Prokaryotic Saturday: everything is always more complex of what it seems

Hi!

Although I skipped last week post, this is the  1^st month of PS!  J

Smiling cupcake time:

And for today’s post let me talk about this paper of Sorek and Cossart. You have to pay to get it, but I have it, so if you feel like reading, email me...... did I just say that?.

So, Sorek and Cossart discuss some findings that have been reported by studying transcriptomes of bacteria. They also spend a couple of paragraphs describing the main approaches for transcriptome studies, RNA-Seq and tilling arrays.

First of all, let us clarify what transcriptomes are. From all the definitions I've read, the one that I liked the most (although I don’t remember where I read it, sorry) said that a transcriptome is the dynamic part of a genome. After getting the genome of an organism, the whole research thing doesn't stop there, you have to know where are those genes located, and most importantly what are they doing and how do they interact with themselves and the environment. That information you get it from a transcriptome analysis.

So, for years many eukaryotes have been widely studied, and there are several genomes and transcriptomes available for model eukaryotes. And as usual, bacterias and archaeas were overlooked. Mainly because of that old belief that microorganisms, for being small, are simple and insignificant (well, poor things, what can I say, sometimes that is also believed of short people, or kids). But guess what?!! that is not only wrong, it is WRONG!!!

But that was not the only reason; actually it is more difficult to study transcriptomes of bacteria than those of eukaryotes. Again, more development is there for techniques to work with eukaryotes. In bacteria the major problem is the need to enrich mRNA in the sample. Prokaryotes lack the 3'-end polyA tail, and >95% of the RNA is ribosomal RNA and you need mRNA, the other 5%. But now it is possible to enrich mRNA and Sorek explains a bit of this process in this really cool paper.

 As I told you, the two main approaches that have been used to study transcripts in prokaryotes are RNA-seq and tilling arrays. RNA-seq can be done in several of the platforms available (you know SOLiD, Illumina, etc etc). First you extract your RNA and synthesize cDNA by reverse transcriptase (RT). As bacteria don't have the polyA tail, then some priming step has to be there for RT to work, with oligo(dt), random hexamers or artificial poly adenylation. Then, don’t forget the fact that we need mRNA and it is only a 5% of the sample. So, several methods are there for enrichment, and in Sorek and Cossart paper they are briefly explained. Here they are :

•  rRNA capture with magnetic beads, so at the end you remove the beads and the sample ends up being mRNA
• degradation of the 5’P RNAs, so mRNAs in bacteria have an analogous to the cap in eukaryotes, a 5’PPP (triphosphate). So the idea is to get rid of all those RNA species, tRNA and rRNA, that does not have a 5’PPP.
•  Polyadenylation of mRNA. Here an artificial polyA is added.
•  Capture the undesirable RNAs with Hfq, a very famous protein originally discovered in E. coli that binds to RNA.

I do not know if there are more enrichment methods. Details about each of these I've told you are presented in the paper, and some other references are suggested in case you want to learn more. Authors also give the name of the kits to do all these protocols, you know, if there are two people in the world doing same thing, there should be a kit for that.

So, once you have your mRNA, you get your cDNA libraries, and again, several methods are there for doing this. And the final output of the RNA-seq are millions of reads (from 20 to 200 bp), that you use to align to a reference genome, and the expression of the genes is measured depending of how many reads are aligned to that particular region in the genome.

And in the other method, tiling arrays, after the cDNA synthesis, the library is hybridized to an array and expression is measured using signal intensities. Enrichment is not necessary here. So both methods require a reference genome, and tilling arrays do not need an enrichment step. Several prokaryotic transcriptomes that have already been completed using these methods include: Listeria monocytogens, Bacillus subtilis, Halobacterium salinarum, Burkholderia canocepacia, Listeria monocytogens, E. coli, Salmonella, Sulpholobus solfataricus…

And what all people found, in general, is that prokaryotic transcriptomes are decidedly complex. Gene structures had to be remodeled as new genes were discovered and specially important ncRNA (non coding RNAs that are well know to play key roles in regulation of gene expression) were found. In this paper they mentioned one study (Wurtzel et al), in which they found 162 transcription start sites equivocally annotated. Although I don't know exactly why ORFs are usually predicted upstream of their actual places, this clearly shows how useful transcriptomes are.

Also, several riboswitches structures (you can read more about riboswitches in this previous PS) were discovered. It looks like 2% of bacterial genes are under riboswitch-mechanisms control. The detection of riboswitches is done by analyzing contiguous regions of the 5’UTRs at different conditions, when expression of such regions is interrupted at one growth condition and not in the other, you have spotted a riboswitch. Although this is a paper from 2010, an important comment they did is that there should be more focus on the 3'UTRs, especially in the case of archaea, which were already reported to have 3'UTR with regulatory roles. It'd be nice to find out what has been done in this regard in the last three years.

And now, to almost conclude this post, the most exciting (in my very personal point of view) part of the paper: gene plasticity. In this paper I found the reference of an experiment conducted in 2007, in which Mycoplasma pneumoniae was grown in 173 different conditions (woooth?!! ….... well, this shouldn't surprise us, we know someone who likes experiments of 100+ reactors ;)

So in this study, the found that operons (the bacterial genes) act in an homologous way to eukaryote genes, emulating alternative splicing… although, bacteria were here before eukaryotes, should we say that eukaryotes are emulating bacteria?

Probably, yes. In this study of the oh-so-many-conditions they saw that polycistronic regions can be transcribed as monocistronic when conditions varied, so, from one part with many genes, sometimes only one was expressed. So operons are versatile entities. And it looks like archaea act similar. So, who is copying who?

Another important discoveries were the definition of some ncRNAs in critical processes such as quorum sensing (more about this amazing bacterial signaling processes here) and the antisense transcription of some genes. So, isn't this cool? Although they are small, with small chromosomes and everything, prokaryotes overlap their genes, and it looks like "this is the rule, rather than the exception". 

I would like to keep writing, but this post is already too long. Please, check out the paper and if you have any comments, leave them down below these lines ↓.
thanks.