Introduction

With the 9progress of treatment technology in neonatal ICU (NICU), the survival rate of preterm infants in China has increased year by year [1,2], but due to the extreme immaturity of preterm infants’ brain and lung and other organ development, preterm infants are prone to respiratory dysfunction, leading to hypoxic brain injury. Its pathophysiological process mainly shows that the early neuronal necrosis and apoptosis after cerebral hypoxia ischemia lead to delayed neuronal death. Moreover, the blood-brain barrier of premature infants is often underdeveloped, and many inflammatory factors are more likely to pass through the underdeveloped Blood-brain barrier to cause or aggravate brain damage, leading to long-term neurodevelopmental disorders such as cerebral palsy and cognitive impairment [3,4]. Treatment of hypoxic/ischemic (HI) brain injury in the NICU is limited, and rehabilitation is primarily used to improve prognosis. At present, it has been found that EPO can resist inflammation and reduce the damage of cell apoptosis to cells, which may have some positive neuroprotective effects on HI brain injury.

Erythropoietin (EPO) is a glycoprotein hormone involved in regulating the differentiation and proliferation of erythrocyte progenitor cells, and plays an important role in maintaining the dynamic balance of erythrocytes in vivo, which has also been widely used in NICU as a drug to prevent anemia in premature infants for the past few years [5]. It is not only a hematopoietic factor but also a neurotrophic factor [6]. EPO has been shown to be produced in human brain regions such as the hippocampus, temporal cortex and amygdala [7,8], and plays a neuroprotective role by binding to EPO receptors on mature neurons, neuronal progenitor cells and other cells. EPO and its receptors appear in the developing human brain as early as 5 weeks after conception, and increase with gestation age [9]. Recombinant human EPO (rhEPO) is a glycoprotein synthesized by genetic recombination technology. It has the same amino acid sequence and similar biological activity as natural EPO, also can cross the blood-brain barrier [10]. Recent years, the protective effect of rhEPO has been proved in many studies in vivo [11] and in vitro [12], however, the role and mechanism of rhEPO on premature immature brain injury induced by HI insult still remain unclear. Thus, in the present study, we investigated the neuroprotective effects of rhEPO as well as its potential mechanism in immature brain injury induced by HI insult.

Methods Animals and materials

Animals used in this study were complied with the recommendations of the National Research Council Guide for the Care and Use of Experimental Animals and were approved by the Institutional Animal Care and Use Committee of Shandong Provincial Hospital affiliated to Shandong First Medical University.6 Pregnant Sprague–Dawley (SD) rats were obtained from Jinan Pengyue Laboratory Animal Breeding Co., Ltd (China) and were allowed to deliver. Total 85 rat pups were delivered and 82 pups were survived at Postnatal day 3 (PD3) to use for this study. Recombinant human erythropoitin injection (CHO cell) was purchased from Sansheng Pharmaceutical Co., Ltd (Approval Number: 10000 units/branch, National drug Approval number: S20010001). Anti- cyclooxygenase-2 (COX-2), and anti-Caspase-3 from Abcam Co., Ltd, anti- phosphorylated-Akt (p-Akt) from Millipore Co., Ltd respectively. Anti-β-actin was purchased from Wuhan Sanying Biotechnology Co., Ltd. HRP conjugation affinity goat anti-mouse IgG antibody was from Proteintech. Hematoxylin-Eosin (HE) staining kit was from Solebo Company. The general two-step kit, EDTA antigen repair solution pH9.0 and diaminobenzidine (DAB) chromogenic kits were from Beijing Zhongshan Jinqiao Biotechnology Co., Ltd, and the bicinchoninic acid (BCA) protein concentration determination kit was from Biosharp.

Animal groups, model establishment and drug administration

The use of immature rats to establish model is now of greater clinical relevance, as brain injury is most common in infants surviving after delivery between 24 and 28 weeks’ gestation. On PD 3 rat model would provide lesions representative of those seen at 24–28 week’s gestation in surviving human infants [13]. PD3 rats were randomly divided into three groups: Sham group, HI group and HI+rhEPO group. The HI group and HI+rhEPO group rats underwent permanent ligation of the right common carotid artery after ether anesthesia. After surgery, the cubs were returned to the cage for 2 h, then exposed to hypoxia (92% N2 + 8% O2) for 2 h, and then returned to the cage for continued breastfeeding. Sham surgical rats terminated the surgery after cervical incision and immediately closed the wound. After surgery, the young rats returned to the cage for 2 h, then were exposed to the air for 2 h, and then returned to the cage to continue breastfeeding. During the model establishment process, a total of 5 offspring died, and the dead offspring were replaced by extra normal offspring to continue modeling, ensuring that each group had 20 offspring. Rats in HI+rhEPO group were immediately given intraperitoneal injection of rhEPO (5000 IU/kg) after hypoxia, with a total of 3 injections every day. Other rats were given the same amount of physiological saline in the same way.

Tissue preparation

Five rats in per group were sacrificed at PD10, and brain tissues were removed, quickly placed in a liquid nitrogen tank and stored in a refrigerator at -80 °C for Western Blot analysis. The other five rats were anesthetized by intraperitoneal injection of pentobarbital (50 mg/kg), then perform perfusion brain extraction.We divide the whole brain into anterior, middle and posterior parts at the optic chiasm and the posterior part of the hypothalamus, which were fixed with 4% paraformaldehyde for 48h to embed in paraffin wax, and then was serially sectioned in the coronal plane into 5um sections, for HE and immunohistochemistry detection.

Neurological behavior

The remaining rats were subjected to Morris water maze test on PD28. Morris water maze was used to evaluate long-term cognitive function of the rats.The experiment was carried out in a circular galvanized steel maze (2 meters in diameter and 75 centimeters in depth), which was filled with water 40 centimeters deep (water temperature 22 °C) and made opaque by adding a non-toxic water-soluble dye. The maze was located in a large and quiet testing chamber, where the rats’ movements were monitored and recorded by video cameras. The pool is divided into four equal quadrants. In the middle of a quadrant, a circular escape platform of 11 centimeter (cm) in diameter, was placed about 1.5 cm below the surface and held in place. Starting on PD28, each rat received swimming training four times a day for five days. In the swimming experiment, each rat was gently lowered into the water from a randomly selected quadrant (except the quadrant of the hidden platform). The rats swim to find and climb to the hidden platform in 60 s. The time it took to climb the invisible platform (Escape Latency, EL) was recorded daily. The rats were allowed to stay on the platform for 20 s, then rest for one minute to the next training session. The rats that could not find the platform within 60 s, were helped to find the platform and stay on the platform for 20 s, whose EL was recorded as 60 s [14,15]. The platform was removed on the fifth day, and the rats were put into the pool from the opposite quadrant of the platform quadrant. The time of searching the platform quadrant and platform crossing number within 60 s were recorded.

Histopathological changes

Brain sections were gone through xylene dewaxing and ethanol gradient dehydration followed by blocking endogenous peroxidase activity in 3% peroxide for 30 min. After washing three times with 0.01mol/L phosphate-buffered saline (PBS) for 5 min each time, the sections were then incubated in DAB substrate for 3–5 min and then stained with hematoxylin and eosin. Rinse and soak in water. Dehydrated, transparent and sealed with neutral gum. then observed under light microscope to find the histopathological changes.

Immunofluorescence assay for neuronal apoptosis

Apoptotic cells were identified by TUNEL staining using an in situ end labeling detection kit (Fluorescein; Roche Applied Science) according to the manufacturer’s instructions. The cells were irradiated with blue excitation light under the fluorescence microscope. Randomly select five regions of the cortex and hippocampus, and use ImageJ software to analyze the number of positive cells emitting green fluorescence.

Immunohistochemistry

Briefly, sections were washed for three times in PBS, then endogenous peroxidase activity was quenched by incubation of the samples in 3% hydrogen peroxide in PBS for 30 min, followed by wash for 10min with PBS. Then the sections were incubated for 2h at room temperature with a rabbit polyclonal antibody against COX-2 (1:100) and Caspase-3 (1:500). Sections were washed with PBS for three times, then Added 100ul reaction enhancement solution, incubate at room temperature for 20min, and rinsed with PBS solution for 3 times before being incubated for 30min at room temperature in the presence of biotinylated goat anti-rabbit IgG secondary antibodies (1:1000). Sections were then washed with PBS and incubated with avidin–peroxidase complex for 30 min before the immunocomplex was visualized using the chromogen 3,3-diaminobenzidine. then counterstained with hematoxylin. Finally, sections were dehydrated in ethanol, cleared in xylene, and covered with neutral balsam. Take pictures under light microscope and the results were analyzed by means of average positive staining area percentage.

Western blot analysis

Tissue protein were extracted according instruction of BCA protein concentration assay kit. Briefly, the freezen brain tissue were homogenized in ice-cold lysis buffer for 30 min. Then the lysates were centrifuged for 15 min at 10 000 x g at 4˚C, and the resulting supernatants were collected and boiled. Protein concentrations were measured, protein samples (30ug/lane) were separated by 10% SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. The membranes were then blocked for 1.5h with 2% skimmed milk, and were incubated with antibodies against anti-Caspase-3 (1:2000), Anti-p-Akt (1:1000), and β-actin (1:5000) at 4˚C overnight, followed by washing with TBST. The membranes were then incubated for 1.5h at room temperature with alkaline phosphatase-labeled anti-rabbit secondary antibody (1:5000). β-actin was used as a loading control. Drop enhanced chemiluminescence to develop and the image were taken into Image J software system for gray analysis. Immunoreactive of Caspase-3 and p-Akt were quantified by protein bands normalized to β-actin expression ratio.

Statistical analysis

Data analysis, including graphing and statistical evaluation, was carried out using GraphPad Prism 9 and IBM SPSS 27.0 software. To compare between two groups, we employed t-test. The one-way ANOVA was employed to assess overall differences among multiple groups, and then Dunnett’s T3 test (when equal variances not assumed) and the Bonferroni (when equal variances assumed) post hoc test were used for pairwise comparisons between individual groups. For Escape Latency on days 1–4 of the water maze, the cox proportional hazards model was used for analysis [16,17]. Statistical significance was established when the P-value was lower than 0.05.

Results Effect of rhEPO on general health

During the experiment, the survival rate of the Sham surgery group was 95%, the survival rate of the HI group was 60%, and the survival rate of the HI+rhEPO group was 80%. There was a significant difference in survival rates among the three groups of rats [F (2, 57) = 3.927, P = 0.0281].

As shown in Fig. 1, the weight gain of the HI and HI+rhEPO groups was slow and smaller than that of the Sham group. Prior to PD10, difference of rats’ weight in three groups was NS. however, after PD15, weight of rats in the HI group significantly lower than Sham group(P < 0.05), while rats in HI+rhEPO group grew faster than those in HI group. These differences of weight among HI group and Sham group or HI+rhEPO group were most obvious at PD20[F (2, 12) = 43.85, P < 0.0001], while were alleviated at PD28.

Fig. 1:

Body weight of rats in every group. *P < 0.05 vs. Sham Group; #P < 0.05 vs. HI group.

Effect of rhEPO on Morris water maze performance

Morris water maze test was used to analyze the pup’s cognitive activities at PD28. As shown in Fig. 2a, which representative typical trajectory of rats in three groups on third day. During the training days, the track of rats in HI group looking for the platform was tortuous, while the intervention of rhEPO could improve rat’s learning ability, which showed significant tend to platform. On the fifth day, after the platform was removed, the rats in sham and HI+rhEPO groups had good memory function and would tend to stay in the quadrant where the original platform was located, while HI rats showed random swimming trajectory. As shown in Fig. 2b and Table 1, the proportion of three groups of rats finding the platform gradually increases with the increase of training days. However, the ratio of rats in HI group to find the platform from the beginning of the fourth test on the second day to the end of the the fourth day was significantly lower than that in Sham group and HI+rhEPO groups (P < 0.05). After removing the platform, search time [F (2, 12) = 7.089, P = 0.0093] and platform crossing number [F (2, 12) = 9.438, P = 0.0034] in HI group were significantly less than other groups (Fig. 2c), as for rats in rhEPO intervention group, search time and platform crossing number were more than HI group.

Table 1 - Cox regression results of Morris water maze B SE P-value Exp(B) CI 95 min CI 95 max Day 1, Swim 1 -1.285 0.989 0.194 0.277 0.040 1.922 Day 1, Swim 2 −0.039 0.251 0.875 0.961 0.588 1.572 Day 1, Swim 3 −0.254 0.235 0.281 0.776 0.489 1.231 Day 1, Swim 4 −0.487 0.228 0.033 0.614 0.393 0.961 Day 2, Swim 1 −0.654 0.236 0.005 0.520 0.327 0.825 Day 2, Swim 2 −0.291 0.209 0.164 0.748 0.496 1.126 Day 2, Swim 3 −0.266 0.206 0.197 0.766 0.511 1.148 Day 2, Swim 4 −0.477 0.221 0.031 0.620 0.402 0.957 Day 3, Swim 1 −0.469 0.211 0.027 0.626 0.413 0.947 Day 3, Swim 2 −0.696 0.201 0.001 0.498 0.336 0.739 Day 3, Swim 3 −0.724 0.211 0.001 0.485 0.320 0.733 Day 3, Swim 4 −0.696 0.201 0.001 0.498 0.336 0.739 Day 4, Swim 1 −0.838 0.190 0.000 0.433 0.298 0.628 Day 4, Swim 2 −0.717 0.188 0.000 0.488 0.338 0.706 Day 4, Swim 3 −0.502 0.205 0.014 0.605 0.405 0.904 Day 4, Swim 4 −0.775 0.200 0.000 0.460 0.311 0.681
Fig. 2:

(a) Locus diagram of rats in each group on third and fifth day (D3, D5), О indicates the platform; (b) Cumulative incidence plots for the first to fourth day escape incubation period of Morris water maze; (c) Search time & Platform crossing number. *P < 0.05 vs. Sham Group; #P < 0.05 vs. HI Group.

Effect of rhEPO on brain pathological morphological damages after HI insult

As Fig. 3 showed, structure of the cerebral cortex and hippocampus in sham group was tight, with the cells orderly arranged and hierarchical clear, meanwhile the cell morphology was normal, with rich cytoplasm, obvious nucleus and clearly centered nucleoli (A1, B1). However, for HI rats, the cells of cerebral cortex and hippocampus showed disorderly arranged, loose and unclear hierarchical, as well as obviously irregular cell morphology. Many cells showed nuclear pyknosis, nucleolysis and even vacuolization (A2, B2). While in rhEPO group, these morphological damages were significantly lightened than HI group, only showing cell layers in the cerebral cortex and hippocampus slightly loose and scattered, and some cell structures were incomplete (A3, B3).

Fig. 3:

Histological analysis of brain (10×, 20×). (a) Hippocampus; (b) Cerebral cortex; (a1, b1) Sham group; (a2, b2) HI group; (a3, b3) HI + rhEPO group.

Effect of rhEPO on neuronal apoptosis

As Fig. 4 (A, a; B, b; C, c) showed, double staining with DNA fragment labeled TUNEL (green, arrow) was used to detect neuronal apoptosis. As shown in Fig. 4d, the proportion of TUNEL positive neurons in the cortex (2.60 ± 1.10%) and Hippocampus (3.70 ± 1.77%) of rats in the Sham surgery group was very low. The proportion of TUNEL positive neurons in the cortex (16.38 ± 4.42%) and Hippocampus (13.98 ± 3.90%) of rats in the HI group was significantly increased (P < 0.05). After rhEPO intervention, the proportion of TUNEL positive neurons in cortex (4.23 ± 2.01%) and Hippocampus (5.65 ± 2.12%) of rats was significantly less than that in HI rats (P < 0.05). It suggests that rhEPO has therapeutic effects on immature brain models of HI.

Fig. 4:

Neuronal apoptosis in every group. (A, a) Sham Group; (B, b) HI Group; (C, c) HI+ rhEPO Group D. Densitometric analysis of TUNEL staining. *P < 0.05 vs. Sham Group; #P < 0.05 vs. HI Group.

Effect of rhEPO on COX-2 and Caspase-3 expression after HI insult

COX-2 and Caspase-3 expressed mainly in neurons of the cortex and hippocampus, which appear as brown-yellow colored deposits. There are significant differences in the expression of Cox-2[F (2132) = 4803, P < 0.0001] and Caspase-3[F (2132) = 3727, P < 0.0001] between the groups. For rats in HI group, positive staining of both proteins appeared significantly stronger than those in sham group (P < 0.05). Compared to HI rats, the expression of COX-2 and Caspase-3 were largely downregulated in HI+rhEPO group (P < 0.05), which only appeared light brown staining in the nucleus and cytoplasm of some cells (Fig. 5).

Fig. 5:

Expression of COX-2 and Caspase-3 protein in every group. (a) and (b) are the representative picture of immunohistochemical staining involving COX-2 and Capase-3 protein (×10, ×20). 1: Sham Group; 2: HI group; 3: HI+rhEPO Group. (c and d) Comparison of both protein positive expression in three groups. *P < 0.05 vs. Sham Group; #P < 0.05 vs. HI Group. (e–g): Western blot analysis of caspase-3 and p-Akt protein in brain tissue (e) and quantification comparison of three groups (f and g). *P < 0.05 vs. Sham Group; #P < 0.05 vs. HI Group.

Activation of PI3K/Akt pathway involve in anti-apoptosis effect of rhEPO after HI insult

The expression of Caspase-3 and p-Akt were further detected by western blot analysis. There are significant differences in the expression of Caspase-3 [F (2, 63) = 54.05, P < 0.0001] and P-Akt [F (2, 63) = 31.15, P < 0.0001] between each group. Compared with sham group, both expression levels of Caspase-3 and p-Akt protein in HI group were increased (P < 0.05). In HI+rhEPO group, the expression level of Caspase-3 protein was decreased by rhEPO intervention, meanwhile, p-Akt protein expression level was significantly increased (P < 0.05) (Fig. 5).

Discussion

For premature infants, the molecular mechanisms of brain injury are very complex, and inflammation-induced neuronal apoptosis is the most critical [18]. In this study, We selected PD3 rats followed by HI insult to establish an animal model of premature immature brain injury. HE and Morris water maze experiments confirmed the model successfully established, in terms of pathological morphology and cognitive function. Cyclooxyganese-2 (COX-2) [19], is most expressed in the brain and mainly exists in the nerve cell membrane. COX-2 expression level is very low under normal physiological conditions, once the membrane potential changes and nerve conduction activity increases, COX-2 gene overexpression can be induced. Hidetoshi et al. [20]found at 24h after HI induced encephalopathy of rats, the expression and activity of COX-2 in brain tissues increased significantly, which induced acute neuronal excitotoxicity and injury. We found that the COX-2 expression in cerebral cortex and hippocampus of rats in HI group was significantly higher than that of sham group, from which we can conclude that inflammatory response in neuro cells is a key mechanism in immature brain injury induced by HI insult.Caspase-3 is a key executor of programmed cell death, which plays a key role in the process of apoptosis and is a reliable indicator for detecting apoptosis in recent years [21]. We found that the positive expression of caspase-3 protein and the proportion of TUNEL positive neurons in the cortex and hippocampus cells of HI rats was significantly increased, while rhEPO intervention could significantly reduced, which also confirmed that neuro apoptosis was also an important cause of immature brain injury induced by HI insult.

PI3K/Akt pathway plays a key role in the cytoprotection of damaged neurons. Akt is an important survival-promoting kinase that phosphorylates to p-Akt through the PI3K/Akt pathway in Ser473, downregulating the downstream substrate Caspase-3 of Akt, which has been shown to play a crucial role in neuronal survival [22,23]. Study reported that the level of p-Akt expression was significantly elevated after nerves injury [24],PI3K-specific inhibitor can significantly aggravate the extent of brain damage [25], while some drugs can downregulate capase-3 expression by activating PI3K/Akt signal pathway to significantly reduce neuronal apoptosis and improve brain function [26]. Yu et al. [27] used PD5 rats to establish HI brain injury model, found that the expression of p-Akt protein in brain tissue of HI rats was increased at 90 h after HI insult, indicating that PI3K/Akt signaling pathway was involved in the stress compensatory response of brain tissue in the early stage of HI. While in present study, we found although the expression of p-Akt in HI rats was higher than those in sham group, but the difference has no statistical significance, which may be related to the different detection time. However, after intervention with rhEPO, the expression of p-Akt was significantly increased compared with that of HI group, meantime the expressions of COX-2 and Caspase-3 in brain were also decreased. All these results indicated that rhEPO played a neuroprotection by activating the PI3K/Akt signaling pathway to inhibit the inflammatory injury and apoptosis of neuro cells induced by HI.

In the present study, histopathological changes such as edema, hemorrhage and necrosis in cerebral cortex and hippocampus of HI rats were significantly alleviated by rhEPO intervention, which were consistent with De-mers’s report [28]. Hippocampus is responsible for function of learning and memory. The hippocampus of preterm infants has selective vulnerability, so the structure and function of the hippocampus are easily damaged after cerebral HI insult, which lead to long-term neurological function and behavior abnormalities. The neural development of a rat at 28–30 days after birth is equivalent to that of human at 2–3 years of age [29], which is the stage that the typical clinical manifestation of cerebral palsy develops. So, it is of great clinical significance to select PD28 rats for neurobehavioral assessment. Morris water maze is a classic experiment to evaluate hippocampal function [30]. The present study showed that HI insult induced significant long-lasting cognitive deficits in rats. Rats suffered HI insult had longer ELs throughout the training days than those in sham group. While rhEPO intervention not only significantly improved the learning ability (shorter ELs), but also increased memory function (more search time and times of crossing target quadrant). All these results indicated that rhEPO significantly improved HI-induced damages of cognitive function occurred in immature brain, which is consistent with previous report [9].

In conclusion, the present study has found that rhEPO may inhibit the expression of inflammatory and apoptosis-related proteins to reduce HI-induced injury in immature brain, then improve long-term cognitive function. At least part of the mechanism is involved in activation of PI3K/Akt signal pathway. All these data may provide help for the clinical use of rhEPO in premature infants suffered from HI brain damage.

Acknowledgements

This article was funded by academic promotion programme of Shandong First Medical University (2019QL023) and Shandong Province key research and development plan (2018GSF118163).

Our animal procedures were complied with the recommendations of the National Research Council Guide for the Care and Use of Experimental Animals and were approved by the Institutional Animal Care and Use Committee of Shandong Provincial Hospital affiliated to Shandong First Medical University.

Conflicts of interest

There are no conflicts of interest.

References 1. Ezenwa B, Ezeaka C, Fajolu I, Ogbenna A, Olowoyeye O, Nwaiwu O, et al. Impact of Erythropoietin in the management of Hypoxic Ischaemic Encephalopathy in resource-constrained settings: protocol for a randomized control trial. BMC Neurol 2020; 20:171. 2. Goisbault M, Simon L, Hanf M, Darmaun D, Rozé JC, Flamant C. Neonatal length growth and height at two years. Neonatol 2016; 110:125–128. 3. Perez JM, Golombek SG, Sola A. Clinical hypoxic-ischemic encephalopathy score of the Iberoamerican Society of Neonatology (Siben): a new proposal for diagnosis and management. Rev Assoc Med Bras (1992) 2017; 63:64–69. 4. Zhu M, Lu M, Li QJ, Zhang Z, Wu ZZ, Li J, et al. Hyperbaric oxygen suppresses hypoxic-ischemic brain damage in newborn rats. J Child Neurol 2015; 30:75–82. 5. McAdams RM, McPherson RJ, Mayock DE, Juul SE. Outcomes of extremely low birth weight infants given early high-dose erythropoietin. J Perinatol 2013; 33:226–230. 6. Kuo-Mao L, Lu-Tai T, Zhengwei C, Shuying L, Yi P, Sachiko T, et al. Erythropoietin ameliorates neonatal hypoxia-ischemia-induced neurobehavioral deficits, neuroinflammation, and hippocampal injury in the juvenile rat. Int J Mol Sci 2016; 17:289. 7. Marti HH, Wenger RH, Rivas LA, Straumann U, Digicaylioglu M, Henn V, et al. Erythropoietin gene expression in human, monkey and murine brain. Eur J Neurosci 1996; 8:666–676. 8. Wenger RH, Kurtz A.Erythropoietin. Compr Physiol 2011; 1:1759–1794. 9. Kumral A, Uysal N, Tugyan K, Sonmez A, Yilmaz O, Gokmen N, et al. Erythropoietin improves long-term spatial memory deficits and brain injury following neonatal hypoxia-ischemia in rats. Behav Brain Res 2004; 153:77–86. 10. Oorschot DE, Sizemore RJ, Amer AR. Treatment of neonatal hypoxic-ischemic encephalopathy with erythropoietin alone, and erythropoietin combined with hypothermia: history, current status, and future research. Int J Mol Sci 2020; 21:1487. 11. Wang H, Chen M, Gao Z, Gong Y, Yu X, Wu H. Recombinant human erythropoietin protects long-term cultured ageing primary nerve cells by upregulating the PI3K/Akt pathway. Neuroreport 2022; 33:186–198. 12. Juul SE, Comstock BA, Wadhawan R, Mayock DE, Courtney SE, Robinson T, et al.; PENUT Trial Consortium. A randomized trial of erythropoietin for neuroprotection in preterm infants. N Engl J Med 2020; 382:233–243. 13. Alfreda S, Anthony J, Ronald F, Fu WY, Michael R, Christopher H. Development of a postnatal 3-day-old rat model of mild hypoxic-ischemic brain injury. Brain Res 2003; 993:101–110. 14. Mutlu O, Akar F, Celikyurt IK, Tanyeri P, Ulak G, Erden F. 7-NI and ODQ disturbs memory in the elevated plus maze, Morris water maze, and radial arm maze tests in mice. Drug Target Insights 2015; 9:1–8. 15. Sun L, Li X, Ma C, He Z, Zhang X, Wang C, et al. Improving effect of the policosanol from ericerus pela wax on learning and memory impairment caused by Scopolamine in Mice. Foods (Basel, Switzerland) 2022; 11:2095. 16. George B, Seals S, Aban I. Survival analysis and regression models. J Nucl Cardiol 2014; 21:686–694. 17. Jahn-Eimermacher A, Lasarzik I, Raber J. Statistical analysis of latency outcomes in behavioral experiments. Behav Brain Res 2011; 221:271–275. 18. Liu F, McCullough LD. Inflammatory responses in hypoxic ischemic encephalopathy. Acta Pharmacol Sin 2013; 34:1121–1130. 19. Hewett SJ, Bell SC, Hewett JA. Contributions of cyclooxygenase-2 to neuroplasticity and neuropathology of the central nervous system. Pharmacol Ther 2006; 112:335–357. 20. Hidetoshi T, Christoph A, Beatriz SG, Qian W, Katrin A. Function of prostaglandin E2 EP receptors in the acute outcome of rodent hypoxic ischemic encephalopathy. Neurosci Lett 2011; 504:185–190. 21. Adachi M, Sohma O, Tsuneishi S, Takada S, Nakamura H. Combination effect of systemic hypothermia and caspase inhibitor administration against hypoxic-ischemic brain damage in neonatal rats. Pediatr Res 2001; 50:590–595. 22. Franke TF, Hornik CP, Segev L, Shostak GA, Sugimoto C. PI3K/Akt and apoptosis: size matters. Oncogene 2003; 22:8983–8998. 23. Takada-Takatori Y, Kume T, Sugimoto M, Katsuki H, Sugimoto H, Akaike A. Acetylcholinesterase inhibitors used in treatment of Alzheimer’s disease prevent glutamate neurotoxicity via nicotinic acetylcholine receptors and phosphatidylinositol 3-kinase cascade. Neuropharmacology 2006; 51:474–486. 24. Li X, Zhang J, Chai S, Wang X. Progesterone alleviates hypoxic-ischemic brain injury via the Akt/GSK-3β signaling pathway. Exp Ther Med 2014; 8:1241–1246. 25. Gu Q, Zhai L, Feng X, Chen J, Miao Z, Ren L, et al. Apelin-36, a potent peptide, protects against ischemic brain injury by activating the PI3K/Akt pathway. Neurochem Int 2013; 63:535–540. 26. Zhang B, Ji X, Zhang S, Ren H, Wang M, Guo C, et al. Hemin-mediated neuroglobin induction exerts neuroprotection following ischemic brain injury through PI3K/Akt signaling. Mol Med Rep 2013; 8:681–685. 27. DaFan Y, LiHua Z, Li J. Recombinant human erythropoietin augments neovascularization responses in a neonatal rat model of premature brain damage by phosphatidylinositol 3 kinase/Akt pathway. Chin Med J 2017; 130:854–858. 28. Demers EJ, McPherson RJ, Juul SE. Erythropoietin protects dopaminergic neurons and improves neurobehavioral outcomes in juvenile rats after neonatal hypoxia-ischemia. Pediatr Res 2005; 58:297–301. 29. Semple BD, Blomgren K, Gimlin K, Ferriero DM, Noble-Haeusslein LJ. Brain development in rodents and humans: Identifying benchmarks of maturation and vulnerability to injury across species. Prog Neurobiol 2013; 106-107:1–16. 30. Feng EC, Jiang L. Effects of leptin on neurocognitive and motor functions in juvenile rats in a preterm brain damage model. Mol Med Rep 2018; 18:4095–4102.