Reverse transcriptases

[Patrick O'Neil]

On 24 May 1995, Bruce Phillips wrote:

> 
> 	The error-proneness of lentivirus RTs are clearly documented.  Does
> anyone with expertise in something other than guns know if there are 
> experimental data concerning the error-proneness of oncovirus RTs?  While
> this isn't my field of expertise, I don't recall ever hearing about RTs
> lack of fidelity until HIV came along.

All RTs from all retroviruses tend to be highly error-prone.  The lab I 
am in works with HIV RT and Murine Leukemia Virus RT.  They both 
demonstrate about the same error rate of about 10^-5 per base per 
replication cycle.

The actual focus of the lab in regards to both is error rate.  We are 
looking at the difference between error rates in vitro vs that seen in 
vivo.  With MLV, the in vivo error rate is about 30-fold lower than 
predicted by in vitro assay, implying either other viral encoded or host 
cell supplied proteins that increase fidelity somewhat.  Nevertheless, 
the in vivo error rate of 3 x 10^-5 mutations per base-pair per 
replication is about the norm for others:  Rous Sarcoma Virus 
demonstrates an error rate in vivo of about 1.4 x 10^-5 per base-pair per 
replication.  Spleen Necrosis Virus has an error rate of 0.7 x 10^-5.  In 
general, the error rates for RTs from any sampling of retroviruses range 
from 2.5 x 10^-5 to 6 x 10^-5.  Basically, you can count on about 0.5 
base substitutions per retroviral genome per replication.  These rates do 
not include other errors such as deletions or insertions due to slippage 
on repetitive sequences.

Patrick