Insight into genetic regulation of mirna in mouse brain

RESEARCH ARTICLE Open Access Insight into genetic regulation of miRNA in mouse brain Gordon Kordas1* , Pratyaydipta Rudra2, Audrey Hendricks3, Laura Saba4† and Katerina Kechris1† Abstract Background m[.]

Kordas et al BMC Genomics (2019) 20:849 https://doi.org/10.1186/s12864-019-6110-6 RESEARCH ARTICLE Open Access Insight into genetic regulation of miRNA in mouse brain Gordon Kordas1* , Pratyaydipta Rudra2, Audrey Hendricks3, Laura Saba4† and Katerina Kechris1† Abstract Background: micro RNA (miRNA) are important regulators of gene expression and may influence phenotypes and disease traits The connection between genetics and miRNA expression can be determined through expression quantitative loci (eQTL) analysis, which has been extensively used in a variety of tissues, and in both human and model organisms miRNA play an important role in brain-related diseases, but eQTL studies of miRNA in brain tissue are limited We aim to catalog miRNA eQTL in brain tissue using miRNA expression measured on a recombinant inbred mouse panel Because samples were collected without any intervention or treatment (naïve), the panel allows characterization of genetic influences on miRNAs’ expression levels We used brain RNA expression levels of 881 miRNA and 1416 genomic locations to identify miRNA eQTL To address multiple testing, we employed permutation p-values and subsequent zero permutation p-value correction We also investigated the underlying biology of miRNA regulation using additional analyses, including hotspot analysis to search for regions controlling multiple miRNAs, and Bayesian network analysis to identify scenarios where a miRNA mediates the association between genotype and mRNA expression We used addiction related phenotypes to illustrate the utility of our results Results: Thirty-eight miRNA eQTL were identified after appropriate multiple testing corrections Ten of these miRNAs had target genes enriched for brain-related pathways and mapped to four miRNA eQTL hotspots Bayesian network analysis revealed four biological networks relating genetic variation, miRNA expression and gene expression Conclusions: Our extensive evaluation of miRNA eQTL provides valuable insight into the role of miRNA regulation in brain tissue Our miRNA eQTL analysis and extended statistical exploration identifies miRNA candidates in brain for future study Keywords: miRNA, eQTL, Hotspots, Mediation, Brain, Bayesian networks Background In recent years, there has been increasing interest in micro RNAs (miRNAs) [1] miRNAs are small (approximately 22 nucleotides in length) non-coding RNA known to influence gene expression by way of targeting messenger RNA (mRNA) Specifically, miRNAs will act to repress mRNA translation or increase mRNA degradation [2] miRNAs contain a small ‘seed’ region which is complementary to the 3′ untranslated region (UTR) * Correspondence: gordon.kordas@cuanschutz.edu † Laura Saba and Katerina Kechris are joint senior authors Department of Biostatistics and Informatics, Colorado School of Public Health, Aurora, CO 80045, USA Full list of author information is available at the end of the article of the mRNA(s) it targets [3] More than 60% of human mRNA genes have such target sites in their 3′ UTR [4] There are various miRNA biogenesis pathways [5] The ‘canonical’ biogenesis of a miRNA starts with primary miRNA (pri-miRNA) being transcribed by either RNA polymerase II or RNA polymerase III miRNA are transcribed from intronic regions (within a host gene) or from intergenic regions [6] The pri-miRNA is further prepared by the Drosha microprocessor complex and the characteristic hairpin is cleaved by the Dicer complex [5] The functional strand of the miRNA then combines with Argonaute proteins to form the RNAinduced silencing complex This complex can then perform cleavage, promote translational repression, or deadenylate target mRNA [5] At any point in this © The Author(s) 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated Kordas et al BMC Genomics (2019) 20:849 pathway there may be alterations or omissions that results in a non-linear pathway to a mature miRNA and thus, there exists various regulatory mechanisms of miRNA expression [5, 7] miRNAs can be down-regulated or upregulated and thereby, positively or negatively regulate gene expression respectively miRNAs are important for cell development (including the vascular, immune, and neurological cells) [8] miRNAs are also known to contribute to a wide variety of brain related diseases, including Alzheimer’s, Parkinson’s, Huntington’s and alcohol use disorders [8, 9] The link between genetic background and miRNA expression can be investigated through expression quantitative trait loci (eQTL) analysis, which examines regions of the genome (loci) that influence a quantitative trait [10] Here, the quantitative trait (i.e., continuous measure) is miRNA expression Most frequently the regions of the genome are represented by single nucleotide polymorphisms (SNPs) [10] eQTL can be placed in one of two categories depending on their genomic location Local eQTL are located near the gene (or miRNA) while distal eQTL are in a region far from the gene (or miRNA) Local and distal are often referred to as cis or trans, where cis implies variants affecting transcription factor binding sites or other regulatory sequences near a gene, and trans implies variants affecting changes in the structure or function of transcription factors or other regulatory proteins for a more ‘global’ effect [11] True cis effects are defined by Gilad as, “Regulatory elements [that] have an allele-specific effect on gene expression” [12] Examples of cis regulatory elements include, promotors and enhancer elements [12] We will assume that local implies cis and distal implies trans, but experimental validation is necessary to confirm these assumptions Many miRNA eQTL studies have been performed [13–19], but few examine miRNA specific to brain tissue [20, 21] Cataloging brain tissue miRNA eQTL in mice provides a way to uncover genetic influence on miRNA expression levels that is difficult to determine in humans because of the challenges of obtaining brain tissue and difficulty in limiting the variability due to environmental exposure Model organisms have the advantages of living in a controlled environment, and RNA samples from brain are easier to collect [22] By combining the information from brain eQTL in mouse models, we can provide candidate miRNAs for future mechanistic studies in animals, which will serve as an accompaniment to the more limited human brain studies Although in some cases specific mouse miRNA may not be conserved in humans, these miRNAs could still reveal biological mechanisms that are relevant in human Furthermore, many miRNA eQTL studies have limited their scope to Page of 14 only cis eQTL [19, 21] We will examine both cis and trans eQTL to gain more information on the regulation of miRNAs in brain The specific data used in this study are obtained from the LXS recombinant inbred (RI) panel This panel was derived from the parental Inbred Long (L) Sleep and Inbred Short (S) Sleep strains [23], which were originally selected to vary in the loss of righting reflex (LORR) behavioral phenotype and were later inbred over many generations The LORR phenotype is defined as the time it takes for a mouse to right itself in a v-shaped tray after being given a dose of ethanol [24] Long sleep strains take a longer time to right themselves compared to the short sleep strains and are, therefore, more sensitive to the hypnotic effects of ethanol RI panels allow for improved mapping power due to their ability to minimize environmental variability and to isolate genetic variability by taking measurements on numerous mice from the same strain [23] Another major advantage of the RI panel is that they are perpetually renewable and allow for the collection of many different traits by collaborating research teams over extended periods of time The LXS panel is also useful for investigating variation in non-alcohol related traits, and has been shown to vary in phenotypes such as longevity [25], and hippocampus weight [26] Furthermore, the advantage of using strains from a RI panel that have no experimental exposure (i.e., to ethanol) is that we can measure RNA expression levels that determine predisposition to a phenotype rather than expression levels that respond to an exposure We performed miRNA eQTL (mi-eQTL) analysis and mRNA, i.e gene, eQTL (g-eQTL) analysis on the LXS RI panel to better understand the role of genetic regulation of miRNA expression in the brain Related work included Rudra et al [24], which used the same miRNA brain expression data, but focused on a few specific alcohol related phenotypes, rather than taking a global approach Therefore, our work is presented as a comprehensive QTL study that is generalizable to other brain related traits This work helps fill the gap in mi-eQTL literature by providing resources specific to brain tissue, which is largely understudied We also reported the results of a hotspot analysis, which has the potential to uncover novel regulators of miRNA expression Finally, we integrated our results with available gene expression data on the same RI panel to examine the relationship between miRNAs and their associated gene targets via Bayesian network analysis The extensive evaluation of mi-eQTL allows us to obtain more information on the role of miRNA regulation in brain and generate a resource for researchers investigating miRNA in brain and brain related diseases Discovered mi-eQTL are available at PhenoGen (http://phenogen.org) Kordas et al BMC Genomics (2019) 20:849 Results mi-eQTL analysis mi-eQTL were obtained via correlation of miRNA expression and the genotype at a given genomic locus (see workflow in Additional file 1: Figure S3 and S4) Because of the multiplicity of SNPs across the RI panel, we test eQTL associations using strain distribution patterns (SDPs) (see Methods) Considering the power of our statistical tests due to the sample size and the nature of our permutation p-value calculation, each miRNA was limited to one genome-wide eQTL (across variants) represented by the maximum logarithm of the odds (LOD) score The LOD Score is a representation of eQTL strength and allows us to compare different types of mi-eQTLs by their statistical strength (Fig 1) 38 miRNAs (4.3% of all miRNAs tested) had a genome-wide significant mi-eQTL Significance was determined via a permutation threshold of 0.05 to account for multiple testing across SDPs and further false discovery rate (FDR) threshold of 0.05 (to adjust for multiple testing across miRNAs) Table contains all significant mi-eQTL and their corresponding Bayes’ 95% credible interval All mi-eQTL tested can be found on PhenoGen (see Data Availability section) and Additional file 1: Figure S1 contains a visualization of eQTLs via a boxplot illustrating the differences in miRNA expression between genetic variant Eight (21%) miRNA involved in mi-eQTL Page of 14 were novel and 14 (37%) were miRNA transcribed from intronic regions (Table 2) The majority of mi-eQTL are cis mi-eQTL (79%), leaving only eight trans mi-eQTL (mmu-miR-677-5p, mmu-miR-193a-3p, mmu-miR-69293p, mmu-miR-6516-5p, mmu-miR-381-5p, mmu-miR3086-5p, mmu-miR-32-3p, novel:chr4_10452) Human orthologs (of miRNA) can be found in Additional file 1: Table S1 Cis mi-eQTL compared to trans mi-eQTL have significantly higher LOD scores (p-value = 0.023; Fig 1a) Additionally, novel miRNAs have significantly higher LOD scores on average, compared to annotated miRNAs (p-value = 0.028; Fig 1b) However, there is no significant difference in mi-eQTL LOD score based on miRNA location (intronic versus non-intronic; Fig 1c) or between highly conserved miRNAs and lowly conserved miRNAs (p-value = 0.169; Fig 1d) The number of validated gene targets, as determined by MultiMiR [27] varied substantially between miRNAs (Table 2) Finally, we find a strong positive correlation between mi-eQTL LOD score and heritability of the miRNA involved (p-value = 3.67e-8; Fig 1e) mi-eQTL enrichment analysis We were only able to perform enrichment analysis on annotated miRNAs (30 of the 38 miRNAs with mi- Fig Comparisons of characteristics of mi-eQTL in brain with statistical significance Log transformed LOD scores are for visualization reasons only The actual calculations were done on untransformed LOD scores a The difference in mi-eQTL strength between cis and trans mi-eQTL (Wilcoxon summed rank test-statistic (W) = 183, p-value = 0.023) b The difference in mi-eQTL strength between mi-eQTL of annotated miRNA compared to mi-eQTL of novel miRNA (W = 59, p-value = 0.028) c The difference in mi-eQTL strength between mi-eQTL with miRNA in intronic locations compared to those in non-intronic locations (W = 229, p-value = 0.067) d The difference in strength between mi-eQTL involving miRNAs that were highly conserved (mean PhastCon conservation score above 0.5) compared to those involving lowly conserved miRNAs (W = 108, pvalue = 0.169) The conservation scores were dichotomized at 0.5 because that were often close to zero or one e The relationship between mieQTL strength and the heritability (measured by the intraclass correlation coefficient) of the miRNA involved (in the mi-eQTL) (rho = 0.82, p-value = 3.67e-8) Kordas et al BMC Genomics (2019) 20:849 Page of 14 Table Significant brain mi-eQTL and their characteristics miRNA eQTL chr eQTL location (Mb) eQTL 95% C.I eQTL LOD Genome-wide p-value FDR Cis/trans mmu-miR-8114 154.3 (152.9, 154.9) 10.11 0.0010 0.0251 C mmu-miR-32-3p 171.8 (171.1, 174.1) 5.51 0.0010 0.0251 T mmu-miR-1981-3p 187.2 (183, 188.7) 7.63 0.0010 0.0251 C mmu-miR-205-5p 193.6 (192.5, 193.6) 6.38 0.0010 0.0251 C mmu-miR-669o-5p 9.8 (5.6, 10.9) 5.32 0.0020 0.0463 C mmu-miR-467e-5p 10.6 (3.2, 10.9) 5.49 0.0010 0.0251 C mmu-miR-669a-5p 10.6 (8.1, 10.9) 6.29 0.0010 0.0251 C mmu-miR-297b-5p 10.6 (10.6, 10.9) 5.83 0.0020 0.0463 C mmu-miR-7674-5p 32.8 (32.8, 34.6) 6.77 0.0010 0.0251 C mmu-miR-466q 28.4 (28, 28.9) 18.86 0.0010 0.0251 C novel:chr4_9669 43.1 (32, 43.1) 12.41 0.0010 0.0251 C novel:chr4_11381 87.1 (87.1, 87.2) 23.47 0.0010 0.0251 C mmu-miR-9769-3p 30.6 (30.6, 30.6) 25.92 0.0010 0.0251 C mmu-miR-5121 43.5 (40.3, 45.5) 17.25 0.0010 0.0251 C mmu-miR-7057-5p 64.6 (57, 67.1) 24.43 0.0010 0.0251 C novel:chr8_23508 125.5 (125.4, 125.6) 21.89 0.0010 0.0251 C novel:chr9_24385 77.3 (77.3, 77.3) 19.99 0.0010 0.0251 C novel:chr4_10452 100.3 (94, 113.4) 5.78 0.0010 0.0251 T novel:chr10_26214 10 4.8 (4, 5.6) 32.53 0.0010 0.0251 C mmu-miR-6905-5p 10 25.3 (24.8, 25.9) 12.75 0.0010 0.0251 C novel:chr10_26328 10 25.3 (24.8, 25.9) 20.74 0.0010 0.0251 C mmu-miR-1934-5p 11 69.0 (67.8, 69.7) 20.39 0.0010 0.0251 C mmu-miR-193a-3p 11 74.3 (74.3, 79.5) 7.29 0.0010 0.0251 T mmu-miR-8103 11 95.1 (95.1, 99.6) 8.04 0.0010 0.0251 C mmu-miR-152-5p 11 96.2 (93.4, 103.4) 6.33 0.0010 0.0251 C mmu-miR-677-5p 11 98.3 (96.4, 101) 20.81 0.0010 0.0251 T mmu-miR-5621-5p 11 115.6 (115.4, 116) 25.18 0.0010 0.0251 C mmu-miR-208b-3p 14 54.7 (54.6, 55) 9.32 0.0010 0.0251 C novel:chr15_40280 15 94.8 (94.8, 95.5) 8.10 0.0010 0.0251 C mmu-miR-6516-5p 16 45.7 (45.7, 51.5) 5.94 0.0020 0.0463 T mmu-miR-381-5p 16 45.8 (45.7, 51.5) 5.59 0.0010 0.0251 T mmu-miR-6929-3p 19 30.0 (29.3, 30.5) 7.46 0.0010 0.0251 T mmu-miR-3086-5p 19 30.2 (29.3, 30.5) 5.06 0.0010 0.0251 T mmu-miR-201-5p X 66.6 (65.4, 66.6) 8.63 0.0010 0.0251 C mmu-miR-465c-5p X 66.6 (65.4, 82.2) 5.44 0.0010 0.0251 C mmu-miR-547-3p X 66.6 (65.4, 66.6) 9.58 0.0010 0.0251 C mmu-miR-871-3p X 66.6 (49, 67.4) 6.28 0.0010 0.0251 C mmu-miR-881-3p X 66.6 (65.4, 67.4) 6.13 0.0010 0.0251 C Abbreviations: Chr Chromosome, pos Position, Mb Megabase, C.I Bayes’ credible interval, LOD Logarithm of the odds score, FDR False Discovery Rate, cis/trans cis (within Mb on either side of the associated SDP) or trans (indicated by C or T) eQTL) Of those 30 miRNAs, three had no related KEGG pathway information for their target genes, and 13 had less than four target genes with KEGG pathways information Of the remaining 14 miRNAs with KEGG pathway information for at least four of their target genes, ten had brain-related KEGG pathways relevant to the nervous system, brain tissue, brain function or neurological/neuropsychiatric disease (Table 3) All results Kordas et al BMC Genomics (2019) 20:849 Page of 14 Table miRNA characteristics of those miRNA with significant mi-eQTL miRNA chr location start location end location miRNA type Annotation conservation ICC No targets mmu-miR-8114 153899989 153900009 I A 0.064 0.71 mmu-miR-32-3p 56895232 56895252 N A 1.000 0.31 mmu-miR-1981-3p 184822409 184822429 I A 0.062 0.56 mmu-miR-205-5p 193507503 193507524 N A 1.000 0.25 31 mmu-miR-669o-5p 10514318 10514340 N A 0.972 0.16 mmu-miR-467e-5p 10505731 10505752 N A NA 0.35 12 mmu-miR-669a-5p 10510185 10510208 N A NA 0.26 13 mmu-miR-297b-5p 10511686 10511707 N A 0.727 0.09 147 mmu-miR-7674-5p 32050946 32050969 I A 0.093 0.40 mmu-miR-466q 28419988 28420007 N A NA 0.80 175 novel:chr4_9669 41640264 41640319 N N 0.525 0.59 novel:chr4_11381 87071780 87071841 I N 0.264 0.88 mmu-miR-9769-3p 30552871 30552892 N A 0.820 0.89 mmu-miR-5121 45126925 45126945 N A 0.799 0.72 17 mmu-miR-7057-5p 66381702 66381719 I A 0.000 0.85 novel:chr8_23508 125837774 125837841 N N 0.003 0.84 novel:chr9_24385 74966743 74966804 I N 0.474 0.75 novel:chr4_10452 132310004 132310065 N N 0.822 0.56 novel:chr10_26214 10 4092814 4092873 I N 0.002 0.87 mmu-miR-6905-5p 10 24910669 24910691 I A 0.000 0.62 novel:chr10_26328 10 25416000 25416061 N N 0.923 0.71 mmu-miR-1934-5p 11 69663055 69663077 N A 0.000 0.60 13 mmu-miR-193a-3p 11 79712009 79712030 I A 1.000 0.33 mmu-miR-8103 11 97063829 97063849 N A 0.001 0.33 mmu-miR-152-5p 11 96850400 96850423 N A 0.996 0.48 mmu-miR-677-5p 10 128085291 128085312 I A 0.858 0.76 51 mmu-miR-5621-5p 11 115795824 115795846 N A 0.003 0.83 mmu-miR-208b-3p 14 54975710 54975731 N A 1.000 0.38 12 novel:chr15_40280 15 95488968 95489024 N N 0.001 0.38 mmu-miR-6516-5p 11 117077370 117077391 N A 0.974 0.51 mmu-miR-381-5p 12 109726829 109726851 I A 1.000 0.31 mmu-miR-6929-3p 11 101419187 101419209 I A 0.165 0.25 mmu-miR-3086-5p 19 58911725 58911744 N A 0.012 0.35 mmu-miR-201-5p X 67988135 67988156 I A 0.001 0.48 31 mmu-miR-465c-5p X 66832566 66832587 N A 0.079 0.39 24 mmu-miR-547-3p X 67988383 67988403 I A 0.025 0.50 14 mmu-miR-871-3p X 66810438 66810460 N A 0.001 0.43 mmu-miR-881-3p X 66801954 66801975 N A 0.045 0.50 28 Abbreviations: Chr Chromosome, annotation Annotated or novel (indicated by A or N), where novel miRNAs were identified by the mirDeep2 software, miRNA type intronic or non-intronic (indicated by I or N) as determined by the UCSC Genome Table Browser, conservation PhastCons Conservation Score (closer to indicates more highly conserved) where Not Applicable (NA) values indicate that a score was not returned by the Table Browser, ICC Intraclass correlation (a measure of miRNA heritability), No targets Number of validated gene targets identified by the MultiMiR R package Kordas et al BMC Genomics (2019) 20:849 Page of 14 Table Brain-related enriched pathways obtained for annotated miRNA with a significant mi-eQTL miRNA Brain Related KEGG Pathway # of genes FDR miR-547-3p Axon guidance 0.0165 mmu-miR-32-3p GABAergic synapse < 0.0001 mmu-miR-208b-3p Glutamatergic synapse 0.0007 Nicotine addiction 0.0045 Morphine addiction 0.0046 Amphetamine addiction 0.0052 Axon guidance 0.0095 Glioma 0.0007 Neurotrophin signaling pathway 0.0346 mmu-miR-8114 Axon guidance 0.0456 mmu-miR-677-5p mTOR signaling pathway 13 0.0037 Cocaine addiction 0.0273 mmu-miR-6929-3p Ubiquitin mediated proteolysis 0.0208 mmu-miR-465c-5p GABAergic synapse < 0.0001 Morphine addiction 10 < 0.0001 Nicotine addiction 0.0028 mmu-miR-193a-3p Glioma 0.0015 mmu-miR-466q Nicotine addiction 0.0038 mmu-miR-7674-5p Axon guidance 0.0010 FDR are the adjusted p-values Only pathways with or more genes and an FDR less than 5% are shown in the table Pathways were deemed brain related if the PubMed search of the pathway name AND the keyword “brain” yielded at least one abstract The abstract(s) were read to confirm brain related research from the enrichment analysis can be found in Additional file Hotspot analysis Figure provides a visualization of the mi-eQTL analysis by physical location of the loci and of the miRNA Although there are many cis mi-eQTL, indicated by points on the diagonal, there are also potential hotspots, indicated by vertical bands Potential hotspots were identified by dividing the genome into non-overlapping bins that were four SDPs wide (total number of bins equal to 354) Assuming mieQTLs were uniformly distributed across the genome, the counts of mi-eQTL in each bin follow a Poisson distribution [28] To obtain a Bonferroni corrected p-value less than 0.05, a hotspot must have contained more than six mi-eQTLs Using this cutoff, we identified seven bins with six or more mi-eQTL (see Fig and Table 4), that were collapsed into four final hotspots There were originally two additional hotspots on chromosome and one additional hotspot on chromosome 11 but they were collapsed with an adjacent hotspot (i.e the ending SDP of the first hotspot resided directly next to the starting SDP of the second hotspot) Three of the four hotspots overlapped addiction related behavioral QTLs We performed an enrichment analysis on the targets of any miRNA with mi-eQTL within a given hotspot using Diana-MirPath [32] (Additional file 1: Table S2) Of the nine miRNAs in the hotspots, seven had enrichment to a variety of functions including signaling and metabolism pathways Bayesian network analysis We tested triplets of SDP, miRNA, gene (i.e mRNA) for evidence of mediation, where the association of the SDP with the miRNA (or gene) is mediated by a gene (or miRNA) respectively Triplets were determined by the overlap of SDPs of the 38 significant mi-eQTL and SDPs of the 2389 significant g-eQTL (data not shown) Of the 175 possible triplets (SDPs, miRNA, mRNA), there were 11 significant triplets (p < 0.05) based on an initial mediation analysis (Additional file 1: Table S3) We then performed Bayesian Network Analysis (BNA) on these top mediation pathway candidates, which consist of four distinct miRNAs Bayesian networks that included all genes and all miRNA associated with a given SDP were fit (Fig 4) The Bayesian network results identified two types of mediation for the four, candidate miRNAs In one type of network, genes are acting as mediators of the effect of the genetic variant on miRNA expression (Fig 4a, b), while in the other miRNAs are acting as mediators of the effect of the genetic variant on gene expression (Fig 4c, d) The strength of associations was typically strong, Kordas et al BMC Genomics (2019) 20:849 Page of 14 Fig Chromosomal position of mi-eQTL Rows are miRNAs and columns are SDPs Scale is based on base pairs (bp) Blue spots indicate significant mi-eQTLs A relaxed p-value threshold of 5e-6 is used to help illustrate potential hotspots chromosome Number of significant eQTL 8 10 11 12 13 14 15 16 17 18 19 x Genome location Fig Brain mi-eQTL hotspots across the genome Locations with more than mi-eQTL cross the dotted line and indicate a significant hotspot is the threshold where the probability of getting more mi-eQTL in a bin is small (less than 0.05 after adjustments) Each color (as indicated by the legend) denotes the chromosome on which the significant mi-eQTL resides Black in the legend denotes there were no significant mi-eQTL The x-axis orders mi-eQTL from chromosome up to chromosome X and is not scaled to physical distance ... extensive evaluation of mi-eQTL allows us to obtain more information on the role of miRNA regulation in brain and generate a resource for researchers investigating miRNA in brain and brain related diseases... advantages of living in a controlled environment, and RNA samples from brain are easier to collect [22] By combining the information from brain eQTL in mouse models, we can provide candidate miRNAs... Many miRNA eQTL studies have been performed [13–19], but few examine miRNA specific to brain tissue [20, 21] Cataloging brain tissue miRNA eQTL in mice provides a way to uncover genetic influence