Vascular nitric oxide and superoxide anion contribute to sex-specific programmed cardiovascular physiology in mice

Abstract

Intrauterine environmental pertubations have been linked to the development of adult hypertension. We sought to evaluate the interrelated roles of sex, nitric oxide, and reactive oxygen species (ROS) in programmed cardiovascular disease. Programming was induced in mice by maternal dietary intervention (DI; partial substitution of protein with carbohydrates and fat) or carbenoxolone administration (CX, to increase fetal glucocorticoid exposure). Adult blood pressure and locomotor activity were recorded by radiotelemetry at baseline, after a week of high salt, and after a week of high salt plus nitric oxide synthase inhibition (by l-NAME). In male offspring, DI or CX programmed an elevation in blood pressure that was exacerbated by N^ω-nitro-l-arginine methyl ester administration, but not high salt alone. Mesenteric resistance vessels from DI male offspring displayed impaired vasorelaxation to ACh and nitroprusside, which was blocked by catalase and superoxide dismutase. CX-exposed females were normotensive, while DI females had nitric oxide synthase-dependent hypotension and enhanced mesenteric dilation. Despite the disparate cardiovascular phenotypes, both male and female DI offspring displayed increases in locomotor activity and aortic superoxide production. Despite dissimilar blood pressures, DI and CX-exposed females had reductions in cardiac baroreflex sensitivity. In conclusion, both maternal malnutrition and fetal glucocorticoid exposure program increases in arterial pressure in male but not female offspring. While maternal DI increased both superoxide-mediated vasoconstriction and nitric oxide mediated vasodilation, the balance of these factors favored the development of hypertension in males and hypotension in females.

the seminal work of barker and colleagues (1) identified a strong inverse relationship between weight in infancy and death from ischemic heart disease. Following replication of these findings worldwide, it is now accepted that decreased perinatal growth is an independent risk factor for the development of adult-onset hypertension, diabetes, and obesity (5, 15). Elucidation of the mechanisms by which perinatal physiological stressors exert long-term cardiovascular effects has worldwide public health implications.

Studies in a variety of animal species have demonstrated that fetal programming of adult metabolic and cardiovascular disease can be induced by dietary manipulation (21). Excessive fetal exposure to maternal glucocorticoids may mediate the interaction between nutritional deficiencies and the programming of hypertension (19), although some investigators refute this (39). Mechanistic links between maternal undernutrition, fetal glucocorticoid exposure, and subsequent programmed hypertension have been bolstered by studies in rats that have shown diet-induced programming can be blocked through inhibition of maternal glucocorticoid synthesis and replicated by direct administration of synthetic glucocorticoids during the last week of pregnancy (3, 11, 12).

We previously characterized the cardiovascular effects of both maternal low-protein diet and late-gestation dexamethasone administration in an isogenic mouse model. Although effects on blood pressure were not apparent by tail cuff blood pressure measurements, both maternal dietary intervention and dexamethasone administration programmed aortic reactivity (32). Of note, resistance vessel reactivity was not assessed, and the study was not powered to evaluate for sex-specific effects (32). With emerging data suggesting 1) female offspring are partially protected from programmed hypertension (9, 22, 30, 35), 2) programmed cardiovascular disease may be mediated by alterations in vascular superoxide anion and nitric oxide production (7, 8, 10, 40), and 3) the hemodynamic consequences of fetal steroid exposure may depend on postnatal influences (25, 37, 38), we used radiotelemetric recordings of arterial pressure and measures of vascular nitric oxide and reactive oxygen species (ROS)-dependent physiology to better characterize programmed murine cardiovascular phenotypes under a broad range of physiological stressors. To better define the role of nitric oxide in programming, arterial pressures were recorded during sequential administration of a high-salt diet and the nitric oxide synthase inhibitor N^ω-nitro-l-arginine methyl ester (l-NAME) (16).

For this study, the programming effects of fetal malnourishment or transplacental glucocorticoid exposure were evaluated by maternal administration of a high-carbohydrate, low-protein diet or the 11-β-hydroxysteroid dehydrogenase inhibitor carbenoxolone, respectively. We demonstrate that maternal dietary intervention (DI) or carbenoxolone administration (CX) program increases arterial pressure and ROS-mediated vascular dysfunction in male mice. For the first time, we further show that 1) female offspring of DI dams have enhanced mesenteric dilation and nitric oxide synthase-dependent hypotension, 2) maternal CX or DI reduces cardiac baroreflex sensitivity in female offspring, and 3) maternal DI increases offspring locomotor activity.

METHODS

Animal model.

All procedures were approved by the University of Iowa Animal Care and Use Committee. The investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication No. 85-23, revised 1996). Adult C57BL/6J mice (Jackson Laboratory, Bar Harbor, ME) were bred until discovery of a maternal plug. Dams were then randomized to either a control diet (18% protein, 6% fat) or a dietary intervention (DI) involving isocaloric substitution of protein with carbohydrate (9% protein, 6% fat), as previously described (31). Some dams on the control diet received daily (embryonic days 12–19) subcutaneous injections of saline or the 11β-hydroxysteroid dehydrogenase inhibitor carbenoxolone (CX; 12.5 mg/kg; Sigma, St. Louis, MO). After delivery, all mice received the control diet. Adult offspring caloric and water intake were measured during telemetry recordings.

Telemetry recordings.

At 5–6 mo of age, carotid radiotelemetry catheters (PA-C10; Data Sciences International, St. Paul, MN) were implanted as previously described (4). For these catheters, manufacturing specifications tolerate an initial offset of less than 3 mmHg and a pressure drift of less than 2 mmHg/mo. In our experience, the routine use of new transmitters for a total implantation time of less than a month produces a nominal offset drift of less than 2 mmHg. After a 7-day recovery period, arterial pressures, heart rate and relative locomotor activity were sampled over 10 s every 5 min for 60 h (encompassing 3 dark cycles and 2 light cycles). Immediately thereafter, arterial waveforms were recorded at 2,000 Hz to calculate spontaneous baroreflex activity, as previously described (13). This sampling occurred at the end of the light cycle, during a 10-min phase of voluntary activity, and event frequency was normalized to 1,000 beats. Baroreflex events were defined as reciprocal changes in blood pressure and heart rate occurring over the course of four cardiac cycles.

Immediately following these baseline recordings, feeds were supplemented to 9% NaCl. From days 4 to 7 of high salt supplementation, radiotelemetry recordings were repeated. Thereafter, while mice continued to receive salt supplementation, the nitric oxide synthase inhibitor l-NAME (1 mg/ml) was added to the drinking water. From days 4 to 7 of l-NAME, a third set of scheduled recordings were obtained. This was followed by 2,000-Hz recordings for reassessment of spontaneous baroreflex sensitivity.

During each of the three recording epochs (baseline, +high salt and +l-NAME), data were sampled every 5 min throughout three dark cycles and two light cycles. With 12 replicates completed each hour for each mouse, this schedule allowed for the averaging of 36 values for each hour of the dark cycle and 24 values for each hour of the light cycle. After calculating mean hourly values for each mouse, descriptive and inferential statistics were calculated.

Mesenteric reactivity.

A second group of control and DI offspring never instrumented for blood pressure recording were killed at 6 mo of age. Second-generation branches of the superior mesenteric artery (∼100 μm internal diameter) were isolated and mounted for wire myography (model 610 M; Danish Myo Technology, Aarhus, Denmark). Passive tension was set at 1.5 mN. Arteries were initially constricted with KCl (90 mmol/l) to provide a standard response for normalization of subsequent responses to noradrenaline (10⁻⁹ to 10⁻⁵ M), ACh (10⁻⁹ to 10⁻⁶ M) or sodium nitroprusside (SNP; 10⁻⁹ to 10⁻⁵ M). ACh concentration-response curves were obtained following preconstriction with 10⁻⁵ M noradrenaline. Some arteries were preincubated in l-NNA (10⁻⁴ M) and/or polyethylene glycol conjugates of superoxide dismutase (SOD; 50 units/ml) and catalase (250 units/ml) throughout the experiment.

Chemiluminescence.

The aortic arch was excised from the noninstrumented group of control and DI-exposed offspring. Basal superoxide anion production was measured by lucigenin (25 μmol/l)-enhanced chemiluminescence as previously described (31). NADPH oxidase-dependent superoxide production was measured as the diphenylene-iodonium (10⁻⁴ mol/l)-inhibitable chemiluminescence measured after the addition of the enzyme substrate NADPH (10⁻⁴ mol/l).

Data analysis.

All values are presented as means ± SE. Locomotor activity, blood pressure, and heart rate data were compared by three-way ANOVA, factoring for time of day, prenatal intervention, and sex. When a significant interaction was present between the programming stimulus and sex (F < 1), two-way repeated-measures ANOVA was utilized to evaluate for independent effects. Post hoc analysis (Holm-Sidak test) was performed if statistically significant differences were detected. All other statistical comparisons were performed by t-test or Fisher exact test, as appropriate. A value of P < 0.05 was considered significant. All analyses were performed using SigmaStat 3.0 (SPSS, Chicago, IL).

RESULTS

Growth characteristics.

Dams receiving the DI had increased feed intake and thus increased total caloric intake during pregnancy (Table 1). Accounting for this difference in intake, DI dams consumed about 36% less protein, 35% more carbohydrate, and 34% more fat than control mice. Although this complex diet intervention did not alter maternal weight gain, DI offspring were significantly smaller than control offspring. While nursing, DI dams had significantly increased weight gain, and their offspring achieved weanling weights approximating control counterparts, consistent with brisk catchup growth. While CX did not alter dam feed intake or pup weight at 3 days of age, pup weight at weaning was significantly reduced.

Table 1.

Growth parameters for dams and their offspring following gestational exposure to control diet, carbenoxolone, or dietary intervention

	Control	CX	DI
Litters (n)	7	8	9
Dam baseline weight, g	21±1	20±1	20±1
Feed intake during pregnancy, kcal·kg−1·day−1	359±29	398±13	468±14*
Weight gain during pregnancy, g	12.7±1.1	12.9±1.1	12.2±0.9
Litter size (pups)	6.6±0.6	7.1±0.4	6.9±0.5
Pup weight, postnatal day 3, g	2.2±0.2	1.9±0.1	1.7±0.1*
Dam feed intake during lactation, kcal·kg−1·day−1	1116±187	1285±39	1175±133
Dam weight gain during lactation, g	3.8±1.1	3.3±1.5	5.1±1.5*
Weanling weight, postnatal day 21, g	10.5±0.4	8.4±0.5*	10.5±0.6

CX, carbenoxolone; DI, dietary intervention.

^*P < 0.05 vs. control.

DI or CX did not significantly alter adult offspring weight or the postmortem weight of the heart, brain, liver, or kidneys (Table 2). CX increased adult male caloric intake, while both DI and CX increased adult male fat pad weights (Table 2). In contrast, DI and CX did not alter adult female caloric intake, but both DI and CX decreased adult female fat pad weights (Table 2).

Table 2.

Growth parameters for adult offspring of dams that received control diet, carbenoxolone, or dietary intervention during pregnancy

	Male Offspring			Female Offspring
	Control	CX	DI	Control	CX	DI
Mice (n)	6	6	7	7	6	6
Weight, g	28±1	29±1	28±1	22±1	22±1	22±1
Feed intake, kcal·kg−1·day−1	403±16	457±35	472±20*	636±34	681±64	686±52
Water intake, ml·kg−1·day−1	312±114	380±58	353±36	578±37	592±69	553±43
LV, including septum, mg/g	3.4±0.3	3.4±0.1	3.6±0.1	3.8±0.1	3.8±0.2	3.7±0.1
RV free wall, mg/g	0.9±0.2	0.8±0.1	0.8±0.1	0.8±0.1	0.9±0.1	0.8±0.1
Brain, mg/g	16±1	15±1	16±1	20±1	17±2	20±1
Liver, mg/g	39±6	44±2	47±1	53±1	51±1	48±3
Kidneys, mg/g	14±1	13±1	14±1	16±2	14±1	15±1
Gonadal fat pads, mg/g	11±2	20±4*	23±2†	27±1	17±2*	19±3*

^*P < 0.05 vs. control.

^†P < 0.01 vs. control.

Blood pressure and activity.

Compared with controls, male DI offspring had increased locomotor activity (Fig. 1A), increased arterial pressure (Fig. 1C), and relative bradycardia (Fig. 1E). Female DI offspring also had increased locomotor activity (Fig. 1B), but in contrast to males, they displayed decreased blood pressures (Fig. 1D) without significant alteration in heart rate (Fig. 1F). Maternal CX did not influence offspring locomotor activity but did increase blood pressure and decrease heart rate in males (Fig. 2, C and E). Interestingly, while control females had higher blood pressures than control males, DI and CX females had lower blood pressures than their male counterparts (Figs. 1 and 2, all P < 0.01). Although sex did not influence heart rate of DI mice (Fig. 1), sexual dimorphism was present in control and CX mice, with control males and CX females having a relative sex-based tachycardia (Fig. 2, both P < 0.01). Following maternal DI or CX exposure, female offspring had similar baroreceptor event frequency (Fig. 3A) but a decrease in baroreflex sensitivity (Fig. 3B), an effect not seen in male offspring.

Fig. 1.

Locomotor activity, mean blood pressure, and heart rate were measured by radiotelemetry in 6-mo-old male (left) and female (right) offspring of dams that received control feed (solid symbols) or dietary intervention (DI; open symbols). Three-way ANOVA revealed a main effect of DI on activity (F = 52, P < 0.001), and a significant interaction between DI and sex for both blood pressure (F = 31) and heart rate (F = 96). Planned contrast testing revealed DI increased blood pressure in male mice (difference of means 5.9 mmHg, P < 0.01) and decreased blood pressure in female mice (difference of means 6.0 mmHg, P < 0.01). Although DI decreased heart rate in male mice [difference of means 33 beats per minute (bpm), P < 0.01], it did not significantly alter female heart rates. n = 6 or 7; *P < 0.05 or **P < 0.01 vs. control by ANOVA with Holm-Sidak testing for multiple comparisons.

Fig. 2.

Locomotor activity, mean blood pressure, and heart rate were measured in adult male (left) and female (right) offspring of dams that received control feed (solid symbols) or control feed plus carbenoxolone (CX; shaded symbols). Three-way ANOVA revealed no significant effect of CX or sex on adult locomotor activity, but a significant interaction between CX and sex for both blood pressure (F = 14) and heart rate (F = 47). For male mice, CX increased blood pressure (difference of means 7.3 mmHg, P < 0.01) and decreased heart rate (difference of means 28 bpm, P < 0.01). CX did not significantly alter blood pressure or heart rate in female mice; n = 6 or 7; *P < 0.05 vs. control by ANOVA with Holm-Sidak testing for multiple comparisons.

Fig. 3.

Spontaneous baroreflex sequences (events) and baroreflex sensitivity (BRS) were recorded in adult offspring of dams that received control feed (solid bars), control feed plus carbenoxolone (gray bars), or dietary intervention (open bars) during pregnancy (A and B). After these baseline recordings, all mice received dietary sodium chloride supplementation plus the nitric oxide synthase inhibitor l-NAME. From days 4 to 7 of combined salt and l-NAME, mean pressure and heart rate were recorded and compared with baseline values (C and D). Baroreflex measurements were repeated on the 7th day of l-NAME administration (E and F); n = 6 or 7. *P < 0.05 or ** P < 0.01 vs. control.

Dietary salt supplementation did not significantly change the blood pressures or heart rates of any group (data not shown). The addition of l-NAME to the drinking water led to significant increases in blood pressure and decreases in heart rate for all groups, except control males (Fig. 3, C and D). Compared with same sex controls, CX males and DI females had an exaggerated pressor response to l-NAME (Fig. 3C). CX males also had an accentuated bradycardic response to l-NAME (Fig. 3D). Although similar trends were noted for DI males, the l-NAME-induced differences did not reach statistical significance (for blood pressure, P = 0.06; for heart rate, P = 0.09 vs. control). While receiving salt supplementation and l-NAME, DI and CX males had increased baroreceptor event frequency (Fig. 3E), increased locomotor activity (Fig. 4A), hypertension (Fig. 4C), and marked bradycardia (Fig. 4E). As seen during baseline recordings (Fig. 3B), DI and CX females had decreased baroreflex sensitivity while receiving high salt plus l-NAME (Fig. 3F). Although the DI females maintained their increase in locomotor activity during l-NAME administration (Fig. 4B), they were no longer hypotensive (Fig. 4D).

Fig. 4.

One week after adding l-NAME to the drinking water and two wk after starting sodium chloride supplementation, locomotor activity (A and B), mean blood pressure (C and D), and heart rate (E and F) were measured in adult male (left) and female (right) offspring of dams that received control feed (solid symbols), control feed plus carbenoxolone (CX, gray symbols), or dietary intervention (DI, open symbols) during pregnancy; n = 6 or 7. *P < 0.05 or **P < 0.01 for DI vs. control. †P < 0.05 or ††P < 0.01 for CX vs. control by ANOVA.

Taken together, these results demonstrate increased arterial pressure and relative bradycardia in free-moving, unstressed male offspring of both DI and CX mothers. The increased arterial pressure is exaggerated by l-NAME, suggesting that enhanced nitric oxide-mediated vasodilation provides partial compensation for some other primary vascular problem. Surprisingly, female DI but not CX offspring are relatively hypotensive, but heart rate is unaltered. This is consistent with the observed reduction in baroreflex sensitivity in these animals. l-NAME treatment increased blood pressure in all female mice and eliminated the relative hypotension seen in DI females, suggesting the programmed hypotension is nitric oxide dependent.

Vascular reactivity.

Mesenteric artery contractile responses to 90 mmol/l KCl were not different between the control and DI mice of either sex (CX not examined). However, mesenteric segments from male DI offspring displayed a sex-specific increase in response to 10⁻⁵ mol/l noradrenaline (Fig. 5A). Preincubation in SOD plus catalase decreased the response of programmed mesenteric arteries to 10⁻⁶ mol/l noradrenaline (P < 0.05 vs. incubation in buffer alone for both male and female mice, Figs. 5, C and D). Antioxidant preincubation reversed the heightened constriction to 10⁻⁵ mol/l noradrenaline seen in programmed males (Fig. 5C) but led to a paradoxical increase in constriction to 10⁻⁵ mol/l noradrenaline for programmed females (Fig. 5D). Arterial segments from both DI male and female mice had significantly increased constriction following inhibition of nitric oxide production (Fig. 5, E and F). The additional inhibition of ROS metabolism using a combination of catalase plus SOD in the presence of l-NNA normalized these contractile responses (Fig. 5, G and H). These data suggest resistance arteries from DI offspring produce more ROS (favoring constriction) and nitric oxide (favoring relaxation), with an imbalance in favor of vasoconstriction in male offspring.

Fig. 5.

Mesenteric artery constrictions to cumulative concentrations of noradrenaline (NA) were measured in male (left) and female (right) offspring of dams that received control diet (solid symbols) or dietary intervention (open symbols). Vessels were preincubated in buffer alone (A and B), polyethylene glycol conjugated superoxide dismutase and catalase (PEG-SOD/catalase; C and D), l-NNA (E and F), or both l-NNA and PEG-SOD/catalase (G and H); n = 6 to 10. *P < 0.05 vs. control.

In vessels from male DI offspring, vasodilatory responses to both ACh and SNP were reduced (Figs. 6, A and C). These differences were eliminated by preincubation in the antioxidants catalase and SOD (Figs. 6E). A different pattern of reactivity was observed in female DI offspring, with enhanced vasorelaxation to SNP (Fig. 6B) and enhanced relaxation to ACh following preincubation in catalase plus SOD (Fig. 6F). These data again suggest resistance arteries from DI offspring produce more ROS (favoring constriction) and nitric oxide (favoring relaxation), with an imbalance in favor of vasodilation in female offspring.

Fig. 6.

Mesenteric dilation to sodium nitroprusside (A and B) or ACh (C and D) was evaluated for male (left) and female (right) offspring of control (solid symbols) or DI (open symbols) dams. Additional arteries were incubated in a combination of polyethylene glycol-conjugated catalase and superoxide dismutase prior to the addition of ACh (+SOD/catalase; E and F); n = 6 to 10. *P < 0.05 or **P < 0.01 vs. control.

The effects of DI, l-NNA, and SOD + catalase on mesenteric resistance vessel function suggest that there are DI-induced increases in both superoxide and nitric oxide production in both sexes. In males, enhanced noradrenaline-induced constriction and reductions in both SNP and ACh-induced dilation combined with the impressive reversal of these abnormalities by antioxidants suggests a primary increase in oxidant production that is consistent with the increased arterial pressure observed in these animals. In females, the l-NNA-induced increase in constriction to noradrenaline, coupled with the ability of SOD + catalase to reverse this, as well as the enhanced responsiveness to SNP and ACh (in the presence of SOD + catalase), suggest less dramatic increases in oxidant production combined with an impressive enhancement in nitrovasodilator function.

Chemiluminescence.

DI-exposed male mice had significantly decreased basal aortic arch superoxide production (Fig. 7A). In the presence of NADPH, the lucigenin signal was augmented to a far greater extent among aorta obtained from male and female DI offspring, compared with control (Fig. 7B). For all groups, greater than 95% of the NADPH-stimulated superoxide production was rapidly inhibited by DPI (data not shown). Superoxide production from CX offspring was not examined.

Fig. 7.

Superoxide production was measured by lucigenin-enhanced chemiluminescence among aortic arches obtained from never instrumented offspring of dams that received control diet (solid bars) or complex dietary intervention (open bars). Following measurement of relative light unit (RLU) production during incubation in buffer alone (A), NADPH and then diphenylene-iodonium (DPI) were serially added (B); n = 9 to 17. *P < 0.05 or **P < 0.01 vs. control.

DISCUSSION

Studies of isogenic animals subjected to different intrauterine conditions, but raised in similar postnatal environments, have shown a critical role of the intrauterine environment in the programming of adult cardiovascular disease (20, 28–30). Both of the murine models described herein were associated with sex-dependent increases in arterial pressure, being seen only in male offspring. CX had minimal effects on female offspring; however, DI females had a clear and novel phenotype that was suggestive of a primary enhancement in nitric oxide-mediated vasodilatation. Our analysis of the DI model was more comprehensive and demonstrated that the increased arterial pressure in males was correlated with ROS-mediated resistance vessel dysfunction. In contrast, female DI mice exposed to the same fetal environment were not only protected from hypertension but were actually hypotensive relative to controls. This reduction in blood pressure was associated with enhanced nitrovasodilator responsiveness of mesenteric microvessels.

Both male and female DI mice exhibited exaggerated responses to the loss of nitric oxide synthase-mediated blood pressure control, suggestive of the presence of nitric oxide-mediated compensation for vascular dysfunction. Consistent with the enhanced nitrovasodilator responsiveness and increased NADPH-stimulated aortic superoxide production in programmed mice of both sexes, responsiveness to endothelium-dependent vasodilators was increased in the presence of antioxidants. These results suggest that a sex-dependent imbalance between vascular ROS and nitric oxide production contribute to the programming of adult cardiovascular phenotypes.

Sex-specific cardiovascular programming.

The increased arterial pressure seen in male offspring of DI or CX-treated dams is consistent with findings across diverse animal models examining the effects of altered in utero environments (21). Although the absolute elevation in blood pressure that we demonstrate in programmed male mice (6 mmHg) approaches the degree of telemeter variance seen by some investigators after 9 wk of implantation (4 mmHg) (16), we minimized this random measurement error by 1) using a freshly calibrated telemeters for each mouse, 2) implanting telemeters for less than 4 wk, and 3) confirming location, lack of encapsulation and absence of significant pressure drift upon telemeter explanation. It is important to note that although programming induces a modest elevation in adult blood pressure, the Framingham Heart Study showed this degree of blood pressure elevation dramatically increases the risk of cardiovascular events (36). Prospective studies have further shown a 2 mmHg reduction in blood pressure can decrease coronary heart disease mortality by 7% (17).

The sexual dimorphism in control blood pressure that we observed is also consistent with data showing female C57BL/6J mice have higher blood pressures than isogenic male C57BL/6J mice (6). Although the mechanisms responsible for this sex-specific programming are not known, our results suggest that alterations in endothelial nitric oxide-dependent relaxation are not causative. Rather, l-NAME administration induced a disproportionate increase in blood pressure among DI and CX offspring, suggesting enhanced nitric oxide production buffers increased blood pressure in programmed mice. This is consistent with our prior observations in prehypertensive, programmed newborn sheep (33). The fact that blood pressure was increased in DI males yet decreased in DI females, despite both groups displaying increased activity, speaks to the complexity of the relationships among the sexes, programming stimuli, behavior, and cardiovascular regulation.

Role of free radicals.

Basal levels of nitric oxide appear to exert powerful effects in programmed mice both in vivo and in vitro. Female DI offspring displayed significantly greater increases in blood pressure with high salt plus l-NAME compared with control or CX offspring. Nitric oxide synthase inhibition removed the differences in blood pressure among groups. Interestingly, these DI females also demonstrated enhanced endothelial dependent (ACh) and independent (SNP) vasorelaxation, consistent with the fact that these animals have lower blood pressures. These data, and the absence of l-NAME-induced hypertension in control male mice, are consistent with studies showing a direct relationship between ovarian function and nitric oxide synthase activity (27).

The role of altered superoxide production in regulating vascular responsiveness was examined by evaluating mesenteric responses in the presence of catalase and SOD. In both male and female offspring, this antioxidant combination unmasked a propensity toward nitric oxide-mediated dilation and normalized the exaggerated constriction to noradrenaline that was seen following nitric oxide inhibition. Lucigenin-enhanced chemiluminescence confirmed a programmed increase in NADPH oxidase-derived superoxide production. Although lucigenin is relatively specific for the detection of superoxide anion (23), we are unable to exclude a role for other reactive oxygen species in the programming of cardiovascular physiology. The antioxidant-elicited increase in programmed female vasoconstriction to high concentrations of noradrenaline does suggest a dilatory reactive oxygen species, potentially a derivative of hydrogen peroxide, may also be involved.

It is possible the DI-induced improvement in female endothelial function is a reflection of the time course of our evaluations, which occurred at an age when the female mice were protected from programmed hypertension. Supporting this hypothesis are data in other model systems using mice, sheep, and humans, suggesting that increased nitric oxide-mediated vasodilation buffers inherent predispositions to hypertension (14, 33, 34). Nonetheless, our findings from these in vitro studies provide evidence of alterations in both ROS and nitric oxide metabolism in these tissues, which are sex dependent. Although the causes of this sexual dimorphism have not been elucidated, previous studies have suggested an estrogen-mediated protection from programmed hypertension, which is lost with ovariectomy or aging (22, 24, 26).

Intriguingly, female CX and DI offspring had decreased baroreflex sensitivity while receiving baseline diet or high salt plus l-NAME. The decreased baroreflex sensitivity suggests a programming of autonomic regulation, specifically parasympathetic impairment, which is independent of blood pressure. While the persistence of the phenotype during l-NAME administration argues against a role of nitric oxide, increased oxidative stress has been shown to inhibit baroreceptor afferent sensitivity (18). Coupled with this uniquely described alteration in spontaneous baroreflex sensitivity, DI female offspring had a long-term resetting of the blood pressure-heart rate relationship, as evidenced by the lack of reflex tachycardia during the evolution of nitric oxide synthase-mediated hypotension.

Importance of locomotor activity.

The increase in baseline locomotor activity among DI male and female mice is a novel and unexpected finding. Prior publications describing a programmed decrease in locomotor activity employed behavioral apparatuses that use finite periods of testing and are associated with animal distress (2, 38). These tests are likely to evaluate a different aspect of behavior than that monitored by the telemetric recording of spontaneous activity among undisturbed animals in their home cage. By extension, the dramatic tail cuff hypertension and inactivity reported by many investigators may be strongly influenced by the interaction between environmental stress and prenatal interventions. These confounding interactions were avoided in the present study through the use of telemetric recordings to more closely reflect baseline cardiovascular status. The interrelationships that we demonstrate between circadian patterns in activity, blood pressure, baroreflexes, and resistance vessel responsiveness further emphasize the ability of radiotelemetry recordings to enhance our understanding of sex-specific programmed cardiovascular physiology.

Perspectives and Significance

The murine models that we describe demonstrate that alterations in the perinatal environment are associated with cardiovascular dysfunction in a species amenable to genetic manipulation. The pathways responsible for the alterations seen in diet- or glucocorticoid-induced fetal programming have yet to be fully elucidated. Sex-specific differences have been seen in many models of developmental programming. With studies showing a reversal of these sex differences following ovariectomy, it is likely that estrogen plays an important role in the buffering of programmed adult phenotypes. Interestingly, altered maternal nutrition (low-protein, high-carbohydrate, high-fat diet) and fetal glucocorticoid exposure (following inhibition of placental 11-β-hydroxysteroid dehydrogenase) induced similar, but not identical adult phenotypes. While the findings support prior studies, suggesting fetal glucocorticoid exposure is a proximal cause of programmed adult disease, alternative hypotheses remain tenable. Further investigations into links between the early environment and later health hold the promise of preventing an important subset of cardiovascular disease.

GRANTS

This study was supported by National Institutes of Health Grants ES012268 (to J. Segar), HD050359 (to R. Roghair), HL04495 (to T. Scholz), and HL62483 (to F. Lamb).