Palmitoylethanolamide attenuates neurodevelopmental delay and early hippocampal damage following perinatal asphyxia in rats

Impaired gas exchange close to labor causes perinatal asphyxia (PA), a neurodevelopmental impairment factor. Palmitoylethanolamide (PEA) proved neuroprotective in experimental brain injury and neurodegeneration models. This study aimed to evaluate PEA effects on the immature-brain, i.e., early neuroprotection by PEA in an experimental PA paradigm. Newborn rats were placed in a 37°C water bath for 19 min to induce PA. PEA 10 mg/kg, s.c., was administered within the first hour of life. Neurobehavioral responses were assessed from postnatal day 1 (P1) to postnatal day 21 (P21), recording the day of appearance of several reflexes and neurological signs. Hippocampal CA1 area ultrastructure was examined using electron microscopy. Microtubule-associated protein 2 (MAP-2), phosphorylated high and medium molecular weight neurofilaments (pNF H/M), and glial fibrillary acidic protein (GFAP) were assessed using immunohistochemistry and Western blot at P21. Over the first 3 weeks of life, PA rats showed late gait, negative geotaxis and eye-opening onset, and delayed appearance of air-righting, auditory startle, sensory eyelid, forelimb placing, and grasp reflexes. On P21, the hippocampal CA1 area showed signs of neuronal degeneration and MAP-2 deficit. PEA treatment reduced PA-induced hippocampal damage and normalized the time of appearance of gait, air-righting, placing, and grasp reflexes. The outcome of this study might prove useful in designing intervention strategies to reduce early neurodevelopmental delay following PA.


Introduction
Transient interruption of oxygen supply close to delivery causes an obstetrical complication known as perinatal asphyxia (PA) (Adcock and Papile, 2008). The estimated incidence of this life-threatening complication ranges from 1 to 8 up to 26 per 1,000 live births in developed and developing countries, respectively (Douglas-Escobar and Weiss, 2015). Neonatal care advances are so far unsuccessful in overcoming the impact of PA, which increases neonatal mortality, neurological morbidity, and neurodevelopmental disorders (NDDs) (Weitzdoerfer et al., 2004;Herrera-Marschitz et al., 2014). The extensively used Bjelke's experimental model (Bjelke et al., 1991) has allowed to study PA neuropathological and behavioral effects (Barkhuizen et al., 2017). We have reported long-term PA-induced deficits (Capani et al., 2009;Galeano et al., 2011;Saraceno et al., 2010Saraceno et al., , 2012Muñiz et al., 2014). We observed behavioral alterations 1 month after experimental PA (Saraceno et al., 2016;Herrera et al., 2018), though little is known about the early impact of the perinatal insult (Horvath et al., 2015;Barkhuizen et al., 2017). Hence, we examined body weight gain and several signs of neurological maturation in asphyctic rats throughout the first 3 weeks of life, corresponding to the first 3 years of human development (Clancy et al., 2007;Semple et al., 2013), a critical period for neurotypical and aberrant neurodevelopment (Meredith, 2015). Developmental reflex testing concerns human infants, while their evaluation in rats provides a translational expression of perinatal injuries, offering genuine developmental traits (Moser, 2001;Nguyen et al., 2017).

Materials and methods
The experimental protocol was approved by the Institutional Animal Care and Use Committee of the University of Buenos Aires (CICUAL#4091/04). The experiments were conducted following the principles of the Guide for the Care and Use of Laboratory Animals (Animal Welfare Assurance, A-3033-01/protocol#S01084).
Only original figures are included in this manuscript. Some of them are cropped for space limitations and shown in full as Supplementary material.

Animals
Twenty pregnant Sprague-Dawley rats from the central vivarium of the School of Veterinary Sciences of the University of Buenos Aires arrived at our local vivarium for environmental adaptation 1 week before delivery.

Induction of perinatal asphyxia
Rat pups were subjected to PA using Bjelke et al.'s original model modified in our laboratory (Bjelke et al., 1991;Capani et al., 2009). This experimental paradigm induces severe asphyxia by submerging rat pups immediately upon delivery in a water bath at 37 • C. After 19 min, intermittent tactile stimulation is given until regular breathing restoration (Saraceno et al., 2010).

Neuroprotection protocol
Within the first hour of life, 75 male rat pups were injected with either vehicle (VHI, 1:1:8 solution of DMSO, Tween 80 and NaCl) or PEA 10 mg/kg (Herrera et al., 2018). Only male pups were used to avoid confounding variables due to estrogens neuroprotective properties (Saraceno et al., 2010). Four experimental groups were studied: rats subjected to PA injected with VHI (PA-VHI, n = 19), rats born vaginally (control, CTL) injected with VHI (CTL-VHI, n = 21), rats subjected to PA injected with PEA (PA-PEA group, n = 17), and rats born vaginally injected with PEA (CTL-PEA group, n = 18). As this model includes euthanasia administration to the mothers, rat pups in all the experimental groups were placed by surrogate mothers, which had delivered vaginally in the previous 24 h. Rats were identified according to the group and placed in the respective litters (Udovin et al., 2020).

Neurobehavioral development examination
Neurodevelopment was assessed from P1 to P21, e.g., during the first 3 weeks of life, between 12:00 and 15:00 p.m. Pups (N = 75) underwent daily weight control and testing of reflexes and signs, symptomatic of nervous system maturation (Kiss et al., 2009). The experimenter was blind to the groups, i.e., unaware of rat treatment.
• Surface righting reflex: pups were placed in the supine position. Time (seconds) to turn over to the prone position placing all four paws on the surface was recorded daily.
• Air-righting reflex: pups were dropped head down onto a bed of shavings from a height of 50 cm. The first day of landing on four paws was recorded (postnatal day of appearance).
• Gait: pups were placed at the center of a 13 cm diameter white paper circle. The test ended if the rat did not leave the circle after the first 30 s. Postnatal day of gait appearance, e.g., the first day the rat moved off the circle with both forelimbs was recorded. Thereafter, test performance was recorded, in seconds, daily.
• Forelimb placing reflex: the back of the forepaw of a suspended pup was touched with the bench edge. The first day of placing the paws on the table was recorded.
• Forelimb grasp reflex: forelimbs were touched with a thin rod. The first day of grasping onto the rod was recorded.
• Negative geotaxis: pups were placed head down, hindlimbs in the middle of a 45 • inclined 30 cm long grid. The test was ended if the rat did not turn round, climbed up the board with their forelimbs, and reached the upper rim within the first 30 s. The first day the rat so did was recorded as the postnatal day of appearance. Thereafter, negative geotaxis performance was recorded, in seconds, daily.
• Eye opening: the first day of both eyes' opening was recorded.
• Auditory startle reflex: the first day of the startle response to a clapping sound was recorded.
• Sensory eyelid reflex: the eyelid was gently touched with a cotton swab. The first day of eyelid contraction was recorded.
Rats are born altricial, so unable to perform complex behaviors. Reflex screening-level assessment appears the only testing available at very young ages (Moser, 2011;Nguyen et al., 2017).

Immunohistochemistry
Three coronal hippocampal sections were cut −480 mm to −530 mm from Bregma along the rostrocaudal axis (Paxinos and Watson, 2007) of each of four rats per group (Saraceno et al., 2016;Herrera et al., 2018). On P21, rats were anesthetized (ketamine 40 mg/kg + xylazine 5 mg/kg, i.p.), and intracardially perfused with 4% paraformaldehyde in 0.1 M phosphate buffer, pH = 7.4. Brains were dissected out and immersed in the same fixative solution at room temperature for 2 h, and in 0.1 M phosphate buffer, pH = 7.4 at 4 • C overnight. Coronal hippocampal sections (50 µm thick) were obtained (Vibratome VT 1000 S, Leica Microsystems, Wetzlar, Germany). Immunohistochemistry was performed on free-floating sections under moderate shaking. Endogenous peroxidase activity was quenched using a 0.3% hydrogen peroxide solution in methanol. Non-specific labeling was blocked with 0.3% normal goat serum diluted in phosphate-buffered saline (PBS) at room temperature (RT) for 1 h. Samples were PBS-washed 5 times for 10 min and incubated with anti-microtubule-associated protein 2 (MAP-2; 1:250, polyclonal rabbit-IgG; Abcam), anti-phosphorylated high and medium molecular weight neurofilaments (pNF H/M; 1:500, monoclonal mouse-IgG; Millipore), or anti-glial fibrillary acidic protein (GFAP; monoclonal rabbit IgG, 1:200, Cell Marque, a Sigma-Aldrich Company) diluted in 0.3% normal goat serum in PBS for 48 h at 4 • C. The following day, samples were PBS-washed 5 times for 10 min and incubated with horseradish peroxidase (HRP) biotinylated secondary antibody (Biotinylated anti-mouse-IgG, 1:500, Vector; Biotinylated antirabbit-IgG, 1:500, Vector) diluted in PBS at room temperature for 2 h. Then, samples were PBS-washed 5 times for 10 min and incubated with an avidin-biotinylated HRP complex (1:500, Dako) in PBS in darkness at RT for 1 h, followed by 5 washes with PBS 10 min. Finally, the sections were incubated at RT for 5 min in 0.05% diaminobenzidine (DAB, Sigma) diluted in Tris-HCl 0.05 M pH = 7.4, containing 0.03% H 2 O 2 for signal detection. After several running water-washes, the sections were transferred to a dish with 1× PBS for mounting. Glass slides were dipped into 1× PBS and a fine paintbrush was used to coax the sections gently towards the slide. After 1-h drying at RT, a drop of mounting medium (1:1 PBS: glycerol) was added barely to cover the tissue-section, and the coverslip was gradually placed starting with one edge against the slide and slowly releasing the coverslip nicely to avoid air bubbles. Finally, a thin nail polish layer was placed to seal the coverslip perimeter of and left to dry at RT (Bachman, 2013;Potts et al., 2020). Samples were observed using a digital camera-coupled Leica microscope, under constant light and brightness/contrast conditions. The images were processed and analyzed using ImageJ software (Image J 1.41o, NIH, United States). Antibody dilutions and DAB chromogen development time were unique for each protein staining. The intensity was determined in a blind fashion, using a semi-quantitative 0 to +++score.

Morphometric analysis
The percentage of immunopositive area for pNF H/M and MAP-2 was estimated by sampling 150 µm 2 per photomicrograph (ImageJ 1.41o, NIH, United States). The number of GFAP immunoreactive astrocytes was estimated in the CA1 hippocampal stratum radiatum area using the optical dissector method (Howard and Reed, 1998) with total section thickness for dissector height (Hatton and von Bartheld, 1999) and a 55 × 55 µm counting frame. A total of 78 counting frames per animal was assessed. Section thickness was measured using a microscope stage-attached digital length measuring device (Heidenhain-Metro MT 12/ND221; Traunreut, Germany). Every cell nucleus of GFAP-immunoreactive cells observed by focusing down through the height of the dissector was counted. Counts were performed on coded sections. Stratum radiatum volume in CA1 was estimated using the point count method (Weibel, 1979). Determinations were made by triplicate (Herrera et al., 2018).

Western blot
The animals were euthanized by decapitation at 21 days of age and whole brains were extracted from the skull (Chiu et al., 2007). Hippocampi were macroscopically dissected out and stored frozen at −80 • C. For protein extraction, specimens were thawed, homogenized in ice-cold lysis buffer (10 mM Tris/HCl, pH = 7.4, 10 mM NaCl, 3 mM MgCl 2 , 0.1% Triton X-100, protease inhibitors), and centrifuged at 14,000 rpm at 4 • C for 15 min. Supernatants were sampled and protein content was quantified by Bradford dosage in a 96-well plate assay using bovine serum albumin (BSA) as standard.

Statistical analysis
Results were expressed as mean ± SEM. Shapiro-Wilk and Levene's tests were used to check for normal distribution and equality of variances. Results underwent a two-way analysis of variance (ANOVAs) with birth condition (CTL or PA) and treatment (VHI or PEA) as main factors. For repeated measure variables like daily body weight and reflexes performance, a two-way ANOVA with group (CTL-VHI, PA-VHI, CTL-PEA, and PA-PEA) and day number as the main factors was used. Two-tailed Student's t-test, adjusted by the Bonferroni correction, was used for post hoc comparisons. p-Value ≤ 0.05 was considered statistically significant, e.g., the probability that the null hypothesis was correct and results were random (type I error, or false positive) was ≤5% (Graphpad Prism version 7.04).

Body weight gain in the first 3 weeks of life
Group (F 3,34 = 18.51, p < 0.0001) and postnatal day (F 20,680 = 3,938, p < 0.0001) were main sources of variation in daily body weight and showed interaction (F 60,680 = 6.634, p < 0.0001). Starting on P5, PA-VHI rats' daily body weight was lower than CTL-VHI rats' , but not different from PA + PEA rats at every time-point studied (Figure 1). All through the first weeks of life, CTL + PEA and CTL-VHI were indistinguishable based on body weight (Figure 1).

Neurobehavioral development over the first 3 weeks of life
The appearance day of the air-righting reflex was affected by PA and PEA treatment (Figure 2A). Birth condition (F 1,66 = 120.6, p < 0.0001) and treatment (F 1,66 = 120.6, p < 0.0001) were main sources of variation, showing interaction (F 1,66 = 157.5; p < 0.0001). The air-righting reflex appeared later in PA-VHI than in CTL-VHI rats (p < 0.0001), earlier in PA + PEA rats than in PA-VHI rats (p < 0.0001), while CTL + PEA and CTL-VHI rats showed no differences (p = 0.67) as post hoc analysis confirmed.

FIGURE 4
Surface righting performance. Results are expressed as mean ± SEM. Statistical analyses were conducted by two-way mixed ANOVA. CTL-VHI, control rats treated with vehicle; PA-VHI, rats subjected to PA and treated with vehicle; PA + PEA, rats subjected to PA and PEA treatment; CTL + PEA, control rats treated with PEA.
Immunohistochemistry and Western blot analysis of pNF H/M reactive area and expression levels allowed axonal function evaluation. Neurofilaments' aberrant phosphorylation is a hallmark of axonal degeneration (Grant and Pant, 2000;Sihag et al., 2007;Dale and Garcia, 2012;Chen et al., 2017) and is found in several human neurological diseases (Hirano, 1994;Mori et al., 1996;Bomont et al., 2000;Shepherd et al., 2002;Douglas-Escobar et al., 2010). Changes in immunoreactivity and phosphorylation status measured by Western blotting for pNF H/M give the pattern of PA-induced alterations in axonal functionality and are in agreement with our previous findings (Saraceno et al., 2010;Herrera et al., 2018). Figure 8A is a representative CA1 hippocampal stratum radiatum section immunostained for pNF H/M. Neither birth condition nor treatment were sources of variation according to two-way ANOVA (F 1,56 = 0.19, p = 0.66; F 1,56 = 0.06, p = 0.8, respectively; Figure 8B). In agreement with these findings, analysis of pNF H/M protein expression confirmed that birth condition and treatment were not sources of data variation (F 1,8 = 0.0006, p = 0.98; F 1,8 = 0.001, p = 0.97, respectively; Figure 8C). Full scans of uncropped blots are presented in Supplementary Figure 2.
Similar results were observed for glial response according to GFAP immunostaining data analysis ( Figure 9A). The hippocampus and dentate gyrus, phylogenetically, of the oldest cortical areas, keep much of the radial orientation of their immature astroglial system (Eckenhoff and Rakic, 1984). In this region, a fusiform or rod-shaped and elongated morphology is observed (Zhou et al., 2019). All experimental groups showed a strikingly regular intense pattern of GFAP immunoreactivity in the hippocampus (Garcia-Segura et al., 1988;Hajós and Kálmán, 1989). Two-way ANOVA showed that the number of GFAP positive astrocytes was unrelated to either birth condition or treatment (F 1,56 = 0.22, p = 0.64; F 1,56 = 0.001, p = 0.97; Figure 9B). In agreement with these findings, GFAP protein expression was not affected by either birth condition or treatment (F 1,8 = 0.007, p = 0.94; F 1,8 = 0.01, p = 0.91; Figure 9C). Full scans of uncropped blots are presented in Supplementary Figure 3.

Discussion
Perinatal asphyxia-induced growth retardation and neurodevelopmental delay In our study, PA for 19 min caused growth retardation, evidenced by low weight gain and severe neurodevelopmental delay by the first 3 weeks of life. The forelimb placing and grasp reflexes, resembling human palmar placing and grasp reflexes (Futagi et al., 2012;Nguyen et al., 2017), were delayed 3-5 days. Asphyctic animals had a 1.5-day delay in eye-opening, a 2day delay in air-righting and auditory startle reflexes, and a 1-day delay in gait, negative geotaxis, and sensory eyelid reflex. Asphyctic rats showed slow negative geotaxis on P11 and P14 that normalized by the third week of life. These results extend earlier evidence after PA for 15 min, where forelimb placing and grasp reflexes were also affected the most. However, rats subjected to PA for 15 min (moderate PA) had body weight gain restored by the second week of life and normal eye opening (Kiss et al., 2009). Our findings show severe neurodevelopmental lag after 19 min of PA. Perinatal hypoxia-ischemia (HI) in Rice-Vanucci's experimental model induced growth retardation (Fan et al., 2005(Fan et al., , 2006Lubics et al., 2005), delayed eye-opening (Fan et al., 2005(Fan et al., , 2006Romero et al., 2017) and grasping onset (Lubics et al., 2005), and slowed gait (Fan et al., 2005(Fan et al., , 2006Lubics et al., 2005), surface righting (Fan et al., 2005(Fan et al., , 2006Lubics et al., 2005), and negative geotaxis (Lubics et al., 2005), early signs of neurobehavioral dysfunction.
The neurodevelopmental delay observed over the first 3 weeks of life after PA precedes the alterations in exploratory activity, anxiety levels, and cognition on P30 (Barkhuizen et al., 2017). One month after PA, we found a decrease in rearing time (Herrera et al., 2018), i.e., vertical exploration in response to novelty, typically dependent on the integrity of the hippocampus (Lever et al., 2006). One-month-old rats subjected to an episode of 19-20 min of PA had reduced locomotion and rearing as well (Chen et al., 1995). Deficits in exploratory locomotion and increased anxiety in the open-field test were reported 1 month after severe PA (21 min). These 1-month-old asphyctic rats showed reduced exploratory locomotion on a squared area on P7 and slowed negative geotaxis on P14, and surface righting on P1 (Farfán et al., 2020).

Perinatal asphyxia-induced hippocampal neuronal degeneration and dendritic alterations on postnatal day 21
Unlike studies on early growth and reflex development after PA (Kiss et al., 2009), we included morphological and biochemical analysis along with neurobehavioral testing. Instead of focusing on hippocampal oxidative stress or neuroinflammation (Farfán et al., 2020), we studied cytoskeletal modifications in CA1 neurons and the corresponding astrocytic response to extend our findings on P30 (Herrera et al., 2018). Neurobehavioral testing showed growth retardation and delayed reflexes over the first 3 weeks of life, and neuropathology examination on P21 confirmed CA1 hippocampal neurons' vulnerability to severe PA (19 min). Besides signs of degeneration, these neurons showed decreased MAP-2 immunostaining and expression. MAP-2 seems an early biomarker of PA-induced neuronal injury, as observed in a birth-asphyxia piglet model (Lingwood et al., 2008), used to assess dendritic cytoskeletal dysfunction induced by HI (Malinak and Silverstein, 1996;Mink and Johnston, 2000;Sánchez et al., 2000;Zhu et al., 2003;Takita et al., 2004;Kühn et al., 2005;Graham et al., 2018). MAP-2 is phosphorylated by protein kinase A (PKA), an ATP-dependent enzyme. Then, PA-induced ATP reduction might explain MAP-2 decrease as phosphorylation might alter its susceptibility to proteolysis (Islam and Burns, 1981;Johnson et al., 1991;Grau et al., 1992;Sánchez et al., 2000;Ashworth et al., 2003).
In this work, the decreased hippocampal MAP-2 immunostaining and protein expression extend our findings on MAP-2 decreased level on P30 (Herrera et al., 2018), observed in the hippocampus as far as on P120 (Saraceno et al., 2010). In contrast, on P21 hippocampal pNF H/M level was stable and was not affected until P30 (Saraceno et al., 2010;Herrera et al., 2018). On P21, we have not found differences in either labeling intensity or the number of GFAP-positive astrocytes in the hippocampus. Once again, our results pose GFAP as a late biomarker of glial hippocampal damage following PA (Saraceno et al., 2016;Herrera et al., 2018), showing a significant increase 4 months following severe PA for 19 min (Saraceno et al., 2010). Likewise, clinical data suggests plain astrogliosis in

Early neuroprotective effects of palmitoylethanolamide treatment
Palmitoylethanolamide (10 mg/kg) administered within the first hour of life, reversed the delay in the appearance of gait, air-righting, forelimb placing and grasp reflexes, and improved negative geotaxis performance on P11 and P12 in rats subjected to severe PA (19 min). PEA treatment reduced CA1 neuronal degeneration and cytoskeletal dendritic alterations on P21, as inferred from MAP-2 immunostaining and protein expression. Neuroprotection by PEA treatment against MAP-2 deficit and early neuromotor dysfunction has been observed in experimental neurodegeneration. PEA blunted Aβ42-induced reduction in MAP-2 labeling in degenerating neurons in vitro (Beggiato et al., 2018) and attenuated MAP-2 deficit in an in vivo Parkinson's disease (PD) model (PEA 10 mg/kg) (Esposito et al., 2012). Therapeutic effects were reported for PEA (10 mg/kg) on limb locomotor rating scale over the first 8 days following experimental spinal cord injury (Genovese et al., 2008). Likewise, PEA prevented short-term limb weakness and altered gait in an experimental autoimmune encephalomyelitis model of multiple sclerosis (Rahimi et al., 2015).

Conclusion
Treatment with PEA (10 mg/kg) within the first hour of life attenuated neurodevelopmental delay in rats subjected to severe PA (19 min), reducing neurodegeneration and MAP-2 deficit in CA1 neurons on P21. Involved in the pathogenesis of several NDDs (Lasser et al., 2018), dendritic protein MAP-2 appears as an early marker of PA-induced hippocampal damage and a novel target for PEA-mediated neuroprotection. The therapeutic properties of this endogenous amide in NDDs have gathered evidence as case reports (Antonucci et al., 2015), experimental rodent models (Cristiano et al., 2018), randomized clinical trials (Khalaj et al., 2018), and comparative studies on animals and humans (Bertolino et al., 2017). Clinical research showed a highsafety profile for PEA (Steels et al., 2019). Therefore, PEA seems to be a promising neuroprotective agent against PA. Further studies should clarify the molecular mechanisms underlying PEA effects and help specify its precise indications.

Data availability statement
The original contributions presented in this study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

Ethics statement
The animal study was reviewed and approved by the Institutional Animal Care and Use Committee of the University of Buenos Aires (CICUAL#4091/04).