The authors have declared that no competing interests exist.
Conceived and designed the experiments: GWS IB MS RH T. Kerner UFM. Performed the experiments: T. Karen GWS IB MS RH CP. Analyzed the data: GWS T. Karen IB RH MS DE MK. Contributed reagents/materials/analysis tools: IB MS GWS RH. Wrote the paper: T. Karen GWS IB MS UF.
Propofol is commonly used as sedative in newborns and children. Recent experimental studies led to contradictory results, revealing neurodegenerative or neuroprotective properties of propofol on the developing brain. We investigated neurodevelopmental short- and long-term effects of neonatal propofol treatment.
6-day-old Wistar rats (P6), randomised in two groups, received repeated intraperitoneal injections (0, 90, 180 min) of 30 mg/kg propofol or normal saline and sacrificed 6, 12 and 24 hrs following the first injection. Cortical and thalamic areas were analysed by Western blot and quantitative real-time PCR (qRT-PCR) for expression of apoptotic and neurotrophin-dependent signalling pathways. Long-term effects were assessed by Open-field and Novel-Object-Recognition at P30 and P120.
Western blot analyses revealed a transient increase of activated caspase-3 in cortical, and a reduction of active mitogen-activated protein kinases (ERK1/2, AKT) in cortical and thalamic areas. qRT-PCR analyses showed a down-regulation of neurotrophic factors (BDNF, NGF, NT-3) in cortical and thalamic regions. Minor impairment in locomotive activity was observed in propofol treated adolescent animals at P30. Memory or anxiety were not impaired at any time point.
Exposing the neonatal rat brain to propofol induces acute neurotrophic imbalance and neuroapoptosis in a region- and time-specific manner and minor behavioural changes in adolescent animals.
Propofol (6,2 Diisopropylphenol) is widely used in paediatric anaesthesia. In 1999 the US Federal Drug administration decreased the approved age for maintenance of anaesthesia with propofol to 2 months, whereas in Germany the use of propofol 1% for induction and maintenance of anaesthesia is approved for children older than 1 month
Positive attributes of propofol are its pharmacokinetic properties that account for its clinical benefits such as rapid onset and of anaesthesia and short recovery time.
Experimental studies revealed several mechanisms of action, which strongly depend on age of animals but also on the dose administered. Propofol potentiates GABAA receptor functioning while at higher concentrations it causes opening of GABAA receptors
In adults propofol has been suggested as an ideal anaesthetic for neurosurgery because of its presumed beneficial effects on cerebral physiology (reduction in cerebral metabolic rate, reduction in cerebral blood flow, and brain relaxation). Experimental investigations revealed that propofol might also protect the brain against ischemic injury
However, in the immature brain general anaesthetics can induce apoptotic cell death in the central nervous system of experimental animals when administered during synaptogenesis that occurs during the first 2 weeks of life
Expression patterns of neurotransmitter receptors such as GABA and NMDA differ significantly between the adult and the newborn brain. This is probably a major reason for the peak vulnerability of the immature brain in the developmental period of rapid synaptogenesis, also known as the brain growth spurt period
To address this issue, we studied the temporal and regional activity of apoptotic and anti-apoptotic proteins and furthermore changes in the expression of neurotrophins following propofol anaesthesia on the developing brain. The second aim was to elucidate the long-term consequences on spontaneous behaviour, learning and memory abilities and its influence on anxiety-like behaviour. We hypothesise that administration of propofol to the immature rat brain causes neuroapoptosis, leading to a neurobehavioural phenotype/deficit in adolescence and adulthood.
All animal experiments were approved and performed in accordance with the guidelines of the Charité-Universitätsmedizin Berlin, Germany and the University hospital Essen, Germany. Animal care and handling were conducted in accordance with the European guidelines for use of experimental animals by certified FELASA fellows (Federation of European Laboratory Animal Science Associations) and with permission of local welfare committees. In the present study we employed a previously characterised animal model
For molecular studies animals were sacrificed 6, 12 and 24 hrs after administration of substances, by i.p. injection of 1.5 g/kg BW chloral hydrate, followed by transcardial perfusion with sterile phosphate buffered saline (PBS). Olfactory bulb and cerebellum were discarded from brain tissue, and cortex and thalamus were micro-dissected, snap frozen in liquid nitrogen, and stored at −80°C until further analysis.
Molecular analyses were focused on changes in apoptotic signalling pathways (caspase-3 activation and the caspase-independent apoptosis-inducing factor (AIF)), changes in neurotrophin expression patterns (BDNF, NT-3, NGF), and neurotrophin-dependent signalling pathways (serine-threonine kinase AKT, extracellular signal-regulated kinase ERK1/2, and their phosphorylated forms p-AKT and p-ERK1/2).
RNA was extracted from cortical and thalamic brain fractions, using Trizol (Invitrogen, Darmstadt, Germany) according to the manufacturer’s recommendations. 1 µg of total RNA was treated with DNase I (Invitrogen) in a total volume of 10 µl and 5 µl (500 ng) of this batch was reverse transcribed with 200 U SuperScript II (Invitrogen) using 500 ng oligo-dT and 250 ng random hexamers. cDNA was diluted and BDNF (GenBank accession no.: NM_012513; forward:
Snap-frozen tissue was homogenised in RIPA (radio-immuno-precipitation assay) buffer (1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EGTA, 1 mM Na3VO4, 20 mM NaF, 0.5 mM DTT, 1 mM PMSF and protease inhibitor cocktail in PBS pH 7.4). The homogenate was centrifuged at 1,050 g (4°C) for 10 min, and the microsomal fraction was subsequently centrifuged at 17,000 g (4°C) for 20 min. Twenty micrograms of the resulting cytosolic protein extracts were heat denaturated in Laemmli sample loading buffer, separated by electrophoresis in 10 or 15% SDS polyacrylamide gels and electro transferred onto a nitrocellulose membrane. Equal loading and transfer of proteins was confirmed by staining the membranes with Ponceau S solution (Fluka, Buchs, Switzerland). Nonspecific protein binding was prevented by treating the membrane with block solution (5% skim-milk, 0.5% BSA in TBST) 1 h at room temperature. The following primary antibodies (Cell Signaling, New England Biolabs GmbH, Frankfurt, Germany) were used for overnight incubation at 4°C: extracellular signal regulated kinase (ERK1/2, rabbit polyclonal p44/42 ERK 1/2, 1∶1000; mouse monoclonal phospho-p44/42 ERK1/2, 1∶500), protein kinase B (AKT, rabbit polyclonal AKT, 1∶1000; rabbit polyclonal phospho-AKT, 1∶1000), cleaved caspase-3 (rabbit polyclonal cleaved caspase-3, 1∶1000) and apoptosis-inducing factor (AIF, rabbit polyclonal AIF, 1∶1000). Primary antibodies were detected with appropriate horseradish-peroxidase labelled secondary antibody (swine anti-rabbit and rabbit anti-mouse 1∶2000-1∶10000, DAKO, Hamburg, Germany) and visualised using enhanced chemiluminescence (ECL, Amersham Biosciences, GE Healthcare, Buckinghamshire, UK). Serial exposures were made to radiographic film (Hyperfilm ECL; Amersham Biosciences, GE Healthcare). Densitometric analysis was performed with the image analysis program BioDocAnalyze (Whatman Biometra, Göttingen, Germany). For stripping, membranes were incubated with stripping buffer (100 mM β-mercaptoethanol, 2% SDS, 62.5 mM Tris-HCl; pH 6.7) at 50°C for 30 min, then washed, blocked and reprobed overnight at 4°C with mouse anti-β-actin monoclonal antibody (1∶1000, SIGMA-ALDRICH, Schnelldorf, Germany).
To assess the effect of propofol treatment on functional long-term outcome we performed behavioural testing on adolescent animals at postnatal day 30 and re-tested them at adult age on postnatal day 120. We focused on Open-Field test (OF)
Animals were placed into the centre of a dimly lit open-field arena (51×51×40 cm), fabricated of biologically inert, infrared (IR) translucent material, placed upon a IR light-box (TSE Systems) emitting IR light (λ∼850 nm). Movements were tracked by an automatic monitoring system (VideoMot2, TSE Systems) for 5 min, the movement threshold was set to 0.5 cm. Activity parameters the time in motion (locomotion), travelled distance (distance) and speed were analysed. Anxiety was calculated as the percentage of time the animal stayed in the border areas of the box in relation to the total time spent in maze. This procedure was repeated every 24 hrs on four consecutive days in order to capture time dependent changes and as training for the NOR.
Following the OF, we performed the NOR test, a two-trial, non-spatial, non-aversive memory test consisting of a “sample” phase and a “choice” phase, separated by a “test-free” interval during which the animals are returned to their home cages
Data analysis and representations were performed within the statistical environment R
Western blot analysis of brain lysates of cortex and thalamus at 6, 12 and 24 hrs after intraperitoneal applications of 3×30 mg/kg BW propofol at 0, 90 and 180 min in P6 rats, showed that propofol induces apoptotic neurodegeneration.
Propofol induced time dependent changes in the activation of caspase-3 in cortical (F(2,29) = 4.32, p = 0.023) and thalamic (F(2,28) = 6.61, p = 0.005) areas. The average difference of cleaved caspase-3 in cortical areas between treatment groups changed significantly from 6 to 24 hrs after the injection (Δlog2FC(6 h: 24 h) = 1.22, SE = 0.42, t(29) = 2.92, q = 0.02). In thalamic brain regions we observed a significant increase in cleaved caspase-3 12 hrs after the last injection (log2FC(12 h) = 0.83, SE = 0.26, t(28) = 2.92, q = 0.01). We further detected a significant change in the average difference between the two treatment groups from 6 to 12 hrs (Δlog2FC(6 h: 12 h) = 1.08, SE = 0.30, t(28) = 3.63, q = 0.003) and 12 to 24 hrs (Δlog2FC (12 h: 24 h) = −0.79, SE = 0.29, t(28) = −2.73, q = 0.02). There was no significant effect on activation of AIF (
Densitometric quantifications of caspase-3 and AIF in cortex and thalamus of P6 rats as analysed by Western blotting. Values represent mean normalised ratios of the densities of caspase-3 and AIF bands compared to densities of the control group (n = 5–6/point+SE). There was an effect of propofol treatment on caspase-3 levels over time, which was significant after 24 hrs in the cortex [F(1,29) = 3.63, p = 0.06] and after 12 hrs in the thalamus [F(1,28) = 3.1, p = 0.09).
To explore potential mechanisms involved in pathogenesis of apoptotic neurodegeneration in the developing brain following propofol exposure, we investigated whether expression patterns of brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3) and nerve growth factor (NGF) in cortex and thalamus of P6 rats (
Densitometric quantifications of mRNA levels of BDNF and NT-3 in cortex and thalamus of P6 rats, analysed by qRT-PCR. Values represent mean normalised ratios of the densities of BDNF and NT-3 bands compared to the density of the control group (n = 6–7/point+SE). There was an effect of propofol treatment with a decrease of BDNF levels over time, which was significant after 6 hrs in the cortex [F(1,30) = 66.5, p<0.001]. There was also a decrease in NT-3 levels, which was significant in the cortex after 6 hrs [F(1,28) = 12.7, p = 0.004] and after 12 hrs in the thalamus [F(1,24) = 3.5, p = 0.06].
Propofol triggered an overall reduction (log2FC(6 h–24 h) = −0.80, SE = 0.12, t(30) = −6.92, p = 1.11×10−7;
In thalamic areas BDNF mRNA expression of propofol treated animals changed significantly over time (F(2,24) = 5.21, p = 0.013;
Propofol treatment led to a significant, time dependent reduction of cortical NT-3 mRNA levels (F(2,28) = 8.45, 1.3×10−3;
Expression of NGF was transiently up-regulated in cortical areas at 6 hrs (log2FC(6 h) = 0.71, SE = 0.19, t(30) = 3.82, p = 1.68×10−3) after the last injection of propofol, followed by a significant down-regulation after 12 and 24 hrs. In thalamic areas we observed a general down regulation (F(1.26) = 16.93, p = 3×10−4) of NGF mRNA over all time points (log2FC(6 h–12 h) = −0.25, SE = 0.06, t(26) = −4.11, p = 3×10−4) (data not shown).
In the following step we investigated the influence of repeated propofol administration on the phosphorylated isoforms of the kinases AKT and ERK1/2. Western blot analysis of pAKT/AKT expression at 6, 12 and 24 hrs post injection, revealed a time dependent regulation of pAKT, in cortical (F(2,29) = 4.48, p = 0.02) and thalamic (F(2,28) = 4.49, p = 0.02) regions compared controls. The levels of pAKT were found to be significantly up-regulated in cortical areas after 12 hrs (log2FC(12 h) = 0.81, SE = 0.16, t(29) = 5.51, q = 5.01×10−5;
Densitometric quantifications of pAKT and pERK1/2 in the cortex and thalamus of P6 rats, analysed by Western blotting. Values represent mean normalised ratios of the densities of pAKT and pERK1/2 bands compared to the density of the control group (n = 6/point+SE). There was an effect of propofol treatment in decrease of pAKT levels over time in the thalamus, which was significant after 12 hrs [F(1,28) = 5.6, p = 0.06]. Post-hoc analysis showed most pronounced decrease after 12 hrs (2-sample t-test). In the cortex there was a significant decrease of pERK1/2 levels over the time, which was significant after 6, 12 and 24 hrs [F(1,29) = 12.7, p = 0.013].
To assess the impact of these time- and region-dependent changes in neurodegeneration and neurotrophin-dependent signalling we investigated long-term behavioural and neurocognitive outcome longitudinal behavioural outcome (OF and NOR) on P30 and P120.
Analysis of OF activity between P30 and P34 revealed an elevated pattern of general activity of propofol treated animals compared to controls (
Propofol treatment did not alter D) anxiety related behavior in adolescent animals [F(1,74) = 0.02, p = 0.89]. An overall change in activity was observed over individual measurements, resulting in a significant decrease in locomotion [F(3,71) = 13.6, p = 4.08×10−7] and distance [F(3,71) = 5.35, p = 2.23×10−3] and a significant increase in speed [F(3,74) = 15.7, p = 5.53×10−8] and the index of anxiety [F(3,74) = 7.25, p = 3×10−4]. (ncontrols = 12 animals, npropofol = 8 animals).
Assessment of OF activity parameters between P120 and P124 (
Apart from a transient effect on locomotion [F(3,71) = 5.92, p = 1.13×10−3], no significant changes over repeated measurements were observed in adult aged animals. (ncontrols = 12 animals, npropofol = 8 animals).
The assessment of cognitive performance in P30 animals (
Propofol (t(7) = −1.44, q = 0.192) as well as control animals (t(10) = −1.92, q = 0.168) failed to do so after a 24 hrs inter-trial interval. At P120 both groups spent a random amount of time with either of the objects after 6 hrs and also after a 24 hrs interval, indicating that they were unable to remember the old object. (ncontrols = 12 animals, npropofol = 8 animals).
Based on the previously well described propofol-induced apoptotic neurodegeneration when administered in the first week of rodent life
Our results indicate that apoptotic neurodegeneration was mainly mediated by caspase-3 activation, whereas there was no difference between experimental groups in expression of apoptosis-inducing factor (AIF). There were also spatial and time-dependent differences in caspase-3 expression with increased levels at 12 hrs in the thalamus, whereas the cortex was affected later (24 hrs), which is in accordance with previous observations of other groups
In this study we have chosen caspase-3 and AIF activation as representative markers for caspase-dependent and independent pathways to further characterise apoptosis in our animal model. These findings imply caspase involvement but do not exclude the possibility of many other death- and survival-promoting factors involved in the apoptotic machinery also contributing to propofol-induced acute injury in the immature brain.
Signalling pathways regulating cell death in development and after brain injury are not fully elucidated. Our findings indicate that propofol anaesthesia acutely depresses endogenous neurotrophins which have been shown to be crucial for neuronal development, synaptic plasticity and survival
Such changes may reflect impairment of survival promoting signals resulting in an imbalance between neuroprotective and neurodestructive mechanisms, which, during a developmental period of ongoing physiological elimination of brain cells, can promote apoptotic death.
Recent studies have emphasised, that anaesthetics caused disturbances in neurotrophin homeostasis in the developing brain. General anaesthesia with midazolam, isoflurane and nitrous oxide caused a decrease of BDNF in thalamus and an increase in cortex
However, propofol anaesthesia in our model resulted in a significant reduction of neurotrophin availability (BDNF, NT-3, and NGF) in cortex and thalamus and significantly decreased activation of ERK1/2 in cortical areas at all time points investigated (6, 12 and 24 hrs). Prevention of cell death by the pERK1/2 pathway has been previously shown in cultured rat hippocampal neurons
It remains unclear why the pERK1/2-pathway in our study was more affected than the pAKT-pathway. Based on presently available data, the variation in neurotrophin expression and function during development of each brain region is time-specific and may explain, at least in part, region-specific differences in an anaesthesia induced insult. Furthermore, our findings on reduced neurotrophin expression may explain anaesthesia-induced damage in the most vulnerable regions of the developing brain (thalamus and cortex)
In addition, results addressing neurotrophin expression under pathological conditions in the immature brain strongly depend on the type of injury and the experimental model used. In traumatic injury to the developing brain, neurotrophin up-regulation has been observed
In order to determine long-term consequences of propofol we investigated whether the previously well-documented propofol induced neuroapoptosis in combination with an acute impairment of neurotrophin dependent signalling is reflected by sustained functional cognitive and motor impairment. A cumulative dose of 90 mg/kg BW i.p. propofol increased locomotive activity in 30 day-old adolescent animals, expressed by an increase in the time spent in motion, which led to an increased travel distance. The speed was not significantly altered between treatment groups, indicating that the observed changes were due to an increased locomotion, rather than speed. We suggest that the observed increased activity was triggered by the novel environment, since behaviour of experimental animals and control animals did not differ in the following days. This hypothesis is further supported by our observations from P120 to P124 which also revealed no alteration in these parameters.
Both treatment groups at P30 and P120 showed a similar change in parameters observed over repeated measurements. We therefore cannot conclude a significant inability to habituate to the testing procedure. This finding stands in contrast to the work from Bercker et al.
Fredriksson et al.
Anaesthesia with propofol on P6 in our study did not result in memory deficits neither on P30 nor on P120, which is in accordance with previous results obtained by Bercker et al.
Similar findings have been described after neonatal exposure to isoflurane alone
The present work suggests that acute propofol-induced neurodegeneration combined with a transient disturbance in neurotrophin availability observed in the thalamus and cortex has no long-term effects on cognitive performance in this model.
Upon the translation of our experimental results into the human situation several questions remain open. The first critical issue concerns the extrapolation of appropriate developmental stages from different animal species to humans
Therefore analgesic or anesthetic treatments should be tailored to the invasiveness or presumed pain intensity of the procedure
The currently available experimental and clinical data addressing the toxic effects of anesthetics and sedatives and the impact of pain and stress on the developing brain are not sufficient and evidence-based enough to make any scientifically based recommendations for pediatric surgery or anesthesia
In conclusion, exposure of propofol to the neonatal rat brain induces acute neurotrophic imbalance and neuroapoptosis in a region- and time specific-manner. In the long-term it resulted in minor behavioural changes in adolescent but not in adult animals. As clinical studies are still lacking, future research has to focus on the investigation of safe anaesthetic strategies possibly in combination with neuroprotective agents. Until then, caution should be taken when using anaesthetic agents alone or in combination in preterm infants, newborns and young children.
The authors thank Mandana Rizazad and Karina Kempe for technical assistance.