Increased precipitation of spasms in an animal model of infantile spasms by prenatal stress exposure

    Increased precipitation of spasms in an animal model of infantile spasms by prenatal stress exposure Xiu-Yu Shi, Jun Ju, Li-Ping Zou, Juan Wang, Ning-Xiu Shang, Jian-Bo Zhao, Jing Wang, Jun-Yan Zhang PII: DOI: Reference:

S0024-3205(16)30196-5 doi: 10.1016/j.lfs.2016.03.047 LFS 14830

To appear in:

Life Sciences

Received date: Revised date: Accepted date:

26 September 2015 24 March 2016 24 March 2016

Please cite this article as: Shi Xiu-Yu, Ju Jun, Zou Li-Ping, Wang Juan, Shang NingXiu, Zhao Jian-Bo, Wang Jing, Zhang Jun-Yan, Increased precipitation of spasms in an animal model of infantile spasms by prenatal stress exposure, Life Sciences (2016), doi: 10.1016/j.lfs.2016.03.047

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Increased precipitation of spasms in an animal model of infantile

PT

spasms by prenatal stress exposure

RI

Xiu-Yu Shia, Jun Jua, Li-Ping Zoua,b*, Juan Wanga, Ning-Xiu Shanga, Jian-Bo Zhaoc, Jing

b c

MA

Department of Pediatrics, Chinese PLA General Hospital, Beijing China, 100853 Center of Epilepsy, Beijing Institute for Brain Disorders, Beijing China, 100069

Department of Neurology, Beijing Children’s Hospital, The Capital Medical University,

D

a

NU

SC

Wanga, Jun-Yan Zhangd

*

Department of Pediatrics, Beijing Haidian Hospital, Beijing China, 100080

AC CE P

d

TE

Beijing, China, 100000

Correspondence: Li-Ping Zou, Department of Pediatric, Chinese PLA General Hospital,

28 Fuxing Road, Beijing, 100853, P.R. China.

(Tel: +86-10-66939770; Fax: +86-10-66939770; E-mail: [email protected])

Word count:5228, Figure count:4

ACCEPTED MANUSCRIPT

Abstract Infantile spasms (IS) represent a serious epileptic syndrome, called West syndrome (WS)

PT

that occurs in the early infantile age. Although several hypotheses and animal models have been proposed to explain the pathogenesis of IS, the pathophysiology of IS has not

RI

been elucidated. Recently, we proposed a hypothesis for IS under prenatal stress exposure

SC

(also called Zou’s hypothesis) by correlating diverse etiologies and prenatal stresses with IS development. This research aims to determine the mechanism through which prenatal

NU

stress affects offspring and establish the potential underlying mechanisms. Pregnant rats were subjected to forced swimming in cold water. Rat pups exposed to prenatal stress

MA

were administered with N-methyl-D-aspartate (NMDA). Exposure to prenatal stress sensitized the rats against development of NMDA-induced spasms. However, this phenomenon was altered by administering adrenocorticotropin. Prenatal stress exposure

D

also altered the hormonal levels and neurotransmitter receptor expression of the

TE

developing rats as well as influenced the tissue structure of the brain. These findings suggest that maternal stress could alter the level of endogenous glucocorticoid, which is

AC CE P

the basis of IS, and cerebral dysplasia, hypoxic-ischemic encephalopathy (HIE), inherited metabolic diseases, and other factors activated this disease in developmental brain.

Keywords: Infantile spasms, prenatal stress, adrenocorticotropin, NMDA

ACCEPTED MANUSCRIPT 1. Introduction Infantile spasms (IS), also called West syndrome (WS), is an age-specific epileptic syndrome associated with many underlying conditions and poor developmental outcomes.

PT

Generally, IS occurs during the first two years of life (Lux and Osborne, 2004). Frost and Hrachovy (Frost and Hrachovy, 2005) studies the etiology of IS by summarizing 400

RI

research papers. The results indicate that many factors interfere with disequilibrium,

SC

which is consistent with age-dependence and multiple-cause nature of IS. However, this finding cannot explain why anti-seizure medications cannot effectively treat the disease,

NU

whereas adrenocorticotropin (ACTH) can produce the optimal effect (Mackay et al., 2004).

MA

Several models have been established to study pathogenic mechanisms and develop novel drugs. For example, intraperitoneal injection of N-methyl-D-aspartate (NMDA) into developing and young rats causes hyperflexion (emprosthotonic) seizures (Kabova et

D

al., 1999; Stafstrom and Sasaki–Adams, 2003). Moreover, the electroencephalogram

TE

(EEG) of the seizure consists of periods of suppression mixed with ictal activity of serrated waves and high-voltage chaotic activity (Mares and Velisek, 1992), which are

AC CE P

manifestations of IS; however, response to the therapeutic agent is incongruent with clinical experience. Velíšek et al. (Velisek et al., 2007) established an IS model of prenatally impaired brains induced by NMDA. The model shows that prenatal exposure to betamethasone sensitizes rats to development of NMDA-induced spasms and renders the spasms sensitive to ACTH therapy. Furthermore, ictal EEG results correspond to human IS (Velisek et al., 2007). There are also a few mouse models for IS, including administration of γ-aminobutyric acid agonists in a Down’s syndrome mouse and an Arx conditional knockout mouse model We found out that some mothers of patients with IS experienced prenatal stress events during long-term clinical work. Using a case control study, we found that the degree of prenatal stress was higher among mothers of patients with IS than that among the control group; within a certain range, the onset risk of IS increased with the degree of prenatal stress (Shang et al., 2010). Yum et al. (Yum et al., 2012) reported that prenatal stress significantly accelerated the onset and increased the number of NMDA-triggered spasms compared with the handled controls. The efficacy of hormone therapy on IS could

ACCEPTED MANUSCRIPT be due to endocrine changes induced by prenatal stress on infants. This fact suggests a final common pathway for these etiologies operating on the immature brain; thus, IS occur only at the maturational state during infancy (Baram et al., 1995). We propose the

PT

hypothesis that maternal stress could alter the levels of endogenous glucocorticoid, which is the basis of IS, and NMDA (induced epilepsy factors) could activate the disease in the

RI

developing brain. Thus, in the present research, experiments were performed on animals

SC

to study the correlation between prenatal stress and the pathogeny of IS and explain the

2. Methods and materials

MA

2.1. Animals and prenatal treatments

NU

underlying mechanism.

Three-month-old specific-pathogen free Wistar female (n = 31) and male (n = 10) rats,

D

weighing 275–345 g, were obtained from the Capital University of Medical Sciences

TE

Laboratory Animal Center. The rats were placed individually in standard polycarbonate cages (35 cm × 30 cm × 17 cm). The rats were also given with ad libitum access to

AC CE P

standard diet and water and kept under controlled conditions on a 12 h light cycle (lights off at 20:00). The temperature and relative humidity of the cage were set at 20  2 °C, and 50%–60%, respectively. The cages were cleaned weekly, but not more than 4 days after prenatal stress began (Drago et al., 1999). All procedures were approved by the authority for laboratory animal care and use. The rats were left undisturbed under the above conditions for a week to allow them adapt to the experimental environment. Female rats were mated with male rats. The presence of vaginal plug was considered as gestational day 1. The pregnant dams were randomly assigned to different experiments and then to repeated forced cold swimming during pregnancy. The remaining pregnant dams were non-stressed. Pregnant rats in the stress group were subjected to physical stress consisting of forced immersion in water maintained at 4 °C in Plexiglass cylinders (50 cm in height; 20 cm in diameter). The water depth was maintained at 35 cm to prevent the rat standing up without touching the bottom with its tail. After 5 min in water, the rats were removed and allowed to dry for 10 min in a heated container before returning to their home cages. Stress was administered daily at 17:00 from day 1 of

ACCEPTED MANUSCRIPT pregnancy until parturition. The number of offspring was counted (n = 312), and their sex was identified. Offspring of mothers that received forced cold swimming during pregnancy were

PT

categorized into the stressed mother group (M-stress), whereas those of non-stressed mothers were classified into the normal mother group (M-normal). A collection of male

RI

and female pups were placed in a separate group (Fig. 1). The weights of the rats were

SC

monitored throughout the experiments. Most pups were kept with their dams until they reach 13 days of age. However, pups at an older age were kept with their dams until they

MA

male and female animals were combined.

NU

reach 21 days of age. Initial statistical evaluation found no differences in sex; therefore,

2.2. NMDA rat model and ACTH treatments

D

Day of birth was considered postnatal 1 (P1). Prenatally stressed (n = 30) and prenatally

TE

non-stressed rats (n = 30) were randomly assigned on P13. The rats were injected intraperitoneally with NMDA (7 mg/kg) to induce spasm-like attacks and continuously

AC CE P

monitored for 3 h. The severity of their attacks was scored (from 0 to 9) and evaluated using the following rule. The latency to the onset of the first occurrence of tail twisting was recorded. Grade 0: no seizures, normally active; grade 1: stable and puffing; grade 2: hyperactive, irritative, twisting tail continuously; grade 3: bothering itself and others; grade 4: repeated flexion, n ≤ 4; grade 5: repeated flexion, 5 ≤ n ≤ 14; grade 6: repeated flexion, 15 ≤ n ≤ 29; grade 7: repeated flexion, n ≥ 30; grade 8: extremities move like thrashing after tonic-clonic seizures; and grade 9: death (Fathollahi et al., 1997). Pups in the prenatally stressed and non-stressed groups on P13 were randomly assigned to ACTH treatment (dose of 20 IU/kg, i.p. full ACTH molecule) to examine the curative effect on attacks induced by NMDA. Animal-derived natural ACTH (Shanghai Pharmaceutical Co., Ltd. China; each 25 IU/2ml ampoule loaded) was dissolved in saline and administered intraperitoneally 30 min before the NMDA treatment. The control group received two intraperitoneal injections of the equivalent amount of saline.

2.3. Plasma hormone measurement

ACCEPTED MANUSCRIPT

Blood samples for ACTH and cortisol analyses were obtained at 9 AM on P1, P7, P14, P21, P28, P35, and P42. Blood samples were collected in prechilled tubes containing

PT

ethylenediaminetetraacetic acid, centrifuged, and aliquoted promptly. Plasma was stored at –70 °C until assayed. Plasma cortisol was measured in the Radioimmunoassay

RI

Inspection Center, Beijing Pediatric Research Institute. Cortisol was directly measured

intra-assay

CV

of

<

SC

by radioimmunoassay with inter-assay coefficient of variation (CV) of 8.3% and 4.5%

(DiaSorin,

Inc.,

Stillwater,

MN).

The

NU

ACTH-immunoradiometric assay used paired monoclonal and polyclonal antibodies, which were reacted with the N terminal and C terminal regions of ACTH, respectively.

D

2.4. Radioligand binding assay

MA

The intra-assay and inter-assay CVs were 6.2% and 10.5%, respectively.

TE

Rats were sacrificed by decapitation on P13, and the brains were quickly removed and placed on an ice-cold glass plate. The forebrain (hippocampus included) was rapidly

AC CE P

dissected, frozen, and stored in a deep freezer at –20 °C until assayed. Binding to NMDA and gamma-aminobutyric acid (GABA) receptor was performed using [3H]-MK-801 and [3H]-muscimol, respectively. Rat brain membranes were routinely prepared from the brains, and binding assays were conducted using previously described procedures (Negro et al., 1995; Yoneda and Ogita, 1991). Briefly, the membranes (circa 0.5 mg protein) were incubated with [3H]-muscimol (20 nM final concentration) in darkness at 4 °C for 30 min. To determine [3H]MK-801 binding, we incubated an aliquot (0.3 mg of protein) of the prepared membranes in the presence or absence of glutamate (10 M), glycine (10 M), and spermidine (1 mM) with 5 nM (+)[3-3H MK-801 (22.5 Ci/mmol; NEN Life Science Products, Boston, MA, USA) in a total volume of 0.5 ml of 50 mM Tris-acetate buffer (pH 7.4) at 30 °C for 16 h. The filters were then incubated with sodium dodecyl sulfate to dissolve the membranes and release any trapped tritium. Radioactivity was measured using FJ-2100 scintillation counter.

ACCEPTED MANUSCRIPT 2.5. Immunohistochemistry and electron microscopy

PT

Anesthetized rats were sequentially perfused through the ascending aorta with 1000 units/ml heparin in normal saline on P13, followed by 40 ml of a mixture of 3.75%

RI

acrolein and 2% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). The brains were removed, postfixed for 30 min, and then cut by using a vibratome. The sections were

SC

collected in phosphate buffer.

For immunohistochemistry, brain tissues were cut into 5 µm-thick sections, rinsed

NU

with 0.01 M phosphate buffered saline two times, and treated with hydrochloric acid for 30 min. The sections were blocked with 5% normal goat serum for 30 min and incubated

MA

with the primary rabbit anti-NMDA receptor antibodies (NMDAR1 1:50, NMDAR2A 1:100, NMDAR2B 1:100; Chemicon International, Inc.) for 48 h at room temperature.

D

Biotinylated goat anti-rabbit IgG was used as secondary antibody (Beijing Biosynthesis

TE

Biotechnology) and applied for 3 h at room temperature and overnight at 4 °C. Positive reactions were visualized in diaminobenzidine (DAB) solution (Beijing Biosynthesis containing

0.1%

H2O2

at

room

temperature.

No

NMDAR

AC CE P

Biotechnology)

immunoreactivity was detected in the controls without the primary antibody. The number of NMDAR immunoreactive cells in the hippocampus, cerebral cortex, hypothalamus, and brain stem nuclear ridge were counted by image analysis. Images were captured using a Leica DCRE microscope with Leica DC500 camera and Leica IM50 Image Manager software (Leica Microsystems, Wetzlar, Germany). The number of cells was counted at 20× magnification in the entire section areas. For electron microscopy, the sections were fixed in 2% osmium tetroxide in PB (1 h), dehydrated, and embedded in EmBed 812 between two plastic sheets. Ultrathin sections (50 nm thick) from the tissue–plastic interface of each flat-embedded section through the area of interest were collected on grids and counterstained with uranyl acetate and lead citrate. Final preparations were examined with JEOL 1200 EXII transmission electron microscope.

2.6. Statistical analysis

ACCEPTED MANUSCRIPT

Statistically analyzed data were expressed as mean ± SEM. All statistical analyses were performed using SSPS11.0 for MacOS10 software. Comparisons of continuous variables

PT

among the groups (more than two) were performed by one-way ANOVA. We analyzed

RI

the results on seizure with two-way ANOVA with prenatal stress versus non-stress as well

SC

as postnatal ACTH versus non-ACTH.

3. Results

NU

3.1 NMDA induced spasm-like seizure in rats and prenatal stress increased

MA

sensitivity to NMDA

Intraperitoneal injection of NMDA into rat pups elicited epilepsy-like symptoms. Generally, curling spasms (rat curls into ball-like shape) were observed. The specific

D

characteristics of the curling spasms included spine curling, head meets the tail, paws curl

TE

against the body, entire body forms a ball-like shape, spasms occur singularly or continuously, and occurrence of tonic seizures or even death. We evaluated the curling

AC CE P

spasms in each pup (Fig. 2A). The semiology of NMDA symptoms was similar between prenatally stressed and naive rat pups, although prenatal stress increased the sensitivity of the rat to NMDA. Seizure latencies, clinical score, and mortality of NMDA-induced spasm seizures significantly differed between the group prenatally exposed to stress (M-stress n = 30) and the prenatally naive group (M-normal n = 30) (P < 0.001, Fig. 2B).

3.2 ACTH reduced the attack and mortality in prenatal stress-exposed NMDA rat pups

In prenatally stress-exposed rat pups, the ACTH-treated NMDA-induced spasms in rats significantly differed in latency, number of spasms and mortality from the saline-treated NMDA-induced spasms (Fig. 2C). Eliminating the confounding effect of stress factors is necessary to determine the effect of ACTH on seizure. Therefore, we analyze the data with two-way ANOVA, which can be used when data follow the normal distribution with homogeneity of variance. However, the present data were not normally

ACCEPTED MANUSCRIPT distributed; hence, the Friedman rank test was adopted, with CMH as test statistics. The results showed that after controlling the influence of stress factors, ACTH significantly delayed latency compared with saline (CMH=34.7, P <0.0001). After controlling the

PT

influence of stress factors, ACTH also decreased the number of spasms (CMH = 23.3, P < 0.0001). To analyze mortality data, we used CMH Chi-square test. After controlling the

RI

influence of stress factors on mortality, the difference between the effect of ACTH and

SC

that of saline on mortality is statistically significant (χ2CMH = 5.5, P = 0.02); this finding

NU

indicated that ACTH significantly reduced the risk of death (Table 1).

MA

3.3. Prenatal stress exposure altered the hormonal levels of developing rats

We compared the serum cortisol concentration on P1, P7, P14, P21, P28, P35, and P42-day-old rats (n = 5 in each time point) in the prenatally exposed, stressed (M-stress

D

group), and control groups (M-normal group). Significant differences were observed

TE

between the control group and M-stress group on P21 and P28 (*P < 0.05) (Fig. 3A). ACTH plasma levels of rats on P1, P7, P14, P21, P28, P35, and P42 (n = 5 in each

AC CE P

time point) in the prenatally exposed stressed groups (M-stress group) were compared with those in the control group. The physiological levels of ACTH decreased in the two groups, namely, the M-normal group and M-stress group on P28-day-old rats. The difference in ACTH plasma levels between the M-normal group and M-stress group in P35-day-old rats was P < 0.05. The lowest level of ACTH plasma in M-stress groups differed from that in the M-normal group (*P <0.05, Fig. 3B).

3.4 Prenatal stress exposure altered neurotransmitter receptor expression

[3H]-Muscimol, a potent GABA agonist used to label GABA receptor sites in brain specifically bound to sites in both groups. Quantitative receptor autoradiography was used to investigate the distribution of high-affinity GABA receptors (GABAA) in the forebrain of P13 rats (n = 7). The binding capacity of [3H]-muscimol (2615 ± 693) significantly decreased in rats in the M-stress group compared with that in the M-normal group (5322 ± 847) (t = 2.403, P = 0.04). [3H]-MK-801, a potent NMDA antagonist used

ACCEPTED MANUSCRIPT to label NMDA receptor sites in the forebrain of P13 rats (n = 7) specifically bound to sites in both groups. The binding capacity of [3H]-MK-801 significantly increased in rats in the M-stress (10537 ± 1838) compared with that in the M-normal group (n = 7, 4927 ±

PT

904) (Fig. 4B, t = 2.591, **P = 0.025). These results suggest that rats in the group exposed to prenatal stress may have modulated [3H]-muscimol and [3H]-MK-801

RI

binding in the brain. For NMDAR subunits, three NMDA receptor subunits were present

SC

in the brain tissues of rats sacrificed on P13. Analysis of the mean number of NMDAR1 (Fig. 4A-1), NMDAR2A (Fig. 4A-2), and NMDAR2B (Table 2, Fig. 4A-3) indicated the

NU

number of cells per 100-micrometer squared sections in rats in the M-stress (n = 6) and M-normal groups (n = 6). Rats prenatally exposed to stress showed altered expression of

MA

NMDA receptor subunits. The M-stress group had a significantly higher number than the M-normal groups in extensive NMDAR1 and NMDAR2B expression of hippocampus, cortex, hypothalamus, and brainstem, as well as extensive NMDAR2A expression of

D

hippocampus (Fig. 4A1-3). However, NMDAR2A expression was not significantly

TE

different among the cortex, hypothalamus, and brainstem.

AC CE P

3.5 Prenatal stress exposure affects tissue structure of the brain

To examine whether prenatal stress exposure affects the tissue structure of the brain, we conducted an experiment where histopathological findings represent the sections of the hippocampus stained with cresyl fast violet in a 13-day-old rat in the prenatally exposed stressed group (M-stress, n = 6), with corresponding sections in the prenatally naive infant rat group (M-normal, n = 6). Two groups of rats showed hippocampal cells without necrosis compared with the two groups with non-discrimination. The CA1 pyramidal cell layer is shown at higher magnification on the right side (Fig. 4C). Swollen mitochondria, reduced cristae, and disappearance of cristae broken ridge phenomenon were observed in P13-day-old rats in the prenatally exposed stressed group (M-stress, n = 3) compared with corresponding sections in the prenatally naive infant rat group (M-normal, n = 3) (Fig. 4D).

4. Discussion

ACCEPTED MANUSCRIPT

The proposed “prenatal stress exposure hypothesis for IS” is derived from the combination of three hypotheses for IS, namely, corticotropin releasing hormone (CRH)

PT

hypothesis, NMDA hypothesis, and serotonin-kynurenine hypothesis (Rho, 2004). We used our hypothesis to explain the pathogenesis of the disease. The results elucidate how

RI

prenatal stress affects offspring and present the potential underling mechanism.

SC

Regarding the influence of prenatal stress exposure on the endocrine system of the offspring, the recently increasing number of studies on animals and humans indicates that

NU

excessive prenatal stress exposure undermines the mother’s adjustment function of hypothalamus-pituitary-adrenal (HPA) axis, and causes too much glucocorticoid to be secreted. It further affects the endocrine system of the fetus, and finally interferes with

MA

their behavior and physiogenesis (Gotz and Stefanski, 2007; Zou et al., 2006). Our research found that maternal stress results in decrease in cortisol and ACTH plasma levels

D

in the offspring, which are important hormones in the HPA axis. Furthermore, such effect

TE

could last four weeks postnatal. The possible mechanism is that too much glucocorticoid and catecholamine produced by the mother cause the uterine placenta’s blood to decrease

AC CE P

and further cause the fetus to lack nutrition; thereby affecting its HPA-axis (de Weerth and Buitelaar, 2005). The fetus has a limited ability to circulate glucocorticoid. Therefore, even a small amount of glucocorticoid flowing into the fetus blood would cause a significant change in the concentration of glucocorticoid, and particularly affect the function of HPA axis of the fetus (Murphy et al., 1974). Brunson et al. believe that the excessive CRH related to prenatal stress exposure is crucial to the occurrence of IS (Brunson et al., 2002). Exposure to prenatal stress elicits damages in hippocampal neuron and neuron ultrastructure of the offspring. Changes in the hippocampal CA1 neuron ultrastructure of rat pup exposed to prenatal stress were observed in the electron microscopy analysis. Injuries, such as mitochondrion swelling, unclear boundary, uneven electron density, enhanced lipofuscin, and nuclei deformation could be due to excessive growth of oxides. Prenatal stress exposure could change the formation and differentiation of the hippocampal neuron of offspring (Fujioka et al., 2006). This phenomenon could also result in high concentrations of NMDA receptors in the cerebral cortex and basal nuclei.

ACCEPTED MANUSCRIPT These findings differ from previous reports, in which maternal stress results in substantial decrease in NR1 and NR2B subunits of the NMDA receptor in offspring when they become adults (Son et al., 2006; Yaka et al., 2007). One of the possible reasons for this

PT

finding is animal age; the present experiment used 13-day-old rats, whereas Son and Yaka used adult animals. The expression of the NMDA receptor is known to be age dependent.

RI

Yang et al. (Yang et al., 2006) found that providing chronic and irregular electric

SC

stimuli to rats in mid-late gestational period could aggravate morphine-induced addictive behavior, and adult offspring showed abnormal behavior during the forced swimming test.

NU

To block the NMDA receptor from the effect of stress exposure, they participate in the formation of the emotional handicap of the offspring (Saal et al., 2003) during the early

MA

period after young rats are born; this finding is consistent with that in the formation period of high amounts of synapses. The peak level of rats appears in the second week after young rats are born; thus, the expression of the NMDA receptor increased

D

remarkably (Represa et al., 1989). In the cortex and hippocampus, GluN2B expression is

TE

initially high in neurons, but decreases during development as the expression of GluN2A increases (Monyer et al., 1994). Several studies focused on the effects of prenatal stress

AC CE P

on the expression of NMDA receptor subunits; some researchers associated the outcome to learning and synaptic plasticity, which in fact, may contribute to poor developmental outcomes in IS.

Frost et al. (Frost and Hrachovy, 2005) concluded that about 200 factors are related to IS, thereby indicating that many brain injuries can increase the onset risk of IS. We believe that all factors are derived from excessively hyper-irritable NMDA receptors, which increase cell calcium ion influx and thus activate a series of enzymes (kinase, prolease, nitric oxide synthase, phosphorylase, etc). Moreover, these factors could enhance the generation of active oxygen substances, resulting in damage to cells and death of programmed cells (apoptosis) (Rho, 2004). In conclusion, prenatal stress exposure sensitizes rats to development of NMDA-induced spasms and renders the experimental model sensitive to ACTH therapy (Yum et al., 2012). These findings verify the hypothesis of Zou (prenatal stress exposure hypothesis for IS), that is, cerebral dysplasia, inherited metabolic diseases, and other factors can lead to excessive expression of the NMDA receptor and trigger IS on the basis

ACCEPTED MANUSCRIPT of the adverse events in the gestational period; these factors affect the growth of the neuroendocrine system of the offspring. Our work presents a mechanism underlying the induction of the disease established a new model that can be used to explain the prevalent

PT

three hypotheses (CRH hypothesis, NMDA hypothesis, and serotonin-kynurenine hypothesis). This study also provides a basis for outbreak of IS and elucidate why ACTH

SC

RI

is the most effective treatment for the disease.

Conflict of interest

NU

None

MA

Acknowledgments

We thank that Zuo PP, Lan ZQ for technical assistance and Pan for statistical assistance This study was supported by the grants from Major State Basic Research Development

D

Program (973; no. 2012CB517903), the National Natural Science Foundation of China

TE

(nos. 81471329, 81211140048, 81201013, 81200463), and the Beijing Municipal Natural

AC CE P

Science Foundation (nos. 7081002 and 7042024).

ACCEPTED MANUSCRIPT

References: Baram, T.Z., Mitchell, W.G., Hanson, R.A., Snead, O.C., 3rd, Horton, E.J., 1995. Cerebrospinal fluid

PT

corticotropin and cortisol are reduced in infantile spasms. Pediatr Neurol. 13, 108-10. Brunson, K.L., Avishai-Eliner, S., Baram, T.Z., 2002. ACTH treatment of infantile spasms: mechanisms of

RI

its effects in modulation of neuronal excitability. Int Rev Neurobiol. 49, 185-97. de Weerth, C., Buitelaar, J.K., 2005. Physiological stress reactivity in human pregnancy--a review. Neurosci

SC

Biobehav Rev. 29, 295-312.

Drago, F., Di Leo, F., Giardina, L., 1999. Prenatal stress induces body weight deficit and behavioural

NU

alterations in rats: the effect of diazepam. Eur Neuropsychopharmacol. 9, 239-45. Fathollahi, Y., Motamedi, F., Semnanian, S., Zardoshti, M., 1997. Examination of persistent effects of repeated administration of pentylenetetrazol on rat hippocampal CA1: evidence from in vitro study

MA

on hippocampal slices. Brain Res. 758, 92-8.

Frost, J.D., Jr., Hrachovy, R.A., 2005. Pathogenesis of infantile spasms: a model based on developmental desynchronization. J Clin Neurophysiol. 22, 25-36.

D

Fujioka, A., Fujioka, T., Ishida, Y., Maekawa, T., Nakamura, S., 2006. Differential effects of prenatal stress

TE

on the morphological maturation of hippocampal neurons. Neuroscience. 141, 907-15. Gotz, A.A., Stefanski, V., 2007. Psychosocial maternal stress during pregnancy affects serum corticosterone, blood immune parameters and anxiety behaviour in adult male rat offspring. Physiol Behav. 90,

AC CE P

108-15.

Kabova, R., Liptakova, S., Slamberova, R., Pometlova, M., Velisek, L., 1999. Age-specific N-methyl-D-aspartate-induced seizures: perspectives for the West syndrome model. Epilepsia. 40, 1357-69.

Lux, A.L., Osborne, J.P., 2004. A proposal for case definitions and outcome measures in studies of infantile spasms and West syndrome: consensus statement of the West Delphi group. Epilepsia. 45, 1416-28.

Mackay, M.T., Weiss, S.K., Adams-Webber, T., Ashwal, S., Stephens, D., Ballaban-Gill, K., Baram, T.Z., Duchowny, M., Hirtz, D., Pellock, J.M., Shields, W.D., Shinnar, S., Wyllie, E., Snead, O.C., 3rd, American Academy of, N., Child Neurology, S., 2004. Practice parameter: medical treatment of infantile spasms: report of the American Academy of Neurology and the Child Neurology Society. Neurology. 62, 1668-81. Mares, P., Velisek, L., 1992. N-methyl-D-aspartate (NMDA)-induced seizures in developing rats. Brain Res Dev Brain Res. 65, 185-9. Murphy, B.E., Clark, S.J., Donald, I.R., Pinsky, M., Vedady, D., 1974. Conversion of maternal cortisol to cortisone during placental transfer to the human fetus. Am J Obstet Gynecol. 118, 538-41. Negro, M., Chinchetru, M.A., Fernandez, A., Calvo, P., 1995. Effect of ethanol treatment on rate and equilibrium constants for [3H] muscimol binding to rat brain membranes: alteration of two affinity

ACCEPTED MANUSCRIPT states of the GABAA receptor. J Neurochem. 64, 1379-89. Represa, A., Tremblay, E., Ben-Ari, Y., 1989. Transient increase of NMDA-binding sites in human hippocampus during development. Neurosci Lett. 99, 61-6. Rho, J.M., 2004. Basic science behind the catastrophic epilepsies. Epilepsia. 45 Suppl 5, 5-11.

PT

Saal, D., Dong, Y., Bonci, A., Malenka, R.C., 2003. Drugs of abuse and stress trigger a common synaptic adaptation in dopamine neurons. Neuron. 37, 577-82.

RI

Shang, N.X., Zou, L.P., Zhao, J.B., Zhang, F., Li, H., 2010. The Association between Prenatal Stress and Infantile Spasms: A Case-Control Study. Epilepsy Res. 42, 181-186.

SC

Son, G.H., Geum, D., Chung, S., Kim, E.J., Jo, J.H., Kim, C.M., Lee, K.H., Kim, H., Choi, S., Kim, H.T., Lee, C.J., Kim, K., 2006. Maternal stress produces learning deficits associated with impairment of

NU

NMDA receptor-mediated synaptic plasticity. J Neurosci. 26, 3309-18. Stafstrom, C.E., Sasaki-Adams, D.M., 2003. NMDA-induced seizures in developing rats cause long-term learning impairment and increased seizure susceptibility. Epilepsy Res. 53, 129-37.

MA

Velisek, L., Jehle, K., Asche, S., Veliskova, J., 2007. Model of infantile spasms induced by N-methyl-D-aspartic acid in prenatally impaired brain. Ann Neurol. 61, 109-19. Yaka, R., Salomon, S., Matzner, H., Weinstock, M., 2007. Effect of varied gestational stress on acquisition

D

of spatial memory, hippocampal LTP and synaptic proteins in juvenile male rats. Behav Brain Res.

TE

179, 126-32.

Yang, J., Li, W., Liu, X., Li, Z., Li, H., Yang, G., Xu, L., Li, L., 2006. Enriched environment treatment

AC CE P

counteracts enhanced addictive and depressive-like behavior induced by prenatal chronic stress. Brain Res. 1125, 132-7.

Yoneda, Y., Ogita, K., 1991. Heterogeneity of the N-methyl-D-aspartate receptor ionophore complex in rat brain, as revealed by ligand binding techniques. J Pharmacol Exp Ther. 259, 86-96. Yum, M.S., Chachua, T., Velíšková, J., Velíšek, L., 2012. Prenatal stress promotes development of spasms in infant rats. Epilepsia. 53, e46-9. Zou, L.P., Zhang, W.H., Wang, H.M., Zen, M., Chen, K., Mix, E., 2006. Maternal IgG suppresses NMDA-induced spasms in infant rats and inhibits NMDA-mediated neurotoxicity in hippocampal neurons. J Neuroimmunol. 181, 106-11.

ACCEPTED MANUSCRIPT

Figure legends

PT

Figure 1. Study design: Pregnant rats were subjected to forced swimming in cold water (M-stress) and non-stress (M-normal). Offspring were grouped into different experiments

SC

RI

at different ages.

Figure 2. NMDA induced spasm seizure in rats. Prenatal stress increased sensitivity

NU

to model.

MA

(A) NMDA induced spasms and assessment of clinical signs. (A-1) Grade 1: freezing with hyperventilation. (A-2) Grade 2: hyperactive, irritable, twisting tail continuously. (A-3) Grade 3: spine curling, head meets the tail. (A-4) Grade 4: paws curl against the

TE

D

body, the entire body forms a ball-like shape.

AC CE P

(B) Prenatal stress increased the sensitivity to NMDA. The differences in seizure latencies (m = minute), number of spasms and mortality of the NMDA-induced seizures between prenatally stress-exposed group (M-stress n = 30) and prenatally naive infant rats group (M-normal n = 30) were significant (P < 0.001).

(C) ACTH-treated NMDA-induced spasms in M-stress group (n=15) demonstrate a very significant delay of the latency, decreased number of spasms and mortality compared with the saline-treated M-stress group’s rats (n=30), P<0.001.

Figure 3. Prenatal stress exposure altered the hormonal levels of developing rats.

(A) Comparison of serum cortisol concentration of rats operated at P1, P7, P14, P21, P28, P35, and P42 days of age (n=5 in each time point) in the prenatally exposed to stress group (M-stress n=35) and control group (M-normal n=35). Significant differences are

ACCEPTED MANUSCRIPT observed between the M-normal group and M-stress group at P21 and P28. *P < 0.05.

PT

(B). Comparison of ACTH plasma levels of rats operated at P1, P7, P14, P21, P28, P35, and P42 days of age (n = 5 in each time point) in M-stress group (n = 35) and M-normal

RI

group (n=35). Significant differences are observed between the M-normal group and

NU

SC

M-stress group at P35. *P < 0.05.

Figure 4. Prenatal stress exposure altered density of central neurotransmitter

MA

receptors and mitochondrial structure.

(A) Prenatal stress exposure alters the expression of NMDA receptor subunits in their

D

brain tissue. The mean numbers of (A-1) NMDAR1, (A-2) NMDAR2A, and (A-3)

TE

NMDAR2B-expressing cells per 100 micrometer-squared hippocampus, cortex, hypothalamus, and brainstem sections from the prenatally exposed to stress group

AC CE P

(M-stress n=6) and prenatally naive infant rats group (M-normal n=6). Extensive NMDAR1 and NMDAR2B expression was detected in the hippocampus, cortex, hypothalamus, and brainstem sections; and NMDAR2A expression was detected in the hippocampus from the M-stress groups. In contrast, only some positive cells were seen in the sections from M-normal rats. The P values refer to comparisons with control animals. *P < 0.05, **P < 0.001.

(B) Binding to the NMDA receptor (n = 7) and GABA receptor (n = 7) was performed using [3H]-MK-801 and [3H] muscimol in the forebrain of the P13 rats. M-stress group rats showed decreased binding capacity of [3H]-muscimol and increased binding capacity of [3H]-MK-801; and the differences were found to be significant compared with the M-normal group rats. **P < 0.01

(C) Histopathological findings represent the sections of the hippocampus stained with cresyl fast violet from a rat operated at P13 of age in the M-stress group (n = 6), with

ACCEPTED MANUSCRIPT sections from the corresponding group of M-normal group (n = 6). In the two groups of rats, no hippocampal cells and necrotic cells were found compared with the two groups of

PT

non-discrimination.

(D) Transmission electron microscopy of the brainstem mitochondria. The mitochondria

RI

appears to be swollen, the cristae are reduced, and the cristae broken ridge phenomenon

AC CE P

TE

D

MA

NU

with sections from the M-normal group (n = 3).

SC

disappears in rats operated at P13 days of age in the M-stress group (n = 3) compared

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

Fig. 1

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

Fig. 2

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Fig. 3

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

Fig. 4

ACCEPTED MANUSCRIPT Table 1 Effect of prenatal exposure and ACTH treatment on seizure CMH: Cochran-Mantel-Haenszel statistics value based on Rank Score.

Stress

ACTH

15

Saline

30

16.5 (15,20)

ACTH

30

27.5 (23,30)

Saline

30

20.0 (18,23)

Outcome index Numbers of spasms(N) Median P (Q1,Q3) 10 (10,15)

PT

Latency(min) Median P (Q1,Q3) 26.0 (18,33)

40 (30,40)

MA D

<0.0001 (ACTH&Saline, CMH=23.3)

Death (N) 0

9

10 (2,22)

0

22 (10,22)

0

NU

<0.0001 (ACTH&Saline, CMH=34.7)

TE AC CE P

Non-stress

Numbers

RI

Treatment

SC

Prenatal exposure

Mortality P

0.02 (ACTH& Saline, 2 χ CMH=5.5)

ACCEPTED MANUSCRIPT

hippocampus

P NMDAR2B value M-stress M-normal

D TE AC CE P

68.6± 6.1 55.8± 8.3 0.009 36.1± 9.9 28.2± 3.8

RI

0.006 52.5±5.3 38.8±3.8

0.0004

MA

P value

cortex

SC

P value NMDAR2A M-stress M-normal

57.8±4.2 37.0±5.1 0.001 42.5±8.2 32.8±6.0

NU

NMDAR1

Group M-stress M-normal

PT

Table 2 The mean numbers of NMDAR1, NMDAR2A and NMDAR2B-expressing cells per 100 micrometers

0.067 57.8±3.9 40.4±2.6

0.0002

hypothalamus

brainstem

51.5±2.8 46.4±4.3 0.015 30.3±8.4 29.5±4.8

77.7±12.5 47.1±4.7 0.001 43.1±5.9 40.8±4.1

0.7 53.5±3.9 31.8±3.7

0.5 60.9±5.4 38.8±3.8

0.0003

0.0002