Am reading a study about a bacteria that can resist UV radiation. They were able to do phylogenetic studies on the 16S rRNA. Then, they exposed the cells to gamma radiation and calculated how many survived.
In another study, they completed draft genome sequence of this bacteria. They did gene annotation to try to determine genes related to the UV resistance and other types of environmental stress.
How feasible and/or useful would it be for the next study to try to perform some type of RNA-sequencing on this bacteria to really pinpoint how the genes change in response to say 3 different levels of UV radiation? Several smaller questions:
1) Is RNA-sequencing possible for bacteria? I have only used eukaryotes before for RNA-sequencing. If it is possible, would any R packages help determine "differentially expressed genes". For instance, with eukaryotes, I used DESeq2 to obtain DEGs. What would one use to determine DEGs when bacteria is exposed to different levels of UV radiation?
2) Would methylation studies be of interest for these 3 different levels of UV radiation in a bacteria? What type of R package would be useful for bacterial methylation?
3) How much would such an RNA-sequencing experiment cost? Say I had 3 reps (3 cultures of cells) in each of these 3 UV radiation level groups. I would have 9 samples.
4) How much would such a methylation experiment cost? Say I had 3 reps (3 cultures of cells) in each of these 3 UV radiation level groups. I would have 9 samples.
Even if anyone can answer one of these questions, I would be grateful. I am new to prokaryotes and feeling overwhelmed with trying to compare and contrast with what I know about eukaryotic RNA-sequencing and methylation studies. Thank you many fold to anyone who can kindly share advice/knowledge on this matter.
Thanks, Mark Ziemann. I really appreciate your responses. I am particularly interested in your answer to #1, which made me start looking for a few additional answers. If you have a moment to reply, I would be very grateful. Thank you so much again for sharing your advice.
1) Are there any other recommended ways to try to determine which genes in a bacteria are "differentially expressed" in response to a treatment (here UV radiation)? Is RNA-sequencing the preferred way to do this? I am only asking because you stated that "most reads will map to the highly abundant rRNA", and I was uncertain if that was a good or bad fact (or just a fact to not worry about).
2) Would the procedure for RNA-sequencing for bacteria be similar to that for eukaryotes? I am assuming I would still want "replication"? I guess I am used to one "replicate" being one specimen for eukaryotes, but in the case of bacteria, one "replicate" would be one culture (a Petri dish perhaps) of bacteria specimens?
3) In prokaryotes, is it still possible to do downstream analyses? Such as GO and pathway analyses of the DEGs?
4) If you happened to have analyzed both eukaryotes and prokaryotes by RNA-sequencing, are there any differences or warnings you would recommend to somebody new to prokaryotes?
I think the best approach would be total RNA sequencing. Your genomics core lab might have a way to enrich bacterial mRNA and deplete rRNA. If they can't then understand that a large proporiton of reads will be rRNA. This means there will be fewer reads representing other genes. The total number of reads required for the experiment will need to be higher if rRNA is not depleted first.
In eukaryotes, it is common to deplete rRNA using either specific degradation (Ribominus/RiboZero) or mRNA enrichment using oligo dT beads. Bacteria don't polyadenylate mRNA so that won't work. rRNA depletion kits are mostly designed for mammalian analysis, you'll need to see if there's one for bacteria. If none exist, then you're options are limited to total RNA seq with the caveat that most reads will map to rRNA.
Yes, although you'll need a database that maps bacterial gene names to GO annotations.
The major consideration is the quality of the reference genome and annotation databases. Bacteria can be very diverse and not all of them have high quality genome assemblies or annotations. If these resources are high quality (like human and most model organisms) then it makes the downstream analysis much easier. If you have good quality reference data, then bacterial analysis could be more simple in that you don't need as many sequence reads to get good transcriptome coverage due to fewer genes.