Re: Can a DNA coding sequence with 5' Kozak and a cap be used for translation?

Question: Can a DNA coding sequence with 5' Kozak and a cap be used for
translation?
From: No name entered.
Grade: grad (science)
City: No city entered., State/Prov.: No state entered. Country: No country
entered.
Area: Molecular Biology Message ID Number: 1176555046.Mb

Dear Anonymous Grad,

I think you are asking whether ribosomes can translate DNA instead of RNA.
This is a really fascinating question. While I can't find any direct
experimental evidence, the answer is almost certainly no. Let's consider
the question from each of the three stages of translation: initiation,
elongation, and termination.

Initiation:
Eukaryotic ribosomes are recruited to the mRNA by the G-cap, as you know.
The key, however, is that the next two nucleotides of the mRNA following
the cap are sometimes methylated on the 2'-OH (to give 2'-OMe). This
methylation would not be possible if the nucleic acid sequence were
deoxyribose based, since the 2'-OH would not exist. Any proteins that
recognize the cap would thus be unable to recognized a DNA-based cap, so
translation could not initiate. Also, the 2'-OH alters many aspects of RNA
structure, so losing it may cause any other mRNA associating proteins to
lose their affinity (these proteins would have low affinity for DNA), even
independent of the cap.

For example, after splicing, the mRNA is associated with exon-junction
complexes. There is also evidence that these complexes are very important
during the first round of translation (also called the initiator round,
where the very first ribosome comes through). A DNA sequence would be
unable to bind these exon-junction complexes, which could potentially cause
problems.

Elongation:
Let's assume, for the sake of argument, that we were able to somehow
pre-load a ribosome, in the middle of translation, with a DNA sequence.
Could the ribosome elongate in this case? The answer is still no. The 2'-OH
is a critically important element in RNA structure. In the ribosome, the
2'-OHs of the mRNA are probably making many important contacts with the
rest of the ribosome. Without it (as in the case of DNA), the
ribosome-mRNA-tRNA complex would be in the wrong conformation for peptidyl
transfer, tRNA recognition, elongation factor binding, etc. Crystal
structures of the ribosome has recently come out from the Ramakrishnan lab,
and are well worth a look.

Another problem would be in anticodon recognition. Correct base pairing to
the third base of each triplet codon is often not required (this is the
cause of redundancy in the genetic code). For this to happen, the base U is
critically important. It turns out that U can form "wobble-pairs" with G,
and this is one of many RNA-specific types of recognition accepted by the
third codon base (other examples involve non-canonical bases, such as
inosine, in the tRNA anticodon loop). DNA does not have U's, so this could
not happen. What if we substitute all the T's on DNA with U's, while
maintaining the deoxyribose backbone? Even then, as mentioned before, the
contacts made by the 2'-OH would be too important to lose.

Termination:
If the ribosome's made it all the way to termination on the DNA, I'd say
that counts as translation. But, due to many of the same structural
problems as in initiation and elongation, release factors would probably
have trouble recognizing a ribosome-DNA complex. Thus, even if the ribosome
could initiate and elongate a DNA sequence (highly highly highly highly
unlikely for the previous reasons!), it could probably not be able to
terminate correctly.

So far, we have been assuming that DNA would be accessible to ribosomes. In
fact, in the context of the eukaryotic cell, DNA is always sequestered to
the nucleus, where it is inaccessible to ribosomes. The only time that DNA
would be exposed to ribosomes would be during mitosis, when the nuclear
envelope dissolves. At that stage, however, DNA is highly condensed and
would be unable to be unravelled. Even if unravelling could happen, protein
synthesis is almost completely turned off during mitosis.

There is actually also an interesting evolutionary reason why the ribosome
does not recognize DNA. It is thought that the very first living organism
was composed of RNA, because RNA can both carry genetic information (in
terms of its sequence) as well as carry out reactions as a catalyst (based
on its structure). Thus, genotype and phenotype are collected in the same
molecule, which can presumably arise spontaneously more easily than
requiring DNA and protein to co-evolve in the prebiotic soup. This is
called the "RNA World" hypothesis. Ribosomes are often cited in support of
this hypothesis, because it is, in essence, a catalytic RNA complex. Thus,
the fact that ribosomes would translate RNA rather than DNA makes sense, if
you believe the RNA World hypothesis.

These are just a sampling of the reasons why DNA would not be able to be
translated by the ribosome. If you are interested in thinking up other
reasons, I recommend looking at a molecular biology textbook (such as
Molecular Biology of the Gene, by Watson, et. al.) and checking out the
ribosome crystal structures (just PubMed search for Ramakrishnan and ribosome).

Cheers!
Kathy