Combined bacterial and fungal intestinal microbiota analyses: Impact of storage conditions and DNA extraction protocols

BackgroundThe human intestinal microbiota contains a vast community of microorganisms increasingly studied using high-throughput DNA sequencing. Standardized protocols for storage and DNA extraction from fecal samples have been established mostly for bacterial microbiota analysis. Here, we investigated the impact of storage and DNA extraction on bacterial and fungal community structures detected concomitantly.MethodsFecal samples from healthy adults were stored at -80°C as such or diluted in RNAlater® and subjected to 2 extraction protocols with mechanical lysis: the Powersoil® MoBio kit or the International Human Microbiota Standard (IHMS) Protocol Q. Libraries of the 12 samples targeting the V3-V4 16S and the ITS1 regions were prepared using Metabiote® (Genoscreen) and sequenced on GS-FLX-454. Sequencing data were analysed using SHAMAN (http://shaman.pasteur.fr/). The bacterial and fungal microbiota were compared in terms of diversity and relative abundance.ResultsWe obtained 171869 and 199089 quality-controlled reads for 16S and ITS, respectively. All 16S reads were assigned to 41 bacterial genera; only 52% of ITS reads were assigned to 40 fungal genera/section. Rarefaction curves were satisfactory in 3/3 and 2/3 subjects for 16S and ITS, respectively. PCoA showed important inter-individual variability of intestinal microbiota largely overweighing the effect of storage or extraction. Storage in RNAlater® impacted (downward trend) the relative abundances of 7/41 bacterial and 6/40 fungal taxa, while extraction impacted randomly 18/41 bacterial taxa and 1/40 fungal taxon.ConclusionOur results showed that RNAlater® moderately impacts bacterial or fungal community structures, while extraction significantly influences the bacterial composition. For combined bacterial and fungal intestinal microbiota analysis, immediate sample freezing should be preferred when feasible, but storage in RNAlater® remains an option under unfavourable conditions or for concomitant metatranscriptomic analysis; and extraction should rely on protocols validated for bacterial analysis, such as IHMS Protocol Q, and including a powerful mechanical lysis, essential for fungal extraction.

**背景** 人类肠道菌群（human intestinal microbiota）包含庞大的微生物群落，目前越来越多地通过高通量DNA测序技术开展相关研究。此前建立的粪便样本储存与DNA提取标准化方案，大多仅针对细菌菌群分析。本研究旨在探究储存与DNA提取流程对同时检测到的细菌与真菌群落结构的影响。

**方法** 收集健康成年人的粪便样本，将样本分为两组：一组直接置于-80℃冻存，另一组用RNAlater®试剂稀释后保存。随后采用两种带有机械裂解步骤的提取流程处理样本：分别为Powersoil® MoBio试剂盒，以及国际人类菌群标准（International Human Microbiota Standard, IHMS）Protocol Q方案。使用Metabiote®（Genoscreen公司）构建针对12份样本的V3-V4区16S rRNA基因与ITS1区测序文库，并在GS-FLX-454测序平台上完成测序。测序数据通过SHAMAN工具（http://shaman.pasteur.fr/）进行分析。从群落多样性与相对丰度两个维度，对细菌菌群与真菌菌群开展比较分析。

**结果** 16S rRNA基因与ITS区域分别获得171869条和199089条经过质量质控的测序读段（reads）。所有16S读段均可注释到41个细菌属；仅52%的ITS读段可注释到40个真菌属/类群。稀疏曲线（rarefaction curves）结果显示，针对3/3受试者的16S测序数据与2/3受试者的ITS测序数据均达到饱和。主坐标分析（Principal Coordinates Analysis, PCoA）结果表明，肠道菌群的个体间差异远大于储存或提取流程带来的影响。使用RNAlater®储存会对7/41个细菌类群与6/40个真菌类群的相对丰度产生下调趋势的影响；而提取流程则随机影响18/41个细菌类群与1/40个真菌类群的相对丰度。

**结论** 本研究结果表明，RNAlater®储存对细菌或真菌群落结构仅产生中等程度影响，而提取流程则会显著改变细菌群落组成。若需同时分析肠道细菌与真菌菌群，在条件允许时优先选择直接冻存样本；但若在不利实验条件下，或需同时开展宏转录组分析时，仍可选择RNAlater®储存样本。此外，提取流程应采用经过细菌菌群分析验证的方案（如IHMS Protocol Q），且需包含高效机械裂解步骤——这是真菌菌群提取的必要环节。