2016 Volume 80 Issue 11 Pages 2388-2396
Background: Extremely preterm infants frequently have patent ductus arteriosus (PDA). Recent recommendations include immediately beginning amino acid supplementation in extremely preterm infants. However, the effect of amino acids on closure of the ductus arteriosus (DA) remains unknown.
Methods and Results: Aminogram results in human neonates at day 2 revealed that the plasma glutamate concentration was significantly lower in extremely preterm infants (<28 weeks’ gestation) with PDA than in those without PDA and relatively mature preterm infants (28–29 weeks gestation). To investigate the effect of glutamate on DA closure, glutamate receptor expression in fetal rats was examined and it was found that the glutamate inotropic receptor, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) type subunit 1 (GluR1), mRNA was highly expressed in the DA compared to the aorta on gestational day 19 (preterm) and gestational day 21 (term). GluR1 proteins were co-localized with tyrosine hydroxylase-positive autonomic nerve terminals in the rat and human DA. Intraperitoneal administration of glutamate increased noradrenaline production in the rat DA. A whole-body freezing method demonstrated that glutamate administration induced DA contraction in both preterm (gestational day 20) and term rat fetuses. Glutamate-induced DA contraction was attenuated by the calcium-sensitive GluR receptor antagonist, NASPM, or the adrenergic receptor α1 blocker, prazosin.
Conclusions: These data suggest that glutamate induces DA contraction through GluR-mediated noradrenaline production. Supplementation of glutamate might help to prevent PDA in extremely preterm infants. (Circ J 2016; 80: 2388–2396)
The ductus arteriosus (DA) is the fetal bypass artery between the pulmonary artery (PA) and the aorta, and is responsible for fetal circulation.1 However, failure of the DA to close after birth, which is called patent DA (PDA), causes a left-to-right shunt that leads to left ventricular volume overload, and is frequently observed in preterm infants. More than 60% of extremely preterm infants (<28 weeks’ gestation) exhibited PDA, which is a major cause of morbidity and mortality, because it leads to severe complications including pulmonary hypertension, right ventricular dysfunction, postnatal infections, and respiratory failure.2 Despite considerable progress in medical management for preterm infants, 25% of PDA in extremely preterm infants is still resistant to pharmacological therapies; that is, cyclooxygenase inhibitors, which have been the only pharmacological therapeutic strategy since the mid-1970 s.1,3–5 Therefore, novel strategies for PDA in extremely preterm infants are required.
Amino acid administration beginning immediately after birth has recently been suggested,6 and the recommended daily intake is calculated based on rectifying cumulative deficits and matching intrauterine growth velocity.7,8 However, the role of amino acids in DA regulation has not been reported. Emerging evidence suggests multiple roles for each amino acid, and it has been shown that glutamate stimulates noradrenaline release.9,10 A recent study demonstrated that norepinephrine content was decreased in the brain of glutamate inotropic receptor α-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) type subunit 1 (GluR1)-deficient mice.11 These data suggest that glutamate induces noradrenaline release through GluR1.
Morphological studies on DA innervation have demonstrated the presence of adrenergic nerves in the vessel wall of the DA and suggested the possibility of some neuronal control.12–16 Many adrenergic nerve terminals were shown in the muscular walls of the human DA, and fluorometric results revealed the presence of noradrenaline in the human DA.15 Ikeda et al performed a quantitative comparison of nerve density in the DA and adjacent vessels and found that noradrenaline content was greater in lamb DA.17 Noradrenaline induced significant constriction of the DA even in the 20-week human fetus.15
Based on these previous studies, we examined the role of glutamate on DA closure. Here, we demonstrate that GluR1 was highly expressed in the DA, and glutamate induced rat DA contraction through production of noradrenaline. Aminogram results from human plasma samples showed that the glutamate concentration was lower in extremely preterm infants with PDA than in those without PDA.
Wistar rat fetuses were obtained from timed-pregnant mothers purchased from Japan SLC, Inc (Shizuoka, Japan). The experiments were approved by the Yokohama City University Institutional Animal Care and Use Committee in accordance with the Guide for the Care and Use of Laboratory Animals (reference number: F-A-13-006).
Human DA Tissue and Plasma SamplesA protocol for using blood samples and human tissues was approved by the Research Ethics Committee at Yokohama City University Hospital and Kanagawa Children’s Medical Center (reference numbers: B100107034 and D13021010), and it conformed to the principles outlined in the Declaration of Helsinki.
Measurement of Plasma Amino Acid LevelsPlasma amino acids from fetal rats on gestational day 21 and preterm human infants were analyzed using an automatic amino acid analyzer (L-8800A; Hitachi, Tokyo, Japan). Briefly, amino acids separated by cation exchange chromatography were detected spectrophotometrically after a postcolumn reaction with a ninhydrin reagent. Blood samples from fetal rats on gestational day 21 were obtained via heart puncture after rats were anesthetized with isoflurane. To obtain fetal rats by cesarean section, the maternal rats were euthanized with 200 mg/kg of pentobarbital.
Human infants were enrolled in the study after parental written informed consent was obtained. Twenty-seven preterm infants at ≥24 and <30 weeks gestation who were admitted to the neonatal intensive care unit of Yokohama City University Hospital were recruited prospectively between August 2011 and February 2014. Exclusion criteria were major congenital anomalies, congenital heart diseases, and death in the first week of life. The patients were divided into 2 groups: those who were ≥28 weeks’ gestation and those who were <28 weeks’ gestation. The <28-week group was subdivided into PDA and non-PDA groups. In the >28-week group, there were no patients with PDA. In this study, PDA was defined as the necessity for indomethacin therapy within the first week of life. Indomethacin is not prophylactically administered at this institution.
Medical history was recorded and is summarized in Table 1. Extremely preterm infants with PDA were treated with an intravenous injection of indomethacin between day 1 and day 6 (average 2.4±0.9 days). Respiratory distress syndrome (RDS) was defined as any situation in which a surfactant was administered for therapy. Blood samples were collected on day 2 after birth (33–53 h after birth). Arterial umbilical cord blood samples from the same patients were obtained at the time of delivery. All samples used in the aminogram test were centrifuged at 845 g for 5 min, and the plasma samples were stored frozen at –80℃ until analysis.
| Variables | A 28–29 weeks (n=5) |
B 24–27 weeks without PDA (n=6) |
C 24–27 weeks with PDA (n=5) |
P value | |||
|---|---|---|---|---|---|---|---|
| A vs. B vs. C |
A vs. B | A vs. C | B vs. C | ||||
| Gestational age (weeks) | 29.3±0.5 | 25.7±0.8 | 25.6±2.0 | 0.007† | 0.008† | 0.007† | 0.927 |
| Birth weight (g) | 1,142.8±86.1 | 774.6±203.7 | 654±221.8 | 0.008† | 0.008† | 0.007† | 0.428 |
| Male (%) | 3 (60) | 3 (50) | 3 (60) | 0.926 | |||
| Amino acid (g·kg–1·day–1) | 1.02±0.06 | 1.16±0.16 | 0.93±0.26 | 0.309 | |||
| RDS no. (%) | 3 (60) | 4 (67) | 3 (60) | 0.965 | |||
| Maximum FiO2 in 12 h | 0.33±0.08 | 0.64±0.24 | 0.56±0.18 | 0.029* | 0.015* | 0.056 | 0.853 |
| SFD no. (%) | 1 (20) | 2 (33.3) | 2 (40) | 0.302 | |||
| Apgar score at 1 min | 5.6±2.6 | 4.1±2.2 | 4.6±1.9 | 0.613 | |||
| Apgar score at 5 min | 8.4±0.89 | 6.8±1.8 | 7.6±1.5 | 0.307 | |||
| Infection no. (%) | 0 (0) | 1 (17) | 2 (20) | 0.265 | |||
| Antenatal steroids no. (%) | 2 (40) | 6 (100) | 4 (80) | 0.069 | |||
| Furosemide no. (%) | 0 (0) | 0 (0) | 0 (0) | ||||
RDS, respiratory distress syndrome; SFD, small for date; *P<0.05; †P<0.01.
DA tissues from fetal rats on gestational day 21 and from a human neonate with coarctation complex, which were obtained during cardiac surgery, were stained using anti-GluR1 (Millipore, Kenilworth, NJ, USA) and anti-tyrosine hydroxylase (Immunostar, Hudson, WI, USA) antibodies, as described previously.18,19
RNA Isolation and Quantitative Reverse Transcription-Polymerase Chain ReactionTotal RNA isolation, cDNA generation, and quantitative reverse transcription-polymerase chain reaction (RT-PCR) analysis were performed as described previously.18,20 The primers were designed based on nucleotide sequences of rat GluR1 (NM_031608.1) (5′-TCCTGTTGACACATCCAATCA-3′ and 5′-CCGTTACCTGCCAGTTCTTC-3′), GluR2 (NM_017261.2) (5′-CAGTCACCAATGCTTTCTGC-3′ and 5′-TGCTCCTTTGAGGTCAGGTC-3′), GluR3 (NM_001112742.1) (5′-GCTTCGTTTTAGGCGTAGCA-3′ and 5′-GCTCCTGAACCGTGTTTCTC-3′), GluR4 (NM_017263.2) (5′-CGATTTGGAGGGCATAAAAA-3′ and 5′AGAGGCATTGAAGACGATGG-3′), and glutamate ionotropic receptor N-methyl-D-aspartate (NMDA) type subunit 1 (NR1) (NM_017010.1) (5′-ACAAGCCCAACGCCATAC-3′ and 5′-CCGTGCGAAGGAAACTCA-3′). Each primer set was designed between multiple exons, and PCR products were confirmed by sequencing. The abundance of each gene was determined relative to the 18S transcript. Each PCR cycle consisted of denaturation at 95ºC for 30 s, annealing and elongation at 60ºC for 30 s, and denaturation at 95ºC for 5 s, and 40 such cycles were performed.
Rapid Whole-Body Freezing MethodA whole-body freezing method was used to examine the in situ morphology and inner diameter of fetal rat DA specimens, as described previously.19–21 Briefly, after the maternal rats were anesthetized with isoflurane, in utero fetuses on gestational day 21 or day 20 were injected intraperitoneally with normal saline (200 μl), 0.7% of glutamate (200 μl; Wako, Osaka, Japan), 0.7% glutamate (200 μl)+1-naphthylacetyl spermine trihydrochloride (NASPM, 10 μg; Sigma-Aldrich, St Louis, MI, USA), or 0.7% glutamate (200 μl)+prazosin (1 μg; Sigma-Aldrich). NASPM or prazosin was injected 15 min before glutamate administration. The adequacy of anesthesia was monitored by assessment of skeletal muscle tone, respiration rate and response to tail pinch. Thirty minutes after glutamate injection, rat fetuses were delivered by cesarean section and immediately frozen in liquid nitrogen before breathing. The frozen specimens were sectioned under a microscope, and the inner diameters of the DA, PA, and aorta were measured. DA/PA ratios were used to evaluate in vivo DA contraction.
Determination of Noradrenaline Concentration in DA TissuesNoradrenaline concentration in rat DA tissues was measured using an ELISA kit purchased from LDN (Nordhorn, Germany). After the maternal rats were anesthetized with isoflurane, fetuses in utero were injected intraperitoneally on gestational day 21 with normal saline (200 μl) or 0.7% of glutamate (200 μl; Wako, Osaka, Japan). Thirty minutes after the glutamate injection, rat fetuses were delivered by cesarean section and DA tissues were isolated and immediately frozen in liquid nitrogen. The frozen DA tissues were homogenized in buffer containing 10 mmol/L HCl, 1 mmol/L EDTA, and 4 mmol/L sodium metabisulfite and were analyzed using the ELISA assay. The assay was carried out following the manufacturer’s instructions. The value of noradrenaline content measured by ELISA was divided by the total amount of protein in the rat DA tissue sample to obtain the noradrenaline concentration in rat DA tissue (expressed as pmol/mg of protein).
Statistical AnalysisAll values are shown as mean±SEM of more than 3 independent experiments, except for Tables 1–3, in which the results are expressed as mean±SD. In Tables 1–3, the data shown as values were statistically analyzed using the Kruskal-Wallis test, and the data shown as ratios were analyzed using the chi-squared test, both followed by Fisher’s least significant difference post-hoc test, the Mann-Whitney U-test or Fisher’s exact test, as appropriate. The decrease in plasma glutamate concentration between at birth and day 2 was analyzed using the Wilcoxon signed-rank test. A one-way ANOVA followed by the Bonferroni correction was used for multiple group comparisons. A two-tailed unpaired Student’s t-test was used for 2 group comparisons. A value of P<0.05 was considered significant.
| Variables | A 28–29 weeks (n=5) |
B 24–27 weeks without PDA(n=6) |
C 24–27 weeks with PDA (n=5) |
P value | |||
|---|---|---|---|---|---|---|---|
| A vs. B vs. C |
A vs. B | A vs. C | B vs. C | ||||
| Glu | 231.8±128.2 | 289.8±165.9 | 136.5±36.5 | 0.162 | |||
| Ser | 117.9±19.0 | 173.2±63.9 | 385.8±70.5 | 0.319 | |||
| Asp | 15.5±5.7 | 87.1±90.8 | 12.7±4.8 | 0.036* | 0.115 | 0.430 | 0.103 |
| Asn | 40.2±8.3 | 71.4±37.2 | 128.6±24.8 | 0.264 | |||
| Ala | 389.0±56.7 | 621.4±395.2 | 1,215.5±220.8 | 0.601 | |||
| His | 65.8±20.2 | 95.5±33.3 | 225.3±23.2 | 0.097 | |||
| Leu | 104.0±38.8 | 148.9±52.3 | 327.2±56.5 | 0.338 | |||
| Lys | 343.0±8.69 | 326.8±105.8 | 744.0±140.2 | 0.817 | |||
| Phe | 73.4±6.62 | 95.5±36.2 | 178.6±34.3 | 0.328 | |||
| Val | 211.4±49.9 | 227.1±71.5 | 487.6±67.8 | 0.793 | |||
| Trp | 61.8±11.9 | 50.1±19.0 | 88.4±11.6 | 0.137 | |||
| Thr | 248.8±46.6 | 284.8±139.2 | 586.7±108.6 | 0.489 | |||
| Met | 28.8±3.9 | 39.8±13.1 | 78.1±10.5 | 0.172 | |||
| Ile | 75.2±21.8 | 86.1±30.7 | 170.4±27.9 | 0.650 | |||
| Arg | 83.9±19.2 | 105.5±28.1 | 167.4±34.4 | 0.147 | |||
| Tyr | 64.9±11.6 | 82.4±19.1 | 164.9±20.9 | 0.295 | |||
| Pro | 182.6±15.2 | 247.3±49.0 | 439.2±32.4 | 0.046* | 0.020* | 0.231 | 0.122 |
| Gly | 227.4±26.2 | 393.0±100.2 | 742.2±93.1 | 0.015* | 0.006† | 0.055 | 0.275 |
| Gln | 345.8±157.3 | 645.3±288.2 | 1,137.2±150.7 | 0.117 | |||
| Cys | 20.7±7.4 | 18.3±11.5 | 34.4±10.7 | 0.281 | |||
| Tau | 213.0±51.9 | 345.2±209.5 | 223.1±105.4 | 0.526 | |||
Values of amino acids are expressed as μmol/L; *P<0.05; †P<0.01. Glu, glutamate; Ser, serine; Asp, asparatic acid; Asn, asparagine; Ala, alanine; His, histidine; Leu, leucine; Lys, lysine; Phe, phenylalanine; Val, valine; Trp, tryptophan; Thr, threonine; Met, methionine; Ile, isoleucine; Arg, arginine; Tyr, tyrosine; Pro, proline; Gly, glycine; Gln, glutamine; Cys, cysteine; Tau, taurine.
| Variables | A 28–29 weeks (n=5) |
B 24–27 weeks without PDA(n=6) |
C 24–27 weeks with PDA (n=5) |
P value | |||
|---|---|---|---|---|---|---|---|
| A vs. B vs. C |
A vs. B | A vs. C | B vs. C | ||||
| Glu | 265.9±196.7 | 202.3±144.8 | 85.8±26.5 | 0.028* | 1.000 | 0.031* | 0.017* |
| Ser | 206.0±57.2 | 178.8±68.6 | 104.3±41.1 | 0.058 | |||
| Asp | 18.5±7.0 | 56.4±47.8 | 34.6±30.6 | 0.426 | |||
| Asn | 47.8±11.7 | 56.6±20.7 | 30.5±8.8 | 0.025 | |||
| Ala | 215.2±66.0 | 379.8±220.5 | 199.3±62.0 | 0.052 | |||
| His | 81.6±20.5 | 92.4±14.1 | 78.8±13.3 | 0.264 | |||
| Leu | 145.0±30.5 | 174.4±29.9 | 111.3±9.2 | 0.010* | 0.177 | 0.055 | 0.004† |
| Lys | 131.2±36.3 | 230.4±91.3 | 120.5±21.7 | 0.028* | 0.051 | 0.547 | 0.017* |
| Phe | 87.3±29.9 | 91.9±30.5 | 56.3±10.3 | 0.020* | 0.930 | 0.055 | 0.004† |
| Val | 155.2±50.7 | 185.7±41.7 | 111.5±15.0 | 0.025* | 0.329 | 0.222 | 0.004† |
| Trp | 32.6±9.9 | 31.3±7.5 | 12.9±11.8 | 0.015* | 0.792 | 0.036* | 0.008† |
| Thr | 111.8±40.3 | 249.8±172.1 | 128.2±86.4 | 0.094 | |||
| Met | 32.9±6.0 | 37.7±16.6 | 18.2±4.3 | 0.014* | 0.662 | 0.015* | 0.008† |
| Ile | 68.7±11.9 | 68.4±12.8 | 44.7±6.9 | 0.011* | 1.000 | 0.015* | 0.004† |
| Arg | 95.8±11.5 | 93.4±57.0 | 55.4±8.9 | 0.038* | 0.930 | 0.007† | 0.082 |
| Tyr | 170.9±133.8 | 156.4±101.8 | 29.9±10.4 | 0.011* | 0.930 | 0.007† | 0.008† |
| Pro | 193.6±30.0 | 352.5±324.1 | 166.2±31.1 | 0.036* | 0.125 | 0.150 | 0.030* |
| Gly | 344.0±83.1 | 460.7±198.4 | 256.1±59.7 | 0.041* | 0.329 | 0.150 | 0.017* |
| Gln | 419.2±234.1 | 569.5±266.0 | 317.4±75.2 | 0.213 | |||
| Cys | 24.5±5.9 | 22.1±15.5 | 12.9±15.1 | 0.390 | |||
| Tau | 150.3±54.2 | 246.1±166.3 | 115.8±34.1 | 0.110 | |||
Values of amino acids are expressed as μmol/L; *P<0.05; †P<0.01. Abbreviations as in Table 2.
We examined the presence of glutamate inotropic receptor AMPA- and NMDA-type subunits in the rat DA on gestational day 21 and found that, among AMPA-type subunits, GluR1 (subunit 1), GluR3 (subunit 3), and GluR4 (subunit 4) were abundantly expressed in the DA. The GluR2 (subunit 2) expression level in the DA was low compared to that in the brain (Figure 1A). The common subunit for glutamate inotropic receptor NMDA-type NR1 (subunit 1) was not detected in the DA, suggesting that NMDA receptors were not expressed in the DA. We then quantified GluR1, GluR3, and GluR4 mRNA expression in the rat DA and aorta on gestational days 19 (e19) and 21 (e21). Quantitative RT-PCR revealed that the GluR1 mRNA expression level was significantly higher in the DA than in the aorta at both e19 and e21 (Figure 1B), although GluR1 expression in the DA was lower than in the brain (Figure 1A). GluR3 and GluR4 mRNA expression levels were not greater in the DA compared to the aorta (Figure 1B). These data suggested that GluR1 was the primary glutamate receptor subtype in the rat DA.
Expression of glutamate receptor subtypes in the rat ductus arteriosus (DA). (A) Semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) using pooled rat DA and rat brain (B) on gestational day 21. N, negative control. (B) Quantitative RT-PCR using the pooled rat DA and aorta on gestational day 19 (e19) and 21 (e21). n=5–8; *P<0.05; ***P <0.001; NS, not significant.
We then performed immunohistochemistry to examine the GluR1 localization in the DA. A strong immunoreaction against GluR1 was detected in the outer layer of tunica media and adventitia in rat and human DA (Figure 2A, Left panels). Proteins of tyrosine hydroxylase, which is an autonomic nerve terminal marker, were similarly detected in the outer layer of tunica media and adventitia in the rat and human DA (Figure 2A, Right panels). Immunofluorescent staining demonstrated that GluR1 was co-localized with tyrosine hydroxylase (Figure 2B). These data suggested that GluR1 was expressed at the DA autonomic nerve terminal.
Localization of glutamate inotropic receptor (GluR1) in the rat and human ductus arteriosus (DA). (A) Immunohistochemistry for anti-GluR1 antibody in the rat (e21) and human DA. The brown color indicates the GluR1 protein expressed in the outer layer of the tunica media and adventitia. (B) Immunofluorescent stain against GluR1 (red) and tyrosine hydroxylase (green). Merged image of GluR1 and tyrosine hydroxylase is shown in the right panels. Scale bars, 50 μm.
Previous reports suggest that glutamate induces noradrenaline production through postsynaptic GluR1.9,10 Thus, we examined whether glutamate promotes noradrenaline production in the rat DA. Glutamate was administered intraperitoneally to rat fetuses (e21), and plasma samples were obtained 30 min after injection. A significant increase in glutamate concentration in the rat plasma was observed (Figure 3A). We obtained DA tissues 30 min after glutamate administration and measured noradrenaline concentrations in the DA tissues using an ELISA. Glutamate administration caused a significant increase in noradrenaline production in the rat DA tissues (Figure 3B).
Glutamate-induced noradrenaline production in the rat ductus arteriosus (DA). (A) Plasma glutamate concentration 30 min after intraperitoneal glutamate administration in e21 rat fetuses. n=5–6. (B) Noradrenaline concentrations in rat DA tissues, which were obtained from e21 rat fetuses that were intraperitoneally administered saline (control) or glutamate (Glu) 30 min before tissue isolation. n=7–8; *P<0.05.
Because it has been suggested that noradrenaline induces DA contraction in humans and other animals,12–16 we examined the effect of glutamate on rat e21 DA contraction. Intraperitoneal glutamate administration promoted DA contraction (Figure 4), and this effect was attenuated by the calcium-sensitive AMPA receptor antagonist, NASPM (Figure 4). Glutamate-induced DA contraction was also attenuated by the adrenergic receptor α1 blocker, prazosin (Figure 4). These data suggested that glutamate induced rat mature DA contraction through the AMPA receptor and noradrenaline.
Effect of glutamate on mature rat ductus arteriosus (DA) contraction. (A) Representative images of e21 rat DA (arrows) 30 min after intraperitoneal administration of saline (control, 200 μl per fetus) or 0.7% glutamate aqueous solution (Glu, 200 μl per fetus). NASPM (10 μg) or prazosin (1 μg) were combined with 0.7% glutamate aqueous solution. The effect of glutamate was evaluated using a whole-body freezing method. Scale bars, 500 μm. (B) Quantification of (A). n=7–13; ***P<0.001.
Our data demonstrated that GluR1 expression was greater in rat preterm DA than in the aorta. We further examined the effect of glutamate on immature DA contraction. Intraperitoneal glutamate administration significantly induced rat e20 DA contraction (Figure 5), which is similar to the effect on e21 mature DA.
Effect of glutamate on preterm rat ductus arteriosus (DA) contraction. (A) Representative images of e20 rat DAs (arrows) 30 min after intraperitoneal administration of saline (control, 150 μl per fetus) or 0.7% glutamate aqueous solution (Glu, 150 μl per fetus). The effect of glutamate was evaluated using a whole-body freezing method. Scale bars, 500 μm. (B) Quantification of (A). n=11–17, *P<0.05.
Rat fetus data demonstrated that glutamate promotes immature DA contraction, and we further examined concentrations of amino acids in plasma from preterm infants who were admitted in Yokohama City University Hospital. Patient profiles are shown in Table 1. The amount of amino acid intake on day 2 was similar between extremely preterm infants (24–27 weeks’ gestation) with or without PDA and more mature preterm infants who were born at 28–29 weeks’ gestation. No change in the amino acid profile was observed in arterial umbilical cord blood samples in all 3 groups (Table 2). On day 2, plasma glutamate concentration was, however, lower in extremely preterm infants with PDA than in extremely preterm infants without PDA and more mature preterm infants (Table 3). The decrease in plasma glutamate concentration between at birth and day 2 was analyzed using the Wilcoxon signed-rank test, and no significant decrease was observed in extremely preterm infants (24–27 weeks’ gestation) without PDA and the more mature preterm infants (28–29 weeks’ gestation) (P=0.31 and 0.63, respectively). However, in plasma samples from extremely preterm infants with PDA, the glutamate concentration tended to be decreased after birth, although the change did not reach statistical significance (cord blood, 136.5±36.5 μmol/L; day 2, 85.8±26.5 μmol/L; P=0.06). Although these data do not indicate a causal relationship between plasma glutamate concentration and the occurrence of PDA, these data together with the data obtained from rat DA at least partially support our hypothesis that glutamate induces DA contraction.
Parenteral amino acid supplementation has been used widely for extremely preterm infants.6 Although the effects of amino acid administration on neonates were investigated,22–25 the roles of each single amino acid have not been fully investigated. In the current study, we demonstrated the contractile role of glutamate in the preterm rat e20 DA. Based on previous studies of DA contractility and remodeling,1,26,27 the DA in e19–e20 rats seems to have similar immaturity as that of extremely and very preterm human infants. Aminogram data revealed that plasma glutamate levels on day 2 were lower in extremely preterm human infants with PDA than in those without PDA, which partly supports the findings of the in vivo rat experiments.
Glutamate is recognized as an excitatory neurotransmitter in the brain.28 In addition, glutamate is a major substrate for the antioxidant, glutathione,29 and is an important gluconeogenic substrate that enters the citric acid cycle though α-ketoglutarate.30 Although the multiple roles of glutamate have been extensively studied,28 most studies were performed in the field of neuroscience. To the best of our knowledge, this is the first study demonstrating that glutamate induces vascular contraction in vivo.
It has been demonstrated that glutamate stimulates the release of noradrenaline from presynaptic glutamate receptors,10,31 and that glutamate-induced noradrenaline release is associated with calcium influx through N- and P-type calcium channels.31 In these studies, the authors measured the release of [3H]-noradrenaline from hippocampal synaptosomes and found that glutamate induced noradrenaline release in a dose-dependent manner (EC50=5.6 μmol/L) and 100–1,000 μmol/L of glutamate caused almost maximum noradrenaline induction.10,31 In the current study, an increase in plasma glutamate concentration ranging from 2,000 to 3,000 μmol/L increased noradrenaline content and induced contraction in rat e21 DA. Based on our previous study, the degree of this response is similar to that of the indomethacin-induced maximum contraction in rat e21 DA.32 We do not know the exact glutamate concentration at autonomic nerve terminals when glutamate was administrated into fetal rats. Further studies are needed to determine the range of the plasma glutamate concentration that induces clinically relevant DA contraction in vivo. In addition to glutamate, our study showed that the plasma tyrosine concentration was significantly lower in extremely preterm human infants with PDA than in those without PDA. Because tyrosine is converted to L-3,4-dihydroxyphenylalanine (L-DOPA), following the conversion into dopamine and noradrenaline at the catecholamine nerve terminals, other amino acid components; that is, tyrosine, may influence noradrenaline release and should be considered together with glutamate for further study.
There are three families of ionotropic glutamate receptors with intrinsic cation-permeable channels; that is, NMDA, AMPA, and kainite.28 Using the calcium-sensitive AMPA receptor antagonist, NASPM, the current study demonstrated that glutamate contributes to DA contraction though AMPA receptors. Among AMPA receptors, GluR1 was highly expressed in immature and mature rat DA compared to the aorta, and was co-localized with autonomic nerve terminals in the outer layer of the tunica media and adventitia in the human and rat DA. Previous studies suggest that AMPA receptors or GluR1 contributed to noradrenaline release.9–11 Based on these findings and previous reports, we speculate that GluR1 possibly mediates DA contraction.
In accordance with previous observations,15,33 our data demonstrated noradrenaline-mediated DA contraction. It has been demonstrated that incubation with indomethacin increased DA sensitivity to noradrenaline 10-fold, and prostaglandin E attenuated the sensitivity of the DA to noradrenaline.33 Glutamate-induced DA contraction may be attenuated under abundant placental prostaglandin E in utero, and glutamate may have a synergistic effect on indomethacin-induced DA contraction after birth.
In the human fetus, histological analysis showed that there are many adrenergic nerve terminals in the muscular wall of the extremely immature DA (eg, at 18–21 weeks’ gestation).15 It has also been demonstrated that noradrenaline induced significant constriction of the DA even at 20 weeks’ gestation, and this contractile effect was attenuated by administration of the α-adrenergic receptor antagonist, phenoxybenzamine.15 Together with these previous findings, our data suggest that noradrenaline-mediated contraction of the DA might be applicable to extremely preterm human infants.
In the current study, plasma glutamate levels in very preterm human infants (28–29 weeks’ gestation) and extremely preterm human infants (24–27 weeks’ gestation) without PDA who received amino acid supplementation were 265.9±196.7 and 202.3±144.8 μmol/L, respectively. Wu et al reported that the glutamate level in plasma samples taken 2 h after breastfeeding was 133.7±51.4 μmol/L (95% confidence interval, 24–243 μmol/L) in human term infants on day 30 (n=16).34 Another report demonstrated that the glutamate level in plasma samples was 243±110 μmol/L (95% confidence interval, 76–551 μmol/L) in human breast-fed term infants on day 11 (n=28).35 Clark et al reported that the glutamate concentration ranged from 110 to 376 μmol/L in very low birth weight (488–1,454 g) infants who received amino acid parenteral nutrition that was increased daily by 0.5 g·kg–1·day–1 to a maximum of 3.0 g·kg–1·day–1.36 Glutamate concentrations in our study were similar to the above-mentioned previous studies, while preterm human infants (24–27 weeks’ gestation) with PDA showed a significantly lower glutamate concentration; that is, 85.8±26.6 μmol/L.
In the enterally fed healthy preterm human infants, the glutamate concentration was similar between preterm, very preterm, and extremely preterm human infants,37 while previous studies demonstrated a change in aminogram results in disease conditions.38,39 Oladipo et al reported that the plasma glutamate level was lower in neonates who were receiving intensive care and who had a diagnosis that included sepsis, respiratory distress, cardiac malfunction/malformation or gastrointestinal complications than in neonates with uncomplicated birth histories.39 However, no study appears to focus on the presence of PDA in the human aminogram analysis. Our data suggested that the plasma glutamate concentration in extremely preterm human infants with or without PDA was similarly maintained until birth, and only the glutamate level in extremely preterm human infants with PDA was decreased after birth. We do not know the mechanisms by which the glutamate level was impaired in extremely preterm human infants with PDA during the postnatal period despite no difference in the amount of amino acid intake. However, our data from in vitro experiments suggested that glutamate supplementation that is maintained at proper levels may promote DA contraction in preterm human infants. Because previous reports confirmed that several parenteral supplementation doses of amino acids significantly increased glutamate concentration in preterm human infants,22,25 the strategy of glutamate-mediated DA contraction would be feasible.
The composition of amino acid mixtures varies between commercially available solutions. Of note, glutamate concentration is markedly different compared to the other components. Glutamate concentrations were 0.8, 5.0, 8.2, and 10.0 g/L in Pleamin-P, Troph-Amine 10%, Aminosyn-PF 10%, and Prinene 10%, respectively. In this study, we used Pleamin-P, which is the only amino acid solution that is available for preterm infants in our country and it has a lower concentration of glutamate compared to the others. No significant difference was shown between Aminosyn-PF and Troph-Amine for the rate of weight gain, nitrogen balance, and nitrogen retention.40 To the best of our knowledge, the effect of each amino acid solution on occurrence of PDA has not been reported. Further investigation is required to seek better amino acid composition for extremely preterm infants who frequently have PDA.
Although the potential toxicity of exogenous glutamate has been discussed in previous reports that suggested that very high doses of glutamate induce acute neuronal damage,41 emerging evidence in humans suggests that there are no serious adverse effects.42 The importance of amino acid composition; that is, glutamate, needs to be reconsidered in postnatal circulatory adaptation, especially among extremely preterm infants who frequently have PDA.
We thank Haruko Horiguchi, Miho Sato, Fumihiko Ishida, Hiroshi Ogo, Maiko Kita, Michisato Hirata, Takahiro Kemmotsu, Mai Hanaki, and Kazuhiro Iwama (NICU at Yokohama City University Hospital) for their cooperation in obtaining the plasma samples, and Yuka Sawada for technical support in histology.
None.
This study was supported, in part, by JSPS (S.F., 26461639; U.Y., 16H05358, H1605358, 15H05761, 26670506, 25293236; Y.I., 24390200, 25670131), MEXT (Y.I., 22136009), NEDO (Y.I., 60890021), NCVC (Y.I., 22-2-3), AMED (Y.I., 66890005, 66890011, 66890001, 66890023), and the MEXT fund for Creation of Innovation Centers for Advanced Interdisciplinary Research Areas Program in the Project for Developing Innovation Systems (U.Y., 15638648).
