Metabolism is a term that describes all the chemical reactions that are involved in keeping a living organism alive. Diseases related to metabolism – such as obesity, heart disease and diabetes – are a major health problem in the Western world. The causes of these diseases are complex and include both environmental factors, such as diet and exercise, and genetics. Indeed, many genetic variants that contribute to obesity have been uncovered in both humans and mice. However, it is only dimly understood how these genetic variants affect the underlying networks of interacting genes that cause metabolic disorders.

Measuring gene activity or expression, and tracing how genetic instructions are carried from DNA into RNA and proteins, can reliably identify groups of genes that correlate with metabolic traits in specific organs. This strategy was successfully used in previous studies to reveal new information about abnormalities linked to obesity in specific tissues such as the liver and fat tissues. It was also shown that this approach might suggest new molecules that could be targeted to treat metabolic disorders.

A brain region called the hypothalamus is key to the control of metabolism, including feeding behavior and obesity. Hasin-Brumshtein et al. set out to explore gene expression in the hypothalamus of 99 different strains of mice, in the hope that the data will help identify new connections between gene expression and metabolism.

This approach showed that thousands of new and known genes are expressed in the mouse hypothalamus, some of which coded for proteins, and some of which did not. Hasin-Brumshtein et al. uncovered two genetic variants that controlled the expression of hundreds of other genes. Further analysis then revealed thousands of genetic variants that regulated the expression of, and type of RNA (so-called "spliceforms") produced from neighboring genes. Also, the expression of many individual genes showed significant similarities with about 150 metabolic measurements that had been evaluated previously in the mice.

This new dataset is a unique resource that can be coupled with different approaches to test existing ideas and develop new ones about the role of particular genes or genetic mechanisms in obesity. Future studies will likely focus on new genes that show strong associations with attributes that are relevant to metabolic disorders, such as insulin levels, weight and fat mass.

The regulation of body weight and appetite are complex processes, in which hypothalamic nuclei play a pivotal role. Genome wide association studies have shown that DNA sequence variants significantly contribute to variation in metabolic traits both in humans and mice. However, in most cases the connection between genetic variant and final phenotype remains unknown (Suhre et al., 2011; Teslovich et al., 2010; Lappalainen et al., 2013). In an effort to better understand how genetic variation results in phenotypic differences, many projects in the last decade have focused on genome wide characterization of sequence variants regulating gene expression in different tissues and organisms (Lappalainen et al., 2013; Majewski and Pastinen, 2011; Grundberg et al., 2011, , 2012). These studies showed that up to 80% of genetic variants associated with complex traits likely act through the regulation of gene expression rather than changing protein sequence and function. Such genes, termed expression quantitative trait loci (eQTLs), offer useful insights into the mechanistic links between genotype and phenotype, providing the eQTLs are characterized with sufficient power and resolution in phenotypically relevant tissues and states (Pickrell et al., 2010; Min et al., 2012; Gaulton et al., 2015).

Mouse is the primary model organism for many cardiovascular traits, including atherosclerosis, metabolic syndrome, obesity, the neural control of metabolism, and diabetes (Lusis, 2000; Billon et al., 2010; Allayee et al., 2003). Dozens of loci contribute genetically to metabolic traits have been identified in mouse models (Parks et al., 2013; Lusis, 2012). Indeed, we and others have extensively characterized eQTLs in metabolically relevant tissues of mice, suggesting potential new genes related to obesity (van Nas et al., 2010; Keane et al., 2011; Huang et al., 2009; Aylor et al., 2011). Integration of transcriptomic data from liver and adipose with genetic mapping and phenotypes led to mechanistic insights into the complexity of metabolic phenotypes. Yet, hypothalamus, which is not a readily accessible tissue, lacks such high resolution expression data. In fact, only one previous study examined eQTLs in mouse hypothalamus, using mice from 39 inbred strains fed chow diet, using microarray data and 5000 SNPs (McClurg et al., 2007). The lack of extensive transcriptome data, which allows mapping of eQTLs and the linking of traits and transcripts, is a major impediment in integrating the hypothalamus with a systems biology of metabolism. In this study, we used RNA-Seq to characterize the hypothalamic transcriptome in 99 inbred strains of mice from the Hybrid Mouse Diversity Panel (HMDP, [Bennett et al., 2010; Ghazalpour et al., 2012]) fed a high fat, high sugar diet (Parks et al., 2013). The advantage of the HMDP is that these strains have all been densely genotyped and carefully phenotyped for about 150 metabolic traits, allowing high resolution genetic mapping of QTLs and eQTLs. We identified thousands of novel isoforms and hundreds of new genes that were not previously annotated, and that may represent variants or transcripts specific to the hypothalamus. The HMDP also allowed us to map QTLs with high resolution and power, identifying both local and trans acting variants. The RNA sequencing data permitted examination of a much broader spectrum of transcriptional features and facilitated analysis not only at the gene level, but also of genetic variants affecting specific isoforms, coding sequences or transcription start sites.

In this study, we explored the transcriptional landscape of mouse hypothalamus using RNAseq from 282 mice, representing 99 inbred and recombinant inbred strains from the HMDP. We first focused on quantifying gene expression in a transcript specific manner, the discovery of novel expressed regions and isoforms, and the contribution of genetic factors to expression variance. We also sought qualitative support for translation of new isoforms and genes. We then examined and quantified RNA modifications, such as alternative splicing events and editing in our data and used the data to map genetic variants affecting these events. Finally, we used the extensive phenotyping available for the HMDP, to look for associations between the expression of genes and transcripts, and physiologic phenotypes. All of the expression data and transcriptome assembly are publicly available from the Gene Expression Omnibus database, accession number GSE79551.

Hypothalamic gene expression and proteomic data reveal multiple new isoforms and novel genes

Similar to other large-scale RNAseq studies (Mutz et al., 2013) we identified thousands of novel transcripts, with the vast majority of them only expressed at low levels in a small subset of samples (Table 1). Nevertheless, in the robust set of transcripts that are expressed at appreciable levels (FPKM >1 in 50+ samples and FPKM>0 in 100+ samples), we still identified 21,234 novel isoforms and 485 transcripts originating from 407 novel expressed genes (Supplementary file 1). Interestingly, the number of novel isoforms in our study is comparable to the number of previously annotated transcripts passing the same filtering criteria (n = 29,305, Table 1).

Table 1

	None	Filter #1*	Filter #2†	NR‡
Loci (genes)	40472	37591	14079	14079
Transcripts	383420	357066	51024	50347
known (=) coding	99658		20721
known (=) non coding			8584
novel isoform (j)	259570		21234
novel locus (u)	11753		485
other status	12439
TSS	100073	94417	32537	20013
CDS	46242	46242	18687	7643
Total features	570207	535316	116327	92082

1. *Filter #1: Expression values of <1e–6 were rounded to zero, and novel transcripts with all zero values were removed both from expression table and from merged file. Also, all transcripts with class code not "=", "j" or "u" were removed.

2. ^†Filter #2: Implementation of detection and expression thresholds (detected in more than 100 samples and expressed (fpkm>1) in more than 50) on each feature separately.

3. ^‡Filter_NR: Non redundant features count (those that do not have 1_2_1 correlation to a gene).

There are several possible interpretations to as why the 407 genes could be missing from the genomic annotation. First, the 407 novel genes expressed in our data potentially constitute transcripts that are unique to the hypothalamus. Second, the GENCODE M2 annotation used in this study was the most recent available when we started to analyze the data. Since then, however, the annotation has been augmented based on more recently published datasets and improved prediction pipelines. Thus, while the 407 genes are novel relative to the M2 version, they may have been added later. To explore that possibility, we compared our assembly to the latest version of annotation released by GENCODE – M10 (released January 2016, http://www.gencodegenes.org/) and find that 193 out of 407 loci are still novel. Some of the genes may also represent genomic DNA contamination. However, we consider this possibility less likely since we used stringent filtering criteria based on the number of samples that the genes were expressed in.

In terms of genomic properties, such as putative open reading frame length, transcript length or splicing complexity, the novel genes seem to resemble known non-coding genes, suggesting that the majority of the novel genes likely belong to this class. On the other hand, the properties of novel isoforms are consistent with known coding transcripts (Figure 1). To explore the translational potential of newly identified isoforms and genes, we compared our RNA-Seq data to proteomic data generated from additional hypothalamus samples from the HMDP. Specifically, we translated all known and newly identified transcripts in 6 potential open reading frames, and compared this dataset to the hypothalamic peptide sequences. As expected, >95% of the identified peptides (Supplementary file 1) matched at least one known transcript with the vast majority of these (>99%) corresponding to known protein coding transcripts, suggesting high accuracy of the peptide data. In addition, 430 peptides matched either novel isoforms (n = 401), or novel genes (n = 29) exclusively (Table 2). Since the genomic properties of the novel genes hint that the majority of them are likely non coding, we do not find their low representation in peptide data surprising. Moreover, manual in-depth characterization of the 29 peptides matching novel genes, with UCSC and NCBI database revealed that almost all of them represent processed pseudogenes, rather than proteins with novel functions. This finding both supports the previously published reports of wide spread transcription of pseudogenes, and their translation (Kim et al., 2014; Tay et al., 2015), while strengthening the suggestion that the majority of novel genes identified in this work are not protein coding.

Figure 1

Download asset Open asset

Table 2

Peptide matching*	All	Uniquely mapped
Known transcripts	9913	1016
Protein_coding	9839	1002
ncRNAs	74	14
Novel isoforms	401	94
Novel genes	29	24
Total number	10343	1134

1. *Peptides are counted towards supporting novel isoforms only if they do not match any known transcript, and towards supporting novel genes if they do not match either known or novel isoform transcripts. All peptides matching known transcripts were also assigned the most likely transcript. Non coding status was assigned only to peptides that do not match any coding transcripts (for full details please see Materials and methods).

We further examined whether multi-mapping reads are a substantial contributor to the measured expression of the novel identified genes. For that purpose, we re-quantified 3 samples from C57BL/6J strain, using uniquely mapping reads only (mapping quality = 255), and compared the quantification of the novel genes between the two approaches. Not surprisingly, the proportion of uniquely mapped reads contributing to the expression of the genes matched very well between the 3 samples, suggesting that it is an intrinsic gene property and not unique to the individual samples. Importantly, while the low uniqueness of mapping reads may indicate false results, we also noted that among the 29 peptides matching to the novel genes exclusively and uniquely, 18 match genes that do not pass our threshold. In fact, many of gene families share sequence motifs and are homologues, and as such some of the reads would be multi-mapping between them. Thus we chose not to remove these genes from the annotation.

Genetic regulation of gene expression

We examined expression QTLs in terms of type of the affected features and identified SNPs affecting expression of all transcripts at a locus (eQTL), specific transcript isoforms (isoQTLs), transcription start sites (tssQTLs) or open reading frame (cdsQTLs) (Hasin-Brumshtein et al., 2016). Since the linkage disequilibrium (LD) is extensive in the mouse genome, we distinguish between three types of QTLs: local (within 2 Mb of the gene), distant (on the same chromosome as the gene, distance >2 Mb) and trans (SNP and gene reside on different chromosomes). The number of identified interactions, and genes affected by such regulation depends on the statistical cutoff of p-value for the interaction one chooses. We examined a set of thresholds ranging from very liberal to stringent. The liberal threshold was previously established by permutation from microarray expression data in the HMDP, i.e. p<1.4E–3 for local variants and p<6E–6 for distant and trans. The most stringent threshold used a Bonferroni corrected threshold (i.e 0.01 divided by the number of tests, p<1E-12). As expected, different thresholds resulted in different numbers of QTLs, with local and distant QTLs being more robust to threshold stringency than trans QTLs. However, 85–95% of the genes regulated by distant variants, were also regulated by local variants, suggesting that most of the distant QTLs reflect local signals emulated at a distance as a result of wide-ranging LD, rather than independent regulatory elements. In contrast, 70% of the trans interactions involved genes lacking local signal over the entire set of thresholds. Since distant regulation was mostly redundant to local, and it would be very difficult to determine whether that signal is a result of independent regulation versus extended LD, we chose to focus our analysis on trans and local interactions only, defining trans as trans-chromosomal interactions. If we look at regulation types over a wide range of thresholds, the local signal predominates at the more stringent thresholds reflecting the larger typical effect size in this group (Figure 2A). We identified local and trans QTLs for all expressed feature types, with the most common being eQTLs and isoQTLs (Figure 2B). Notably, while we used kinship matrix specific for our strains, still several genes in our dataset show extensive trans regulation (horizontal lines in Figure 2B) suggesting a residual influence of population structure on our mapping results. We also note that regulation of gene expression often occurs at the gene level, than at transcript specific levels (Figure 2C).

Figure 2 with 1 supplement see all

Download asset Open asset

Classical genome-wide eQTL studies used association analysis of total gene expression levels to map local eQTL, assuming that variation linked to proximal genetic variant indicates cis regulation. Recently, several studies have exploited the single base resolution of RNA-Seq to examine this assumption. RNA-Seq permits the identification allele expression (ASE), a hallmark of functional cis acting regulation, in heterozygotes, such as humans or in mouse F1s. Notably, these studies generally show poor concordance between ASE ratios and previously identified local eQTLs, which had been attributed mostly to different technical aspects of RNAseq and ASE analysis (Hasin-Brumshtein et al., 2014; Lappalainen et al., 2013; Grundberg et al., 2012). To examine the concordance between local eQTLs obtained from ASE and genome-wide association, we performed RNA-Seq on brain tissue from 20 mice representing two F1 crosses between three of the classical inbred strains used to construct the HMDP. The parental strains were C57Bl/6 (B), A/J (A) and C3H (H), and the F1 offspring were BxA and HxB. We compared the effect size of ~2600 local eQTLs, to the average ASE effect in both crosses. While local eQTLs were significantly positively correlated with ASE (p<2E-16) correlation estimates were modest (R² = 0.2), and lower than between the ASEs in the two sets of F1 mice (R² = 0.4, Figure 2—figure supplement 1).

RNA-Seq shows extensive alternative splicing in the hypothalamus, with complex regulation pattern

Using a previously established pipeline for analysis of alternative splicing events, SUPPA (see Materials and methods), we identified 7564 alternative splicing events affecting 3599 genes (Table 3). An alternative first exon was most common, accounting for the majority of alternative splicing events with several genes exhibiting multiple alternative first exons (Figure 3A). For other types of events, each most often affected one gene, with some of the genes exhibiting a combination of alternative splicing events. The extent of alternative splicing in each sample was quantified as a percent spliced in (PSI) of every event (Figure 3B). This quantification can be regarded as a quantitative trait; however, the distribution of PSI for every type of event suggests an excess of 0 or 1 values (never or always spliced in, respectively Figure 3E). This observation is consistent with alternative splicing being often bimodal rather than a normally distributed continuous measure (e.g. present or absent splice site), however, it also shows that quantitative regulation plays a significant role in ratio of isoforms that arise from the inclusion or exclusion of a splicing event.

Figure 3

Download asset Open asset

Table 3

	All events	Cis QTL events
Alternative 3' splice site (A3)	316 (266)	155 (134)
Alternative 5' splice site (A5)	365 (304)	220 (189)
Alternative first exon (AF)	5288 (1874)	2776 (1278)
Alternative last exon (AL)	507 (320)	283 (208)
Mutually exclusive exons (MX)	44 (36)	29 (27)
Retained Intron (RI)	645 (476)	356 (285)
Skipped exon (SE)	399 (323)	221 (189)
Total	7564 (3599)	4040 (2310)

1. Table 2: Number in parenthesis indicates number of distinct genes affected by the events.

Contrary to gene expression, all forms of alternative splicing were predominantly regulated by local variants (Figure 3C,D). This is not unexpected, since variation in alternative splicing is most likely to arise from particular sequence variants in the RNA itself. Across all examined categories of alternative splicing, 50–60% of the events were significantly associated with local variants, the strongest signal often residing within 1 Mb of the event (Figure 3F). We observe that for many of the genetically regulated splicing events the data show an excess of 0 or 1 PSI values that correlate with the genetic variant (e.g Figure 3G), rather than a shift in a continuous quantitative distribution. This observation suggests that many of the genetically regulated events in alternative splicing sequence play a deterministic role.

Trans eQTL hotspots

Trans eQTLs are not uniformly distributed across the genome, clustering into potential hotspots of genome wide regulation. Such hotspots have been observed in several datasets (Orozco et al., 2015; Orozco et al., 2012; Tian et al., 2015), and they are thought to represent cases where a cis acting variant affects a regulatory gene, e.g. transcription factor, subsequently affecting expression of multiple targets. To identify trans hotspots we divided the genome into bins of 100 kb, and counted the number of distinct genes which exhibit a trans interaction associated with any of the SNPs in the bin (Figure 4A). Importantly, the bins do not necessarily contain the same number of SNPs, and are an order of magnitude shorter than the typical LD blocks in the inbred mouse genome. Just counting the number of SNPs in a bin that are associated in trans would be confounded by SNP density. However, since LD essentially recapitulates the same interactions over and over again, counting the number of genes per bin, rather than the number of SNPs, is not significantly affected by LD.

Figure 4

Download asset Open asset

There are clearly two very strong trans acting hotspots on chromosome 1 and 15, which are observable for eQTLs, but not for isoQTLs, tssQTLs or cdsQTLs. Each of the two hotspots span 5–6 Mb and include >500 SNPs regulating >400 genes in trans (Figure 4B, Supplementary file 2). Functional enrichment analysis of gene targets of these hotspots suggests that the hotspot on chr1 regulates multiple genes involved in nucleotide binding, while genes regulated by chr15 locus are clearly associated with ion transport in synapse activity (Figure 4C, Supplementary file 2). Moreover, consistent with the hypothesis that trans regulation is mediated by local effects, the trans acting SNPs in these hotspots, are also associated with expression of several local genes (Figure 4B,C, Supplementary file 2).

To identify potential mediators of trans regulation, we examined the trans eQTL hotspot on chr15 in more detail. In this region a cluster of SNPs, which share high linkage disequilibrium, is associated with the expression of the majority of the genes mapping to this locus in trans (Figure 4B). We focused on the SNP with most trans associations (rs31703733), and found that while the hotspot itself contains dozens of genes, rs31703733 constitutes a local eQTL for only 10 of them (Figure 4B). Moreover, the expression levels of genes regulated in trans by the hotspot as well as 6 genes that were local eQTLs were closely correlated (Figure 4C), forming a coexpression module. Further, this module was significantly associated with triglycerides and cholesterol measurements in the mice. Interestingly, the two strongest local eQTL genes for this hotspot, Endou (an endonuclease) and Rapgef3 (also known as EPAC, cAMP activated guanine exchange factor involved in Ras signaling) were significantly associated with cholesterol and triglyceride measures in this study as well.

Genetic association does not necessarily imply that the associated allele is causative for the change in expression. In fact, the majority of the eQTL SNPs fall into regions with poor or no annotation, making it unlikely to be the actual causative variant. However, surprisingly, peak SNPs in the chr15 hotspot (SNPs rs31703733 and rs31780670) are located within a H3K4me1 histone methylation mark associated with enhancers, and which was detected in all neuronal tissues probed by ENCODE in adult mice (cortex, cerebellum, olfactory bulb), but not in other metabolic tissues such as liver, heart, intestine or lung (UCSC genome browser). Based on this preliminary and indirect evidence one may hypothesize that rs31703733 and rs31780670 potentially affect the enhancer activity and the expression of nearby genes. However, to reach any functional conclusion based on our data would require experimental validation in model systems.

Long non-coding RNAs in the mouse hypothalamus

Long non coding RNA (lncRNAs) are a generally poorly characterized class of RNA molecules, with sometimes unclear classification (St Laurent et al., 2015). Commonly used criteria for identification of lncRNAs are a transcript >200 bp long without an obvious potential open reading frame. In contrast to the relative paucity of data regarding the functionality of most lncRNAs, several specific lncRNAs (e.g. Xist, HOTAIR or H19) had been studied extensively and shown to play an essential role in cellular function (Quek et al., 2015). Recent large RNA-Seq studies and integrative projects such as ENCODE suggest that lncRNAs likely constitute up to 60% of the transcribed RNAs (Djebali et al., 2012). Moreover, in recent years an increasing number of functional studies have shown that lncRNAs play an important role in the regulation of transcription and translation, as well as in cell differentiation, signaling and other processes (Sun et al., 2013; Guttman et al., 2011; Ramos et al., 2015). Further, lncRNAs are enriched for genetic association signals in genome wide association studies (Iyer et al., 2015; Kumar et al., 2013).

GENCODE M2 annotation classifies 4540 lncRNA transcripts into at least 5 biotypes, based on their overlap with protein coding genes (http://www.gencodegenes.org/mouse_releases/2.html). The two most common biotypes - long intergenic non coding RNAs (lincRNAs) and antisense RNAs, account for 97% of all the annotated lncRNAs. Since our data were generated without retaining strand specificity, we focused on lincRNAs only. LincRNAs have been shown to be highly tissue specific (Cabili et al., 2011), and therefore it is not surprising that although there are 2417 annotated lincRNA transcripts, we found only 381 expressed in the mouse hypothalamus at our filtering criteria. The 381 transcripts represent 237 known and 144 novel isoforms of 181 lincRNA genes, with both novel and known isoforms showing similar expression levels (Figure 5A,B). We also used length and open reading frame criteria to examine the novel loci. Notably, the RNA length criteria of >200 bp is a commonly accepted parameter, while the length of minimal potential open reading frame varies between 30–100 aa in different studies. Based on these criteria we can identify up to 129 potentially novel lincRNAs expressed in the mouse hypothalamus, with 49 of them being spliced. Since lincRNAs have been shown to be often highly tissue specific, we consider the novel lincRNAs to be good candidates for hypothalamus specific transcripts.

Figure 5

Download asset Open asset

We used gene expression levels to explore the correlation of 310 lincRNAs (181 known and 129 novel genes) with metabolic traits in HMDP. Up to 35% of lincRNA transcripts significantly (p<1e–3) correlated to at least one phenotype in the HMDP, with a few lincRNAs associated with a multitude of related phenotypes (Figure 5C,D). The statistical threshold was determined by permutations in previous studies of gene expression in HMDP, is less stringent than a Bonferroni correction, and reflects interdependencies among phenotypes. Expression of >30% of lincRNAs is associated with a significant local eQTL, suggesting that a considerable number of lincRNAs in the hypothalamus are playing an important role in translation of genetic regulatory variance into physiologic phenotypes. Notably, six lincRNAs – Meg3, Gm26924, Snhg4, Miat, 6330403K07Rik, and Malat1– were highly expressed in most strains (Figure 5A upper cluster). Meg3 is a known imprinted tumor suppressor gene (Zhang et al., 2010) and was recently implicated in hepatic insulin resistance (Zhu et al., 2016). Miat and Malat1 are both part of nuclear bodies. Knockout models of Malat1 are all viable and normal (Zhang et al., 2012; Eißmann et al., 2012; Nakagawa et al., 2012), however their response to metabolic challenges, such as high fat diet, had not been assessed. Myocardial infarction associated transcript (Miat) was initially linked to myocardial infarction through a genetic association study(Ishii et al., 2006). Subsequently, Miat was shown to regulate development of neuronal progenitors, involved in schizophrenia pathogenesis and fear response(Aprea et al., 2013; Liao et al., 2016). Although all of the six highly expressed lincRNAs are not novel, none of them were previously reported to be expressed in the hypothalamus, or to play an established role in the metabolic or reproductive system.

Previous investigations have documented several examples of lincRNAs that code for short translated open reading frames (Anderson et al., 2015). In addition, recent work has shown that lincRNAs can be associated with ribosomes, and their occupancy is similar to that of protein coding transcripts (Ingolia et al., 2014; Ruiz-Orera et al., 2014). We detected 6 peptides which mapped uniquely to 3 different lincRNAs (Gm26825, Gm16295, and Gm26593), suggesting that translation of lincRNAs occurs rarely, and that their association with ribosomes is more likely to be in the context of translational regulation of other protein coding transcripts.

RNA editing in the mouse hypothalamus

RNAseq provides the opportunity to look at RNA sequence modifications in a quantitative manner. In this study, we examined RNA editing patterns of all possible substitution types over the genome. A recent study showed that genetic variation affects C to U editing in the intestine, both in site specific and genome wide manners in the mouse Diversity Outbred cross (Gu et al., 2015). We hypothesized that genetic variation may contribute to the extent of RNA editing either in a site specific (e.g. by altering editing sites in cis) or genome wide manner (e.g. by altering the expression or specificity of editing enzymes). Altogether we detected 8462 potential editing sites in 3319 genes. As expected, the majority of edits (>70%) were A-to-G modifications, and the total number of detected sites in each strain varied between several hundred to over a thousand (Figure 6A). A comparison of our findings to previously reported editing sites in two databases - DARNED and RADAR - suggests that 75% of these are novel. We then used the proportion of edited reads per site, or per substitution type, across strains as a quantitative trait for mapping genetic variants that contribute to editing. We did not detect any significant genetic association for genome-wide levels of editing, which is consistent with both lack of genetic variation in ADAR in our panel and its very stable expression level across samples (49 ± 0.27 FPKM).

Figure 6

Download asset Open asset

Most editing sites were detected only in a small number of strains, which precluded meaningful mapping. For the site specific mapping we therefore focused on 90 sites that were detected in >70% of the strains (Figure 6C). We detected editing QTLs for 3 specific sites. For example, one of the editing sites in Ociad1 (on chr5 at 73312444) had a strong association (p<1e–8) with genetic variation residing on the same chromosome (Figure 6D). Altogether our data suggest that RNA editing occurs at low level in thousands of genes, however the impact of genetic variation on the editing level in mouse brain is limited.

Association of hypothalamic expression and phenotypes

There are 150 phenotypes available for the HMDP, reflecting many clinically relevant traits as well as various metabolic measures (Parks et al., 2013). We used several approaches to examine the potential role of hypothalamic expression in the various phenotypes. We began by identifying the top 500 transcripts that were significantly correlated with each of the phenotypes (correlation p values ranged from 4e–3 to <2e–16, Supplementary file 3). We then analyzed each of these expression sets for enrichment of potential pathways and functional annotation, using DAVID (Huang et al., 2009a, 2009b). Surprisingly, only a few of those expression sets showed moderate to strong enrichment (Figure 7B). For example, fat mass accumulation between 4 and 8 weeks, was distinctly associated with pathways related to oxidative phosphorylation and energy metabolism, while genes associated with food intake were enriched for ribosome related pathways. Clustering of phenotypes based on the proportion of shared genes associated with each trait (Figure 7A), clearly showed that related phenotypes also shared expression dependencies. For example, up to 70% of the genes most correlated with 'esterified cholesterol', were also correlated with 'total cholesterol', and more than 30% were shared with 'HDL'. Together these data validate the notion that related phenotypes share underlying molecular mechanisms, yet these shared genetic correlations may not necessarily correspond to specific readily identifiable pathways.

Figure 7

Download asset Open asset

We then analyzed whether the novel genes and isoforms we uncovered potentially contribute to phenotypic diversity. We found 232 novel transcripts associated with at least one phenotype (Supplementary file 3). For example TCONS_00279690, was strongly associated with several metabolic traits, such as total mass in response to a high fat, high carbohydrate dietary challenge, initial mouse insulin levels, as well as weight and fat mass (Supplementary file 3). Similar to known isoforms, expression of ~30% of the novel isoforms was significantly correlated with phenotypes. Correlation coefficients and p values were also similar between known and novel isoforms, suggesting that novel isoforms are as likely to contribute to mouse diversity as previously identified transcripts.

In this work we present a comprehensive picture of the transcriptome of the mouse hypothalamus and its genetic variation and regulation. We identify thousands of new isoforms, and >400 new genes, and show independent support for these being translated into protein, which further validates our data. Notably, transcription of pseudogenes had been noticed previously, and likely plays a role in gene regulation. Recent shotgun proteomic studies of the human proteome strongly suggest that a sizable fraction of pseudogenes and lncRNAs is translated (Ji et al., 2015; Kim et al., 2014). The peptide data from our study supports low level translation of processed pseudogenes and is in line with these results.

The hypothalamus is a highly heterogeneous tissue with multiple nuclei and cell types acting in concert. A recent RNA-seq study of fed and food-deprived mice showed that cell type specific transcripts in hypothalamic Agrp and Pomc neurons exhibited specific co-expression networks associated with feeding (Henry et al., 2015). In contrast, due to the practical constrains of the study, our data were collected on the entire hypothalamus, and as such would be less sensitive to cell type specific signals. Nevertheless, expression of cell-specific markers and functional molecules showed that our approach recapitulates known correlations between genes and metabolic phenotypes, and identifies new ones. In addition, our data capture the various non neuronal cell types, such as microglia or astrocytes, which are often overlooked in the mostly neuron focused studies of the hypothalamus. These cells are important mediators of hypothalamic inflammation and other processes induced by a high fat diet. Regulation of gene expression in this population impacts every aspect of metabolism. While cell type specific transcriptomics is valuable for understanding cellular processes in hypothalamic neurons, our data provide a robust framework recapitulating transcriptional processes affecting multiple cell populations. Our approach is thus complementary to the cell type-specific transcriptomic efforts.

In this study, we showed extensive genetic regulation of transcription and alternative splicing in the hypothalamus and identified two loci which influence transcription of hundreds of genes in trans. In addition, our data indicate that a considerable proportion of the new isoforms and transcripts are significantly correlated to physiological phenotypes. While human studies generally lack the power to address trans regulation on a genome-wide scale, the HMDP provides a powerful resource for such analysis. Indeed, we identified two very strong trans acting hotspots that seem to harbor major regulators of gene expression in hypothalamus. We suggest that the trans effects of genetic variation in these regions are likely mediated by local interactions, which is consistent with previously observed cases (Small et al., 2011; Heinig et al., 2010). Enrichment analyses clearly suggest that each trans eQTLs hotspot regulates a set of functionally enriched genes (e.g. the hotspot on chromosome 15 is strongly associated with ion transport in synaptic areas) suggesting a new link between genetic variants in these loci and specific cellular function. We further showed that the genes associated with these hotspots correlate to physiological phenotypes, such as HDL and triglyceride levels providing insight into the mechanism behind correlation of these genotypes with complex traits. The connection between the associated genes and traits does not imply direct causality. For example, ion transport is regulated by circadian rhythm (Ko et al., 2009), which in turn is associated with many other aspects of metabolism (Tsuneki et al., 2016).

Notably, although several examples of trans eQTL hotspots were published and analyzed (Small et al., 2011, Heinig et al., 2010; Tian et al., 2016), authenticity of such hotspots remains somewhat controversial, and trans eQTL hotspots may arise due to uncontrolled batch effects (Breitling et al., 2008; Kang et al., 2008; Joo et al., 2014), which are difficult to distinguish from real interactions. To minimize technical batch effects arising from sequencing, we used a round robin approach where each sample was sequenced as part of several pools (see details in Materials and methods). Batch effects may also arise from random environmental exposures, rather than technical sample preparation. In such case, it is unlikely that those effects would be tissue specific or affect only gene level analysis. We found no indication of these trans eQTL hotspots in the adipose or liver data from the same cohort of mice (Parks et al., 2013), nor were the hotspots detected for isoQTL. Thus, while we cannot exclude the possibility that trans eQTL hotspots described in this study may had arisen from unaccounted environmental or technical effects, we believe that this is unlikely, but further molecular studies are required to validate our results.

RNA-Seq allowed us to quantify transcripts at different levels of analysis- from the total expression of all isoforms in a locus, to transcript specific estimates and their combinations. Our analysis showed that regulation of the total number of transcripts from a gene is far more common than isoform specific regulation. Nevertheless, we were able to identify specific interactions at every level of detail we explored, namely eQTLs, isoQTLs, tssQTLs and cdsQTLs. In addition, we identified and quantified >7000 alternative splicing events affecting >3500 genes in the hypothalamus, and showed that these events are mostly affected by extensive local genetic regulation. In many cases of alternative splicing QTLs, the associated variant resulted in an excess of either 0 or 1 splice PSI values, suggesting that variants affecting splicing act as strong determinants rather than weak contributors to a complex trait.

RNA editing is a result of post transcriptional deamination processes whereby an adenosine is converted to inosine (A-to-I), or cytosine to uracil (C-to-U). Both types of RNA editing are mediated by specific enzymes members of the ADAR family facilitate the A-to-I editing which is most commonly observed in neuronal tissues, while Apobec1 mediates the C-to-U editing, and is primarily expressed in the intestine and liver. Editing is a tissue specific process which usually results in modification of only fraction of the transcripts, and therefore can be regarded as a quantitative trait. Previous studies showed that polymorphisms either in the editing enzymes or in the sequence proximal to editing site affect the extent of editing. Specifically, Gu et al recently reported a sequence variant in Apobec1 which affects in trans multiple C-to-U editing sites in the mouse liver (Gu et al., 2015). As expected from previous studies, A-to-I was by far the most common RNA editing event in our study. Further, our data show that RNA editing occurs at low levels in thousands of sites, but is highly variable. Less than <1% of the sites were consistently detected in the adult mouse brain across >75% HMDP strains. Consistent with the lack of coding genetic variation in the ADAR enzymes in our mouse panel, and with their invariable expression levels, we did not observe any trans editing QTL that would affect editing levels of multiple sites. However, we found a few local associations that seem to affect editing of specific sites. For example a strong local editing QTL (p<e–7) was observed for one of the editing sites in the Ociad1 gene (also known as Asrij). This editing QTL is likely due to a sequence variant in or near the Ociad1 gene. Notably, Ociad1 is expressed in stem cells, where it regulates pleuropotency via the JAK/STAT pathway (Sinha et al., 2013). Editing of Ociad1 or its expression in neuronal tissue was not reported before.

Another significant aspect of our data relates to long non-coding RNAs. LincRNAs have been shown to play an important role in various cellar functions. Recent genome annotations include thousands of known lincRNAs, yet most of them remain functionally uncharacterized, and only a few studies have examined the genetic control of their expression in detail. Moreover, lincRNAs are often expressed in a tissue specific manner, and therefore are not readily identifiable from general expression datasets. In this study, we detected expression of 381 known and novel lincRNA isoforms, and also identified 129 novel, potentially hypothalamus- specific lincRNAs. Our data indicate that lincRNAs are subject to similar variation in expression, and exhibit similar overall genetic control as the coding genes. Specific lincRNAs have been implicated in a variety of phenotypes (Guttman et al., 2011; Huarte et al., 2010; Kumar et al., 2013), and our data indicate strong correlations between some of the hypothalamic lincRNAs and metabolic phenotypes, such as body weight.

The hypothalamus is a very heterogeneous tissue, and one of the major drawbacks of our analysis is that we used whole hypothalamus, rather than dissecting specific nuclei. This limitation stems from practical considerations – meaningful expression QTL analysis requires sacrifice of hundreds of mice, while the dissection of specific hypothalamic nuclei is delicate and time consuming and thus was not feasible within the constraints of this study. Still, this shortcoming is likely to limit our power to detect meaningful associations, rather than introducing spurious ones. Moreover, since our study mostly focuses on genetic regulation of transcription, which was shown to be largely shared among tissues and cell types, we believe that the data presented in this work are not substantially confounded by heterogeneity.

To summarize, our data fill a substantial gap and will be useful to the research community. The hypothalamus is composed of multiple nuclei, which are distinct in morphology and function, and many laboratories focus on disentangling the complex interactions that ultimately affect metabolism and behavior. However, genome wide transcriptome analysis of this tissue has not been published, and genetic regulation of transcription as well as tissue specific transcripts remains largely obscure. We believe that our study is complimentary to physiological studies and will facilitate research into the crosstalk between the brain and other metabolic tissues. All of the expression data described in this paper are publicly available from NCBI archives GEO (GSE79551) and SRA (project number PRJNA314533).

1. 1. Yehudit Hasin-Brumshtein
   2. Arshad H Khan
   3. Farhad Hormozdiari
   4. Calvin Pan
   5. Brian W Parks
   6. Vladislav A Petyuk
   7. Paul D Piehowski
   8. Anneke Bruemmer
   9. Matteo Pellegrini
   10. Xinshu Xiao
   11. Eleazar Eskin
   12. Richard D Smith
   13. Aldons J Lusis
   14. Desmond J Smith

   (2016) Data from: Hypothalamic transcriptomes of 99 mouse strains reveal trans eQTL hotspots, splicing QTLs and novel non-coding genes

   Available at Dryad Digital Repository under a CC0 Public Domain Dedication.

   http://doi.org/10.5061/dryad.vm525
2. 1. Lusis AJ
   2. Smith D
   3. Hasin-Brumshtein Y

   (2016) Hypothalamic transcriptome of male mice on high fat diet, from 99 strains

   Publicly available at the NCBI Gene Expression Omnibus (accession no: GSE79551).

   http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE79551

1. 1. Ahn J
   2. Xiao X

   (2015) RASER: reads aligner for SNPs and editing sites of RNA

   Bioinformatics 31:3906–3913.

   https://doi.org/10.1093/bioinformatics/btv505

   • Google Scholar
2. 1. Allayee H
   2. Ghazalpour A
   3. Lusis AJ

   (2003) Using mice to dissect genetic factors in atherosclerosis

   Arteriosclerosis, Thrombosis, and Vascular Biology 23:1501–1509.

   https://doi.org/10.1161/01.ATV.0000090886.40027.DC

   • Google Scholar
3. 1. Anderson DM
   2. Anderson KM
   3. Chang CL
   4. Makarewich CA
   5. Nelson BR
   6. McAnally JR
   7. Kasaragod P
   8. Shelton JM
   9. Liou J
   10. Bassel-Duby R
   11. Olson EN

   (2015) A micropeptide encoded by a putative long noncoding RNA regulates muscle performance

   Cell 160:595–606.

   https://doi.org/10.1016/j.cell.2015.01.009

   • Google Scholar
4. 1. Aprea J
   2. Prenninger S
   3. Dori M
   4. Ghosh T
   5. Monasor LS
   6. Wessendorf E
   7. Zocher S
   8. Massalini S
   9. Alexopoulou D
   10. Lesche M
   11. Dahl A
   12. Groszer M
   13. Hiller M
   14. Calegari F

   (2013) Transcriptome sequencing during mouse brain development identifies long non-coding RNAs functionally involved in neurogenic commitment

   The EMBO Journal 32:3145–3160.

   https://doi.org/10.1038/emboj.2013.245

   • Google Scholar
5. 1. Aylor DL
   2. Valdar W
   3. Foulds-Mathes W
   4. Buus RJ
   5. Verdugo RA
   6. Baric RS
   7. Ferris MT
   8. Frelinger JA
   9. Heise M
   10. Frieman MB
   11. Gralinski LE
   12. Bell TA
   13. Didion JD
   14. Hua K
   15. Nehrenberg DL
   16. Powell CL
   17. Steigerwalt J
   18. Xie Y
   19. Kelada SN
   20. Collins FS
   21. Yang IV
   22. Schwartz DA
   23. Branstetter LA
   24. Chesler EJ
   25. Miller DR
   26. Spence J
   27. Liu EY
   28. McMillan L
   29. Sarkar A
   30. Wang J
   31. Wang W
   32. Zhang Q
   33. Broman KW
   34. Korstanje R
   35. Durrant C
   36. Mott R
   37. Iraqi FA
   38. Pomp D
   39. Threadgill D
   40. de Villena FP
   41. Churchill GA

   (2011) Genetic analysis of complex traits in the emerging collaborative cross

   Genome Research 21:1213–1222.

   https://doi.org/10.1101/gr.111310.110

   • Google Scholar
6. 1. Bennett BJ
   2. Farber CR
   3. Orozco L
   4. Kang HM
   5. Ghazalpour A
   6. Siemers N
   7. Neubauer M
   8. Neuhaus I
   9. Yordanova R
   10. Guan B
   11. Truong A
   12. Yang WP
   13. He A
   14. Kayne P
   15. Gargalovic P
   16. Kirchgessner T
   17. Pan C
   18. Castellani LW
   19. Kostem E
   20. Furlotte N
   21. Drake TA
   22. Eskin E
   23. Lusis AJ

   (2010) A high-resolution association mapping panel for the dissection of complex traits in mice

   Genome Research 20:281–290.

   https://doi.org/10.1101/gr.099234.109

   • Google Scholar
7. 1. Billon N
   2. Kolde R
   3. Reimand J
   4. Monteiro MC
   5. Kull M
   6. Peterson H
   7. Tretyakov K
   8. Adler P
   9. Wdziekonski B
   10. Vilo J
   11. Dani C

   (2010) Comprehensive transcriptome analysis of mouse embryonic stem cell adipogenesis unravels new processes of adipocyte development

   Genome Biology 11:R80.

   https://doi.org/10.1186/gb-2010-11-8-r80

   • Google Scholar
8. 1. Breitling R
   2. Li Y
   3. Tesson BM
   4. Fu J
   5. Wu C
   6. Wiltshire T
   7. Gerrits A
   8. Bystrykh LV
   9. de Haan G
   10. Su AI
   11. Jansen RC

   (2008) Genetical genomics: spotlight on QTL hotspots

   PLoS Genetics 4:e1000232.

   https://doi.org/10.1371/journal.pgen.1000232

   • Google Scholar
9. 1. Cabili MN
   2. Trapnell C
   3. Goff L
   4. Koziol M
   5. Tazon-Vega B
   6. Regev A
   7. Rinn JL

   (2011) Integrative annotation of human large intergenic noncoding RNAs reveals global properties and specific subclasses

   Genes & Development 25:1915–1927.

   https://doi.org/10.1101/gad.17446611

   • Google Scholar
10. 1. Djebali S
    2. Davis CA
    3. Merkel A
    4. Dobin A
    5. Lassmann T
    6. Mortazavi A
    7. Tanzer A
    8. Lagarde J
    9. Lin W
    10. Schlesinger F
    11. Xue C
    12. Marinov GK
    13. Khatun J
    14. Williams BA
    15. Zaleski C
    16. Rozowsky J
    17. Röder M
    18. Kokocinski F
    19. Abdelhamid RF
    20. Alioto T
    21. Antoshechkin I
    22. Baer MT
    23. Bar NS
    24. Batut P
    25. Bell K
    26. Bell I
    27. Chakrabortty S
    28. Chen X
    29. Chrast J
    30. Curado J
    31. Derrien T
    32. Drenkow J
    33. Dumais E
    34. Dumais J
    35. Duttagupta R
    36. Falconnet E
    37. Fastuca M
    38. Fejes-Toth K
    39. Ferreira P
    40. Foissac S
    41. Fullwood MJ
    42. Gao H
    43. Gonzalez D
    44. Gordon A
    45. Gunawardena H
    46. Howald C
    47. Jha S
    48. Johnson R
    49. Kapranov P
    50. King B
    51. Kingswood C
    52. Luo OJ
    53. Park E
    54. Persaud K
    55. Preall JB
    56. Ribeca P
    57. Risk B
    58. Robyr D
    59. Sammeth M
    60. Schaffer L
    61. See LH
    62. Shahab A
    63. Skancke J
    64. Suzuki AM
    65. Takahashi H
    66. Tilgner H
    67. Trout D
    68. Walters N
    69. Wang H
    70. Wrobel J
    71. Yu Y
    72. Ruan X
    73. Hayashizaki Y
    74. Harrow J
    75. Gerstein M
    76. Hubbard T
    77. Reymond A
    78. Antonarakis SE
    79. Hannon G
    80. Giddings MC
    81. Ruan Y
    82. Wold B
    83. Carninci P
    84. Guigó R
    85. Gingeras TR

    (2012) Landscape of transcription in human cells

    Nature 489:101–108.

    https://doi.org/10.1038/nature11233

    • Google Scholar
11. 1. Eißmann M
    2. Gutschner T
    3. Hämmerle M
    4. Günther S
    5. Caudron-Herger M
    6. Groß M
    7. Schirmacher P
    8. Rippe K
    9. Braun T
    10. Zörnig M
    11. Diederichs S

    (2012) Loss of the abundant nuclear non-coding RNA MALAT1 is compatible with life and development

    RNA Biology 9:1076–1087.

    https://doi.org/10.4161/rna.21089

    • Google Scholar
12. 1. Gaulton KJ
    2. Ferreira T
    3. Lee Y
    4. Raimondo A
    5. Mägi R
    6. Reschen ME
    7. Mahajan A
    8. Locke A
    9. Rayner NW
    10. Robertson N
    11. Scott RA
    12. Prokopenko I
    13. Scott LJ
    14. Green T
    15. Sparso T
    16. Thuillier D
    17. Yengo L
    18. Grallert H
    19. Wahl S
    20. Frånberg M
    21. Strawbridge RJ
    22. Kestler H
    23. Chheda H
    24. Eisele L
    25. Gustafsson S
    26. Steinthorsdottir V
    27. Thorleifsson G
    28. Qi L
    29. Karssen LC
    30. van Leeuwen EM
    31. Willems SM
    32. Li M
    33. Chen H
    34. Fuchsberger C
    35. Kwan P
    36. Ma C
    37. Linderman M
    38. Lu Y
    39. Thomsen SK
    40. Rundle JK
    41. Beer NL
    42. van de Bunt M
    43. Chalisey A
    44. Kang HM
    45. Voight BF
    46. Abecasis GR
    47. Almgren P
    48. Baldassarre D
    49. Balkau B
    50. Benediktsson R
    51. Blüher M
    52. Boeing H
    53. Bonnycastle LL
    54. Bottinger EP
    55. Burtt NP
    56. Carey J
    57. Charpentier G
    58. Chines PS
    59. Cornelis MC
    60. Couper DJ
    61. Crenshaw AT
    62. van Dam RM
    63. Doney AS
    64. Dorkhan M
    65. Edkins S
    66. Eriksson JG
    67. Esko T
    68. Eury E
    69. Fadista J
    70. Flannick J
    71. Fontanillas P
    72. Fox C
    73. Franks PW
    74. Gertow K
    75. Gieger C
    76. Gigante B
    77. Gottesman O
    78. Grant GB
    79. Grarup N
    80. Groves CJ
    81. Hassinen M
    82. Have CT
    83. Herder C
    84. Holmen OL
    85. Hreidarsson AB
    86. Humphries SE
    87. Hunter DJ
    88. Jackson AU
    89. Jonsson A
    90. Jørgensen ME
    91. Jørgensen T
    92. Kao WH
    93. Kerrison ND
    94. Kinnunen L
    95. Klopp N
    96. Kong A
    97. Kovacs P
    98. Kraft P
    99. Kravic J
    100. Langford C
    101. Leander K
    102. Liang L
    103. Lichtner P
    104. Lindgren CM
    105. Lindholm E
    106. Linneberg A
    107. Liu CT
    108. Lobbens S
    109. Luan J
    110. Lyssenko V
    111. Männistö S
    112. McLeod O
    113. Meyer J
    114. Mihailov E
    115. Mirza G
    116. Mühleisen TW
    117. Müller-Nurasyid M
    118. Navarro C
    119. Nöthen MM
    120. Oskolkov NN
    121. Owen KR
    122. Palli D
    123. Pechlivanis S
    124. Peltonen L
    125. Perry JR
    126. Platou CG
    127. Roden M
    128. Ruderfer D
    129. Rybin D
    130. van der Schouw YT
    131. Sennblad B
    132. Sigurðsson G
    133. Stančáková A
    134. Steinbach G
    135. Storm P
    136. Strauch K
    137. Stringham HM
    138. Sun Q
    139. Thorand B
    140. Tikkanen E
    141. Tonjes A
    142. Trakalo J
    143. Tremoli E
    144. Tuomi T
    145. Wennauer R
    146. Wiltshire S
    147. Wood AR
    148. Zeggini E
    149. Dunham I
    150. Birney E
    151. Pasquali L
    152. Ferrer J
    153. Loos RJ
    154. Dupuis J
    155. Florez JC
    156. Boerwinkle E
    157. Pankow JS
    158. van Duijn C
    159. Sijbrands E
    160. Meigs JB
    161. Hu FB
    162. Thorsteinsdottir U
    163. Stefansson K
    164. Lakka TA
    165. Rauramaa R
    166. Stumvoll M
    167. Pedersen NL
    168. Lind L
    169. Keinanen-Kiukaanniemi SM
    170. Korpi-Hyövälti E
    171. Saaristo TE
    172. Saltevo J
    173. Kuusisto J
    174. Laakso M
    175. Metspalu A
    176. Erbel R
    177. Jöcke KH
    178. Moebus S
    179. Ripatti S
    180. Salomaa V
    181. Ingelsson E
    182. Boehm BO
    183. Bergman RN
    184. Collins FS
    185. Mohlke KL
    186. Koistinen H
    187. Tuomilehto J
    188. Hveem K
    189. Njølstad I
    190. Deloukas P
    191. Donnelly PJ
    192. Frayling TM
    193. Hattersley AT
    194. de Faire U
    195. Hamsten A
    196. Illig T
    197. Peters A
    198. Cauchi S
    199. Sladek R
    200. Froguel P
    201. Hansen T
    202. Pedersen O
    203. Morris AD
    204. Palmer CN
    205. Kathiresan S
    206. Melander O
    207. Nilsson PM
    208. Groop LC
    209. Barroso I
    210. Langenberg C
    211. Wareham NJ
    212. O'Callaghan CA
    213. Gloyn AL
    214. Altshuler D
    215. Boehnke M
    216. Teslovich TM
    217. McCarthy MI
    218. Morris AP
    219. William Rayner N
    220. Luan Jian'an
    221. Jørgensen T
    222. Männistö S
    223. Stančáková A
    224. Jöcke KH
    225. Pedersen O
    226. Morris AP
    227. DIAbetes Genetics Replication And Meta-analysis (DIAGRAM) Consortium

    (2015) Genetic fine mapping and genomic annotation defines causal mechanisms at type 2 diabetes susceptibility loci

    Nature Genetics 47:1415–1425.

    https://doi.org/10.1038/ng.3437

    • Google Scholar
13. 1. Ghazalpour A
    2. Bennett B
    3. Petyuk VA
    4. Orozco L
    5. Hagopian R
    6. Mungrue IN
    7. Farber CR
    8. Sinsheimer J
    9. Kang HM
    10. Furlotte N
    11. Park CC
    12. Wen PZ
    13. Brewer H
    14. Weitz K
    15. Camp DG
    16. Pan C
    17. Yordanova R
    18. Neuhaus I
    19. Tilford C
    20. Siemers N
    21. Gargalovic P
    22. Eskin E
    23. Kirchgessner T
    24. Smith DJ
    25. Smith RD
    26. Lusis AJ

    (2011) Comparative analysis of proteome and transcriptome variation in mouse

    PLoS Genetics 7:e1001393.

    https://doi.org/10.1371/journal.pgen.1001393

    • Google Scholar
14. 1. Ghazalpour A
    2. Rau CD
    3. Farber CR
    4. Bennett BJ
    5. Orozco LD
    6. van Nas A
    7. Pan C
    8. Allayee H
    9. Beaven SW
    10. Civelek M
    11. Davis RC
    12. Drake TA
    13. Friedman RA
    14. Furlotte N
    15. Hui ST
    16. Jentsch JD
    17. Kostem E
    18. Kang HM
    19. Kang EY
    20. Joo JW
    21. Korshunov VA
    22. Laughlin RE
    23. Martin LJ
    24. Ohmen JD
    25. Parks BW
    26. Pellegrini M
    27. Reue K
    28. Smith DJ
    29. Tetradis S
    30. Wang J
    31. Wang Y
    32. Weiss JN
    33. Kirchgessner T
    34. Gargalovic PS
    35. Eskin E
    36. Lusis AJ
    37. LeBoeuf RC

    (2012) Hybrid mouse diversity panel: a panel of inbred mouse strains suitable for analysis of complex genetic traits

    Mammalian Genome 23:680–692.

    https://doi.org/10.1007/s00335-012-9411-5

    • Google Scholar
15. 1. Grundberg E
    2. Adoue V
    3. Kwan T
    4. Ge B
    5. Duan QL
    6. Lam KC
    7. Koka V
    8. Kindmark A
    9. Weiss ST
    10. Tantisira K
    11. Mallmin H
    12. Raby BA
    13. Nilsson O
    14. Pastinen T

    (2011) Global analysis of the impact of environmental perturbation on cis-regulation of gene expression

    PLoS Genetics 7:e1001279.

    https://doi.org/10.1371/journal.pgen.1001279

    • Google Scholar
16. 1. Grundberg E
    2. Small KS
    3. Hedman ÅK
    4. Nica AC
    5. Buil A
    6. Keildson S
    7. Bell JT
    8. Yang TP
    9. Meduri E
    10. Barrett A
    11. Nisbett J
    12. Sekowska M
    13. Wilk A
    14. Shin SY
    15. Glass D
    16. Travers M
    17. Min JL
    18. Ring S
    19. Ho K
    20. Thorleifsson G
    21. Kong A
    22. Thorsteindottir U
    23. Ainali C
    24. Dimas AS
    25. Hassanali N
    26. Ingle C
    27. Knowles D
    28. Krestyaninova M
    29. Lowe CE
    30. Di Meglio P
    31. Montgomery SB
    32. Parts L
    33. Potter S
    34. Surdulescu G
    35. Tsaprouni L
    36. Tsoka S
    37. Bataille V
    38. Durbin R
    39. Nestle FO
    40. O'Rahilly S
    41. Soranzo N
    42. Lindgren CM
    43. Zondervan KT
    44. Ahmadi KR
    45. Schadt EE
    46. Stefansson K
    47. Smith GD
    48. McCarthy MI
    49. Deloukas P
    50. Dermitzakis ET
    51. Spector TD
    52. Multiple Tissue Human Expression Resource (MuTHER) Consortium

    (2012) Mapping cis- and trans-regulatory effects across multiple tissues in twins

    Nature Genetics 44:1084–1089.

    https://doi.org/10.1038/ng.2394

    • Google Scholar
17. 1. Gu T
    2. Gatti DM
    3. Srivastava A
    4. Snyder EM
    5. Raghupathy N
    6. Simecek P
    7. Svenson KL
    8. Dotu I
    9. Chuang JH
    10. Keller MP
    11. Attie AD
    12. Braun RE
    13. Churchill GA

    (2015) Genetic architectures of quantitative variation in RNA editing pathways

    Genetics 202:787–798.

    https://doi.org/10.1534/genetics.115.179481

    • Google Scholar
18. 1. Guttman M
    2. Donaghey J
    3. Carey BW
    4. Garber M
    5. Grenier JK
    6. Munson G
    7. Young G
    8. Lucas AB
    9. Ach R
    10. Bruhn L
    11. Yang X
    12. Amit I
    13. Meissner A
    14. Regev A
    15. Rinn JL
    16. Root DE
    17. Lander ES

    (2011) lincRNAs act in the circuitry controlling pluripotency and differentiation

    Nature 477:295–300.

    https://doi.org/10.1038/nature10398

    • Google Scholar
19. 1. Hasin-Brumshtein Y
    2. Hormozdiari F
    3. Martin L
    4. van Nas A
    5. Eskin E
    6. Lusis AJ
    7. Drake TA

    (2014) Allele-specific expression and eQTL analysis in mouse adipose tissue

    BMC Genomics 15:471.

    https://doi.org/10.1186/1471-2164-15-471

    • Google Scholar
20. Data

    1. Hasin-Brumshtein Y
    2. Khan AH
    3. Hormozdiari F
    4. Pan C
    5. Parks BW
    6. Petyuk VA
    7. Piehowski PD
    8. Bruemmer A
    9. Pellegrini M
    10. Xiao X
    11. Eskin E
    12. Smith RD
    13. Lusis AJ
    14. Smith DJ

    (authors) (2016) Data from: Hypothalamic transcriptomes of 99 mouse strains reveal trans eQTL hotspots, splicing QTLs and novel non-coding genes

    Dryad Digital Repository.

    https://doi.org/10.5061/dryad.vm525
21. 1. Heinig M
    2. Petretto E
    3. Wallace C
    4. Bottolo L
    5. Rotival M
    6. Lu H
    7. Li Y
    8. Sarwar R
    9. Langley SR
    10. Bauerfeind A
    11. Hummel O
    12. Lee YA
    13. Paskas S
    14. Rintisch C
    15. Saar K
    16. Cooper J
    17. Buchan R
    18. Gray EE
    19. Cyster JG
    20. Erdmann J
    21. Hengstenberg C
    22. Maouche S
    23. Ouwehand WH
    24. Rice CM
    25. Samani NJ
    26. Schunkert H
    27. Goodall AH
    28. Schulz H
    29. Roider HG
    30. Vingron M
    31. Blankenberg S
    32. Münzel T
    33. Zeller T
    34. Szymczak S
    35. Ziegler A
    36. Tiret L
    37. Smyth DJ
    38. Pravenec M
    39. Aitman TJ
    40. Cambien F
    41. Clayton D
    42. Todd JA
    43. Hubner N
    44. Cook SA
    45. Cardiogenics Consortium

    (2010) A trans-acting locus regulates an anti-viral expression network and type 1 diabetes risk

    Nature 467:460–464.

    https://doi.org/10.1038/nature09386

    • Google Scholar
22. 1. Henry FE
    2. Sugino K
    3. Tozer A
    4. Branco T
    5. Sternson SM

    (2015) Cell type-specific transcriptomics of hypothalamic energy-sensing neuron responses to weight-loss

    eLife 4:e09800.

    https://doi.org/10.7554/eLife.09800

    • Google Scholar
23. 1. Huang DW
    2. Sherman BT
    3. Lempicki RA

    (2009a) Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists

    Nucleic Acids Research 37:1–13.

    https://doi.org/10.1093/nar/gkn923

    • Google Scholar
24. 1. Huang DW
    2. Sherman BT
    3. Lempicki RA

    (2009b) Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources

    Nature Protocols 4:44–57.

    https://doi.org/10.1038/nprot.2008.211

    • Google Scholar
25. 1. Huang GJ
    2. Shifman S
    3. Valdar W
    4. Johannesson M
    5. Yalcin B
    6. Taylor MS
    7. Taylor JM
    8. Mott R
    9. Flint J

    (2009) High resolution mapping of expression QTLs in heterogeneous stock mice in multiple tissues

    Genome Research 19:1133–1140.

    https://doi.org/10.1101/gr.088120.108

    • Google Scholar
26. 1. Huarte M
    2. Guttman M
    3. Feldser D
    4. Garber M
    5. Koziol MJ
    6. Kenzelmann-Broz D
    7. Khalil AM
    8. Zuk O
    9. Amit I
    10. Rabani M
    11. Attardi LD
    12. Regev A
    13. Lander ES
    14. Jacks T
    15. Rinn JL

    (2010) A large intergenic noncoding RNA induced by p53 mediates global gene repression in the p53 response

    Cell 142:409–419.

    https://doi.org/10.1016/j.cell.2010.06.040

    • Google Scholar
27. 1. Ingolia NT
    2. Brar GA
    3. Stern-Ginossar N
    4. Harris MS
    5. Talhouarne GJ
    6. Jackson SE
    7. Wills MR
    8. Weissman JS
    9. Gaëlle JS

    (2014) Ribosome profiling reveals pervasive translation outside of annotated protein-coding genes

    Cell Reports 8:1365–1379.

    https://doi.org/10.1016/j.celrep.2014.07.045

    • Google Scholar
28. 1. Ishii N
    2. Ozaki K
    3. Sato H
    4. Mizuno H
    5. Saito S
    6. Takahashi A
    7. Miyamoto Y
    8. Ikegawa S
    9. Kamatani N
    10. Hori M
    11. Saito S
    12. Nakamura Y
    13. Tanaka T

    (2006) Identification of a novel non-coding RNA, MIAT, that confers risk of myocardial infarction

    Journal of Human Genetics 51:1087–1099.

    https://doi.org/10.1007/s10038-006-0070-9

    • Google Scholar
29. 1. Iyer MK
    2. Niknafs YS
    3. Malik R
    4. Singhal U
    5. Sahu A
    6. Hosono Y
    7. Barrette TR
    8. Prensner JR
    9. Evans JR
    10. Zhao S
    11. Poliakov A
    12. Cao X
    13. Dhanasekaran SM
    14. Wu YM
    15. Robinson DR
    16. Beer DG
    17. Feng FY
    18. Iyer HK
    19. Chinnaiyan AM

    (2015) The landscape of long noncoding RNAs in the human transcriptome

    Nature Genetics 47:199–208.

    https://doi.org/10.1038/ng.3192

    • Google Scholar
30. 1. Ji Z
    2. Song R
    3. Regev A
    4. Struhl K

    (2015) Many lncRNAs, 5’UTRs, and pseudogenes are translated and some are likely to express functional proteins

    eLife 4:e08890.

    https://doi.org/10.7554/eLife.08890

    • Google Scholar
31. 1. Joo JW
    2. Sul JH
    3. Han B
    4. Ye C
    5. Eskin E

    (2014) Effectively identifying regulatory hotspots while capturing expression heterogeneity in gene expression studies

    Genome Biology 15:r61.

    https://doi.org/10.1186/gb-2014-15-4-r61

    • Google Scholar
32. 1. Kang HM
    2. Ye C
    3. Eskin E

    (2008) Accurate discovery of expression quantitative trait loci under confounding from spurious and genuine regulatory hotspots

    Genetics 180:1909–1925.

    https://doi.org/10.1534/genetics.108.094201

    • Google Scholar
33. 1. Keane TM
    2. Goodstadt L
    3. Danecek P
    4. White MA
    5. Wong K
    6. Yalcin B
    7. Heger A
    8. Agam A
    9. Slater G
    10. Goodson M
    11. Furlotte NA
    12. Eskin E
    13. Nellåker C
    14. Whitley H
    15. Cleak J
    16. Janowitz D
    17. Hernandez-Pliego P
    18. Edwards A
    19. Belgard TG
    20. Oliver PL
    21. McIntyre RE
    22. Bhomra A
    23. Nicod J
    24. Gan X
    25. Yuan W
    26. van der Weyden L
    27. Steward CA
    28. Bala S
    29. Stalker J
    30. Mott R
    31. Durbin R
    32. Jackson IJ
    33. Czechanski A
    34. Guerra-Assunção JA
    35. Donahue LR
    36. Reinholdt LG
    37. Payseur BA
    38. Ponting CP
    39. Birney E
    40. Flint J
    41. Adams DJ

    (2011) Mouse genomic variation and its effect on phenotypes and gene regulation

    Nature 477:289–294.

    https://doi.org/10.1038/nature10413

    • Google Scholar
34. 1. Kim MS
    2. Pinto SM
    3. Getnet D
    4. Nirujogi RS
    5. Manda SS
    6. Chaerkady R
    7. Madugundu AK
    8. Kelkar DS
    9. Isserlin R
    10. Jain S
    11. Thomas JK
    12. Muthusamy B
    13. Leal-Rojas P
    14. Kumar P
    15. Sahasrabuddhe NA
    16. Balakrishnan L
    17. Advani J
    18. George B
    19. Renuse S
    20. Selvan LD
    21. Patil AH
    22. Nanjappa V
    23. Radhakrishnan A
    24. Prasad S
    25. Subbannayya T
    26. Raju R
    27. Kumar M
    28. Sreenivasamurthy SK
    29. Marimuthu A
    30. Sathe GJ
    31. Chavan S
    32. Datta KK
    33. Subbannayya Y
    34. Sahu A
    35. Yelamanchi SD
    36. Jayaram S
    37. Rajagopalan P
    38. Sharma J
    39. Murthy KR
    40. Syed N
    41. Goel R
    42. Khan AA
    43. Ahmad S
    44. Dey G
    45. Mudgal K
    46. Chatterjee A
    47. Huang TC
    48. Zhong J
    49. Wu X
    50. Shaw PG
    51. Freed D
    52. Zahari MS
    53. Mukherjee KK
    54. Shankar S
    55. Mahadevan A
    56. Lam H
    57. Mitchell CJ
    58. Shankar SK
    59. Satishchandra P
    60. Schroeder JT
    61. Sirdeshmukh R
    62. Maitra A
    63. Leach SD
    64. Drake CG
    65. Halushka MK
    66. Prasad TS
    67. Hruban RH
    68. Kerr CL
    69. Bader GD
    70. Iacobuzio-Donahue CA
    71. Gowda H
    72. Pandey A

    (2014) A draft map of the human proteome

    Nature 509:575–581.

    https://doi.org/10.1038/nature13302

    • Google Scholar
35. 1. Kleinridders A
    2. Schenten D
    3. Könner AC
    4. Belgardt BF
    5. Mauer J
    6. Okamura T
    7. Wunderlich FT
    8. Medzhitov R
    9. Brüning JC

    (2009) MyD88 signaling in the CNS is required for development of fatty acid-induced leptin resistance and diet-induced obesity

    Cell Metabolism 10:249–259.

    https://doi.org/10.1016/j.cmet.2009.08.013

    • Google Scholar
36. 1. Ko GY
    2. Shi L
    3. Ko ML

    (2009) Circadian regulation of ion channels and their functions

    Journal of Neurochemistry 110:1150–1169.

    https://doi.org/10.1111/j.1471-4159.2009.06223.x

    • Google Scholar
37. 1. Kumar V
    2. Westra H-J
    3. Karjalainen J
    4. Zhernakova DV
    5. Esko T
    6. Hrdlickova B
    7. Almeida R
    8. Zhernakova A
    9. Reinmaa E
    10. Võsa U
    11. Hofker MH
    12. Fehrmann RSN
    13. Fu J
    14. Withoff S
    15. Metspalu A
    16. Franke L
    17. Wijmenga C

    (2013) Human disease-associated genetic variation impacts large intergenic non-coding RNA expression

    PLoS Genetics 9:e1003201.

    https://doi.org/10.1371/journal.pgen.1003201

    • Google Scholar
38. 1. Lappalainen T
    2. Sammeth M
    3. Friedländer MR
    4. 't Hoen PA
    5. Monlong J
    6. Rivas MA
    7. Gonzàlez-Porta M
    8. Kurbatova N
    9. Griebel T
    10. Ferreira PG
    11. Barann M
    12. Wieland T
    13. Greger L
    14. van Iterson M
    15. Almlöf J
    16. Ribeca P
    17. Pulyakhina I
    18. Esser D
    19. Giger T
    20. Tikhonov A
    21. Sultan M
    22. Bertier G
    23. MacArthur DG
    24. Lek M
    25. Lizano E
    26. Buermans HP
    27. Padioleau I
    28. Schwarzmayr T
    29. Karlberg O
    30. Ongen H
    31. Kilpinen H
    32. Beltran S
    33. Gut M
    34. Kahlem K
    35. Amstislavskiy V
    36. Stegle O
    37. Pirinen M
    38. Montgomery SB
    39. Donnelly P
    40. McCarthy MI
    41. Flicek P
    42. Strom TM
    43. Lehrach H
    44. Schreiber S
    45. Sudbrak R
    46. Carracedo A
    47. Antonarakis SE
    48. Häsler R
    49. Syvänen AC
    50. van Ommen GJ
    51. Brazma A
    52. Meitinger T
    53. Rosenstiel P
    54. Guigó R
    55. Gut IG
    56. Estivill X
    57. Dermitzakis ET
    58. Monlong J
    59. Rivas MA
    60. Geuvadis Consortium

    (2013) Transcriptome and genome sequencing uncovers functional variation in humans

    Nature 501:506–511.

    https://doi.org/10.1038/nature12531

    • Google Scholar
39. 1. Lee JH
    2. Ang JK
    3. Xiao X

    (2013) Analysis and design of RNA sequencing experiments for identifying RNA editing and other single-nucleotide variants

    RNA 19:725–732.

    https://doi.org/10.1261/rna.037903.112

    • Google Scholar
40. 1. Liao J
    2. He Q
    3. Li M
    4. Chen Y
    5. Liu Y
    6. Wang J
    7. Li M
    8. Chen Y
    9. Liu Y

    (2016) LncRNA MIAT: Myocardial infarction associated and more

    Gene 578:158–161.

    https://doi.org/10.1016/j.gene.2015.12.032

    • Google Scholar
41. 1. Lusis AJ

    (2000) Atherosclerosis

    Nature 407:233–241.

    https://doi.org/10.1038/35025203

    • Google Scholar
42. 1. Lusis AJ

    (2012) Genetics of atherosclerosis

    Trends in Genetics 28:267–275.

    https://doi.org/10.1016/j.tig.2012.03.001

    • Google Scholar
43. 1. Majewski J
    2. Pastinen T

    (2011) The study of eQTL variations by RNA-seq: from SNPs to phenotypes

    Trends in Genetics 27:72–79.

    https://doi.org/10.1016/j.tig.2010.10.006

    • Google Scholar
44. 1. McClurg P
    2. Janes J
    3. Wu C
    4. Delano DL
    5. Walker JR
    6. Batalov S
    7. Takahashi JS
    8. Shimomura K
    9. Kohsaka A
    10. Bass J
    11. Wiltshire T
    12. Su AI

    (2007) Genomewide association analysis in diverse inbred mice: power and population structure

    Genetics 176:675–683.

    https://doi.org/10.1534/genetics.106.066241

    • Google Scholar
45. 1. Min JL
    2. Nicholson G
    3. Halgrimsdottir I
    4. Almstrup K
    5. Petri A
    6. Barrett A
    7. Travers M
    8. Rayner NW
    9. Mägi R
    10. Pettersson FH
    11. Broxholme J
    12. Neville MJ
    13. Wills QF
    14. Cheeseman J
    15. Allen M
    16. Holmes CC
    17. Spector TD
    18. Fleckner J
    19. McCarthy MI
    20. Karpe F
    21. Lindgren CM
    22. Zondervan KT
    23. GIANT Consortium
    24. MolPAGE Consortium

    (2012) Coexpression network analysis in abdominal and gluteal adipose tissue reveals regulatory genetic loci for metabolic syndrome and related phenotypes

    PLoS Genetics 8:e1002505.

    https://doi.org/10.1371/journal.pgen.1002505

    • Google Scholar
46. 1. Morari J
    2. Anhe GF
    3. Nascimento LF
    4. de Moura RF
    5. Razolli D
    6. Solon C
    7. Guadagnini D
    8. Souza G
    9. Mattos AH
    10. Tobar N
    11. Ramos CD
    12. Pascoal VD
    13. Saad MJ
    14. Lopes-Cendes I
    15. Moraes JC
    16. Velloso LA

    (2014) Fractalkine (CX3CL1) is involved in the early activation of hypothalamic inflammation in experimental obesity

    Diabetes 63:3770–3784.

    https://doi.org/10.2337/db13-1495

    • Google Scholar
47. 1. Mutz KO
    2. Heilkenbrinker A
    3. Lönne M
    4. Walter JG
    5. Stahl F

    (2013) Transcriptome analysis using next-generation sequencing

    Current Opinion in Biotechnology 24:22–30.

    https://doi.org/10.1016/j.copbio.2012.09.004

    • Google Scholar
48. 1. Nakagawa S
    2. Ip JY
    3. Shioi G
    4. Tripathi V
    5. Zong X
    6. Hirose T
    7. Prasanth KV

    (2012) Malat1 is not an essential component of nuclear speckles in mice

    RNA 18:1487–1499.

    https://doi.org/10.1261/rna.033217.112

    • Google Scholar
49. 1. Orozco LD
    2. Bennett BJ
    3. Farber CR
    4. Ghazalpour A
    5. Pan C
    6. Che N
    7. Wen P
    8. Qi HX
    9. Mutukulu A
    10. Siemers N
    11. Neuhaus I
    12. Yordanova R
    13. Gargalovic P
    14. Pellegrini M
    15. Kirchgessner T
    16. Lusis AJ

    (2012) Unraveling inflammatory responses using systems genetics and gene-environment interactions in macrophages

    Cell 151:658–670.

    https://doi.org/10.1016/j.cell.2012.08.043

    • Google Scholar
50. 1. Orozco LD
    2. Morselli M
    3. Rubbi L
    4. Guo W
    5. Go J
    6. Shi H
    7. Lopez D
    8. Furlotte NA
    9. Bennett BJ
    10. Farber CR
    11. Ghazalpour A
    12. Zhang MQ
    13. Bahous R
    14. Rozen R
    15. Lusis AJ
    16. Pellegrini M

    (2015) Epigenome-wide association of liver methylation patterns and complex metabolic traits in mice

    Cell Metabolism 21:905–917.

    https://doi.org/10.1016/j.cmet.2015.04.025

    • Google Scholar
51. 1. Parks BW
    2. Nam E
    3. Org E
    4. Kostem E
    5. Norheim F
    6. Hui ST
    7. Pan C
    8. Civelek M
    9. Rau CD
    10. Bennett BJ
    11. Mehrabian M
    12. Ursell LK
    13. He A
    14. Castellani LW
    15. Zinker B
    16. Kirby M
    17. Drake TA
    18. Drevon CA
    19. Knight R
    20. Gargalovic P
    21. Kirchgessner T
    22. Eskin E
    23. Lusis AJ
    24. Simon T
    25. Hui CP

    (2013) Genetic control of obesity and gut microbiota composition in response to high-fat, high-sucrose diet in mice

    Cell Metabolism 17:141–152.

    https://doi.org/10.1016/j.cmet.2012.12.007

    • Google Scholar
52. 1. Pickrell JK
    2. Marioni JC
    3. Pai AA
    4. Degner JF
    5. Engelhardt BE
    6. Nkadori E
    7. Veyrieras JB
    8. Stephens M
    9. Gilad Y
    10. Pritchard JK
    11. Joseph K

    (2010) Understanding mechanisms underlying human gene expression variation with RNA sequencing

    Nature 464:768–772.

    https://doi.org/10.1038/nature08872

    • Google Scholar
53. 1. Quek XC
    2. Thomson DW
    3. Maag JL
    4. Bartonicek N
    5. Signal B
    6. Clark MB
    7. Gloss BS
    8. Dinger ME

    (2015) lncRNAdb v2.0: expanding the reference database for functional long noncoding RNAs

    Nucleic Acids Research 43:D168–173.

    https://doi.org/10.1093/nar/gku988

    • Google Scholar
54. 1. Ramos AD
    2. Attenello FJ
    3. Lim DA

    (2016) Uncovering the roles of long noncoding RNAs in neural development and glioma progression

    Neuroscience Letters 625:70–79.

    https://doi.org/10.1016/j.neulet.2015.12.025

    • Google Scholar
55. 1. Ruiz-Orera J
    2. Messeguer X
    3. Subirana JA
    4. Alba MM
    5. Mar Alba M

    (2014) Long non-coding RNAs as a source of new peptides

    eLife 3:e03523.

    https://doi.org/10.7554/eLife.03523

    • Google Scholar
56. 1. Shah R
    2. O’Neill SM
    3. Hinkle C
    4. Caughey J
    5. Stephan S
    6. Lynch E
    7. Bermingham K
    8. Lynch G
    9. Ahima RS
    10. Reilly MP

    (2015) Metabolic effects of CX3CR1 deficiency in diet-induced obese mice

    Plos One 10:e0138317.

    https://doi.org/10.1371/journal.pone.0138317

    • Google Scholar
57. 1. Sinha A
    2. Khadilkar RJ
    3. S VK
    4. Roychowdhury Sinha A
    5. Inamdar MS
    6. Vinay KS

    (2013) Conserved regulation of the Jak/STAT pathway by the endosomal protein asrij maintains stem cell potency

    Cell Reports 4:649–658.

    https://doi.org/10.1016/j.celrep.2013.07.029

    • Google Scholar
58. 1. Small KS
    2. Hedman AK
    3. Grundberg E
    4. Nica AC
    5. Thorleifsson G
    6. Kong A
    7. Thorsteindottir U
    8. Shin SY
    9. Richards HB
    10. Soranzo N
    11. Ahmadi KR
    12. Lindgren CM
    13. Stefansson K
    14. Dermitzakis ET
    15. Deloukas P
    16. Spector TD
    17. McCarthy MI
    18. Spector TD
    19. GIANT Consortium
    20. MAGIC Investigators
    21. DIAGRAM Consortium
    22. MuTHER Consortium

    (2011) Identification of an imprinted master trans regulator at the KLF14 locus related to multiple metabolic phenotypes

    Nature Genetics 43:561–564.

    https://doi.org/10.1038/ng.833

    • Google Scholar
59. 1. St Laurent G
    2. Wahlestedt C
    3. Kapranov P
    4. Laurent S
    5. Georges CW

    (2015) The Landscape of long noncoding RNA classification

    Trends in Genetics 31:239–251.

    https://doi.org/10.1016/j.tig.2015.03.007

    • Google Scholar
60. 1. Suhre K
    2. Shin SY
    3. Petersen AK
    4. Mohney RP
    5. Meredith D
    6. Wägele B
    7. Altmaier E
    8. Deloukas P
    9. Erdmann J
    10. Grundberg E
    11. Hammond CJ
    12. de Angelis MH
    13. Kastenmüller G
    14. Köttgen A
    15. Kronenberg F
    16. Mangino M
    17. Meisinger C
    18. Meitinger T
    19. Mewes HW
    20. Milburn MV
    21. Prehn C
    22. Raffler J
    23. Ried JS
    24. Römisch-Margl W
    25. Samani NJ
    26. Small KS
    27. Wichmann HE
    28. Zhai G
    29. Illig T
    30. Spector TD
    31. Adamski J
    32. Soranzo N
    33. Gieger C
    34. CARDIoGRAM

    (2011) Human metabolic individuality in biomedical and pharmaceutical research

    Nature 477:54–60.

    https://doi.org/10.1038/nature10354

    • Google Scholar
61. 1. Sun L
    2. Goff LA
    3. Trapnell C
    4. Alexander R
    5. Lo KA
    6. Hacisuleyman E
    7. Sauvageau M
    8. Tazon-Vega B
    9. Kelley DR
    10. Hendrickson DG
    11. Yuan B
    12. Kellis M
    13. Lodish HF
    14. Rinn JL

    (2013) Long noncoding RNAs regulate adipogenesis

    PNAS 110:3387–3392.

    https://doi.org/10.1073/pnas.1222643110

    • Google Scholar
62. 1. Tay AP
    2. Pang CN
    3. Twine NA
    4. Hart-Smith G
    5. Harkness L
    6. Kassem M
    7. Wilkins MR

    (2015) Proteomic validation of transcript isoforms, including those assembled from RNA-seq data

    Journal of Proteome Research 14:3541–3554.

    https://doi.org/10.1021/pr5011394

    • Google Scholar
63. 1. Teslovich TM
    2. Musunuru K
    3. Smith AV
    4. Edmondson AC
    5. Stylianou IM
    6. Koseki M
    7. Pirruccello JP
    8. Ripatti S
    9. Chasman DI
    10. Willer CJ
    11. Johansen CT
    12. Fouchier SW
    13. Isaacs A
    14. Peloso GM
    15. Barbalic M
    16. Ricketts SL
    17. Bis JC
    18. Aulchenko YS
    19. Thorleifsson G
    20. Feitosa MF
    21. Chambers J
    22. Orho-Melander M
    23. Melander O
    24. Johnson T
    25. Li X
    26. Guo X
    27. Li M
    28. Shin Cho Y
    29. Jin Go M
    30. Jin Kim Y
    31. Lee JY
    32. Park T
    33. Kim K
    34. Sim X
    35. Twee-Hee Ong R
    36. Croteau-Chonka DC
    37. Lange LA
    38. Smith JD
    39. Song K
    40. Hua Zhao J
    41. Yuan X
    42. Luan J
    43. Lamina C
    44. Ziegler A
    45. Zhang W
    46. Zee RY
    47. Wright AF
    48. Witteman JC
    49. Wilson JF
    50. Willemsen G
    51. Wichmann HE
    52. Whitfield JB
    53. Waterworth DM
    54. Wareham NJ
    55. Waeber G
    56. Vollenweider P
    57. Voight BF
    58. Vitart V
    59. Uitterlinden AG
    60. Uda M
    61. Tuomilehto J
    62. Thompson JR
    63. Tanaka T
    64. Surakka I
    65. Stringham HM
    66. Spector TD
    67. Soranzo N
    68. Smit JH
    69. Sinisalo J
    70. Silander K
    71. Sijbrands EJ
    72. Scuteri A
    73. Scott J
    74. Schlessinger D
    75. Sanna S
    76. Salomaa V
    77. Saharinen J
    78. Sabatti C
    79. Ruokonen A
    80. Rudan I
    81. Rose LM
    82. Roberts R
    83. Rieder M
    84. Psaty BM
    85. Pramstaller PP
    86. Pichler I
    87. Perola M
    88. Penninx BW
    89. Pedersen NL
    90. Pattaro C
    91. Parker AN
    92. Pare G
    93. Oostra BA
    94. O'Donnell CJ
    95. Nieminen MS
    96. Nickerson DA
    97. Montgomery GW
    98. Meitinger T
    99. McPherson R
    100. McCarthy MI
    101. McArdle W
    102. Masson D
    103. Martin NG
    104. Marroni F
    105. Mangino M
    106. Magnusson PK
    107. Lucas G
    108. Luben R
    109. Loos RJ
    110. Lokki ML
    111. Lettre G
    112. Langenberg C
    113. Launer LJ
    114. Lakatta EG
    115. Laaksonen R
    116. Kyvik KO
    117. Kronenberg F
    118. König IR
    119. Khaw KT
    120. Kaprio J
    121. Kaplan LM
    122. Johansson A
    123. Jarvelin MR
    124. Janssens AC
    125. Ingelsson E
    126. Igl W
    127. Kees Hovingh G
    128. Hottenga JJ
    129. Hofman A
    130. Hicks AA
    131. Hengstenberg C
    132. Heid IM
    133. Hayward C
    134. Havulinna AS
    135. Hastie ND
    136. Harris TB
    137. Haritunians T
    138. Hall AS
    139. Gyllensten U
    140. Guiducci C
    141. Groop LC
    142. Gonzalez E
    143. Gieger C
    144. Freimer NB
    145. Ferrucci L
    146. Erdmann J
    147. Elliott P
    148. Ejebe KG
    149. Döring A
    150. Dominiczak AF
    151. Demissie S
    152. Deloukas P
    153. de Geus EJ
    154. de Faire U
    155. Crawford G
    156. Collins FS
    157. Chen YD
    158. Caulfield MJ
    159. Campbell H
    160. Burtt NP
    161. Bonnycastle LL
    162. Boomsma DI
    163. Boekholdt SM
    164. Bergman RN
    165. Barroso I
    166. Bandinelli S
    167. Ballantyne CM
    168. Assimes TL
    169. Quertermous T
    170. Altshuler D
    171. Seielstad M
    172. Wong TY
    173. Tai ES
    174. Feranil AB
    175. Kuzawa CW
    176. Adair LS
    177. Taylor HA
    178. Borecki IB
    179. Gabriel SB
    180. Wilson JG
    181. Holm H
    182. Thorsteinsdottir U
    183. Gudnason V
    184. Krauss RM
    185. Mohlke KL
    186. Ordovas JM
    187. Munroe PB
    188. Kooner JS
    189. Tall AR
    190. Hegele RA
    191. Kastelein JJ
    192. Schadt EE
    193. Rotter JI
    194. Boerwinkle E
    195. Strachan DP
    196. Mooser V
    197. Stefansson K
    198. Reilly MP
    199. Samani NJ
    200. Schunkert H
    201. Cupples LA
    202. Sandhu MS
    203. Ridker PM
    204. Rader DJ
    205. van Duijn CM
    206. Peltonen L
    207. Abecasis GR
    208. Boehnke M
    209. Kathiresan S

    (2010) Biological, clinical and population relevance of 95 loci for blood lipids

    Nature 466:707–713.

    https://doi.org/10.1038/nature09270

    • Google Scholar
64. 1. Tian J
    2. Keller MP
    3. Broman AT
    4. Kendziorski C
    5. Yandell BS
    6. Attie AD
    7. Broman KW

    (2016) The dissection of expression quantitativetrait locus hotspots

    Genetics 202:1563–1574.

    https://doi.org/10.1534/genetics.115.183624

    • Google Scholar
65. 1. Tian J
    2. Keller MP
    3. Oler AT
    4. Rabaglia ME
    5. Schueler KL
    6. Stapleton DS
    7. Broman AT
    8. Zhao W
    9. Kendziorski C
    10. Yandell BS
    11. Hagenbuch B
    12. Broman KW
    13. Attie AD

    (2015) Identification of the bile acid transporter Slco1a6 as a candidate Gene that broadly affects gene expression in mouse pancreatic islets

    Genetics 201:1253–1262.

    https://doi.org/10.1534/genetics.115.179432

    • Google Scholar
66. 1. Tsuneki H
    2. Sasaoka T
    3. Sakurai T

    (2016) Sleep control, GPCRs, and glucose metabolism

    Trends in Endocrinology & Metabolism 27:633–642.

    https://doi.org/10.1016/j.tem.2016.06.011

    • Google Scholar
67. 1. van Nas A
    2. Ingram-Drake L
    3. Sinsheimer JS
    4. Wang SS
    5. Schadt EE
    6. Drake T
    7. Lusis AJ

    (2010) Expression quantitative trait loci: replication, tissue- and sex-specificity in mice

    Genetics 185:1059–1068.

    https://doi.org/10.1534/genetics.110.116087

    • Google Scholar
68. 1. Zhang B
    2. Arun G
    3. Mao YS
    4. Lazar Z
    5. Hung G
    6. Bhattacharjee G
    7. Xiao X
    8. Booth CJ
    9. Wu J
    10. Zhang C
    11. Spector DL

    (2012) The lncRNA Malat1 is dispensable for mouse development but its transcription plays a cis-regulatory role in the adult

    Cell Reports 2:111–123.

    https://doi.org/10.1016/j.celrep.2012.06.003

    • Google Scholar
69. 1. Zhang X
    2. Rice K
    3. Wang Y
    4. Chen W
    5. Zhong Y
    6. Nakayama Y
    7. Zhou Y
    8. Klibanski A

    (2010) Maternally expressed gene 3 (MEG3) noncoding ribonucleic acid: isoform structure, expression, and functions

    Endocrinology 151:939–947.

    https://doi.org/10.1210/en.2009-0657

    • Google Scholar
70. 1. Zhu X
    2. Wu YB
    3. Zhou J
    4. Kang DM

    (2016) Upregulation of lncRNA MEG3 promotes hepatic insulin resistance via increasing FoxO1 expression

    Biochemical and Biophysical Research Communications 469:319–325.

    https://doi.org/10.1016/j.bbrc.2015.11.048

    • Google Scholar

Author details

1. Yehudit Hasin-Brumshtein

   1. Department of Human Genetics, University of California, Los Angeles, Los Angeles, United States
   2. David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, United States
   3. Department of Microbiology, University of California, Los Angeles, Los Angeles, United states
   4. Department of Immunology and Molecular Genetics, University of California, Los Angeles, Los Angeles, United States

   Contribution

   YH-B, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   yehudit.hasin@gmail.com

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0001-7528-603X

2. Arshad H Khan

   Department of Molecular and Medical Pharmacology, University of California, Los Angeles, Los Angeles, United States

   Contribution

   AHK, Conception and design, Acquisition of data, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

3. Farhad Hormozdiari

   Department of Computer Science, University of California, Los Angeles, Los Angeles, United States

   Contribution

   FH, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

4. Calvin Pan

   1. Department of Human Genetics, University of California, Los Angeles, Los Angeles, United States
   2. David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, United States

   Contribution

   CP, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

5. Brian W Parks

   1. Department of Human Genetics, University of California, Los Angeles, Los Angeles, United States
   2. David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, United States
   3. Department of Microbiology, University of California, Los Angeles, Los Angeles, United states
   4. Department of Immunology and Molecular Genetics, University of California, Los Angeles, Los Angeles, United States

   Contribution

   BWP, Acquisition of data

   Competing interests

   The authors declare that no competing interests exist.

6. Vladislav A Petyuk

   Biological Sciences Division, Pacific Northwest National Laboratory, Richland, United States

   Contribution

   VAP, Acquisition of data, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

7. Paul D Piehowski

   Biological Sciences Division, Pacific Northwest National Laboratory, Richland, United States

   Contribution

   PDP, Acquisition of data

   Competing interests

   The authors declare that no competing interests exist.

8. Anneke Brümmer

   Department of Integrative Biology and Physiology, University of California, Los Angeles, Los Angeles, United States

   Contribution

   AB, Analysis and interpretation of data

   Competing interests

   The authors declare that no competing interests exist.

9. Matteo Pellegrini

   Department of Molecular, Cell, and Developmental Biology, University of California, Los Angeles, Los Angeles, United States

   Contribution

   MP, Conception and design

   Competing interests

   The authors declare that no competing interests exist.

10. Xinshu Xiao

    Department of Integrative Biology and Physiology, University of California, Los Angeles, Los Angeles, United States

    Contribution

    XX, Analysis and interpretation of data

    Competing interests

    The authors declare that no competing interests exist.

11. Eleazar Eskin

    Department of Computer Science, University of California, Los Angeles, Los Angeles, United States

    Contribution

    EE, Conception and design, Analysis and interpretation of data

    Competing interests

    The authors declare that no competing interests exist.

12. Richard D Smith

    Biological Sciences Division, Pacific Northwest National Laboratory, Richland, United States

    Contribution

    RDS, Conception and design, Acquisition of data

    Competing interests

    The authors declare that no competing interests exist.

13. Aldons J Lusis

    1. Department of Human Genetics, University of California, Los Angeles, Los Angeles, United States
    2. David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, United States
    3. Department of Microbiology, University of California, Los Angeles, Los Angeles, United states
    4. Department of Immunology and Molecular Genetics, University of California, Los Angeles, Los Angeles, United States

    Contribution

    AJL, Conception and design, Analysis and interpretation of data, Drafting or revising the article, Contributed unpublished essential data or reagents

    Competing interests

    The authors declare that no competing interests exist.

14. Desmond J Smith

    Department of Molecular and Medical Pharmacology, University of California, Los Angeles, Los Angeles, United States

    Contribution

    DJS, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article, Contributed unpublished essential data or reagents

    Competing interests

    The authors declare that no competing interests exist.

• 2,983

  views

• 509

  downloads

• 35

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.