I’m interested in finding out the minimum number of adult worms necessary for extracting sufficient RNA for RNA-seq transcriptome profiling.
I would be grateful to learn about any attempts at conducting such analyses on small number of individuals (less than 400, ideally), the specific number of worms, the kits/protocols used, tips to follow, and pitfalls to avoid.
Some context: I’m hoping to examine how the transcriptome changes with age under experimental conditions. As my worms age, I’ll lose individuals to escapes and senescent deaths, and since I’m not working with C. elegans, I’ll have to separate F1s manually (no sterile mutants have been characterized for this species). All told, I’m foreseeing some difficulties maintaining large numbers of worms in the face of these issues.
Any help you can give me before I embark on this fool’s errand will be greatly appreciated!
We have prepared RNA-Seq libraries from 200 hand-picked sterile C. elegans adults for differential gene expression, and the results were comparable to bulk samples. I believe our recovery was in the range of 100 ng total RNA, which is sufficient for a variety of the standard library prep kits. There are also linear-amplification techniques for lower amounts of input, but those should not be necessary.
You should also look at some of Aviv Regev’s work. They’ve been doing beautiful single-cell RNA-seq in dendritic cells. She was strongly advocating sequencing multiple single cells to limited depth (250 K reads/cell). Would probably need more worms, but her data was convincing at finding differential expression of highly regulated genes between two experimental conditions. For lowly expressed genes, the noise from the various sources of experimental bias (i.e. amplification) were much larger than the likely biological variation. I also suppose it may depend on which tissues you are interested. I wonder if this could be applied to single worms or small numbers of worms. Theoretically one should be able to do single-worm seq, if one wants to identify the most significant, reproducibly changed genes under your experimental condition.
Without actually weighing in on the fraught issue of the legitimacy of different methods of amplifying message, I’d assumed the question was about doing RNAseq without amplification. Single-cell profiling can be great (I’ll give a shout-out to the home team, as it were, as an example), but may not be wholly relevant to the question.
Harold, I’d be interested in hearing more details on the RNA extraction method you employed.
Is there a paper you can direct me to that shows these methods/results (so I can convince the powers-that-be of its feasibility)?
And w.r.t. the other replies, no, I’m not planning on using an amplification step (if possible).