Comparison of globin RNA processing methods for genome-wide transcriptome analysis from whole-blood

Whole blood rather than purified peripheral blood mononuclear cells is likely to become the prime tissue using expression microarrays for disease predication or prognosis however excess of globin mRNA may reduce probe detection sensitivity. In our study, we assessed whether whole-blood or globin-reduced RNA gives the most robust and sensitive results to detect small gene expression changes in response to hormone replacement therapy exposure. Each sample (N = 12) were hybridized according to 3 different protocols: no globin reduction (controls), globin reduction using peptid nucleic acids (PNA) and using magnetic beads in the GlobinClear kit from Ambion. Finally, 7 and 4 technical replicates were conducted in no globin reduction and PNA groups, respectively. Both globin reduction approaches were mostly efficient at reducing globin RNA from cRNA. Samples processed by GlobinClear kit gave a very distinct gene expression profiles from the controls while samples processed with PNA gave an intermediary profile closest to the non globin reduction group with a slight increased sensitivity of transcript detection but a loss of reproducibility. Overall, no sign of higher sensitivity in detection of gene expression changes followed by hormone exposure was observed after globin reduction which was therefore judged not beneficial. Keywords: Groups comparaison To assess effect of globin reduction protocols on gene expression from whole-blood, we selected 12 postmenopausal women (6 HRT users and 6 non-HRT users) who were not using other medication than HRT at the time of blood sampling and in order to cover a wide body mass index (BMI) range in both HRT and non HRT users. The samples were also selected to contain the highest concentration of total extracted RNA. Each sample was hybridized according to 3 different protocols: no globin reduction (controls), globin reduction using PNAs, globin reduction using magnetic beads in the GlobinClear™ kit from Ambion. The reproducibility of the globin reduction method compared to the non globin reduction approach was investigated for the PNA method only since this method was the most effective and specific. We planned to conduct 7 technical replicates in each group (i.e. no globin reduction and PNA groups). During amplification process with PNA, 4 samples failed to reverse transcript long fragment of mRNA certainly due to inhibitory contamination and these arrays were therefore excluded from our analyses. One sample (sample 34) was amplified with PNA twice the same day to study technical reproducibility without amplification date effect. Finally, a total of 47 arrays were conducted including 7 and 4 technical replicates in no globin reduction and PNA group, respectively.

相较于纯化的外周血单个核细胞（purified peripheral blood mononuclear cells），全血更有可能成为表达芯片（expression microarrays）用于疾病预测或预后的首选组织，但珠蛋白mRNA过量可能会降低探针检测灵敏度。本研究旨在评估，全血或经珠蛋白消减处理的RNA，哪种样本能获得最稳定且灵敏的检测结果，以精准检测激素替代疗法（hormone replacement therapy, HRT）暴露诱导的微小基因表达变化。本研究共纳入12份样本（N=12），分别采用3种不同的实验方案进行杂交：未进行珠蛋白消减处理的对照组、使用肽核酸（peptide nucleic acids, PNA）进行珠蛋白消减的实验组，以及使用安毕恩（Ambion）公司GlobinClear™试剂盒中的磁珠进行珠蛋白消减的实验组。最终，未进行珠蛋白消减组与PNA组分别完成7次和4次技术重复（technical replicates）。两种珠蛋白消减方法均能有效降低互补RNA（complementary RNA, cRNA）中的珠蛋白RNA水平。经GlobinClear™试剂盒处理的样本，其基因表达谱与对照组差异显著；而经PNA处理的样本则呈现出介于对照组与未消减组之间的表达谱，其转录本检测灵敏度略有提升，但重复性有所下降。总体而言，经珠蛋白消减处理后，并未观察到激素暴露诱导的基因表达变化检测灵敏度有所提升，因此认为珠蛋白消减并无益处。
关键词：组间比较 为评估珠蛋白消减方案对全血基因表达的影响，我们招募了12名绝经后女性（6名HRT使用者与6名非使用者），所有受试者在采血时仅使用HRT相关药物，且两组均覆盖了较宽的体质量指数（body mass index, BMI）范围。
同时，我们选取了总提取RNA浓度最高的样本。每份样本均采用3种不同方案进行杂交：未进行珠蛋白消减的对照组、使用PNA进行珠蛋白消减的实验组，以及使用Ambion公司GlobinClear™试剂盒磁珠进行珠蛋白消减的实验组。鉴于PNA法是目前最有效且特异性最优的珠蛋白消减方法，我们仅针对PNA法，探究了其与未消减方案相比的重复性表现。原计划在未消减组与PNA组中各完成7次技术重复。在使用PNA进行扩增的过程中，有4份样本因存在抑制性污染而无法完成mRNA长片段的逆转录，因此对应的基因芯片阵列被排除在本次分析之外。另有1份样本（样本34）在同一天内使用PNA进行了两次扩增，以探究不受扩增日期影响的技术重复性。最终，本次实验共完成47张基因芯片阵列，其中未消减组与PNA组分别包含7次和4次技术重复。