO
(19). It is conceivable that elevated EPO might influence both the
marrow and spleen to stimulate erythropoietic activity. In addition, it
has long been known that the spleen can influence various aspects of
hematopoiesis, including the erythropoiesis in the marrow (5, 11, 21),
and, therefore, some other unknown mechanisms cannot be excluded.
Another characteristic of the polycythemic responses to hypoxia of both
the intact and splenectomized animals was a sustained macrocytosis
associated with a higher mean corpuscular hemoglobin content after 24 h
of hypoxic exposure. Because there was also a sustained reticulocytosis
under these conditions in both groups of animals (Fig. 8), and
reticulocytes have a greater average diameter than the average
circulating RBC (8, 22), the persistent reticulocytosis may account for
the sustained macrocytosis. In addition, reticulocytes released under
severe stress (such as high EPO titers or severe anemia) are
considerably larger than ordinary reticulocytes, and these larger
reticulocytes will also have greater hemoglobin content (22). Thus the
abnormal and persistently elevated level of EPO seen in the Hilltop
strain of rats during hypoxic exposure (17, 19) may contribute to the
morphological features of the hematologic responses to hypoxia in this
particular rat strain.
The present observations of normal circulating WBC levels and the
development of thrombocytopenia after prolonged hypoxia in intact
animals corroborated our early findings (17). Splenectomy appeared not
to affect this chronic hypoxic effect. The transient leukocytosis and
thrombocytosis during the acute phase of the hypoxic exposure are also
consistent with observations in humans and animals (1, 25). The causes
of these changes remain unclear. Blood volume contracts during acute
hypoxia (25) and expands during chronic hypoxia (16). These blood
volume changes could conceivably affect the blood cell counts. The life
span of WBC and platelets remains unchanged under hypoxic conditions
(1, 14). The transient circulating leukocytosis coexisted with a decrease in WBC population in the marrow, suggesting that increased release of marrow leukocytes contributed to the circulating
leukocytosis in intact animals. The rate of thrombocyte production
increases in acute hypoxia and decreases during chronic hypoxia (2). Thus it seems likely that the transient early thrombocytosis during hypoxia resulted from increased production, and the thrombocytopenia during chronic hypoxia resulted from depressed platelet production.
Splenectomy augmented both the hypoxia-induced transient leukocytosis
and thrombocytosis. Because there was a conspicuous transient rise in
the marrow WBC population after splenectomy (Fig. 9), increased
production and release are likely involved, at least for the WBC.
Splenectomy induces prolonged leukocytosis and thrombocytosis under SL
conditions (7, 11, 15), whereas leukocytosis and thrombocytosis were
short lived under hypoxic conditions. All blood cells are derived from
multipotent stem cells (12). Competition among different cell lines
occurs under conditions when a particular cell line is specifically
stimulated (2, 10). Thus, when erythropoiesis is stimulated under
hypoxic conditions, the common stem cells are directed toward
production of RBCs, and fewer stem cells are available to support
production of other cells types. This may explain the short-lived
leukocytosis and the development of thrombocytopenia during chronic
hypoxia after splenectomy.
In summary, radioactive-iron and morphological analyses in both intact
and splenectomized rats during 30 days of exposure to hypoxia revealed
an important role of splenic erythropoiesis in the excessive
polycythemic response in the Hilltop strain of Sprague-Dawley rats.
Although exaggerated gene expression and production of EPO is a
characteristic response to hypoxia in this rat strain, the capacity for
splenic erythropoiesis also appears to be essential, either as an
extramedullary site or an extra source of progenitor cells.
| |
ACKNOWLEDGEMENTS |
This study was supported by National Heart, Lung, and Blood
Institute Grants HL-21159 (L. C. Ou) and HL-14127 (R. P. Smith).
| |
FOOTNOTES |
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Address for reprint requests and other correspondence: L. C. Ou, Dept.
of Physiology, Borwell Bldg., Dartmouth Medical School, 1 Medical
Center Drive, Lebanon, NH 03756-001 (E-mail:
lo.chang.ou{at}dartmouth.edu).
Received 23 April 1999; accepted in final form 21 July 1999.
| |
REFERENCES |
1.
Birks, J. W.,
L. W. Klassen,
and
C. W. Gurney.
Hypoxia-induced thrombocytopenia in mice.
J. Clin. Lab. Med.
86:
230-238,
1975[Medline].
2.
Cooper, G. W.,
and
B. Cooper.
Relationships between blood platelet and erythrocyte formation.
Life Sci.
20:
1571-1580,
1977[Medline].
3.
Crosby, W. H.
Hematopoiesis in human spleen.
Arch. Intern. Med.
143:
1321-1322,
1984.
4.
Donohue, D. M.,
B. W. Gabrico,
and
C. A. Finch.
Quantitative measurement of hematopoietic cells of the marrow.
J. Clin. Invest.
37:
1564-1570,
1958.
5.
Fruhman, G. J.
Splenic erythropoiesis.
In: Regulation of Hematopoiesis, edited by A. S. Gordon. New York: Appleton-Century-Crofts, 1970, p. 339-368.
6.
Fruhman, G. J.,
and
A. S. Gordon.
Quantitative effects of corticosterone on rat bone marrow.
Proc. Soc. Exp. Biol. Med.
88:
130-131,
1955.
7.
Hirsh, J.,
and
J. V. Dacie.
Persistent post-splenectomy thrombocytosis and thrombo-embolism: a consequence of continuing anaemia.
Br. J. Haematol.
12:
44-53,
1966[Medline].
8.
Hoffman, R.,
E. J. Benz, Jr.,
S. J. Shattil,
B. Furrie,
H. J. Cohen,
and
L. E. Silberstein.
Biology of erythropoietin, erythroid differentiation and maturation.
In: Hematology. Basic Principles and Practice. New York: Churchill Livingstone, 1995, p. 242-254.
9.
Keene, W. R.,
and
J. H. Jandl.
Studies of the reticuloendothelial mass and sequestering function of rat bone marrow.
Blood
26:
157-175,
1965[Abstract/Free Full Text].
10.
Lawrence, J. S.,
and
C. G. Craddock, Jr.
Stem cell competition: the response to antimetabolic serum as affected by hemorrhage.
J. Lab. Clin. Med.
72:
731-738,
1968[Medline].
11.
Lipson, R. L.,
E. D. Bayrd,
and
C. H. Watkins.
The post-splenectomy blood picture.
Am. J. Clin. Pathol.
32:
526-532,
1959[Medline].
12.
Loeffler, M.,
P. Herkenrath,
H. E. Wichmann,
B. I. Lord,
and
M. J. Murphy.
The kinetics of hematopoietic stem cells during and after hypoxia.
Blut
49:
427-439,
1984[Medline].
13.
Markoe, A. M.,
J. P. Okunewick,
and
L. M. Schiffer.
Kinetic analysis of splenic erythropoiesis in mice under prolonged hypoxic stress.
Exp. Hematol.
1:
340-349,
1973[Medline].
14.
Matter, M.,
J. R. Hartman,
J. Kautz,
Q. B. Demarsh,
and
C. A. Finch.
A study of thrombopoiesis in induced acute thrombocytopenia.
Blood
15:
174-185,
1960[Free Full Text].
15.
McBride, J. A.,
J. V. Dacie,
and
R. Shapley.
The effect of splenectomy on the leukocyte counts.
Br. J. Haematol.
14:
225-231,
1968[Medline].
16.
Ou, L. C.,
Y. N. Cai,
and
S. M. Tenney.
Responses of blood volume and red cell mass in two strains of rats acclimatized to high altitude.
Respir. Physiol.
62:
85-94,
1985[Medline].
17.
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[Abstract/Free Full Text].
18.
Ou, L. C.,
D. Kim,
W. Layton,
and
R. P. Smith.
The role of splenic erythropoiesis in the polycythemic response to the rat at high-altitude exposure.
J. Appl. Physiol.
48:
857-861,
1980[Abstract/Free Full Text].
19.
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:
1242-1251,
1998[Abstract/Free Full Text].
20.
Ou, L. C.,
and
R. P. Smith.
Hemoglobinemia in rats exposed to high altitude.
Exp. Hematol.
6:
473-478,
1978[Medline].
21.
Palmer, J. G.,
I. Kemp,
G. E. Cartwright,
and
M. M. Wintrobe.
Studies on the effect of splenectomy on the total leukocyte count in the albino rat.
Blood
6:
3-15,
1951[Abstract/Free Full Text].
22.
Papayannopoulou, T.,
and
C. A. Finch.
Measurements of red cell maturation.
Blood Cells
1:
535-546,
1975.
23.
Seifert, M. F.,
and
S. C. Marks, Jr.
The regulation of hematopoiesis in the spleen.
Experientia
41:
192-199,
1985[Medline].
24.
Smith, R. P.,
R. Kruszyna,
and
L. C. Ou.
Hemoglobinemia in rats exposed to high altitude is not due to an overload of catabolic mechanisms.
Aviat. Space Environ. Med.
50:
9-13,
1979[Medline].
25.
Van Liere, E. J.,
and
J. C. Stickney.
Hypoxia. Chicago, IL: University of Chicago Press, 1963.
26.
Wilcoxon, F.,
and
R. A. Wilcox.
Some Rapid Approximate Statistical Procedures. Pearl River, NY: Lederle Laboratories, American Cyanamid, 1960.
J APPL PHYSIOL 87(5):1901-1908
8570-7587/99 $5.00
Copyright © 1999 the American Physiological Society
TOP
ABSTRACT
INTRODUCTION
