| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
DEVELOPMENT AND TISSUE PLASTICITY
1Department of Medicine, Division of Cardiology, and Departments of 2Physiology and Pharmacology, 4Surgery, and 5Obstetrics and Gynecology, at Oregon Health Science University, and 3Division of Cardiology, Portland Veterans Affairs Medical Center, Oregon Health & Science University, Portland, Oregon 97201
Submitted 25 November 2002 ; accepted in final form 22 May 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
phlebotomy; end-systolic elastance
There is growing evidence that the prenatal environment can predispose a person to develop endothelial dysfunction and cardiovascular disease later in life (4). It is also clear that decreased fetal oxygen delivery can lead to intrauterine growth retardation (20) and can potentially affect the cardiovascular system later in life. However, it is possible that fetal programming may be an advantage for adult cardiovascular function. For example, an individual with higher maximal coronary conductance might maintain a greater coronary flow during severe hypoxic stress and maintain cardiac performance.
Aoyagi, Flanagan, and colleagues (2, 13) have shown that aortic banding in the adult sheep heart results in decreased maximal coronary conductance, whereas aortic banding in newborn lambs does not change coronary conductance. Taken together, these studies demonstrate that the cardiovascular system of the newborn lamb responds differently to increased systolic pressure load from the adult. Thus a physiological adaptation early in life may confer a beneficial phenotype at least in response to certain stresses in adulthood.
Increased coronary reserve resulting from fetal or neonatal remodeling may be associated with greater cardiac functional reserve as an adult. However, because resting coronary blood flow can increase approximately threefold in the normal adult heart during hypoxic conditions (15), the potential value of an increased flow reserve would not be observed at rest. A physiological difference, if present, would be unmasked only during a cardiovascular stress, when the effects of maximum coronary flow are challenged.
Therefore, we sought to determine whether a 3-wk period of anemia induced in utero would lead to a difference in adult cardiac function in response to acute hypoxemia. We reasoned that fetal anemia would increase coronary reserve in the adult and lead to an improved cardiac functional response to acute hypoxic stress in adulthood. Thus we predicted that sheep hearts remodeled by anemia in utero would better maintain or improve their cardiac function under stress. We also hypothesized that there would be a measurable difference in coronary arteriolar density between control adult hearts and those that were anemic during fetal life.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
Fetal anemia. The day after catheterization, blood gas, oxygen
content, and hematocrit were measured by a 482 CO-oximeter and a 1610 pH and
blood gas analyzer (Instrumentation Laboratories, Lexington, MA) calibrated at
39°C. Fetuses were assigned to two groups on the basis of ear clips and
catheter labels. One twin was selected for isovolumic hemorrhage. In the in
utero anemia group, anemia was induced by withdrawing 25-100 ml of blood daily
and replacing an equal volume of normal saline, as previously described
(11). Arterial pH, blood gas
tensions, oxygen content, and hematocrit were measured before each hemorrhage
and volume replacement. Phlebotomy volumes were chosen to reach a target
hematocrit of <50% of the initial value and to maintain this for at least
10 days, based on our laboratory's prior experience
(9). The total period of anemia
was 20 days. The withdrawn blood was stored in sterile citrate phosphate
dextrose solution, to which penicillin was added. At 
138 days of
gestation, animals were autotransfused with 120 ml of stored packed red blood
cells through a high-efficiency leukocyte filter (Y-type blood set, McGaw,
Irvine, CA) to remove aggregate material. Control animals underwent similar
measurements of hematocrit and arterial blood gas approximately every 4 days.
After completion of the anemic period, the catheters were tied off at the
ewe's flank. All animals were then allowed to continue to spontaneous term
delivery, during which the inserted catheters were pulled out of the fetus
spontaneously. The day after delivery, lambs were weighed, hematocrit and
venous blood gas were measured, and 1 ml of iron dextran containing 100 mg of
elemental iron (Butler, Columbus, OH) was given intramuscularly. One week
later, the lambs and ewes were returned to standard habitation at a local
farm.
Adult protocol. At 7 mo of age, general anesthesia was induced with intravenous ketamine-diazepam to allow tracheal intubation and was then maintained with 2% isoflurane and supplemental oxygen. Temperature, end-tidal PCO2 and percent oxygen saturation were monitored continuously during the study. An 8-Fr catheter was placed in the left common carotid artery and advanced into the aorta. A V8 4-Fr catheter was placed in the left internal jugular vein and advanced to the right atrium. A left thoracotomy was performed, and a 10-mm inflatable vascular occluder (In Vivo Metric Systems, Ukiah, CA) was placed around the intrathoracic inferior vena cava (IVC). The ascending aorta was then mobilized by blunt dissection, and a size 20 or 24A Doppler flow probe (Transonic Systems, Ithaca, NY) was placed around the proximal aorta. A silastic hydraulic occluder was also placed around the aorta distal to the flow probe. The pericardial sac was opened, and a high-fidelity pressure transducer (Konigsberg Instruments, Pasadena, CA) was placed in the left ventricular (LV) chamber through an apical stab wound. A fluid-filled V8 catheter was also placed in the ventricle through the apex to serve as a reference catheter for the Konigsberg transducer. Omnidirectional ultrasound crystals (Sonometrics, London, Canada) were placed at the endocardial layer through myocardial punctures at the short and long axes of the LV. A fifth crystal was placed on the anterior epicardium to determine wall thickness. The thoracotomy incision was covered but remained open during the remainder of the study. Arterial blood pH, PO2 (PaO2), and PCO2 (PaCO2) were measured to ensure adequate ventilation.
Data collection. To minimize cardiac depression, halothane was discontinued, fentanyl (0.03-0.1 mg · kg-1 · h-1 iv) was initiated, and time was allowed for sufficient washout (minimum of 15 min) of the halothane. Throughout the study, a trained animal-care technician, who was not involved in data collection or interpretation, maintained anesthesia at a surgical plane. Right atrial, aortic, and LV hydrostatic pressures were measured with Transpac transducers (Abbott Critical Care Systems, Chicago, IL) that were calibrated by using a mercury manometer and zeroed to atmosphere at the level of the right atrium. LV pressure was measured with the high-fidelity transducer calibrated by using the fluid-filled LV reference. LV pressure, right atrial pressure, and aortic pressure were recorded continuously with a Gould 2800S poly-graph (Gould, Valley View, OH) throughout the experiment, as well as displayed on-line. Signals from ultrasonic crystals, flow meter, and pressure tracings were recorded digitally by using Sonolab measurement system (Sonometrics) at a sample rate of 320 Hz. All data were obtained with the ventilator held at end expiration. Animals underwent autonomic blockade with 4 mg of propranolol iv and 4 mg of atropine iv at least 15 min before any data collection to minimize autonomic responses during the study. First, oxygen content was measured and hemodynamic data were recorded during steady-state conditions. Then, data were recorded during a graded occlusion of the IVC over 10-15 s, and thereafter during a graded occlusion of the aorta to zero flow. Pressure and flow were allowed to return to baseline between and after the IVC and aortic occlusion.
Hypoxic stress. The oxygen supplied to the ventilator was replaced with 7% oxygen-93% nitrogen. After peripheral oxygen saturation had declined to 30% (over a period of 5-7 min), data were again obtained before and during IVC occlusion and aortic occlusion as described above in Data collection. An oxygen saturation of 30% corresponded to a desired oxygen content of 2.5 ml/dl. If the measured oxygen content was not <4.0 ml/dl, the hypoxic period was repeated after a recovery phase to reach the target low oxygen content. If >60 min had elapsed since previous administration of atropine and propranolol, this was repeated before IVC and aortic occlusion.
Heart preparation. After all physiological data had been collected, intravenous heparin and pentobarbital sodium were given. The heart was arrested in diastole by using intravenous potassium chloride, dissected from the chest cavity, trimmed of great vessels, and weighed. A blunt-end steel syringe was sutured into the left main coronary artery, and the atria were closed with suture. A solution of paraformaldehyde (2%) and adenosine (1%) with air bubbles removed was hung 60 cm above the heart to perfuse passively at half mean arterial pressure over 30 min. Hearts were then stored in phosphate buffer at 4°C.
Histological examination. After all animals had been studied, full thickness transections were taken from the midanterior LV wall of each heart. The samples were embedded in paraffin, and 5-µm sections were fixed and stained with hematoxylin and eosin by using standard protocols. Each slide was scanned at a high resolution (2 pixels/µm, Aperio Technologies, Vista, CA) and digital images were viewed by using MrSID Geoviewer software (International Land Systems), which allows on-screen measurement of dimensions and polygonal areas. A micrometer was also scanned and measured identically to ensure calibration.
Arteriolar measurement. All slides were analyzed by a single scientist who was blinded to treatment group. The sample was viewed systematically from epicardium to endocardium, and all arteries and arterioles were identified by size, wall thickness relative to lumen size, and presence of a muscularis layer. The shortest and longest diameter of each vessel lumen were measured. The two diameters were used to calculate the cross-sectional area of an ellipse. Vessels with a minimum diameter of <8 µm were excluded. Total area of tissue scanned was also measured by circumscribing the tissue sample with a polygon. For each animal, we calculated the total number of vessels counted, total cross-sectional area, and length density per square area. Vessels were also categorized into groups based on minimum diameter.
Analysis. Data were reviewed by a single scientist blinded to treatment group and analyzed by using Cardiosoft software (Sonometrics). Ectopic beats were excluded. Heart rate and LV pressure were obtained from stored data. LV volume was calculated by using the short and long axes' crystal dimensions in a two-axis ellipsoid model. The maximum and minimum derivative of pressure with respect to time (dP/dtmax and dP/dtmin, respectively) were calculated from stored pressure tracings. Pressure-volume loops were constructed to calculate LV end-systolic elastance (Emax), defined as the slope of the regression line connecting points at end systole during graded IVC occlusion. Data obtained during graded occlusion of the aorta were used to calculate wall stress by using standard formulas (6). This was plotted against fractional shortening, giving the stress-shortening relationship whose slope and intercept was calculated by linear regression. The time constant of relaxation was calculated by using the least-squares exponential fit during isovolumic relaxation from peak negative (dP/dtmin) to 5 mmHg above end-diastolic pressure LVP(t) = LVP0 · e-t/tau, where tau is the pressure decay coefficient and LVP0 is the LV pressure at end systole. Means and standard deviations were calculated for each group at baseline and hypoxia. For each LV performance index, we calculated the change for each animal that occurred from baseline to hypoxia and compared these differences between the two groups by using a nonparametric rank sum test. For other parameters with apparent post hoc differences, the data were evaluated with ANOVA or two-way ANOVA for repeated measures, followed when indicated by Newman-Keuls inference testing.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
50% of the initial hematocrit for
12.8 ± 2.4 days. In control fetuses, hematocrit was maintained during
gestation (36.8 ± 4.3%), as expected. The gestational age at delivery,
birth weight, and newborn hematocrit were not different between the two groups
(Table 1).
|
At the time of final experimentation, the two adult groups did not differ in age, hematocrit, body weight, or heart weight (Table 1). Both groups had similar hemodynamic parameters at baseline, and these did not change after autonomic blockade (Table 2). Oxygen content was the same at baseline. During hypoxia, both groups reached a similar oxygen content (2.6 ± 1.0 ml/dl in the in utero anemia group vs. 2.5 ± 1.0 ml/dl in the control group).
|
|
|
dP/dtmax was similar in the two groups under basal conditions (Table 2). However, during hypoxia, the control group dP/dtmax tended to decrease (2,036 ± 500 mmHg/s at baseline vs. 1,533 ± 289 mmHg/s in hypoxia). Conversely, dP/dtmax for the in utero anemia group was 1,912 ± 426 mmHg/s at baseline and 2,398 ± 1,111 mmHg/s in hypoxia. The difference for each animal from normoxia to hypoxia was on average -503 ± 471 mmHg/s in controls vs. 486 ± 962 mmHg/s in the in utero anemic animals (P < 0.03).
At baseline, the slope of the stress-shortening relationship was similar between groups (-0.108 ± 0.094 cm2/g in the control group vs. -0.131 ± 0.106 cm2/g in the in utero anemia group) and there was no difference in y-intercept. However, the mean difference in the change of the slope from baseline to hypoxia was -0.143 ± 0.145 cm2/g in the control group vs. 0.038 ± 0.094 cm2/g in the in utero anemia group as shown in Fig. 2 (P < 0.03).
Diastolic function was evaluated by calculating the pressure decay coefficient. No difference was found at baseline or in response to hypoxia in either group, nor between groups (Table 2). Likewise, there seemed to be no effect of perinatal treatment or hypoxia on dP/dtmin, another index of diastolic function.
Histological analysis of tissue samples showed no difference in total number of arterioles, distribution of artery or arteriolar size, luminal area of vessels, or length-density calculations between groups (Table 3).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
We chose Emax as a primary gauge of systolic function because it is a load-independent measure of contractility (21, 22). Emax increased to a greater extent in the in utero anemia group than nonanemic controls. Other functional parameters such as dP/dtmax, although objective and simple to obtain, are load and rate dependent (7). The stress-shortening relationship is independent of load but has its own limitations, including certain geometric assumptions that are required for calculating wall stress. Changes in these two other measures of cardiac function in our study were also supportive of the Emax findings.
Other investigators have shown that myocardial function in response to stress is different when the stress occurs in the perinatal period compared with the adult. Aoyagi et al. (2, 13) found that month-old lambs submitted to 6 wk of aortic banding maintained coronary conductance and LV function, whereas aortic banding in adult sheep decreased maximal coronary conductance and LV shortening indexes. Similarly, these investigators noted that there was no difference in baseline dP/dtmax, yet shortening indexes, such as rate-corrected circumferential shortening, increased above normal in banded lambs and decreased in banded adults.
In this study, we chose to measure cardiac function during acute hypoxia for three reasons. First, hypoxia can be titrated acutely in anesthetized animals. Second, it does not require pharmacological manipulation that may directly alter contractility. Third, it can reliably reduce myocardial oxygen supply to the point where coronary conductance becomes functionally limiting, and would thereby be an ideal method to distinguish differences between our experimental and control groups. Because coronary blood flow is capable of increasing three- to fourfold in response to hypoxia under normal circumstances (14), under resting conditions no differences in ventricular function would be expected. At the other extreme, reduced oxygen content beyond a critical point may cause both groups to deteriorate remarkably. To reach a level at which oxygen delivery would be sufficiently decreased to affect myocardial performance acutely, we estimated that the oxygen content would have to be reduced to less than one-fourth of basal levels. We therefore targeted an oxygen content of 2.5 ml/dl during hypoxia. There was an expected hemoconcentration during hypoxia (see Table 2) but no difference in hemoglobin by post hoc testing between experimental groups.
There are multiple possible mechanisms for our observed findings. We
initially hypothesized that a difference in arteriolar density would account
for the changes in cardiac function seen, as our previous work suggests.
However, our analysis detected no change in arteriolar density, even though
explanted hearts were infused with adenosine, quite possibly because our study
was not sensitive enough to detect a difference. If indeed no difference
exists, other possible explanations are worth exploring. First, changes in
conductance may have been more at the capillary level. Jayaweera et al.
(17) showed that during
hyperemia capillary resistance accounted for roughly three-fourths of total
resistance. Second, differences in maximal coronary conductance in response to
adenosine may be secondary to subcellular mechanisms, such as changes in
endothelial responses to nitric oxide, ATP-sensitive K+ channels
(16), or intracellular calcium
concentration (1). Thus
experimental arterioles may have dilated more than control arterioles during
stress due to prenatally programmed differences in these mechanisms. Third, in
other studies using various models of hypertrophy in young animals, including
aortic banding (5,
12), arteriovenous anastamoses
(8), and anemia
(19), arteriolar density did
not change. Arteriolar growth was presumed to have occurred because arteriolar
density did not decrease with concurrent hypertrophy. Because there was no
difference in heart weight in either group in our study, it is not surprising
that we did not observe a difference in arteriolar density. Fourth, because
our autonomic blockade did not include alpha block, it is possible that
increased sympathetic innervation in the anemic adults may have played a role
during hypoxia. 
-Adrenergic vasoconstriction during adrenergic
stimulation is thought to maintain coronary blood flow to the LV inner wall
during exercise by reducing retrograde coronary blood flow
(18). Perhaps prior anemia
affected -adrenergic regulation of vasoconstriction in ways that became
functionally important during hypoxia.
A number of limitations inherent in this study must be mentioned. First, we studied animals in early adulthood without comorbid problems. Thus we cannot conclude that the physiological patterns we observed would be found in a more geriatric adult. The programming that offers an advantage in early adult life could become increasingly disadvantageous with age (3).
Second, because our adult study was performed under general anesthesia, it is likely that cardiac function could have been suppressed to some degree. Indeed, cardiac output did not increase significantly during acute hypoxia in either group. We attempted to limit the effects of cardiac depression by using narcotic anesthesia in combination with nitrous oxide. However, the lack of increase in cardiac output may have been secondary to a depressed central nervous system hypoxic response often seen with anesthesia (23). Because fentanyl was adjusted by an independent animal-care technician during the study, we do not know whether significant differences in required fentanyl dosage existed between groups but have no reason to so believe.
Third, our study required a sensitive balance of autonomic blockade. Insufficient blockade would cloud efforts to measure changes in contractility because of reflex activation during IVC occlusion, whereas excessive blockade would blunt all response to stress. Indeed, we did not show a statistical difference in hemodynamic parameters before and after atropine or propranolol. The dose of atropine/propranolol we used was based on our own experience in how to best reach this balance.
In summary, we have shown that in utero anemia predisposes the adult to respond to acute hypoxia with increased indexes of LV systolic function. Thus fetal anemia can alter contractile responses to hypoxic stress even in the mature, normocythemic adult. Based on our laboratory's earlier work (9), this may in part be mediated through an increase in maximal coronary conductance. In addition, a number of other possible subcellular mechanisms may play a role. This is relevant to the renewed interest in the fetal origins of adult disease hypothesis, i.e., that the intrauterine environment can program life-long disease (3). We have demonstrated that, during conditions of acute hypoxia, previous perinatal anemia may be beneficial in the adult. Thus some aspects of perinatal programming may even be advantageous at least under specific conditions. It remains to be determined whether these changes persist into old age.
|
DISCLOSURES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
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. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
-Adrenergic vasoconstriction reduces systolic retrograde
coronary blood flow. Am J Physiol Heart Circ Physiol
273: H2746-H2755,
1997.This article has been cited by other articles:
|
|
 |
|
  J. W. Calvert, D. J. Lefer, S. Gundewar, L. Poston, and W. A. Coetzee Developmental programming resulting from maternal obesity in mice: effects on myocardial ischaemia\#8211;reperfusion injury Exp Physiol, July 1, 2009; 94(7): 805 - 814. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Jiang, K. M. Godfrey, C. N. Martyn, and C. R. Gale Birth Weight and Cardiac Structure in Children Pediatrics, February 1, 2006; 117(2): e257 - e261. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. E. Mascio, A. K. Olison, J. C. Ralphe, R. J. Tomanek, T. D. Scholz, and J. L. Segar Myocardial vascular and metabolic adaptations in chronically anemic fetal sheep Am J Physiol Regulatory Integrative Comp Physiol, December 1, 2005; 289(6): R1736 - R1745. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. C. Mcmillen and J. S. Robinson Developmental Origins of the Metabolic Syndrome: Prediction, Plasticity, and Programming Physiol Rev, April 1, 2005; 85(2): 571 - 633. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Zhang Prenatal Hypoxia and Cardiac Programming Reproductive Sciences, January 1, 2005; 12(1): 2 - 13. [Abstract] [PDF]
|
|
|
|
|
|
L. Bennet and A. J. Gunn Memories of the fetal heart Am J Physiol Regulatory Integrative Comp Physiol, September 1, 2003; 285(3): R517 - R518. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) in left ventricular function
[Emax (A), maximal derivative of pressure with respect to
time (dP/dtmax; B), and stress shortening slope
(C)] in response to acute hypoxia (relative to baseline). Solid bars
are control animals, and open bars are the in utero anemic animals. Bars show
means ± SD. Significant difference between groups: *P <
0.03; **P < 0.01.