研究思路

研究方法

实验样本的选取

选择肿瘤细胞丰富的区域，用手动打孔机钻取组织样本，用数码相机拍照，然后用ImageJ软件测量出组织体积。

提取试剂盒的选择

一共14个试剂盒，包含了16种提取DNA和RNA的方法，其中RecAll和AllPrep既可以提取RNA又可以提取DNA。

核酸的提取及定量

样本前处理：从4块归档的FFPE组织石蜡块中钻取了120个样，共计37.2mm³。用二甲苯室温脱蜡5分钟两次，无水乙醇室温复水5分钟两次，然后使用匀浆机匀浆。

核酸提取：每份样本0.68mm³,按照各个试剂盒说明书提取，每个试剂盒三重复。其中RecAll和AllPrep试剂盒提取时，在匀浆后的组织中添加蛋白酶K，然后取上清提取RNA，沉淀提取DNA。

核酸产量及纯度检测：使用dsDNA HS和RNA BR在Qubit3.0上分别对DNA及RNA定量；使用分光光度计测定A260/280和A260/230判定DNA及RNA纯度。

Limitations of this study

This study has some notable limitations. Kit performance was tested on cores but not on tissue sections, which may be better suited for certain applications. Nevertheless, cores are a popular and efficient method of selecting tissue of interest for molecular assays. In addition, the head to-head kit comparison implemented was based on relatively recent tissue samples. Older samples were evaluated with the AllPrep kit but not with any of the other kits.

研究结果

1. 核酸产量和完整性

By spectrophotometric assessment, all of the RNA extraction kits produced A260/280 ratios close to 2.0, consistent with highly “pure” samples. In contrast, with the exception of the RNeasy kit, all of the RNA extraction kits produced A260/230 ratios that indicated significant impurities (Table 2). Similarly for DNA, the A260/280 ratios were near or above 1.8 for all kits. In contrast, the A260/230 ratios for the RecAll, AllPrep, QIAamp, and DNeasy kits indicated potential organic contaminants in the eluent (Table 2). Contaminants such as EDTA, phenol, heme, and carbohydrates all have absorbances near 230 nm. Given that these contaminants can inhibit downstream applications, we undertook a series of assays to quantify their inhibitory effect.

2. PCR抑制剂效果检测

The inhibition assay was designed to quantify the impact of any inhibitors in eluted RNA and DNA samples on downstream applications. Extracted RNA and DNA were spiked into a non-target PCR reaction, and the observed delay in the Cq value of the RNA- or DNA-spiked reaction relative to that of the control (water-spiked) was interpreted as the extent of the inhibitory effect. 单因素方差分析结果显示，8个RNA提取试剂盒的Cq值和Control的Cq值基本在同一水平，表明RNA纯度较高；6个DNA提取试剂盒中有5个试剂盒的Cq值和Control的Cq值基本在同一水平，但AllPrep和对照组有明显差异。

AllPrep试剂盒PCR抑制剂补充实验

额外选取了前列腺癌和乳腺癌样本进行抑制实验，发现抑制现象只在前列腺癌样本中发生。前列腺癌样本DNA分别稀释十倍和二十倍再进行实验，在稀释十倍时抑制情况有明显改观。

结论：
测试的试剂盒中，AllPrep提取的DNA中存在PCR抑制剂（仅在前列腺癌样本中），但可以通过稀释消除抑制效应；其余试剂盒提取的DNA和RNA与对照相比，没有明显的抑制效应。

3. 核酸片段大小的评估

方案1：RNA样本使用RNA 6000 Pico 试剂盒在2100生物分析仪上检测，用大于200bp碱基对百分比（DV200*）表示。

结果见表2，相同样本用不同试剂盒抽提获得的DV200值相差很大。

*DV200值表示的是片段大于200bp的相对数量，而不是绝对数量，并不能用于反映进入PCR反应的核酸的性能。

方案2：评估RNA片段分布时先进行逆转录，然后用扩增子大小从92bp到386bp（大约以50bp大小递增）的6对碱基对和cDNA进行PCR扩增；评估DNA片段分布时用扩增子大小从102bp到300bp（大约以65bp大小递增）的4对碱基对和DNA进行PCR扩增；扩增后产物用3%凝胶电泳质检。

In the current study, no clear correlation was observed in the fragment size distribution as determined by end-point PCR versus Bioanalyzer analysis. It is noteworthy, however, that extensive fragmentation of both RNA and DNA, which is typical for FFPE samples, made the interpretation of the Bioanalyzer electropherogram unreliable(S1 Fig).

Based on yield and purity, five DNA protocols (AllPrep, RecAll, DNeasy, QIAamp, and GenJet) and four RNA protocols (AllPrep, RecAll, PuLink, and RNeasy) were prioritised for further quality assessment. The prioritised kits were further evaluated using NanoString, RT-qPCR expression, and MS-PCR assays.

4. NanoString和实时荧光定量法对mRNA 检测

The NanoString platform has previously been used for direct, digital quantitation of specific mRNA transcripts through hybridization to two sequence-specific, color-coded probes . As this technology is strictly hybridization-based, it avoids the use of reverse transcription and amplification, and thereby eliminates potential amplification bias common to PCR.

Pairwise comparisons between RNA extracted from the cell line and each of the four selected kits showed no significant difference in total signal counts (Fig 2A and S2 Fig).

By RT-qPCR, RNA from all four prioritized RNA protocols were successfully used to amplify the three house-keeping genes tested (PGK1, KRT8 and HPRT1), showing the overall compatibility of these protocols with downstream gene expression assays. Of these, the AllPrep protocol yielded significantly higher Cq values for all three genes (two-way ANOVA, p < 0.0001; Fig 2B). 原因是RNA中有PCR抑制剂。

5. 甲基化特异性PCR检测DNA

We targeted three genes known to be hypermethylated in prostate cancer, namely GSTP1, ABCB1, and RASSF1, along with the Alu repeat sequences as a positive control. DNA from the AllPrep, DNeasy, and GenJet kits resulted in similar Cq values (Fig 2C). Compared to the other kits, and using equivalent amount of DNA input, RecAll and QIAamp had significantly higher Cq values for each of the three gene targets (two-way ANOVA, p < 0.05).

These results indicated that AllPrep DNA was most compatible with the methylation-specific PCR protocol used in this study.

**6. AllPrep提取方案重现性

Inter-laboratory validation of a molecular protocol or assay is performed by molecular pathology laboratories to ensure its reproducibility and accuracy, which are critical components of competent patient care. We aimed to identify the optimal RNA and DNA extraction protocols and then determine whether they are reproducible across multiple laboratories.

We placed a high priority on the metrics routinely used in pre-analytical quality control, such as spectrophotometric absorbance and fragment distribution, and selected Qiagen’s AllPrep DNA/RNA FFPE kit as the optimal one for assessing the reproducibility of extraction protocols between three laboratories.

One-way ANOVA analysis did not detect significant differences between laboratories in yield of DNA or RNA (p > 0.05, Fig 3A), or in the absorbance at 280, 260, and 230 nm (results not shown). The results of the MS-PCR (S3C Fig) and NanoString (S3D Fig) assays using nucleic acids extracted independently by the three laboratories were highly correlated (all Pearson’s R2 ≥ 0.94; p < 0.0001).

Thus, the AllPrep kit produced nucleic acids that strongly correlated between the three laboratories in yield, quality, and compatibility with downstream applications.

S3 Fig. Comparisons of nucleic acid yield and endpoint assays across three independent labs. (A) DNA and (B) RNA yields across the three labs. Results of the (C) MS-PCR and (D) NanoString assays performed using serial extractions of RNA and DNA from 12 FFPE prostate cancer samples. R2 values are Pearson’s correlation coefficients (all p values < 0.05).

7. Effect of sample age on yield and downstream molecular applications

The 12 FFPE samples used in the inter-laboratory study on the AllPrep kit ranged in age from 7 to 15 years. Pearson’s correlation analysis of the data failed to detect any correlation between the sample age and the yield (Fig 3B). Likewise, sample age was not significantly correlated with MS-PCR Cq values (Fig 3C) or with NanoString total mRNA counts (Fig 3D, p ≥ 0.05).

样本保存年份与核酸产量之间的没有相关性；与 MS-PCR Cq values 或 NanoString total mRNA counts之间也没有相关性。

参考资料：

1. Patel, P.G. et al. Reliability and performance of commercial RNA and DNA extraction kits for FFPE tissue cores. PLoS One 12, e0179732 (2017).