Previous Rapid Response Next Rapid Response Top
Nicholas Bennett,
Infectious Disease Postdoc/Clinician
Department of Pediatrics, University Hospital, Syracuse NY

Send response to journal:

I'm impressed that the Perth Group turn to me for answers to questions that frankly ought to be already known to them. Admittedly some of the points are at the cutting edge of current science, but others are quite frankly undergraduate genetics.

However, yes or no answers are not sufficient (even though this squares with the Perth Group's dogmatic approach to science). There are, as always, qualifiers and exceptions.

1. Irrespective of its origin, any RNA molecule (indeed, any molecule) can be inserted into a cell. The same procedures used for DNA insertion can be used or modified for RNA (eg lipofectamine transfection).

2. The right conditions for reverse transcription include an appropriate primer (tRNA for the retroviruses, components of telomerase for the telomeres) and actual RT enzyme. Certainly adding any old RNA to a cell is far more likely to result in non-specific degradation as non- specific anti-viral defense mechanisms kick in than the formation of cDNA. The primer would need to match both the RNA sequence and the reverse- transcriptase enzyme in question.

3. Integration into the cellular DNA is even more unlikely. Unlike RT, there are no cellular correlates of integrase and the whole integration phenomenon isn't yet completely understood. I would say that in the abscence of reverse-transcription that occured due to a retroviral RT/IN enzymatic duplex (as occurs in nature) the resulting cDNA doesn't stand a chance of getting integrated. The RT in retroviruses acts as a polyprotein including the integrase enzyme, so the function is sterically close to the DNA to be integrated. Integrase supplied in trans will not function unless it is packaged with the virus (eg as a vpr-IN fusion protein in some studies) and then subsequently cleaved to remove the extra peptide. It is a very unusual process.

4. cDNA transcription is exceedingly unlikely unless the appropriate transcription factor binding sites are present, and the matching transcription factors are also present. HIV with mutated regulatory regions will not produce RNA , and integrated HIV in the abscence of NF-kB will not produce RNA. Adding the gene is not sufficient, and remember that most cDNAs produced include only the mRNA sequence and not the regulatory regions. They require extra processing (in the lab) to add the appropriate sequences to allow them to be transcribed. It's like making a car but forgetting to add the engine under the hood! The RNA produced would initially be similar to the RNA introduced, but without appropriate correctional mechanisms the regulatory sequences at either end will be lost. Retroviruses employ a circular intermediate for their reverse- transcription to ensure replication of the terminal repeats that include the transcription/translation controls (and viral packaging!). They also prevent splicing of their mRNAs using specific RNA structures or additional proteins (rev, in HIV) to ensure that full-length RNA is produced for genomes. These function ONLY for the RNA which which they are intended.

5. Protein translation is even more unlikely: not only do you need similar regulatory elements to transcription, but you also need a true open reading frame. Random RNA sequences will produce only short peptides as the genetic code produces frequent STOP codons if ordered randomly. The absence of things like poly A tails will mean the mRNA is prematurely degraded. The absence of a 5` cap will not allow the ribosome to even bind (unless a IRES is present, such as found in polio virus etc).

So, in summary, while RNA from any source can indeed be inserted into a cell, the odds are that it will be degraded. Without the appropriate sequences (usually only found in exogenous retroviruses) and non-cellular enzymes, the RNA will not be reverse transcribed, or integrated, or transcribed, or translated. It will simply be destroyed.

If the Perth Group want a likelihood-based response to their questions it would be:

Yes, no, no, no, no.

One of the intellectual attractions of retroviruses, to me, was that they were able to do all of this and produce an RNA/lipid/protein structure that could do it all again. And all with only 9 genes...

Competing interests: None declared