Transcriptome analysis and differentially expressed gene screening for hypoxic-ischemic brain damage in rats treated with acupuncture

Hypoxic-ischemic brain damage (HIBD) is one of the most common critical diseases in neonates with high mortality and disability rates. The latest research showed that long non-coding RNAs(lncRNA) played an important role in the development of HIBD. Recently, acupuncture therapy has been found to be effective in the treatment of HIBD. However, the mechanism of lncRNA in acupuncture treatment of HIBD is still unclear. In this study, we investigated the role of lncRNA in acupuncture treatment of HIBD in detail. We demonstrated behavioral performance similar to cognitive decits in HIBD rat models in the new object recognition experiment and pathological lesion of the prefrontal cortex in nissl staining. Acupuncture treatments at acupoints DU24 and GB13 was proved to be effective in alleviating behavioral decits and brain injury. A whole transcriptome analysis was applied to investigate transcriptome changes caused by acupuncture in PFC of HIBD rats. A total of 48 mRNAs and 65 lncRNAs was identied relate to acupuncture group and model group. According to Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) analysis, we found several lncRNAs and their target mRNAs were related to PI3K-Akt signaling pathway, TNF signaling pathway and NOD-like receptor signaling pathway, etc. The results of our research may provide new perspectives on the mechanism of acupuncture and affect the diagnosis and treatment of HIBD. acetylation Reln synaptic plasticity and memory retention in the medial PFC 63,64 . Reln also has a role in PI3-kinase signaling in neuronal growth cones, and it contributes to nal neuron positioning in the mammalian brain by local modulation of protein kinase B and glycogen synthase kinase 3beta kinase activities 65 . The results of our study suggested that the increasing expression of TCONS_00173912 and TCONS_00057072 and the decreasing expression of TCONS_00088551 and TCONS_00097914 may inhibit Reln, which then works via the PI3K-Akt signaling pathway in neonatal HIBD.


Introduction
Neonatal hypoxic-ischemic brain disorder (HIBD) is an important cause of neonatal death and birth defects 1. Reportedly, the incidence of brain injury caused by perinatal asphyxia is 0.2%-0.4%2. What is worse, 20%-50% of the children with HIBD die in the neonatal period, and >25% of the survivors develop permanent neurological disorders, such as mental retardation, cerebral palsy, epilepsy, and autism2,3.
Presently, no international guidelines for the treatment of HIBD have been established, and the relatively common hypothermia therapy, stem cell transplantation, and melatonin pharmacological treatments 4,5 cannot fundamentally change the nerve defects and ischemic and hypoxic internal environment of children with HIBD. In China, acupuncture, as an alternative therapy, is widely used to treat various sequelae caused by HIBD, with signi cant clinical e cacy and no obvious side effects 6-9. Researchers have been constantly exploring the speci c mechanisms of acupuncture treatment for HIBD 10-12. However, no de nite conclusion has been reached on the target of acupuncture. Therefore, it is of great scienti c signi cance to strengthen the research on the pathogenesis of neonatal ischemic and hypoxic encephalopathy and the therapeutic effect of acupuncture treatment.
Long non-coding RNA (lncRNA) is a type of non-coding RNA (ncRNA) with lengths ranging from 200 nt to 100 kb that lack a signi cant open reading frame (ORF). Studies have shown that lncRNAs could regulate the potential molecular mechanisms of various biological processes, including chromatin tissue, epigenetic regulation, gene transcription, RNA conversion, and genome defenses [13][14][15]. Additionally, lncRNA is involved in many important physiological functions of neurons in the nervous system playing an important role in the human brain development and the growth and evolution of higher cognitive functions 16. In recent years, a number of studies have realized the expression spectrum determination of lncRNAs in the brain tissue of HIBD-induced hypoxia and ischemia in neonatal rats through highthroughput sequencing technology, and some lncRNAs with signi cant changes have been identi ed. Ruibin Zhao et al 17 found that HIBD-induced brain injury changed the expression pro le of lncRNA in newborn rats; for example, ENSRN OG00000021987 was down-regulated, and silencing of IT aggravated the apoptosis of cells in the hippocampus. Li H et al 18 found that lncRNA TCONS_00044595 L was involved in the regulation of CLOCK in a post-transcriptional manner, which was likely to lead to a circadian rhythm disorder in newborns after HIBD. Fengyan Zhao et al 19 found that knocking out lncRNA BC088414 could reduce cell apoptosis and promote cell proliferation. However, there was no study on whether acupuncture can cause changes in the expression pro le of lncRNAs in the prefrontal lobe of the brains of hypoxic-ischemic newborn rats caused by HIBD.
To investigate whether acupuncture can improve the cognitive ability of HIBD rats and reduce their brain injury, and to detail the gene expression level of RNA regulated by acupuncture, we conducted a new object recognition experiment and Nissl staining after acupuncture treatment at the acupoints DU24 and GB13 in HIBD rats. Additionally, we used RNA-sequencing (RNA-seq) technology to identify the changes in lncRNAs, as well as the mRNAs and circRNAs, in the prefrontal cortex (PFC) of HIBD newborn rats after treatment with acupuncture. Furthermore, the interaction network of differentiated miRNA-lncRNA or circRNA-mRNA was studied to nd new pathways that participate in the mechanisms of neonatal HIBD.
The results of this study may provide a theoretical basis for speci c molecular biomarkers and targets of acupuncture therapy for neonatal HIBD.

Novel object recognition
After acupuncture intervention, there was no signi cant difference in the exploration time for left object (Tl) and exploration time for right object (Tr) of the rats in the normal, model, and acupuncture groups which indicated that the rats in each group didn't show preference for the location of the objects. The exploration time for familiar object (Tf) and exploration time for novel object (Tn) of the rats between the normal, model, and acupuncture groups in the test period were statistically signi cantly different. Furthermore, the difference in the relative discrimination index (DI) value between the model and normal groups was statistically signi cant (P < 0.001). The difference in the DI value between the acupuncture and model groups was also statistically signi cant (P < 0.001). The results of novel object recognition suggested that acupuncture could signi cantly improve the new object discrimination ability of the HIBD rats ( Fig. 1).

Nissl staining
Nissl staining was used to observe the level of neuronal injury. There was no signi cant change in the thickness of the PFC in the normal, model, and acupuncture groups. However, compared with the normal group, the number of Nissl bodies was signi cantly decreased and disordered in the model group. The amount of Nissl bodies in the acupuncture group was also less than that in the normal group but was more than that in the model group (Fig. 2).

RNA-seq of PFC
To fully elucidate the mechanism of action of acupuncture on lncRNA and mRNA of HIBD rats, RNA-seq technology was applied to the acupuncture, model, and normal groups of young rats (four samples per group). We selected the PFC, which is a brain area with a high association of HIBD, to be the sampling part. After strict control of data quality, mapping, and assembling of transcripts, a total of 1,371,417,384 clean reads were separated from the total reads of 1,304,758,138 raw reads. Then, we predicted the coding potential of lncRNA by CPC2, Pfam, and CNCI. Finally, a total of 15,525 lncRNAs were selected for further study.

Identi cation and characteristic comparisons of lncRNAs and mRNAs
A total of 174 differentially expressed mRNAs were identi ed in total. The results of the clustering analysis of differentially expressed mRNAs were illustrated with a heat map (Fig. 3A). According to the restrictively screened condition FDR < 0.05 and P < 0.05, 103 mRNAs were signi cantly and differentially expressed in the model rats relative to the normal rats with 53 up-regulated and 53 down-regulated (Fig.  3B); 85 mRNAs were signi cantly and differentially expressed in acupuncture rats relative to model rats with 34 up-regulated and 51 down-regulated (Fig. 3C). Venn analysis showed that acupuncture treatment reversed 10 down-regulated mRNA (ENSRNOT00000021857, ENSRNOT00000071615,  ENSRNOT00000088376, ENSRNOT00000080832, ENSRNOT00000081497, ENSRNOT00000089292,   ENSRNOT00000089938, ENSRNOT00000088188, TCONS_00345929, ENSRNOT00000003297) and 7 upregulated mRNAs (ENSRNOT00000042539, ENSRNOT00000092470, ENSRNOT00000086014, ENSRNOT00000061858, TCONS_00150509, TCONS_00143419, ENSRNOT00000086397) of the model group ( Fig. 3D-E).
There were 258 differentially expressed lncRNAs identi ed. The results of the clustering analysis of differentially expressed lncRNAs were illustrated with a heat map (Fig. 4A). Screening with the same method as above, 202 lncRNAs were signi cantly and differentially expressed in the model rats relative to the normal rats with 158 up-regulated and 44 down-regulated (Fig. 4B); 77 lncRNAs were signi cantly and differentially expressed in the rats that received acupuncture relative to the model rats with 31 upregulated and 46 down-regulated (Fig. 4C). As the Venn analysis showed, the 14 up-regulated lncRNAs (TCONS_00038506, TCONS_00105149, TCONS_00021584, TCONS_00216724, TCONS_00310736,  TCONS_00216955, TCONS_00068680, TCONS_00163999, TCONS_00004602, TCONS_00013568,   TCONS_00118348, TCONS_00025912, TCONS_00181752, TCONS_00092759) decreased in the Page 5/29 acupuncture group, whereas 7 down-regulated lncRNAs (TCONS_00206366, TCONS_00097918, TCONS_00049434, TCONS_00364333, TCONS_00010371, TCONS_00250258, TCONS_00001104) increased in the acupuncture group ( Fig. 4D-E). Table 1 lists the parts of the potential targets of lncRNAs that are common and were signi cantly differentially expressed. Moreover, we found that the distinct characteristic of lncRNAs and mRNAs was that lncRNAs had fewer exons, shorter lengths, and shorter lengths of the ORF than mRNAs (Fig. 5). CircRNA is a new type of RNA that comprises a covalently closed continuous loop structure with better structural stability than traditional linear RNA and could be highly expressed in eukaryotic transcriptome. The tissue and developmental speci c expression were shown by circRNA, suggesting that they may play a key role in a variety of cellular processes. Remarkably, it has been reported that circRNAs could act as sponges of microRNA and participate in the regulation of genetic transcription.

Bioinformatics
To evaluate the potential cellular function of overlapped mRNAs, lncRNAs, and circRNAs between experimental groups, we conducted GO enrichment analysis across three domains, including molecular functions (MF), cellular components (CC), and biological processes (BP).

Enrichment analysis: GO and KEGG of mRNAs
Through the GO survey between the acupuncture and model groups, 1,522 GO terms were respectively enriched with 36 GO terms having distinct statistical signi cance (P < 0.05; Fig. 7A), with 29 in MF, 25 in CC, and 236 in BP. The mRNAs that were differentially expressed between the acupuncture and model groups were involved in the positive regulation of glutamate metabolic processes, regulation of interleukin-2 production, histone H3-R17 methylation, and intracellular cAMP activated cation channel activity. In the GO enrichment analysis of the model and normal groups, 2,870 GO terms were respectively enriched with 52 GO terms having distinct statistical signi cance (P < 0.05), with 24 in MF, 11 in CC, and 17 in BP. The mRNAs that were differentially expressed between the model and normal groups were involved in the oxidation-reduction process, regulation of nuclease activity, proton-transporting ATP synthase complex, oxidoreductase activity, and transmembrane transporter activity (Fig. 7B).
We used the KEGG database for pathway analysis. The results showed that mRNAs that were differentially expressed between the acupuncture and model groups participated in multiple pathways, including morphine addiction, cholinergic synapses, primary bile acid biosynthesis, and glycosaminoglycan biosynthesis-heparan sulfate/heparin. In the comparison between the model and normal groups, the differentially expressed lncRNAs were related to chemical carcinogenesis, phototransduction, steroid hormone biosynthesis, and retinol metabolism. Figure 7C-D shows the most prominent pathways.

Enrichment analysis: GO and KEGG of lncRNAs
To evaluate the potential function of lncRNAs, we searched the target mRNAs of lncRNAs in cis and trans. Through the GO survey between the acupuncture and model groups, 4,619 GO terms were respectively enriched with 36 GO terms having distinct statistical signi cance (P < 0.05), with 8 in MF, 8 in CC, and 20 in BP. The lncRNAs that were differentially expressed between the acupuncture and model groups were involved in neurological system processes, the plasma membrane, olfactory receptor activity, and receptor activity (Fig. 8A). In the GO enrichment analysis of the model and normal groups, 6,210 GO terms were respectively enriched with 37 GO terms having distinct statistical signi cance (P < 0.05), with 10 in MF, 7 in CC, and 20 in BP. The lncRNAs differentially expressed between the model and normal groups were involved in olfactory receptor activity, G-protein-coupled receptor activity, and sensory perception (Fig. 8B).
We used the KEGG database for pathway analysis. The results showed that lncRNAs that were differentially expressed between the acupuncture and model groups participated in multiple pathways, including olfactory transduction, ribosomes, cytokine-cytokine receptor interactions, and taste transduction. In the comparison between the model and normal groups, the differentially expressed lncRNAs were related to pathways, such as olfactory transduction pathway, ribosomes, drug metabolism and other enzymes, and biosynthesis of amino acids. Figure 8C-D shows the most prominent pathways.

Enrichment analysis: GO and KEGG analysis of circRNAs
Through the GO survey, we found 2,145 GO terms that were respectively enriched between the acupuncture and model groups, whereas 2,613 GO terms were respectively enriched between the model and normal groups ( Figure 9A-B). However, no GO term showed a signi cant statistical difference.
The results of the KEGG pathway analysis showed that circRNAs that were differentially expressed between the acupuncture and model groups participated in multiple pathways, including axon guidance, Wnt signaling pathway, B cell receptor signaling pathway, and GABAergic synapse. In the comparison between the model and normal groups, the differentially expressed lncRNAs may be related to pathways, such as adrenergic signaling in cardiomyocyte pathway, mTOR signaling pathway, neurotrophin signaling pathway, and MAPK signaling pathway. However, no pathway showed a signi cant statistical difference. Figure 9C-D shows the most prominent pathways.
miRNA-lncRNA/circRNA-mRNA interaction network MicroRNAs (miRNAs) are small endogenous RNAs that comprise 19 to 25 nucleotides and can regulate gene expression post-transcriptionally. Since the spongy peculiarity of miRNA, both lncRNA and mRNA can regulate miRNA function through the competing endogenous RNA (ceRNA) network In this study, differentially expressed mRNAs, lncRNAs, circRNAs and miRNAs between model and acupuncture groups were systematically identi ed by the whole transcriptome sequencing (RNA-seq). 30 circRNAs, 47 lncRNAs, and 11 mRNAs with differential expression were found that shared the same binding site for miRNA. In the same way, we found 30 mRNAs, 162 lncRNAs and 88 circRNAs with differential expression were on the common binding site of miRNA between the model and normal groups (Fig. 10).

Discussion
HIBD is a common disease found in newborns which often presents as a series of neurological sequelae. Acupuncture, as a therapy of alternative medicine, has achieved a remarkable curative effect in the treatment of this disease 20,21 . Meanwhile, several studies have documented the important neuroprotective roles of acupuncture in cerebral functional compensation in animal models with HIBD [22][23][24] . With the rise of RNA-seq technology, people gradually became interested in the gene-related mechanism of HIBD and began to do research 25,26 . However, the role of acupuncture in regulating RNAs in HIBD rats remains unknown.
In this study, HIBD rats showed signi cantly longer new object recognition times, suggesting that HIBD rats may be worse than normal rats in terms of learning and memory. This result is consistent with a number of studies indicated that HIBD rats perform poorly in behavioral tests of learning and memory abilities, such as water maze, shuttle box, T maze, and radial maze [27][28][29] . As a common neonatal nervous system disease, HIBD can lead to different degrees of brain damage, such as brain metabolism disorder, neuronal apoptosis, reduced brain ow, or even cerebral hemorrhage in severe cases, resulting in liquefaction and necrosis of brain tissue [30][31][32] . Our results were in line with these ndings in that we found that there were signi cantly fewer Nissl corpuscles in the prefrontal lobe of HIBD rats, indicating that HIBD can indeed cause damage to neurons in the brains of rats. Acupuncture, as an alternative therapy, is attracting increasing attention for its ideal curative effect. Recently, it has been reported that acupuncture can improve locomotor activity and learning-memory ability by improving hippocampal cellular autophagy in rats with HIBD, suggesting that acupuncture could effectively alleviate the brain injury caused by HIBD 33 . Additionally, Yuan Q et al. found that acupuncture could signi cantly reduce the expression of CYT-C and caspase-3 in the cerebral cortex of HIBD rats and reduce the apoptosis of neurons to protect the brain tissue 11 . Consistent with these results, we found that the number of Neisseria bodies in HIBD rats treated with acupuncture was signi cantly higher than that in untreated rats. Additionally, the performance of the new object recognition experiment of the rats after acupuncture was better than that of the untreated HIBD rats, indicating that acupuncture is an effective therapy for HIBD.
In order to explore the target of acupuncture, we identi ed RNA differentially expressed in the PFC of three groups of rats and found something interesting: 1) lncRNAs had fewer exons, shorter lengths, and shorter lengths of the ORF than mRNAs; 2) from the gene heat map, we found that the effect of acupuncture intervention on lncRNA was more obvious than that of mRNA and circRNA. Moreover, the Venn diagram showed that the number of genes in the overlap part of lncRNA was higher than that of mRNA and circRNA; 3) 877 differentially expressed lncRNAs were identi ed, 65 lncRNAs were signi cantly and differentially expressed in acupuncture rats relative to model rats with 17 up-regulated and 48 downregulated, and 410 lncRNAs were signi cantly and differentially expressed in the model rats relative to the normal rats with 372 up-regulated and 38 down-regulated; 4) the GO analysis showed that lncRNAs that were differentially expressed between the acupuncture and model groups were involved in neurological system processes, plasma membrane, olfactory receptor activity, and receptor activity, whereas the lncRNAs that were differentially expressed between the model and normal groups were involved in olfactory receptor activity, G-protein-coupled receptor activity, and sensory perception. The KEGG pathway analysis revealed that lncRNAs that were differentially expressed between the acupuncture and model groups participated in multiple pathways, including olfactory transduction and ribosomes, cytokinecytokine receptor interactions, and taste transduction. In the comparison between the model and normal groups, the differentially expressed lncRNAs were associated with pathways, such as olfactory transduction, ribosomes, drug metabolism and other enzymes, and biosynthesis of amino acids; and 5) 30 circRNAs, 47 lncRNAs, and 11 mRNAs with differential expression shared a common binding site for miRNA between the acupuncture and model groups. Eighty-eight circRNAs, 162 lncRNAs, and 30 mRNAs with differential expression shared a common binding site for miRNA between the model and normal groups.
TCONS_00013568 was an up-regulated lncRNA in the model group but was decreased in the acupuncture group. Through KEGG pathway analysis, we found that the target gene of TCONS_00013568, Epo, was related to the Jak-STAT signaling pathway, HIF-1 signaling pathway, and PI3K-Akt signaling pathway.
Reportedly, the inhibition of the JAK/STAT pathway can not only promote brain-derived neurotrophic factors and hippocampal neuron proliferation to alleviate autism symptoms 34 but also protect against αsynuclein-induced neuroin ammation and dopaminergic neurodegeneration 35 . Some researchers have found that the improvement in EPO was associated with the neuroprotective effect of the JAK/STAT pathway in cerebral palsy 36 . The HIF-1 signaling pathway was found to be associated with stroke, vascular dementia, and chronic hypoxia [37][38][39] . Activating the HIF-1 signaling pathway could exert angiogenic and anti-apoptotic effects against cerebral ischemia-reperfusion injury 40 . Hippocampal neurons could be protected against excessive autophagy and apoptosis through the PI3K/Akt signaling pathway 41 . EPO was found to promote cerebrovascular regeneration by activating the PI3K/Akt signaling pathway in premature brain damage 42 . In our results, Cdkn2a, the other target gene of TCONS_00013568, was also related to glioma and the p53 signaling pathway. The results of several studies showed that the p53 signaling pathway was closely related to Alzheimer's disease, glioblastoma multiforme, and ischemic stroke [43][44][45] . It has been reported that Cdkn2a plays an important role in apoptosis by activating the p53 signaling pathway 46 . The new target genes of TCONS_00013568 in the above pathways may provide new therapeutic targets for the treatment of HIBD. However, the detailed mechanism needs further study.
TCONS_00068680 was another up-regulated lncRNA in the model group with down-regulation in the acupuncture group. Several target genes of TCONS_00068680 participated in the Jak-STAT signaling pathway, such as Il9r and Csf3. The expressed protein of Il9r is the receptor of interleukin (IL)-9. Some researchers found that IL-9 was important in driving immune responses to chronic in ammation and that the blockade of IL-9 signaling in the Jak-STAT signaling pathway may be novel therapeutic targets for in ammatory diseases in the mucosal immune system 47 . Furthermore, IL-9 is closely associated with multiple sclerosis and could reduce in ammation and provide protection from neurodegeneration 48 . The expressed protein of Csf3 is a potent activator of neutrophil ROS, which is related to stroke. It was found to be neuroprotective in acute and chronic settings through various mechanisms, including the enhancement of neurogenesis and angiogenesis, as well as the apoptosis and suppression of in ammation 49 . Moreover, Csf3 was involved in the PI3K-Akt signaling pathway. However, there are few studies on Csf3 in the PI3K-Akt signaling pathway, and the mechanism still needs further study. Fshr was another target gene of TCONS_00068680, which was associated with the p53 signaling pathway. Fshr was reported to be related to apoptosis and cell proliferation 50 , but few studies investigated the mechanism of Fshr in the p53 signaling pathway in HIBD.
TCONS_00310736 was a lncRNA found to be up-regulated in the model group and down-regulated in the acupuncture group. Its target mRNAs, Th and Slc6a3, were related to dopaminergic synapses and Parkinson's disease. Dopaminergic synapses were found to be closely related to schizophrenia and Parkinson's disease 51,52 . The expressed protein of Th is the tyrosine hydroxylase, which was reported to inhibit the synthesis of dopamine and cause Parkinson's symptoms by decreasing its enzymatic activity 53 . Moreover, the inhibition of Th in the substantia nigra could decrease the movement frequency 54 .
Reportedly, Slc6a3 is strongly linked to dopamine transport. Hyperactivity, reduced sensitivity to reward, and impaired social behavior were measured in mutant rats with Slc6a3 knock-out 55 . Htr1f was another target mRNA of TCONS_00310736 and was found to participate in the p53 signaling pathway. However, there are few studies on Htr1f in the p53 signaling pathway, and the mechanism still needs further study.
TCONS_00163999 with its target mRNA, Ccl12, associated with the TNF signaling pathway and NOD-like receptor signaling pathway, was also a lncRNA that was up-regulated in the model group and downregulated in the acupuncture group. TNF plays an important role in nerve in ammation, and the TNF signaling pathway is closely related to ischemic stroke, Alzheimer's disease, Parkinson's disease, amyotrophic lateral sclerosis, and multiple sclerosis 56 . Reportedly, the functions of NOD-like receptors could be divided into four parts, namely, the in ammasome assembly, signaling transduction, transcription activation, and autophagy 57 . Moreover, the NOD-like receptor protein 3 (NLRP3) in ammasome has been found to be closely associated with proptosis in encephalomyelitis 58 . However, the mechanism of Ccl12 in the TNF signaling pathway and NOD-like receptor signaling pathway is not clear.
TCONS_00001104 was lncRNA down-regulated in the model group and up-regulated in the acupuncture group. Its target gene, Gykl1, is involved in the PPAR signaling pathway. Reportedly, activating the PPAR signaling pathway could signi cantly attenuate HI-induced brain injury by reducing neuronal apoptosis 59 . However, there are a few studies on the function of Gykl1 in the PPAR signaling pathway. Il24, the other target gene of TCONS_00001104, was associated with the Jak-STAT signaling pathway. The expressed protein of IL-24 is interleukin-24, which was found to be closely related to apoptosis 60 .
Additionally, we found that two lncRNAs (TCONS_00173912 and TCONS_00057072) up-regulated in the model group and two lncRNAs (TCONS_00088551 and TCONS_00097914) down-regulated in the acupuncture group jointly regulate a target gene, Reln. Interestingly, Reln was found to be up-regulated in the model group and down-regulated in the acupuncture group and was closely related to the PI3K-Akt signaling pathway. Reln is reported to play an important role in neuronal structure maintenance of mature neurons and can regulate neuronal migration and synaptogenesis 61,62 . Moreover, some researchers found that Reln de ciency may be involved in the development of remote cognitive impairments and epigenetic regulation of DNA demethylation and histone acetylation of Reln might underlie the mechanisms of synaptic plasticity and memory retention in the medial PFC 63,64 . Reln also has a role in PI3-kinase signaling in neuronal growth cones, and it contributes to nal neuron positioning in the mammalian brain by local modulation of protein kinase B and glycogen synthase kinase 3beta kinase activities 65 . The results of our study suggested that the increasing expression of TCONS_00173912 and TCONS_00057072 and the decreasing expression of TCONS_00088551 and TCONS_00097914 may inhibit Reln, which then works via the PI3K-Akt signaling pathway in neonatal HIBD.
In summary, we found several signi cant lncRNAs whose corresponding target genes might be the targets of acupuncture therapy for HIBD. At the same time, we investigated the miRNA-mRNA-lncRNA interaction network of the normal, HIBD, and acupuncture groups using RNA-seq analysis. To the best of our knowledge, this is the rst research to study the expression of lncRNAs, circRNAs, and mRNAs between HIBD and acupuncture. The ndings further expanded our understanding of ceRNA networks and will help us explore the functions of acupuncture. These novel networks may be potential biomarkers or therapeutic targets of acupuncture in HIBD. The research provides new perspectives on the mechanism of acupuncture and may affect the diagnosis and therapy of HIBD. However, this study also has its limitations because of its small sample size. In future studies, we will further expand the sample size and introduce clinical samples to provide a more su cient basis for acupuncture treatment of HIBD.

Model
On the postnatal day 21 (PND21), the pregnant rats received a 10 min-delayed cesarean section after 12 h of preoperative fasting. They were weighed rst and then placed in a supine position on a surgical board after anesthesia with the abdominal cavity opened along the midline and the uterus exposed. Then, the bilateral uterine aortas were clamped for 10 min before the uterus was cut open and squeezed out in 2 min with their umbilical cords retained; the amniotic uid in the mouths and noses of the newborn rats was cleaned to promote breathing with dry cotton swabs. Subsequently, the newborn rats were placed on a hot plate covered with moist gauze at 33°C, and their backs were rubbed repeatedly with a cotton swab moistened with warm physiological saline for rescue. Only the baby rats that could breathe autonomously and were curled up with skin that could turn into pink were selected for our study. Then, the selected baby rats were suckled by the other female rats who delivered naturally. After 21 days of lactation, the male and female offspring were weaned and raised in different cages separately.

Groups
The neonatal rats underwent a 10 min-delayed cesarean section were assigned to model and acupuncture groups randomly. In the acupuncture group, neonatal rats underwent a 10 min-delayed cesarean section and received an acupuncture treatment(n=13). In the model group, neonatal rats only received a 10 min-delayed cesarean section(n=13). In the normal group, neonatal rats were all delivered naturally(n=13).
Treatments Normal group and model group received no treatments but be xed the same as the other groups. The neonatal rats in acupuncture group received acupuncture treatments. According to Jin's three-needle theory, three acupuncture point groups were selected: the 3-Points for Intelligence (DU24, GB13), the 3-Occipital Points (DU17, GB19). The rats were kept awake, and their hair were stroked gently by the intervention staff to make them mentally stable. Beauty acupuncture needles were used (0.25 mm × 10 m, Huatuo; Medical Instrument Co., Ltd., Suzhou, China) and acupuncture was applied at the corresponding points with the rat's neck was xed by index and middle ngers of the left hand. The needles were kept for 10mins after de qi with the rats placed on the platform 20cm away from the ground without environmental interference. The acupuncture treatments were started from PND14 and applied once a day for 2 weeks. What's more, all intervention were accomplished by the same person.

Novel object recognition
The rats in each group were subjected to a new object recognition experiment after 14 days of acupuncture intervention to observe their ability to recognize old and new objects. An 80 × 80 × 40 cm open box was prepared to have a black bottom and an HD camera 145 cm directly above it. The bottom of the box was divided into 16 squares of equal size with white lines. The environment was kept dark, and the open-eld box was clean throughout the experiment. During the adaptation period, the animals were put into an experimental box without any objects for 10 min a day for 3 days continuously. Familiar stage experiments were started on the fourth day, and the rats were adapted to the box for 2 min without any object inside before being taken out. Then, two identical objects (base can be xed) were put into the box, and the rats were placed into the box again. The exploration times for left objects (Tl) and right objects (Tr) for each rat within 5 min were recorded. After a familiarization period, the rats were put back into the cage, and the test period began 10 min later. During the test period, two different objects were put in the box, one of which was identical to the object from a familiar period and the other was a novel object. Then, the rats were placed in the box for 5 min, and the exploration times for the familiar object (Tf) and novel object (Tn) were recorded. Exploration activities were de ned as the rats' noses pointing at objects closely or smelling or licking objects directly. Non-exploratory activity occurred when other parts of the body touched an object without the nose pointing or the rat standing on it.

Nissl staining
After behavior testing, 3 rats were randomly selected from each group and perfused intracardially with 0.1 mol/L PBS and 4% paraformaldehyde before brains taken out. After xation in 4% paraformaldehyde at 4℃ for 24h, brains were dehydrated, embedded in para n, and then sectioned into a thickness of 4μm for Nissl staining. Morphological observation of brain tissue was performed under high-power light microscopy.

RNA isolation, library preparation, and sequencing
Total RNA was extracted from brain samples (prefrontal lobe) with Trizol reagent (Invitrogen) according to the manufacturer's instructions. RNA degradation and contamination were monitored on 1% agarose gels. The RNA purity, concentrations and integrity were assessed by NanoPhotometer® spectrophotometer (IMPLEN, CA, USA), Qubit® RNA Assay Kit in Qubit® 2.0 Fluorometer (Life Technologies, CA, USA) and RNA Nano 6000 Assay Kit for a Bioanalyzer 2100 system (Agilent Technologies, CA, USA).
A total of 3 µg of RNA per sample was used as input material for the RNA sample preparations. Under the manufacturer's recommendations, NEBNext® UltraTM RNA Library Prep Kit for Illumina® (NEB, USA) was used to generate the sequencing libraries with index codes added to each sample. Brie y, mRNA was puri ed from the total RNA using poly-T oligo-attached magnetic beads.The RNA fragmentation, rst and second strand cDNA synthesis, end repair, adaptor connection and PCR ampli cation were conducted with the manufacturer's protocol. After the library was constructed, Qubit 2.0 and Agilent 2100 were rst used for validation of RNA integrity and quantity. Subsequently, qRT-PCR was performed to accurately quantify the effective concentration of the library (the effective concentration of the library was higher than 3nm) to ensure the quality of the library.
According to the manufacturer's protocol, TruSeq PE Cluster Kit v3-cBot-HS (Illumina) was used to cluster of the index-coded samples on the cBot Cluster Generation System. the library preparations were sequenced on an Illumina Hiseq platform after cluster generation, and 125 bp/150 bp paired-end reads were generated.
Quality control, mapping, and quanti cation Raw data were rst processed through in-house perl scripts. Reads containing adapter, ploy-N, and lowquality reads were removed from raw data to obtain clean reads. At the same time, Q20, Q30, and the GC content of the clean data were calculated (Table S1). All the succeeding analyses were based on the clean data with high quality. STAR (v2.5.1b) was used to build the index of the reference genome downloaded from the genome website directly and to connect paired-end clean reads to the reference genome (Table S2). HTSeq v0.6.0 was used to count the reads mapped to each gene. Then, the FPKM of each gene was calculated to estimate gene expression levels (Table S3).

LncRNA functional prediction and expression analysis
Cis and trans genes of lncRNA were detected for further functional analysis. The cis gene is lncRNA that acts on adjacent target genes. The coding genes 10k/100k upstream and downstream of lncRNA were searched and analyzed for their function. The trans gene is lncRNA that can be used to identify each other by the expression level. The expressed correlation between coding genes and lncRNAs were calculated with custom scripts. Cuffdiff (v2.