There were no significant differences in baseline blood gas data, pH, glucose and lactate between groups (Table 1). Glucose concentrations were significantly elevated from baseline in the ischaemia-vehicle and -low dose groups between two hours and one day after ischaemia (p < 0.05). In the ischaemia-high dose group glucose levels were only significantly elevated from baseline at day one (p < 0.05). There was a significant increase in lactate concentrations after ischaemia in all groups at day one (p < 0.05), which was not significantly different between groups. A significant reduction in PaO₂ was seen in the ischaemia-high dose group at day one (p < 0.05).

All animals in the ischaemia-vehicle and -low dose groups survived the occlusion and recovery periods (7/7 and 6/6 respectively). In the ischaemia-high dose group all animals survived the occlusion period but two animals died between days two and three after ischaemia (2/7). Both of the animals that died had very high plasma lactate concentrations at day one (18.5 and 20.8 mmol/L) and at day two (8.11 and 11.9 mmol/L) before death. A third fetus developed an elevated lactate level at day one (11.2 mmol/L) but this resolved by day two (0.9 mmol/L).

There were no significant differences in absolute baseline EEG power between groups (ischaemia-vehicle 22.3 ± 1.3 vs. ischaemia-low dose 19.8 ± 3.1 vs. ischaemia-high dose 20.9 ± 1.1 dB). Cerebral ischaemia was associated with rapid, profound suppression of EEG power in all groups (Figure 1), with no significant difference in EEG power during ischaemia. After release of occlusion, EEG activity gradually increased from its nadir, but remained suppressed compared to baseline for approximately eight hours in the ischaemia-vehicle group and then increased simultaneously with the onset of a period of intense seizure activity that lasted until approximately 48 h after ischaemia. In both of the groups that received peptide, EEG power was significantly higher between 2–12 h after ischaemia compared to the ischaemia-vehicle group (p < 0.05). There was no significant difference in the time of seizure onset between the ischaemia-vehicle (285 ± 63 min), ischaemia-low dose (258 ± 32 min) and ischaemia-high dose (373 ± 63 min) groups (p > 0.05). There was a weak inverse correlation between the average amplitude or total duration of seizures, during the peak period of 22–26 h, with a peak lactate concentration at day one in the ischaemia-high dose group (r² = 0.37 and r² = 0.39, respectively). Following the resolution of seizures, total EEG power fell to below baseline levels in all groups. However, in the ischaemia-low dose group EEG power returned towards baseline and was significantly higher than the ischaemia-vehicle and-high dose groups from day five until the end of the experiment.

The normal near-term fetal EEG characteristically shows sleep state cycling (Figure 2A). In the ischaemia-vehicle and-high dose groups EEG power remained reduced and no sleep state cycling was evident at day seven (Figure 2B, D). In contrast, in the ischaemia-low dose group EEG power was near baseline values and clear sleep state cycling was present (Figure 2C).

There were no significant differences in baseline EEG spectral edge between groups (Figure 3). Ischaemia was associated with rapid suppression of spectral edge to below baseline levels, which persisted to day seven in all groups. An apparent transient decrease in spectral edge was seen between 48–72 h after occlusion in the ischaemia-high dose group, which was associated with data dropout due to the death of two fetuses.

Cerebral impedance (a measure of intracellular oedema [32]) increased progressively from approximately five minutes after the start of occlusion (Figure 3). There was no significant difference between the maximum impedance during ischaemia in the ischaemia-vehicle (152.8 ± 3.7%), ischaemia-low dose (145.3 ± 5.5%) and ischaemia-high dose (154.8 ± 3.4%) groups (p > 0.05). After release of occlusion, impedance resolved to near baseline levels over 30 to 45 min. From approximately 18 h onwards a secondary rise in impedance developed in all groups, to a significantly greater peak in the ischaemia-high dose group than the other groups (p < 0.05). By day seven, impedance had dropped below baseline in the ischaemia-vehicle and -high dose groups, whereas it appeared to remain at approximately baseline values in the ischaemia-low dose group (N.S.).

In all groups, carotid artery occlusion was associated with a significant fall in carotid artery blood flow and an increase in fetal heart rate and mean arterial pressure compared to baseline (p < 0.05, Figure 4). Following the release of the occlusion these parameters returned to baseline values. A secondary rise in fetal heart rate, mean arterial pressure, nuchal electromyogram (EMG) and carotid blood flow was seen at approximately 18–36 h. There was no significant difference in any parameter between the groups. Extradural temperature was approximately 39.5 °C, with a small diurnal fluctuation, with no significant difference between groups.

Consistent with our previous report, a low-dose dose mimetic peptide infusion started 90 min after reperfusion from severe cerebral ischaemia was associated with markedly improved recovery of EEG power [13]. In contrast, high-dose infusion of this peptide was not associated with any apparent improvement in EEG recovery, but instead with a significantly greater secondary rise in cerebral impedance (brain swelling), lower blood oxygen levels, and apparent trends to increased plasma lactate values and fetal death. Bilateral carotid artery occlusion is an exclusively cerebral insult, and although it is associated with severe neural injury, there is no systemic compromise [4,32].

There is increasing evidence to suggest that astrocytes play an important role in synaptic function. Astrocytes respond to the synaptic release of glutamate with an increase in intracellular Ca²⁺ and can also modulate neuronal excitability via the release of neuroactive substances called gliotransmitters [33–35]. These gliotransmitters include glutamate, ATP, D-serine and tumour necrosis factor α [36–39]. Such research has led to the concept of the “tripartite synapse”, consisting of astrocytes along with the pre- and post-synaptic terminals [33]. This has been shown to occur when an astrocyte releases glutamate that acts on both the pre- and post-synaptic terminals to modulate synaptic transmission and excitability at excitatory and inhibitory synapses [40–43]. Astrocytic ATP release can modulate excitability via purinergic receptors, whilst ATP that is degraded to adenosine can mediate depression of excitatory synapses [44,45]. It is possible that the early increase in EEG power seen in both the low- and high-dose peptide infusion groups was a result of reduced ATP release from astrocytes via connexin hemichannels. Intriguingly, this early increase in EEG power does not seem to have affected the longer-term recovery of EEG power, which was significantly improved after low-dose but not high-dose infusion. This is consistent with reports both in human infants and in large animals that EEG amplitude during the immediate recovery phase has relatively limited predictive value [46,47].

The relationship between hemichannel and gap junctional activity likely varies with the precise environmental conditions. In cell culture, hypoxia enhanced connexin hemichannel opening but reduced gap junction communication [18]. Conversely, Cotrina and colleagues have demonstrated that gap junction communication remains intact during ischaemia in brain slices [37,48]. Physiological increases in Ca²⁺ have been shown to inhibit gap junction permeability in HeLa cells [49]. Although intracellular Ca²⁺ levels increase rapidly during ischaemia [50], which could reduce gap junction coupling [49], after reperfusion, intracellular Ca²⁺ concentration rapidly falls to baseline level, and is therefore unlikely to affect gap junction coupling in the latent phase.

The mechanism of loss of protection with high-dose infusion of connexin mimetic peptide is unknown, but it is highly likely to be related to blockade of the connexin 43 gap junctions in addition to connexin hemichannels. This is supported by evidence from O’Carroll et al., who showed loss of dye transfer between cells, when the same connexin mimetic peptide as used in this study was applied at 500 μM but not 5 μM doses [51]. In addition to the dose response, there may also be a time course response for mimetic peptide blockade of connexin hemichannels and gap junctions. Gap26, a mimetic peptide which binds to the first extracellular loop of connexin 43, has been shown to block connexin 43 hemichannel currents in less than five minutes in single cells and then gap junction currents after approximately 30 min in cell pairs, following application at a relatively high dose (175 μM) [52,53]. However, that study did not test whether this sequence also occurred at lower doses. Indeed, low-dose Gap26 (0.5 μM) applied either 10 min before or 30 min after the initiation of 40 min of myocardial ischaemia was shown to reduce myocardial infarct size by 48% and 55%, respectively, suggesting hemichannel blockade after ischaemia might be of relatively greater importance than during ischaemia [54,55].

Although transient blockade of gap junctions following ischaemia or injury may be beneficial [9,10,56], long-term blockade, such as that seen in the connexin 43 knockout mouse, has been associated with greater brain injury following ischaemia [11]. Thus, blockade of connexin 43 gap junctions may compromise the astrocytic syncytium and impair vital functions. These functions include glutamate uptake from the extracellular space via the transporters GLT-1 and GLAST, and conversion of glutamate to glutamine which is then released for re-uptake by neurons [57]. Astrocytes are also responsible for spatial buffering of K⁺, whereby uptake of K⁺, cytosolic diffusion through the astrocytic syncytium, followed by release at a distant site, prevents ionic imbalances in active regions [58]. Extracellular K⁺ under normal conditions is approximately 2.7–3.5 mM but can reach 50–80 mM during ischaemia and sustained exposure can trigger death in fetal rat neuronal cultures through enhanced Na⁺ entry into cells [59,60]. Astrocytes may also play a role in energy supply to neurons under compromised conditions via the “lactate shuttle”, the process by which astrocytes convert glycogen to glucose and glucose to lactate (as reviewed in [61]), which can then be transported to neurons and used for oxidative metabolism.

The excessive rise in lactate concentration seen in three animals in the ischaemia-high dose group, despite no significant increase in seizure activity, may be due to impaired glucose trafficking through the astrocytic syncytium from the blood to the neurons. In turn this may result in astrocytes promoting greater glycogenolysis to supply lactate as a fuel for local neuronal function. Alternatively, the decrease in blood PaO₂ raises the possibility that the combination of seizure activity with greater background neural activity may have exceeded the capacity for oxidative metabolism, with lactate being generated from both neurons and astrocytes by anaerobic metabolism. The secondary increase in impedance, a measure of cytotoxic oedema [32], after brain ischaemia is a result of increased intracellular water uptake, predominantly in the glial cells of the grey matter [32,62,63]. Thus the present data suggest that high-dose infusion was associated with disruption of the astrocytic syncytium, which potentiated intracellular swelling, most likely through impaired local ion homoeostasis.

In conclusion, this study suggests that while low dose infusion of a connexin 43 mimetic peptide that blocks gap junction hemichannel opening may prove to be a clinically useful therapeutic intervention following perinatal ischaemia [13], higher doses may be associated with adverse effects on neural recovery and on survival, speculatively, by uncoupling gap junction communication within the astrocytic syncytium.

All procedures were approved by the Animal Ethics Committee of The University of Auckland. In brief, 20 time-mated Romney/Suffolk fetal sheep were instrumented using sterile technique at 118–124 days gestation (term is 145). Food, but not water was withdrawn 18 h before surgery. Ewes were given 5 mL of Streptocin (procaine penicillin (250,000 IU/mL) and dihydrostreptomycin (250 mg/mL, Stockguard Labs Ltd, Hamilton, New Zealand)) intramuscularly for prophylaxis 30 min prior to the start of surgery. Anesthesia was induced by I.V. injection of Alfaxan (Alphaxalone, 3 mg/kg, Jurox, Rutherford, New South Wales, Australia), and general anesthesia maintained using 2–3% isoflurane in O₂. The depth of anesthesia, maternal heart rate and respiration were constantly monitored by trained anesthetic staff. Ewes received a constant infusion isotonic saline drip (at an infusion rate of approximately 250 mL/h) to maintain fluid balance.

Following a maternal midline abdominal incision and exteriorization of the fetus, both fetal brachial arteries were catheterized with polyvinyl catheters to measure mean arterial blood pressure. An amniotic catheter was secured to the fetal shoulder. ECG electrodes (Cooner Wire Co., Chatsworth, California, USA) were sewn across the fetal chest to record fetal heart rate. The vertebral-occipital anastamoses were ligated and inflatable carotid occluder cuffs were placed around both carotid arteries [4,64]. A 3S Transonic ultrasonic flow probe (Transonic systems, Ithaca, NY) was placed around the right carotid artery. Using a 7 stranded stainless steel wire (AS633-5SSF; Cooner Wire Co.), two pairs of EEG electrodes (AS633-5SSF; Cooner Wire Co.) were placed on the dura over the parasagittal parietal cortex (10 mm and 20 mm anterior to bregma and 10 mm lateral) and secured with cyanoacrylate glue. A reference electrode was sewn over the occiput. A further two electrodes were sewn in the nuchal muscle to record electromyographic activity as a measure of fetal movement and a reference electrode was sewn over the occiput. A thermistor was placed over the parasagittal dura 30 mm anterior to bregma. An intracerebroventricular catheter was placed into the left lateral ventricle (6 mm anterior and 4 mm lateral to bregma). The uterus was then closed and antibiotics (80 mg Gentamicin, Pharmacia and Upjohn, Rydalmere, New South Wales, Australia) were administered into the amniotic sac. The maternal laparotomy skin incision was infiltrated with a local analgesic, 10 mL 0.5% bupivacaine plus adrenaline (AstraZeneca Ltd., Auckland, New Zealand). All fetal catheters and leads were exteriorized through the maternal flank. The maternal long saphenous vein was catheterized to provide access for post-operative maternal care and euthanasia.

Sheep were housed together in separate metabolic cages with access to food and water ad libitum. They were kept in a temperature-controlled room (16 ± 1 °C, humidity 50 ± 10%), in a 12 h light/dark cycle. Antibiotics were administered daily for four days I.V. to the ewe (600 mg Benzylpencillin Sodium, Novartis Ltd, Auckland, New Zealand, and 80 mg Gentamicin, Pharmacia and Upjohn). Fetal catheters were maintained patent by continuous infusion of heparinized saline (20 U/mL at 0.15 mL/h) and the maternal catheter maintained by daily flushing.

Data recordings began 24 h before the start of the experiment and continued for the remainder of the experiment. Data were recorded and saved continuously to disk for off-line analysis using custom data acquisition programs (LabView for Windows, National Instruments, Austin, Texas, USA). Arterial blood samples were taken for pre-ductal pH, blood gas, base excess (Ciba-Corning Diagnostics 845 blood gas analyser and co-oximeter, Massachusetts, USA), glucose and lactate measurements (YSI model 2300, Yellow Springs, Ohio, USA). All fetuses had normal biochemical variables for their gestational ages [65,66].

Experiments were performed at 128 ± 1 days gestation. Ischaemia was induced by reversible inflation of the carotid occluder cuffs with saline for 30 min. For Cx43 hemichannel blocking, a peptide (H-Val-Asp-Cys-Phe-Leu-Ser-Arg-Pro-Thr-Glu-Lys-Thr-OH (Auspep, Vic, AU)) that mimics the second extracellular loop of Cx43 (“Peptide 5” reported in [51]) was infused into the lateral ventricle via the intracerebroventricular catheter attached to an external pump. Control fetuses received cerebral ischaemia followed by infusion of the vehicle (ischaemia-vehicle, n = 7). The ischaemia-low dose group (n = 6) received 50 μmol/kg/h for one h followed by 50 μmol/kg/24 h for 24 h, and the ischaemia-high dose group (n = 6) received 50 μmol/kg/h for 25 h dissolved in artificial cerebrospinal fluid (aCSF), at a rate of 1 mL/h for 25 h starting 90 min after the end of the occlusion. Animals were killed at seven days with an overdose of sodium pentobarbitone (9 g I.V. to the ewe; Pentobarb 300, Chemstock International, Christchurch, N.Z.).

Seizure activity was quantified as number, duration and peak amplitude of seizures between 22–26 h (time of peak EEG power) for each animal in the ischaemia-high dose group. This was correlated with lactate concentration at day one. The time of onset of the first seizure in each animal in both groups was determined. Seizures were defined as previously described by Scher et al. [67]. Seizures were visually defined on the raw EEG recording as sudden repetitive, evolving stereotyped waveforms lasting at least 10 seconds with an amplitude greater than 20 μV. Sleep state cycling was identified as alternating periods of high (non-rapid eye movement) and low (rapid eye movement) amplitude activity, with each period lasting approximately 20 min. Data was analyzed using ANOVA or repeated measures ANOVA, followed by the Tukey post-hoc test when a significant difference was found. Statistical significance was accepted when p < 0.05.