From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
Blood Flow to Bone Marrow During Development of Anemia or Polycythemia
in the Rat
By Per Ole Iversen, Gunnar Nicolaysen, and Haakon 6. Benestad
We applied the radioactive microsphere method to follow
the magnitude and time course (0 to 96 hours) of blood flow
changes during development and recovery from anemia in
awake rats. Blood flow was also monitored during a 96-hour
periodafter polycythemia was induced (erythropoietinadministered subcutaneously[SC]). The possible influence of innervation was also examined. After a blood loss of approximately 50% (hypovolemia),blood flow to the femoral marrow
tripled within 12 hours and remained elevated for the entire
96-hour period. The relative increase in blood flow to the
femoral bone was even greater. Similar findings were obtained in rats with phenylhydrazine (PHZ) hemolytic anemia
(normovolemia).Denervationhad no detectable effect on the
increased blood flow to either marrow or bone. The aug-
mented blood flow during hemolytic anemia was accompanied by a doubling of the oxygen consumption rate by the
marrow, while the glucose uptake was not detectably altered. Erythropoietin supplements (3 x 1.000 IU/kg, SC,
6-hour intervals) increased blood flow to the marrow by
approximately25% after 48 hours, and at 72 hours the blood
flow had reached a value twice that obtained under control
conditions. These results indicate that blood flow to bone
marrow is highly variable and hormonally and/or locally
regulated.This may have practicalconsequencesfor marrow
transplantation technology and for administration of drug
therapyto patientswith insufficient bone marrow hematopoiesis.
o 1992 by The American Society of Hematology.
M
by inhalation.’ These studies have not been sufficiently
extensive to clarify the principles of blood flow regulation.
The findings have also been contradictory, in that marrow
blood flow has been claimed to be dependent on: as well as
unrelated to,7the bone marrow cellularity. Technical limitations can explain the slow progress in this field.
We wanted to use an experimental animal model and a
technique that could satisfy the following requirements: (1)
the marrow cavities to be investigated should be filled with
red marrow and absent of yellow, fatty marrow; (2) controlled and reproducible perturbations of hematopoiesis
should be possible; (3) serial measurements of blood flow in
awake animals should be feasible; and (4) a distinction
between hormonal/local and neural mechanisms of blood
flow regulation should be made. Consequently, we chose
rats and the radioactive microsphere method. To our
knowledge, no other study has been published that satisfies
all four of the above conditions.
The purpose of the present study was to find out how
blood flow to bone marrow is regulated during initial (0 to
96 hour) development and repair of a hypovolemic or
normovolemic anemia, as well as during induction of
polycythemia.
UCH HAS BEEN learned recently about how growth
and differentiation factors affect hematopoiesis in
vitro after the advent of sophisticated cell culture methods,
purified recombinant cytokines, well-characterized hemic
cell lines, and other techniques of molecular biology.’ We
know far less about the physiology of the bone marrow, that
is, about the in situ regulatory interplay of hematopoietic
cells with the above-mentioned and other factors, such as
nerves, marrow vasculature, and other stromal elements.
Yet, such integrated knowledge may have more than
theoretical interest, since it might lead to improved acceptance of bone marrow grafts and increased efficiency of
marrow cell stimulation with recombinant growth factors.
Therefore, we studied an aspect of bone marrow physiology
that has been neglected, namely marrow blood flow and its
regulation.
Because it is encased in bone, the marrow vasculature is
not easily accessible. Nevertheless, vital microscopy of
marrow blood flow has been performed’ and controlled
perfusion through the nutrient femoral marrow artery of
normal and anemic rabbits3 has given some clues to the
understanding of blood flow regulation. More recently,
blood flow indicators that do not necessitate extensive
target organ dissection have been used. 1311-antipyrine
has
been administered into the marrow cavity: InXe intraven ~ u s l yradioactive
,~
microspheres intraarterially,6 and ”CO,
From the Department of Physiology, Institute of Basic Medical
Sciences, University of Oslo, Norway.
Submitted December 6, 1990; accepted September 24, 1991.
Supported by NAVF, The Norwegian Research Council on Cardiovascular Diseases, and Nordisk Insulin Foundation Committee. P. 0.I.
holds a research fellowship with The Norwegian Research Council for
Science and the Humanities (NAVF).
Address reprint requests to Per Ole Iversen, MD, Department of
Physiology, Institute of Basic Medical Sciences, PO Box 1103,
Blindem, N-0317 Oslo, Norway.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1992 by The American Society of Hematology.
0006-497119217903-0018$3.00/0
594
MATERIALS AND METHODS
Rats
One hundred fifty-five male Wistar rats (384 to 423 g) were
included in the study. Before surgery, rats were kept in cages with
an ambient temperature of approximately 19°C. Lights went on at 7
A M and off at 11 PM.Rats had free access to water and pellet food.
Permission was given by the Animal Experimentation Committee
to perform experiments on awake rats.
Surgery and Postoperative Surveillance
General anesthesia was achieved with Equithesin (chloralhydrate, 42.5 mg/mL; magnesium sulphate, 21 mg/mL; and pentobarbital, 9.7 mg/mL; 0.4 mL/100 g body weight, intraperitoneally
[IP]). Supplementary oxygen was provided if necessary.
In groups A to F (see below), a PE 50 catheter for infusion of
radioactive tracers was inserted into the left common carotid artery
of each rat and positioned within the left ventricle, guided by
pressure tracing. The other end of the catheter was led subcutaneously (SC) to the neck, filled with heparin (2,000 IU), exteriorized,
Blood, Vol79, No 3 (February 1). 1992: pp 594-601
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
MARROW BLOOD FLOW DURING ERYTHROPOIESIS
and heat-sealed. Similarly, a PE 10 catheter for reference sampling
of blood (groups A to E, see below) was placed into one brachial
artery and advanced proximally so that its tip was just within the
axillary region. This catheter was also used for measurement of
mean arterial blood pressure. The other end of this catheter was
led SC to the neck, filled with heparin, exteriorized adjacent to the
infusion catheter, and heat-sealed. In this manner, atraumatic
access to the catheters was ensured. The exact positions of the
catheters were verified by dissection postmortem.
In each of the 32 rats in groups C and D (see below), one hind leg
was denervated. The sciatic nerve on one side was exposed at the
level of the sciatic notch and cut. Likewise, the ipsilateral femoral
nerve was exposed just below the inguinal ligament and cut. The
contralateral hind leg served as control.
All 155 rats were given ampicilline (15 mg/kg body weight,
intramuscularly [IM]). Rats in groups A to F (see below) were
returned to their cages to recover. Weight loss during the first 24
hours was at most 2.2% of initial body weight. Weight reduction
per 24 hours was smaller during the remaining experimental
period. Values for leukocytes, arterial blood gases, and arterial pH
all remained within the normal range during the experimental
period in all 155 rats.
Blood Flow Measurements
The microsphere method described by Heymann et al’ was
adopted to the present study as detailed by Iversen et al? Evidence
for the validity and reliability of our version of this method has
recently been presented.’ I o Briefly, microspheres (diameter, 16
pm; New England Nuclear, Boston, MA) labeled with either I4’Ce,
51Cr, lU3Ru,or y5Nband stored in 0.9% NaCl containing 0.2%
Tween 80, were ultrasonicated for 2 minutes. Shortly after and
immediately before an infusion, the microspheres were vigorously
mixed for 1 minute with a vortexer. The microspheres were then
transferred to a 10-mL syringe and placed on the infusion pump
(Harvard Instruments, Cambridge, MA). The infusion system was
designed to ensure adequate mixing of the microspheres during the
entire infusion period. This was verified in a previous study where
we found that less than 0.5% of the microspheres were deposited
as pairs or clusters? This latter result stems from microscopic
examinations of skeletal muscles from a rabbit that had received
6.5 X l@microspheres. Each infusion contained approximately
3.0 x lo5 microspheres labeled with only one nuclide (randomly
chosen). The infusions each lasted 30 seconds at a constant rate of
3 mL/min. The reference catheter in the brachial artery was
connected to a similar pump. Withdrawal of blood through this
catheter, at a rate of 1 mL/min, started 30 seconds before the
infusion and was stopped 1 minute after the infusion was ended.
The volume deficit of 0.5 mL (volume in reference sample minus
the volume infused) was immediately corrected with 0.9% NaCl
(intraarterially). The 143 unrestrained rats in groups A to F (see
below) were all standing quietly during the periods of tracer
infusions.
Organ blood flow (Fs, mL/100 g x min) was calculated according to the equation: F, = F, x Q,/Q,, where F, is the reference
sample flow (1 mL/min), Q, is the number of microspheres
deposited per 100g tissue, and Q, is the number of microspheres in
the reference sample. The lowest number of microspheres deposited in any sample was 572.”
Assessment of Glucose Uptake
The deoxyglucose method was used to determine the uptake of
glucose from plasma into the marrow cells from the femurs of
group F rats (see below).” We gave 0.74 MBq (20 pCi) of
3H-deoxyglucose (specific activity, 3.8 GBq/mg, 103 mCi/mg;
595
Amersham, Buckinghamshire, England) as an intraarterial bolus
injection. The ’H-deoxyglucose was then allowed to circulate 30
minutes before the rats were killed. All marrow cells from one
femur in each rat were then removed as described below, and the
radioactivity determined. The value for glucose uptake for all
marrow cells sampled from one femur was expressed as a percentage of the injected dose of ’H-deoqglucose. This technique thus
yielded one estimate of the glucose uptake into the marrow in each
of the 36 rats, at six time points, six of the rats serving as
preintervention controls. In four separate rats, we perfused the
whole body postmortem and then determined the marrow glucose
uptake. The values for glucose uptake in these rats were similar to
those in the six preintervention control rats, indicating that the
measured values represent true uptake of glucose by the marrow
cells.
Determination of Oxygen Consumption Rate
Oxygen consumption rate in group G rats (see below), six test
rats, and six control rats was estimated from the changes in oxygen
tension in an incubated marrow cell suspension as described by
Opdahl et al.” Immediately after each rat was killed, all marrow
cells were removed from one femur as described below. All marrow
cells were then suspended in Fisher’s medium (Gibco, Paisley,
Scotland) containing 0.5% bovine albumin (Sigma, St Louis, MO)
and adjusted to pH 7.35 with 10 mmol/L HEPES buffer (Sigma).
One milliliter of this suspension was enclosed in a chamber. The
suspension was kept at 37°C and constantly stirred, its oxygen
tension being measured continuously with a polarographic electrode (MSE, UK). In this way, one measurement could be obtained
from each rat. The values for oxygen consumption rate were
expressed as nmol O,/min.
Analyses of Blood Variables
Hemoglobin (Hb), hematocrit (Hct), and reticulocytes were
measured in duplicate with standard procedures. The values for
reticulocytes were corrected according to the formula: Corrected
reticulocytes (%) = actual reticulocytes (%) x (actual Hct/control
Hct). When several measurements of either Hb, Hct, or reticulocytes were made from one rat, the person performing the measurements was ignorant about the order in which the samples had been
collected.
Levels of erythropoietin were determined in plasma from groups
A and E rats with a commercial kit (Medac, Hamburg, Germany)
based on an enzyme-linked immunosorbent assay (ELISA) procedure as outlined by Goto et al.I4
Emenmental Protocol
GroupA (bleeding and hypovolemic anemia). Forty-eight hours
after surgery, one batch of microspheres was infused into 30 rats
(control infusion). Hypovolemic anemia was then induced by
bleeding through the intracardial catheter at a rate of 2 mL/min
with a withdrawal pump (Harvard Instruments). The blood loss
was made to be approximately 1.5% of the body weight and was
typically between 5 and 7 mL. Microspheres were then infused into
10 rats 5 minutes, 24 hours, and 48 hours after the control infusion,
into another 10 rats after 6, 17, and 96 hours, and finally into
another 10 rats after 12, 72, and 84 hours. Into five of these latter
rats, we also infused 0.19 MBq (5 pCi) 59Fe-citrate(specific activity,
1.3 GBa/me. 36 mCi/me. IP: New Eneland Nuclear, Dreieich.
Germany) 5 hours before the bleeding. Five other rats (sham
anemic) were bled in a similar manner as described above, but the
blood was immediately reinfused. These five rats also received 0.19
MBq 5yFe-citrate5 hours before the reinfusion. Microspheres were
. -
I
I
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
IVERSEN, NICOLAYSEN, AND BENESTAD
596
infused just before bleeding (control infusion) and then after 5
minutes, 24 hours, and 96 hours.
Group B (hemolysis and normovolemic anemia). Microspheres
were infused into 20 rats 48 hours after surgery. The hemolyzing
agent phenylhydrazine (Phz; 50 mg/kg body weight, IP) was then
administered immediately. Microspheres were then infused into 10
of the rats 12 and 48 hours after the control infusion and into the
remaining 10 rats after 36,72, and 96 hours.
Group C (denervation and hemolysis). Twenty rats with one
hind leg denervated were each given one batch of microspheres 96
hours after surgery. Normovolemic anemia was induced as described for group B and microspheres were administered into 10 of
the rats 48 and 96 hours after the control infusion. The remaining
10 rats received microspheres after 12,36, and 72 hours.
Group D (denervation, bleeding and plasma transfusion; normovolemic anemia). From each of 12 rats, we removed approximately 6 mL plasma 2 weeks before surgery. The plasma was stored
at -70°C until use. These 12 rats had one hind leg denervated and
were each given one batch of microspheres 96 hours after surgery.
The rats were then bled as described for group A rats. Immediately
following the bleeding, the blood loss was replaced with an equal
amount of autologous plasma. Microspheres were then given to six
of the rats 3, 6, and 12 hours after the control infusion. The other
six rats received microspheres after 24 hours.
Group E (polycythemia). Twenty rats were all given one infusion of microspheres 48 hours after surgery. Recombinant human
erythropoietin (rHuEPO, Integrated Genetics, Boston, M A 1,000
IU/kg body weight, SC) was then administered three times at
6-hour intervals. Ten rats received microspheres 6, 24, and 72
hours after the control infusion. The remaining 10 rats were given
microspheres after 12,48, and 96 hours.
Group F (glucose uptake and hemolysis; normovolemic anemia).
Forty-eight hours after surgery, six control rats received one
intraarterial bolus injection of 'H-deoxyglucose. In another 30 rats,
normovolemic anemia was induced as described for group B.
Deoxyglucose was injected either 5 minutes, or 12, 24, 48, or 96
hours after injection of Phz, with six rats in each group. The rats
were killed 30 minutes after the deoxyglucose injection. The
activity of 3H in plasma was then virtually zero.
Group G (oxygen consumption rate and hemolysis; normovolemic
anemia). Forty-eight hours after the injection of ampicilline, six
rats were killed, and the marrow cells within one femur from each
rat were analyzed with respect to the oxygen consumption rate.
Hemolytic anemia was induced as described for group B rats in six
other rats 24 hours before the oxygen consumption rate was
determined.
Preparation of Samples
Experiments were terminated by giving rats a lethal dose of
pentobarbital.
One femur was removed from rats belonging to groups B, E, F,
and G and both femurs from group A rats. Both tibiae were
removed from rats in groups C and D. The bone was cleared of all
visible soft tissue by dissection. The femurs and the tibiae were
then weighed before the marrow was flushed out. Marrow cells
were retrieved as completely as possible from the bones by flushing
the marrow cavity with 0.9% NaCl and by scraping the cavity wall.
Flushing was performed several times with a total amount of 60 mL
saline. In addition, one femur from each of the group A rats was
flushed twice with 20 mL 1% collagenase in saline (37°C Sigma
type ,).I5 Between these flushes, the collagenase-containingfemurs
were incubated for 5 minutes at 37°C. Finally, all bones, except
from group G rats, were flushed once with 20 mL acetone and dried
with air before they were reweighed.
From 11 specimens of decalcified bone taken from group A rats,
histological preparations (20-pn cross-sections) were made and
stained with hematoxylin and eosin.
One nutritional artery was removed from both tibiae in four rats
in group E. These arteries were prepared for fluorescence histochemical demonstration of catecholamines according to the method
of Furness and Costa.16
Assessment of Radioactivity
Gamma-emitting nuclides within the bones and marrow samples
from groups A to E were counted in an Auto-Gamma 5220
(Packard, Massachusetts). Cross-over between the various y-emitting nuclides was corrected for by stripping, neglecting the minute
( < 0.5%) overlap of the lower energy nuclides into the higher ones.
Bones from the rats in group B were counted before and after they
had been dissolved in 2 mL ammonium hydroxide (Soluene-350,
Packard). The count rates were identical and thus there was no
self-absorption in the bone.
Bone and marrow samples from group F rats were also dissolved
in 2 mL ammonium hydroxide. Then, 2 mL scintillation liquid
(Hionic Fluor, Packard) was added. The 'H activity was then
recorded in a p-counter (Tri Carb 460, Packard). The degree of
quenching of the 'H-nuclide (usually 17%) was determined with
an internal (3H) standard.
The smallest number of counts above a low background for any
sample was 6,900, from either a y- or a P-emitting nuclide.
-
Statistics
Values are expressed as medians with their 95% confidence
intervals, determined with a nonparametric method. If the median
of a control group was not included in the 95% confidence interval
of the test group, and vice versa, then the two groups of data were
considered significantly different at the 5% level. In the figures, the
median symbols sometimes cover one or both ends of the confidence bars.
RESULTS
Validation Studies
Group A. Compared with the five sham-operated rats,
the uptake of 59Feinto the marrow was approximately 60%
higher in the anemic rats 96 hours after bleeding. No
appreciable activity of "Fe was detected in any of the
flushed bones, indicating that the removal of marrow by
flushing the bones had been complete. No significant
differences were found in blood flow to bone between those
femurs that had been flushed with collagenase and those
that had not.
Histological preparations from the interface between the
bone and the marrow cavity showed only occasional and
small fragments of marrow. Microscopic examination of the
marrow content of 10 rats showed only red and no yellow
marrow.
Group E. The arteries supplying denervated tibias were
only faintly stained with glyoxylic acid, in marked contrast
to the brightly stained arteries from the contralateral
innervated bones. Thus, catecholamine stores had been
depleted, indicating that denervation had been fairly complete.
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
597
MARROW BLOOD FLOW DURING ERYTHROPOIESIS
Hypovolemic Anemia
Group A. Approximately 12 hours after bleeding, both
Hb and Hct were reduced by approximately 50% (Fig 1).
During the remaining experimental period, Hb and Hct
remained fairly constant. The number of reticulocytes was
markedly elevated approximately 48 hours after bleeding
(Fig 1).
Ninety-six hours after bleeding, the plasma concentration of erythropoietin had increased markedly to 140 (135
to 155) mU/mL, as compared with the control value of 7
(5.5 to 8.5) mU/mL.
Values for mean arterial blood pressure and heart rate
are presented in Table 1. During bleeding, the mean
arterial blood pressure decreased, from approximately 95
mm Hg to approximately 60 mm Hg. The rats then
remained hypotensive throughout the observation period.
A concomitant increase in heart rate from approximately
450 to 530 bpm persisted for approximately 24 hours after
bleeding. The heart rate then gradually decreased.
Blood flow to the marrow had increased substantially by 6
hours after the bleeding, and it remained elevated approximately three times above the control level for the entire
96-hour period (Fig 2). A surprising increase (P < .05) in
blood flow to the bone was even more pronounced in
relative terms, but showed a more transient course (Fig 2).
Normovolemic Anemia
Group B. Data for Hb, Hct, and reticulocytes were
similar to those presented for group A (Fig l), but with a
slower development of the anemia.
No major changes in either mean arterial blood pressure
or heart rate were detected (Table 1).
The temporal changes in blood flow to the marrow and
bone following Phz injection are depicted in Fig 3. In these
rats, blood flow to the marrow increased more slowly, thus
reflecting the slow development of anemia. However, marrow blood flow reached a higher peak value than in group A
rats (50 v 35 mL/100 g x min).
Group C. The blood flow pattern for both the marrow
and bone was similar in the innervated and denervated hind
leg (Fig 4); consequently, denervation had no detectable
effect on the initiation and maintenance of a high level of
blood flow to marrow and bone. The tibial bones showed a
more marked hyperemia than the femoral bones in group B
rats (Fig 3 v Fig 4); however, no direct comparison has been
performed of the blood flow rates to these bones.
Group D. Blood flow increased rapidly to marrow and
bone in both the innervated and denervated hind leg (Fig
4). Thus, the sympathetic vasomotor activity did not play
any appreciable role for the rapid induction of the hyperemic response to the acute normovolemic anemia. Neither
mean arterial blood pressures nor heart rates were significantly different from the values obtained during control
conditions in rats of groups C and D.
Injection of rHuEP0
GroupE. Hb, Hct, and reticulocytes all increased within
72 hours after rHuEPO injections (Fig l), indicating an
HEMOGLOBIN
gllOOml
20€
t
Y-----HEMATOCRIT
*I*
50 r
40
Erythropoietin inject.
30
Hemolysis
20
Bleeding
R ET1 CULOCY TE S
Erythropoietin inject.
OL
f
I
, , I
10
1
1
1
1
1
30
1
1
1
1
1
1
50
1
1
1
1
70
1
1
1
1
1
1
1
90 h
Time
Fig 1. Time course of changes in Hb, Hct, and corrected reticulocyte counts for 30 rats in group A (bleeding, hypovolemic anemia), 20
rats in group B (Phz-hemolysis, normovolemic anemia), and 20 rats in
group E (rHuEPO, polycythemia). The arrows denote the time point
when bleeding (group A), Phz injection (group B), or injection of
rHuEPO (group E) was initiated. Data points corresponding to the
arrows are preintervention control values. Median values with their
95% confidence intervalsare shown.
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
598
IVERSEN, NICOLAYSEN, AND BENESTAD
Table 1. Hemodynamic Characteristicsof Experimental Rats During Control and Experimental Phases
Before One Infusion
Group
After One Infusion
Before Last Infusion
After Last Infusion
MABP
HR
MABP
HR
MABP
HR
MABP
HR
91-101
90-106
94-115
85-105
90-105
433-462
429-453
425-449
432-452
425-446
90-99
95-103
93-106
90-108
93-104
429-455
433-452
429-454
425-445
435-452
76-85
93-102
90-105
94-110
92-110
449-467
425-455
430-445
425-456
423-446
79-87
93-110
94-104
92-110
90-105
454-462
430-462
438-451
425-456
435-450
~~
A
B
C
D
E
Mean arterial blood pressure (MABP, mm Hg) and heart rate (HR, bpm) obtained before and after the first and last infusion of microspheres,
respectively, in groups A, B, C, D, and E. Values are the 95% confidence intervals of the medians. Data obtained before the first infusion represent
baseline values.
increased erythropoietic activity. A concomitant hyperemia
was observed both in the marrow and in the bone (Fig 5).
However, the blood flow increase was not significant until
48 hours after the first rHuEPO injection, and not marked
before the 72-hour sampling. The plasma concentration of
erythropoietin increased from 6.0 (4.5 to 7.5) mU/mL
under control conditions to 9.5 (8.5 to 11.5) mU/mL at 48
hours, further to 55 (35 to 85) mU/mL at 72 hours, and
finally to 145 (130 to 155) mU/mL at 96 hours.
Marrow Uptake of Glucose
Group F. Glucose uptake remained at control levels
throughout the experimental period (P > .05), measurements being made just before or either 5 minutes, or 12,24,
48, or 96 hours after Phz injection. The median uptake of
glucose into the retrieved marrow cells from one femur was
approximately 2.5 x
of the injected dose (36 rats).
Thus, the induced hemolytic, normovolemic anemia was
not accompanied by a detectable change in marrow metabolic activity as measured by glucose uptake.
Marrow Oxygen ConsumptionRate
Group G . The oxygen consumption rate of femoral cells
(see group F above) doubled (P < .05) from 15.3 (12.4 to
Blood flow
mI/lOOg.min
18.6) to 32.4 (30.8 to 36.7) nmol O,/min (six rats) during
transition from the control to the anemic condition 24 hours
after induction of hemolysis. The enhanced erythropoietic
activity within the marrow thus appeared to be accompanied by an increase in both blood flow and oxygen consumption rate.
DISCUSSION
Blood Flow to Marrow and Bone During Stimulated
Eythropoiesk
We found that bone and bone marrow blood flow
increased markedly in rats after induction of both a
hypovolemic (bleeding) and a nonnovolemic (hemolytic)
anemia and after polycythemia elicited by rHuEPO injections. The increase in blood flow occurred early after
bleeding, being significant after 6 hours, when the Hb was
still only slightly below normal. The start of the increased
blood flow was intermediate in time under the hemolytic
regimen, but late after the rHuEPO injections. This succession in time was reflected by the three reticulocyte curves,
in that their half-maximum values were reached approximately 24 hours after the half-maxima of the respective
marrow blood flow curves.
To our knowledge, this is the first report on changes in
bone marrow blood flow occurring during perturbations of
Blood f l o w
m l / 100g.min
Femoral marrow
E
Femoral bone
t
lo
30
50
70
90 h
Time
Fig 2. Blood flow to the femoral marrow and bone of bled rats
(group A). *Denotes a measurement 5 minutes after bleeding. Animals
and other symbols as in Fig 1.
Fig 3. Blood flow to the femoral marrow and bone of group B rats
(hemolytic, normovolemic anemia). Symbols as in Fig 2.
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
MARROW BLOOD FLOW DURING ERYTHROPOIESIS
599
and neurally. However, a rapidly induced normovolemic,
hemorrhagic anemia (group D, bleeding and plasma infusion) and a more slowly developing hemolytic anemia
(group B) were accompanied by a rapidly and slowly
occurring blood flow increase, respectively, that were not
detectably different in the innervated and the denervated
tibiae. This indicates that the augmented marrow and bone
blood perfusion was not dependent on an intact innervation
of bone and marrow in any of our experimental models.
This conclusion is supported by our finding of a seemingly
constant blood flow during the first 5 minutes after rapid
bleeding (Fig 2), since a nervous regulatory mechanism
would presumably have been apparent within this time
period.
Erythropoietin might be a regulator of blood flow to bone
marrow directly or indirectly. This suggestion has been
made by earlier investigators'Y.21
for bone marrow and for
spleen blood flow. When the increased marrow blood flow
was similar in the bled and in the rHuEPO-injected groups
of rats (A and E), 96 hours after the start of erythropoietic
perturbations, we found similar and markedly greater than
normal plasma concentrations of erythropoietin in the two
groups. It could be that increased metabolism triggered by
the enhanced erythropoiesis might lead to a larger blood
flow, as is the case in, for example, skeletal muscle, putative
chemical mediators here being K', H', phosphates, or
endothelium-derived relaxing factor^.^.^^ Perhaps some more
specific mediator, such as a cytokine or an intermediary of
heme metabolism, is released in the marrow when erythropoiesis is accelerated. In any case, we observed a doubling
of oxygen consumption rate in erythropoietically stimulated
marrow, which indicates a higher metabolic rate. We
cannot explain the lack of any detectable increase in the
measured glucose uptake by the marrow, but it could mean
that the erythroid cells compete with the myeloid cells for
glucose as an energy substrate, so that the total glucose
uptake remains approximately the same during increased
Fig 4. Blood flow to the tibial marrow and bone with intact and cut
nerves. Results are from group C rats (hemolyticnormovolemia), and
group D rats (bleeding and plasma transfusion, normovolemic anemia).
steady-state hematopoiesis, with serial measurements obtained from one animal. Comparisons with previous data
based on single measurements during steady-state, mostly
on patients,'." are therefore of limited value.
Michelsen3 showed that blood flow to a denervated,
perfused femur of a rabbit with Phz hemolytic anemia and
accelerated erythropoiesis was approximately 2.5 times
higher than the mean control values. Chen et a16confirmed
with the microsphere method that an increased marrow
blood flow was present after 3 days and persisted for about
1 week after one large (25 mg/kg body weight, IP) dose of
Phz to rabbits.
The mechanism behind the increased blood flow is
unknown. Bone marrow contains sympathetic, noradrenergic, and vasoconstrictor nerves." In theory, the blood flow
regulation is therefore possibly both hormonally/locally
Fig 5. Blood flow to the marrow and bone in group E rats given
three rHuEPO injections (1,000 IU/ kg, SC, 6-hour intervals). Symbols
as in Fig 2.
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
600
erythropoiesis and decreased myelopoiesis.24However, this
hypothesis also rests on the assumption that anaerobic
metabolism is more prevalent in myelopoietic than in
erythropoietic cells.
We found concomitant changes in blood flow to marrow
and to bone, the only exception being the late phase of
hypovolemic anemia, when marrow blood flow stayed high,
while bone blood flow had declined. The parallel blood flow
increase could be due to the fact that the nutritional artery
looses its muscular tunica media shortly after penetration
to the marrow cavity.25Branches given off thereafter to
either the haversian canals of the bone or to the marrow
would receive increased blood supply if the total blood flow
is enhanced by dilatation of the muscular parts of the
vessels.
Why the increase in blood flow to the bone after bleeding
was transient is not easily explained. Chen et a16 did not
detect any significant increase in cortical blood flow at all in
anemic rabbits, and Michelsen’s results led him to postulate
that the blood flow increase to the marrow was at least
partly due to decreased venous resistance.’ New approaches therefore appear necessary to clarify this confusing picture. The microsphere method allowed accurate
determination of blood flow to the bone marrow. This
method does not give information as to which vessels
contribute to any changes in vascular resistance. Furthermore, it does not give any information about events within
the microcirculation, except for their contribution to overall
vascular resistance.
The Microsphere Method
The validity and reliability of the microsphere method for
estimation of cardiac output and organ blood flows in
conscious rats have been firmly established.2b29However,
there are conflicting data concerning hemodynamic stability after infusion of high numbers of microspheres. While
Tsuchiya et a13’ reported a significant decrease in cardiac
IVERSEN, NICOLAYSEN, AND BENESTAD
output and an increase in heart rate when the number of
infused microspheres exceeded 1 x lo5, Stanek et al”
observed a significant decrease in heart rate only when as
many as 1.4 x 10“microspheres had accumulated within the
rat. In the present study, we infused at most a total dose of
1.2 x lo6 microspheres into a single rat (divided over
several infusions). No changes in either arterial blood
gases, heart rate, or mean arterial blood pressure were
observed following any infusion. We therefore conclude
that the infusion of microspheres did not interfere with the
circulation in these rats.
Possible Clinical Implications
In bone marrow transplantation, for marrow failure, and
in the future also probably as a remedy for inborn errors of
metabolism, as many as possible of the intravenously
introduced stem cells should “home to fertile soil” in the
marrow. Here, pretreatment of the patient to secure a high
blood flow to the bone marrow during the transplantation
may be worthwhile. On the other hand, during chemotherapy of neoplasms without suspected marrow deposits, a
sparse blood flow would be desirable. Some recombinant
cytokines (colony-stimulating factors) have had unexpected
granulocyte-mobilizing effects, which might depend on
blood flow changes in the marrow, since earlier experiments
suggested a relationship between the rate of bone marrow
perfusion and the rate of cell release.” To clarify these and
other issues, the present work should be continued; for
example, the possible effects of recombinant cytokines on
bone marrow blood flow should be investigated.
ACKNOWLEDGMENT
Recombinant human erythropoietin was kindly provided by P.V.
Zarins at Norske Hoechst. The technical assistance of I. StrgmGundersen, T. Flatebo, and I.S.M. Johannessen is gratefully
appreciated. We also thank A.G. Bechensteen, MD, for performing the erythropoietin analyses.
REFERENCES
1. Metcalf D: The Molecular Control of Blood Cells. Cam9. Iversen PO, Standa M, Nicolaysen G: Marked regional
heterogeneity in blood flow within a single skeletal muscle at rest
bridge, MA, Harvard University Press, 1988
2. Brinemark P: Vital microscopy of bone marrow in rabbit.
and during exercise hyperemia in the rabbit. Acta Physiol Scand
Scand J Clin Lab Invest 115,1959 (suppl38)
136:17,1989
10. Iversen PO, Nicolaysen G: Heterogeneous blood flow distri3. Michelsen K Hemodynamics in the bone marrow of anemic
butions within single skeletal muscles in the rabbit: Role of
rabbits with increased hematopoiesis. Acta Physiol Scand 77:52,
vasomotion, sympathetic nerve activity and effect of vasodilation.
1969
Acta Physiol Scand 137:125, 1989
4. Semb H: Bone marrow blood flow studied by iodoantipyrine
11. Buckberg GD, Luck JC, Payne B, Hoffman JIE, Archie JP,
clearance technique. Surg Gynecol Obstet 133:472,1971
F d e r DE: Some sources of error in measuring regional blood flow
5. Lahtinen R, Lahtinen T, Romppanen T: Bone and bonewith radioactive microspheres. J Appl Physiol31:598,1971
marrow blood flow in chronic granulocytic leukemia and primary
12. Sokoloff L, Reivich M, Kennedy C, Des Rosiers MH, Patlak
myelofibrosis. J Nucl Med 23:218,1982
CS, Pettigrew KD, Sakurada 0, Shinohara M: The (“C) deoxyglu6. Chen LT, Chen MF, Porter V L Increased bone marrow
cose method for the measurement of local cerebral glucose
blood flow in rabbits with acute hemolytic anemia. Am J Hematol
utilization: Theory, procedure, and normal values in the conscious
22:35,1986
and anesthetized albino rat. J Neurochem 282397, 1977
7. Martiat P, Ferrant A, Cogneau M, Bo1 A, Michel C, Rodhain
13. Opdahl H, Benestad HB, Nicolaysen G: Differences and
J, Michaux JL, Sokal G: Assessment of bone marrow blood flow
similarities between neutrophil granulocyte responses in vitro: The
using positron emission tomography: No relationship with bone
effects of zymosan-activated plasma, phorbol myristate acetate and
marrow cellularity. Br J Hematol66307,1987
n-formyl-methionyl-leucyl-phenylalanine.
Acta Anaesthesiol Scand
8. Heymann MA, Payne BD, Hoffman JIE, Rudolph AM: Blood
31:491,1987
flow measurements with radionuclide-labeled particles. Prog Car14. Goto M, Murakami A, Akai K, Kawanishi G, Ueda M, Chiba
diovasc Dis 20:55,1977
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
MARROW BLOOD FLOW DURING ERYTHROPOIESIS
H, Sasaki R: Characterization and use of monoclonal antibodies
directed against human erythropoietin that recognize different
antigenic determinants. Blood 74:1415,1989
15. Gupta V Differential secretion of stimulatory-inhibitory
CFU-GM activity by stromal cells from different compartments of
mouse bone femur. Exp Hematol17188,1989
16. Furness JB, Costa M: The use of glyoxylic acid for the
fluorescence histochemical demonstration of peripheral stores of
noradrenaline and 5-hydroxytryptamine in whole mounts. Histochemistry 41:335, 1975
17. Lahtinen R, Lahtinen T, Hyodynmaa S: Increased bone
marrow blood flow in polycythemia vera. Eur J Nucl Med 3:19,
1983
18. Weiss RA, Root WS: Innervation of the vessels of the
marrow cavity of certain bones. Am J Physiol197:1255,1959
19. McClugage SG Jr, McCusky RS, Meineke HA. Microscopy
of the living bone marrow in situ. 11. Influence of the microenvironment on hematopoiesis. Blood 38:96,1971
20. McCusky RS, Meineke HA, Townsend SF: Studies of the
hemopoietic microenvironment. I. Changes in the microvascular
system and stroma during erythropoietic regeneration and suppression in the spleens of CF1 mice. Blood 39:697, 1972
21. McCusky RS, Meineke H A Studies of the hemopoietic
microenvironment. V. Erytropoietin-induced release of vasoactive
substance(s) from erythropoietin-responsive stem cells. Proc SOC
Exp Biol Med 156:181,1977
22. Kiens B, Saltin B, Wallae L, Wesche J: Temporal relationship between blood flow changes and release of ions and metabolites from muscles upon single weak contractions. Acta Physiol
Scand 13631.1989
60 1
23. Furchgott RF: Studies on endothelium-dependent vasodilation and the endothelium-derived relaxing factor. Acta Physiol
Scand 139:257,1990
24. Ventura GJ, Hester JP, Buescher ES, Vadhan-Raj S, Durrett A, Reading C L Hematopoiesis in limiting dilution cultures:
Influence of cytokines on human hematopoietic progenitor cells.
Exp Hematol18:878,1990
25. Yoffey JM: Structural peculiarities of the blood vessels of
the bone marrow. Bib1 h a t 7:298,1965
26. Malik AB, Kaplan JE, Saba TM: Reference sample method
for cardiac output and regional blood flow determinations in the
rat. J Appl Physiol40:472, 1976
27. Tsuchiya M, Ferrone RA, Walsh GM, Frohlich ED: Regional blood flows measured in conscious rats by combined Fick
and microsphere methods. Am J Physiol235:H357,1978
28. Ishise S, Pegram BL, Yamamoto J, Kitamura Y, Frohlich
ED: Reference sample microsphere method: Cardiac output and
blood flows in conscious rats. Am J Physiol239:H443,1980
29. Laughlin MH, Armstrong RB, White J, Rouk K: A method
for using microspheres to measure muscle blood flow in exercising
rats. J Appl Physiol52:1629, 1982
30. Tsuchiya M, Walsh GM, Frohlich ED: Systemic hemodynamic effects of microspheres in conscious rats. Am J Physiol
233:H617,1977
31. Stanek KA, Smith TL, Murphy WR, Coleman TG: Hemodynamic disturbances in the rat as a function of the number of
microspheres injected. Am J Physiol245:H920,1983
32. Dornfest BS: Perfusion techniques-A tool in elucidating
blood cell release phenomena, in Gordon AS (ed): Regulation of
Hematopoiesis. New York, NY, Appleton-Century-Crofts, 1970, p
237
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
1992 79: 594-601
Blood flow to bone marrow during development of anemia or
polycythemia in the rat
PO Iversen, G Nicolaysen and HB Benestad
Updated information and services can be found at:
http://www.bloodjournal.org/content/79/3/594.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.