Prenatal Hypoxia Leads to Increased Muscle Sympathetic Nerve Activity, Sympathetic Hyperinnervation, Premature Blunting of Neuropeptide Y Signaling, and Hypertension in Adult LifeNovelty and Significance
Adverse conditions prenatally increase the risk of cardiovascular disease, including hypertension. Chronic hypoxia in utero (CHU) causes endothelial dysfunction, but whether sympathetic vasoconstrictor nerve functioning is altered is unknown. We, therefore, compared in male CHU and control (N) rats muscle sympathetic nerve activity, vascular sympathetic innervation density, and mechanisms of sympathetic vasoconstriction. In young (Y)-CHU and Y-N rats (≈3 months), baseline arterial blood pressure was similar. However, tonic muscle sympathetic nerve activity recorded focally from arterial vessels of spinotrapezius muscle had higher mean frequency in Y-CHU than in Y-N rats (0.56±0.075 versus 0.33±0.036 Hz), and the proportions of single units with high instantaneous frequencies (1–5 and 6–10 Hz) being greater in Y-CHU rats. Sympathetic innervation density of tibial arteries was ≈50% greater in Y-CHU than in Y-N rats. Increases in femoral vascular resistance evoked by sympathetic stimulation at low frequency (2 Hz for 2 minutes) and bursts at 20 Hz were substantially smaller in Y-CHU than in Y-N rats. In Y-N only, the neuropeptide Y Y1-receptor antagonist BIBP3226 attenuated these responses. By contrast, baseline arterial blood pressure was higher in middle-aged (M)-CHU than in M-N rats (≈9 months; 139±3 versus 126±3 mm Hg, respectively). BIBP3226 had no effect on femoral vascular resistance increases evoked by 2 Hz or 20 Hz bursts in M-N or M-CHU rats. These results indicate that fetal programming induced by prenatal hypoxia causes an increase in centrally generated muscle sympathetic nerve activity in youth and hypertension by middle age. This is associated with blunting of sympathetically evoked vasoconstriction and its neuropeptide Y component that may reflect premature vascular aging and contribute to increased risk of cardiovascular disease.
It is well established that adverse conditions in utero induce fetal programming and increase the risk of cardiovascular disease in the offspring.1 Notably, placental insufficiency, which leads to impaired delivery of oxygen and nutrients to the fetus, is a major cause of intrauterine growth restriction and has been associated with hypertension in adult life.2
Studies involving animal models of fetal programming have already shown that endothelial dysfunction occurs in the offspring when pregnant rats are exposed to systemic hypoxia during pregnancy to induce chronic hypoxia in utero (CHU) or nutrient restriction.3 However, the effects of CHU are more dramatic and occur earlier. Thus, male CHU offspring showed endothelial dysfunction at 3 to 4 months of age that was associated with reduced bioavailability of nitric oxide and oxidative stress3 and prevented by treating the pregnant dam with antioxidant.3–5 By contrast, endothelial dysfunction did not occur in nutrient restriction offspring until 7 months of age when nitric oxide still contributed to endothelium-dependent dilation.3
Endothelial dysfunction may contribute to development of hypertension6 as may enhanced myogenic tone, another characteristic of male CHU, but not nutrient restriction rats.7 However, elevated blood pressure in patients with essential hypertension is considered to be initiated by sympathetic hyperactivity.8 Accordingly, young men and women who were small for gestational age at birth had raised muscle sympathetic nerve activity (MSNA).9 Further, chick embryos made hypoxic during incubation showed augmented responses to tyramine, which releases norepinephrine (NE) and sympathetic hyperinnervation of femoral arteries.10 Whether similar changes occur in mammals is not known.
Sympathetic hyperactivity in essential hypertension in young humans and spontaneously hypertensive rats (SHRs) is accompanied by exaggerated muscle vasoconstrictor responses to sympathetic activation and higher plasma levels of the cotransmitter neuropeptide Y (NPY).11,12 But chronically raised MSNA also occurs during aging and is accompanied by reduced reflex vasoconstriction and loss of NPY involvement in skin.13,14 However, in young men at high altitude, chronic hypoxia caused an increase in tonic MSNA, tonic muscle vasoconstriction, and hypertension that persisted on descent to sea level.15,16 Further, vasoconstrictor responses evoked by sympathetic activation were blunted in acute and chronic hypoxia in adult rats and humans17–19; this was attributed to impaired vasoconstriction to NE and adenosine triphospate, but preservation of the NPY component.19,20
In view of these findings, we hypothesized that CHU rats have raised MSNA, develop hypertension early during aging, and show augmented muscle vasoconstrictor responses to sympathetic nerve activation, but preservation of the NPY component. By contrast, we hypothesized that control (N) rats show impaired sympathetic vasoconstriction in muscle with aging and loss of the NPY component. Thus, we performed experiments on male control rats and CHU rats at 10 to 12 and at 36 to 37 weeks of age, which may be compared with humans in young adulthood and early middle age (≈16–18 and 30–35 years, respectively).21
Methods and Materials
For a complete description of the Materials and Methods, see the online-only Data Supplement.
Experiments were performed on male N and CHU Wistar rats; the latter were offspring of rat dams housed in a hypoxic chamber at 12% O2 from day 10 to 20 of pregnancy.4 All experiments were approved by the Biomedical Research Ethics Committee of the University of Birmingham and performed under the UK Animal (Scientific Procedures) Act.
Recordings of MSNA
Under anesthesia, MSNA was recorded from the surface of arterial vessels of spinotrapezius muscle,22 together with arterial blood pressure (ABP), heart rate (HR), and respiratory frequency (Rf) from 27 Y-N and 17 Y-CHU rats under baseline conditions (i) during progressive acute hypoxia and (ii) baroreceptor unloading induced by sodium nitroprusside.
Sympathetic Innervation Density
Responord Countses Evoked by Sympathetic Nerve Stimulation
In 12 Y-N rats and 11 Y-CHU and 13 mol/L-N and 13 mol/L-CHU rats, sympathetic nerve stimulation (SNS) was applied via the right sympathetic chain as a 60 sec train of 120 pulses at 2 Hz and as 6 bursts of 20 pulses at 20 Hz at 10 sec intervals (120 pulses in total).19 Changes in femoral vascular resistance (FVR) were recorded before and during infusion of the NPY Y1-receptor antagonist BIBP-3226 (10 mg/kg/min).
All data are presented as mean±SEM. Single comparisons were made by Student’s paired or unpaired t test. Multiple within-group comparisons were made by repeated measures analysis of variance, with Dunnet’s post hoc. For responses evoked by SNS, the maximum change in FVR (ΔFVR Max) evoked by SNS was calculated. In addition, the integral of FVR during SNS, relative to a 60 sec period before stimulation (Δint FVR) was calculated for each pattern of stimulation. Comparisons were made before and after BIBP-3226 by using mixed model analysis of variance with Scheffés post hoc test. P<0.05 was considered statistically significant.
Cardiac and respiratory rhythmicity was present in all single units discriminated from MSNA in Y-N and Y-CHU rats. Mean ongoing firing rate recorded from single units was ≈70% higher in CHU than in N rats (0.56±0.075 versus 0.33±0.036 Hz). Impulses generally occurred singly in both Y-N and Y-CHU rats; couplets, with an interpulse interval of <0.1 sec, seldom occurred, that is, the instantaneous firing frequencies were low. However, there was a lower proportion of low instantaneous frequencies (<1 Hz) and greater proportion of higher instantaneous frequencies (1–10 Hz) in CHU than in N rats (Figure 1), reflecting more couplets of high instantaneous frequency in Y-CHU rats. The percent of spikes occurring within the phase-related peaks and nadirs of the cardiac and respiratory cycles were fully comparable in Y-N and Y-CHU rats. For further details of MSNA, see Figures S1–S4 and Table S1 in the online-only Data Supplement.
Progressive systemic hypoxia induced greater increases in Rf and HR in Y-CHU than in Y-N rats; the fall in ABP was greater in Y-N rats (P<0.05; Figure 2A). Concomitantly, MSNA increased, the change in firing frequency evoked by 8% O2 being similar in Y-N and Y-CHU rats (see Figure 2A). When SNP was infused to induce a fall in ABP comparable to that of hypoxia; the reflex increase in MSNA firing frequency was similar in Y-N rats and Y-CHU rats, as was the increase in HR (Figure 2B).
Sympathetic Nerve Density
The density of sympathetic innervation on tibial arteries was ≈50% greater in Y-CHU than in Y-N rats (Figure 3).
Responses Evoked by Sympathetic Nerve Stimulation
Continuous SNS stimulation at 2 Hz and bursts of impulses at 20 Hz evoked smaller increases in FVR in Y-CHU than in Y-N rats; these differences reaching statistical significance for FVR Max (Figure 4, top right). The Y1-receptor antagonist attenuated the increases in IntFVR evoked by 2 Hz and by bursts at 20 Hz in Y-N, but not Y-CHU rats (Figure 4, top left).
Baseline ABP was higher in M-CHU than in Y-CHU rats (139±3 versus 131±3 mm Hg; P<0.05), but not different between M-N and Y-N rats (126±3 versus 127±3 mm Hg); it was also higher in M-CHU than in M-N rats (P<0.05). IntFVR increases evoked in M-N rats by SNS at 2 Hz and bursts at 20 Hz were blunted relative to Y-N rats (P<0.05), such that evoked increases in IntFVR were similar in M-N and M-CHU rats (Figure 4, bottom left). The Y1-receptor antagonist did not affect the sympathetically evoked responses in either group. For further details, see online-only Data Supplement.
The present study demonstrates for the first time that prenatal hypoxia leads to a substantial increase in the frequency of single unit MSNA and in sympathetic nerve density of skeletal muscle arteries in young adult male rats (Y-CHU) relative to control (Y-N) rats. Muscle vasoconstrictor responses evoked by SNS at low and high frequencies were blunted in Y-CHU relative to Y-N rats, and the contribution of NPY was absent in Y-CHU rats. By early middle age, M-CHU rats had raised ABP relative to M-N rats, but muscle vasoconstrictor responses evoked by SNS were also blunted in M-N rats; NPY made no contribution in either M-N or M-CHU rats.
Prenatal Hypoxia and MSNA
We recorded MSNA from the surface of muscle arterial vessels; it displayed both cardiac and respiratory rhythmicities as expected of muscle vasoconstrictor nerves.22,25 We analyzed our data as single unit activity rather than as multiunit MSNA because direct comparisons between experimental groups are difficult when the number of fibers adjacent to the electrode or recruitment of silent fibers can affect gross activity. In line with our hypotheses, mean frequency in single units was ≈70% greater in Y-CHU than in Y-N rats. Further, MSNA in Y-CHU rats contained a higher proportion of couplets at instantaneous frequencies of 1 to 5 Hz and 5 to 10 Hz and a lower proportion at <1 Hz. In view of these findings, it is tempting to suggest that raised MSNA recorded in young men and women who were small for gestational age at birth9 may arise partly from prenatal hypoxia.
Ours is the first direct evidence that basal sympathetic activity is raised in any experimental model of prenatal programming. In the only other study in which gross renal sympathetic nerve activity was recorded in male offspring of dams fed a low protein diet, baseline renal sympathetic nerve activity was comparable to that of controls, but showed larger increases during static muscle contraction.26
In the few reports of single unit MSNA in humans, mean frequency of single unit MSNA was low in normotensives (≈0.3 Hz), higher in established essential hypertensives (≈0.6 Hz), and higher still in borderline hypertensives (≈0.75 Hz), consistent with sympathetic hyperactivity preceding the full development of essential hypertension.27 Further, in normotensives, couplets or higher multiples of impulses were uncommon, instantaneous frequencies being <1 Hz in ≈50% of units,28 but far more common in essential hypertension.29 On the other hand, in obesity-related hypertension, there was raised multiunit MSNA but not single unit MSNA, indicating greater recruitment of individual sympathetic fibers. Patients with congestive heart failure, or obstructive sleep apnea (OSA) and hypertension, had raised single-unit MSNA (0.98, 0.96 Hz), but only those with OSA showed increased numbers of multiple impulses.29,30 Thus, prenatal hypoxia shares some characteristics with essential hypertension and OSA: raised single-unit MSNA and impulses in couplets or multiples.
Cross-correlograms of respiratory- and cardiac-related coupling of single-unit MSNA showed they were similar in Y-N and Y-CHU rats. Thus, there is no reason to suggest that the higher MSNA in Y-CHU rats reflected stronger respiratory modulation or weaker baroreceptor-reflex inhibition of MSNA. By contrast, in working heart–brain stem preparations, the augmented sympathetic activity in SHRs reflected amplified respiratory coupling of sympathetic nerve activity.31 Thus, our findings compare better with those of essential hypertension or chronic obstructive pulmonary disease, which did not show augmented respiratory coupling of MSNA even though they had raised MSNA.32
Nevertheless, cardiac-related rhythmicity was lower in chronic obstructive pulmonary disease patients than hypertensives, suggesting the raised MSNA in chronic obstructive pulmonary disease reflects depressed baroreceptor inhibition and increased tonic drive from peripheral chemoreceptors.32 A similar suggestion was made for the raised MSNA and multiple firing in OSA.30 In our study, the increases in MSNA evoked by baroreceptor unloading and the hypoxia-induced fall in ABP were comparable in Y-N and Y-CHU rats, suggesting baro- and chemoreceptor reflex effects on MSNA were similar. Indeed, the simplest interpretation is that the augmented MSNA in Y-CHU rats is neurogenic, reflecting greater central neural output to individual muscle sympathetic neurons. The baroreceptor reflex effect on MSNA apparently operates around a raised set point, and the peripheral chemoreceptor input modulates, rather than contributes to, this raised discharge.
Long-term hypoxia in adult life, lasting from 20 minutes o several months, causes an increase in MSNA that outlasts return to normoxia.16,33 Further, chronic intermittent hypoxia induced experimentally and occurring in OSA causes a persistent increase in MSNA.30,34 It has been suggested that this arises because hypoxia induces a long-term increase in the activity of the premotor sympathetic neurons of rostroventrolateral medulla initiated by chemoafferent input and accompanied by resetting of the baroreceptor reflex effect on rostroventrolateral medulla to higher levels.33,35 The present findings raise the possibility that prenatal hypoxia evokes a similar long-term potentiation of rostroventrolateral medulla that raises MSNA in adult Y-CHU rats.
Prenatal Hypoxia and Sympathetic Vasoconstriction
Higher mean frequency of single unit MSNA with more couplets at higher instantaneous frequencies would be expected to exert a greater tonic vasoconstrictor influence in Y-CHU rats.36,37 The greater sympathetic nerve density revealed in tibial arteries of Y-CHU might be expected to further augment the number of impulses reaching arterial smooth muscle. And yet, baseline FVR was not different between Y-N and Y-CHU rats.
As expected, SNS with bursts at 20 Hz evoked a greater increase in FVR than stimulation at 2 Hz in both Y-N20 and Y-CHU rats, but these vasoconstrictor responses were blunted, not exaggerated in Y-CHU rats (in contrast to our hypotheses). Moreover, blockade of NPY Y1-receptors attenuated responses evoked by both patterns of stimulation in Y-N rats, but not Y-CHU rats. Because the cotransmitters NPY, ATP, and NE facilitate one another’s actions,12,19,36,38 it is difficult to quantify contributions of individual transmitters with selective receptor antagonists. Nevertheless, as responses evoked by both patterns were similar in N and CHU rats after Y1-receptor antagonism, it is likely the blunted vasoconstrictor responses in Y-CHU rats reflected lack of an NPY component. Such a deficit could also explain why baseline FVRs were similar in Y-N and Y-CHU rats even though MSNA in Y-CHU rats contained more high instantaneous frequencies.20
Clearly, these effects of prenatal hypoxia differ from acute hypoxia in adult life when the contribution of NPY to sympathetically evoked vasoconstriction is maintained while the contributions of ATP and NE are attenuated.19 At first sight, they also differ from essential hypertension in humans and SHRs. NPY has been implicated in the development of hypertension not only because plasma levels are raised in young SHRs, but because NPY is released by high sympathetic discharge, facilitates noradrenergic vasoconstrictor responses, and induces vascular hypertrophy.12 However, although at 4 weeks of age, NPY content was higher in mesenteric arteries of SHRs than Wistar Kyoto controls (WKY), by 4 months of age, it was lower in SHRs than Wistar Kyoto controls.39 Moreover, at ≈3 months, antioxidant treatment augmented NPY release from sympathetic nerve fibers in SHRs but not in Wistar Kyoto controls, such that NPY then contributed to sympathetic vasoconstriction in SHRs.40 Thus, it may be that oxidative stress in essential hypertension, SHRs, and prenatal hypoxia3,5,41 has the common effect of limiting NPY release such that, in Y-CHU rats, the NPY contribution to sympathetic muscle vasoconstriction is minimal. Oxidative stress may also blunt the NE-evoked component of sympathetic vasoconstriction in Y-CHU rats by limiting NE synthesis, as in aging skin.14
Effects of Aging
M-N rats showed blunted vasoconstrictor responses to sympathetic stimulation relative to Y-N rats that were not attenuated by Y1-receptor blockade. This is consistent with the finding that, in humans, reflex muscle vasoconstriction was attenuated with aging, even though MSNA was increased and that cutaneous vasoconstriction was blunted by aging and no longer included an NPY component.13,14 To our knowledge, ours is the first evidence that even modest aging attenuates the NPY component of sympathetic muscle vasoconstriction. The mechanisms are not clear.
From the fragmentary evidence available, sympathetic innervation density on femoral arteries of rats increased from 10 to 12 weeks to 9 months,24 but the NPY content in arterial vessels was similar at 4 and 16 weeks.39 Moreover, NPY neuropepidase activity increased from 2 to 6 months of age in skeletal muscle arterioles of rats and greatly attenuated sympathetically evoked NPY release at 6 months,42 whereas constrictor responses evoked by NPY in rat mesenteric arteries in vitro were similar at 4 weeks and 6 months.39 However, vasoconstriction evoked in muscle circulation by α- adrenoceptor agonists was smaller in old than young human,43 whereas in dogs, muscle vasoconstriction evoked by adrenoceptor, ATP, or NPY Y1-receptor stimulation did not change with age, neither did receptor expressions.44 Given these findings, we simply propose that attenuation of sympathetically mediated vasoconstriction in M-N rats reflects loss of the NPY component and, possibly, reduced release or action of NE and ATP; the underlying mechanisms may include oxidative stress associated with aging.14,40
Aging in CHU Rats
The lack of NPY contribution to sympathetically evoked vasoconstriction persisted in M-CHU rats. Indeed, M-CHU rats showed no obvious age-related attenuation of sympathetically evoked vasoconstrictor responses. Rather, it seems Y-CHU rats (≈3-months-old) showed premature aging, displaying sympathetic vasoconstriction comparable to M-N rats (≈9-months-old). This is reminiscent of the finding that endothelium-dependent relaxation was blunted in CHU rats at 4 months and did not show the age-dependent attenuation seen in N rats between 4 and 7 months.3 Because the nitric oxide contribution to endothelium-dependent relaxation was prevented in CHU rats at 4 months by treating the pregnant dam with Vitamin C,5 although endothelium-dependent relaxation at 7 months was restored by acute antioxidant treatment,3 it may be that oxidative stress not only led to early endothelial dysfunction, but depressed sympathetically evoked vasoconstriction in Y-CHU and M-CHU rats by blunting release and action of NPY and NE.14,40 This should be tested in future studies.
Prenatal Hypoxia and Hypertension
As we hypothesized, the raised MSNA in Y-CHU rats was associated with a progressive increase in baseline ABP: ABP was higher in M-CHU than in Y-CHU rats with no age-related change in N rats. As far as we are aware, this is a novel finding. It will be important to establish whether MSNA is also raised in M-CHU rats and whether ABP is raised in nonanesthetized Y-CHU and M-CHU rats. Whether raised MSNA in male Y-CHU rats and hypertension in M-CHU rats are causally linked remains speculative. Baseline FVR was not raised in M-CHU rats; therefore, increased vasoconstrictor tone in muscle did not directly contribute to the raised ABP. However, the relationships between MSNA, vascular resistance, and ABP in healthy men are not straightforward and do not change systematically with age or hypertension. Nevertheless, much evidence supports a sympathetic mechanism for the genesis of hypertension.8,21,27,45
We have provided novel evidence that fetal programming induced by prenatal hypoxia leads to increased MSNA in young adult males with no change in respiratory or cardiac-related rhythmicities, nor in the effects of baroreceptor unloading or peripheral chemoreceptor stimulation. We therefore suggest that prenatal hypoxia leads to an increase in centrally generated sympathetic outflow to skeletal muscle. Our findings also add to evidence3 that prenatal hypoxia induces premature vascular aging: vasoconstrictor responses evoked by sympathetic nerve activation were blunted in young CHU rats, apparently because the NPY Y1-receptor contribution was lacking, a change that occurred in control rats by middle-age. Because baseline hypertension was present in CHU rats by middle age, we propose that raised MSNA and altered NPY signaling, together with endothelial dysfunction,3–6,39,40 collude to increase the risk of cardiovascular disease in individuals who experience prenatal hypoxia.1
Sources of Funding
This study was supported by a British Heart Foundation PhD Studentship (FS/08/068/25798).
The online-only Data Supplement is available with this article at http://hyper.ahajournals.org/lookup/suppl/doi:10.1161/HYPERTENSIONAHA.114.04374/-/DC1.
- Received August 1, 2014.
- Revision received August 20, 2014.
- Accepted September 2, 2014.
- © 2014 American Heart Association, Inc.
- Hemmings DG,
- Williams SJ,
- Davidge ST
- Esler M
- Rouwet EV,
- Tintu AN,
- Schellings MW,
- van Bilsen M,
- Lutgens E,
- Hofstra L,
- Slaaf DW,
- Ramsay G,
- Le Noble FA
- Hodges GJ,
- Jackson DN,
- Mattar L,
- Johnson JM,
- Shoemaker JK
- Seals DR,
- Dinenno FA
- Doyle MP,
- Walker BR
- Folkow B,
- Svanborg A
- Omar NM,
- Marshall JM
- Häbler HJ,
- Jänig W,
- Krummel M,
- Peters OA
- Mizuno M,
- Siddique K,
- Baum M,
- Smith SA
- Greenwood JP,
- Stoker JB,
- Mary DA
- Macefield VG,
- Wallin BG
- Elam M,
- McKenzie D,
- Macefield V
- Xie A,
- Skatrud JB,
- Puleo DS,
- Morgan BJ
- Querido JS,
- Wehrwein EA,
- Hart EC,
- Charkoudian N,
- Henderson WR,
- Sheel AW
- Johnson CD,
- Coney AM,
- Marshall JM
- Macarthur H,
- Westfall TC,
- Wilken GH
- Touyz RM
- Delorey DS,
- Clifford PS,
- Mittelstadt S,
- Anton MM,
- Kluess HA,
- Tune JD,
- Dincer UD,
- Buckwalter JB
- Joyner MJ,
- Charkoudian N,
- Wallin BG
Novelty and Significance
What Is New?
These are the first recordings of muscle sympathetic nerve activity and sympathetically evoked muscle vasoconstriction in young adult rats made hypoxic in utero rats.
Resting muscle sympathetic nerve activity was raised while muscle vasoconstrictor responses to sympathetic activation were blunted in chronic hypoxia in utero rats and lacked a neuropeptide Y component.
Similar changes in sympathetic vasoconstrictor responses did not occur in control rats until middle age, by which time chronic hypoxia in utero rats were hypertensive.
What Is Relevant?
Adverse conditions during pregnancy increase risk of cardiovascular disease in adult life.
Raised muscle sympathetic nerve activity in youth is causally implicated in the genesis of hypertension.
Impaired NPY release and blunted sympathetic vasoconstriction have been associated with hypertension and aging.
Fetal programming induced by hypoxia in utero increases muscle sympathetic nerve activity, leads to hypertension by middle age, and modulates sympathetic vasoconstriction in ways that suggest premature aging.