Main

HI encephalopathy after perinatal asphyxia remains a significant cause of severe neurologic disability. NMDA receptor is considered important in the development of postasphyxial neuronal injury. Because there are a far greater proportion of NMDA receptor sites in both the immature rat brain cortex and the developing human brain compared with the adult brain of the same species(1), the immature brain may be particularly vulnerable to the excessive glutamate release. Not only glutamates, but also monoamine has become considered to have an important role in hypoxic brain damage(2–5). In a previous study we examined anoxic effects on monoamine metabolism in 5-d-old rat brain by measuring extracellular NE, DA and its metabolites, as well as 5-HIAA, and succeeded in producing a DA surge by anoxic insult without additional provocations(6).

Mg2+ was a first pharmacologic tool used to inhibit selectively the effect of NMDA. It has been shown that glutamates stimulate DA release through an NMDA receptor both in vivo and in vitro(7, 8), and that NMDA causes a concentration-dependent stimulation of DA release(9). In striatal slices, the absence of Mg2+ potentiates the stimulatory effect of NMDA on the synthesis(10) and the release(7) of DA. In addition, the NMDA-evoked release of DA was completely blocked by Mg2+ in the rat striatum(8, 11), and intrastriatal perfusion of Mg2+ totally blocked the effect of NMDA on the release of DA in adult rats(9). However, in neonatal animals, the effect of Mg2+ on DA release during anoxia and after NMDA administration has not been determined. To examine the effect of Mg2+ on monoamine metabolism in anoxic immature brain, we administered MgSO4 to 5-d-old rat pups by IS and by IP. We compared the effects of Mg2+ administered in two different ways after exposure to 16 min of anoxia. We hypothesized that MgSO4 suppressed the monoamine metabolism induced by transient anoxia.

METHODS

Animals. Pregnant Wistar rats (Nihon SLC, Shizuoka, Japan) were housed in separate cages after 15 d of gestation, and pups were born on d 20 or 21 of gestation. The day of birth was considered d 0. Eight rat pups were culled by random selection from a litter 2 d after birth. They were housed in a room with 12-h light-dark cycles, with food and water available ad libitum. One hundred and seventy-six 5-d-old pups from 22 litters were used for the experiment. The littermates were matched by weight and were assigned to all three study groups. The protocol for animal experimentation had been previously reviewed and approved by the Animal Research Committee, Akita University School of Medicine.

Surgery and brain microdialysis. A rat pup was anesthetized for about 2 min with ether, then placed in a Kopf stereotaxic instrument. The skull was exposed, and a hole was drilled directly above the striatum. An AG-4 guide cannula (Eicom, Kyoto, Japan) was implanted into the right striatum. Coordinates for probe placement in the striatum were relative to bregma: A+0.5, L +2.5, V -4.0, according to the atlas of Paxinos and Watson(12). A dialysis probe (A-1-4-02; Eicom) was inserted into the guide cannula so that 2.0 mm of the probe were embedded in the tissue of the striatum. We completed the implantation of the probe within 10 min. Pups were kept separated from their dams and in a fasting state throughout the experiment. On applying the microdialysis technique to the immature brain, one major limitation is nourishment. We adopted a 3-h stabilization period because pups could not be returned to their dams after probe insertion. However, Benveniste et al.(13) reported that 24-h stabilization after dialysis probe insertion was necessary, because insertion of a microdialysis probe caused regional damage, namely, increased glucose metabolism or decreased blood flow, especially in the first 2 h(13). According to this report, baseline levels may be artificially elevated in our experiment.

Probes were perfused with Ringer's solution (NaCl 147 mM, KCl 4 mM, CaCl2 1.2 mM, pH 7.4) at a flow rate of 4.0 μL/min using a microinfusion pump (EP-60; Eicom). After stabilization for 3 h to minimize the effect of the insertion, the pups were divided into three groups. 1) IS group (n = 20): striatum was perfused with MgSO4 (10 mM) from the beginning of the sampling to the end of anoxia. MgSO4 was dissolved in Ringer's solution. Flow rate was 4.0 μL/min. 2) IP group (n = 20): MgSO4 was injected intraperitoneally at the beginning of sampling. MgSO4 was dissolved in isotonic saline(pH 7.4, injection volume = 0.05 mL). The 0.3 mg/kg MgSO4 was injected intraperitoneally using a 1-mL syringe with a 30-gauge needle. Based on preliminary administration of different concentrations of Mg2+, the maximum dose at which the pups' general conditions could be maintained was determined in both the IS and the IP groups. 3) Control group(n = 40): 20 pups received NaCl by intraperitoneal injection, whereas another 20 pups received an intrastriatal infusion of Ringer's solution.

After the treatment, samples of the dialysates were collected in microtubes at 16-min intervals and placed in an ice bath. The first two samples were used as a basal level of extracellular neurotransmitters. During the next sampling, the pups were exposed to anoxia.

Anoxia was achieved by placing the pups in a 3.0-L airtight plastic box and flushing it with warmed and humidified 100% nitrogen delivered at 0.5 L/min. The oxygen concentration in the plastic box was measured at regular intervals with an oxygen analyzer to maintain it at 0 ≈ 2%. The final eight samples were obtained in 21% oxygen. Body temperature was monitored throughout the experiment with a rectal probe by a thermometer, Thermotrack (Create Medic, Yokohama, Japan). From the surgery to the final sampling, pups were placed in an infant incubator V-75 (Atom, Tokyo, Japan), the temperature of which was adjusted manually to about 35 °C to maintain the pups' body temperature at 37.5 °C. Although the localization of the probe was histologically verified in some pups, it was confirmed mostly by visual examination of horizontal brain sections at the end of each experiment.

To examine alteration in Mg2+ levels in both plasma and in the perfusates, a further two sets of pups 32 and 30 animals, respectively, were subjected to the same treatment without anoxia, namely, 32 pups received MgSO4 intraperitoneally and 30 received MgSO4 intrastriatally. After brief anesthesia with ether, blood was withdrawn from a cardiac puncture and centrifuged at various time points, before, during, and after Mg2+ administration. Separate animals were used for each cardiac puncture. The number of pups used for each time point was four. Samples for Mg2+ were immediately processed in an electrolyte analyzer, IONOLYTE EM-5100 (Taiyo Yuden, Tokyo, Japan) by the ion selective electrode method.

During the experiment a total of 23 pups died. These animals were excluded from the study.

HPLC with electrochemical detection. The dialysates were kept at -80 °C until analysis. Monoamines were quantified by HPLC with electrochemical detection. Using an autoinjector (model-231 XL; Gilson, France), 40 μL of dialysate from each microtube were submitted to HPLC every 30 min to determine the concentrations of monoamines. The HPLC system consisted of a liquid chromatograph pump (655 A-11; Hitachi, Tokyo, Japan), an ECD-100 electrochemical detector (Eicom), and an integrator (Chromotopac C-R6A, Shimadzu, Kyoto, Japan). The detector potential was set at 650 mV versus Ag/AgCl reference electrode. A reverse phase column (150× 3.9 mm) filled with Micro Bondasphere C 18 (Nihon Waters, Tokyo, Japan) was used. The mobile phase was a mixture of 0.1 M NaH2PO4, 0.1 mM EDTA, 2.6 mM octyl sodium sulfate, 0.25 mM triethylamine adjusted to pH of 3.4 with phosphate, and 16% methanol. Separations were conducted isocratically at 27 °C. The flow rate was set at 0.7 mL/min. The detection thresholds of the assay are shown in Table 1.

Table 1 Basal levels and detection thresholds of monoamines
Full size table

Statistical analysis. Comparisons were made among the IP, IS, and control groups using two-way repeated measures ANOVA for monoamine analyses. If the results of ANOVA showed statistical significance, pairs of data sets were compared using the Mann Whitney U test at each time point for monoamine analyses, whereas in Mg2+ analyses, differences from basal levels were compared using Wilcoxon's ranked signed sum test. We also performed one-way repeated measures ANOVA to show the changes with time in both monoamine and Mg2+ concentrations in each group. Differences of p < 0.01 were considered significant.

RESULTS

Alteration in Mg2+concentration (Fig. 1,A and B). The basal level of plasma Mg2+ in the IP and the IS group and that of Mg2+ in perfusates in both groups was 0.37 ± 0.02 and 0.4 ± 0.03 mmol/L, respectively. Neither of the basal levels of Mg2+ differed significantly between the two groups. In the IP group, the plasma Mg2+ level began to elevate as soon as MgSO4 was injected, reaching a peak(1.56 ± 0.07 mmol/L) 32 min after Mg2+ injection, whereas in the IS group, there was no significant change (Fig. 1A). The level of Mg2+ in the perfusates in the IS group reached a peak (7.9± 0.07 mmol/L) 48 min after Mg2+ administration, and was significantly higher (p < 0.01) than baseline until 80 min after Mg2+ perfusion; although in the IP group it did not alter significantly (Fig. 1B).

Figure 1
figure 1

(A) Plasma Mg2+ level in both Mg2+ IP and IS groups. In the IP group (closed square, n = 32), MgSO4 was injected intraperitoneally at the beginning of sampling(arrow). In the IS (closed circle, n = 30) group, striatum was perfused by MgSO4 for 48 min from the beginning of the sampling. After pups were anesthetized briefly with ether, blood was withdrawn from a cardiac puncture and collected at 16-min intervals for 128 min. Separate animals were used for each cardiac puncture and the number of pups used for each time point was four. Statistical analyses were performed in each group by one-way repeated measures ANOVA, and if the results were significant, differences from the basal level were compared using Wilcoxon's ranked signed sum test.*p < 0.01, for difference from the basal levels by Wilcoxon's ranked signed sum test. (B) Mg2+ concentration in perfusates in both the Mg2+ IP and the IS group. Sampling of perfusates was initiated 3 h after probe insertion, and perfusates were collected at 16-min intervals. Flow rate was 4.0 μL/min. Perfusates were collected consecutively until blood was taken to measure plasma Mg2+. The experimental protocol was as in A.

Full size image

Alteration of monoamines. The average of the first two samples was considered the basal level and was taken as 100%. All subsequent samples were expressed in percentages relative to the basal levels. The basal level of each metabolite did not differ significantly among the three groups, as shown in Table 1.

Alteration in DA levels induced by anoxia significantly differed among the three groups (two-way ANOVA). A “DA surge,” a sharp and marked increase in striatal DA, was detected in controls during anoxia(Fig. 2, p < 0.001, one-way ANOVA). The DA level reached 1493 ± 408% of the basal level during anoxia, declined rapidly to 321 ± 58% in the next 16 min, and returned to the basal level over the following 48 min. Although the peak concentration of plasma Mg2+ was achieved at the beginning of anoxia in the IP group (Fig. 1), DA increased (the peak level was 1608 ± 402% of the basal level) as in the control group. In contrast, the peak level of DA was only 185 ± 49% in the IS group. The DA surge was significantly and markedly suppressed (p < 0.01, Mann Whitney U test).

Figure 2
figure 2

Effect of MgSO4 on striatal extracellular DA during anoxia in 5-d-old rats. Sampling of perfusates was initiated 3 h after probe insertion, and perfusates were collected at 16-min intervals. The protocol included a basal level period (32 min), anoxic period while animals were exposed to 100% nitrogen (16 min), and a recovery period in 21% oxygen(128 min). In the IP group (closed square, n = 20), MgSO4 was injected intraperitoneally at the beginning of sampling (arrow). In the IS(closed circle, n = 20) group, striatum was perfused by MgSO4 from the beginning of the sampling to the end of anoxia (painted area, 48 min). In the control group (open triangle, n = 40), NaCl and Ringer's solution were used in place of MgSO4. Flow rate was 4.0μL/min. Data are expressed as percentages of the basal level and are mean± SEM (bars) values. Statistical analyses were performed using two-way repeated measures ANOVA among the IP, IS, and the control groups, and one-way repeated measures ANOVA for the changes with time in monoamine concentrations in each group. If the results of two-way repeated measures ANOVA showed statistical significance, pairs of data sets were compared using Mann Whitney U test at each time point in monoamine analyses. Differences of p < 0.01 were considered significant. ‡p < 0.001, by one-way ANOVA. *p < 0.01, for difference from the basal levels by Mann Whitney U test.

Full size image

The levels of other metabolites were also altered. The striatal extracellular DOPAC decreased transiently (Fig. 3), and HVA in each group declined significantly (Fig. 4,p < 0.001, one-way ANOVA). NE (Fig. 5) increased during anoxia, whereas the level of 5-HIAA gradually rose from the end of anoxia, reaching about 140% of the basal level by the end of the recovery period (Fig. 6, p < 0.001, one-way ANOVA, respectively). The alterations in the level of these metabolites induced by anoxia did not significantly differ among the three groups (two-way ANOVA).

Figure 3
figure 3

Effect of MgSO4 on striatal extracellular DOPAC during anoxia in 5-d-old rats. See the legend of Fig. 2 for a description of the experimental protocol. ‡p < 0.001 and † p < 0.01, by one-way ANOVA, respectively.

Full size image
Figure 4
figure 4

Effect of MgSO4 on striatal extracellular HVA during anoxia in 5-d-old rats. See the legend of Fig. 2 for a description of the experimental protocol. ‡p < 0.001, by one-way ANOVA.

Full size image
Figure 5
figure 5

Effect of MgSO4 on striatal extracellular NE during anoxia in 5-d-old rats. See the legend of Fig. 2 for a description of the experimental protocol. ‡p < 0.001, by one-way ANOVA.

Full size image
Figure 6
figure 6

Effect of MgSO4 on striatal extracellular 5-HIAA during anoxia in 5-d-old rats. See the legend of Fig. 2 for a description of the experimental protocol. ‡p < 0.001, by one-way ANOVA.

Full size image

DISCUSSION

In this study we showed that Mg2+ administered intrastriatally but not intraperitoneally inhibited the DA surge induced by anoxia in 5-d-old rat. To our knowledge, this is the first study to examine the effect of Mg2+ on monoamine metabolism in the anoxic immature brain. Many studies have shown that Mg2+ can limit neuronal injury resulting from hypoxia, ischemia, and traumatic brain damage in vitro and in vivo(14–17). In addition, recent clinical studies suggest that antenatal exposure to Mg2+ therapy may be associated with improved neonatal survival and lower risk of cerebral palsy(18–20).

Several mechanisms could account for magnesium's neuroprotective properties. First, Mg2+ is a selective noncompetitive antagonist of the NMDA receptor(21) and a calcium channel antagonist(22). Mg2+ can prevent excessive neurotransmitter release(23). Physiologic or high concentrations of Mg2+ reduce responses to NMDA by blocking the NMDA-associated ionophore(21, 24). Indeed, a relief of the Mg2+ block has been proposed as the primary event resulting in NMDA receptor activation(25). Second, supplements of Mg2+ itself may have a neuroprotective effect. Mg2+ is essential for many critical enzymatic reactions, including those of glycolysis and oxidative phosphorylation(26). Furthermore, optimum Mg2+ concentrations are required for DNA transcription and protein synthesis(27), as well as a variety of plasma membrane functions.

In addition, there is evidence that DA increase has a deleterious effect on ischemic-hypoxic brain. Because the inactivation of DA by oxidative deamination and by reuptake into presynaptic neurons requires oxygen, the effects of DA released within the ischemic area might be potentiated(2). The oxidized DA generated hydroxydopamine, a neurotoxin(5), moreover, prior substantia nigra lesioning attenuated postischemic striatal neuronal death, and the protective effect was related to inhibition of excessive DA(4). Thus, preventing DA increase might protect immature brain from anoxic neuronal injury.

Two possible causes of the significant increase of extracellular DA are reduced reuptake activity of DA(28), and leakage from intracellular compartments(29, 30). Hypoxia and acidosis contributed to the brain cell membrane damage(31, 32), which occurred coincidentally with calcium influx and ATP loss during hypoxia(33). Also in our previous study anoxia produced significant acidosis and an increase in plasma Ca2+ in 5-d-old rat pups (data not shown). Therefore, it is possible that anoxia damaged the brain cell membrane and induced the greater leakage of DA, although Mg2+ maintained Na+,K+-ATPase activity against the hypoxic condition in newborn piglets in vivo(17) and in cortical slices of adult rat brain(34). Na+,K+-ATPase is a marker of neuronal membrane function. Maintenance of membrane function accordingly may have been one mechanism by which the striatal DA surge in the IS group during anoxia was inhibited.

The alteration with time of each metabolite in each group differed significantly. However, alterations in DOPAC, HVA, 5-HIAA, and NE did not significantly differ among the three groups. It is not clear why the alteration was not inhibited by Mg2+.

Although we did not directly measure Mg2+ in striatal tissue, we did in both plasma and the striatal perfusates. In the IS group, only Mg2+ in perfusates was elevated significantly; the DA surge was significantly suppressed. In the IP group, Mg2+ in the plasma but not in the perfusates was significantly elevated during anoxia, whereas the DA surge was not inhibited. Because the striatal tissue around the probe in the IS group was considered to contain high concentrations of Mg2+, the elevated level of Mg2+ in the striatum might have a supplemental effect of Mg2+ itself.

Systemically administered Mg2+ must cross the BBB to have a neuroprotective effect. There has been controversy whether Mg2+ enters the CSF from plasma in sufficient concentration to alter CNS function(35, 36). Oppelt et al.(36) found that plasma Mg2+ levels after 3-4 h were as high as 300-400% above normal in adult dogs, whereas Mg2+ in CSF rose to a maximum of only 21% above control. However, recent studies have shown that Mg2+ can cross the BBB in neonates(37–39). In addition, BBB permeability is considered greater in infants during seizures and HI. When administered Mg2+ to neonatal swine with i.v. infusion for 60 min, the plasma Mg2+ level was greatest at the end of the infusion period(37). In our experiment, the peak level of plasma Mg2+ was obtained 32 min after IP. The finding that the DA surge in the IP group was not inhibited in spite of the increase in the plasma Mg2+ level does not imply that Mg2+ did not cross the BBB. We did not examine magnesium's ability to penetrate the BBB in this study.

As to why Mg2+ did not inhibit the DA surge in the IP group, we speculate that the reason is that a high concentration of Mg2+ was not maintained in the striatum. In a study of newborn piglets by Hoffman et al.(17), larger doses of MgSO4 given before and during hypoxia by a continuous infusion had a neuroprotective effect. There is a major difference between their study and ours, namely, their study was performed in cortex, whereas ours was done in the striatum. However, compared with their study, in which plasma Mg2+ levels were elevated from 1.9 mg/dL (1.56 mmol/L) to 17.7 mg/dL (14.6 mmol/L), it is possible that our peak plasma Mg2+ level (1.54 mmol/l) was too low and lasted too briefly. Intraperitoneal Mg2+ administration did not have any effect on DA inhibition in our study, yet clinical studies with systemic Mg2+ administration have shown a neuroprotective benefit. One possible reason for this contradiction is the various concentrations of striatal Mg2+ due to different methods of administration (whether by an injection or by a continuous infusion and the size of doses used). Although we considered that excessive DA has harmful effects on hypoxic brain, many mechanisms and various conditions contribute to the development of brain damage in clinical cases besides DA surge. Therefore only failure in inhibiting the DA surge may not necessarily mean that systemically administered Mg2+ has no protective effect on hypoxic brain damage.

In a recent study that evaluated acute effects of two different doses of MgSO4 in 15 infants with birth asphyxia, the larger dose(400 mg/kg) had an unacceptable risk of hypotension and a neuromuscular blocking effect(40). When administering Mg2+ to asphyxial infants, one should be careful of side effects that may result in an even more serious condition. Considering that a DA surge has harmful effects on the anoxic brain, we expect that Mg2+ administration may have some protective effects partly by inhibiting the DA surge. Whether Mg2+ crosses the BBB or not, maintaining a high Mg2+ concentration in the brain would enable the clinical use of Mg2+ as an effective and safe drug. It could remove the threat of side effects such as bradycardia, hypotension, and muscle paralysis, including respiratory muscle paralysis.