Exp Brain Res (1997) 115:105–115

© Springer-Verlag 1997

R E S E A R C H A RT I C L E

&roles:Elisabetta Dell′Anna · Yong Chen Ephrem Engidawork · Kurt Andersson · Gert Lubec Johan Luthman · Mario Herrera-Marschitz

Delayed neuronal death following perinatal asphyxia in rat

&misc:Received: 16 August 1996 / Accepted: 6 December 1996

&p.1:Abstract The consequences of perinatal asphyxia on the rat brain were studied 80 min to 8 days after birth with hematoxylin-eosin and in situ DNA double-strandbreaks labeling histochemistry. Asphyxia was induced by immersing fetus-containing uterus horns, removed from ready-to-deliver Sprague-Dawley rats, in a water bath at 37°C for various time periods (0–22 min). Spontaneousand cesarean-delivered pups were used as controls. Perinatal asphyxia led to a decrease in the rate of survival, depending upon the length of the insult. No gross morphological changes could be seen in the brain of either control or asphyctic pups at any of the studied time points after delivery. However, in all groups, nuclear chromatin fragmentation, corresponding to in situ detection of DNA fragmentation, was observed at different stages. Nuclear fragmentation in control pups showed a specific distribution that appeared to be related to brain maturation, thus indicating programmed cell death. A progressive and delayed increase in nuclear fragmentation was found in asphyctic pups, which was dependent upon the length of the perinatal insult. The most evident effect was seen in frontal cortex, striatum, and cerebellum at postnatal day 8, although changes were also found in ventral-posterior thalamus, at days 1 and 2. Thus, nuE. Dell’Anna · Y. Chen · E. Engidawork · M. Herrera-Marschitz (✉) Department of Physiology and Pharmacology, Karolinska Institutet, S-17177 Stockholm, Sweden; Fax: +46–8 324969, e-mail: [email protected] Y. Chen · K. Andersson Department of Internal Medicine, Karolinska Institute, S-17177 Stockholm, Sweden E. Dell’Anna Institute of Neurology, Department of Experimental and Clinical Pathology and Medicine, University of Udine, I-33100 Udine, Italy G. Lubec Department of Paediatrics, University of Vienna, A-1090 Vienna, Austria J. Luthman Behavioral and Biochemical Pharmacology, Astra Arcus AB, S-15185 Södertälje, Sweden&/fn-block:

clear chromatin fragmentation in asphyctic pups indicates a delayed post-asphyctic neuronal death. The absence of signs of inflammation or necrosis suggests that delayed neuronal cell death following perinatal asphyxia is an active, apoptosis-like phenomenon. &kwd:Key words Perinatal asphyxia · Apoptosis · Necrosis · Hematoxylin-eosin · DNA fragmentation · Rat&bdy:

Introduction Asphyxia can be a complication of birth that, when not resulting in death, is responsible for, or a concurrent factor in, behavioral and neurological deficits with onset at different stages of life. Mental retardation, epilepsy, cerebral palsy, and dystonia may develop soon after severe perinatal asphyctic insults, while learning disabilities and attention deficits may first become evident at school age after mild birth anoxia (Amiel-Tyson and Ellison 1986; Younkin 1992). Thus, the characteristics of the neurological deficits following perinatal asphyxia seem to depend upon severity of the insult and the brain region predominantly affected. Therefore, the relationship between degree of perinatal asphyxia and brain damage has to be further investigated. In our laboratory, an experimental model of perinatal asphyxia in rats has been developed, consisting of the exposure of fetus-containing uterus horns to different asphyctic periods (Andersson et al. 1992; Bjelke et al. 1991; Herrera-Marschitz et al. 1993). The rate of survival decreases with increasing length of asphyxia (Herrera-Marschitz et al. 1993). Also, depending upon the length of the insult, levels of amino acids and metabolism products, and neuronal Fos expression are changed during the immediate postasphyctic period (Dell’Anna et al. 1995a). Furthermore, long-term changes in basal ganglia monoamine systems and motor behavior have been observed in animals surviving perinatal asphyxia (Andersson et al. 1992, 1995; Bjelke et al. 1991; Chen et al. 1995; Herrera-Marschitz et al. 1994; Loidl et al. 1994). Thus, this model seems to be

106

suitable for investigating in detail the relationship between severity of perinatal asphyxia and brain damage. Pathological changes at cellular level may occur either immediately, or with some delay, following hypoxia or ischemia (Ferrer et al. 1994; Kihara et al. 1994; MacManus et al. 1993; Okamoto et al. 1993). These changes can be detected with histological techniques such as hematoxylin-eosin staining (Auer et al. 1984; Ferrer et al. 1994; Kihara et al. 1994) or histochemical labeling of nuclear DNA fragmentation (Gavrieli et al. 1992; Kihara et al. 1994). Thus, in the present study, morphological changes in various brain regions were analyzed 80 min to 8 days following perinatal asphyxia, by using semiquantitative hematoxylin-eosin staining and in situ DNA double-strand-breaks labeling histochemistry.

Materials and methods Animals and induction of perinatal asphyxia Asphyxia was induced in pups delivered by cesarean section on Sprague-Dawley rats. Within the last day of gestation, assessed by estabularium protocols and clinical palpation, the rats were anesthetized with ether and hysterectomized. The uterus horns, still containing the fetuses, were taken out and placed in a water bath at 37°C for various periods (2–22 min) of time. Cesarean-delivered control and asphyctic pups were obtained from the same mother, since each rat delivered approximately 10–14 pups. Following asphyxia, the uterus horns were rapidly opened and the pups removed. They were stimulated to breath by cleaning the amniotic fluid and by tactile stimulation of the oral region with pieces of medical wipes. The umbilical cord was ligated and the animals were left to recover on a heating pad. Rats spontaneously delivering during the course of the experiments were used as surrogate mothers and some of their pups as spontaneously delivered controls. Only litters with pups weighing an average of 5.5 g at the time of delivery were included in the experiments (see Table 1). The experiments reported in this study were approved by the Local Ethical Committee (Stockholms Norra Djurförsöksetiska Nämnd). Behavioral and neurological assessment Survival was assessed during an 80-min postasphyctic period. At the 40- to 80-min period after delivery, color of the skin, respiratory frequency, presence of gasping, vocalization, and spontaneous movements were examined in surviving pups in order to evaluate respiratory, cardiovascular, and neurological status (see Table 1). A general neurological assessment was also performed before perfusion for neuroanatomical studies (1–8 days after birth). Hematoxylin-eosin histology Twenty-four hours (day 1), 2 days, 4 days, and 8 days after delivery, asphyctic and control animals were anesthetized (Mebumal 50 mg/kg i.p.), and perfused transcardially with 30–50 ml of 0.1 M pH 7.4 phosphate-buffered saline (PBS) containing 4% paraformaldehyde. Approximately 50–60 min after perfusion, the brains were removed, postfixed in the same solution for 12–18 h, and then kept in PBS containing 20% sucrose at +4°C until sectioning. Control and asphyctic pups studied 80 min after delivery were not transcardially perfused, but killed by decapitation, then the brain was immediately removed and immersed in 4% buffered paraformaldehyde for 12–18 h. A progressive number was given to each brain processed, which corresponded to the order of perfusion/dissection. Thus, the

analyzers were kept blind to the perinatal treatment of the animal during preparation of histochemical sections and under microscopical observations. Sagittal 20-µm serial sections of the left hemisphere were cut on a cryostat, mounted on gelatin-coated slides, and stored at –20°C until analysis. Every third section was processed for hematoxylin-eosin staining (Bancroft and Cook 1984). In situ DNA double-strand breaks labeling histochemistry Adjacent sections were processed for in situ DNA double-strandbreaks labeling with the ApopTag kit (in situ apoptosis detection kit-peroxidase; Oncor, Gaithesburg, Md., USA), based on Gavrieli and coworkers’ procedure (Gavrieli et al. 1992). Briefly, endogenous peroxidases were inhibited by incubating the sections with 2% hydrogen peroxide for 5 min at room temperature. Sections were then rinsed in PBS and immersed in terminal deoxynucleotydil transferase (TdT) buffer (supplied ready for use with the ApopTag kit, as well as the TdT enzyme, stop-wash buffer, and antidigoxigenin peroxidase) for 10–15 min at room temperature, and in TdT enzyme for 60 min in a humidified chamber at 37°C, following the instructions of the supplier. The reaction was stopped by transferring the sections to the stop-wash buffer for 30 min at 37°C. The sections were rinsed in PBS and incubated with antidigoxigenin peroxidase for 30 min at room temperature. After rinsing in PBS, color development was obtained with 0.05% diaminobenzidine-hydrogen peroxide reaction. Control procedures consisted in pretreatment with DNAase I (positive control) or by omitting TdT enzyme (negative control). Quantification Sections stained for hematoxylin-eosin and in situ DNA doublestrand-breaks labeling were analyzed under light microscopy. Hematoxylin-eosin-stained sagittal sections corresponding to lateral levels 2.4, 3.4, 3.9, and 4.2 of the Paxinos and Watson atlas (1986) were analyzed at a ×40 magnification to detect cellular alterations. Two to three sections at each level were analyzed for each animal. The cellular alterations observed in hematoxylin-eosin-stained sections consisted of nuclear fragmentation, or punctate chromatin condensation, associated with eosinophilic cytoplasm, without evidence of loss of cell membrane integrity, vacuolization, incrustations, or neighboring inflammation. Such aspects are considered as typical of apoptosis (see Bonfoco et al. 1995; CharriautMarlangue et al. 1996). The distribution of cells showing the above-indicated changes was recorded by camera lucida drawings (Fig. 1). The number of cells with a fragmented nucleus for each brain region was determined by averaging first the measurements obtained from the sections taken at the same level and also by averaging the means obtained from the various levels. A preliminary observation under dark-field illumination was necessary to clearly identify the various brain regions observed in sections processed for in situ DNA double-strand-breaks labeling, since no counterstaining was performed. The presence and distribution of positive staining was analyzed at a ×40 magnification. Positive nuclei were easily identified by their intense and selective brown staining, as a result of the peroxidase reaction (see Fig. 3b,d). No quantitative analysis was performed on this material, but it was used to confirm nuclear fragmentation found in the adjacent hematoxylin-eosin sections. A one-way analysis of variance (ANOVA) was used for statistical analysis. Post hoc comparisons were done, when appropriate, with Fisher’s protected least-significant difference (LSD) test at a significance level of P<0.05, for the two-tailed test.

Results As shown in Table 1, a decrease in the rate of survival was seen as the length of perinatal asphyxia was in-

107 Table 1 Short-term effects of perinatal asphyxia performed at 37°C, monitored by direct observation at the 40- to 80-min period following delivery. Respiratory frequency is per minute, from several samples taken during the 40- to 80-min period. Spontaneous movements were scored by the following scale during the 40- to 80-min period following delivery: 0, akinesia and rigidity (mainly

on the hind legs); 1, single movements of front legs, hind legs, or head alone; 2, movements of two of the previous indicated body structures; 3, movements of all body structures; 4, intensive movements shown by wriggling. Gasping refers to an effort to maintain respiration by opening of the mouth and movements of the diaphragm&/tbl.c:&

Experimental condition

Body weight (g)

Survival (%)

Respiratory Gasping (%) frequency

Color of skin

Vocalization (%)

Spontaneous movements

Spontaneous delivery (n = 14) Cesarean delivery (n = 13) 2–3 min asphyxia (n = 12) 5–6 min asphyxia (n = 12) 10–11 min asphyxia (n = 14) 15–16 min asphyxia (n = 21) 19–20 min asphyxia (n = 32) 20–21 min asphyxia (n = 29) 21–22 min asphyxia (n = 25)

5.5±0.2 5.4±0.2 5.5±0.3 5.3±0.2 5.9±0.2 6.0±0.2 5.5±0.1 5.2±0.2 5.3±0.1

100 100 100 100 100 100 64 40 5

67±2 65±2 65±6 62±3 60±2 51±4 31±5 27±5 21±5

Pink Pink Pink Pink Pink Pink/pale Pale Pale Pale

100 100 100 100 100 69 0 0 0

4±0 3.8±0.2 3.8±0.5 3.5±0.2 3.4±0.1 2.5±0.3 0.4±0.2 0±0 0±0

0 0 0 0 0 0 50 90 95

&/tbl.: Fig. 1 Reconstruction of camera lucida drawings showing the distribution of neurons with nuclear fragmentation in various brain regions of a cesareandelivered control pup at postnatal day 1. Two sections, corresponding to lateral 2.4-mm and 3.4-mm levels, according to the Paxinos and Watson (1986) atlas, are shown as an example. Various cortical areas (P para-, presubiculum, Fr frontal, Pir piriform, Occ occipital), subcortical regions (VP ventralposterior thalamus, Hip hippocampus, Amy amygdala, Str striatum) and cerebellum (Cb) are indicated. Each dot represents one neuron with a fragmented nucleus&ig.c:/f

creased. During the immediate post-delivery period (0–80 min), the effect of asphyxia was evident on breathing, motility, color of skin, and vocalization. Spontaneous- and cesarean-delivered controls started regular breathing immediately after birth and showed a pink-colored skin, intense vocalization, and motility. Asphyctic pups recovered very slowly. Pups exposed to more than 16 min asphyxia were akinetic and did not vocalize, even 80 min after delivery. Furthermore, they showed gasping, cyanosis, tremor, and sporadic clonic movements. Figure 2 shows body weights of control and asphyctic pups 80 min to 8 days after delivery. Surviving asphyctic pups increased body weight in a manner similar to control pups. However, some neurological symptoms such as slight tremor and rigidity affecting the hind legs could be assessed in pups exposed to severe asphyxia (more than 19 min) at postnatal days 4 and 8. No gross morphological changes could be seen in the brain of asphyctic pups killed at different time points

(80 min to 8 days) after delivery. However, in hematoxylin-eosin-stained sections, punctate chromatin condensation and, mostly, irregular fragmentation of nuclear chromatin, sometimes associated with eosinophilic cytoplasm, were observed at different stages in control and asphyctic pups. No signs of inflammation, or other neuronal alterations such as loss of cell membrane integrity, shrinkage, vacuolization, or incrustations could be detected in any group. In the regions where nuclear fragmentation was identified by hematoxylin-eosin staining, specific labeling was observed in adjacent sections stained for in situ double-strand breaks, indicating DNA fragmentation (Fig. 3b,d). Histological changes in spontaneousand cesarean-delivered (control) pups In spontaneous- and cesarean-delivered pups, neurons with nuclear chromatin fragmentation were seen only in

108 Table 2 Neurons showing nuclear fragmentation in various cortical areas of spontaneous- and cesarean-delivered controls and perinatal asphyctic pups 1–8 days after delivery. Values are expressed as mean±SEM&/tbl.c:& Experimental conditions Spontaneous delivery (n = 5–16) Cesarean delivery (n = 5–9) 2–3 min asphyxia (n = 5–11) 5–6 min asphyxia (n = 3–8) 10–11 min asphyxia (n = 5–11) 15–16 min asphyxia (n = 2–12) 19–20 min asphyxia (n = 4–11) 20–21 min asphyxia (N = 3–7) 21–22 min asphyxia (N = 2–5)

Frontal cortex

Para- and presubiculum

Piriform cortex

Day 1 Day 2 Day 4 Day 8

Day 1

Day 2 Day 4 Day 8

Day 1 Day 2 Day 4 Day 8

1±0 0±0 0±0 0±0 1±0 0±0 0±0 0±0 0±0

61±13 63±5 62±12 46±14 33±16 44±15 50±7 44±3 60±8

26±6 29±4 36±8 51±6 34±16 47±8 42±10 35±7 25±1

0±0 1±0 0±0 1±0 1±1 0±0 0±0 1±1 0±0

0±0 2±1 0±0 0±0 1±0 0±0 0±0 0±0 0±0

1±0 2±2 1±0 (0) 1±0 0±0 2±1 0±0 1±1

3±1 4±2 8±1 8±3 18±3* 18±2* 18±3* 22±5* 18±5*

11±2 11±1 13±3 11a 15±2 11±2 10±2 12±1 10±2

1±1 1±0 1±1 0±0 0±0 1±1 2±1 2±1 2±1

1±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0

0±0 1±1 0±0 0±0 0±0 0±0 0±0 0±0 0±0

0±0 0±0 1±0 1±0 1±0 1±1 1±0 1±0 1±1

* P<0.05 compared with: controls; 2- to 3-min asphyxia; 5- to 6-min asphyxia a n=1 &/tbl.:

certain regions, such as para- and pre-subiculum, hippocampus, thalamus, striatum, amygdala, and cerebellum (Fig. 1). Mesencephalic structures were also evaluated, but no nuclear fragmentation could be detected. The distribution of fragmented nuclei in pups killed 80 min after delivery was similar to that seen at postnatal day 1 (data not shown). No differences between spontaneous- and cesarean-delivered pups were seen at any developmental stage (Tables 2–5). One day after delivery, numerous fragmented nuclei were found in para- and pre-subiculum (Fig. 3A, Table 2), ventral-posterior thalamic nuclei (Fig. 3C, Table 3) and amygdala (Fig. 3E, Table 4), while only one to four fragmented nuclei were present in striatum, hippocampus, cerebellum, and various cortical areas (Fig. 3G,H, Tables 2–5). At postnatal day 2, a discrete number of fragmented nuclei was still present in para- and presubiculum (Table

2) and in ventral-posterior thalamic nuclei (Table 3), while one to four fragmented nuclei were present in frontal cortex, striatum, hippocampus, amygdala, and cerebellum (Tables 2–5). At postnatal day 4, 5–13 neurons showing nuclear fragmentation were found in para- and presubiculum (Table 2), thalamus (Table 3), and cerebellum (Table 5), while one to four fragmented nuclei were observed in the other brain regions (Tables 2–5). At postnatal day 8, only one to five neurons showing nuclear fragmentation could be seen in the various cortical (Fig. 4A, Table 2) and subcortical (Fig. 4C, Tables 3, 4) regions. On the other hand, abundant chromatin fragmentation was observed in cerebellum (Fig. 4E, Table 5). Histological changes in pups exposed to perinatal asphyxia In asphyctic pups, neurons with fragmented nuclei were seen in the same brain regions as in controls. A progressive increase in the number of cells showing chromatin fragmentation was, however, observed with increasing ▲

Fig. 2 Increase in body weight following delivery in spontaneouscesarean-delivered, and asphyctic pups. Values are expressed as mean±SEM. Circle, spontaneous delivery; square, cesarean delivery; shadow triangle, 2- to 3-min perinatal asphyxia; inverted triangle, 5- to 6-min asphyxia; rhomboid, 10- to 11-min perinatal asphyxia; hexagon, 15- to 16-min perinatal asphyxia; shadow circle, 19- to 20-min perinatal asphyxia; shadow square, 20- to 21-min perinatal asphyxia; triangle, 21- to 22-min perinatal asphyxia (D delivery time)&ig.c:/f

Fig. 3A–H Microphotograph of neurons showing nuclear fragmentation (arrows), as detected with hematoxylin-eosin staining (A–H) and in situ DNA fragmentation labeling (b,d), in various brain regions at postnatal day 1, in control (A, C, E, G, H) and perinatal asphyxia-exposed pups (B, D, E). In A parasubiculum (cesarean-delivered pup) and B presubiculum (15- to 16-min perinatal asphyctic pup), numerous fragmented nuclei are seen in all groups. In adjacent sections, in situ DNA fragmentation labeling is also observed (b). Abundant chromatin fragmentation is observed at this age in ventral-posterior thalamus in C control (cesarean-delivered pup) and D asphyctic (20- to 21-min perinatal asphyctic pup with hematoxylin-eosin staining; d, 20- to 21-min perinatal asphyctic pup; in situ DNA fragmentation labeling) pups. An increase in the number of neurons showing nuclear fragmentation is found in ventral-posterior thalamus following perinatal asphyxia. Some fragmented nuclei are also observed in E amygdala (cesarean-delivered pup), F hippocampus (15- to 16-min perinatal asphyctic pup), and G cerebellum (cesarean-delivered pup), of control and asphyctic pups. In the striatum, fragmented nuclei were very rarely observed at H postnatal day 1 (cesarean-delivered pup). Primary magnification ×40. Scale bar 100 µm&ig.c:/f

109

110 Table 3 Nuclear fragmentation in neurons of thalamus ventral-posterior and striatum in spontaneous- and cesarean-delivered controls and perinatal asphyctic pups 1–8 days after delivery. Values are expressed as mean±SEM&/tbl.c:& Experimental conditions Spontaneous delivery (n = 5–16) Cesarean delivery (n = 5–9) 2–3 min asphyxia (n = 5–11) 5–6 min asphyxia (n = 3–8) 10–11 min asphyxia (n = 5–11) 15–16 min asphyxia (n = 2–12) 19–20 min asphyxia (n = 4–11) 20–21 min asphyxia (n = 3–7) 21–22 min asphyxia (n = 2–5)

Thalamus ventral-posterior

Striatum

Day 1

Day 2

Day 4

Day 8

Day 1

Day 2

Day 4

Day 8

11±5 14±3 25±4* 24±5* 25±9* 32±4* 36±4* 28±5* 30±2*

8±1 9±1 9±1 12±3 11±2 11±2 21±1*** 25±3*** 22±2***

5±1 7±2 6±1 7±2 8±1 8±2 8±1 6±2 5±2

0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0 0±0

2±2 3±1 3±1 1±0 1±1 3±2 2±1 1±1 2±1

2±1 2±1 1±1 6±2 4±3 2±2 3±1 2±1 2±1

3±1 3±1 3±2 4±2 3±1 3±2 2±1 2±1 3±1

2±1 2±2 2±1 4±1 8±2** 10±2** 9±1** 13±1*** 17±3***

*

P<0.05 compared with controls. ** P<0.05 compared with: controls; 2- to 3-min asphyxia. *** P<0.05 compared with: controls; 2- to 3-min asphyxia; 5- to 6-min asphyxia &/tbl.: Table 4 Nuclear fragmentation in neurons of hippocampus and amygdala in spontaneous- and cesarean-delivered controls and perinatal asphyctic pups 1–8 days after delivery. Values are expressed as mean±SEM&/tbl.c:&

Experimental conditions Spontaneous delivery (n = 5–16) Cesarean delivery (n = 5–9) 2–3 min asphyxia (n = 5–11) 5–6 min asphyxia (n = 3–8) 10–11 min asphyxia (n = 5–11) 15–16 min asphyxia (n = 2–12) 19–20 min asphyxia (n = 4–11) 20–21 min asphyxia (n = 3–7) 21–22 min asphyxia (n = 2–5)

Hippocampus

Amygdala

Day 1

Day 2

Day 4

Day 8

Day 1

Day 2

Day 4

Day 8

3±1 1±0 0±0 0±0 0±0 2±1 1±0 2±1 2±2

1±0 1±0 0±0 0±0 2±1 2±1 2±0 1±1 2±1

2±1 3±1 3±1 3±1 2±0 3±1 3±1 3±2 3±2

1±0 2±1 1±0 1±0 2±1 1±1 1±1 3±2 0±0

12±3 13±2 21±6 10±4 14±2 13±5 12±2 8±3 10±1

4±0 4±1 3±1 4±1 8±4 4±1 6±1 3±2 5±2

4±1 4±1 6±1 4±2 4±1 5±1 5±1 5±1 5±2

0±0 1±1 2±0 1±0 1±0 1±1 2±1 3±2 0±0

&/tbl.: Table 5 Number of neurons showing nuclear fragmentation in the cerebellum of spontaneous- and cesarean-delivered controls and perinatal asphyctic pups 1–8 days after delivery. Values are expressed as mean±SEM&/tbl.c:& * P<0.05 compared with: controls; 2- to 3-min asphyxia. ** P<0.05 compared with: controls; 2- to 3-min asphyxia; 5- to 6-min asphyxia; a n=1 &/tbl.:

Experimental conditions Spontaneous delivery (n = 5–16) Cesarean delivery (n = 5–9) 2–3 min asphyxia (n = 5–11) 5–6 min asphyxia (n = 3–5) 10–11 min asphyxia (n = 5–11) 15–16 min asphyxia (n = 2–12) 19–20 min asphyxia (n = 4–11) 20–21 min asphyxia (n = 3–7) 21–22 min asphyxia (n = 2–5)

length of perinatal asphyxia in certain, but not all, brain structures (Figs. 3C,D, 4, Tables 2–5). No differences between control and asphyctic pups were seen at 80 min after delivery (data not shown). At day 1, an approximately twofold increase in the number of neurons showing chromatin fragmentation in the ventral-posterior thalamic nuclei was seen in asphyctic, as compared to spontaneous- and cesarean-delivered control pups (Fig. 3C,D; Table 3). No changes were observed in other brain regions (Tables 2–5). At postnatal day 2, the number of fragmented nuclei in ventral-posterior thalamic nuclei of pups exposed to 19 min or more perinatal asphyxia was still increased,

Cerebellum Day 1

Day 2

Day 4

Day 8

4±2 1±1 5±2 2±2 5±2 2±1 2±1 3±1 4±2

3±2 5±0 4±2 0±0 3±1 2a 8±3 6±1 1±1

7±3 10±2 6±1 7±2 9±2 10±2 8±2 6±2 8±2

29±12 26±18 23±2 54±16 71±12* 86±5* 81±4* 112±1** 94±2*

but not following shorter asphyctic periods (Table 3). No changes were observed in other regions (Tables 2–5). No differences between asphyctic and control pups could be seen at postnatal day 4 in any region (Tables 2–5). At postnatal day 8, very few, if any, neurons with fragmented nuclei could be observed in para- and presubiculum, piriform cortex, thalamus, and amygdala of asphyctic pups (Tables 2–4). However, an increase in the number of neurons with chromatin fragmentation was found in frontal cortex, striatum, and cerebellum (Figs. 4, 5, Tables 2–5). In frontal cortex, a light increment in chromatin fragmentation was observed after 2- to 6-min perinatal asphyxia, compared with the controls, while an

111

Fig. 4 Microphotograph of neurons showing nuclear fragmentation (arrows), as detected with hematoxylin-eosin staining, at postnatal day 8 in cesarean-delivered control (A, C, E) and perinatal asphyxia exposed pups (B, D, F). Frontal cortex (A, cesareandelivered pup; B, 15- to 16-min perinatal asphyctic pup); striatum (C, cesarean-delivered control pup; D, 20- to 21-min perinatal asphyctic pup), and cerebellum (E, cesarean-delivered control pup; F, 20- to 21-min perinatal asphyctic pup) are shown to evidence the increase in the number of neurons with a fragmented nucleus following perinatal asphyxia. Primary magnification ×40. Scale bar 100 µm&ig.c:/f

served. The number of neurons with fragmented nuclei increased approximately threefold after 10- to 20-min perinatal asphyxia, and more than sixfold after asphyctic periods longer than 20 min (Fig. 4C,D, Table 3). In addition, following 10 min or more of perinatal asphyxia, a progressive increase in chromatin fragmentation was found in the cerebellum (Fig. 4E,F, Table 5).

Discussion approximately threefold increase in the number of fragmented nuclei was found after perinatal asphyctic periods longer than 10 min (Fig. 4A,B, Table 2). In the striatum, a progressive increase in chromatin fragmentation along with the length of perinatal asphyxia was ob-

Morphological changes were studied in the rat brain 80 min to 8 days after birth, in spontaneous- and cesarean-delivered control and perinatal asphyctic pups. As previously reported (Andersson et al. 1992; Bjelke et al. 1991; Herrera-Marschitz et al. 1993), perinatal asphyxia

112

Fig. 5 Perinatal asphyxia-induced changes in the number of neurons showing nuclear fragmentation in frontal cortex (A), striatum (B), and cerebellum (C) at postnatal day 8. Values are expressed as mean±SEM. a, P<0.05 versus spontaneous- and cesarean-delivered pups; b, P<0.05 versus 2- to 3-min perinatal asphyxia; c, P<0.05 versus 5- to 6-min perinatal asphyxia&ig.c:/f

leads to a decrease in neonatal survival, which is strictly related to the length of the insult and to the temperature at which it occurs (Herrera-Marschitz et al. 1993, 1994). At 37°C, the rate of survival rapidly decreases following exposure to perinatal asphyxia longer than 16 min, and no survival is observed following asphyctic periods longer than 22 min. Several parameters related to respiratory and cardiovascular functions were evaluated during the first 80 min of the postnatal period. Following perinatal asphyxia, there was a decrease in respiratory frequency, an increas-

ing occurrence of gasping-like breathing, and a change in the color of the skin, from a pink to a pale, cyanoticlike color. These changes were accompanied by a decrease in motility and vocalization, indicating a general compromise of the central nervous system (CNS), as previously demonstrated by concomitant modifications in neuronal Fos expression (Dell’Anna et al. 1995a). During the postnatal period, pups surviving perinatal asphyxia are normally accepted by surrogate mothers and can develop in parallel with control rats (Andersson et al. 1992, 1995; Bjelke et al. 1991; Herrera-Marschitz et al. 1994; Loidl et al. 1994). In the present study asphyctic pups did not differ from controls in body weight 80 min to 8 days following delivery (Fig. 1), indicating a normal nurture and a normal physical development. Some neurological symptoms could, however, be assessed in very asphyctic pups at postnatal days 4 and 8, in agreement with biochemical and behavioral changes observed 4 weeks (Andersson et al. 1992, 1995; Bjelke et al. 1991, Chen et al. 1995) and 6–12 months following perinatal asphyxia (Andersson et al. 1992; Herrera-Marschitz et al. 1994; Loidl et al. 1994). Hence, the effect of perinatal asphyxia on neurobiological functions is expressed immediately after the insult and later during development, particularly following more extensive insults. Neuronal changes following perinatal asphyxia were investigated by analyzing hematoxylin-eosin staining in various brain regions and at different time-points after delivery. Hematoxylin-eosin staining is a useful histological technique for detection of cellular damage (Auer et al. 1984), allowing identification of cytoplasmatic and nuclear changes such as shrinkage, swelling, karyorrhexis, pyknosis, and nuclear fragmentation (Ferrer et al. 1994; Kihara et al. 1994). Neurons showing nuclear fragmentation with hematoxylin-eosin staining also specifically stain for DNA fragmentation, when using in situ DNA double-strand-breaks labeling (CharriautMarlangue et al. 1996; Kihara et al. 1994). In the present study, no obvious gross morphological abnormalities, sign of inflammation or necrosis, or differences in ventricular dimension and in the size of various brain structures could be seen in the brain of control and/or asphyctic pups killed 80 min to 8 days after delivery, although quantitative measurements of these parameters were not performed. In parallel experimental series, in which the brains were dissected for biochemistry, no differences in whole (Engidawork et al. 1997) or regional (Andersson et al. 1995; Ungethum et al. 1996) brain weight were observed between asphyctic and control pups. Some cells, clearly identified as neurons, showing punctate chromatin condensation and irregular nuclear fragmentation, sometimes associated with eosinophilic cytoplasm, were, however, observed in various brain regions, in control and asphyctic pups. In adjacent sections, these cells were positive for in situ DNA double-strand-breaks labeling.

113

Nuclear chromatin fragmentation in control pups reflects maturation-related, programmed cell death In pups born by spontaneous- or cesarean-delivery, numerous neurons with fragmented chromatin were present in para- and presubiculum, ventral-posterior thalamic nuclei, and amygdala, 1–4 days after birth. Their number in these structures decreased progressively and, at postnatal day 8, chromatin fragmentation could be detected only sporadically. Few fragmented nuclei were seen in hippocampus, striatum, frontal and piriform cortices, but never in cortical or subcortical brain regions other than those indicated above. In the cerebellum, very few fragmented nuclei were seen at postnatal days 1–2, but were increased at day 4, and abundant chromatin fragmentation was evident at day 8. These changes largely follow maturational patterns of various brain regions. There is a delayed maturation of ventral-posterior compared to antero-lateral thalamic nuclei (Altman and Bayer 1979a,b; Minciacchi and Granato 1989), a late maturation of para- and presubiculum (Bayer and Altman 1991), and a major postnatal development of cerebellum (Merry et al. 1994). For instance, in ventral-posterior thalamic nuclei, intensive changes in terms of somatotopic organization, thalamocortical projection (Killackey et al. 1995), and neurotrophin expression (Bentivoglio et al. 1993) have been demonstrated to occur during the first postnatal days, which may be associated with nuclear modifications in individual cells. Chromatin fragmentation in these regions could therefore reflects changes related to programmed cell death, also designated as apoptosis (Bright and Khar 1994; Spreafico et al. 1995). Changes in nuclear chromatin fragmentation produced by perinatal asphyxia While no effect of perinatal asphyxia could be seen 80 min after delivery in any brain structure, at postnatal day 1 the number of neurons with chromatin fragmentation was increased in ventral-posterior thalamic nuclei in all asphyctic groups. Two days after delivery, this increase was only associated with perinatal asphyctic insults longer than 19 min. At that period, however, the amount and distribution of fragmented nuclei in cortex, striatum, hippocampus, amygdala, and cerebellum was similar, in both asphyctic and control pups (Tables 2–5). At postnatal day 4, no asphyxia-related changes could be observed in any brain region. A pronounced effect of perinatal asphyxia was observed at postnatal day 8, in frontal cortex, striatum, and cerebellum. While no changes were observed following short asphyctic periods (2–6 min), a progressive increase in the number of neurons showing nuclear fragmentation was found after asphyctic periods longer than 10 min. The most evident effect was observed following 20 min or more of perinatal asphyxia, particularly in striatum, where, as compared to controls, the num-

ber of fragmented nuclei was increased by more than sixfold. Does nuclear fragmentation following perinatal asphyxia represent programmed cell death? Nuclear fragmentation could be observed in both control and asphyctic pups, depending upon brain region and the stage of development. In asphyctic, as in control pups, nuclear fragmentation was enhanced in ventral-posterior thalamic nuclei at postnatal days 1–2 and in cerebellum at day 8. No effect of perinatal asphyxia was observed in para- and presubiculum or amygdala, where abundant chromatin fragmentation was present in control pups. In frontal cortex and striatum, however, nuclear fragmentation was only seen after severe asphyxia, with a delayed occurrence (8 days after delivery). Thus, the increment in nuclear fragmentation following perinatal asphyxia seems to reflect a phenomenon that only partially overlaps with programmed cell death. Besides programmed cell death, chromatin fragmentation can result from activation of endonucleases by “death signal” (Bright and Khar 1994; Thompson 1995) as, e.g., exposure to cytotoxic agents (Bonfoco et al. 1995), growth factor deprivation (Bright and Khar 1994), or free radicals (Thompson 1995). This process develops with a delay after the insult and is followed by death, degradation, and phagocytosis of the cells. Dying cells are then detectable for a restricted time-window of 8–11 h, before being removed by microglia (see Spreafico et al. 1995). Nuclear fragmentation has been seen in neurons 2–5 days following hypoxia-ischemia in infant and adult rodents (Charriaut-Marlangue et al. 1996; Ferrer et al. 1994; Kihara et al. 1994; MacManus et al. 1993; Okamoto et al. 1993), while other cellular alterations, such as shrinkage, swelling, membrane lysis, which are typically associated with necrosis, can occur within the first hours after the insult (Bonfoco et al. 1995; Charriaut-Marlangue et al. 1996; Ferrer et al. 1994). Furthermore, following ischemia, swollen necrotic neurons have been early detected in the ischemic core, while apoptotic nuclei might appear later at the boundary of the infarcted tissue (Charriaut-Marlangue et al. 1996). The intensity of the inducing event might favor the prevalence of each of these phenomena, leading to the hypothesis that short-lasting anoxic insults induce immediate cell necrosis, while prolonged hypoxic insults induce delayed death, with some features resembling apoptosis (Bonfoco et al. 1995; Chariaut-Marlangue et al. 1996; Schreiber and Baudry 1995). In this case, increase in intracellular calcium influx, probably mediated by sustained, high extracellular glutamate levels (Szatkowki and Attwell 1994), and subsequent activation of potentially lethal metabolic pathways, including activation of proteases and endonucleases and formation of oxygenfree radicals (Schreiber and Baudry 1995), initiate a metabolic cascade leading to delayed neuronal death (Bright and Khar 1994; Haba et al. 1991; Okamoto et al. 1993;

114

Schreiber and Baudry 1995). Hence, neuronal chromatin fragmentation seen following perinatal asphyxia can be related to delayed, postasphyctic neuronal death (Haba et al. 1991). Progressive and prolonged decrease in oxygenation induced by perinatal asphyxia In the present model of perinatal asphyxia, a progressive and prolonged decrease in oxygenation may occur even after delivery. In fact, at 80 min following delivery, respiratory frequency of the very asphyctic pups was reduced and sustained by gasping. Furthermore, the skin of these pups showed a cyanotic color. Thus, a general cardiorespiratory failure appears to occur, suggesting that the resulting brain damage is secondary to impairment in oxygenation lasting for a long period after delivery. Furthermore, we have recently found (Engidawork et al. 1997) that perinatal asphyxia produces a significant decrease in heart and brain pH, which is directly correlated to the length of the insult. Severe oxygen deprivation and persistent impairment of metabolism may thus lead to delayed neuronal death in brain regions sensitive to anoxic or ischemic conditions. Vulnerability probably depends upon the actual development stage and metabolism level of the particular region (see Beal et al. 1993; Lipton and Kater 1989). Delayed neuronal death depends upon the length of perinatal asphyxia The occurrence of delayed neuronal death appears to depend upon the length of perinatal asphyxia. A progressive increase in the number of neurons showing nuclear fragmentation could be observed following perinatal asphyctic periods longer than 10 min, and mostly 20 min. Changes in neuronal activity depending upon the length of perinatal asphyxia have been evidenced at 80 min after delivery by immunocytochemical detection of Fos protein in cortical neurons (Dell’Anna et al. 1995a). In that study, an increase in Fos was observed following short asphyctic periods (2–5 min), while a progressive decrease in Fos immunostaining occurred after prolonged asphyxia (more than 19 min), indicating an ongoing impairment in cellular functions, such as those required for gene-product expression. These early alterations could be implicated in processes related to delayed neuronal cell death. Indeed, reduction in Fos expression has been suggested to be associated with early intracellular changes leading to activation of apoptosis (see Pittman et al. 1994; Schreiber and Baudry 1995). Therefore, effects of different perinatal asphyctic periods on metabolism and pH may induce rapid changes in immediateearly gene c-fos expression, which can lead to progressive neuronal death in vulnerable brain regions.

Conclusion Perinatal asphyxia produces an extensive and regionally selective increase in nuclear chromatin fragmentation 1–8 days after delivery. The increase in nuclear fragmentation occurs in absence of signs of inflammation or other cellular alterations that could be related to necrosis, indicating that perinatal asphyxia produces a delayed apoptosis-like neuronal cell death. This probably leads to permanent alterations in discrete brain regions, reflecting, later in life, as functional deficits (see Haba et al. 1991; Dell’Anna et al. 1995b). The probability that such deficits occur increases with the severity of the perinatal insult. Perinatal asphyxia can then be directly associated with alteration of functions subserved by the affected regions or by those targeted by primarily damaged structures. Thus, it has been shown that a hypoxic-ischemic insult to the striatum leads to changes in the substantia nigra (see Oo et al. 1995). Indeed, changes in neuronal expression and neurotransmitter release in striatum and substantia nigra have been observed 1–12 months following perinatal asphyxia (Andersson et al. 1992, 1995; Bjelke et al. 1991; Chen et al. 1995; Herrera-Marschitz et al. 1994; Loidl et al. 1994). The morphological changes together with the neurochemical and behavioral alterations observed following perinatal asphyxia may have implications for the understanding of mechanisms involved in the pathogenesis of motor and cognitive deficits seen in animals and humans suffering asphyctic insults during the intrapartum period. The delayed appearance of neuronal death may perhaps provide an extended time-window for therapeutic interventions to prevent brain damage following perinatal asphyxia. &p.2:Acknowledgements The authors express their gratitude to Professor G. Macchi (Institute of Neurology, Catholic University, Rome, Italy) for his critical advice on the histopathological evaluation. The excellent technical assistance of Ms. A. Schönbeck is acknowledged. This study was supported by grants from the Swedish Medical Research Council (8669, 10797, 10362), Karolinska Institutet fonder, the Åke Wibergs Stiftelsen. C.Y. is a recipient of a Karolinska Institutet scholarship.

References Altman J, Bayer SA (1979a) Development of the diencephalon in the rat. IV. Quantitative study of the time of origin of neurons and the internuclear chronological gradients in the thalamus. J Comp Neurol 188: 455–472 Altman J, Bayer SA (1979b) Development of the diencephalon in the rat. V. Thymidine-radiographic observations on internuclear and intranuclear gradients in the thalamus. J Comp Neurol 188: 473–500 Amiel-Tyson C, Ellison P (1986) Birth asphyxia in the fullterm newborn: early assessment and outcome. Dev Med Child Neurol 28: 671–682 Andersson K, Bjelke B, Bolme P, Ögren SÖ (1992) Asphyxia-induced lesion of the rat hippocampus (CA1, CA3) and of the nigro-striatal dopamine system. In: Gross J (ed) Hypoxia and ischemia. (CNS Wissenschafliche, vol 41) Humboldt-Universitat zu Berlin, Berlin, pp 71–76

115 Andersson K, Blum M, Chen Y, Eneroth P, Gross J, Herrera-Marschitz M, Bjelke B, Bolme P, Diaz R, Jamison L, Loidl F, Ungethum U, Åstrom G, Ögren SÖ (1995) Perinatal asphyxia increases bFGF mRNA levels and DA cell body number in the mesencephalon of the rat. Neuroreport 6: 375–378 Auer RN, Wieloch T, Olsson B, Siesjö BK (1984) The distribution of hypoglycemic brain damage. Acta Neuropathol (Berl) 64: 177–191 Bancroft JD, Cook HC (1984) Manual of histological techniques. Churchill Livingstone, Edinburgh, pp 18-19 Bayer SA, Altman J (1991) Neocortical development. Raven, New York Beal MF, Hyman BT, Koroshetz W (1993) Do defects in mitochondrial energy metabolism underlie the pathology of neurodegenerative diseases? Trends Neurosci 16: 125–131 Bentivoglio M, Chen S, Peng Z-C, Bertini G, Ringstedt T, Persson H (1993) Thalamus, neurotrophins and their receptors. In: Minciacchi D, Molinari M, Macchi G, Jones EG (eds) Thalamic networks for relay and modulation. Pergamon, Oxford, pp 309–320 Bjelke B, Andersson K, Ögren SÖ, Bolme P (1991) Asphyctic lesion: proliferation of tyrosine hydroxylase immunoreactive nerve cell bodies in the rat substantia nigra and functional changes in dopamine neurotransmission. Brain Res 543: 1–9 Bonfoco E, Krainc D, Ankarcrona M, Nicotera P, Lipton SA (1995) Apoptosis and necrosis: two distinct events induced, respectively, by mild and intense insults with N-methyl-D-aspartate or nitric oxide/superoxide in cortical cell cultures. Proc Natl Acad Sci USA 92: 7162–7166 Bright J, Khar A (1994) Apoptosis: programmed cell death in health and disease. Biosci Reports 14: 67–81 Charriaut-Marlangue C, Aggon-Zouaoui D, Represa A, Ben-Ari Y (1996) Apoptotic features of selective neuronal death in ischemia, epilepsia and gp120 toxicity. Trends Neurosci 19: 109–114 Chen Y, Ögren SÖ, Bjelke B, Bolme P, Eneroth P, Gross J, Loidl F, Herrera-Marschitz M, Andersson K (1995) Nicotine treatment counteracts perinatal asphyxia-induced changes in the mesostriatal/limbic dopamine systems and in motor behavior in the four-week-old-male rat. Neuroscience 68: 531–538 Dell’Anna E, Chen Y, Loidl F, Andersson K, Luthman J, Goiny M, Rawal R, Lindgren T, Herrera-Marschitz M (1995a) Shortterm effects of perinatal asphyxia studied with fos-immunocytochemistry and in vivo microdialysis in the rat. Exp Neurol 131: 279–287 Dell’Anna E, Geloso MC, Draisci G, Luthman J (1995b) Transient changes in Fos and GFAP immunoreactivity precede neuronal cell loss in the rat hippocampus following neonatal anoxia. Exp Neurol 131:144–156 Engidawork E, Chen Y, Dell’Anna E, Goiny M, Lubec G, Ungerstedt U, Andersson K, Herrera-Marschitz M (1997) Effect of perinatal asphyxia on systemic and intracerebral pH and glycolysis metabolism in the rat. Exp Neurol (in press) Ferrer I, Tortosa A, Macaya A, Sierra A, Moreno D, Munell F, Blanco R, Squier W (1994) Evidence of nuclear DNA fragmentation following hypoxia-ischemia in the infant rat brain, and transient forebrain ischemia in the adult gerbil. Brain Pathol 4: 115–122 Gavrieli Y, Sherman Y, Ben-Sasson SA (1992) Identification of programmed cell death in situ via specific labelling of nuclear DNA fragmentation. J Cell Biol 119: 493–501 Haba K, Ogawa N, Mizukawa K, Mori A (1991) Time course of changes in lipid peroxidation, pre- and post-synaptic cholinergic indices, NMDA receptors binding and neuronal death in the gerbil hippocampus following transient ischemia. Brain Res 540: 116–122

Herrera-Marschitz M, Loidl CF, Andersson K, Ungerstedt U (1993) Prevention of mortality induced by perinatal asphyxia: hypothermia or glutamate antagonism? Amino Acids 5: 413– 419 Herrera-Marschitz M, Loidl FC, You Z-B, Andersson K, Silveira R, O’Connor WT, Goiny M (1994) Neurocircuitry of the basal ganglia studied by monitoring neurotransmitter release. Effects of intracerebral and perinatal asphyctic lesions. Mol Neurobiol 9: 171–182 Kihara S, Shiraishi T, Nakagawa S, Toda K, Tabuchi K (1994) Visualization of DNA double strand breaks in the gerbil hippocampal CA1 following transient ischemia. Neurosci Lett 175: 133–136 Killackey HP, Rhoades RW, Bennett-Clarke CA (1995) The formation of a cortical somatotopic map. Trends Neurosci 18: 402–407 Lipton SA, Kater SB (1989) Neurotransmitter regulation of neuronal outgrowth, plasticity and survival. Trends Neurosci 12: 265–270 Loidl CF, Herrera-Marschitz M, Andersson K, You Z-B, Goiny M, O’Connor WT, Silveira R, Rawal R, Bjelke B, Chen Y, Ungerstedt U (1994) Long-term effects of perinatal asphyxia on basal ganglia neurotransmitter systems studied with microdialysis in rat. Neurosci Lett 175: 9–12 MacManus JP, Buchan AM, Hill IE, Rasquinha I, Preston E (1993) Global ischemia can cause DNA fragmentation indicative of apoptosis in rat brain. Neurosci Lett 164: 89–92 Merry DE, Veis DJ, Hickey WF, Korsmeyer SJ (1994) bcl-2 Protein expression is widespread in the developing nervous system and retained in the adult PNS. Development 120: 301– 311 Minciacchi D, Granato A (1989) Development of the thalamocortical system: transient crossed projections to the frontal cortex in neonatal rats. J Comp Neurol. 281: 1–12 Okamoto M, Matsumoto M, Ohtsuki T, Taguchi A, Mikoshiba K, Yagagihara T, Kamada T (1993) Internucleosomal DNA cleavage involved in ischemia-induced neuronal death. Biochem Biophys Res Commun 196: 1356–1362 Oo TF, Henchliffe C, Burke RE (1995) Apoptosis in substantia nigra following developmental hypoxic-ischemic injury. Neuroscience 69: 893–901 Paxinos G, Watson C (1986) The rat brain in stereotaxic coordinates, 2nd edn. Academic, New York Pittman RN, Mills JC, Dibenedetto AJ, Hynicka WP, Wang S (1994) Neuronal cell death: searching for the smoking gun. Curr Opin Neurobiol 4: 87–94 Schreiber SS, Baudry M (1995) Selective neuronal vulnerability of the hippocampus – a role for gene expression? Trends Neurosci 18: 446–451 Spreafico R, Frassoni C, Arcelli P, Selvaggio M, De Biasi S (1995) In situ labeling of apoptotic cell death in the cerebral cortex and thalamus of rats during development. J Comp Neurol 363: 281–295 Szatkowski M, Attwell D (1994) Triggering and execution of neuronal death in brain ischemia: two phases of glutamate release by different mechanisms. Trends Neurosci 17: 359–365 Thompson CB (1995) Apoptosis in the pathogenesis and treatment of disease. Science 267: 1456–1462 Ungethum U, Chen Y, Gross J, Bjelke B, Bolme P, Eneroth P, Heldt J, Loidl CF, Herrera-Marschitz M, Andersson K (1996) Effects of perinatal asphyxia on the meso-striatal/-limbic dopamine system of neonatal and 4-week-old male rats. Exp Brain Res 112: 403–410 Younkin DP (1992) Hypoxic-ischemic brain injury of the newborn – statement of the problem and overview. Brain Pathol 2: 209–210