Received: 14 April 2021
Revised: 30 July 2021
Accepted: 9 August 2021
FULL- LENGTH ORIGINAL RESEARCH
Long- term outcome in a noninvasive rat model of birth
asphyxia with neonatal seizures: Cognitive impairment,
anxiety, epilepsy, and structural brain alterations
This is an open access article under the terms of the Creat ive Commo ns Attri butio n- NonCo mmerc ial- NoDerivs License, which permits use and distribution in any
medium, provided the original work is properly cited, the use is non- commercial and no modifications or adaptations are made.
© 2021 The Authors. Epilepsia published by Wiley Periodicals LLC on behalf of International League Against Epilepsy.
1Department of Pharmacology,
Toxicology, and Pharmacy, University
of Veterinary Medicine, Hannover,
2Center for Systems Neuroscience,
3Molecular and Integrative Biosciences,
University of Helsinki, Helsinki,
4Neuroscience Center (HiLIFE),
University of Helsinki, Helsinki,
Wolfgang Löscher, Department
of Pharmacology, Toxicology, and
Pharmacy, University of Veterinary
Medicine, Bünteweg 17, D- 30559
Grant/Award Number: Lo 274/15- 1
Objective: Birth asphyxia is a major cause of hypoxic– ischemic encephalopathy
(HIE) in neonates and often associated with mortality, neonatal seizures, brain
damage, and later life motor, cognitive, and behavioral impairments and epilepsy.
Preclinical studies on rodent models are needed to develop more effective thera-
pies for preventing HIE and its consequences. Thus far, the most popular rodent
models have used either exposure of intact animals to hypoxia- only, or a combi-
nation of hypoxia and carotid occlusion, for the induction of neonatal seizures
and adverse outcomes. However, such models lack systemic hypercapnia, which
is a fundamental constituent of birth asphyxia with major effects on neuronal
excitability. Here, we use a recently developed noninvasive rat model of birth
asphyxia with subsequent neonatal seizures to study later life adverse outcome.
Methods: Intermittent asphyxia was induced for 30min by exposing male and
female postnatal day 11 rat pups to three 7+3- min cycles of 9% and 5% O2 at con-
stant 20% CO2. All pups exhibited convulsive seizures after asphyxia. A set of be-
havioral tests were performed systematically over 14months following asphyxia,
that is, a large part of the rat's life span. Video- electroencephalographic (EEG)
monitoring was used to determine whether asphyxia led to the development of
epilepsy. Finally, structural brain alterations were examined.
Results: The animals showed impaired spatial learning and memory and in-
creased anxiety when tested at an age of 3– 14months. Video- EEG at ~10months
showed an abundance of spontaneous seizures, which was paralleled by neuro-
degeneration in the hippocampus and thalamus, and by aberrant mossy fiber
Significance: The present model of birth asphyxia recapitulates several of the
later life consequences associated with human HIE. This model thus allows eval-
uation of the efficacy of novel therapies designed to prevent HIE and seizures
GAILUS et al.
Birth asphyxia (BA; also known as perinatal asphyxia)
is a common cause of hypoxic– ischemic encephalopathy
(HIE), which is characterized by clinical and laboratory
evidence of acute or subacute brain injury, often associated
with neonatal seizures.1,2 BA is the most common cause of
death and disability in human neonates and often leads to
poor later life outcome, including persistent motor, sen-
sory, and cognitive impairment, behavioral alterations,
and epilepsy, characterized by spontaneous recurrent
seizures (SRS).3– 5 At least in part, these later life conse-
quences of perinatal asphyxia are thought to be a conse-
quence of asphyxia- induced brain damage. Asphyxia is a
global insult on the whole organism,6 and in the brain, it
affects some regions more than others, including the cer-
ebral cortex, hippocampus, basal ganglia, thalamus, and
brain stem nuclei.3– 5,7 Despite the high disability burden
associated with surviving neonatal HIE in patients, the
only evidence- based therapy that is available to reduce
BA- induced mortality and brain injury is moderate post-
asphyxial hypothermia.5,8 However, therapeutic hypo-
thermia can only be used to treat HIE in full- term infants,
in whom it has a protective effect in only approximately
one of six individuals, and survivors remain at high risk
for a wide spectrum of neurodevelopmental abnormali-
ties as a result of residual brain injury, highlighting the
need for additional, novel therapies to be used in conjunc-
tion.5,8There is an ongoing debate on whether asphyxia-
induced neonatal seizures may exacerbate HIE- induced
brain injury and its long- term consequences.2,9– 11Notably,
these seizures are often not sufficiently suppressed by hy-
pothermia,9 thus necessitating the development of more
effective seizure- suppressing therapies.2,10,12
To develop novel therapies, animal models of BA and
neonatal seizures that mimic the mechanisms of HIE
and its clinical short- and long- term consequences are
essential.5,13 The widely used Rice– Vanucci model of
hypoxia– ischemia is invasive, based on unilateral ligation
of one of the carotid arteries and subsequent exposure
to hypoxia.5 In another popular model, intact neonatal
rodents are exposed to hypoxia- only.14 In both types of
models, hypoxia triggers seizures that already commence
during the insult,5,13,14 whereas neonatal seizures in hu-
mans usually occur within the first 2– 24h after BA.1,15
Here, it should be noted that asphyxia is a combination
of a decrease in systemic O2 (hypoxia) and an increase
in CO2 (hypercapnia), and these two components of
asphyxia have distinct— often functionally opposite—
actions on the physiology and excitability of the neonatal
brain.16,17 Recently, we have described the first noninva-
sive rat model of BA with seizure generation after, not
during, the insult.17 In this model, intermittent asphyxia
is induced at postnatal day 11 (P11), which in terms of
cortical development corresponds to term human ba-
bies.18,19The present study aimed to evaluate (1) the long-
term outcome in this model as seen in later life motor,
behavioral, cognitive, and structural brain alterations;
and (2) whether and how these alterations resemble the
clinical situation. Furthermore, we examined whether
the animals developed epilepsy in the 14months after as-
phyxia examined in this study.
MATERIALS AND METHODS
Details on animals and all experimental techniques and
statistical methods are described in Appendix S1. In short,
the experiments were carried out in male and female
P11Wistar Han rats that were bred in our laboratory. As
in our previous experiments,20 intermittent asphyxia was
induced for 30min by exposing male and female P11 rat
pups to three 7+3- min cycles of 9% and 5% O2 at constant
following asphyxia, and of how such therapies might alleviate long- term adverse
• Birth asphyxia often leads to HIE and neonatal
seizures, which are important causes of an ad-
verse neurodevelopmental outcome
• Current therapies do not prevent this outcome
in the majority of the patients
• Rodent models of HIE and neonatal seizures
reflecting the clinical syndrome and its adverse
outcome are needed to develop more effective
• Here, we used a novel rat model of birth as-
phyxia with neonatal seizures and character-
ized later life motor, cognitive, and behavioral
• We found persistent cognitive and behavioral
impairment, epilepsy, and structural brain al-
terations resembling the outcome in human
GAILUS et al.
20% CO2. Sham asphyxia was used for comparison. As
shown in Figure 1, a set of behavioral tests was used over
14months following asphyxia, that is, a large part of the
rat's life span. The tests included a modified Irwin test;
a test of developmental motor responses; the chimney
and rotarod tests to study motor function; the adhesive-
removal test to analyze sensorimotor skills; the open field,
elevated plus maze and light– dark box tests to determine
anxiety- related behavior; and the radial- arm water maze
(RAWM) test of spatial learning and memory. Ten months
after asphyxia, video- electroencephalographic (EEG)
monitoring was used to determine whether asphyxia led
to the development of epilepsy. Finally, at the end of the
experimental period, structural brain alterations were ex-
Behavioral alterations during and
after asphyxia in P11 rats
A total of 25 P11 rat pups (12 rats for sham asphyxia and 13
rats for asphyxia) of both sexes were used for the present
experiments. Additional P11 rat pups were used for con-
trol of body temperature (see Appendix S1). As described
recently,17,20 exposing the P11 rat pups to intermittent as-
phyxia did not produce any obvious behavioral response,
apart from brief agitation on initiation of the exposure.
Seizures were never seen during asphyxia. However, in
line with previous work,17,20 following establishment of
normocapnic conditions, all rat pups exhibited convulsive
seizures, with a latency to the first seizure of approximately
1.2min on average (range = .7– 2.3min). These seizures,
which often had a focal onset, occurred repeatedly over
5.2min (range = 1.3– 8.8min) and ended 6.3min (range
= 2.2– 9.9min) after the asphyxia. Subsequently, all ani-
mals resumed apparently normal behavior. Seizures were
scored by a modified Racine scale20 as shown in Figure
S1 and described in Appendix S1. In agreement with
previous findings,17,20 the predominant seizure type was
generalized convulsive (Stage III– V) seizures seen in all
animals, whereof generalized tonic– clonic (Stage V) sei-
zures were observed in 9 of 13 rat pups (Figure S2). No
animal died during or after asphyxia, and no sex differ-
ences were observed. Rectal body temperature was simi-
lar before and after asphyxia and not different from sham
controls (Table S1).
Body weight gain after asphyxia
As shown in Figure S3, body weight gain in male and fe-
male rats did not differ between animals with asphyxia
and those with sham asphyxia.
Estrous cycle after asphyxia
Because cognitive and anxiety functions may be affected
by the female estrous cycle,21 we determined whether
the postasphyxial female rats had regular estrous cycles.
As shown in Table S2, at approximately 6 months after
asphyxia or sham asphyxia, all female rats exhibited a
regular estrous cycle without any obvious difference be-
tween sham controls and postasphyxial rats. The estrous
cycle was not synchronized among the individual animals
FIGURE Timeline of the experiments on long- term consequences of asphyxia. Approximately 3weeks following electrode
implantation at 10months after asphyxia, video- electroencephalographic (EEG) monitoring was performed for 1 week. The last behavioral
testing was done 14months after asphyxia to exclude that the anesthesia used for electrode implantation or the video- EEG monitoring
affected the outcome of the behavioral experiments
GAILUS et al.
Long- term motor and behavioral
consequences of asphyxia with
Much of the full behavioral repertoire of rats is observed
only in adolescence and adulthood.22,23Thus, in contrast
to previous studies in other rat models, which evaluated
consequences of HIE or hypoxia- only over a few days or
weeks after the insult,5,13,14 we evaluated the behavioral
consequences of asphyxia with neonatal seizures over
14months, that is, a large part of the rat's life span. As
shown in Figures 1 and S4, the two groups of rats, that
is, the sham group (n =12; sixmales, six females) and
the asphyxia group (n = 13; seven males, six females)
were repeatedly tested for behavioral alterations. Figure
S4summarizes the outcome of the various tests, which are
illustrated in Figure S5. Males and females are separately
shown in all figures, and sex differences are noted in fig-
Developmental motor responses
Tests of developmental motor responses (see Figure S5A),
which were performed only once (at 7 days after asphyxia),
were not affected in the postasphyxia group (Figure S4).
Modified Irwin test at P18 to
As shown in Figure S4, the modified Irwin test (see Figure
S5A) did not indicate any obvious intergroup differences
in general behavior at any time after asphyxia.
Chimney test at P22
In the chimney test, neurological deficit is indicated by
the inability of the animals to climb backwards through
a tube within 30s. As shown in Figure S6, all rats of all
groups were able to do this.
Rotarod at P19 to 14months
In the rotarod test, which was repeatedly performed over
14 months following asphyxia (Figure S4), not all rats
were capable (or motivated) of staying on the rotating rod
for 60s despite previous training, but there was no sig-
nificant difference between groups (Figure S7). Over the
14months after asphyxia, both groups improved their per-
formance on the rod.
Sensorimotor function (adhesive-
removal test) at P25 to 14months
In this test, neither time to first contact with the adhesive
tape (see Figure S5D) nor time from the first contact to the
removal of the tape was different between sham- treated
and postasphyxial rats (Figure S8). In both experimental
groups, animals improved over the experimental period.
Tests for anxiety- related behavior at
P18 to 14months
Although measurements of anxiety- related behavior in
rodents are often done using a single test, experience from
neuropharmacological work has shown it is better to use
several kinds of tests that measure anxiety under different
conditions.24,25 Here, three unconditioned response tests
(which require no training and usually have a high eco- /
ethological validity) were used: the open field test (P18 to
14months), the elevated plus maze test (P26 to 14months),
and the light– dark box test (P27 to 14months). All three
tests are based on the conflict of rodents between explo-
ration and natural aversions to illuminated, open, and/
or elevated areas.26Thus, as shown in Figure S9 for the
open field test, rats stayed much longer in the outer zone
of the field compared to the aversive middle and center
zones, without any significant intergroup difference.
Furthermore, both groups exhibited the same explorative
activity as indicated by the distance moved and velocity.
Similarly, in the elevated plus maze test (Figure S10),
rats stayed longer in the closed arms than in the center or
open arms of the maze, without any significant intergroup
difference. Furthermore, both groups exhibited the same
explorative activity as indicated by the distance moved and
velocity (Figure S4). However, it should be noted that in
both open field (Figure S9) and elevated plus maze (Figure
S10), the time spent in the aversive locations of these tests
was already quite low in sham controls, which makes it
difficult to determine an increase in anxiety- related be-
havior in the postasphyxial rats.
In the light– dark box test, significantly increased
anxiety- related behavior was observed in postasphyxial rats
at 6 and 14months after asphyxia. As shown in Figure2,
sham controls stayed a much shorter period of time in the
aversive light compartment of the test than in the dark
compartment, which did not change over the 14 months
of the trial period. Both the duration of stay in the light
GAILUS et al.
compartment and the crossing between the dark and light
compartments were significantly reduced in postasphyx-
ial rats, indicating increased anxiety- related behavior.
Interestingly, the data in Figure 2 indicate that this behav-
ior of postasphyxial rats developed slowly, with significant
effects only reached at 6– 14months after the insult.
Long- term cognitive alterations
at 3 and 6months after asphyxia with
Alterations in learning and memory were determined by
the RAWM.27The RAWM is a hybrid of the Morris water
FIGURE The behavior of rats in the light– dark box test at different developmental or time periods following asphyxia. As described
in Appendix S1, time spent in each compartment, and the crossings between compartments, were measured for 3min by a video tracking
system. Data are shown as boxplots with whiskers from minimum to maximal values; the horizontal line in the boxes represents the median
value. In addition, individual data are shown. Male and female rats are shown by different symbols (see key in the figure). For each period,
the duration of stay in the aversive light compartment and the number of crossings between the light and dark compartments are shown.
Data were analyzed with a two- way analysis of variance (ANOVA) mixed- effects model (see Appendix S1), followed by Sidak's multiple
comparisons test. For duration of stay in the light compartment, results from two- way ANOVA were F3.153, 70.95=3.084 for time (p<.05)
and F1, 23=10.3 (p<.01) for column factor. For crossings between compartments, results from two- way ANOVA were F1.934, 42.55=26.47 for
time (p<.00001) and F1, 23=12.55 (p<.01) for column factor. Within each group, sham controls did not significantly differ across the five
periods, whereas postasphyxial rats stayed for a significantly shorter time in the light box at 14months after asphyxia and exhibited fewer
crossings between the dark and light box at 6 and 14months compared to 2 weeks, indicating the development of anxiety- related behavior.
These intragroup differences are also mirrored by significant differences from sham controls, which are indicated by asterisks (*p<.05;
**p<.01). No significant sex differences were observed
GAILUS et al.
maze and a radial- arm maze (Figure S5H), which takes
advantage of the simple motivation provided by immer-
sion into water, together with the benefits of scoring er-
rors (rather than time or proximity to platform location)
associated with the radial- arm maze.27 At 3 and 6months
after asphyxia, sham controls exhibited significant learn-
ing on the second day of the test (Figure 3A,B) in the
2- day reference memory version of this task (Figure S5H;
see Appendix S1). Learning was not observed in the as-
phyxia group, indicating that asphyxia with neonatal sei-
zures induce long- term deficits in reference memory or
other cognitive processes required for the task. It is worth
mentioning that after the training period (Trials 1– 12,
Day 1), all rats of both groups were able to find the plat-
form and almost all trials were completed successfully
within 60s (sham: 3months=98.1%, 6months=100%;
asphyxia: 3 months = 97.9%, 6 months = 99.5%).
However, as shown in Figure 3C,D, the postasphyxial
rats made significantly more errors in finding the correct
arm of the maze, and this did not improve throughout
Apart from that, the described deficits were also ob-
served just by analyzing the rat´s exploration strategies.
Both groups started to systematically search for the plat-
form in one arm after another; however, at the end of the
task, most of the sham- controls swam into the center and
turned around for better orientation to directly choose the
goal arm. In contrast, post- BA rats stuck to the system-
atic searching approach described before, thus finding the
goal arm by chance after several errors. In addition, they
sometimes did not swim to the very end of one arm but
rather turned around before reaching a possible platform
location, which could reflect enhanced impulsivity. Figure
3C,D summarizes the cumulative errors recorded in the
two groups of rats, substantiating the marked cognitive
impairment of the asphyxia group.
Development of spontaneous
recurrent seizures after asphyxia
with neonatal seizures (recorded at
~10.5months of age)
As described in Appendix S1, two of the 25 rats (one male
sham rat, one male postasphyxial rat) had to be eutha-
nized at ~6 and 9months, respectively. The remaining
23 animals (11shams, 12 asphyxia) were implanted with
EEG electrodes at 10months following asphyxia (Figure
1; Figure S11) to avoid the insult produced by electrode
implantation from affecting the behavioral and cognitive
alterations after asphyxia. Continuous (24/7) video- EEG
monitoring for 1 week started ~3weeks after the electrode
implantation. A total of 3864h of video- EEG recording in
23 rats was visually analyzed for the occurrence of spikes,
spike clusters, electrographic or electroclinical seizures,
and other abnormal epileptiform activity by two experi-
enced observers, who were not aware of whether the rats
were from the sham or asphyxia groups. In case of any ab-
normal EEG activity, the concomitant video was viewed
for behavioral alterations. Based on the commonly used
definition of seizures,28 a seizure was defined as paroxys-
mal EEG alteration consisting of epileptiform spikes and
sharp wave trains with an amplitude at least two times
greater than the background, a frequency of at least 2Hz,
and a duration of at least 5 s. Three types of SRS were
found: (1) electrographic seizures without any obvious be-
havioral alteration in the concurrent videos (Figure 4B);
(2) nonconvulsive electroclinical seizures (Figure 4C,D)
that had an associated behavioral correlate, such as sud-
den behavioral arrest, staring episodes, head- jerking, and
facial automatisms; and (3) generalized convulsive elec-
troclinical seizures (Figure 4E). As shown in Figure 4,
the EEG alterations during these three seizure types were
polymorphic. Examples of such EEG seizures and their
behavioral correlates are shown in Video S1.
SRS were recorded in all postasphyxial rats (Figure
5A). Median SRS frequency was 92/week, with a wide in-
terindividual range from four to 1213/week. As shown in
Figure 5C,D, electrographic seizures were more frequent
than electroclinical seizures. Most electroclinical seizures
were nonconvulsive; only one postasphyxial rat exhib-
ited generalized convulsive seizures with loss of righting
reflexes and clonic movements of both hindlimbs. The
occurrence of the SRS over the 7 days of the recording pe-
riod is illustrated in Figure S2. Most SRS had a duration of
5– 10s (please note the lower limit of duration was set at
5s). In addition to these seizures, spikes and spike clus-
ters with a duration of 2– 4s were frequently observed (not
However, some epileptiform EEG alterations were also
observed in approximately one third of the sham control
rats (Figure 5A). They occurred at much lower frequen-
cies than in the postasphyxial rats (Figures 5B– D, S2).
Furthermore, none of the sham controls exhibited gen-
eralized convulsive seizures. The presence of these rare
low- frequency epileptiform discharges in control rats is
consistent with previous reports29 and may be either ge-
netically inherent or a consequence of the brain injury in
response to the EEG electrode implantation (see Section
4). As described in Appendix S1, no obvious histological
alterations were observed in the cortex below the epidural
screw electrodes in any rat. Regardless, both the incidence
and frequency of SRS were much higher in postasphyxial
rats than in the controls (Figure 5). The background EEG
(Figure 4A) did not show obvious differences between the
control and postasphyxial rats.
GAILUS et al.
FIGURE Performance of rats in the radial- arm water maze test. The test was repeatedly performed over 2 days at 3months (A, C)
and 6months (B, D) following asphyxia. In A and B, data are shown as mean±SEM. In C and D, data are shown as boxplots with whiskers
from minimum to maximal values; the horizontal line in the boxes represents the median value. In addition, individual data are shown.
In A and B, data from both sexes were poooled, whereas in C and D, male and female rats are shown by different symbols (see key in D).
For comparison of different trials within one group, the repeated measures one- way analysis of variance (ANOVA) test for nonparametric
data (Friedman test), followed post hoc by Dunn's multiple comparison test was used. For comparing data of sham- and asphyxia- treated
animals, the two- way ANOVA test followed post hoc by Sidak's multiple comparisons test was used. Group differences in cumulative
errors were analyzed by the t- test. (A) The average number of errors in sham controls and rats that had experienced asphyxia and neonatal
seizures 3 months before the radial- arm water maze test. The Friedman test indicated significant differences within each group for sham
rats (Q value = 33.97, df = 9, p<.0001) but not postasphyxial rats (Q value = 10.87, df = 9, p=.285). Two- way ANOVA indicated significant
differences between both groups: F5.7, 131.1=3.194 (p=.0067) for time and F1, 23=26.6 (p<.0001) for column factor. Data from post hoc
analysis indicating significant learning within the sham group (vs. Trials 1– 3) are indicated by asterisks. Significant differences between
both groups are indicated by hash signs. (B) The same as A but 6months after asphyxia. The Friedman test indicated significant differences
within each group for sham rats (Q value = 39.14, df = 9, p<.0001) but not postasphyxial rats (Q value = 4.687, df = 9, p=.8607). Two-
way ANOVA indicated significant differences between both groups: F4.63, 101.9=4.701 (p=.0009) for time and F1, 22=16.15 (p=.0006)
for column factor. (C) Cumulative errors within each group over the whole duration of the 30 individual trials (2 days) at 3 months after
asphyxia. Significant differences from sham controls are indicated by asterisks. (D) Cumulative errors within each group over the whole
duration of the 30 individual trials (2days) at 6months after asphyxia. Significant differences from sham controls are indicated by asterisks.
There were no significant differences between male and female rats. #p<.05, ##p<.01, ####p<.00001, *p<.05, **p<.01, ***p<.001,
GAILUS et al.
hippocampus and thalamus, and aberrant
mossy fiber sprouting in postasphyxial rats
(determined at ~14months of age)
Survivors of HIE are at increased risk of cognitive im-
pairment, which is thought to be at least in part a con-
sequence of hippocampal damage.7,30 Thus, the deficit
in spatial learning and memory observed following as-
phyxia with neonatal seizures in the RAWM indicated
that affected rats might exhibit neurodegeneration in the
hippocampus, which plays a crucial role in spatial mem-
ory.31 Given the functional differences between the dorsal
and ventral hippocampus,32– 34 we first examined sections
of the dorsal hippocampus, which is involved with cogni-
tive and memory functions in rats. As shown in Figure
6A, the dorsal hippocampal formation of the two groups
of rats displayed normal features when NeuN- and DAPI-
stained sections were compared, but at higher magni-
fication (Figure 6B) differences in neuronal density can
be seen. Quantification of NeuN- positive neurons in the
dentate hilus demonstrated a significant neuronal loss in
postasphyxial rats (Figure 7A). No significant intergroup
differences were obtained for the area of the hilus (Figure
7B). In addition to the cell loss in the dentate hilus, a sig-
nificant reduction of neurons in CA3c was determined
in postasphyxial rats (Figure 7C). In contrast, no obvious
neurodegeneration was observed in CA1 of the dorsal hip-
pocampus (Figure 7D).
In addition to the dorsal hippocampus, we examined
sections of the ventral hippocampus, which is involved
in affective processes such as anxiety in rodents.33,34 As
shown in Figure S12, findings were similar to those in the
dorsal hippocampus, with significant loss of neurons in
dentate hilus and CA3c.
The dentate hilus is a polymorphic layer with two
major cell types, glutamatergic mossy cells and diverse γ-
aminobutyric acidergic (GABAergic) interneurons, which
exhibit different vulnerabilities to brain injuries.35 This
prompted us to differentiate mossy cells and different
subtypes of GABAergic interneurons by immunohisto-
chemistry (Figure S13). As shown in Figure S14A,C, in
both dorsal and ventral hippocampus, mossy cells and
parvalbumin- positive GABAergic interneurons were pref-
erentially lost after asphyxia, whereas no significant loss
of somatostatin- positive GABAergic interneurons was
observed. We also determined the percentage of the three
subpopulations of hilus neurons of all NeuN- positive neu-
rons counted in the hilus, confirming that the mossy cells
and somatostatin- positive interneurons were the major
cell types in the hilus and illustrating the effect of as-
phyxia (Figure S14B,D).
Loss of mossy cells in the dentate hilus may lead to
aberrant sprouting of mossy fibers.36 Granule cell axons
(mossy fibers) project into the hilus of the dentate gyrus
and stratum lucidum of CA3 in rodents and other spe-
cies, including humans, where they synapse with hilar
mossy cells, inhibitory interneurons, and CA3 pyrami-
dal cells, but only very rarely with other granule cells.36
Accordingly, when synaptoporin was used to label mossy
fibers in the dorsal hippocampus, the highest density of
labeling was observed in the dentate hilus and stratum lu-
cidum of CA3 (Figure 6C). In postasphyxial rats, aberrant
mossy fiber sprouting (MFS) was observed in the stratum
FIGURE Representative electroencephalographic (EEG)
recordings of rats at ~10.5months after asphyxia. The EEG was
recorded with the epidural screw electrodes located above the
motor cortex close to the hippocampal formation. (A) Normal
baseline (background) EEG. (B) An electrographic seizure that
was not associated with any obvious behavioral abnormality in the
corresponding video. However, subtle behavioral abnormalities
may have been missed in the video. (C) A nonconvulsive
electroclinical seizure. During the paroxysmal EEG activity, a
behavioral arrest was observed in the corresponding video. (D)
shows an extended part (5×) of the seizure illustrated in C. (E)
A generalized convulsive electroclinical seizure. Note the much
higher amplitude of the paroxysmal EEG activity. During the
paroxysmal EEG activity, the loss of righting reflexes and clonic
movements of both hindlimbs were observed in the corresponding
GAILUS et al.
oriens of the CA3 hippocampal region (Figure 6C,D).
Quantification of these data by image analysis is shown in
Figure 7E. In contrast to post- status epilepticus models of
temporal lobe epilepsy,36 no aberrant MFS was observed
in the inner molecular layer of the dentate gyrus (Figure
6C), which is similar to findings in hypoxia- only models
As described above, aberrant MFS is often a conse-
quence of loss of hilar neurons36 as observed here. Thus,
we examined the correlation between the hilar neuron
FIGURE Incidence and frequency of electrographic and electroclinical seizures in rats at ~10.5months after asphyxia. Because
paroxysmal activity was also observed in sham controls (but with much lower incidence and frequency), this is shown in comparison. Group
differences were analyzed by the Barnard test (seizure incidence) or the Mann– Whitney test (seizure frequencies). Significant differences
between sham controls and postasphyxial rats are indicated by asterisks (*p<.05; **p<.01). Data in B– D are shown as boxplots with
whiskers from minimum to maximal values; the horizontal line in the boxes represents the median value. In addition, individual data are
shown. In A, data from both sexes were poooled, whereas in B– D, male and female rats are shown by different symbols (see key in B). (A)
Percent incidence of all seizure types in sham controls and postasphyxial rats. (B) The number of all seizures recorded by continuous (24/7)
monitoring in 1 week. (C) Number of electrographic seizures/week. (D) Number of electroclinical seizures/week. Note that the majority
of these seizures were nonconvulsive (see text). Only paroxysmal electroencephalographic events with a duration of at least 5s were
considered seizures. No sex differences were observed
GAILUS et al.
GAILUS et al.
density and MFS in the 23 rats of the two groups used for
this analysis. As shown in Figure 7F, a significant correla-
tion coefficient (r=−.7074, p=.0002) was obtained, indi-
cating that the most marked MFS occurred in rats with the
highest loss of hilar neurons.
In addition to the hippocampus, NeuN- stained sections
of several other brain areas, including the cerebral cortex,
globus pallidus, caudate– putamen, subthalamic nucleus,
and different thalamic areas, were visually examined for
any obvious neuronal damage. These brain areas were
chosen based on clinical findings following asphyxia3– 5,7
as described in Section 1 and on findings in hypoxia–
ischemia models of HIE.5,13 Differences between sham
controls and postasphyxial rats were present in thalamic
areas below the third ventricle, that is, the paraventricular
and mediodorsal thalamus (Figure S15). Quantification
of the density of NeuN- positive cell bodies demonstrated
significant neuronal loss in the postasphyxial rats (Figure
Correlation between seizure
frequency and structural brain alterations
In Figure S2, individual seizure frequencies, neuronal
loss in the hilus, and aberrant MFS are illustrated for all
23 rats that were finally used for video- EEG monitoring.
Nonparametric Spearman correlation analysis showed
that both neuronal loss in the hilus and MFS were signifi-
cantly correlated with the frequency of SRS (Figure S17).
The correlation coefficient r was −.6667 for hilar neuronal
density versus SRS frequency (p=.0005) and .4770 for in-
tensity of MFS versus SRS frequency (p=.0214).
This is to our best knowledge the first study that describes
the long- term consequences of asphyxia (not pure hy-
poxia) and subsequent neonatal seizures in a rodent model.
Rodent models of HIE/neonatal seizures used in previous
work were either hypoxia– ischemia models or hypoxia-
only models,14,15 which lack the respiratory acidosis (hy-
percapnia) that is a fundamental constituent of BA.37
Hypercapnia has numerous effects on brain functions
and neuronal excitability.16,17 One of the striking differ-
ences between hypoxia– ischemia or hypoxia- only models
and the present asphyxia model is that seizures start dur-
ing hypoxia in the former.13,14This is in line with in vitro
data showing that hypoxia as such promotes neuronal
excitability and synaptic potentiation.38,39Moreover, the
highly artificial condition of pure hypoxia in vivo (which
never occurs during cardiorespiratory failure or local/
global ischemia under nonexperimental conditions) leads
to a brain alkalosis that further promotes excitability (for
data and references, see Pospelov et al.16 and Ala- Kurikka
et al.17). In contrast, hypercapnia, that is, an elevation of
systemic CO2, produces a fall in brain pH and a conse-
quent decrease in neuronal excitability (see Ruusuvuori
and Kaila40 and references cited therein), which explains
the finding that neonatal seizures do not occur during as-
phyxia but only during the subsequent establishment of
normocapnic conditions and brain pH recovery.16,17This
is observed both in human neonates and also in relevant
large- animal models, in which neonatal seizures are trig-
gered after a period of moderate or severe asphyxia.5Thus,
the brain is normoxic during neonatal seizures, which is
a unique, translationally relevant feature of the present
model compared to all rat and mouse hypoxia– ischemia
or hypoxia- only models. Notably, the present data in-
dicate that the differences in the mechanisms involved
in HIE and seizure generation affect later life outcome,
which obviously is an important aspect of translational re-
search of this kind.
In Table 1, the observations in our rat model of BA are
compared with those obtained in hypoxia- alone, hypoxia–
ischemia, and ischemia rat and mouse models of HIE and
neonatal seizures. The only commonality across most
models is the development of epilepsy, with SRS and ab-
errant MFS in the hippocampus. The lack of any obvious
hippocampal damage in rat hypoxia- only models, which
is in sharp contrast to the consequences of HIE in hu-
mans,5may explain why no consistent cognitive impair-
ment has been reported in these models, whereas such
impairment is seen in all other models in Table 1.
A point worth emphasizing in the present context is
the developmental stage of brain development of rat and
mouse pups used in the various models. Traditional mod-
els of developmental brain injury have utilized rodents at
P7, which, based on numerous milestones of cortical (i.e.,
FIGURE Representative photomicrographs of sections of the dorsal hippocampus from sham controls and postasphyxial rats at
~14months after asphyxia. (A) Overview of the left dorsal hippocampus illustrated by NeuN- and DAPI- stained sections. (B) NeuN- stained
sections at higher magnification. Note the obvious cell los in the dentate hilus and CA3c region of postasphyxial rats. (C) Synaptoporin-
stained sections. (D) Synaptoporin- stained sections at higher magnification to illustrate the aberrant mossy fiber sprouting into the stratum
oriens of the CA3 region (arrows). For better visualization, a twofold enlargement of this region of interest was inserted into the bottom right
GAILUS et al.
hippocampal and neocortical) development, corresponds
to a preterm neonate with a mean of the estimated post-
conceptional age of approximately 25– 30 weeks.41,42 In
contrast to this, rats and mice at the age of P10– P12 ex-
hibit many of those developmental brain characteristics
that are achieved by the human term neonate (i.e., when
finishing 40gestational weeks), including continuity and
other features of the EEG,18,19,43 as well as similar cortical
expression patterns of important genes/proteins,44such as
the neuron- specific KCC2, which is an excellent indicator
of neuronal maturity.45,46
In the present study, the consequences of intermittent
asphyxia with neonatal seizures for motor, behavioral,
and cognitive functions as well as for brain structure were
evaluated over a period of ~14 months after asphyxia,
that is, a large part of the life span of the Han:Wistar
outbred rats used here, which have a median life ex-
pectancy of 30– 33 months (females) and 33– 36months
(males).47 Following asphyxia with subsequent seizures,
no later life motor or somatosensory abnormalities were
observed, but the postasphyxial rats developed increased
anxiety, a striking cognitive decline, a high incidence and
FIGURE Neurodegeneration and aberrant mossy fiber sprouting (MFS) in the dorsal hippocampus of postasphyxial rats, determined
~14months after asphyxia. Data in A– E are shown as boxplots with whiskers from minimum to maximal values; the horizontal line in the
boxes represents the median value. In addition, individual data are shown. Male and female rats are shown by different symbols (see key
in A). Group differences were analyzed by the Mann– Whitney test. Significant differences between sham controls and postasphyxial rats
are indicated by asterisks (**p<.01; ****p<.0001) and were determined from six hippocampi with three slices taken from each animal
at section levels −3.0, −3.16, and −3.32mm from bregma. (A) Neuronal density in the hilus. (B) Average area of the hilus. (C) Neuronal
density in CA3c. (D) Neuronal density in CA1. (E) Aberrant MFS into the stratum oriens of the CA3. (F) Correlation between MFS and
neuronal density in the hilus. Note that all rats that exhibited MFS had lower neuronal densities than sham controls; the higher the
neuronal loss, the higher the MFS (r=−.7074; p=.0002). One postasphyxial rat had much higher MFS than the rest of the group. When
excluding this rat as an outlier, the Spearman rank method resulted in a correlation coefficient (r) of −.6927 (p=.0004). No significant
differences in the long- term effects of asphyxia between male and female rats were detected in any of the data illustrated in this figure.
However, the hilus area in male sham rats was significantly bigger than that of female sham rats (p=.0498). Furthermore, neuronal density
in the CA3c of sham controls was significantly lower in female than male rats (p=.0455)
GAILUS et al.
TABLE Comparison of later life consequences in rat and mouse models of HIE and neonatal seizures with such consequences following prolonged birth asphyxia in human neonates
Later life consequences
Rat and mouse models of birth asphyxia/HIE and neonatal seizures
Noninvasive models Invasive models
[hypoxia and hypercapnia] Hypoxia- alone
Hypoxia– ischemia [Rice–
Vannucci] models of HIE
[with unilateral carotid artery
Prolonged birth asphyxia
resulting in HIE with neonatal
seizures in human neonatesa
Species and postnatal age P11 rat P8– P12 rat;
P7 rat; P10– P12 rat; P7mouse P7– P12mice (neonatal
rats do not exhibit
Neurodegeneration Yes (in hippocampus, and
Yes (in hippocampus
Yes (gray matter injury observed in
cortex, hippocampus, thalamus,
and basal ganglia)
Yes (gray matter injury
particularly observed in cortex,
hippocampus, thalamus, basal
ganglia, and brain stem)
Mossy fiber sprouting in
Yes Yes Yes N/A N/A (but seen in adult patients
with mesial temporal lobe
Developmental delays No N/A N/A N/A Yes
Learning and memory
Yes (impaired spatial learning
Mixed results Yes (impaired spatial learning and
Yes Yes (related to smaller
Hyperactivity No Yes N/A Yes Yes
Increased aggression No Yes N/A N/A Yes
ADHD NT Yes Yes N/A Yes
NT Yes (social deficit) Yes (reduced attention) No (but large
NT Yes N/A N/A Yes
Anxiety Yes Yes Yes No (but large
Disturbed sleep NT Yes N/A Yes Yes
No (adhesive removal test) N/A Yes N/A Yes
Hearing and vision loss NT N/A N/A N/A Yes
GAILUS et al.
Later life consequences
Rat and mouse models of birth asphyxia/HIE and neonatal seizures
Noninvasive models Invasive models
[hypoxia and hypercapnia] Hypoxia- alone
Hypoxia– ischemia [Rice–
Vannucci] models of HIE
[with unilateral carotid artery
Prolonged birth asphyxia
resulting in HIE with neonatal
seizures in human neonatesa
Epilepsy with SRS Yes Yes Yes N/A Yes
Motor disabilities No N/A Yes N/A Yes
CP and other motor
No No No No Yes (but more recent work has
shown that birth asphyxia does
not play a major role in the
incidence of CP)
Disadvantages of the
New model ⇒ only few data
on variability across
High variability across
hypoxia- only does
not replicate birth
High variability across
laboratories; the invasive
nature of severing the common
carotid artery does not replicate
Rarely used in the
context of neonatal
of severing the
artery does not
References Present study Sun et al.,14Millar
et al.,5 Quinlan
et al.,78 Hamdy
Kadam and Dudek,75Kadam
et al.,79Millar et al.,5 Hamdy
de Haan et al.,30 van Handel
et al.,3 Ahearne et al.,4Millar
et al.,5Korzeniewski et al.81
Note: Rodent models in which neonates are asphyxiated in utero are not included, because such models do not allow studying neonatal seizures.13
Abbreviations: ADHD, attention- deficit/hyperactivity disorder; CP, cerebral palsy; HIE, hypoxic– ischemic brain injury; N/A, no data found; NT, not tested yet; P, postnatal day; SRS, spontaneous recurrent seizures.
aThe long- term effects of birth asphyxia depend on the age at the time of insult (which also has a major impact on which region of the brain is injured) and the severity of the injury; there is a critical threshold of
asphyxia beyond which brain damage occurs82; furthermore, outcomes are poorer for those with more severe asphyxia.30
GAILUS et al.
frequency of SRS, neurodegeneration in the hippocampus
and thalamus, and aberrant MFS in the stratum oriens of
the CA3sector of the hippocampus. As shown in Table 1,
these alterations resemble the clinical situation, demon-
strating that the present model recapitulates several of
the later life consequences of BA in human neonates. No
obvious sex differences were observed in most readouts;
however, such comparison between sexes was affected
by the relatively low sample size of the male and female
Interestingly, the increased anxiety- related behavior
observed in the light– dark box developed slowly; that
is, significant differences from sham controls were seen
only after 6– 14 months. Thus, this behavior would not
have been detected by a shorter observation period or
the use of only one anxiety model, as done in most previ-
ous studies using other models of BA/HIE and neonatal
seizures (Table 1). The present finding of the protracted
development of anxiety after asphyxia matches strikingly
with what has been observed at an age of 38years in the
Dunedin Multidisciplinary Health and Development
Study, a prospective longitudinal study of a representa-
tive (n=1037) birth cohort.48,49 In terms of human de-
velopment, the rat age span of 6– 14months corresponds
roughly to 18– 40 years,50 with a shift to somewhat older
human age windows in mice of the same age.51
Among the most common consequences of perinatal
asphyxia in humans are cognitive abnormalities later in
life, ranging from learning disabilities to developmental
delay, mental retardation, and autism.3– 5,30 In children
following perinatal asphyxia, impairments in memory
were found to be associated with a reduced volume of the
hippocampus, a brain region that is specifically vulnera-
ble to asphyxia.4 Similar to the clinical findings, marked
cognitive impairment, as evidenced by deficits in learning
and memory, was a later life consequence in the present
model. Furthermore, neurodegeneration was observed in
the dorsal hippocampus, which, together with the aber-
rant MFS in the CA3 region, explains the decline in spa-
tial learning.32,33,52 Although, to our knowledge, aberrant
MFS has not yet been described as a later life consequence
of BA in humans, it is a well- known structural alteration
in patients with mesial temporal lobe epilepsy, which is
the most common type of epilepsy in adults.30
The behavior of postasphyctic rats in the six- arm
RAWM model of spatial learning and memory used here
resembled impulsive behavior as observed in young rats
(P30– P76) in an eight- arm RAWM paradigm following
hypoxic– ischemic brain injury in P7 rats.53 Impulsivity is
a core symptom of attention- deficit/hyperactivity disor-
der— a diagnosis common in children with BA/HIE.3,54
Following BA/HIE, impulsive behavior can occur along
with memory impairment, most likely as a consequence
of neurodegeneration in the hippocampus and striatum,
which have been associated with specific cognitive func-
tions such as memory and attention.3,54More specific tests
of impulsivity, as previously used in hypoxia– ischemia
and hypoxia- only rat models,55– 57 are needed to character-
ize in more detail the altered behavior of the postasphyxial
rats in the RAWM observed here.
In addition to the morphological alterations in the
hippocampus, neuronal loss was observed in the paraven-
tricular and mediodorsal thalamus of the postasphyxial
rats. Neuroimaging and neuropathological studies have
revealed that the thalamus is among the selectively vul-
nerable brain regions following HIE in the human new-
born and may contribute to sensorimotor deficits.5,30,58
In human neonates with HIE, injury to the thalamus is
typically not observed in isolation, but in association with
damage to the hippocampus,5,30 as also observed in the
present animal model.
Epilepsy represents a common outcome of newborns
with neonatal seizures, especially in those with severe
brain injury and additional neurodevelopmental disabili-
ties.59 A high incidence of severe types of epilepsy has been
described following HIE/neonatal seizures, including in-
fantile spasms (West syndrome), early myoclonic enceph-
alopathy, and early infantile epileptic encephalopathy, but
the occurrence of such syndromes is strongly determined
by the underlying etiology of HIE.59 Other seizure types
that have been described include focal and generalized
seizures similar to those observed here, often coexisting in
the same patient and poorly responding to therapy.59– 61 In
the present study, 100% of the postasphyxial rats exhibited
SRS, recorded ~10 months after asphyxia/neonatal sei-
zures. Three types of SRS were observed: electrographic
seizures without obvious clinical correlates, nonconvul-
sive electroclinical seizures with focal behavioral seizure
signs, and, rarely, convulsive electroclinical seizures.
The frequency of the SRS varied widely from a few sei-
zures to hundreds of seizures per week. Electrographic
and nonconvulsive electroclinical seizures occurred at
low frequency also in sham controls, which has been re-
ported in previous studies in other rodent models, includ-
ing the hypoxia- only model of HIE in rats.62 Paroxysmal
EEG patterns have been observed in various rat strains.29
Such patterns may, at least in part, result from structural
or functional alterations caused by the invasive electrodes
during chronic EEG recordings.29,63,64
Interestingly, in postasphyxial rats, we found signifi-
cant correlations between the frequency of SRS and loss
of neurons in the dentate hilus and the extent of aberrant
MFS in the CA3. The correlation between seizure fre-
quency and cellular effects begs the question of whether
it was the asphyxia and/or neonatal seizures per se that
initiated a very long maturational process that culminated
GAILUS et al.
in epilepsy; or whether there was a much earlier onset
of spontaneous electrographic seizures that, over time,
led to progressive cellular damage and to the behavioral
changes. This is an important issue that should be ad-
dressed in further studies.
As described above, aberrant sprouting of mossy
fibers is often a consequence of neuronal loss in the
dentate hilus,36 which is the most susceptible region of
the hippocampal formation following different types
of brain injury35,65 but is relatively resistant to dam-
age in hypoxia- alone and hypoxia– ischemia models of
HIE.14 However, data concerning damage to the hilus
in hypoxia- alone and hypoxia– ischemia models of HIE
are limited and not quantitative. Furthermore, the P7 rat
(used in many HIE models) is less susceptible to hippo-
campal neurodegeneration in response to hypoxia/as-
phyxia and ischemia than the P10– P12 rat.43Moreover,
specific populations of hilar neurons could be lost, and
this would not be known unless appropriate staining
procedures were combined with quantitative analyses to
address this question. This prompted us to count mossy
cells and different subtypes of GABAergic interneu-
rons in the hilus, which indicated that mossy cells and
parvalbumin- positive GABAergic interneurons were
preferentially lost following asphyxia.
Hilar cell loss as well as increased and aberrant MFS
are observed in cases of human temporal lobe epilepsy and
many experimental models of epilepsy.35MFS has been as-
sociated with epileptogenesis in animal models, although
a direct cause– effect relationship has long been a matter
of debate.35,36,65,66More recent preclinical studies suggest
that the new (synaptoporin- positive) neurites establish
functional synaptic connections and may contribute to the
state of hyperexcitability that either provokes or facilitates
abnormal discharges.35,67 In a hypoxia- only model of neo-
natal seizures, aberrant MFS was observed in the stratum
oriens of the CA3hippocampal region,62,68similar to the
MFS in the present model. The altered mossy fiber distri-
bution was a likely explanation of the enhanced synaptic
activity seen at CA1:CA3synapses following neonatal sei-
zures69 and the disturbance in cognitive function seen in
later life.70,71Although in the present model and hypoxia-
only models of HIE, MFS was restricted to the CA3 re-
gion, aberrant MFS into the molecular layer of the dentate
gyrus has been observed in more severe models of HIE.14
We recently reported a loss of neurons in the CA1 re-
gion of the hippocampus shortly (24 h) after asphyxia,
which seemed to be a consequence of neonatal seizure-
induced apoptosis.72This is consistent with findings in
other HIE models that the CA1 (and CA3) regions are
particularly sensitive to hypoxia and ischemia.73– 75 In the
present study, we found a significant loss of CA3c neurons
in both dorsal and ventral hippocampus, but unexpectedly,
we did not observe a corresponding fall in the number of
CA1 neurons at 14months after asphyxia. One possible
explanation would be an adaptive change in the overall
morphology of the hippocampus, with long- term rewiring
of the hippocampal circuitry and rearrangement of CA1
neurons. This explanation is, however, difficult to verify
In conclusion, there is increasing preclinical and clin-
ical evidence that neonatal seizures increase the risk of
adverse outcomes following HIE.12Thus, new treatments
that more efficaciously suppress neonatal seizures and the
later life consequences of HIE and neonatal seizures are
urgently needed. The rat model of BA17 employed in the
present study provides a novel tool to develop such treat-
ments. As shown here, this model recapitulates a number
of salient later life clinical outcomes of BA and subsequent
neonatal seizures, including damage to susceptible brain
regions, cognitive decline, anxiety, and epilepsy. Results
from animal research suggest that neonatal seizures may
exacerbate HIE- induced brain injury and neurodevelop-
mental outcome,9,76,77 whereas this is a matter of debate
in humans.2,9,11,12,59 Because seizures occur after asphyxia
in the present model, it can be employed in future stud-
ies to examine the controversial and important question
of whether antiseizure drugs that block the postasphyxia
seizures20 will decrease the incidence or severity of later
We thank Larsen Kirchhoff, Edith Kaczmarek, Ivo
Denden, Maike Krüger, Denise Riebschläger, Jessica
Nachtwey, and Ricardo Schmidt for technical support and
Dr. Christopher Käufer for instruction on how to perform
the RAWM. The study was supported in part by a grant (LO
274/15- 1) from the Deutsche Forschungsgemeinschaft
(Bonn, Germany). B.G. was supported by a PhD schol-
arship from the Studienstiftung des Deutschen Volkes
(Bonn, Germany), and M.J. by a PhD scholarship from
the Konrad Adenauer Stiftung (Berlin, Germany). The
open access publication was supported by the Deutsche
Forschungsgemeinschaft and University of Veterinary
CONFLICT OF INTEREST
None of the authors has any conflict of interest to disclose.
We confirm that we have read the Journal′s position on
issues involved in ethical publication and affirm that this
report is consistent with those guidelines.
Björn Gailus https://orcid.org/0000-0002-1980-7133
Marie Johne https://orcid.org/0000-0002-0490-2648
Wolfgang Löscher https://orcid.org/0000-0002-9648-8973
GAILUS et al.
1. Guidelines on neonatal seizures. Geneva, Switzerland: World
Health Organization; 2011.
2. Soul JS. Acute symptomatic seizures in term neonates: etiologies
and treatments. Semin Fetal Neonatal Med. 2018;23:183– 90.
3. van Handel M, Swaab H, De Vries LS, Jongmans MJ. Long- term
cognitive and behavioral consequences of neonatal encepha-
lopathy following perinatal asphyxia: a review. Eur J Pediatr.
4. Ahearne CE, Boylan GB, Murray DM. Short and long term
prognosis in perinatal asphyxia: an update. World J Clin Pediatr.
5. Millar LJ, Shi L, Hoerder- Suabedissen A, Molnár Z. Neonatal
hypoxia ischaemia: mechanisms, models, and therapeutic chal-
lenges. Front Cell Neurosci. 2017;11:78.
6. LaRosa DA, Ellery SJ, Walker DW, Dickinson H. Understanding
the full spectrum of organ injury following intrapartum as-
phyxia. Front Pediatr. 2017;5:16.
7. Annink KV, De Vries LS, Groenendaal F, van den Heuvel
MP, van Haren NEM, Swaab H, et al. The long- term effect
of perinatal asphyxia on hippocampal volumes. Pediatr Res.
8. Zhou KQ, Davidson JO, Bennet L, Gunn AJ. Combination treat-
ments with therapeutic hypothermia for hypoxic- ischemic neu-
roprotection. Dev Med Child Neurol. 2020;62:1131– 7.
9. Chapman KE, Raol YH, Brooks- Kayal A. Neonatal seizures:
controversies and challenges in translating new therapies from
the lab to the isolette. Eur J Neurosci. 2012;35:1857– 65.
10. Zhou KQ, McDouall A, Drury PP, Lear CA, Cho KHT,
Bennet L, et al. Treating seizures after hypoxic- ischemic
encephalopathy— current controversies and future directions.
Int J Mol Sci. 2021;22:7121.
11. De Haan TR, Langeslag J, van der Lee JH, van Kaam AH. A
systematic review comparing neurodevelopmental outcome in
term infants with hypoxic and vascular brain injury with and
without seizures. BMC Pediatr. 2018;18:147.
12. Pressler RM, Lagae L. Why we urgently need improved
seizure and epilepsy therapies for children and neonates.
13. Hamdy N, Eide S, Sun HS, Feng ZP. Animal models for neo-
natal brain injury induced by hypoxic ischemic conditions in
rodents. Exp Neurol. 2020;334:113457.
14. Sun H, Juul HM, Jensen FE. Models of hypoxia and ischemia-
induced seizures. J Neurosci Methods. 2016;260:252– 60.
15. Boylan GB, Stevenson NJ, Vanhatalo S. Monitoring neonatal
seizures. Semin Fetal Neonatal Med. 2013;18:202– 8.
16. Pospelov AS, Puskarjov M, Kaila K, Voipio J. Endogenous
brain- sparing responses in brain pH and PO2 in a rodent model
of birth asphyxia. Acta Physiol. 2020;229:e13467.
17. Ala- Kurikka T, Pospelov AS, Summanen M, Alafuzoff A, Kurki
S, Voipio J, et al. A physiologically validated rat model of term
birth asphyxia with seizure generation after, not during, brain
hypoxia. Epilepsia. 2021;62:908– 19.
18. Romijn HJ, Hofman MA, Gramsbergen A. At what age is the
developing cerebral cortex of the rat comparable to that of the
full- term newborn human baby? Early Hum Dev. 1991;26:61– 7.
19. Erecinska M, Cherian S, Silver IA. Energy metabolism in
mammalian brain during development. Prog Neurobiol.
20. Johne M, Römermann K, Hampel P, Gailus B, Theilmann W,
Ala- Kurikka T, et al. Phenobarbital and midazolam suppress
neonatal seizures in a non- invasive rat model of birth asphyxia
while bumetanide is ineffective. Epilepsia. 2021;62:920– 34.
21. ter Horst JP, de Kloet ER, Schächinger H, Oitzl MS. Relevance
of stress and female sex hormones for emotion and cognition.
Cell Mol Neurobiol. 2012;32:725– 35.
22. Vanderschuren LJ, Trezza V. What the laboratory rat has
taught us about social play behavior: role in behavioral devel-
opment and neural mechanisms. Curr Top Behav Neurosci.
23. Enquist M, Lind J, Ghirlanda S. The power of associative learn-
ing and the ontogeny of optimal behaviour. R Soc Open Sci.
24. Ramos A. Animal models of anxiety: do I need multiple tests?
Trends Pharmacol Sci. 2008;29:493– 8.
25. Steimer T. Animal models of anxiety disorders in rats and
mice: some conceptual issues. Dialogues Clin Neurosci.
26. Lezak KR, Missig G, Carlezon WA Jr. Behavioral meth-
ods to study anxiety in rodents. Dialogues Clin Neurosci.
27. Vorhees CV, Williams MT. Assessing spatial learning and mem-
ory in rodents. ILAR J. 2014;55:310– 32.
28. Fisher RS, Scharfman HE, Decurtis M. How can we iden-
tify ictal and interictal abnormal activity? Adv Exp Med Biol.
29. Kadam SD, D'Ambrosio R, Duveau V, Roucard C, Garcia-
Cairasco N, Ikeda A, et al. Methodological standards and in-
terpretation of video- electroencephalography in adult control
rodents. A TASK1- WG1 report of the AES/ILAE Translational
Task Force of the ILAE. Epilepsia. 2017;58(Suppl 4):10– 27.
30. de Haan M, Wyatt JS, Roth S, Vargha- Khadem F, Gadian D,
Mishkin M. Brain and cognitive- behavioural development after
asphyxia at term birth. Dev Sci. 2006;9:350– 8.
31. Bird CM, Burgess N. The hippocampus and memory: insights
from spatial processing. Nat Rev Neurosci. 2008;9:182– 94.
32. Moser MB, Moser EI. Distributed encoding and retrieval of spa-
tial memory in the hippocampus. J Neurosci. 1998;18:7535– 42.
33. Kheirbek MA, Drew LJ, Burghardt NS, Costantini DO,
Tannenholz L, Ahmari SE, et al. Differential control of learning
and anxiety along the dorsoventral axis of the dentate gyrus.
Neuron. 2013;77:955– 68.
34. Jimenez JC, Su K, Goldberg AR, Luna VM, Biane JS, Ordek
G, et al. Anxiety cells in a hippocampal- hypothalamic circuit.
Neuron. 2018;97:670– 83.
35. Scharfman HE. The dentate gyrus and temporal lobe epilepsy:
an "exciting" era. Epilepsy Curr. 2019;19:249– 55.
36. Cavarsan CF, Malheiros J, Hamani C, Najm I, Covolan L. Is
mossy fiber sprouting a potential therapeutic target for epi-
lepsy? Front Neurol. 2018;9:1023.
37. Rainaldi MA, Perlman JM. Pathophysiology of birth asphyxia.
Clin Perinatol. 2016;43:409– 22.
38. Hsu KS, Huang CC. Characterization of the anoxia- induced
long- term synaptic potentiation in area CA1 of the rat hippo-
campus. Br J Pharmacol. 1997;122:671– 81.
39. Zanelli SA, Rajasekaran K, Grosenbaugh DK, Kapur J. Increased
excitability and excitatory synaptic transmission during in vitro
ischemia in the neonatal mouse hippocampus. Neuroscience.
GAILUS et al.
40. Ruusuvuori E, Kaila K. Carbonic anhydrases and brain pH
in the control of neuronal excitability. Subcell Biochem.
41. Semple BD, Blomgren K, Gimlin K, Ferriero DM, Noble-
Haeusslein LJ. Brain development in rodents and humans:
identifying benchmarks of maturation and vulnerability to in-
jury across species. Prog Neurobiol. 2013;106– 107:1– 16.
42. Workman AD, Charvet CJ, Clancy B, Darlington RB, Finlay BL.
Modeling transformations of neurodevelopmental sequences
across mammalian species. J Neurosci. 2013;33:7368– 83.
43. Patel SD, Pierce L, Ciardiello AJ, Vannucci SJ. Neonatal enceph-
alopathy: pre- clinical studies in neuroprotection. Biochem Soc
Trans. 2014;42:564– 8.
44. Kang HJ, Kawasawa YI, Cheng F, Zhu Y, Xu X, Li M, et al.
Spatio- temporal transcriptome of the human brain. Nature.
45. Vanhatalo S, Palva JM, Andersson S, Rivera C, Voipio J, Kaila K.
Slow endogenous activity transients and developmental expres-
sion of K+- Cl- cotransporter 2 in the immature human cortex.
Eur J Neurosci. 2005;22:2799– 804.
46. Sedmak G, Jovanov- Milosevic N, Puskarjov M, Ulamec M,
Kruslin B, Kaila K, et al. Developmental expression patterns
of KCC2 and functionally associated molecules in the human
brain. Cereb Cortex. 2016;26:4574– 89.
47. Deerberg F, Rapp K, Rehm S, Pittermann W. Genetic and envi-
ronmental influences on lifespan and diseases in Han:Wistar
rats. Mech Ageing Dev. 1980;14:333– 43.
48. Shalev I, Caspi A, Ambler A, Belsky DW, Chapple S, Cohen HJ,
et al. Perinatal complications and aging indicators by midlife.
Pediatrics. 2014;134:e1315– 23.
49. Poulton R, Moffitt TE, Silva PA. The Dunedin Multidisciplinary
Health and Development Study: overview of the first 40 years,
with an eye to the future. Soc Psychiatry Psychiatr Epidemiol.
50. Andreollo NA, Santos EF, Araujo MR, Lopes LR. Rat's age ver-
sus human's age: what is the relationship? Arq Bras Cir Dig.
51. Ferando I, Faas GC, Mody I. Diminished KCC2 confounds
synapse specificity of LTP during senescence. Nat Neurosci.
52. Danzer S. Mossy fiber sprouting in the epileptic brain: taking
on the Lernaean Hydra. Epilepsy Curr. 2017;17:50– 1.
53. Smith AL, Hill CA, Alexander M, Szalkowski CE, Chrobak JJ,
Rosenkrantz TS, et al. Spatial working memory deficits in male
rats following neonatal hypoxic ischemic brain injury can be
attenuated by task modifications. Brain Sci. 2014;4:240– 72.
54. Schreglmann M, Ground A, Vollmer B, Johnson MJ. Systematic
review: long- term cognitive and behavioural outcomes of neo-
natal hypoxic- ischaemic encephalopathy in children without
cerebral palsy. Acta Paediatr. 2020;109:20– 30.
55. Ikeda T, Mishima K, Yoshikawa T, Iwasaki K, Fujiwara M, Xia
YX, et al. Selective and long- term learning impairment follow-
ing neonatal hypoxic- ischemic brain insult in rats. Behav Brain
Res. 2001;118:17– 25.
56. Oorschot DE, Voss L, Covey MV, Bilkey DK, Saunders SE.
ADHD- like hyperactivity, with no attention deficit, in adult rats
after repeated hypoxia during the equivalent of extreme prema-
turity. J Neurosci Methods. 2007;166:315– 22.
57. Miguel PM, Schuch CP, Rojas JJ, Carletti JV, Deckmann I,
Martinato LH, et al. Neonatal hypoxia- ischemia induces
attention- deficit hyperactivity disorder- like behavior in rats.
Behav Neurosci. 2015;129:309– 20.
58. Ghei SK, Zan E, Nathan JE, Choudhri A, Tekes A, Huisman TA,
et al. MR imaging of hypoxic- ischemic injury in term neonates:
pearls and pitfalls. Radiographics. 2014;34:1047– 61.
59. Pisani F, Facini C, Pavlidis E, Spagnoli C, Boylan G. Epilepsy
after neonatal seizures: literature review. Eur J Paediatr Neurol.
60. Watanabe K, Kuroyanagi M, Hara K, Miyazaki S. Neonatal sei-
zures and subsequent epilepsy. Brain Dev. 1982;4:341– 6.
61. Pisani F, Piccolo B, Cantalupo G, Copioli C, Fusco C, Pelosi A,
et al. Neonatal seizures and postneonatal epilepsy: a 7- y fol-
low- up study. Pediatr Res. 2012;72:186– 93.
62. Rakhade SN, Klein PM, Huynh T, Hilario- Gomez C, Kosaras B,
Rotenberg A, et al. Development of later life spontaneous sei-
zures in a rodent model of hypoxia- induced neonatal seizures.
Epilepsia. 2011;52:753– 65.
63. Polikov VS, Tresco PA, Reichert WM. Response of brain tissue
to chronically implanted neural electrodes. J Neurosci Methods.
64. Groothuis J, Ramsey NF, Ramakers GM, van der Plasse G.
Physiological challenges for intracortical electrodes. Brain
Stimul. 2014;7:1– 6.
65. Scharfman HE, Myers CE. Hilar mossy cells of the den-
tate gyrus: a historical perspective. Front Neural Circuits.
66. Buckmaster PS. Does mossy fiber sprouting give rise to the epi-
leptic state? Adv Exp Med Biol. 2014;813:161– 8.
67. Hendricks WD, Westbrook GL, Schnell E. Early detonation by
sprouted mossy fibers enables aberrant dentate network activ-
ity. Proc Natl Acad Sci U S A. 2019;116:10994– 9.
68. Lippman- Bell JJ, Rakhade SN, Klein PM, Obeid M, Jackson
MC, Joseph A, et al. AMPA receptor antagonist NBQX attenu-
ates later- life epileptic seizures and autistic- like social deficits
following neonatal seizures. Epilepsia. 2013;54:1922– 32.
69. Rakhade SN, Zhou C, Aujla PK, Fishman R, Sucher NJ,
Jensen FE. Early alterations of AMPA receptors mediate syn-
aptic potentiation induced by neonatal seizures. J Neurosci.
70. Crusio WE, Schwegler H, Lipp HP. Radial- maze performance
and structural variation of the hippocampus in mice: a correla-
tion with mossy fibre distribution. Brain Res. 1987;425:182– 5.
71. Schwegler H, Crusio WE, Lipp HP, Heimrich B. Water- maze
learning in the mouse correlates with variation in hippocampal
morphology. Behav Genet. 1988;18:153– 65.
72. Johne M, Käufer C, Römermann K, Gailus B, Gericke B,
Löscher W. A combination of phenobarbital and the bumeta-
nide derivative bumepamine prevents neonatal seizures and
subsequent hippocampal neurodegeneration in a rat model of
birth asphyxia. Epilepsia. 2021;62:1460– 71.
73. Towfighi J, Mauger D, Vannucci RC, Vannucci SJ. Influence of
age on the cerebral lesions in an immature rat model of cere-
bral hypoxia- ischemia: a light microscopic study. Brain Res Dev
Brain Res. 1997;100:149– 60.
74. Williams PA, Dou P, Dudek FE. Epilepsy and synaptic reorgani-
zation in a perinatal rat model of hypoxia- ischemia. Epilepsia.
75. Kadam SD, Dudek FE. Neuropathogical features of a rat model
for perinatal hypoxic- ischemic encephalopathy with associated
epilepsy. J Comp Neurol. 2007;505:716– 37.
GAILUS et al.
76. Björkman ST, Miller SM, Rose SE, Burke C, Colditz PB. Seizures
are associated with brain injury severity in a neonatal model of
hypoxia- ischemia. Neuroscience. 2010;166:157– 67.
77. Kang SK, Kadam SD. Neonatal seizures: impact on neurodevel-
opmental outcomes. Front Pediatr. 2015;3:101.
78. Quinlan SMM, Rodriguez- Alvarez N, Molloy EJ, Madden SF,
Boylan GB, Henshall DC, et al. Complex spectrum of pheno-
barbital effects in a mouse model of neonatal hypoxia- induced
seizures. Sci Rep. 2018;8:9986.
79. Kadam SD, White AM, Staley KJ, Dudek FE. Continuous elec-
troencephalographic monitoring with radio- telemetry in a rat
model of perinatal hypoxia- ischemia reveals progressive post-
stroke epilepsy. J Neurosci. 2010;30:404– 15.
80. Kang SK, Ammanuel S, Thodupunuri S, Adler DA, Johnston
MV, Kadam SD. Sleep dysfunction following neonatal ischemic
seizures are differential by neonatal age of insult as determined
by qEEG in a mouse model. Neurobiol Dis. 2018;116:1– 12.
81. Korzeniewski SJ, Slaughter J, Lenski M, Haak P, Paneth N. The com-
plex aetiology of cerebral palsy. Nat Rev Neurol. 2018;14:528– 43.
82. Low JA. Relationship of fetal asphyxia to neuropathology and
deficits in children. Clin Invest Med. 1993;16:133– 40.
Additional supporting information may be found online
in the Supporting Information section.
How to cite this article: Gailus B, Naundorf H,
Welzel L, Johne M, Römermann K, Kaila K, et al.
Long- term outcome in a noninvasive rat model of
birth asphyxia with neonatal seizures: Cognitive
impairment, anxiety, epilepsy, and structural brain
alterations. Epilepsia. 2021;00:1– 19. https://doi.