INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

IN THE ADULT, adaptive growth of the heart in response to hypertrophic stimuli involves functional and morphological changes that are predictable depending on the stimuli (32, 33). Concomitant with the development of cardiac hypertrophy in the adult are qualitative and quantitative changes in gene expression, including upregulation of the so-called fetal cardiac genes, including -myosin heavy chain, -actin, and atrial natriuretic factor, and a reversion to fetal metabolic pathways for energy production (1, 32, 33, 43). In the fetal heart, however, the molecular triggers for increasing ventricular mass in response to increased load are poorly understood and may be quite different from those in the adult, particularly given the ability of the fetal heart to increase mass by both hypertrophy and hyperplasia (2, 25, 29).

In response to increased ventricular load and wall stress, coronary vasodilation occurs to meet the increased metabolic needs of the myocardium. Despite maintaining a greater resting myocardial blood flow than the adult, the late-gestation fetus is able to increase myocardial flow during increases in ventricular pressure, suggesting myocardial blood flow is closely linked to metabolic needs even early in development (8, 29). With prolonged or progressive pressure overload, vascular remodeling, including angiogenesis and capillary proliferation, occurs to varying degrees depending on the maturity of the heart and the nature of the stimulus (43). As the adult heart progresses to hypertrophy, there is a relative decrease in microvascular density and an increase in intercapillary diffusion distance and coronary vascular resistance (43). In contrast, young animals with pressure overload display angiogenesis proportionate to hypertrophy such that capillary density and coronary conductance remain normal (12).

It is well established that a variety of factors influence the process of myocardial vascularization, including the extracellular matrix, mechanical forces, and growth factors (5, 11, 14, 16). Vascular endothelial growth factor (VEGF) is a potent vasculogenic and angiogenic factor that appears to play an important role in compensatory coronary vascular growth. Increased collateral vessel growth in response to coronary occlusion is associated with enhanced expression of VEGF (38). Moreover, gene transfer of VEGF enhances collateral vessel formation and improves myocardial function in the ischemic porcine heart (19). In addition to its role in normal vasculogenesis, VEGF may serve as a signal for physiologically necessary compensatory angiogenesis to ameliorate relative myocardial ischemia induced by cardiac hypertrophy. Myocardial expression of VEGF is upregulated in the hypertrophied hearts of spontaneously hypertensive rats compared with age-matched controls (22) and in several animal models of acute stretch or pressure overload (13, 18).

Little is known about the cellular and molecular "triggers" and responses that regulate compensatory vascular growth in the fetal or neonatal heart and how they differ from those in the adult. Transcriptional control and expression of these genes in the developing myocardium has not been extensively studied. The first objective of this study was to investigate developmental differences in acute regulation of coronary blood flow by defining gestation-related changes in coronary flow and coronary flow reserve after acute increases in right ventricular (RV) pressure in the fetal sheep heart. Our second goal was to determine changes in the expression of select genes that may be involved in the early phase of coronary vascular remodeling and growth in response to increased ventricular load.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals and Surgical Preparation

Pregnant ewes (term 145 days) with twin fetal pregnancies at 126-128 days of gestation (n = 6 ewes) and at 98-102 days of gestation (n = 6 ewes) were used for the study. Anesthesia was induced with 12 mg/kg thiopental sodium (Abbott Laboratories; Abbott Park, IL) and maintained with a mixture of halothane (1%), oxygen (33%), and nitrous oxide. Under sterile conditions, the uterus was opened, and polyethylene catheters (0.86-mm inner diameter, 128-day fetus; 0.58-mm inner diameter, 100-day fetus) were placed into the fetal external jugular vein and common carotid artery and advanced into the right atrium and ascending aorta, respectively. Via a third interspace thoracotomy, the pericardium was incised, and the pulmonary artery (PA) was exposed proximal to the ductus arteriosus. For each twin pair, one fetus had the main PA double wrapped with an umbilical tape ligature to constrict the external diameter of the artery by 50% as measured using fine calipers. Previous studies by us demonstrated this degree of constriction produces a pressure gradient of 12 ± 1 mmHg across the constriction along with a significant increase in RV mass within 7 days (35). A catheter was also placed within the left atrial appendage for injection of fluorescent-labeled microspheres. A fourth catheter was secured to the fetal skin for measurement of amniotic pressure. The fetal chest was closed, and the hysterotomy was repaired. The second twin fetus was exposed through a separate hysterotomy, fully instrumented, and served as a non-PA-banded control. All catheters were exteriorized through a subcutaneous tunnel and placed in a cloth pouch on the ewe's flank. Ampicillin sodium (Wyeth Laboratories; Philadelphia, PA) was administered intra-amniotically at the completion of surgery (2 g) and to the ewe before surgery (2 g). At the completion of surgery, ewes were returned to individual pens and allowed free access to food and water.

Experimental Protocol

Baseline myocardial blood flow was measured using 15-µm fluorescent-labeled polystyrene microspheres (Triton Technology; San Diego, CA). For each blood flow determination, ~1.2 × 10⁶ (128-day fetus) or 1.0 × 10⁶ (100-day fetus) microspheres, suspended in 1 ml of 0.9% NaCl with 0.02% Tween and 0.02% thimerosol, were injected into the left atrium and flushed with 2 ml of 0.9% NaCl over a 30-s period. Beginning 25 s before the microsphere injection, reference blood samples were withdrawn from the carotid catheter (tip in the ascending aorta) into heparinized glass syringes for 2.5 min at a rate of 2 ml/min. After a 15-min recovery period, a continuous right atrial infusion of adenosine (150 µg · kg¹ · min¹) was started. Preliminary studies using Doppler flowmeter measurements of blood flow velocity in the left circumflex artery demonstrated that this concentration achieved maximal coronary vasodilation (Fig. 1). After 10 min of continuous adenosine infusion, hemodynamic and myocardial blood flow measurements (using different fluorescent colored microspheres) were repeated without interruption of adenosine administration. This procedure was first performed in the PA-banded fetus and then repeated in the control fetus.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Change in coronary blood flow velocity (expressed as the %change from baseline and measured by Doppler signal from a flow probe placed on the left circumflex artery) during graded infusion of adenosine into the right atrium of 100- and 128-day (d) fetuses (n = 3 for each group). The %change in coronary blood flow velocity was calculated from the following formula: [Doppler shift (in kHz) during vasodilation with adenosine  Doppler shift (in kHz) at baseline]/Doppler shift (in kHz) at baseline.

Upon completion of the physiological studies, the ewe was returned to the surgical area, and, under general anesthesia, the fetuses were exteriorized individually. The fetus was euthanized with an overdose of pentobarbital sodium, and the body weight was recorded. Hearts were removed for determination of total weight and RV and left ventricular (LV) free wall weights. Tissue from the RV and LV free walls obtained approximately midway between the apex and the atrioventricular groove were isolated, and duplicate samples of the epicardium and endocardium were sectioned and stored in preweighed glass vials for subsequent blood flow analysis. The remaining RV and LV free walls were sectioned, snap-frozen in liquid nitrogen, and stored at 80°C.

Laboratory Methods

Microsphere recovery. Blood and tissue sample digestion and filtration recovery of microspheres were performed using previously published methodologies ("Manual for Using Fluorescent Microspheres to Measure Regional Organ Perfusion," Fluorescent Microsphere Resource Center, University of Washington; http://fmrc.pulmcc.washington.edu/frmc/fmrc.html). Briefly, tissue samples were allowed to digest overnight at room temperature in 5 ml of 4 N KOH/g tissue. The blood reference samples were digested in a solution of 16 N KOH (1 ml/4 ml blood). After digestion, the microspheres were filtered from the solutions using 10-µm filter membranes (Triton Technology). Microspheres were dissolved with 1,000 µl of Cellosolve acetate (Fisher Scientific; Fair Lawn, NJ), 200 µl of solution was transferred to a wellplate, and fluorescence was determined. Fluorescent measurements of experimental and "standard curve" samples were determined with a luminescence spectrophotometer (LS-50B, Perkin-Elmer; Wellesley, MA) using the appropriate excitation/emission wavelengths and slit widths for the given colored sphere. With the use of the appropriate mix of fluorescent colors, quenching of the fluorescent signal and spillover of the signals to adjacent colors have little effect on flow determinations (Fluorescent Microsphere Resource Center, University of Washington). Standard curves were generated to determine the amount of fluorescence intensity per microsphere. This process also allowed for testing of the linearity of the fluorescent signal. The intensity of the fluorescence is thus proportional to the number of microspheres in the sample. Tissue blood flow (using ascending aorta reference samples) was calculated using the following formula

(1)

where Q_sample and Q_ref are the blood flows (in ml/min) in the specific sample and reference sample, respectively, and F_sample and F_ref are the fluorescent intensities in the specific sample and reference sample, respectively. These techniques have been validated by a number of investigators in the chronically catheterized fetal sheep model (4, 40). The blood flow values from the duplicate samples for each region were averaged together to obtain the final value.

Preparation of ovine-specific partial cDNA clones. Primers for each of the genes listed in Table 1 were designed from consensus regions of previously published sequences found in GenBank. For the three genes listed, no ovine sequences were available. All PCR primers were designed using the Primer 3 program (http://www.genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi) and synthesized by Integrated DNA Technologies (Coralville, IA). All PCR reactions were performed using GIBCO-BRL Superscript One-Step RT-PCR for Long Templates (Life Technologies; Guttenberg, MD). Starting RNA was obtained from the fetal LV free wall as described in Isolation of RNA. The PCR product was purified with a QIAQuick PCR Purification Kit (Qiagen; Valencia, CA), ligated into the TopoPCR4 vector, and cloned into Top10F' cells (InVitrogen; Carlsbad, CA). The plasmid was isolated and purified using a Qiagen Plasmid MaxiKit and sequenced to determine its orientation in the TopoPCR4 vector.

Isolation of RNA. Total cellular RNA was isolated from 250 mg tissue using a modification of the RNEasy Midi Kit (Qiagen), which entailed tissue homogenization in 5 ml of 7.2 µl/ml 2-mercaptoethanol in 4 M guanidine thiocyanate. RNA was quantitated spectrophotometrically by absorbance at 260 and 320 nm. RNA samples were dissolved in diethylpyrocarbamate-treated water and stored in ethanol at 80°C until further analysis.

Northern blot hybridization. The technique used for Northern blot hybridization has been described previously (34). Briefly, a 1% agarose-6.3% formaldehyde gel was prepared in 20 mM MOPS, 5 mM sodium acetate, and 0.25 mM EDTA (pH 7.0). Approximately 10 µg total RNA were applied to each well and electrophoresed overnight in a running buffer of 40 mM MOPS, 10 mM sodium acetate, and 0.5 mM EDTA (pH 7.0). The gels were stained with ethidium bromide and photographed as a reference for RNA sizing. RNA was then transferred to a 0.45-µm Nytran membrane (Schleicher and Schuell; Keene, NH). The membrane were prehybridized for 1 h at 65°C in a solution containing 50% formamide, 5× SSPE (875 mM sodium chloride, 50 mM sodium phosphate, and 5 mM EDTA), 5× Denhardt's reagent, 0.5% SDS, and 200 µg/ml denatured salmon sperm DNA. Hybridization was performed at 65°C for 12-18 h with the addition of 2 × 10⁶ counts · min¹ · ml¹ radiolabeled probe. Four washes were performed at 65°C using three low-stringency washes (1× SSPE and 0.5% SDS) and one high-stringency wash (0.1× SSPE and 0.5% SDS). Hybridization signals were quantitated using a phosphorimager (Molecular Dynamics). Blots were then stripped and rehybridized with ³²P-labeled probe to the 28S subunit of rRNA. Signals from the 28S-probed blots were used to correct for variable RNA loading.

Quantitative immunoblots. A monoclonal antibody to hypoxia inducible factor (HIF)-1 (amino acids 432-528, Novus Biologicals; Littleton, CO) was obtained for immunoblot studies. Immunoblots were prepared as previously described (34). Briefly, the LV or RV myocardium was homogenized in the presence of protease inhibitors including soybean trypsin inhibitor, leupeptin, and phenylmethylsulfonyl fluoride in 50 mM Tris-10 mM EDTA-150 mM NaCl-0.1% mercaptoethanol and then sonicated for 20 s. After centrifugation, total protein of the supernatant was quantitated spectrophotometrically. Protein (20 µg) was separated by SDS-PAGE and transferred to a nitrocellulose membrane. Membranes were blocked with 5% nonfat milk protein for 1 h and then incubated in primary antibody overnight at 5°C. After washes with EDTA-0.5% Tween in phosphate-buffered saline, incubation of secondary antibody conjugated with horseradish peroxidase at room temperature was performed for 1 h. Detection of the secondary antibody using the Pierce SuperSignal Kit was performed and used to expose Kodak XAR film at room temperature. Films were digitized and signals quantitated using NIH Image (Wayne Rasband, NIH). Serial protein dilutions were tested with each antibody to assure quantitated signals were in the linear range for added protein.

Data Analysis

Comparisons among the different groups were performed using two-way ANOVA factoring for gestational age (100 vs. 128 days) and treatment group (PA band vs. no PA band). Comparisons within a gestational age were performed by one-way ANOVA. When the ANOVA indicated a significant difference among groups, as indicated by the F statistic, comparison among means was performed using Tukey's F-test. Statistical significance was defined as P < 0.05, and all data are expressed as means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Ovine-Specific cDNA Probes

Fetal Hemodynamics, Arterial Blood Gas Values, and Weights

Effects of Gestational Age on Myocardial Blood Flow

Effects of PA Banding on Myocardial Blood Flow

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Endocardial blood flow in control and pulmonary artery (PA)-banded (PAB) fetuses at baseline and during adenosine administration (Adeno). A: responses in 100-day gestation fetal sheep (term 145 day). B: responses in 128-day gestation fetal sheep. PAB fetuses had the PA diameter constricted by 50%. Values were obtained 6 h after banding or sham procedure was performed. Values are means ± SE; n = 6 for each group. *P < 0.05 compared with control values for corresponding ventricle and gestational age; P < 0.05 compared with all other ventricle-specific flows in the same age group. All left ventricular (LV) values are significantly less than corresponding right ventricular (RV) values for both age groups (P < 0.05, ANOVA).

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Ventricle-specific endocardial blood flows in 128- and 100-day gestation fetuses expressed as percentages of control values (defined as baseline flows in control animals for specific gestational ages) at baseline and during adenosine administration. PAB fetuses had the pulmonary artery diameter constricted by 50%. Values were obtained 6 h after banding or sham procedure was performed. Values are means ± SE; n = 6 for each group. *P < 0.05 compared with control value; P < 0.05 compared with all other values.

Effects of Adenosine on Myocardial Blood Flow

1

1

1

1

Effects of PA Banding on Selected Gene and Protein Expression

View larger version (36K): [in this window] [in a new window]   Fig. 4.   Steady-state mRNA levels of vascular endothelial growth factor (VEGF) in 128-day gestation fetal ovine RV (left) and LV myocardium (right) in control (Ctrl) and PAB (Band) fetuses. Membranes (top) were sequentially probed with an ovine-specific probe for VEGF mRNA and 28S rRNA, and the resulting radioactive signal was quantitated to determine the abundance of RNA (bottom). The VEGF signal was normalized to the 28S signal. Northern blot analysis was done in triplicate with the values for each paired tissue averaged, n = 6 for each group. Values are means ± SE. *P < 0.05 compared with control for the same ventricle.

View larger version (37K): [in this window] [in a new window]   Fig. 5.   Steady-state mRNA levels of Flt-1 in 128-day gestation fetal ovine RV (left) and LV myocardium (right) in control and PAB fetuses. Membranes (top) were sequentially probed with an ovine-specific probe for Flt-1 mRNA and 28S rRNA, and the resulting radioactive signal was quantitated to determine the abundance of RNA (bottom). The Flt-1 signal was normalized to the 28S signal. Northern blot analysis was done in triplicate with the values for each paired tissue averaged, n = 6 for each group. Values are means ± SE. *P < 0.05 compared with control for the same ventricle.

View larger version (32K): [in this window] [in a new window]   Fig. 6.   Steady-state mRNA levels of Flk-1/KDR in 128-day gestation fetal ovine RV (left) and LV myocardium (right) in control and PAB fetuses. Membranes (top) were sequentially probed with an ovine-specific probe for Flk-1/KDR mRNA and 28S rRNA, and the resulting radioactive signal was quantitated to determine the abundance of RNA (bottom). The Flk signal was normalized to the 28S signal. Northern blot analysis was done in triplicate with the values for each paired tissue averaged, n = 6 for each group. Values are means ± SE.

Because HIF-1 and TGF-₁ have been implicated in regulation of the expression of VEGF, we determined the effects of PA banding on the expression of these factors in 128-day gestation animals. Both RV HIF-1 protein expression and TGF-₁ mRNA levels were unaltered by PA banding (Fig. 7).

View larger version (32K): [in this window] [in a new window]   Fig. 7.   Top: representative immunoblot for hypoxia inducible factor (HIF)-1 protein in 128-day gestation fetal ovine RV myocardium from control (C) and PAB (B) fetuses. Membranes were stained with Ponceau S to assure equal protein loading. HIF-1 protein levels were detected using standard immunoblot techniques, and the resulting fluorescence signal was quantitated by densitometry. No significant differences in protein levels were detected between groups (n = 6 for each group, P > 0.5). Bottom: representative Northern blot analysis of RV myocardial RNA hybridized with ovine-specific transforming growth factor (TGF)-1 RNA and 28S rRNA probes. Myocardial samples were obtained from 128-day gestation control and PAB fetuses (n = 6 for each group, arranged by twin pairs). Membranes were sequentially probed for TGF-1 RNA and 28S rRNA, and the resulting radioactive signal was quantitated to determine the abundance of RNA. The TGF-1 signal was normalized to the 28S signal. Northern blot analysis was done in triplicate and the values for each tissue sample were averaged. No significant differences in mRNA levels were detected between groups (P > 0.4).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The mechanisms regulating myocardial blood flow in the fetal heart are not well understood. The myocardial flow data obtained from the study of late-gestation fetal sheep are consistent with the findings reported by others (8, 29) and confirm that 1) RV blood flow, expressed per gram of myocardium, exceeds LV blood flow; 2) endocardial blood flow exceeds epicardial blood flow; and 3) acute RV pressure load increases both RV and LV blood flow. Novel findings of the present study relate specifically to the determination of myocardial blood flow in the 100-day gestation fetus, including 1) myocardial blood flow is greater in young compared with older fetuses, although the predominance of RV to LV blood flow is still present; and 2) RV and LV myocardial blood flow increases in the young fetal heart in response to acute increases in RV systolic load. However, whereas acute pressure loading increases RV blood flow in both the young and old fetuses, pharmacological coronary vasodilation with adenosine further increases RV blood flow in older but not younger fetuses. Thus early fetal hearts may be operating at their maximal coronary flow capacity, possible limiting the ability of the early-gestation fetal heart to respond to increased loads. The finding that early-gestation fetuses have no demonstrable coronary flow reserve after RV pressure loading suggests that factors responsible for creating/preserving coronary flow reserve are developmentally regulated and mature late in gestation.

The limited body of information regarding coronary flow regulation in fetal sheep has recently been reviewed (41). Early studies demonstrated that resting myocardial blood flow is greater in the late-gestation fetus than in the adult sheep (8, 10). This difference appears related to the lower oxygen content of fetal blood because myocardial oxygen consumption is similar (9). In contrast to the newborn and adult, fetal RV blood flow is higher than LV flow. This finding is not surprising given the increased workload and resting wall stress of the fetal RV compared with the LV (27). Our findings in the late-gestation fetal sheep as they pertain to RV and LV free wall blood flows and the predominance of RV to LV flow as well as endocardial to epicardial blood flow are consistent with those of other investigators (8, 29).

The studies in the 100-day gestation fetus demonstrate that at this stage of development, a pattern of myocardial blood flow similar to the older fetus exists. A notable finding, however, is that baseline RV and LV blood flows are significantly higher in the younger compared with the older fetuses. The reasons for this difference are unclear. Although precise measurements were not made, we estimate that basal coronary vascular conductance is markedly greater in the 100- compared with 128-day fetus. Assuming right atrial pressures are relatively similar in both age groups (4 mmHg), then coronary conductance in the RV endocardium [calculated as RV endocardial blood flow/(mean carotid artery pressure  right atrial pressure)] is ~10.6 ml · min¹ · 100 g¹ · mmHg¹ [341 ml · min¹ · 100 g¹/(36  4 mmHg)] in the 100-day fetus and 5.9 ml · min¹ · 100 g¹ · mmHg¹ [230 ml · min¹ · 100 g¹/(43  4 mmHg)] in the 128-day fetus. Blood viscosity, which significantly influences fetal coronary blood flow and conductance (47), likely contributes little to the observed developmental differences because the hematocrit was similar in the two gestational age groups. Oxygen-carrying capacity and hemoglobin-oxygen affinity are relatively similar in the two age groups and therefore unlikely to significantly influence our results (24). It is possible that myocardial metabolism and oxygen consumption are greater in the early- compared with late-gestation fetal heart, although, to our knowledge, no studies have been performed to investigate this. The roles of other determinants of fetal coronary flow resistance, including shear-stress dependent vasodilation, neural and humoral factors, and blood vessel microdomains, have yet to be thoroughly evaluated in the fetus. Differences may also exist at the anatomic level, although myocardial capillary luminal area remains constant between 0.78 and 0.97 of gestation (39). Similar measurements in the 0.6 gestation fetus (100 days of gestation) have not been made.

Previous studies have investigated the response in myocardial blood flow to increased RV systolic pressure load in late-gestation fetal sheep. Reller et al. (29) demonstrated that acute (10-30 min) increases in RV systolic pressure load were associate with increased myocardial blood flow. LV blood flow also increased with changes in the RV pressure; however, no change in the ratio of endocardial to epicardial flow occurred with increasing load in either ventricle. Similar results were seen in the late-gestation fetuses we studied. However, we found that infusion of adenosine into the nonloaded fetal heart resulted in an increase in blood flow similar to that seen with PA banding, whereas Reller et al. (29) found that adenosine increases RV blood flow significantly more than a pressure load. These differences in findings may be related to studying animals of slightly different gestational ages, the degree of load imposed on the ventricle, and duration of the RV systolic load (6 h in the present study vs. 10-30 min). Finally, we infused adenosine into the right atrium, as opposed to the left atrium, and may not have achieved complete pharmacological coronary vasodilation. However, our preliminary studies in fetuses of both ages demonstrated that doses of adenosine in excess of that used in the present study did not result in further increases in steady-state coronary blood flow velocity in the proximal left circumflex coronary artery (Fig. 1).

In the late-gestation fetal heart subjected to increased load, blood flow was further augmented by administration of adenosine. Thus a coronary flow reserve was present after an increased pressure load created by PA banding. In contrast, in the 100-day PA-banded fetus, adenosine failed to further increase myocardial blood flow, suggesting an absence of coronary flow reserve in the loaded ventricle of these animals. Whereas it is possible that the attenuated adenosine-induced increase in myocardial blood flow in the younger fetus is related to maturational differences in the sensitivity of the coronary vascular bed to adenosine, we believe that this is not the case. First, our own data demonstrated maximal coronary vasodilation occurred at a lower dose of adenosine in the 100- compared with 128-day fetus (Fig. 1). Second, in separate studies, Matherne et al. (21) and Toma et al. (42) found in the guinea pig heart that the immature coronary vascular bed exhibited greater maximal response to infused adenosine. Vasodilatory adenosine receptors have also been shown to be functional in the brain vasculature in fetal sheep at 0.6 gestation (~100 days) (17). Thus there appears to be a distinct developmentally related difference in the coronary blood flow response to increased load in the fetal heart, suggesting that the factors responsible for preserving coronary flow reserve mature late in gestation.

Whereas resting myocardial blood flow was greater in the 100- compared with 128-day gestation fetus, the RV flows (in absolute value, ml · min¹ · 100 g¹) were relatively similar after imposition of the RV load and during adenosine infusion. When the changes in blood flow are expressed as percentage of baseline values in non-PA-banded animals, it is apparent that the early-gestation fetus is unable to increase myocardial blood flow to the extent seen in the older fetus. This finding suggests that the heart of the younger fetus may be more susceptible or at risk for impaired function or failure from myocardial ischemia resulting from increased load.

For the heart to successfully adapt to pressure overload resulting in myocardial hypertrophy, coronary growth must also occur to meet the increased oxygen and nutrient requirements of the myocardium. However, adult animals display attenuated myocardial capillary growth compared with young postnatal animals with pressure-overload hypertrophy (43, 45, 46). In the young, concurrent growth of the myocardium and coronary microvasculature is present, such that normal capillary density and coronary blood flow reserve are preserved (43). Little is known about the effect of increased load on myocardial vascularization before birth. Recently, Tomanek et al. (44) reported that embryonic chick heart myocardial vascularization is proportionately accelerated relative to the increase in ventricular mass resulting from pressure overload. To our knowledge, there are no published data reporting changes in vascularization in the pressure-loaded fetal sheep heart, although A. M. Rudolph (unpublished observations) found that capillary diameter is increased at least twofold and capillary density is reduced in both the RV and LV of fetuses with chronic pulmonary stenosis (31). Consistent with findings in animals, Rakusan et al. (28) demonstrated in humans that the hearts of children with pressure-overload LV hypertrophy resulting from congenital aortic stenosis and coarctation of the aorta display proportional angiogenesis, whereas adults with pressure-overload hypertrophy display diminished coronary capillary density. Elucidating the mechanisms controlling these developmentally different coronary angiogenic responses has obvious clinical relevance.

A number of growth factors, including VEGF, have been identified that influence the process of coronary vascularization and angiogenesis. VEGF is a 46-kDa heparin-binding glycoprotein that enables basement membrane degradation, increases microvascular permeability, and induces endothelial cell proliferation and migration (7, 23). The effects of VEGF are mediated through specific phosphotyrosine kinase receptors, the two best described being Flt-1 (VEGFR1) and KDR/Flk-1 (VEGFR2) (6). In the current study, we observed an ~50% increase in RV mRNA levels for VEGF and Flt-1 after 6 h of pressure overload in the 128- but not 100-day fetus. Studies in young pigs have similarly shown increases in VEGF and Flk-1 mRNA within several hours of RV pressure overload induced by PA banding (3). Whether this increase in mRNA results from increased transcription or mRNA half-life was not determined. Nonetheless, the increased RV VEGF and Flt-1 mRNA levels in the 128-day fetus suggest that acute load imposed on the RV by PA banding induces an early angiogenic response. Whether induction of vascular growth-promoting genes confers a morphological response and physiological advantage remains to be determined.

The mechanism(s) by which VEGF and Flt-1 expression is upregulated by pressure overload is unclear, although tissue hypoxia, resulting from increased myocardial work, may be involved. A number of studies have suggested that VEGF and Flt-1 mRNA levels are upregulated by hypoxia and that activation of gene transcription is likely mediated by HIF-1 (20, 37). In its active form, HIF-1 exists as a heterodimer of HIF-1 and HIF-1 that binds to DNA at specific sites with a hypoxia response element. HIF-1 appears to be constitutively expressed, whereas the biological activity of HIF-1 appears to be mediated by HIF-1. Target genes activated by HIF-1 include genes whose protein products are involved in energy metabolism, erythropoiesis, cell proliferation, angiogenesis, and vascular remodeling, including VEGF and Flt-1 (for a review, see Ref. 37). We therefore hypothesized that the increased levels of VEGF and Flt-1 would be associated with increased expression of HIF-1 protein. Cardiac HIF-1 is upregulated in fetal sheep subjected to isovolemic anemia, suggesting it is sensitive to hypoxia even in the low oxygen state of the in utero environment (20). However, we were unable to demonstrate a change in HIF1- protein expression, suggesting factors other than HIF-1 are responsible for the upregulation of VEGF and its associated receptor, FLT-1, observed in our study. The lack of change in HIF-1 expression in the RV also suggests that myocardial hypoxia does not occur with the degree of RV systolic pressure load generated in this model.

Mechanical stretch also enhances VEGF mRNA expression, which appears to be mediated in part by TGF-₁ (18, 36). Although cardiac TGF-₁ expression has been shown to increase in response to pressure overload (3), we saw no changes in myocardial TGF-₁ mRNA levels after 6 h of RV systolic load. However, stretch-induced release of stored TGF-₁ may still participate in upregulating the VEGF response (18).

There are several possible explanations for the lack of change in RV VEGF and Flt-1 mRNA levels in the 100-day fetuses subjected to pressure overload. First, the degree of stimulus (i.e., pressure load) may have been less than that observed in the older fetuses and insufficient to generate a response at the transcriptional level. Against this reason is the observed large increase in myocardial blood flow suggesting a significant physiological load occurred with the banding procedure. Second, growth factors that initiate and maintain vascular growth in the heart may already be maximally expressed in the midgestation fetus. Finally, expression of VEGF and Flt-1 may be preprogrammed and uncoupled to transcriptional regulation at this stage of development.

Several limitations in the present study are noted. First, coronary perfusion pressure was not controlled or measured. Although adenosine infusion had no significant effect on MABP and other investigators have seen no effect of adenosine on right atrial pressure (29), small differences in coronary perfusion pressure (aortic blood pressure  RA pressure) may have been present. Second, results may have been influenced by nonbasal physiological conditions because the studies were performed on the day of surgery. Nonetheless, because the study conditions were similar among all groups of animals, we believe comparisons among and between the different age groups are valid.

In summary, this study demonstrates significant developmentally related differences in the adaptive response of the fetal heart to acute systolic pressure load. Coronary flow reserve is preserved in 128- but not 100-day gestation fetuses after PA banding. Furthermore, acute increases in RV pressure load upregulated factors believed to be involved in angiogenesis in the near-term but not younger fetuses. Factors responsible for preserving coronary flow reserve after acute pressure loading are not present early during development but mature late in gestation. These findings may have important implications related to the compensatory changes in the myocardium of congenitally abnormal hearts as well as understanding the divergent angiogenic responses seen in young compared with adults hearts with pressure-overload hypertrophy.