Identification of lncRNA expression profile in the spinal cord of mice following spinal nerve ligation-induced neuropathic pain

Neuropathic pain is one of the most common chronic pain in humans and characterized by an increase in the responsiveness of nociceptive neurons in the peripheral and central nervous system (CNS) [1]. Peripheral and central sensitization represents the altered functional status of nociceptive neurons and results from changes of a vast amount of functional protein and signaling pathways in the neuron and glial cell [2, 3]. Recent pharmaceutical research and discovery activities focus on well-characterized molecular targets, such as ion channels, G-protein-coupled receptors, and kinases in neurons and glial cells localized along the nociceptive pathways, which are regarded as direct contributors to the sensitization of pain signaling systems [4, 5]. However, the transcriptional or translational regulatory mechanisms underlying the expression and functional changes of these molecules are poorly defined.

RNAs that do not code for a protein (noncoding RNAs, ncRNAs) consist of two major classes: the small ncRNAs, which include microRNAs (miRNAs) and other noncoding transcripts of less than 200 nucleotides, and long noncoding RNAs (lncRNAs), which are a novel class of non-protein coding transcripts longer than 200 nucleotides [6]. LncRNAs were initially considered as transcriptional by-products, but recent data suggest that lncRNAs can regulate gene expression via interfering with transcription, post-transcriptional processing, chromatin remodeling, miRNA sequestration, and generating small ncRNAs [7, 8]. Also, lncRNAs are involved in various aspects of cell biology and disease etiology, such as development [9], immune [10], cardiovascular disease [11], oncogenesis [12], and neurological disease [13]. LncRNAs are highly expressed in the CNS, and their expression profiles are associated with specific neuroanatomical regions, cell types, or subcellular compartments suggesting their potential functional roles in the nervous system [14‐16]. It was reported that sciatic nerve resection induced differential expression of lncRNAs in dorsal root ganglia (DRG) [17]. Moreover, Zhao et al. have recently identified a functional lncRNA Kcna2, which contributed to neuropathic pain by silencing Kcna2 in DRG neurons [18]. These findings indicate the involvement of lncRNAs in neuropathic pain.

The spinal cord is responsible for receiving input from nociceptors and projecting to the brain, and plays an important role in the integration and modulation of pain-related signals. To clarify the molecular mechanisms underlying neuropathic pain and explore novel approaches for analgesic strategies, herein, we investigated the genome-wide expression of lncRNAs in the spinal cord following L5 spinal nerve ligation (SNL)-induced neuropathic pain. We found a large number of differentially expressed (DE) lncRNAs and mRNAs in the spinal cord after SNL. Among them, 39 correlated lncRNA-mRNA pairs, consisting of DE lncRNAs and mRNAs with adjacent or overlapping position relationship, were screened out. Our findings will provide new insights into the roles of lncRNAs in the regulation of neuropathic pain-associated genes.

Chronic neuropathic pain is a somatosensory disorder caused by nerve injury or disease that affects the nervous system [35]. Evidence suggested that the particular patterns of gene expression at different levels of the nociceptive system play important roles in the development and maintenance of neuropathic pain [2, 36]. Over the past decades, the molecular mechanisms underlying neuropathic pain have been extensively studied; however, the pathophysiological process of pain is still vague. LncRNAs were recently shown to regulate gene expression [37] and traffic cellular protein complexes, genes, and chromosomes to appropriate locations [8]. Their function in regulating gene expression switching in the maintenance phase of neuropathic pain is poorly understood. In this study, we for the first time identified the global expression changes in lncRNAs and analyzed their characteristics and possible relation with coding genes in the spinal cord under neuropathic pain condition. The 24,833 lncRNAs were detected in the spinal cord of mice. Among them, 366 lncRNAs were up-regulated and 145 lncRNAs were down-regulated at 10 days after SNL. These DE lncRNAs are consistently altered in a high percentage of analyzed spinal cords from SNL and sham mice, suggesting that lncRNAs may be involved in neuropathic pain processing. So far, most DE lncRNAs have not been functionally characterized. Although it was still too early to translate this knowledge into the development of novel analgesic agents for better pain relief, these findings may likely provide novel insight into the molecular basis of pain.

In this study, the expression profiles of mouse genome-wide mRNAs were also detected using lncRNA Microarray Chip at the same time. Among DE mRNAs, the up-regulated mRNAs are far more numerous than the down-regulated in SNL samples, which reflects the emergence of new biology processes and pathways in pathological conditions. A number of reported pain-related genes, including Cacna1g, Trpv1, Ccl5, Cx3cr1 and Irf5 were dramatically increased after SNL. Moreover, a lot of other mRNAs, such as Sprr1a, Anxa10, Kng1, and Gpr151 (G-protein-coupled receptor 151), whose functions are unclear in the spinal cord were also screened out. As the expression changes for some genes may be related to nerve damage and homeostatic responses to denervation, further studies are needed to identify whether they are involved in neuropathic pain processing.

Based on the GO term enrichment analyses of DE mRNA, we found that significantly enriched molecular functions and biological processes of up-regulated gene in SNL vs sham were mainly involved in chemokine activity, inflammation, and immunity. These findings are consistent with previous studies showing that neuroinflammation, manifested as infiltration of immune cells [38], activation of glial cells [39] and production of inflammatory mediators [40] in the peripheral and CNS, plays an important role in the induction and maintenance of chronic pain [41]. Additionally, our immunostaining of GFAP and IBA-1 showed dramatic glial activation in the spinal cord at 10 days after SNL. From significant pathway analyses of DE gene, the third most significant enriched pathway of the up-regulated genes in SNL vs sham is the toll-like receptor signaling pathway. Indeed, Tlr2 [42], Tlr4 [43], and Tlr7 [44] have been implicated as potential therapeutic targets in neuropathic and other pain models. The data collectively indicate that anti-neuroinflammation may be an effective strategy for the treatment of neuropathic pain.

Previous studies utilizing cDNA microarrays to analyze gene expression profiles primarily focus on pain models in rats, rarely in mice [2]. The overlap analysis showed little overlap between rat and mice spinal cord gene expression patterns under neuropathic pain states, suggesting the species difference in gene expression. However, we found that there were 15 overlap genes between our current data and meta-analysis results reported by LaCroix-Fralish et al. [2]. These overlap genes including Atf3, Sprr1al and Nefm can be induced by nerve damage, which contribute to chronic pain [45‐47]. In addition, gene ontology-based functional annotation clustering analyses of the previous gene chip study revealed strong evidence for regulation of immune-related genes in pain states, which was consistent with our data.

Although lncRNAs play important roles in the regulation of gene expression [48], there is a large gap between the number of existing lncRNAs and their known association with a particular molecular or cellular function [49]. Regulatory mechanisms and major functional principles of lncRNAs are complex and quite obscure. Unlike microRNA, there are no common languages that can be used to predict lncRNAs’ target genes and function by their sequence information or secondary structure. Accumulating evidence suggests that a number of lncRNAs function locally to activate or repress their neighboring or overlapping genes’ expression [18, 27, 50]. In this study, we found that intergenic lncRNAs (lincRNAs) were the largest category in all DE lncRNAs after SNL. In reality, lincRNAs are found to be conserved across multiple vertebrate species [51] and perform important functions in many cellular processes, from cell proliferation to cancer progression [52]. Furthermore, lincRNAs can function through different types of mechanisms, including cis or trans transcriptional regulation, translational control, splicing regulation, and other post-transcriptional regulation [33]. We examined whether their neighboring or overlapping protein-coding genes in the genome are simultaneously DE in the spinal cord after SNL, and found that there are 39 DE lncRNA-mRNA pairs. Our further analysis showed that an up-regulated lincRNA, MM9LINCRNAEXON10576− in the spinal cord after SNL was found to be located near Cxcl10, Cxcl9 and Cxcl11 gene cluster in mice chromosome 5. All the four RNAs have the same expression trends and increased more than twofold after SNL. Recently, studies using animal models have shown that upregulation of chemokines in the spinal cord play a vital role in the development and maintenance of chronic pain [41, 53, 54]. Indeed, recent research found that Cxcl10 and its receptor Cxcr3 were involved in inflammatory pain and cancer pain [55‐57]. Therefore, lncRNA MM9LINCRNAEXON10576− may contribute to neuropathic pain through regulation of chemokines Cxcl10, Cxcl9 and Cxcl11.

In our microarray results, 12 DE mRNA have their corresponding DE sense-overlap lncRNAs, and the change patterns of these lncRNA were same as that of their accompanying protein-coding genes. Di et al. found that a sense-overlap lncRNA arising from the CCAAT/enhancer-binding protein alpha (Cebpa) gene locus can bind to DNA methyltransferase 1 (DNMT1) and prevent Cebpa gene locus methylation, then to increase the expression of Cebpa gene. Their deep sequencing of transcripts associated with DNMT1 combined with genome-scale methylation and expression profiling extend the generality of this finding to numerous gene loci. [27]. Given that the 12 DE mRNA and their DE sense-overlap lncRNAs were both increased after SNL, it’s possible that the DE sense-overlap lncRNAs regulate the expression of their sense-overlapping mRNAs via demethylation after SNL.

Our results demonstrated that lncRNA transcripts were highly enriched and hundreds of lncRNAs were differentially expressed in the spinal cord after SNL. Dozens of DE lncRNAs were observed to have neighboring or overlapping DE mRNAs in genome. These lncRNAs may locally regulate their related protein-genes expression and play key roles in the pathogenesis of neuropathic pain. Further studies are required to clarify the molecular and cellular functions of DE lncRNAs and determine whether they can serve as novel analgesic targets in neuropathic pain.