In a recent paper by Wu and Sharp the authors discuss the
concept of Divergent Transcription. Simply this is a study of all the
transcription that generally goes nowhere but from time to time does go somewhere
and in this case the development of new genes.
A few decades ago when we looked at the DNA world we thought
of it in terms of the Dogma: DNA to RNA to Proteins. Then the proteins did
things.
Then we found that we had about 3 Billion base pairs and
only about 20,000 genes. That means that we used only about 1-2% of our bases
and the other 98-99% were not really used, but not really. That unused DNA was
actually used in bits and pieces. There was a ton on non-coding RNA floating
all over.
Thus we might think that decades ago the cell was filled
with some well-organized proteins, coming from a well-orchestrated RNA process
of translation. It was like an airport in the US, with all the people, base
pairs, and lining up with the TSA, the promoters, and moving thorough the
scanners in order, each being read for proper ID, the scanner being the RNA
polymerase and coming out as ticked passengers grouping at each waiting area
for the assigned flight. Each waiting area was the proteins composed of the
translated bases now nucleic acids. Organized, controlled, and no unverified
interlopers.
But now we look again and it really appears like Penn
Station in New York. Doors all over, no lining up, people going on Amtrak,
LIRR, NJ Transit, subways, no waiting, no seats, no tickets, no security. Then
there are vagrants checking out the trash bins, and dozens of other types just
wandering and looking. Order may be there but not the type we see at say
Newark. There are, if you will, big RNAs and little RNAs, RNAs destined to
become proteins, namely passengers on some transport, but there are also just lots
of little segments of RNA going nowhere. These are the equivalent of non-coding
RNAs just wandering around, crowding up the floor, slowing down the passengers,
and at time changing who goes where.
The Model
Wu and Sharp conclude:
we propose that divergent transcription at promoters and
enhancers results in changes of the transcribed DNA sequences that over
evolutionary time drive new gene origination in the transcribed regions.
Although the models proposed here are consistent with significant available
data, systematic tests of these models await further advances such as in-depth
characterization of additional genomes and experiments designed to test
specific hypothesis. Over evolutionary times, genes formed through divergent
transcription can be shuffled to other locations losing their evolutionary
context. We envision future studies will uncover more functional surprises from
divergent transcription, and illuminate how intergenic transcription is
integrated into the cellular transcriptome.
Divergent Transcription is transcription that follows a
different path than the organized transcription that we think of in a highly
organized structure. As Seila et al stated:
Transcription initiation by RNA polymerase II (RNAPII) is
thought to occur unidirectionally from most genes. Here, we present evidence of
widespread divergent transcription at protein-encoding gene promoters.
Transcription start site–associated RNAs (TSSa-RNAs) nonrandomly flank active
promoters, with peaks of antisense and sense short RNAs at 250 nucleotides
upstream and 50 nucleotides downstream of TSSs, respectively. Northern analysis
shows that TSSa-RNAs are subsets of an RNA population 20 to 90 nucleotides in
length. Promoter-associated RNAPII and H3K4-trimethylated histones,
transcription initiation hallmarks, colocalize at sense and antisense TSSa-RNA
positions; however, H3K79-dimethylated histones, characteristic of elongating
RNAPII, are only present downstream of TSSs. These results suggest that
divergent transcription over short distances is common for active promoters and
may help promoter regions maintain a state poised for subsequent regulation.
As Wu and Sharp recount the classic transcription we use
their description:
In the textbook model of a eukaryotic promoter, the
directionality is set by the arrangement of an upstream cis-element region
followed by a core promoter. The cis-elements are bound by sequence-specific
transcription factors, whereas the core promoter is bound by TATA-binding
protein (TBP) and other factors that recruit the core transcription machinery.
Most mammalian promoters lack a TATA element (TATA-less)
and are CpG rich. For these promoters, TBP is recruited through
sequence-specific transcription factors such as Sp1 that bind CpG-rich sequences
and components of the TFIID complex that have little sequence specificity.
Thus, in the absence of strong TATA elements such as for
CpG island promoters, TBP-complexes are recruited on both sides of the
transcription factors to form preinitiation complexes in both orientations.
This model is supported by the observation that divergent
transcription occurs at most promoters that are associated with CpG islands in
mammals, whereas promoters with TATA elements in mammals and worm are
associated with unidirectional transcription
We demonstrate the two concepts below using a modified
graphic from Wu and Sharp. We show the TATA binding site on the gene and we
show the TBP, the TATA binding protein and a mediator and ultimately the RNA
Pol II. This is a classic unidirectional process moving across the exons and
generating mRNA which is then cleaned and changed to a protein. The related
cDNA do not show any of this underlying complexity.
Now below this is a second process, but now we show both
forward and backward transcription. This requires a bi-directional promoter
which Wu and Sharp discuss.
Wu and Sharp then argue that the model can be characterized
by the system below. We have modified their graphic so that we may take a small
step further. The Figure below depicts the four processes they consider:
1. Transcription: This is the classic transcription process
of taking DNA and changing it to RNA, usually an mRNA.
2. G+T Content: This is the G and T content of the intron
and the propensity for mutations to occur in that area and thus setting up a
region for the introduction of new gene type sequences.
3. U1 Process: There are small nuclear RNAs used to splice
RNA segments together and these are called spliceosomes. One of them us the U1
snRNA. As the mRNA segments are produced they get spliced together by these nuclear
RNA segments. They are powerful elements found in the nucleus.
4. PAS Process: The poly(A) is described as, from Baynes
& Dominiczak, pp 430-432, as: At the 3' end of all eukaryotic mRNAs
(with the exception of histone mRNAs), a polyadenosine track is added, termed
the polyA tail. The adenosine residues are not encoded by the DNA but instead
are added by the action of poly(A) polymerase using ATP as a substrate. This
polyA tail is frequently >250 nucleotides in length. Although it is still
susceptible to the action of exo-RNases, the presence of the polyA tail
significantly increases the lifetime of mRNA. The presence of the polyA tail
has historically been used to isolate mRNA from eukaryotic cells.
We now combine these elements into the Wu and Sharp dynamic,
as modified, below:
We can then represent this model by the meta-equation below:
Here we represent GT, T, U and PAS as some measure of each
of the four processes represented in the diagram. Admittedly this is at best an
ad hoc representation but it does demonstrate that indeed we have some form of
dynamical system and in turn this system depending on whatever the constants
are can become an unstable and ever growing process.
New Genes
Out of this process Wu and Sharp argue that new genes can be
born. This is an ingenious and compelling argument. The time scale for such a development
is not specified but perhaps it may be intuited. Also actual changes have yet
to be fully observed from beginning to end. Yet the pieces are logically
consistent and are all supported by the evidence.
First a brief summary of the splicesome (from Baynes and Dominiczak,
pp 430-432)
In the more complicated posttranscriptional processing of
eukaryotic mRNAs, sequences called introns (intravening sequences) are removed
from the primary transcript and the remaining segments, termed exons (expressed
sequences), are ligated to form a functional RNA.
This process involves a large complex of proteins and
auxiliary RNAs called small nuclear RNAs (snRNAs), which interact to form a
spliceosome. The function of the five snRNAs (U1, U2, U4, U5, U6) in the
spliceosome is to help position reacting groups within the substrate mRNA
molecule, so that the introns can be removed and the appropriate exons can be
spliced together precisely. The snRNAs accomplish this task by binding, through
base-pairing interactions, with the sites on the mRNA that represent
intron/exon boundaries. Accompanying protein factors are responsible for
holding the reacting components together to facilitate the reaction.
We summarize the U below:
snRNA
|
Size
|
Function
|
U1
|
165 nt
|
Binds the 5' exon/intron boundary
|
U2
|
185 nt
|
Binds the branch site on the intron
|
U4
|
145 nt
|
Helps assemble the spliceosome
|
U5
|
116 nt
|
Binds the 3' intron/exon boundary
|
U6
|
106 nt
|
Displaces U1 after first
rearrangement
|
We explain how this may work in the Figure below adapted
from Wu and Sharp. There is on the left a progression of changes in a segment
of DNA which would normally read left to right with the inclusion of a new
segment from right to left. The PAS sites, three as shown in the Figure below,
are covered by RNA segments ultimately allowing the creation of an Exon and Intron.
The process is further elucidated on the right where are gene is putatively
relocated from one chromosome to another or even just duplicated.
Let us use Wu and Sharp’s text and go through the argument. They
proceed as follows:
One consequence of transcription is that it can cause mutations,
especially on the coding (nontranscribed) strand.
During transcription, transient R loops can be formed
behind the transcribing RNA polymerase II, exposing the coding strand as
single-stranded DNA, whereas the noncoding strand is base paired with and thus
protected by the nascent RNA.
The lack of splicing signals in the divergent transcript
also makes it more vulnerable to R loop formation, as splicing factors have
been implicated in suppressing R loop formation.
In addition,
divergent transcription generates negative supercoiling at promoters, which
facilitates DNA unwinding and promotes R loop formation.
As a consequence of R loop formation, the single-stranded
coding strand is vulnerable to mutagenic processes, such as cleavage,
deamination, and depurination. Genomics studies have shown that during mammalian
evolution, transcribed regions accumulate G and T bases on the coding strand,
relative to the noncoding strand or nontranscribed regions.
Evidence suggests that such strand bias may result from
passive effects of deamination, transcription-coupled repair, and somatic
hypermutation pathways in germ cell-transcribed genes, in the absence of selection.
Accumulation of G and T content on the coding strand will
strengthen the U1-PAS axis.
A-rich sequences such as PAS (AATAAA) are likely to be
lost when the genomic DNA accumulates G and T.
In contrast, G+T-rich sequences, such as U1 snRNP-binding
sites (e.g., resembling 50 splice sites, G|GTAAGT and G|GTGAGT), are likely to
emerge in these regions. Since promoter-proximal PAS reduces transcriptional activity,
the loss of PAS and gain of U1 sites should contribute to lengthening of the
transcribed region as well as its more robust transcription.
The gain of U1 sites could also enhance transcription by
recruiting basal transcription initiation factors or elongation factors.
Therefore a positive feedback loop is formed: active
transcription causes the coding strand to accumulate sequence changes favoring
higher transcription activity.
As noted above, strengthening of the U1-PAS axis also
favors extension of the transcribed region. Being longer gives the transcript several
advantages: by chance longer RNAs are more likely to contain additional
splicing signals such as a 30 splice site to become spliced, or binding sites
for splicing-independent nuclear export factors, thus escaping nuclear exosome
degradation by packaging and exporting to cytoplasm .
Longer RNAs are also more likely to carry an open reading
frame, either generated de novo or by incorporation of gene remnants.
Once in the
cytoplasm, the RNA should at some frequency be translated into short
polypeptides due to widespread translational activity.
Some of the polypeptides may provide advantage to the
organism and become fixed in the population, thereby forming a new gene.
Thus we have seen a mechanism for new gene creation and
insertion.
Observations
These are a very powerful set if insights and observations. They
have significant conclusions as has been articulated by those in Sharp’s Lab. The
metaphor of a train station with wandering fragments of often “useless” RNA has
certain merit. However all too often those fragments are not useless but have
ways of interfering and disrupting the normal progress of cellular dynamics.
We now pose a few observation which may have some merit.
1. Somatic vs Germline: These changes seem to be mitotic in
nature and thus are reflected in somatic cells. What is the impact in meiosis
and germ line cells? Namely can these mutations be carried forward and be
selected out in subsequent generations? Or is this process one almost
exclusively found in somatic cells and thus may be causes for such diseases as
the cancers? I could not find a clear path to follow here.
2. Causation: What causes some of these processes. Many if
not most of the links are presented and explained but ultimate causality is
missing.
3. Frequency: How frequently do these changes occur? Are
they rare or common and at what rate do they occur? What are the overall temporal dynamics of
these processes. Can we examine genomes and ascertain where they might occur. We
all too often just skip over the Introns, focusing on the Exons and their
resultant expression. There also are many regions of the Exons that are not
expressed, and are they part of this phenomenon as well?
4. Reaction Dynamics: The actual reaction dynamics could
possibly be explained and modelled. We have presented a meta model solely for
the visualization of what may happen. It is expected that the model is most
likely non-linear and more complex. In fact the actual metrics being measured and
modelled are still in question. However not withstanding that we can envision a
dynamic model exhibiting not only stability issues but also oscilliatory
effects.
5. Methylation and Epigenetic Factors: Clearly the CpG
islands play an important factor. Methylation has become a significant area of
study over the past decade and the processes described herein rely on many of
these CpG islands as well. Is methylation a competing process, an allied
process, a controlling or mediating process?
6. What are all these RNA fragments doing?: Ultimately we
find that a cell may have not only well understood Dogma based proteins and
pathways but also a mass of disconnected non coding RNA spinning about in the
nucleus and throughout the cell. Thus we ask; what do these snippets do? Are
they just wanderers going nowhere and possibly just bumping into those going somewhere
or are the truly entities which have predictable effects on pathways? Are they
noise or an aberrant signal?
This is a very compelling paper and it presents in an
elegant manner the results of the efforts to date. This effort demands to be
followed and examined in detail as it progresses.
References
Baynes Dominiczak:
Medical Biochemistry 3E, Mosby (New York) 2013.
Seila, A., et al, Divergent Transcription from Active
Promoters, Science VOL 322 19 December 2008 1849.
Wu, X., P. Sharp, Divergent Transcription: A Driving Force
for New Gene Origination? Cell 155, November 21, 2013.