Transcription profiling and QTL mapping of 60 mice from an F2 sample segregating for diabetes to generate expression trait correlations and expression quantitative trait locus mapping

Coordinated regulation of gene expression levels across a series of experimental conditions provides valuable information about the functions of correlated transcripts. To map gene regulatory pathways, we used microarray-derived gene expression measurements in 60 individuals of an F2 sample segregating for diabetes. We performed correlation analysis among ~40,000 expression traits. By combining correlation among expression traits and linkage mapping information, we were able to identify regulatory networks, make functional predictions to uncharacterized genes, and characterize novel members of known pathways. Using 36 seed traits, we found evidence of coordinate regulation of 160 G-protein coupled receptor (GPCR) pathway expression traits. Of the 160 traits, 50 had their major LOD peak within 8 cM of a locus on chromosome 2, and 81 others had a secondary peak in this region. A previously uncharacterized Riken cDNA clone, which showed strong correlation with stearoyl CoA desaturase 1 expression, was experimentally validated to be responsive to conditions that regulate lipid metabolism. Using linkage mapping, we identified multiple genes whose expression is under the control of transcription regulatory loci. Trait-correlation combined with linkage mapping can reveal regulatory networks that would otherwise be missed if we only studied mRNA traits with statistically significant linkages in this small cross. The combined analysis is more sensitive compared with linkage mapping only. References: ; Kendziorski C., M. Chen, M. Yuan, H. Lan, and A.D. Attie. Statistical Methods for Expression Quantitative Trait Loci (eQTL) Mapping. Biometrics, to appear, 2005. Lan H, Chen M, Flowers JB, Yandell BS, Stapleton DS, et al. (2006) Combined Expression Trait Correlations and Expression Quantitative Trait Locus Mapping. PLoS Genet 2(1): e6. Experiment Overall Design: The F2-ob/ob mice were chosen from a mapping panel that we created to map diabetes related physiological phenotypes (Stoehr et al. 2000). About 110 of these F2-ob/ob mice were also used to map mRNA abundance traits derived by quantitative real-time RT-PCR (Lan et al. 2003). The sixty F2-ob/ob mice that were used to generate microarray-derived mRNA abundance traits were selected from the 110 mice based on a selective phenotyping algorithm (Jin et al. 2004). The F2-ob/ob mice were housed at weaning at the University of Wisconsin-Madison animal care facility on a 12-h light/dark cycle. Mice were provided Purina Formulab Chow 5008 (6.5% fat) and acidified water ad libitum. Mice were killed at 14 weeks of age by CO2 asphyxiation after a 4-hour fast. The livers, along with other tissues, were immediately foil wrapped and frozen in liquid nitrogen, and subsequently transferred to -80 °C freezers for storage. Liver samples were taken from 29 male and 31 females. Total RNA was isolated with RNAzol Reagent (Tel-Test, Inc.) using a modification of the single-step acid guanidinium isothiocyanate phenol-chloroform extraction method according to the manufacturer's protocol. The extracted RNA was purified using RNeasy (Qiagen, Inc.). RNA samples were evaluated by UV spectroscopy for concentration. RNA quality was monitored by visualization on an ethidium bromide-stained denaturing formaldehyde agarose gel. RNA samples were converted to cDNA, and then biotin-labeled cRNA according to Affymetrix Expression Analysis Technical Manual. The labeled samples were hybridized to the M430A, and subsequently the M430B array. The hybridization, washing and scanning steps were carried out by Hong Lan using the Affymetrix core facility at the Gene Expression Center of University of Wisconsin-Madison.在一系列实验条件下对基因表达水平开展协同调控分析，可获取关于关联转录本功能的宝贵信息。为绘制基因调控通路图谱，我们针对60名糖尿病分离F2群体个体，采用了微阵列（microarray）检测得到的基因表达量数据。我们对约40000个表达性状进行了相关性分析。通过整合表达性状间的相关性与连锁定位信息，我们得以识别调控网络、对未注释基因开展功能预测，并鉴定已知通路中的新型成员。借助36个种子性状，我们发现了160个G蛋白偶联受体（G-protein coupled receptor, GPCR）通路表达性状存在协同调控的证据。在这160个性状中，50个性状的主要LOD峰值位于2号染色体某位点8厘摩（cM）范围内，另有81个性状在该区域存在次级峰值。一个此前未被注释的理化学研究所（Riken）cDNA克隆，与硬脂酰辅酶A去饱和酶1（stearoyl CoA desaturase 1）的表达呈强相关性，经实验验证其可响应调控脂质代谢的环境条件。通过连锁定位，我们鉴定出多个其表达受转录调控位点控制的基因。若仅在该小型杂交群体中研究存在统计学显著性连锁的mRNA性状，则会遗漏部分调控网络，而将性状相关性与连锁定位相结合的分析方法则可揭示这类网络。相较于仅采用连锁定位的分析方法，联合分析的灵敏度更高。参考文献：Kendziorski C.、Chen M.、Yuan M.、Lan H.及Attie AD.：《表达数量性状位点（Expression Quantitative Trait Loci, eQTL）定位的统计方法》，《Biometrics》，即将刊出，2005年。Lan H.、Chen M.、Flowers JB.、Yandell BS.、Stapleton DS.等（2006）：《联合表达性状相关性与表达数量性状位点定位分析》，《PLoS Genet》2(1): e6。实验总体设计：本研究的F2-ob/ob小鼠取自我们为绘制糖尿病相关生理表型图谱而构建的定位群体（Stoehr等人，2000年）。其中约110只F2-ob/ob小鼠曾被用于通过实时定量逆转录聚合酶链式反应（quantitative real-time RT-PCR）检测得到的mRNA丰度性状的定位分析（Lan等人，2003年）。我们基于选择性表型分型算法（Jin等人，2004年），从这110只小鼠中筛选出60只，用于获取微阵列检测得到的mRNA丰度数据。F2-ob/ob小鼠于断奶后在威斯康星大学麦迪逊分校动物护理设施中饲养，采用12小时光照/12小时黑暗的光照周期。小鼠自由摄食普瑞纳Formulab Chow 5008饲料（脂肪含量6.5%）与酸化水。小鼠在14周龄时经过4小时禁食后，以二氧化碳窒息法实施安乐死。肝脏及其他组织立即用铝箔包裹，置于液氮中速冻，随后转移至-80℃冰箱保存。本次研究共采集29只雄性与31只雌性小鼠的肝脏样本。总RNA提取采用改良的一步法酸胍异硫氰酸酯-酚-氯仿萃取法，按照RNAzol Reagent（Tel-Test公司）的说明书操作。提取得到的RNA通过RNeasy试剂盒（Qiagen公司）进行纯化。采用紫外分光光度法检测RNA样品的浓度，通过溴化乙锭染色的变性甲醛琼脂糖凝胶电泳可视化观察来评估RNA质量。按照Affymetrix表达分析技术手册的步骤，将RNA样品反转录为cDNA，随后合成生物素标记的cRNA。将标记好的样品分别与M430A及后续的M430B基因芯片进行杂交。杂交、洗涤与扫描步骤由Hong Lan在威斯康星大学麦迪逊分校基因表达中心的Affymetrix核心实验室完成。