Effect of pregnane steroids on electrocortical activity and somatosensory evoked potentials in fetal sheep

Neuroscience Letters 253 (1998) 111–114

Effect of pregnane steroids on electrocortical activity and somatosensory evoked potentials in fetal sheep M.B. Nicol*, J.J. Hirst, D.W. Walker Department of Physiology, Monash University, Clayton, Victoria, 3168 Australia Received 19 March 1998; accepted 24 July 1998

Abstract The effect of infusing the neuroactive steroids pregnanolone and iso-pregnanolone on somatosensory evoked potentials (SEP) and electrocortical (ECoG) activity was studied in unanaesthetised fetal sheep, 130–135 days gestation. Intravenous infusion of pregnanolone (6 mg/kg per h) significantly increased the proportion of high voltage ECoG (56.1 ± 4.8% vs. control 43.5 ± 3.2%, P , 0.05), and decreased low voltage ECoG (43.9 ± 4.8% vs. control 56.6 ± 3.2%, P , 0.05). Pregnanolone treatment decreased the amplitude of the N25 peak of the SEP (89.9 ± 2.8% of control, P , 0.05) evoked following stimulation of the skin of the upper lip. In contrast, iso-pregnanolone treatment had no effect on ECoG activities, or on the amplitude and latency of peaks in the SEP. We conclude that 3a-hydroxy pregnane steroids are active at GABAA receptors in fetal sheep and can modulate sleep/wake activity before birth.  1998 Elsevier Science Ireland Ltd. All rights reserved

Keywords: Pregnane steroids; Sleep/wake; GABAA receptors; Fetus; Evoked potentials

During late gestation in sheep the placenta secretes increasing amounts of progesterone [1], resulting in high plasma concentrations of metabolites in the fetal circulation [5]. From approximately 115–120 days gestation (GA) fetal sleep states become evident as co-ordinated changes of electrocortical (ECoG), electro-ocular (EOG), postural muscle electromyographic (EMG) activities, and fetal breathing movements (FBM) [2]. These parameters indicate that the late gestation fetal sheep spends approximately 95% of the time in utero in either SWS or REM sleep. The remainder of the time occupied by brief periods of a more ‘aroused’ or ‘wakeful’ state that are distinguished by postural muscle EMG activity during periods of low voltage (LV) ECoG [4,10]. The incidence of fetal arousal and EOG activities is reduced by administration of progesterone to the ewe [8]. These findings indicate that changes in circulating progesterone concentrations can produce marked alterations in fetal sleep states.

* Corresponding author. Tel.: +61 3 99052562; fax: +61 3 99052547; e–mail: [email protected]

Certain progesterone metabolites, particularly the 3ahydroxy pregnanes, bind to the steroid site on the GABAA receptor [7] and are therefore potentially capable of producing effects on CNS activity. Placental progesterone metabolites that are present during late gestation may act at GABAA receptors in the fetal brain. The aim of this study was to examine the effect of the infusion of one of the most potent GABA-active steroids, pregnanolone (5b-pregnane3a-hydroxy-20-one), on the composition of fetal sleep states, and somatosensory evoked potentials (SEP) obtained by stimulation of the skin of the upper lip. To verify that the actions were specific to pregnanolone, the effect of the inactive isomer [6], iso-pregnanolone (5a-pregnane-3b-hydroxy-20-one), was also investigated. The SEP is a useful tool in assessing the function and integrity of the CNS [3]. In this study, we have used the SEP to gauge the ability of the fetus to respond to sensory stimulation. We assume that the degree of arousal of the fetus will effect its ability to respond to these types of stimuli, and the SEP can therefore be used to assess any possible change in responsiveness produced by treatment with steroids. HV and LV ECoG states are known to have an

0304-3940/98/$19.00  1998 Elsevier Science Ireland Ltd. All rights reserved PII S0304- 3940(98) 00627- 2

112

M.B. Nicol et al. / Neuroscience Letters 253 (1998) 111–114

effect on the latency of certain peaks produced by upper lip stimulation in the fetal sheep, with longer latencies observed during LV ECoG state, compared to those observed during HV [3]. Eight pregnant Merino-Border Leicester cross ewes of known gestational age were studied. The use of the animals and the procedures performed was approved by the Monash University Standing Committee on Ethics and Animal Experimentation. Fetal and maternal vascular catheters were implanted at 125 days GA as previously described [8]. Insulated multistranded stainless steel wire (Cooner Wire, Chatsworth, CA, USA) electrodes were implanted in the skull above the fetal parietal cortex to record ECoG, as previously described [2]. In addition, a six-wire electrode bundle was implanted to record the fetal SEP. These electrodes were placed in holes drilled bilaterally through the skull, so as to rest on the dura overlying the somatosensory cortex, using the co-ordinates of Cook et al. [3], and were secured with cyanoacrylate glue. Six stimulating electrodes were sewn into the skin of the upper lip of the fetus, with three electrodes placed 5–10 mm apart on either side of the nose. Signals from the pair of ECoG electrodes over the parietal cortex were recorded using high input impedance, differential AC amplifiers with the bandpass filters set at 0.3–15 Hz, and with the output signal displayed on a polygraph (Grass Instrument, Quincy, MA, USA) using a chart speed of 5 mm/min. Signals from the ECoG electrodes used to record the SEP were passed to an analogue-todigital converter (MacLab; AD Instruments, Castle Hill, NSW, Australia) and displayed using software (Scope; AD Instruments) configured to display the signal averaged response to repetitive stimulation of the lip. Two of the six stimulating electrodes sewn into the upper lip of the fetus were used to deliver the stimulus train. Preliminary studies were performed to determine the pair that gave the most consistent response at the level of stimulation that did not alter the ongoing pattern of ECoG activity. The Scope program was used to generate a stimulus train of variable amplitude (0.2–2.0 V), but fixed duration (2 ms) and frequency (1 Hz). Each stimulus triggered a Scope recording from the somatosensory cortex electrodes of 200 ms duration. Each stimulus train consisted of 64 individual pulses delivered to the upper lip at 1 Hz at 4-min intervals, and the resultant sweeps from the somatosensory cortex electrodes were averaged and displayed in real time using the Scope program. Two hours of control recording was obtained before infusions during which the SEP was recorded every 4 min, as described above. Each fetus then received a 2-h infusion of pregnanolone (20 mg/h) or iso-pregnanolone (20 mg/h) via the jugular vein catheter. Both steroids were dissolved in the parenteral nutrition solution, Intralipid (Baxter Healthcare, Sydney, Australia). The infusion rate was 6 mg/kg per h, based on fetal weight derived from a weight for age nomogram. Each averaged response was analysed using a routine

written in IgorPro (Wavemetrics, Lake Oswego, OR, USA) that allowed identification of the positive or negative peaks in the SEP falling within a given time period. All SEP sweeps were analysed by focusing on two peaks, labelled P10 and N25 (Fig. 1), from which the amplitude and latency were determined. The proportion of time occupied by HV and LV ECoG state was calculated for each hour of the control period and the subsequent infusion. Changes in the SEP latency and amplitude, and the incidence of each ECoG state before and after infusion of pregnanolone or iso-pregnanolone were tested using the Student’s paired t-test. P ≤ 0.05 was considered to be significant. The mean incidence of HV and LV ECoG activity from 15 infusions in eight animals given separate pregnanolone and iso-pregnanolone infusions is shown in Table 1. The pregnanolone infusion produced a significant increase in the incidence of HV activity, along with a significant decrease in the incidence in LV activity (P , 0.05). In contrast, iso-pregnanolone infusion had no effect on the incidence of HV and LV ECoG (Table 1). Fig. 1 shows an example of the effect of pregnanolone on SEP. The left hand column of sweeps shows SEP recorded during one hour of control, and the right hand column of sweeps shows SEP recorded during the second hour of a two hour pregnanolone infusion. The SEP was reasonably consistent throughout the control period, with no marked dif-

Fig. 1. Effect of pregnanolone on SEP during HV and LV ECoG state. SEP recorded during HV are shown in solid black lines, whilst those recorded in LV are shown in broken black lines. The left hand side shows 1 h of control recording, with the time prior to beginning the pregnanolone infusion shown on the vertical axis (min), and the horizontal axis showing the actual time of the individual sweep. Arrows indicate the first positive peak, to the right of the stimulus artifact, that was designated P10, and the next negative peak to the right, labelled N25. Each sweep is separated by 4 min. The right hand side of the figure shows the second hour of a 2-h pregnanolone infusion. There was a decrease in amplitude of the N25 peak recorded during HV in the second hour of pregnanolone infusion, compared to control.

113

M.B. Nicol et al. / Neuroscience Letters 253 (1998) 111–114

ferences for the P10 and N25 amplitude and latencies between the periods of HV and LV ECoG activity. However, during the second hour of the pregnanolone infusion there was a clear reduction in amplitude of the N25 peak during HV ECoG activity (Fig. 1). The mean data for eight infusions in eight animals is shown in Table 2. Pregnanolone infusion produced a significant decrease in the amplitude of the N25 peak of the SEP during HV ECoG activity (compared to the amplitude observed during HV activity in the control period, P , 0.05). There was also a tendency for the amplitude of N25 peak to decrease during LV after infusion of pregnanolone and its amplitude to increase during iso-pregnanolone infusion, but these results were variable and the mean changes for the group were not significant. There was no significant effect of the pregnanolone infusion on the amplitude of the P10 peak during either the HV or LV ECoG state. The iso-pregnanolone treatment had no significant effect on the amplitude of either the P10 or N25 peaks during both HV and LV activity. The P10 and N25 peaks of the SEP occurred consistently at 10 and 25 ms during the control periods preceding the two steroid treatments. Pregnanolone treatment produced a significant increase in the P10 latency during HV (12.3 ± 0.9 vs. control, 11.7 ± 0.9; P = 0.02), but had no effect on the latency of P10 during LV (12.2 ± 1.1 vs. control 11.6 ± 0.9; P = 0.09). The mean increase in P10 latency observed during HV and LV was the same (0.6 ms) during both states. In contrast, iso-pregnanolone infusion had no effect on the latency of either peak during LV or HV (data not shown). This study shows that the neuroactive steroid, pregnanolone, has effects on the amplitude and latency of early peaks in the fetal SEP, whereas its inactive isomer, iso-pregnanolone, had no significant effect on the latency. However, isopregnanolone tended to increase N25 peak amplitude, an effect that might reflect its action as a weak competitive agonist that displaces active endogenous steroids. It is well documented that the 3a-hydroxy pregnane steroids are far more potent modulators of the GABAA receptor, whereas the 3b-hydroxy steroids have some affinity, but little efficacy [6]. The stereospecificity and short timecourse of the pregnanolone effect is consistent with an action at the GABAA receptor. Short latency somatosensory evoked potentials such as Table 1 Incidence of high (HV) and low (LV) voltage ECoG in fetal sheep before and after infusion of either pregnanolone (20 mg/h, n = 8) or iso-pregnanolone (20 mg/h, n = 7)

HV LV

Control

Pregnanolone

Control

Iso-pregnanolone

43.5 ± 3.2 56.6 ± 3.2

56.1 ± 4.8* 43.9 ± 4.8*

35.6 ± 3.2 64.3 ± 3.2

40.0 ± 5.5 59.9 ± 5.5

Each fetus received infusions on separate days. Data are % of total recording time occupied by HV or LV. *P , 0.05.

Table 2 Amplitude of SEP observed during high (HV) and low (LV) voltage ECoG in fetal sheep before and after infusion of either pregnanolone (20 mg/h, n = 8) or iso-pregnanolone (20 mg/h, n = 7) Pregnanolone

Isopregnanolone

HV LV

99.7 ± 20.7 96.7 ± 13.5

104.1 ± 21.9 91.7 ± 15.7

HV LV

89.9 ± 2.8* 84.3 ± 6.8

167.0 ± 30.3 128.6 ± 21.0

P10 peak amplitude

N25 peak amplitude

Data are % of SEP amplitude during the pre-treatment control period. *P , 0.05.

those recorded here have been recorded in fetal sheep in other studies [3], where it was suggested that peaks with a latency .20 ms were most likely to be of cortical origin, whilst peaks with latencies ,20 ms were said to be of subcortical origin. There are GABAA receptors present of the cortex of the fetal sheep [11], which could have a direct influence on the N25 peak. Cook et al. [3] suggested that the N10 peak of the SEP recorded in their study could represent activity in the trigeminal sensory system, and this is likely to be the case for P10 observed in the current study. A second important and new finding was that infusion of a GABAA agonist steroid produced alterations in the composition of the fetal sleep patterns. Pregnanolone infusion produced a significant shift in the proportions of HV to LV ECoG activity. The increased incidence of the HV state is indicative of an increase in the presence of SWS, and a decrease in either REM or arousal, or a combination of the two. The composition of sleep states is known to vary with the gestational age of the ovine fetus, with the incidence of REM sleep decreasing, and SWS steadily increasing with age [9]. The development of organised fetal sleep states begins around 115–120 days GA, when the ECoG differentiates into HV and LV states, and breathing activity appears during LV [2,10]. Maternal progesterone concentrations reach their highest levels in the pregnant sheep between 115 and 130 days GA [1]. The activity of progesterone metabolites at GABAA receptors in the fetal brain may contribute to the increase in HV ECoG observed during the last 25 days of gestation. In summary, treatment of fetal sheep with the GABAA agonist steroid pregnanolone, produced a marked increase in the incidence of HV ECoG activity, a significant decrease in the amplitude of the N25 peak of the SEP, and an increase in the latency of the P10 peak, during HV ECoG. These actions were specific to the 3a-hydroxy GABA agonist steroid, as the 3b-isomer had no effect. The present findings support our contention that endogenous 3a-hydroxy pregnane steroids of placental origin may influence the composition of sleep/wake behaviour, and have a suppressive action on the fetal CNS during late gestation.

114

M.B. Nicol et al. / Neuroscience Letters 253 (1998) 111–114

[1] Bassett, J.M., Oxborrow, T.J., Smith, I.D. and Thorburn, G.D., The concentration of progesterone in the peripheral plasma of the pregnant ewe, J. Endocrinol., 45 (1969) 449–457. [2] Clewlow, F., Dawes, G.S., Johnston, B.M. and Walker, D.W., Changes in breathing, electrocortical and muscle activity in unanaesthetized fetal lambs with age, J. Physiol., 341 (1983) 463–476. [3] Cook, C.J., Gluckman, P.D., Johnston, B.M. and Williams, C., The development of the somatosensory evoked potential in the unanaesthetized fetal sheep, J. Dev. Physiol., 9 (1987) 441– 445. [4] Dawes, G.S., Fox, H.E., Leduc, B.M., Liggins, G.C. and Richards, R.T., Respiratory movements and rapid eye movement sleep in the foetal lamb, J. Physiol., 220 (1972) 119– 143. [5] Dolling, M. and Seamark, R.F., Progestagen metabolites in fetal sheep plasma: the effect of fetal nephrectomy, J. Dev. Physiol., 1 (1979) 399–413.

[6] Harrison, N.L., Majewska, M.D., Harrington, J.W. and Barker, J.L., Structure-activity relationships for steroid interaction with the g-aminobutyric acidA receptor complex, J. Pharmacol. Exp. Ther., 241 (1987) 346–353. [7] Harrison, N.L. and Simmonds, M.A., Modulation of the GABA receptor complex by a steroid anaesthetic, Brain Res., 323 (1984) 287–292. [8] Nicol, M.B., Hirst, J.J., Walker, D. and Thorburn, G.D., Effect of maternal progesterone administration on fetal behavioural state during late gestation, J. Endocrinol., 152 (1997) 379–386. [9] Szeto, H., Behavioural states and their ontogeny: animal studies, Semin. Perinatol., 16 (1992) 211–216. [10] Szeto, H. and Hinman, D., Prenatal development of sleep-wake patterns in sheep, Sleep, 8 (1985) 347–355. [11] Villiger, J.W., Taylor, K.M. and Gluckman, P.D., Multiple benzodiazepine receptors in the ovine brain: ontogenesis, properties, and distribution of 3H-diazepam binding, Paediatr. Pharmacol., 2 (1982) 179–187.