- Email: [email protected]
Differential effects of neonatal hypoxic–ischemic brain injury on brainstem serotonergic raphe nuclei
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 –13 3
available at www.sciencedirect.com
www.elsevier.com/locate/brainres
Research Report
Differential effects of neonatal hypoxic–ischemic brain injury on brainstem serotonergic raphe nuclei Hanna E. Reinebrant a , Julie A. Wixey a , Glenda C. Gobe b , Paul B. Colditz a , Kathryn M. Buller a,⁎ a
Perinatal Research Centre, Clinical Neuroscience, University of Queensland Centre for Clinical Research, Brisbane, Queensland, Australia Molecular and Cellular Pathology, School of Medicine, University of Queensland, Brisbane, Queensland, Australia
b
A R T I C LE I N FO
AB S T R A C T
Article history:
Serotonergic fibres have a pervasive innervation of hypoxic–ischemic (HI)-affected areas in
Accepted 23 January 2010
the neonatal brain and serotonin (5-HT) is pivotal in numerous neurobehaviours that match
Available online 1 February 2010
many HI-induced deficits. However, little is known about how neonatal HI affects the serotonergic system. We therefore examined whether neonatal HI can alter numbers of
Keywords:
serotonergic raphe neurons in specific sub-divisions of the midbrain and brainstem since
Brainstem
these nuclei are the primary sources of serotonin throughout the central nervous system
Hypoxia–ischemia
(CNS). We utilised an established neonatal HI model in the postnatal day 3 (P3) rat pup (right
Neonate
common carotid artery ligation + 30 min 6% O2) and determined the effects of P3 HI on 5-HT
Raphe nuclei
counts in 5 raphe sub-divisions in the midbrain and brainstem one and six weeks later. After
Serotonin
P3 HI, numbers of 5-HT-positive neurons were significantly decreased in the dorsal raphe
Serotonin transporter
dorsal, dorsal raphe ventrolateral and dorsal raphe caudal nuclei on P10 but only in the dorsal raphe dorsal and dorsal raphe ventrolateral nuclei on P45. In contrast, P3 HI did not alter counts in the dorsal raphe interfascicular and raphe magnus nuclei. We also discovered that P3 HI significantly reduces brainstem SERT protein expression; the key regulator of 5-HT in the CNS. In conclusion, neonatal HI injury caused significant disruption of the brainstem serotonergic system that can persist for up to six weeks after the insult. The different vulnerabilities of serotonergic populations in specific raphe nuclei suggest that certain raphe nuclei may underpin neurological deficits in HI-affected neonates through to adulthood. © 2010 Elsevier B.V. All rights reserved.
1.
Introduction
Hypoxic–ischemic brain injury is one of the major factors contributing to neonatal morbidity and mortality in preterm
neonates. Neurological deficits after preterm neonatal HI include cerebral palsy, mental retardation, learning disabilities and epilepsy (Graham et al., 2008; Volpe, 2001, 2009). In the premature infant white matter damage is a characteristic
⁎ Corresponding author. Perinatal Research Centre, UQ Centre for Clinical Research, Building #71/918, Royal Brisbane & Women's Hospital, Herston, QLD 4029, Australia. Fax: +61 7 33465594. E-mail address: [email protected] (K.M. Buller). Abbreviations: 5-HIAA, 5-hydroxyindoleacetic acid; 5-HT, serotonin; Aq, aqueduct (Sylvius); CNS, central nervous system; DRc, dorsal raphe caudal nucleus; DRd, dorsal raphe dorsal nucleus; DRif, dorsal raphe interfascicular nucleus; DRvl, dorsal raphe ventrolateral nucleus; HI, hypoxia–ischemia; P, postnatal day; RM, raphe magnus nucleus; SERT, serotonin transporter; SIDS, sudden infant death syndrome; xscp, decussation of the superior cerebellar peduncle 0006-8993/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2010.01.065
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 – 1 33
feature of HI injury, but extensive neuronal damage also occurs (Arvin et al., 2002; Fan et al., 2006; Gluckman et al., 2001; Hawker and Lang, 1990; Johnston et al., 2001; Sizonenko et al., 2003). However little is known about the precise neuronal phenotypes altered after neonatal HI. Serotonin (5-HT) is a major neurotransmitter and serotonergic fibres have a pervasive network throughout the central nervous system (CNS). Many forebrain regions innervated by serotonergic fibres are damaged after HI such as the cortex, striatum, thalamus and amygdala (Stadlin et al., 2003; Towfighi et al., 1997). Furthermore 5-HT is a neurochemical that is involved in a vast array of functions and alterations to the 5-HT system can cause long-term deficits; many of which match those manifested after neonatal HI (Dayan and Huys, 2009). It is therefore important to establish how neonatal HI might alter serotonergic neurons in the brain and begin to decipher whether these specific nuclei constitute primary candidate networks that underpin neonatal HI-induced neurological deficits. Serotonergic neurons are localised primarily in the raphe nuclei of the midbrain and brainstem. The major raphe complex of 5-HT-positive neurons is the dorsal raphe and this complex can be further divided into several sub-divisions located in the midbrain. The dorsal raphe nuclei project to and innervate many target areas in the forebrain via the dorsal, medial and ventral ascending pathways, including the thalamus, hypothalamus, several cortical areas, septum, amygdala nuclei and hippocampus (Michelsen et al., 2008). The raphe magnus nucleus, located more caudally in the brainstem, predominantly projects to the spinal cord (Tanaka et al., 2006). Although damage to the brainstem after HI is evident in human neonates (Leech and Brumback, 1988; Peters et al., 2000) and animal models (Buller et al., 2008; Peng et al., 2005; Tomimatsu et al., 2002), information about the specific effects of the raphe nuclei are scarce. One study has shown that total asphyxia for 30 min in P2 mice results in a significant loss of 5HT-positive neurons in specific raphe nuclei in the brainstem up to 60 days after injury. Asphyxia induced permanent changes in the rostral raphe system, while the caudal raphe system appeared unaffected in terms of 5-HT-positive numbers (Takeuchi et al., 1992). Whether these changes are evident after neonatal HI has not been investigated. Also 5-HT levels in forebrain regions are reduced after P7 hypoxia in rats (Hadjiconstantinou et al., 1990) but the effects of HI on 5-HTpositive neurons residing in the raphe nuclei have not been examined. It is also not known if neonatal HI can alter the expression of the serotonin transporter (SERT); the key regulator 5-HT levels in the CNS. Thus, despite the knowledge that serotonergic fibres innervate areas affected by neonatal HI and the importance of the serotonergic system throughout the brain, it remains to be determined if 5-HT raphe nuclei are damaged after neonatal HI, whether specific sub-divisions are more vulnerable to HI injury than others and if SERT is altered in the brainstem after neonatal HI. In the present study we hypothesised that HI in the immature brain affects numbers of brainstem serotonergic neurons in the raphe nuclei one week (P10) and six weeks (P45) after P3 HI. We used a well established rat model of preterm HI, the P3 rat pup (Buller et al., 2008; Carty et al., 2008; Tai et al., 2009). Using immunohistochemistry we phenotypically identified 5-HT-positive neurons in five specific raphe nuclei and
125
determined if neonatal HI changes numbers of 5-HT-positive neurons in the dorsal raphe nucleus (dorsal, ventrolateral, interfascicular and caudal sub-divisions) and medullary raphe nuclei (raphe magnus) on P10 and P45. We also examined whether P3 HI alters SERT protein expression in the brainstem on P10 and P45.
2.
Results
2.1. Numbers of 5-HT-positive neurons in the raphe nuclei of control animals on P10 and P45 In control animals developmental changes in the number of 5HT-positive neurons in certain raphe nuclei were apparent over the five-week period from P10 to P45. We found a higher number of 5-HT-positive neurons in the dorsal raphe interfascicular and caudal nuclei and raphe magnus nucleus on P10 compared to P45. For the interfascicular, caudal and raphe magnus nuclei there was a significant 30.9%, 39.3% and 32.5% decrease in numbers of 5-HT-positive neurons from P10 to P45, respectively (Fig. 1). In contrast, there was no difference in the number of 5-HT-positive neurons in the dorsal raphe ventrolateral and dorsal raphe dorsal nuclei on P10 compared to P45 (Fig. 1).
2.2.
Effect of P3 HI on brain hemisphere size
We determined cerebral hemisphere size in the brainstem and forebrain in control and P3 HI animals as a measure of injury. In the brainstem, there was no difference in hemisphere size ipsilateral to the carotid ligation in comparison to the contralateral non-ligated side in animals subjected to P3 HI
Fig. 1 – Total numbers of 5-HT-positive neurons in the dorsal raphe dorsal (DRd), dorsal raphe ventrolateral (DRvl), dorsal raphe interfascicular (DRif), dorsal raphe caudal (DRc) and raphe magnus (RM) nuclei of control animals on P10 and P45. See Fig. 2 for localisation and immunolabelling of the raphe nuclei sub-division. The total number of 5-HT-positive neurons in the DRif, DRc and RM nuclei significantly decreased on P45 compared to P10 by 31%, 39% and 32%, respectively. In contrast, there was no difference in total counts on P10 and P45 in the DRd and DRvl nuclei. *p < 0.05, **p < 0.01.
126
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 –13 3
Fig. 2 – Schematic diagrams modified from the rat atlas (Paxinos and Watson, 1997) indicating the location of photomicrographs of coronal sections immunolabelled for 5-HT in selected raphe nuclei of control and P3 HI animals on P10. Photomicrographs have been taken through the dorsal raphe dorsal (A–B), the dorsal raphe ventrolateral (C–D), the dorsal raphe interfascicular (E–F), the dorsal raphe caudal (G–H) and the raphe magnus (I–J) nuclei ipsilateral to the ligated side in control animals (A, C, E, G, I) and P3 HI animals (B, D, F, H, J). Scale bars represent 25 µm. Rostro-caudal distance (mm) given relative to bregma (Paxinos and Watson, 1997). Aq, aqueduct (Sylvius); DRd, dorsal raphe dorsal nucleus; DRc, dorsal raphe caudal nucleus; DRif, dorsal raphe interfascicular nucleus; DRvl, dorsal raphe ventrolateral nucleus; RM, raphe magnus nucleus; xscp, decussation of the superior cerebellar peduncle.
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 – 1 33
on P10 (− 0.13% ± 0.85%) or P45 (− 0.38% ± 1.01%). Further, compared to the left side in control animals there was no difference in brainstem area on the non-ligated side of P3 HI animals. However, in the forebrain after P3 HI, there was a significant reduction in cerebral hemisphere size on the ligated side in comparison to the non-ligated side on both P10 (18.4% ± 3.43%) and P45 (11.7% ± 1.22%). These results are consistent with our previous studies using the same P3 HI rodent model where brainstem area was unchanged after P3 HI (Buller et al., 2008) and forebrain hemisphere size was decreased after neonatal hypoxic–ischemic injury (Buller et al., 2008; Wixey et al., 2009). Thus we were confident we reproduced a certain severity of brain injury after P3 HI.
127
P3 HI, there was a significant 18.6% reduction (Fig. 3). There was no difference between 5-HT-positive neuronal counts on the P3 HI non-ligated side and the corresponding side in control animals on P10 and P45.
2.3.3.
Dorsal raphe interfascicular nucleus
2.3.
Effects of P3 HI on 5-HT-positive dorsal raphe nuclei
On P10, one week after P3 HI, there was no change in 5-HTpositive neurons ipsilateral to the ligation compared to the non-ligated side in the dorsal raphe interfascicular nucleus (Figs. 2 and 3). Similarly on P45, six weeks after P3 HI, there was no difference in the number of 5-HT-positive neurons on the ipsilateral side compared to the non-ligated side (Fig. 3). Also no difference in 5-HT counts was found between the nonligated side and the corresponding side in control animals on P10 and P45.
2.3.1.
Dorsal raphe dorsal nucleus
2.3.4.
On P10, one week after P3 HI, there was a significant 21.4% reduction in the number of 5-HT-positive neurons ipsilateral to the ligation compared to the non-ligated side in the dorsal raphe dorsal nucleus (Figs. 2 and 3). In addition, on P45, six weeks after P3 HI, there was a significant 14.1% reduction after P3 HI on the ipsilateral side compared to the non-ligated side (Fig. 3). There was no difference in numbers of 5-HT-positive neurons between the non-ligated side and the corresponding side in control animals.
2.3.2.
Dorsal raphe caudal nucleus
On P10, one week after P3 HI there was a significant 10.7% reduction in the number of 5-HT-positive neurons ipsilateral to the ligation compared to the non-ligated side in the dorsal raphe caudal nucleus (Figs. 2 and 3). In contrast on P45 in the dorsal raphe caudal nucleus, six weeks after P3 HI, there was no difference in the number of 5-HT-positive neurons ipsilateral to the ligation compared to the non-ligated side (Fig. 3). There was no difference in 5-HT-positive neuronal counts between the non-ligated side and the corresponding side in control animals.
Dorsal raphe ventrolateral nucleus
On P10, one week after P3 HI there was a significant 20.2% reduction of 5-HT-positive neurons ipsilateral to the ligation compared to the non-ligated side in the dorsal raphe ventrolateral nucleus (Figs. 2 and 3). On P45, six weeks after
2.3.5.
Raphe magnus nucleus
On P10 and P45 there was no difference in the number of 5-HTpositive neurons ipsilateral to the ligation compared to the non-ligated side in the raphe magnus nucleus (Fig. 3). In
Fig. 3 – The effects of P3 HI on numbers of 5-HT-positive neurons in specific raphe nuclei on P10 (A) and P45 (B). In P3 HI animals on P10 (A), there was a significant loss of 5-HT-positive neurons in the dorsal raphe dorsal, the dorsal raphe ventrolateral and the dorsal raphe caudal nuclei ipsilateral to the carotid ligation. Counts in the dorsal raphe interfascicular and the raphe magnus nuclei remained unchanged relative to the non-ligated side. In control animals there was no difference in 5-HT-positive neuron counts on the left side compared to the right side (A). On P45, in P3 HI animals, there was a significant loss of 5-HT-positive neurons on the ligated side compared to the non-ligated side in the dorsal raphe dorsal and the dorsal raphe ventrolateral nuclei. The dorsal raphe interfascicular, the dorsal raphe caudal and the raphe magnus remained unchanged on the ligated side compared to the non-ligated side. In control animals there was no difference in 5-HT counts between the left side and right side (B). DRd, dorsal raphe dorsal nucleus; DRc, dorsal raphe caudal nucleus; DRif, dorsal raphe interfascicular nucleus; DRvl, dorsal raphe ventrolateral nucleus; RM, raphe magnus nucleus *p < 0.05, **p < 0.01, ***p < 0.005.
128
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 –13 3
addition, there was no difference in numbers of 5-HT-positive neurons between the non-ligated side and the corresponding side in control animals.
2.4.
Effects of P3 HI on SERT protein expression
Western blotting was used to detect possible changes in SERT protein expression in the brainstem after P3 HI. On P10, one week after P3 HI, there was a significant 23.4% reduction in SERT protein expression compared to control animals (Figs. 4A–B). In contrast, on P45 there was no difference in SERT protein expression between P3 HI and control animals (Figs. 4A–B). Representative sections were also labelled for SERT and we found that SERT immunoreactivity mirrored the changes in 5-HT-positive neurons in the dorsal raphe
(Figs. 4C–D). To further confirm that SERT is concentrated on 5-HT-positive raphe neurons in the brainstem (Fujita et al., 1993; Qian et al., 1995) we performed dual immunofluorescent labelling for 5-HT and SERT and demonstrated a strong colocalization of SERT on 5-HT-positive neurons in the dorsal raphe (Figs. 4E–G).
3.
Discussion
This is the first study to demonstrate that neonatal HI alters the number of 5-HT-positive raphe neurons and that P3 HI also decreases SERT protein expression in the brainstem. We found that certain sub-populations of 5-HT-positive raphe neurons are more vulnerable to injury after P3 HI than others.
Fig. 4 – The effect of P3 HI on SERT protein expression ipsilateral to the carotid ligation on P10 and P45 in the brainstem. Example Western blots to detect SERT are shown (A). The amount of SERT protein detected is expressed as a percentage change from control levels. Beta-actin was used as the loading control (A). After P3 HI, SERT protein expression was significantly decreased on P10 (B). However, on P45, there was no difference in SERT expression between control and P3 HI animals (B). Immunoreactivity for SERT in the dorsal raphe of control (C) and P3 HI animals (D) demonstrated that numbers of SERT-positive nuclei mirrored the changes in immunolabelling for 5-HT-positive neurons in the dorsal raphe. To confirm the localization of SERT on brainstem raphe neurons, examples of dorsal raphe sections dual-labelled for 5-HT (E, green) and SERT (F, red) are shown. Colocalization (G, yellow) of the SERT-positive marker was virtually found exclusively on 5-HT-positive neurons in the dorsal raphe nucleus. *p < 0.05; scale bars represent 25 µm.
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 – 1 33
In addition, we observed developmental changes in the number of 5-HT-positive raphe neurons in normal control animals from P10 to P45.
3.1. nuclei
Developmental change in total 5-HT-positive raphe
Interestingly, in normal unoperated rats, ontogenic differences in numbers of 5-HT-positive neurons were observed over the five-week period examined (from P10 to P45). These changes only occurred in certain raphe nuclei sub-divisions whereby a lower number of 5-HT-positive neurons were evident on P45 compared to P10 in the dorsal raphe caudal, dorsal raphe interfascicular and raphe magnus nuclei. Counts in the dorsal raphe ventrolateral and dorsal raphe dorsal nuclei remained unchanged over this five-week period. Thus total numbers of 5-HT neurons did not increase over time in any of the nuclei raphe examined. To the best of our knowledge only one quantitative study of 5-HT-positive neurons in the raphe magnus nucleus has been performed (Tanaka et al., 2006). Over a similar developmental period to the present study, Tanaka et al. (2006) reported a significant decrease in the number of 5-HT-positive raphe magnus neurons (Tanaka et al., 2006). The decreased number of 5HT-positive neurons in specific raphe nuclei could be due to normal neuronal pruning during development. Tanaka et al. (2006) also demonstrated that, despite the decrease in number of 5-HT neurons in the raphe magnus nucleus, a larger number of neurons were back-labelled after tracer injection to the spinal cord, suggesting axonal growth and dendritic arborisation of descending raphe neurons (Tanaka et al., 2006). Previous studies have shown increased serotonergic innervation density, increased arborization and axonal length and higher expression of the SERT in the forebrain over time, indicating the developmental plasticity and axonal sprouting of the serotonergic system (Lidov and Molliver, 1982; Tarazi et al., 1998). It has also been demonstrated that SERT, mainly localised on 5-HT-positive neurons, increases expression in the early postnatal days, reaches its highest expression in the dorsal raphe nucleus on P14, after which the SERT expression decreases until P21 then remains steady at ‘adult’ levels (Galineau et al., 2004). These results support the developmental changes observed in the dorsal raphe in the present study. Taken together, we speculate that normal pruning of serotonergic nuclei and developmental growth has occurred in the raphe sub-divisions during the five-week period.
3.2.
Brain hemisphere size after P3 HI
We utilised an established rat model of preterm HI (P3 HI) that we, and others, have shown previously to cause white matter damage, loss of neurons and behavioural deficits (Buller et al., 2008; Wixey et al., 2009; Stadlin et al., 2003). As an index of brain injury in the present study we determined forebrain and brainstem hemisphere size ipsilateral to the carotid ligation relative to the non-ligated side. Consistent with our previous studies forebrain cerebral hemisphere size was significantly reduced after P3 HI. In addition, since the primary region of interest was the midbrain and brainstem where the majority of raphe neurons reside, we measured the brainstem hemi-
129
sphere size at the level of the dorsal raphe nuclei. Unlike the forebrain outcome, there was no effect of P3 HI on brainstem hemisphere size. Using the same P3 HI rodent model we have demonstrated previously, at a more caudal level of the medulla oblongata, hemisphere area is also unchanged (Buller et al., 2008). Thus not only were we confident that we reproduced a certain severity of brain injury after P3 HI consistent with our previous studies but we reinforced the phenomenon that P3 HI does not incur gross changes in brainstem hemisphere size.
3.3.
Effects of P3 HI on specific raphe nuclei
There was an overall significant loss of 5-HT-positive raphe neurons after P3 HI, consistent with previous findings (Takeuchi et al., 1992; Shirashi et al., 2008) however the neuronal loss was greater in specific raphe sub-populations than others. Thus certain serotonergic raphe nuclei may be more vulnerable to HI-induced injury than others. The neuronal loss was also more prominent on P10 than on P45. On P10 after P3 HI, 5-HT-positive losses were observed in the dorsal raphe caudal, dorsal raphe ventrolateral and dorsal raphe dorsal nuclei. The dorsal raphe interfascicular and the raphe magnus nuclei, showed no reduction in number of 5HT-positive neurons on P10. In contrast, on P45, only the dorsal raphe ventrolateral and the dorsal raphe dorsal elicited prolonged damage and maintained a significant loss of 5-HTpositive neurons. The topographical distribution of raphe nuclei may dictate their vulnerability to HI injury. Mostly the more anterior raphe sub-divisions were affected by P3 HI as opposed to the more posterior and caudally located nuclei such as the dorsal raphe interfascicular and the raphe magnus nuclei. The positional clustering of different raphe sub-divisions in the midbrain and brainstem also represents differential connectivity patterns in the brain. The dorsal raphe caudal, dorsal raphe ventrolateral and dorsal raphe dorsal nuclei predominantly send axonal projections to the cerebral cortex, basal ganglia, thalamus, hypothalamus, hippocampus and amygdala and are involved in influencing or regulating numerous functions. Each serotonergic sub-division is integral in producing characteristic functional effects based largely on their afferent and efferent connections in the brain. Specific losses of these 5-HT raphe nuclei in these sub-divisions can therefore affect numerous functions and might underpin regulatory mechanisms that are responsible for anxiety, movement, depression, cardiovascular, respiratory deficits reported in HI-affected preterm infants. Serotonin levels are reduced in the brain after neonatal P7 hypoxia (Hadjiconstantinou et al, 1990) but it is not known if there is regional variation in the effect of neonatal HI on serotonin release. Nonetheless the loss of dorsal raphe neurons after HI may explain reduced serotonergic release in the forebrain. In direct contrast to the dorsal raphe nuclei, on both P10 and P45, the number of 5-HTpositive neurons in the raphe magnus nucleus was unchanged after P3 HI. Thus we propose that these neurons are not susceptible to HI injury and remain functionally viable after P3 HI. Retrograde axonal transport studies have shown that the medullary raphe system, including the raphe magnus, provides the major descending serotonergic input to the spinal
130
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 –13 3
cord but that serotonergic nuclei of the raphe magnus send very few ascending projections to the forebrain (Bowker et al., 1981; Loewy and McKellar, 1981). Therefore it is likely, at least in our model, that the serotonergic innervation of the spinal cord remains relatively intact after P3 HI. This suggestion is reinforced by evidence that the spinal cord appears to endure damage only after severe neonatal HI insults (Clancy et al., 1989; Groenendaal et al., 2008). Thus determining the loss of specific raphe nuclei and their ascending and descending neural connections may predict their vulnerability to P3 HI injury as well as their wider influence on other brain regions after P3 HI. It remains to be investigated whether specific serotonergic changes in the dorsal raphe nuclei lead to selective serotonergic HI deficits in the forebrain and associated changes in serotonin-mediated functions.
3.4.
Reduced SERT protein expression after P3 HI
Since distinct losses of 5-HT-positive neurons were detected after P3 HI we wished to establish whether another serotonergic neuronal marker, SERT, was also affected in the brainstem. SERT is a pivotal regulator of serotonin levels in the CNS and thus its expression is an indicator of serotonergic synaptic function. Importantly, we found that SERT protein expression in the brainstem was significantly decreased on P10 after P3 HI. The loss of SERT protein on P10 could be explained by a loss of 5-HT-positive neurons in specific raphe nuclei since we confirmed the localization of SERT on 5-HTpositive neurons (Fujita et al., 1993; Qian et al., 1995). In contrast, on P45 the SERT protein levels were not significantly lower than control levels. The overall recovery of SERT protein expression did not reflect the overall maintained loss of raphe neurons after P3 HI. It is possible that remaining neurons increased their expression of SERT protein to compensate for neuronal losses over time. It has been shown that SERT function is maintained with as little as 20% of its original population intact (Montanez et al., 2003). It is now important for future studies to investigate the functional relevance of the P3 HI-induced losses of 5-HT neurons and SERT expression not only at the neuronal level but also in terms of behavioural deficits.
3.5. Possible mechanisms responsible for brainstem serotonergic injury The observed decreases in numbers of 5-HT-positive dorsal raphe neurons and the associated reduction in SERT protein expression in the brainstem affirm that P3 HI resulted in losses of serotonergic neurons. In P7 rat pups with hypoxiainduced brain injury, both 5-HT and its metabolite, 5hydroxyindoeacetic acid (5-HIAA) are decreased in the forebrain, suggesting neuronal loss occurs rather than metabolic changes (Hadjiconstantinou et al., 1990). Moreover, losses cannot simply be explained by developmental decreases in numbers of 5-HT-positive neurons since effects of P3 HI were calculated as a percentage of the non-ligated side in each animal for the specific postnatal day. Thus each animal essentially served as its own control. The mechanisms underpinning the observed losses of 5HT-positive neurons are not known. The brainstem is located
outside the vascular field of the ipsilateral hemisphere and should not be directly affected by immediate changes in perfusion after common carotid artery occlusion. In fact an increase in blood flow occurs in the brainstem after such an insult (Vannucci et al., 1988). Further we did not detect any change in brainstem hemisphere area after P3 HI, in contrast to significant reductions in forebrain cerebral hemisphere area. The losses of 5-HT-positive neurons in the brainstem after P3 HI are therefore likely to occur via secondary mechanisms. Candidate mechanisms responsible for this secondary injury after P3 HI include excitotoxic mechanisms, neuroinflammation and axonal degeneration. Serotonergic afferent and efferent projections connect major areas of the forebrain, largely matching or at least overlapping with areas known to be damaged after HI, such as the striatum, thalamus and amygdala (Michelsen et al., 2008; Stadlin et al., 2003; Towfighi et al., 1997). Thus it is plausible that damage to these forebrain areas is likely to affect serotonergic fibres, innervating or coursing through these areas as fibres of passage, and cause retrograde cell death in remote areas such as the raphe serotonergic nuclei. It follows, that the regional vulnerability of 5-HT-positive neurons in the dorsal raphe nucleus after P3 HI in this study could be explained by the serotonergic innervation pattern of the forebrain. Possibly the innervation is more dense from certain sub-divisions in the dorsal raphe nucleus to the HI-affected areas in the forebrain, leading to more pronounced damage in these areas. As alluded to earlier, whether there is corresponding serotonergic neuropathology in the forebrain after neonatal HI remains to be determined. The early neuronal injury after HI is believed to be primarily via necrosis resulting from excitotoxic damage; most likely as a result of excessive release of glutamate from presynaptic nerve terminals and astrocytes into the extracellular space, which causes calcium overload followed by necrosis (Choi, 2001). The loss of 5-HT-labelling in the brainstem raphe neurons could be due to apoptosis or necrosis of the neurons. Increased apoptosis has previously been shown in remote areas after P7 HI (Tomimatsu et al., 2002). Secondary neuroinflammation events, such as increased numbers of activated microglia after HI secrete a variety of chemokines and pro-inflammatory cytokines including tumour necrosis factor alpha and interleukin-1 beta, which further exacerbate neuronal injury (Leonardo and Pennypacker, 2009). Neuroinflammation could contribute to brainstem injury via remote actions at the primary injury sites, where inflammatory mediators may damage afferent and efferent fibres of dorsal raphe nuclei and subsequently decrease the survival of brainstem nuclei. Alternatively, neuroinflammation may occur in the brainstem itself and therefore more directly affect serotonergic raphe neuronal function.
3.6.
Conclusions
In the present study we have demonstrated that the serotonergic system in the brainstem is disrupted significantly in a rat model of preterm HI brain injury. Specific losses of 5-HTpositive neurons in select raphe nuclei were observed and an associated decrease in brainstem SERT protein expression. These changes were apparent both one and six weeks after injury, indicating both short-term and long-lasting effects
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 – 1 33
after P3 HI. Disruptions of the serotonergic system are believed to contribute to a multitude of disorders, including SIDS, autism, depression and obsessive–compulsive disorder (Nemeroff and Owens, 2009; Pardo and Eberhart, 2007; Paterson et al., 2006; Stengler-Wenzke et al., 2004). We propose that the HI-induced disruption of the serotonergic system in the brainstem may contribute to functional deficits observed in HI-affected children and adults. For example, decreased serotonergic function is a hallmark feature of depression and depressed patients show 31% loss of dorsal raphe neurons (Baumann et al. 2002). Cerebral palsy is a notable deficit in HIaffected neonates and cerebral palsy patients have been reported to suffer depression (Rone and Ferrando, 1996; Krakovsky et al., 2007). Whether reduced serotonergic function accounts for HI-induced neurological deficits is not known. A plausible avenue of investigation would be to firstly characterise the effects of neonatal HI on serotonergic fibres that innervate specific brain regions.
4.
Experimental procedures
4.1.
Subjects
In this study Sprague–Dawley rat dams and their pups (10–12 per litter) were used. The animals were housed at a constant temperature of 22 °C with a 12 h light cycle and provided standard rat chow and water ad libitum. Experiments were carried out between 8.00 am and 11.00 am in order to minimize possible fluctuations in circadian rhythms. All procedures were approved by the University of Queensland Animal Ethics Committee. Efforts were made to minimize the number of animals used and their suffering.
4.2.
tone (Lethabarb, Virbac, France; 1.5 g/kg i.p) and perfused intracardially with 1% sodium nitrite solution followed by 4% formaldehyde (in 0.1 M PBS, pH 7.4). The brain was excised and post-fixed for 2 h in formaldehyde, and the brainstem was separated from the forebrain at level − 6.7 to −12 mm relative to bregma and then cryo-protected in 10% sucrose (in 0.1 M PBS, pH 7.4, 4 °C) overnight. Serial sections (50 µm for the brainstem and 40 µm for the forebrain) were collected using a sliding microtome. For immunolabelling, brainstem sections from P10 (1-in-4 series, 200 µm intervals) and P45 (1-in-5 series, 250 µm intervals) were incubated with rabbit anti-5-HT antibody (gift from Professor D. Pow) or goat anti-SERT (serotonin transporter, 1:2500; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) in 1% bovine serum albumin at 4 °C for 36 h followed by biotinylated secondary antibody (donkey anti-rabbit or donkey anti-sheep, 1:400; Jackson ImmunoResearch, West Grove, PA, USA). For forebrain hemisphere size, forebrain sections from P10 and P45 (1-in-4 series, 160 µm intervals) were incubated with mouse anti-NeuN antibody (neuronal marker, 1:10,000; Chemicon, Millipore, Autralia) followed by donkey anti-mouse (1:400; Santa Cruz Biotechnology). Sections were then immersed in avidin– biotin–horseradish peroxidase complex (Vector Elite Kit, Burlingame, CA, USA) and the activity was visualized using diaminobenzidine solution. Sections from control and P3 HI animals were processed simultaneously to minimize variations in immunohistochemistry. Sections were then mounted on chrome-alum subbed slides, dehydrated in a series of alcohol, cleared in xylene and coverslipped. For double immunofluorescent labelling, after incubation with primary antibodies (rabbit anti-5-HT, 1:200 and goat anti-SERT, 1:200) brainstem sections were incubated with appropriate Alexa Fluor labels (568 for SERT, 488 for 5-HT, both at 1:200; Invitrogen, Carlsbad, CA), mounted and coverslipped.
Hypoxic–ischemic brain insult 4.4.
On postnatal day 3 (P3) rat pups underwent surgery to induce HI injury. Pups were anaesthetized with Isoflurane (Baxter, 2% in inhalation air, air-flow 3.5 L/min) and the right common carotid artery was exposed, carefully dissected from surrounding tissue and permanently ligated with 5.0 silk thread. Pups were allowed to recover for 15 min on a heating pad and were then placed in a humidified hypoxia chamber for 30 min with constant infusion of 6% O2 and temperature kept at 37 °C. Following hypoxia, pups were again recovered for 15 min on a heating pad and then returned to the dam. Control pups were treated the same way pup but did not undergo surgery and breathed room air. The mortality rate for the HI experiments was <3.5%. The P3 HI rat model is an established preterm animal model that shows white matter damage and neuronal losses (Buller et al., 2008; Carty et al., 2008; Sizonenko et al., 2003; Stadlin et al., 2003) and is analogous to preterm human brain at approximately 24 to 28 weeks gestation in terms of number of synapses, cortical organization and neurochemical development (Clancy et al., 2001; Hagberg et al., 2002).
4.3.
131
Immunohistochemistry
On P10 (control n = 7; P3 HI n = 7) and P45 (control n = 6; P3 HI n = 6), animals were anaesthetized with sodium pentobarbi-
Western blot analyses
On P10 (control n = 9; P3 HI n = 9) and P45 (control n = 10; P3 HI n = 10), animals were anaesthetized with sodium pentobarbitone (Lethabarb, Virbac, France; 1.5 g/kg i.p). The brain was rapidly removed and the brainstem was isolated at level −6.7 to −12 mm relative to bregma and snap-frozen on dry ice. The tissue was then gently homogenised in ice-cold dH2O and centrifuged for 5 min at 4 °C to remove cell debris. The supernatant was recovered and frozen at −80 °C. Total protein was determined using a commercial BCA kit (Pierce, Rockford, IL, USA). Samples (20 µg) from the tissue preparations were mixed with sample buffer at room temperature and separated by 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis using non-denaturing conditions. After separation, the proteins were transferred to a polyvinylidene difluoride membrane (Bio-Rad, Hercules, CA, USA), blocked with 5% non fat dry milk in TBS with 0.1% Tween-20 for 1 h. The membrane was then incubated with SERT antibody (goat anti-SERT, 1:2000; Santa Cruz Biotechnology, Inc.) in 2% non fat dry milk in TBS with 0.1% Tween-20 overnight at 4 °C. The membrane was then incubated with HRP-conjugated antigoat IgG (1:10,000; Sigma, Castle Hill, NSW, Australia) for 2 h and visualized using enhanced chemiluminescence (ECL Plus, Amersham, Sydney, NSW, Australia) on X-ray film. Following
132
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 –13 3
visualization, each blot was stripped using Re-blot Plus (Millipore, Australia) and reprobed with beta-actin antibody (mouse anti-beta-actin, 1:20,000; Imgenex, San Diego, CA, USA) in order to quantify SERT levels relative to beta-actin protein. Chemiluminescence was eliminated for SERT when the antibody was pre-incubated with the control peptide (1:2000; Santa Cruz, CA, USA).
4.5.
Data analysis
To determine the effects of P3 HI, 5-HT neurons were counted in four sub-divisions of the dorsal raphe and the raphe magnus nucleus on P10 and P45. Only 5-HT-positive neurons with a clearly defined nucleus were counted. Raphe nuclei were identified and defined according to the following coordinates in rostro-caudal order: dorsal raphe dorsal −7.04 to −7.64 mm relative to bregma over 3 sections, dorsal raphe ventrolateral −7.64 to −8.3 mm relative to bregma over 3 sections, dorsal raphe interfascicular − 8.0 to − 9.16 mm relative to bregma over 4 sections, dorsal raphe caudal −8.72 to −9.3 mm relative to bregma over 4 sections and raphe magnus −9.16 to −11.6 mm relative to bregma over 8 sections (Hornung, 2003; Paxinos and Watson, 1997). The percentage change in 5-HT-positive neurons ipsilateral to the carotid ligation relative to the non-ligated side was calculated for each animal. To determine the effect of P3 HI on brainstem and forebrain hemisphere size, the outlines of the right and left hemispheres were traced separately over 3 consecutive sections using the software program Analysis Life Science Research. The brainstem hemisphere size was determined at the level of −7.64 to −8.3 mm relative to bregma and the forebrain hemisphere size was calculated at the level 3.7 to 2.7 mm relative to bregma. The percentage change ipsilateral to the ligation relative to contralateral was calculated as previously described (Buller et al., 2008). For analysis of SERT protein expression in Western blotting, bands were analysed for optical density using commercial software (NIH Image J) and adjusted for amount of protein expressed in the loading control of beta-actin. Statistical analysis between control and P3 HI animals were performed in PRISM 5 (PRISM 5, GraphPad, San Diego, CA, USA). All data was analysed using the Kruskal– Wallis non-parametric t-test. Data are expressed as mean ± SEM and p < 0.05 was considered statistically significant.
Acknowledgments We would like to thank the National Health and Medical Research Council of Australia for funding this study. K.M.B is supported by a Lions Medical Research Fellowship and H.E.R. is supported by a University of Queensland International Research Tuition Award.
REFERENCES
Arvin, K.L., Han, B.H., Du, Y., Lin, S.Z., Paul, S.M., Holtzman, D.M., 2002. Minocycline markedly protects the neonatal brain against hypoxic–ischemic injury. Ann. Neurol. 52, 54–61.
Baumann, B., Bielau, H., Krell, D., Agelink, M.W., Diekmann, S., Wurthmann, C., Trubner, K., Berstein, H.G., Danos, P., Bogerts, B., 2002. Circumscribed numerical deficit of dorsal raphe neurons in mood disorders. Psychol. Med. 32, 93–103. Bowker, R.M., Westlund, K.N., Coulter, J.D., 1981. Origins of serotonergic projections to the spinal cord in rat: an immunocytochemical-retrograde transport study. Brain Res. 226, 187–199. Buller, K.M., Wixey, J.A., Pathipati, P., Carty, M., Colditz, P.B., Williams, C.E., Scheepens, A., 2008. Selective losses of brainstem catecholamine neurons after hypoxia–ischemia in the immature rat pup. Pediatr. Res. 63, 364–369. Carty, M.L., Wixey, J.A., Colditz, P.B., Buller, K.M., 2008. Post-hypoxia–ischemia minocycline treatment attenuates neuroinflammation and white matter injury in the neonatal rat; a comparison of two different dose regimens. Int. J. Dev. Neurosci. 26, 477–485. Choi, D.W., 2001. Excitotoxicity, apoptosis, and ischemic stroke. J. Biochem. Mol. Biol. 34, 8–14. Clancy, R.R., Sladky, J.T., Rorke, L.B., 1989. Hypoxic–ischemic spinal cord injury following perinatal asphyxia. Ann. Neurol. 25, 185–189. Clancy, B., Darlington, R.B., Finlay, B.L., 2001. Translating developmental time across mammalian species. Neuroscience 105, 7–17. Dayan, P., Huys, Q.J., 2009. Serotonin in affective control. Ann. Rev. Neurosci. 32, 95–126. Fan, L.W., Lin, S., Pang, Y., Rhodes, P.G., Cai, Z., 2006. Minocycline attenuates hypoxia–ischemia-induced neurological dysfunction and brain injury in the juvenile rat. Eur. J. Neurosci. 24, 341–350. Fujita, M., Shimada, S., Maeno, H., Nishimura, T., Tohyama, M., 1993. Cellular localization of serotonin transporter mRNA in the rat brain. Neurosci. Lett. 162, 59–62. Galineau, L., Kodas, E., Guilloteau, D., Vilar, M.P., Chalon, S., 2004. Ontogeny of the dopamine and serotonin transporters in the rat brain: an autoradiographic study. Neurosci. Lett. 363, 266–271. Gluckman, P.D., Pinal, C.S., Gunn, A.J., 2001. Hypoxic–ischemic brain injury in the newborn: pathophysiology and potential strategies for intervention. Semin. Neonatol. 6, 109–120. Graham, E.M., Ruis, K.A., Hartman, A.L., Northington, F.J., Fox, H.E., 2008. A systematic review of the role of intrapartum hypoxia–ischemia in the causation of neonatal encephalopathy. Am. J. Obstet. Gynecol. 199, 587–595. Groenendaal, F., Vles, J., Lammers, H., De Vente, J., Smit, D., Nikkels, P.G., 2008. Nitrotyrosine in human neonatal spinal cord after perinatal asphyxia. Neonatology 93, 1–6. Hadjiconstantinou, M., Yates, A.J., Neff, N.H., 1990. Hypoxia-induced neurotransmitter deficits in neonatal rats are partially corrected by exogenous GM1 ganglioside. J. Neurochem. 55, 864–869. Hagberg, H., Peebles, D., Mallard, C., 2002. Models of white matter injury: comparison of infectious, hypoxic–ischemic, and excitotoxic insults. Ment. Retard. Dev. Disabil. Res. Rev. 8, 30–38. Hawker, K., Lang, A.E., 1990. Hypoxic–ischemic damage of the basal ganglia. Case reports and a review of the literature. Mov. Disord. 5, 219–224. Hornung, J.P., 2003. The human raphe nuclei and the serotonergic system. J. Chem. Neuroanat. 26, 331–343. Johnston, M.V., Trescher, W.H., Ishida, A., Nakajima, W., 2001. Neurobiology of hypoxic–ischemic injury in the developing brain. Pediatr. Res. 49, 735–741. Krakovsky, G., Huth, M.M., Lin, L., Levin, R.S., 2007. Functional changes in children, adolescents, and young adults with cerebral palsy. Res. Dev. Disabil. 28, 331–340. Leech, R.W., Brumback, R.A., 1988. Massive brain stem necrosis in the human neonate: presentation of three cases with review of the literature. J. Child. Neurol. 3, 258–262.
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 – 1 33
Leonardo, C.C., Pennypacker, K.R., 2009. Neuroinflammation and MMPs: potential therapeutic targets in neonatal hypoxic–ischemic injury. J. Neuroinflam. 6, 13. Lidov, H.G.W., Molliver, M.E., 1982. An immunohistochemical study of serotonin neuron development in the rat: ascending pathways and terminal fields. Brain Res. Bull. 8, 389–430. Loewy, A.D., McKellar, S., 1981. Serotonergic projections from the ventral medulla to the intermediolateral cell column in the rat. Brain Res. 211, 146–152. Michelsen, K.A., Prickaerts, J., Steinbusch, H.W., 2008. The dorsal raphe nucleus and serotonin: implications for neuroplasticity linked to major depression and Alzheimer's disease. Prog. Brain Res. 172, 233–264. Montanez, S., Daws, L.C., Gould, G.G., Frazer, A., 2003. Serotonin (5-HT) transporter (SERT) function after graded destruction of serotonergic neurons. J. Neurochem. 87, 861–867. Nemeroff, C.B., Owens, M.J., 2009. The role of serotonin in the pathophysiology of depression: as important as ever. Clin. Chem. 55, 1578–1579. Pardo, C.A., Eberhart, C.G., 2007. The neurobiology of autism. Brain Pathol. 17, 434–447. Paterson, D.S., Trachtenberg, F.L., Thompson, E.G., Belliveau, R.A., Beggs, A.H., Darnall, R., Chadwick, A.E., Krous, H.F., Kinney, H.C., 2006. Multiple serotonergic brainstem abnormalities in sudden infant death syndrome. J.A.M.A. 296, 2124–2132. Paxinos, G., Watson, C., 1997. The Rat Brain in Stereotaxic Coordinates. Academic Press, San Diego. Peng, J.-H.F., Feng, Y., LeBlanc, M.H., Rhodes, P.G., Parker, J.J.C., 2005. Apoptosis and necrosis in developing cerebellum and brainstem induced after focal cerebral hypoxic–ischemic injury. Brain Res. Dev. Brain Res. 156, 87–92. Peters, B., Walka, M.M., Friedmann, W., Stoltenburg-Didinger, G., Obladen, M., 2000. Hypoxic–ischemic encephalopathy with cystic brain stem necroses and thalamic calcifications in a preterm twin. Brain Dev. 22, 265–271. Qian, Y., Melikian, H.E., Rye, D.B., Levey, A.I., Blakely, R.D., 1995. Identification and characterization of antidepressant-sensitive serotonin transporter proteins using site-specific antibodies. J. Neurosci. 15, 1261–1274. Rone, L.A., Ferrando, S.J., 1996. Serotonin reuptake inhibitorrelated extrapyramidal side effects in two patients with cerebral palsy. Psychosomatics 37, 165–166. Shirashi, M., Takizawa, Y., Ide, S., Obonai, T., Goto, Y., Itoh, M., 2008. Brainstem monoamine pathology of neonatal hypoxic–ischemic brain damage: a model of acute stage of neonatal asphyxia. Brain Res. 1213, 120–126.
133
Sizonenko, S.V., Sirimanne, E., Mayall, Y., Gluckman, P.D., Inder, T., Williams, C., 2003. Selective cortical alteration after hypoxic–ischemic injury in the very immature rat brain. Pediatr. Res. 54, 263–269. Stadlin, A., James, A., Fiscus, R., Wong, Y.F., Rogers, M., Haines, C., 2003. Development of a postnatal 3-day-old rat model of mild hypoxic–ischemic brain injury. Brain Res. 993, 101–110. Stengler-Wenzke, K., Muller, U., Angermeyer, M.C., Sabri, O., Hesse, S., 2004. Reduced serotonin transporter-availability in obsessive–compulsive disorder (OCD). Eur. Arch. Psychiatry Clin. Neurosci. 254, 252–255. Tai, W.C., Burke, K.A., Dominguez, J.F., Gundamraj, L., Turman Jr., J.E., 2009. Growth deficits in a postnatal day 3 rat model of hypoxic–ischemic brain injury. Behav. Brain Res. 202, 40–49. Takeuchi, Y., Fujiwara, K., Ishimura, K., Aoki, S., Okano, S., Yamazoe, I., Yoshioka, H., Sawada, T., 1992. Effects of neonatal asphyxia on the serotonin neuron system in the developing brain studied by immunohistochemistry. Dev. Neurosci. 14, 394–402. Tanaka, H., Amamiya, S., Miura, N., Araki, A., Ohinata, J., Fujieda, K., 2006. Postnatal development of brainstem serotonin-containing neurons projecting to lumbar spinal cord in rats. Brain Dev. 28, 586–591. Tarazi, F.I., Tomasini, E.C., Baldessarini, R.J., 1998. Postnatal development of dopamine and serotonin transporters in rat caudate–putamen and nucleus accumbens septi. Neurosci. Lett. 254, 21–24. Tomimatsu, T., Fukuda, H., Endoh, M., Mu, J., Watanabe, N., Kohzuki, M., Fujii, E., Kanzaki, T., Oshima, K., Doi, K., Kubo, T., Murata, Y., 2002. Effects of neonatal hypoxic–ischemic brain injury on skilled motor tasks and brainstem function in adult rats. Brain Res. 926, 108–117. Towfighi, J., Mauger, D., Vannucci, R.C., Vannucci, S.J., 1997. Influence of age on the cerebral lesions in an immature rat model of cerebral hypoxia–ischemia: a light microscopic study. Brain. Res. Dev. Brain Res. 100, 149–160. Vannucci, R.C., Lyons, D.T., Vasta, F., 1988. Regional cerebral blood flow during hypoxia–ischemia in immature rats. Stroke 19, 245–250. Volpe, J.J., 2001. Neurobiology of periventricular leukomalacia in the premature infant. Pediatr. Res. 50, 553–562. Volpe, J.J., 2009. Brain injury in premature infants: a complex amalgam of destructive and developmental disturbances. Lancet Neurol. 8, 110–124. Wixey, J.A., Reinebrant, H.E., Carty, M.L., Buller, K.M., 2009. Delayed P2X4R expression after hypoxia–ischemia is associated with microglia in the immature rat brain. J. Neuroimmunol. 212, 35–43.
available at www.sciencedirect.com
www.elsevier.com/locate/brainres
Research Report
Differential effects of neonatal hypoxic–ischemic brain injury on brainstem serotonergic raphe nuclei Hanna E. Reinebrant a , Julie A. Wixey a , Glenda C. Gobe b , Paul B. Colditz a , Kathryn M. Buller a,⁎ a
Perinatal Research Centre, Clinical Neuroscience, University of Queensland Centre for Clinical Research, Brisbane, Queensland, Australia Molecular and Cellular Pathology, School of Medicine, University of Queensland, Brisbane, Queensland, Australia
b
A R T I C LE I N FO
AB S T R A C T
Article history:
Serotonergic fibres have a pervasive innervation of hypoxic–ischemic (HI)-affected areas in
Accepted 23 January 2010
the neonatal brain and serotonin (5-HT) is pivotal in numerous neurobehaviours that match
Available online 1 February 2010
many HI-induced deficits. However, little is known about how neonatal HI affects the serotonergic system. We therefore examined whether neonatal HI can alter numbers of
Keywords:
serotonergic raphe neurons in specific sub-divisions of the midbrain and brainstem since
Brainstem
these nuclei are the primary sources of serotonin throughout the central nervous system
Hypoxia–ischemia
(CNS). We utilised an established neonatal HI model in the postnatal day 3 (P3) rat pup (right
Neonate
common carotid artery ligation + 30 min 6% O2) and determined the effects of P3 HI on 5-HT
Raphe nuclei
counts in 5 raphe sub-divisions in the midbrain and brainstem one and six weeks later. After
Serotonin
P3 HI, numbers of 5-HT-positive neurons were significantly decreased in the dorsal raphe
Serotonin transporter
dorsal, dorsal raphe ventrolateral and dorsal raphe caudal nuclei on P10 but only in the dorsal raphe dorsal and dorsal raphe ventrolateral nuclei on P45. In contrast, P3 HI did not alter counts in the dorsal raphe interfascicular and raphe magnus nuclei. We also discovered that P3 HI significantly reduces brainstem SERT protein expression; the key regulator of 5-HT in the CNS. In conclusion, neonatal HI injury caused significant disruption of the brainstem serotonergic system that can persist for up to six weeks after the insult. The different vulnerabilities of serotonergic populations in specific raphe nuclei suggest that certain raphe nuclei may underpin neurological deficits in HI-affected neonates through to adulthood. © 2010 Elsevier B.V. All rights reserved.
1.
Introduction
Hypoxic–ischemic brain injury is one of the major factors contributing to neonatal morbidity and mortality in preterm
neonates. Neurological deficits after preterm neonatal HI include cerebral palsy, mental retardation, learning disabilities and epilepsy (Graham et al., 2008; Volpe, 2001, 2009). In the premature infant white matter damage is a characteristic
⁎ Corresponding author. Perinatal Research Centre, UQ Centre for Clinical Research, Building #71/918, Royal Brisbane & Women's Hospital, Herston, QLD 4029, Australia. Fax: +61 7 33465594. E-mail address: [email protected] (K.M. Buller). Abbreviations: 5-HIAA, 5-hydroxyindoleacetic acid; 5-HT, serotonin; Aq, aqueduct (Sylvius); CNS, central nervous system; DRc, dorsal raphe caudal nucleus; DRd, dorsal raphe dorsal nucleus; DRif, dorsal raphe interfascicular nucleus; DRvl, dorsal raphe ventrolateral nucleus; HI, hypoxia–ischemia; P, postnatal day; RM, raphe magnus nucleus; SERT, serotonin transporter; SIDS, sudden infant death syndrome; xscp, decussation of the superior cerebellar peduncle 0006-8993/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2010.01.065
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 – 1 33
feature of HI injury, but extensive neuronal damage also occurs (Arvin et al., 2002; Fan et al., 2006; Gluckman et al., 2001; Hawker and Lang, 1990; Johnston et al., 2001; Sizonenko et al., 2003). However little is known about the precise neuronal phenotypes altered after neonatal HI. Serotonin (5-HT) is a major neurotransmitter and serotonergic fibres have a pervasive network throughout the central nervous system (CNS). Many forebrain regions innervated by serotonergic fibres are damaged after HI such as the cortex, striatum, thalamus and amygdala (Stadlin et al., 2003; Towfighi et al., 1997). Furthermore 5-HT is a neurochemical that is involved in a vast array of functions and alterations to the 5-HT system can cause long-term deficits; many of which match those manifested after neonatal HI (Dayan and Huys, 2009). It is therefore important to establish how neonatal HI might alter serotonergic neurons in the brain and begin to decipher whether these specific nuclei constitute primary candidate networks that underpin neonatal HI-induced neurological deficits. Serotonergic neurons are localised primarily in the raphe nuclei of the midbrain and brainstem. The major raphe complex of 5-HT-positive neurons is the dorsal raphe and this complex can be further divided into several sub-divisions located in the midbrain. The dorsal raphe nuclei project to and innervate many target areas in the forebrain via the dorsal, medial and ventral ascending pathways, including the thalamus, hypothalamus, several cortical areas, septum, amygdala nuclei and hippocampus (Michelsen et al., 2008). The raphe magnus nucleus, located more caudally in the brainstem, predominantly projects to the spinal cord (Tanaka et al., 2006). Although damage to the brainstem after HI is evident in human neonates (Leech and Brumback, 1988; Peters et al., 2000) and animal models (Buller et al., 2008; Peng et al., 2005; Tomimatsu et al., 2002), information about the specific effects of the raphe nuclei are scarce. One study has shown that total asphyxia for 30 min in P2 mice results in a significant loss of 5HT-positive neurons in specific raphe nuclei in the brainstem up to 60 days after injury. Asphyxia induced permanent changes in the rostral raphe system, while the caudal raphe system appeared unaffected in terms of 5-HT-positive numbers (Takeuchi et al., 1992). Whether these changes are evident after neonatal HI has not been investigated. Also 5-HT levels in forebrain regions are reduced after P7 hypoxia in rats (Hadjiconstantinou et al., 1990) but the effects of HI on 5-HTpositive neurons residing in the raphe nuclei have not been examined. It is also not known if neonatal HI can alter the expression of the serotonin transporter (SERT); the key regulator 5-HT levels in the CNS. Thus, despite the knowledge that serotonergic fibres innervate areas affected by neonatal HI and the importance of the serotonergic system throughout the brain, it remains to be determined if 5-HT raphe nuclei are damaged after neonatal HI, whether specific sub-divisions are more vulnerable to HI injury than others and if SERT is altered in the brainstem after neonatal HI. In the present study we hypothesised that HI in the immature brain affects numbers of brainstem serotonergic neurons in the raphe nuclei one week (P10) and six weeks (P45) after P3 HI. We used a well established rat model of preterm HI, the P3 rat pup (Buller et al., 2008; Carty et al., 2008; Tai et al., 2009). Using immunohistochemistry we phenotypically identified 5-HT-positive neurons in five specific raphe nuclei and
125
determined if neonatal HI changes numbers of 5-HT-positive neurons in the dorsal raphe nucleus (dorsal, ventrolateral, interfascicular and caudal sub-divisions) and medullary raphe nuclei (raphe magnus) on P10 and P45. We also examined whether P3 HI alters SERT protein expression in the brainstem on P10 and P45.
2.
Results
2.1. Numbers of 5-HT-positive neurons in the raphe nuclei of control animals on P10 and P45 In control animals developmental changes in the number of 5HT-positive neurons in certain raphe nuclei were apparent over the five-week period from P10 to P45. We found a higher number of 5-HT-positive neurons in the dorsal raphe interfascicular and caudal nuclei and raphe magnus nucleus on P10 compared to P45. For the interfascicular, caudal and raphe magnus nuclei there was a significant 30.9%, 39.3% and 32.5% decrease in numbers of 5-HT-positive neurons from P10 to P45, respectively (Fig. 1). In contrast, there was no difference in the number of 5-HT-positive neurons in the dorsal raphe ventrolateral and dorsal raphe dorsal nuclei on P10 compared to P45 (Fig. 1).
2.2.
Effect of P3 HI on brain hemisphere size
We determined cerebral hemisphere size in the brainstem and forebrain in control and P3 HI animals as a measure of injury. In the brainstem, there was no difference in hemisphere size ipsilateral to the carotid ligation in comparison to the contralateral non-ligated side in animals subjected to P3 HI
Fig. 1 – Total numbers of 5-HT-positive neurons in the dorsal raphe dorsal (DRd), dorsal raphe ventrolateral (DRvl), dorsal raphe interfascicular (DRif), dorsal raphe caudal (DRc) and raphe magnus (RM) nuclei of control animals on P10 and P45. See Fig. 2 for localisation and immunolabelling of the raphe nuclei sub-division. The total number of 5-HT-positive neurons in the DRif, DRc and RM nuclei significantly decreased on P45 compared to P10 by 31%, 39% and 32%, respectively. In contrast, there was no difference in total counts on P10 and P45 in the DRd and DRvl nuclei. *p < 0.05, **p < 0.01.
126
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 –13 3
Fig. 2 – Schematic diagrams modified from the rat atlas (Paxinos and Watson, 1997) indicating the location of photomicrographs of coronal sections immunolabelled for 5-HT in selected raphe nuclei of control and P3 HI animals on P10. Photomicrographs have been taken through the dorsal raphe dorsal (A–B), the dorsal raphe ventrolateral (C–D), the dorsal raphe interfascicular (E–F), the dorsal raphe caudal (G–H) and the raphe magnus (I–J) nuclei ipsilateral to the ligated side in control animals (A, C, E, G, I) and P3 HI animals (B, D, F, H, J). Scale bars represent 25 µm. Rostro-caudal distance (mm) given relative to bregma (Paxinos and Watson, 1997). Aq, aqueduct (Sylvius); DRd, dorsal raphe dorsal nucleus; DRc, dorsal raphe caudal nucleus; DRif, dorsal raphe interfascicular nucleus; DRvl, dorsal raphe ventrolateral nucleus; RM, raphe magnus nucleus; xscp, decussation of the superior cerebellar peduncle.
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 – 1 33
on P10 (− 0.13% ± 0.85%) or P45 (− 0.38% ± 1.01%). Further, compared to the left side in control animals there was no difference in brainstem area on the non-ligated side of P3 HI animals. However, in the forebrain after P3 HI, there was a significant reduction in cerebral hemisphere size on the ligated side in comparison to the non-ligated side on both P10 (18.4% ± 3.43%) and P45 (11.7% ± 1.22%). These results are consistent with our previous studies using the same P3 HI rodent model where brainstem area was unchanged after P3 HI (Buller et al., 2008) and forebrain hemisphere size was decreased after neonatal hypoxic–ischemic injury (Buller et al., 2008; Wixey et al., 2009). Thus we were confident we reproduced a certain severity of brain injury after P3 HI.
127
P3 HI, there was a significant 18.6% reduction (Fig. 3). There was no difference between 5-HT-positive neuronal counts on the P3 HI non-ligated side and the corresponding side in control animals on P10 and P45.
2.3.3.
Dorsal raphe interfascicular nucleus
2.3.
Effects of P3 HI on 5-HT-positive dorsal raphe nuclei
On P10, one week after P3 HI, there was no change in 5-HTpositive neurons ipsilateral to the ligation compared to the non-ligated side in the dorsal raphe interfascicular nucleus (Figs. 2 and 3). Similarly on P45, six weeks after P3 HI, there was no difference in the number of 5-HT-positive neurons on the ipsilateral side compared to the non-ligated side (Fig. 3). Also no difference in 5-HT counts was found between the nonligated side and the corresponding side in control animals on P10 and P45.
2.3.1.
Dorsal raphe dorsal nucleus
2.3.4.
On P10, one week after P3 HI, there was a significant 21.4% reduction in the number of 5-HT-positive neurons ipsilateral to the ligation compared to the non-ligated side in the dorsal raphe dorsal nucleus (Figs. 2 and 3). In addition, on P45, six weeks after P3 HI, there was a significant 14.1% reduction after P3 HI on the ipsilateral side compared to the non-ligated side (Fig. 3). There was no difference in numbers of 5-HT-positive neurons between the non-ligated side and the corresponding side in control animals.
2.3.2.
Dorsal raphe caudal nucleus
On P10, one week after P3 HI there was a significant 10.7% reduction in the number of 5-HT-positive neurons ipsilateral to the ligation compared to the non-ligated side in the dorsal raphe caudal nucleus (Figs. 2 and 3). In contrast on P45 in the dorsal raphe caudal nucleus, six weeks after P3 HI, there was no difference in the number of 5-HT-positive neurons ipsilateral to the ligation compared to the non-ligated side (Fig. 3). There was no difference in 5-HT-positive neuronal counts between the non-ligated side and the corresponding side in control animals.
Dorsal raphe ventrolateral nucleus
On P10, one week after P3 HI there was a significant 20.2% reduction of 5-HT-positive neurons ipsilateral to the ligation compared to the non-ligated side in the dorsal raphe ventrolateral nucleus (Figs. 2 and 3). On P45, six weeks after
2.3.5.
Raphe magnus nucleus
On P10 and P45 there was no difference in the number of 5-HTpositive neurons ipsilateral to the ligation compared to the non-ligated side in the raphe magnus nucleus (Fig. 3). In
Fig. 3 – The effects of P3 HI on numbers of 5-HT-positive neurons in specific raphe nuclei on P10 (A) and P45 (B). In P3 HI animals on P10 (A), there was a significant loss of 5-HT-positive neurons in the dorsal raphe dorsal, the dorsal raphe ventrolateral and the dorsal raphe caudal nuclei ipsilateral to the carotid ligation. Counts in the dorsal raphe interfascicular and the raphe magnus nuclei remained unchanged relative to the non-ligated side. In control animals there was no difference in 5-HT-positive neuron counts on the left side compared to the right side (A). On P45, in P3 HI animals, there was a significant loss of 5-HT-positive neurons on the ligated side compared to the non-ligated side in the dorsal raphe dorsal and the dorsal raphe ventrolateral nuclei. The dorsal raphe interfascicular, the dorsal raphe caudal and the raphe magnus remained unchanged on the ligated side compared to the non-ligated side. In control animals there was no difference in 5-HT counts between the left side and right side (B). DRd, dorsal raphe dorsal nucleus; DRc, dorsal raphe caudal nucleus; DRif, dorsal raphe interfascicular nucleus; DRvl, dorsal raphe ventrolateral nucleus; RM, raphe magnus nucleus *p < 0.05, **p < 0.01, ***p < 0.005.
128
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 –13 3
addition, there was no difference in numbers of 5-HT-positive neurons between the non-ligated side and the corresponding side in control animals.
2.4.
Effects of P3 HI on SERT protein expression
Western blotting was used to detect possible changes in SERT protein expression in the brainstem after P3 HI. On P10, one week after P3 HI, there was a significant 23.4% reduction in SERT protein expression compared to control animals (Figs. 4A–B). In contrast, on P45 there was no difference in SERT protein expression between P3 HI and control animals (Figs. 4A–B). Representative sections were also labelled for SERT and we found that SERT immunoreactivity mirrored the changes in 5-HT-positive neurons in the dorsal raphe
(Figs. 4C–D). To further confirm that SERT is concentrated on 5-HT-positive raphe neurons in the brainstem (Fujita et al., 1993; Qian et al., 1995) we performed dual immunofluorescent labelling for 5-HT and SERT and demonstrated a strong colocalization of SERT on 5-HT-positive neurons in the dorsal raphe (Figs. 4E–G).
3.
Discussion
This is the first study to demonstrate that neonatal HI alters the number of 5-HT-positive raphe neurons and that P3 HI also decreases SERT protein expression in the brainstem. We found that certain sub-populations of 5-HT-positive raphe neurons are more vulnerable to injury after P3 HI than others.
Fig. 4 – The effect of P3 HI on SERT protein expression ipsilateral to the carotid ligation on P10 and P45 in the brainstem. Example Western blots to detect SERT are shown (A). The amount of SERT protein detected is expressed as a percentage change from control levels. Beta-actin was used as the loading control (A). After P3 HI, SERT protein expression was significantly decreased on P10 (B). However, on P45, there was no difference in SERT expression between control and P3 HI animals (B). Immunoreactivity for SERT in the dorsal raphe of control (C) and P3 HI animals (D) demonstrated that numbers of SERT-positive nuclei mirrored the changes in immunolabelling for 5-HT-positive neurons in the dorsal raphe. To confirm the localization of SERT on brainstem raphe neurons, examples of dorsal raphe sections dual-labelled for 5-HT (E, green) and SERT (F, red) are shown. Colocalization (G, yellow) of the SERT-positive marker was virtually found exclusively on 5-HT-positive neurons in the dorsal raphe nucleus. *p < 0.05; scale bars represent 25 µm.
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 – 1 33
In addition, we observed developmental changes in the number of 5-HT-positive raphe neurons in normal control animals from P10 to P45.
3.1. nuclei
Developmental change in total 5-HT-positive raphe
Interestingly, in normal unoperated rats, ontogenic differences in numbers of 5-HT-positive neurons were observed over the five-week period examined (from P10 to P45). These changes only occurred in certain raphe nuclei sub-divisions whereby a lower number of 5-HT-positive neurons were evident on P45 compared to P10 in the dorsal raphe caudal, dorsal raphe interfascicular and raphe magnus nuclei. Counts in the dorsal raphe ventrolateral and dorsal raphe dorsal nuclei remained unchanged over this five-week period. Thus total numbers of 5-HT neurons did not increase over time in any of the nuclei raphe examined. To the best of our knowledge only one quantitative study of 5-HT-positive neurons in the raphe magnus nucleus has been performed (Tanaka et al., 2006). Over a similar developmental period to the present study, Tanaka et al. (2006) reported a significant decrease in the number of 5-HT-positive raphe magnus neurons (Tanaka et al., 2006). The decreased number of 5HT-positive neurons in specific raphe nuclei could be due to normal neuronal pruning during development. Tanaka et al. (2006) also demonstrated that, despite the decrease in number of 5-HT neurons in the raphe magnus nucleus, a larger number of neurons were back-labelled after tracer injection to the spinal cord, suggesting axonal growth and dendritic arborisation of descending raphe neurons (Tanaka et al., 2006). Previous studies have shown increased serotonergic innervation density, increased arborization and axonal length and higher expression of the SERT in the forebrain over time, indicating the developmental plasticity and axonal sprouting of the serotonergic system (Lidov and Molliver, 1982; Tarazi et al., 1998). It has also been demonstrated that SERT, mainly localised on 5-HT-positive neurons, increases expression in the early postnatal days, reaches its highest expression in the dorsal raphe nucleus on P14, after which the SERT expression decreases until P21 then remains steady at ‘adult’ levels (Galineau et al., 2004). These results support the developmental changes observed in the dorsal raphe in the present study. Taken together, we speculate that normal pruning of serotonergic nuclei and developmental growth has occurred in the raphe sub-divisions during the five-week period.
3.2.
Brain hemisphere size after P3 HI
We utilised an established rat model of preterm HI (P3 HI) that we, and others, have shown previously to cause white matter damage, loss of neurons and behavioural deficits (Buller et al., 2008; Wixey et al., 2009; Stadlin et al., 2003). As an index of brain injury in the present study we determined forebrain and brainstem hemisphere size ipsilateral to the carotid ligation relative to the non-ligated side. Consistent with our previous studies forebrain cerebral hemisphere size was significantly reduced after P3 HI. In addition, since the primary region of interest was the midbrain and brainstem where the majority of raphe neurons reside, we measured the brainstem hemi-
129
sphere size at the level of the dorsal raphe nuclei. Unlike the forebrain outcome, there was no effect of P3 HI on brainstem hemisphere size. Using the same P3 HI rodent model we have demonstrated previously, at a more caudal level of the medulla oblongata, hemisphere area is also unchanged (Buller et al., 2008). Thus not only were we confident that we reproduced a certain severity of brain injury after P3 HI consistent with our previous studies but we reinforced the phenomenon that P3 HI does not incur gross changes in brainstem hemisphere size.
3.3.
Effects of P3 HI on specific raphe nuclei
There was an overall significant loss of 5-HT-positive raphe neurons after P3 HI, consistent with previous findings (Takeuchi et al., 1992; Shirashi et al., 2008) however the neuronal loss was greater in specific raphe sub-populations than others. Thus certain serotonergic raphe nuclei may be more vulnerable to HI-induced injury than others. The neuronal loss was also more prominent on P10 than on P45. On P10 after P3 HI, 5-HT-positive losses were observed in the dorsal raphe caudal, dorsal raphe ventrolateral and dorsal raphe dorsal nuclei. The dorsal raphe interfascicular and the raphe magnus nuclei, showed no reduction in number of 5HT-positive neurons on P10. In contrast, on P45, only the dorsal raphe ventrolateral and the dorsal raphe dorsal elicited prolonged damage and maintained a significant loss of 5-HTpositive neurons. The topographical distribution of raphe nuclei may dictate their vulnerability to HI injury. Mostly the more anterior raphe sub-divisions were affected by P3 HI as opposed to the more posterior and caudally located nuclei such as the dorsal raphe interfascicular and the raphe magnus nuclei. The positional clustering of different raphe sub-divisions in the midbrain and brainstem also represents differential connectivity patterns in the brain. The dorsal raphe caudal, dorsal raphe ventrolateral and dorsal raphe dorsal nuclei predominantly send axonal projections to the cerebral cortex, basal ganglia, thalamus, hypothalamus, hippocampus and amygdala and are involved in influencing or regulating numerous functions. Each serotonergic sub-division is integral in producing characteristic functional effects based largely on their afferent and efferent connections in the brain. Specific losses of these 5-HT raphe nuclei in these sub-divisions can therefore affect numerous functions and might underpin regulatory mechanisms that are responsible for anxiety, movement, depression, cardiovascular, respiratory deficits reported in HI-affected preterm infants. Serotonin levels are reduced in the brain after neonatal P7 hypoxia (Hadjiconstantinou et al, 1990) but it is not known if there is regional variation in the effect of neonatal HI on serotonin release. Nonetheless the loss of dorsal raphe neurons after HI may explain reduced serotonergic release in the forebrain. In direct contrast to the dorsal raphe nuclei, on both P10 and P45, the number of 5-HTpositive neurons in the raphe magnus nucleus was unchanged after P3 HI. Thus we propose that these neurons are not susceptible to HI injury and remain functionally viable after P3 HI. Retrograde axonal transport studies have shown that the medullary raphe system, including the raphe magnus, provides the major descending serotonergic input to the spinal
130
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 –13 3
cord but that serotonergic nuclei of the raphe magnus send very few ascending projections to the forebrain (Bowker et al., 1981; Loewy and McKellar, 1981). Therefore it is likely, at least in our model, that the serotonergic innervation of the spinal cord remains relatively intact after P3 HI. This suggestion is reinforced by evidence that the spinal cord appears to endure damage only after severe neonatal HI insults (Clancy et al., 1989; Groenendaal et al., 2008). Thus determining the loss of specific raphe nuclei and their ascending and descending neural connections may predict their vulnerability to P3 HI injury as well as their wider influence on other brain regions after P3 HI. It remains to be investigated whether specific serotonergic changes in the dorsal raphe nuclei lead to selective serotonergic HI deficits in the forebrain and associated changes in serotonin-mediated functions.
3.4.
Reduced SERT protein expression after P3 HI
Since distinct losses of 5-HT-positive neurons were detected after P3 HI we wished to establish whether another serotonergic neuronal marker, SERT, was also affected in the brainstem. SERT is a pivotal regulator of serotonin levels in the CNS and thus its expression is an indicator of serotonergic synaptic function. Importantly, we found that SERT protein expression in the brainstem was significantly decreased on P10 after P3 HI. The loss of SERT protein on P10 could be explained by a loss of 5-HT-positive neurons in specific raphe nuclei since we confirmed the localization of SERT on 5-HTpositive neurons (Fujita et al., 1993; Qian et al., 1995). In contrast, on P45 the SERT protein levels were not significantly lower than control levels. The overall recovery of SERT protein expression did not reflect the overall maintained loss of raphe neurons after P3 HI. It is possible that remaining neurons increased their expression of SERT protein to compensate for neuronal losses over time. It has been shown that SERT function is maintained with as little as 20% of its original population intact (Montanez et al., 2003). It is now important for future studies to investigate the functional relevance of the P3 HI-induced losses of 5-HT neurons and SERT expression not only at the neuronal level but also in terms of behavioural deficits.
3.5. Possible mechanisms responsible for brainstem serotonergic injury The observed decreases in numbers of 5-HT-positive dorsal raphe neurons and the associated reduction in SERT protein expression in the brainstem affirm that P3 HI resulted in losses of serotonergic neurons. In P7 rat pups with hypoxiainduced brain injury, both 5-HT and its metabolite, 5hydroxyindoeacetic acid (5-HIAA) are decreased in the forebrain, suggesting neuronal loss occurs rather than metabolic changes (Hadjiconstantinou et al., 1990). Moreover, losses cannot simply be explained by developmental decreases in numbers of 5-HT-positive neurons since effects of P3 HI were calculated as a percentage of the non-ligated side in each animal for the specific postnatal day. Thus each animal essentially served as its own control. The mechanisms underpinning the observed losses of 5HT-positive neurons are not known. The brainstem is located
outside the vascular field of the ipsilateral hemisphere and should not be directly affected by immediate changes in perfusion after common carotid artery occlusion. In fact an increase in blood flow occurs in the brainstem after such an insult (Vannucci et al., 1988). Further we did not detect any change in brainstem hemisphere area after P3 HI, in contrast to significant reductions in forebrain cerebral hemisphere area. The losses of 5-HT-positive neurons in the brainstem after P3 HI are therefore likely to occur via secondary mechanisms. Candidate mechanisms responsible for this secondary injury after P3 HI include excitotoxic mechanisms, neuroinflammation and axonal degeneration. Serotonergic afferent and efferent projections connect major areas of the forebrain, largely matching or at least overlapping with areas known to be damaged after HI, such as the striatum, thalamus and amygdala (Michelsen et al., 2008; Stadlin et al., 2003; Towfighi et al., 1997). Thus it is plausible that damage to these forebrain areas is likely to affect serotonergic fibres, innervating or coursing through these areas as fibres of passage, and cause retrograde cell death in remote areas such as the raphe serotonergic nuclei. It follows, that the regional vulnerability of 5-HT-positive neurons in the dorsal raphe nucleus after P3 HI in this study could be explained by the serotonergic innervation pattern of the forebrain. Possibly the innervation is more dense from certain sub-divisions in the dorsal raphe nucleus to the HI-affected areas in the forebrain, leading to more pronounced damage in these areas. As alluded to earlier, whether there is corresponding serotonergic neuropathology in the forebrain after neonatal HI remains to be determined. The early neuronal injury after HI is believed to be primarily via necrosis resulting from excitotoxic damage; most likely as a result of excessive release of glutamate from presynaptic nerve terminals and astrocytes into the extracellular space, which causes calcium overload followed by necrosis (Choi, 2001). The loss of 5-HT-labelling in the brainstem raphe neurons could be due to apoptosis or necrosis of the neurons. Increased apoptosis has previously been shown in remote areas after P7 HI (Tomimatsu et al., 2002). Secondary neuroinflammation events, such as increased numbers of activated microglia after HI secrete a variety of chemokines and pro-inflammatory cytokines including tumour necrosis factor alpha and interleukin-1 beta, which further exacerbate neuronal injury (Leonardo and Pennypacker, 2009). Neuroinflammation could contribute to brainstem injury via remote actions at the primary injury sites, where inflammatory mediators may damage afferent and efferent fibres of dorsal raphe nuclei and subsequently decrease the survival of brainstem nuclei. Alternatively, neuroinflammation may occur in the brainstem itself and therefore more directly affect serotonergic raphe neuronal function.
3.6.
Conclusions
In the present study we have demonstrated that the serotonergic system in the brainstem is disrupted significantly in a rat model of preterm HI brain injury. Specific losses of 5-HTpositive neurons in select raphe nuclei were observed and an associated decrease in brainstem SERT protein expression. These changes were apparent both one and six weeks after injury, indicating both short-term and long-lasting effects
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 – 1 33
after P3 HI. Disruptions of the serotonergic system are believed to contribute to a multitude of disorders, including SIDS, autism, depression and obsessive–compulsive disorder (Nemeroff and Owens, 2009; Pardo and Eberhart, 2007; Paterson et al., 2006; Stengler-Wenzke et al., 2004). We propose that the HI-induced disruption of the serotonergic system in the brainstem may contribute to functional deficits observed in HI-affected children and adults. For example, decreased serotonergic function is a hallmark feature of depression and depressed patients show 31% loss of dorsal raphe neurons (Baumann et al. 2002). Cerebral palsy is a notable deficit in HIaffected neonates and cerebral palsy patients have been reported to suffer depression (Rone and Ferrando, 1996; Krakovsky et al., 2007). Whether reduced serotonergic function accounts for HI-induced neurological deficits is not known. A plausible avenue of investigation would be to firstly characterise the effects of neonatal HI on serotonergic fibres that innervate specific brain regions.
4.
Experimental procedures
4.1.
Subjects
In this study Sprague–Dawley rat dams and their pups (10–12 per litter) were used. The animals were housed at a constant temperature of 22 °C with a 12 h light cycle and provided standard rat chow and water ad libitum. Experiments were carried out between 8.00 am and 11.00 am in order to minimize possible fluctuations in circadian rhythms. All procedures were approved by the University of Queensland Animal Ethics Committee. Efforts were made to minimize the number of animals used and their suffering.
4.2.
tone (Lethabarb, Virbac, France; 1.5 g/kg i.p) and perfused intracardially with 1% sodium nitrite solution followed by 4% formaldehyde (in 0.1 M PBS, pH 7.4). The brain was excised and post-fixed for 2 h in formaldehyde, and the brainstem was separated from the forebrain at level − 6.7 to −12 mm relative to bregma and then cryo-protected in 10% sucrose (in 0.1 M PBS, pH 7.4, 4 °C) overnight. Serial sections (50 µm for the brainstem and 40 µm for the forebrain) were collected using a sliding microtome. For immunolabelling, brainstem sections from P10 (1-in-4 series, 200 µm intervals) and P45 (1-in-5 series, 250 µm intervals) were incubated with rabbit anti-5-HT antibody (gift from Professor D. Pow) or goat anti-SERT (serotonin transporter, 1:2500; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) in 1% bovine serum albumin at 4 °C for 36 h followed by biotinylated secondary antibody (donkey anti-rabbit or donkey anti-sheep, 1:400; Jackson ImmunoResearch, West Grove, PA, USA). For forebrain hemisphere size, forebrain sections from P10 and P45 (1-in-4 series, 160 µm intervals) were incubated with mouse anti-NeuN antibody (neuronal marker, 1:10,000; Chemicon, Millipore, Autralia) followed by donkey anti-mouse (1:400; Santa Cruz Biotechnology). Sections were then immersed in avidin– biotin–horseradish peroxidase complex (Vector Elite Kit, Burlingame, CA, USA) and the activity was visualized using diaminobenzidine solution. Sections from control and P3 HI animals were processed simultaneously to minimize variations in immunohistochemistry. Sections were then mounted on chrome-alum subbed slides, dehydrated in a series of alcohol, cleared in xylene and coverslipped. For double immunofluorescent labelling, after incubation with primary antibodies (rabbit anti-5-HT, 1:200 and goat anti-SERT, 1:200) brainstem sections were incubated with appropriate Alexa Fluor labels (568 for SERT, 488 for 5-HT, both at 1:200; Invitrogen, Carlsbad, CA), mounted and coverslipped.
Hypoxic–ischemic brain insult 4.4.
On postnatal day 3 (P3) rat pups underwent surgery to induce HI injury. Pups were anaesthetized with Isoflurane (Baxter, 2% in inhalation air, air-flow 3.5 L/min) and the right common carotid artery was exposed, carefully dissected from surrounding tissue and permanently ligated with 5.0 silk thread. Pups were allowed to recover for 15 min on a heating pad and were then placed in a humidified hypoxia chamber for 30 min with constant infusion of 6% O2 and temperature kept at 37 °C. Following hypoxia, pups were again recovered for 15 min on a heating pad and then returned to the dam. Control pups were treated the same way pup but did not undergo surgery and breathed room air. The mortality rate for the HI experiments was <3.5%. The P3 HI rat model is an established preterm animal model that shows white matter damage and neuronal losses (Buller et al., 2008; Carty et al., 2008; Sizonenko et al., 2003; Stadlin et al., 2003) and is analogous to preterm human brain at approximately 24 to 28 weeks gestation in terms of number of synapses, cortical organization and neurochemical development (Clancy et al., 2001; Hagberg et al., 2002).
4.3.
131
Immunohistochemistry
On P10 (control n = 7; P3 HI n = 7) and P45 (control n = 6; P3 HI n = 6), animals were anaesthetized with sodium pentobarbi-
Western blot analyses
On P10 (control n = 9; P3 HI n = 9) and P45 (control n = 10; P3 HI n = 10), animals were anaesthetized with sodium pentobarbitone (Lethabarb, Virbac, France; 1.5 g/kg i.p). The brain was rapidly removed and the brainstem was isolated at level −6.7 to −12 mm relative to bregma and snap-frozen on dry ice. The tissue was then gently homogenised in ice-cold dH2O and centrifuged for 5 min at 4 °C to remove cell debris. The supernatant was recovered and frozen at −80 °C. Total protein was determined using a commercial BCA kit (Pierce, Rockford, IL, USA). Samples (20 µg) from the tissue preparations were mixed with sample buffer at room temperature and separated by 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis using non-denaturing conditions. After separation, the proteins were transferred to a polyvinylidene difluoride membrane (Bio-Rad, Hercules, CA, USA), blocked with 5% non fat dry milk in TBS with 0.1% Tween-20 for 1 h. The membrane was then incubated with SERT antibody (goat anti-SERT, 1:2000; Santa Cruz Biotechnology, Inc.) in 2% non fat dry milk in TBS with 0.1% Tween-20 overnight at 4 °C. The membrane was then incubated with HRP-conjugated antigoat IgG (1:10,000; Sigma, Castle Hill, NSW, Australia) for 2 h and visualized using enhanced chemiluminescence (ECL Plus, Amersham, Sydney, NSW, Australia) on X-ray film. Following
132
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 –13 3
visualization, each blot was stripped using Re-blot Plus (Millipore, Australia) and reprobed with beta-actin antibody (mouse anti-beta-actin, 1:20,000; Imgenex, San Diego, CA, USA) in order to quantify SERT levels relative to beta-actin protein. Chemiluminescence was eliminated for SERT when the antibody was pre-incubated with the control peptide (1:2000; Santa Cruz, CA, USA).
4.5.
Data analysis
To determine the effects of P3 HI, 5-HT neurons were counted in four sub-divisions of the dorsal raphe and the raphe magnus nucleus on P10 and P45. Only 5-HT-positive neurons with a clearly defined nucleus were counted. Raphe nuclei were identified and defined according to the following coordinates in rostro-caudal order: dorsal raphe dorsal −7.04 to −7.64 mm relative to bregma over 3 sections, dorsal raphe ventrolateral −7.64 to −8.3 mm relative to bregma over 3 sections, dorsal raphe interfascicular − 8.0 to − 9.16 mm relative to bregma over 4 sections, dorsal raphe caudal −8.72 to −9.3 mm relative to bregma over 4 sections and raphe magnus −9.16 to −11.6 mm relative to bregma over 8 sections (Hornung, 2003; Paxinos and Watson, 1997). The percentage change in 5-HT-positive neurons ipsilateral to the carotid ligation relative to the non-ligated side was calculated for each animal. To determine the effect of P3 HI on brainstem and forebrain hemisphere size, the outlines of the right and left hemispheres were traced separately over 3 consecutive sections using the software program Analysis Life Science Research. The brainstem hemisphere size was determined at the level of −7.64 to −8.3 mm relative to bregma and the forebrain hemisphere size was calculated at the level 3.7 to 2.7 mm relative to bregma. The percentage change ipsilateral to the ligation relative to contralateral was calculated as previously described (Buller et al., 2008). For analysis of SERT protein expression in Western blotting, bands were analysed for optical density using commercial software (NIH Image J) and adjusted for amount of protein expressed in the loading control of beta-actin. Statistical analysis between control and P3 HI animals were performed in PRISM 5 (PRISM 5, GraphPad, San Diego, CA, USA). All data was analysed using the Kruskal– Wallis non-parametric t-test. Data are expressed as mean ± SEM and p < 0.05 was considered statistically significant.
Acknowledgments We would like to thank the National Health and Medical Research Council of Australia for funding this study. K.M.B is supported by a Lions Medical Research Fellowship and H.E.R. is supported by a University of Queensland International Research Tuition Award.
REFERENCES
Arvin, K.L., Han, B.H., Du, Y., Lin, S.Z., Paul, S.M., Holtzman, D.M., 2002. Minocycline markedly protects the neonatal brain against hypoxic–ischemic injury. Ann. Neurol. 52, 54–61.
Baumann, B., Bielau, H., Krell, D., Agelink, M.W., Diekmann, S., Wurthmann, C., Trubner, K., Berstein, H.G., Danos, P., Bogerts, B., 2002. Circumscribed numerical deficit of dorsal raphe neurons in mood disorders. Psychol. Med. 32, 93–103. Bowker, R.M., Westlund, K.N., Coulter, J.D., 1981. Origins of serotonergic projections to the spinal cord in rat: an immunocytochemical-retrograde transport study. Brain Res. 226, 187–199. Buller, K.M., Wixey, J.A., Pathipati, P., Carty, M., Colditz, P.B., Williams, C.E., Scheepens, A., 2008. Selective losses of brainstem catecholamine neurons after hypoxia–ischemia in the immature rat pup. Pediatr. Res. 63, 364–369. Carty, M.L., Wixey, J.A., Colditz, P.B., Buller, K.M., 2008. Post-hypoxia–ischemia minocycline treatment attenuates neuroinflammation and white matter injury in the neonatal rat; a comparison of two different dose regimens. Int. J. Dev. Neurosci. 26, 477–485. Choi, D.W., 2001. Excitotoxicity, apoptosis, and ischemic stroke. J. Biochem. Mol. Biol. 34, 8–14. Clancy, R.R., Sladky, J.T., Rorke, L.B., 1989. Hypoxic–ischemic spinal cord injury following perinatal asphyxia. Ann. Neurol. 25, 185–189. Clancy, B., Darlington, R.B., Finlay, B.L., 2001. Translating developmental time across mammalian species. Neuroscience 105, 7–17. Dayan, P., Huys, Q.J., 2009. Serotonin in affective control. Ann. Rev. Neurosci. 32, 95–126. Fan, L.W., Lin, S., Pang, Y., Rhodes, P.G., Cai, Z., 2006. Minocycline attenuates hypoxia–ischemia-induced neurological dysfunction and brain injury in the juvenile rat. Eur. J. Neurosci. 24, 341–350. Fujita, M., Shimada, S., Maeno, H., Nishimura, T., Tohyama, M., 1993. Cellular localization of serotonin transporter mRNA in the rat brain. Neurosci. Lett. 162, 59–62. Galineau, L., Kodas, E., Guilloteau, D., Vilar, M.P., Chalon, S., 2004. Ontogeny of the dopamine and serotonin transporters in the rat brain: an autoradiographic study. Neurosci. Lett. 363, 266–271. Gluckman, P.D., Pinal, C.S., Gunn, A.J., 2001. Hypoxic–ischemic brain injury in the newborn: pathophysiology and potential strategies for intervention. Semin. Neonatol. 6, 109–120. Graham, E.M., Ruis, K.A., Hartman, A.L., Northington, F.J., Fox, H.E., 2008. A systematic review of the role of intrapartum hypoxia–ischemia in the causation of neonatal encephalopathy. Am. J. Obstet. Gynecol. 199, 587–595. Groenendaal, F., Vles, J., Lammers, H., De Vente, J., Smit, D., Nikkels, P.G., 2008. Nitrotyrosine in human neonatal spinal cord after perinatal asphyxia. Neonatology 93, 1–6. Hadjiconstantinou, M., Yates, A.J., Neff, N.H., 1990. Hypoxia-induced neurotransmitter deficits in neonatal rats are partially corrected by exogenous GM1 ganglioside. J. Neurochem. 55, 864–869. Hagberg, H., Peebles, D., Mallard, C., 2002. Models of white matter injury: comparison of infectious, hypoxic–ischemic, and excitotoxic insults. Ment. Retard. Dev. Disabil. Res. Rev. 8, 30–38. Hawker, K., Lang, A.E., 1990. Hypoxic–ischemic damage of the basal ganglia. Case reports and a review of the literature. Mov. Disord. 5, 219–224. Hornung, J.P., 2003. The human raphe nuclei and the serotonergic system. J. Chem. Neuroanat. 26, 331–343. Johnston, M.V., Trescher, W.H., Ishida, A., Nakajima, W., 2001. Neurobiology of hypoxic–ischemic injury in the developing brain. Pediatr. Res. 49, 735–741. Krakovsky, G., Huth, M.M., Lin, L., Levin, R.S., 2007. Functional changes in children, adolescents, and young adults with cerebral palsy. Res. Dev. Disabil. 28, 331–340. Leech, R.W., Brumback, R.A., 1988. Massive brain stem necrosis in the human neonate: presentation of three cases with review of the literature. J. Child. Neurol. 3, 258–262.
BR A I N R ES E A RC H 1 3 2 2 ( 2 01 0 ) 1 2 4 – 1 33
Leonardo, C.C., Pennypacker, K.R., 2009. Neuroinflammation and MMPs: potential therapeutic targets in neonatal hypoxic–ischemic injury. J. Neuroinflam. 6, 13. Lidov, H.G.W., Molliver, M.E., 1982. An immunohistochemical study of serotonin neuron development in the rat: ascending pathways and terminal fields. Brain Res. Bull. 8, 389–430. Loewy, A.D., McKellar, S., 1981. Serotonergic projections from the ventral medulla to the intermediolateral cell column in the rat. Brain Res. 211, 146–152. Michelsen, K.A., Prickaerts, J., Steinbusch, H.W., 2008. The dorsal raphe nucleus and serotonin: implications for neuroplasticity linked to major depression and Alzheimer's disease. Prog. Brain Res. 172, 233–264. Montanez, S., Daws, L.C., Gould, G.G., Frazer, A., 2003. Serotonin (5-HT) transporter (SERT) function after graded destruction of serotonergic neurons. J. Neurochem. 87, 861–867. Nemeroff, C.B., Owens, M.J., 2009. The role of serotonin in the pathophysiology of depression: as important as ever. Clin. Chem. 55, 1578–1579. Pardo, C.A., Eberhart, C.G., 2007. The neurobiology of autism. Brain Pathol. 17, 434–447. Paterson, D.S., Trachtenberg, F.L., Thompson, E.G., Belliveau, R.A., Beggs, A.H., Darnall, R., Chadwick, A.E., Krous, H.F., Kinney, H.C., 2006. Multiple serotonergic brainstem abnormalities in sudden infant death syndrome. J.A.M.A. 296, 2124–2132. Paxinos, G., Watson, C., 1997. The Rat Brain in Stereotaxic Coordinates. Academic Press, San Diego. Peng, J.-H.F., Feng, Y., LeBlanc, M.H., Rhodes, P.G., Parker, J.J.C., 2005. Apoptosis and necrosis in developing cerebellum and brainstem induced after focal cerebral hypoxic–ischemic injury. Brain Res. Dev. Brain Res. 156, 87–92. Peters, B., Walka, M.M., Friedmann, W., Stoltenburg-Didinger, G., Obladen, M., 2000. Hypoxic–ischemic encephalopathy with cystic brain stem necroses and thalamic calcifications in a preterm twin. Brain Dev. 22, 265–271. Qian, Y., Melikian, H.E., Rye, D.B., Levey, A.I., Blakely, R.D., 1995. Identification and characterization of antidepressant-sensitive serotonin transporter proteins using site-specific antibodies. J. Neurosci. 15, 1261–1274. Rone, L.A., Ferrando, S.J., 1996. Serotonin reuptake inhibitorrelated extrapyramidal side effects in two patients with cerebral palsy. Psychosomatics 37, 165–166. Shirashi, M., Takizawa, Y., Ide, S., Obonai, T., Goto, Y., Itoh, M., 2008. Brainstem monoamine pathology of neonatal hypoxic–ischemic brain damage: a model of acute stage of neonatal asphyxia. Brain Res. 1213, 120–126.
133
Sizonenko, S.V., Sirimanne, E., Mayall, Y., Gluckman, P.D., Inder, T., Williams, C., 2003. Selective cortical alteration after hypoxic–ischemic injury in the very immature rat brain. Pediatr. Res. 54, 263–269. Stadlin, A., James, A., Fiscus, R., Wong, Y.F., Rogers, M., Haines, C., 2003. Development of a postnatal 3-day-old rat model of mild hypoxic–ischemic brain injury. Brain Res. 993, 101–110. Stengler-Wenzke, K., Muller, U., Angermeyer, M.C., Sabri, O., Hesse, S., 2004. Reduced serotonin transporter-availability in obsessive–compulsive disorder (OCD). Eur. Arch. Psychiatry Clin. Neurosci. 254, 252–255. Tai, W.C., Burke, K.A., Dominguez, J.F., Gundamraj, L., Turman Jr., J.E., 2009. Growth deficits in a postnatal day 3 rat model of hypoxic–ischemic brain injury. Behav. Brain Res. 202, 40–49. Takeuchi, Y., Fujiwara, K., Ishimura, K., Aoki, S., Okano, S., Yamazoe, I., Yoshioka, H., Sawada, T., 1992. Effects of neonatal asphyxia on the serotonin neuron system in the developing brain studied by immunohistochemistry. Dev. Neurosci. 14, 394–402. Tanaka, H., Amamiya, S., Miura, N., Araki, A., Ohinata, J., Fujieda, K., 2006. Postnatal development of brainstem serotonin-containing neurons projecting to lumbar spinal cord in rats. Brain Dev. 28, 586–591. Tarazi, F.I., Tomasini, E.C., Baldessarini, R.J., 1998. Postnatal development of dopamine and serotonin transporters in rat caudate–putamen and nucleus accumbens septi. Neurosci. Lett. 254, 21–24. Tomimatsu, T., Fukuda, H., Endoh, M., Mu, J., Watanabe, N., Kohzuki, M., Fujii, E., Kanzaki, T., Oshima, K., Doi, K., Kubo, T., Murata, Y., 2002. Effects of neonatal hypoxic–ischemic brain injury on skilled motor tasks and brainstem function in adult rats. Brain Res. 926, 108–117. Towfighi, J., Mauger, D., Vannucci, R.C., Vannucci, S.J., 1997. Influence of age on the cerebral lesions in an immature rat model of cerebral hypoxia–ischemia: a light microscopic study. Brain. Res. Dev. Brain Res. 100, 149–160. Vannucci, R.C., Lyons, D.T., Vasta, F., 1988. Regional cerebral blood flow during hypoxia–ischemia in immature rats. Stroke 19, 245–250. Volpe, J.J., 2001. Neurobiology of periventricular leukomalacia in the premature infant. Pediatr. Res. 50, 553–562. Volpe, J.J., 2009. Brain injury in premature infants: a complex amalgam of destructive and developmental disturbances. Lancet Neurol. 8, 110–124. Wixey, J.A., Reinebrant, H.E., Carty, M.L., Buller, K.M., 2009. Delayed P2X4R expression after hypoxia–ischemia is associated with microglia in the immature rat brain. J. Neuroimmunol. 212, 35–43.