A flowering of RNA

Curiouser and curiouser: down the genomic rabbit hole with regulatory RNAs

One of the biggest molecular biology stories of the last decade was about RNA. The more we look, the more functions biologists find for RNA in the cell. We thought we understood RNA decades ago- the floppy and unstable molecule used in only three places- as rRNA, forming the structural scaffold of the translating ribosome; as tRNA to link amino acids to their codons during protein translation; and as mRNA- the message translated from DNA into RNA, which likewise enters the ribosome to serve as the template for translation.

That was it- nice and neat, with three functions all centered around translation. It was surprising enough to find out that the rRNA was not just a scaffold, but actually the catalytic center of the ribosome- further evidence for the centrality of RNA in that way-back eon when DNA hadn't been invented yet, before the full advent of life as we know it.

But RNA has kept turning up in the oddest places over the last decade or two, clearly not willing to be neatly boxed up and put on the shelf of molecular biology knowledge. A flood of new prefixes and other variations tell the tale:

miRNA
ncRNA
lincRNA
snoRNA
piRNA
mirtron
RNAi
Xist

Clearly things (or people) are going a little nuts. No one knew about these RNAs before because they are typically small, or of modest significance. Small genes, especially of non-coding RNAs, are hard to find, since with the flood of genomic DNA sequence, we use trained computers to find genes by traditional rules, which prominently include the genetic code to conceptually translate DNA to proteins, and statistics to decide whether a gene is "real" or not (i.e. large). But nature is not so tidy, and doesn't need any stinkin' rules. It makes things up as it goes along.

The central "dogma" of DNA -> mRNA -> protein still holds for most functions of the cell. The new RNAs don't make structures in the cell, but regulate other genes, adding to an already elaborate network of control. They control, but don't stick around.

A recent paper in science (reviewed here; general review) provided a fascinating example of one of these RNAs in action, operating in that eternally interesting pathway that plants have to figure out whether to flower or not. First off, I should say that this pathway is not fully understood. The actors are only gradually coming into focus, and this paper is just one scene in a long-running production.

Vernalization is the process by which plants sense the alternation of cold to warm in order to flower properly in the spring, when we all want our daffodils to bloom. No winter, no tulips, crocuses, or other delightful flowers. The model plant used in this paper and elsewhere is the small weed Arabidopsis thaliana. We can assume that its processes more or less generalize to all plants, especially to most temperate flowering plants.

A central gene the vernalization process is FLC (flowering locus C), which represses a battery of other genes involved in flowering by repressing their transcription. It is a central regulator, and typifies the very common motif of repression, which often occurs in extensive circuits of double-negatives. Biological circuits are logical, but that doesn't mean they are intelligently designed!

Anyhow, FLC is usually on, repressing the whole flowering razmatazz in most tissues and at most times. During extended cold temperatures, however, FLC is turned off in gradual, progressive fashion in selected tissues, which, when maintained as temperatures rise again, allows flowering. This shutoff of FLC has to be maintained through many cell divisions, as the flower grows, so its mechanism has to be quite robust. A central question, then, has been how FLC is turned off in response to cold in the gradual way required to sense the winter season. And that is what this paper is about.

Prior papers showed that FLC is turned off by a venerable mechanism known as the polycomb system, named for the fruit fly where it was first found to have subtle effects on the male sex comb (don't ask!). Polycomb proteins form large complexes linearly spread over the chromatin (DNA plus histones, etc.) and also chemically modify histones, shutting off nearby genes in a permanent fashion- just the thing when you are, say, a skin cell, and don't ever want those liver genes turned on again. Polycomb proteins are used frequently in developmental processes once final decisions have been made. Their repression is carried along through mitosis to progeny cells, as required by the vernalization and flowering process.

Prior work also found one gene (VIN3) that is necessary (though not  sufficient) for establishing FLC repression in cold temperatures, as well as other genes encoding proteins of the generic polycomb complexes which are required to maintain the repression once temperatures rise again. VIN3 is another transcriptional regulator, making this whole story rather humdrum and typical, so far.

But the thing about polycomb proteins is that they don't glomb onto a zone of chromatin and do their thing without some extra help ... they need direction from RNA guides, whose sequence directly mates with the complementary DNA and then attracts the polycomb proteins. Only in the last few years has this necessity for RNA been realized, opening a new field looking for such guide RNAs all over the genome that service these developmental repression processes. The VIN3 protein may help turn the FLC gene off, but it can't by itself set up more durable repression by the polycomb complexes. And that is where the new paper comes in, finding an RNA, (named COLDAIR), which is encoded by the first intron within the FLC locus itself, and which seems to guide polycomb repression of FLC.

COLDAIR non-coding RNA is transcribed from the first intron of the FLC gene. VRE stands for "vernalization response element", which is a DNA site controlling transcription of COLDAIR, presumably by binding activating proteins.

The researchers show that COLDAIR is essential for the vernalization process, is expressed at the right time, and associates with the polycomb proteins as hypothesized, both in the test tube and in plants. They somewhat acidly note that a nearby RNA found by others (COOLAIR) neither has any known role in vernalization (by deletion or other functional test) nor associates with polycomb proteins, despite a proposed role in the process. The current authors had found COLDAIR by intensively searching through the FLC gene for stray transcription products appearing under the right conditions, hypothesizing that as a subject of a polycomb repression process, such an RNA would be required to direct its location/nucleation. COLDAIR is about 1100 bases long- very long for minor regulatory RNAs in general, but typical for these polycomb complex guide RNAs.

Expression of relevant genes during vernalization, expressed as number of days of cold temperature (V) or warm (T).

How all these dots connect isn't entirely clear, unfortunately. What regulates the expression of the COLDAIR RNA at the right time and place in response to cold temperature isn't known. What connects VIN3 protein binding and repression (and local histone de-acetylation) with COLDAIR RNA expression or recruitment is also not known, though it is likely that VIN3 represses FLC transcription directly and partially modifies the local histones. And the details of how RNA in combination with the VIN3 DNA binding protein can guide the polycomb complex growth around the FLC gene is not quite clear, though direct triplex formation between the RNA and the DNA duplex is a leading theory.

While a work in progress, the polycomb story is most interesting and general. The master regulators of mammalian body plan development, (capable of cutting off limbs, digits, and vertebral segments when misexpressed), the HOX genes, are regulated in a similar fashion, turned off in various areas of the body by the polycomb system using locally produced guide RNAs, similar to what is described above. (Landmark paper.) Indeed it is becoming apparent that our cells are full of stray RNAs that may add substantially to the count of "genes" in the human genome. These don't code for protein components of the physical body, but regulate how other genes operate, lending support to the theme that our complexity arises less out of the hardware of what we are made of, and more out of the vastly complicated (though also quite junky) software controlling how, where, and when the limited number of pieces are put together.

• Some GOP commentary on Egypt.
• GOP announces new climate strategy: Abandon Earth.
• Does anyone do background checks at Freddie Mac?
  
• A bit of ham & jazz.
• Bill Mitchell quote of the week: (Quote taken from a recent report criticizing the economists at the IMF, showing how ideologically blind they were and remain, here in regard to Iceland.)

"In spite of a banking sector that had grown from about 100 percent of GDP in 2003 to almost 1,000 percent of GDP, financial sector issues were not the focal point of the 2007 Article IV discussions. The massive size of the banking sector was noted, but this was not highlighted as a key vulnerability that needed to be addressed urgently. Instead, the IMF worried about the possibility of overheating, and the staff report was sanguine about Iceland’s overall prospects. For example, the headline sentences in the staff appraisal were “Iceland’s medium term prospects remain enviable. Open and flexible markets, sound institutions … have enabled Iceland to benefit from the opportunities afforded by globalization.” The report presented a positive picture of the banking sector itself, noting that “the banking sector appears well-placed to withstand significant credit and market shocks” and “[B]anks took important steps over the past year to reduce vulnerabilities and increase resilience.”"

• Lastly, we are all Egyptians this week, with high hopes for the future.