| Download datasets and supplementary data files |
|---|
Summary
These hypothalamic gene expression data were generated by Khyobeni Mozhui, Lu Lu, and Robert W. Williams and colleagues with funding support from NIAAA. The data set includes samples from 50 strains, including 46 BXDs, both parental strains, and reciprocal F1 hybrids. Expression data were generated using the Affymetrix Mouse Gene 1.0 ST exon-style microarray (multiple probes in all known exons) by Lorne Rose in the UTHSC Molecular Resources Center (MRC), Memphis TN. The table below provides a summary of cases, sex, and age. Hypothalamic tissue was dissected by K. Mozhui (description to follow) with special attention to time of day (every sample has time stamp). RNA was extracted by K. Mozhui. All other processing steps by the UTHSC MRC by L. Rose. Data were processed by Arthur Centeno.
Data released initially Nov 25, 2010, updated March 7, 2011 by A. Centeno and K. Mozhui to add two additional arrays. Data appear to be error-free in terms of sex and strain assignments shown in the table below.
Dissection protocol:
- Animals were sacrificed by quick cervical dislocation and brains were removed and stored in RNAlater (www.ambion.com) for 2 to 3 days
- Brain was placed with ventral side up and a partial cut was made with a blade at -2.5 from Bregma (just a little rostral from the pontine fibres when viewed from the ventral side)
- The brain was then place in a coronal matrix and a 2 mm section was made rostral to the first cut
- The 2mm Section was placed on a clean glass slide and hypothalamus was sliced out and placed in a tube on dry ice.
- To dissect out the BLA, the temporal lobes were detached by placing a scalple in the lateral ventricles and teasing it apart. The cortical amygdala was removed and the BLA was then sliced out and placed in a tube on dry ice.
- Tissues from two mice of the same strain and sex were pooled. The only exceptions to this are the BLA samples for strains BXD5, BXD13, BXD16, BXD19, BXD25, BXD38, BXD51, and BXD61 (tissue from only one animal).
The hypothalamus contains nuclei and cell populations that are critical in reproduction and that differ significantly between the sexes in structure and function. To examine the molecular and genetic basis for these differences, we quantified gene expression in the hypothalamus of 39 pairs of adult male and female mice belonging to the BXD strains. This experimental design enabled us to define hypothalamic gene coexpression networks and provided robust estimates of absolute expression differences. As expected, sex has the strongest effect on the expression of genes on the X and Y chromosomes (e.g., Uty, Xist, Kdm6a).Transcripts associated with the endocrine system and neuropeptide signaling also differ significantly. Sex-differentiated transcripts often have well delimited expression within specific hypothalamic nuclei that have roles in reproduction. For instance, the estro-gen receptor (Esr1) and neurokinin B (Tac2) genes have intense expression in the medial preoptic and arcuate nuclei and comparatively high expression in females. Despite the strong effect of sex on single transcripts, the global pattern of covariance among transcripts is well preserved, and consequently, males and females have well matched coexpression modules. However, there are sex-specific hub genes in functionally equivalent modules. For example, only in males is the Y-linked gene, Uty, a highly connected transcript in a network that regulates chromatin modification and gene transcription. In females, the X chromo-some paralog, Kdm6a, takes the place of Uty in the same network. We also find significant effect of sex on genetic regulation and the same network in males and females can be associated with markedly different regulatory loci. With the exception of a few sex-specific modules, our analysis reveals a system in which sets of functionally related transcripts are organized into stable sex-independent networks that are controlled at a higher level by sex-specific modulators.
Experiment design
Hypothalamus was dissected from adult male and female mice and process for expression analysis.
RNA isolation
Total RNA was purified using the RNAeasy micro kit on the QIAcube system (www.qiagen.com). RNA purity and concentration was checked using 260/280 nm absorbance ratio and RNA integrity was analyzed using the Agilent Bioanalyzer 2100 (Agilent Technologies).
This preliminary data set is associated with 430 eQTLs with LOD scores above 10. Peak LRS is 146 for Trait ID 10513604 (Hdhd3).
About cases
The BXD recombinant inbred strains are derived from crossing the C57BL/6J (B6) and DBA/2J (D2) parental strains and inbreeding for over 20 generations (Taylor et al., 1999; Peirce et al., 2004). All mice used in this study were born and housed at the University of Tennessee Health Science Center. Mice were kept at an average of 3–4 per cage in a temperature-controlled vivarium (22 deg C) and maintained at a 12 h light/dark cycle. All animal protocols were approved by the Animal Care and Use Committee. We studied a total of 50 BXD strains, but only acquired matched male–female data pairs for 39 strains (35 BXD strains, parental B6 and D2, and two reciprocal F1 hybrids, B6D2F1 and D2B6F1). The average age of mice was 78 days. We provide more detail on the experimen-tal design and precise age and time of sacrifice of all cases at stored in 4 deg C for 2–3 days. To dissect the hypothalamus, the brain was placed with the ventral side facing up in a coronal "brain cutting" matrix. Using the medial mammillary body as landmark, a vertical cut was made right along the posterior boundary of the hypothala-mus. A second vertical cut was made 2 mm from the first cut. This edge lies slightly caudal to the optic chiasm and cuts through the retrochiasmatic nucleus. The hypothalamus was then sliced out from this 2 mm thick section. Each of the 39 mouse strains is represented by male and female samples (total of 78 microarray samples). Each individual sample consisted of tissue pooled from two mice of the same strain and sex that are usually littermates. The total number of mice used was 78 females and 78 males. RNA was purified using the RNAeasy micro kit on the QIAcube system (Qiagen)2. RNA purity and concen-tration was checked using 260/280 nm absorbance ratio and RNA integrity was analyzed using the Agilent Bioanalyzer 2100 (Agilent Technologies).
|
About platform
[MoGene-1_0-st] Affymetrix Mouse Gene 1.0 ST Array [transcript (gene) version] (GPL6246)
About data processing
Total RNA was processed using standard protocols and hybridized to the Affymetrix Mouse Gene 1.0 ST arrays4. This array contains 34,700 probe sets that target ∼29,000 well-defined transcripts (RefSeq mRNA isoforms). A single probe set is a collection of about 27 probes that target known exons within a single gene. The multiple probes design provides a more comprehensive cov-erage of transcripts from a single gene. Male and female samples were interleaved and processed together to avoid batch confounds. Details on the strain, age, and sex of each sample can be obtained from the information available for the INIA Hypothalamus Affy MoGene 1.0 ST (Nov10) data set on www.genenetwork.org. Probe set level intensity values were extracted from the CEL files using the Affymetrix GeneChip Operating Software. Data nor-malization was performed using the R package “Affy” available from www.Bioconductor.org. The Robust Multichip Averaging protocol was used to process the expression values. The array data was then log transformed and rescaled using a z-scoring procedure to set the mean of each sample at eight expression units with a SD of 2 units. The entire data set can be down-loaded from www.genenetwork.org using the accession number GN281 (http://www.genenetwork.org/webqtl/main.py?FormID = sharinginfo&GN_AccessionId = 281) and also from the NIH Gene Expression Omnibus5 using the GEO accession number GSE36674. For this study we used a subset of cases that were repre-sented by both male and female samples – 78 sex-balanced arrays. Statistical power provided by this sample size (N = 39 strains) was estimated using the R function power.t.test6 with the SD set at 0.17, which is the average value. Only transcripts with above aver-age expression (minimum expression value above 8.5 on a log2 scale) were included for further analysis (17,192 probe sets). We used a two-tailed paired t -test to identify transcripts with signifi-cant expression difference between males and females. We applied the Benjamini and Hochberg false discovery rate (FDR) method and selected differentially expressed transcripts using a 10% FDR criterion.
Contributors
Mozhui K, Lu L, Armstrong WE, Williams RW (2012) Sex-specific modulation of gene expression networks in murine hypothalamus. Front Neurosci. 2012 May 11;6:63. doi: 10.3389/fnins.2012.00063. eCollection 2012.
Abstract
The hypothalamus contains nuclei and cell populations that are critical in reproduction and that differ significantly between the sexes in structure and function. To examine the molecular and genetic basis for these differences, we quantified gene expression in the hypothalamus of 39 pairs of adult male and female mice belonging to the BXD strains. This experimental design enabled us to define hypothalamic gene coexpression networks and provided robust estimates of absolute expression differences. As expected, sex has the strongest effect on the expression of genes on the X and Y chromosomes (e.g., Uty, Xist, Kdm6a). Transcripts associated with the endocrine system and neuropeptide signaling also differ significantly. Sex-differentiated transcripts often have well delimited expression within specific hypothalamic nuclei that have roles in reproduction. For instance, the estrogen receptor (Esr1) and neurokinin B (Tac2) genes have intense expression in the medial preoptic and arcuate nuclei and comparatively high expression in females. Despite the strong effect of sex on single transcripts, the global pattern of covariance among transcripts is well preserved, and consequently, males and females have well matched coexpression modules. However, there are sex-specific hub genes in functionally equivalent modules. For example, only in males is the Y-linked gene, Uty, a highly connected transcript in a network that regulates chromatin modification and gene transcription. In females, the X chromosome paralog, Kdm6a, takes the place of Uty in the same network. We also find significant effect of sex on genetic regulation and the same network in males and females can be associated with markedly different regulatory loci. With the exception of a few sex-specific modules, our analysis reveals a system in which sets of functionally related transcripts are organized into stable sex-independent networks that are controlled at a higher level by sex-specific modulators.
eQTL; hypothalamus; sex-specificity; weighted gene coexpression networks
- PMID: 22593731 PMCID: PMC3350311 DOI: 10.3389/fnins.2012.00063
Citation
Mozhui K, Lu L, Armstrong WE, Williams RW. Sex-specific modulation of gene expression networks in murine hypothalamus. Front Neurosci 2012;6:63. PMID:22593731
Acknowledgment
This work was supported by Integrative Neuroscience Initiative on Alcoholism grants U01AA13499, U01AA017590, U01AA0016662. The authors are also grateful to Arthur Centeno and Lorne Rose, and the Molecular Resource Center at UTHSC.