RPA or Northerns????

Richard Rohan rrohan at world.std.com
Thu Sep 14 21:21:25 EST 1995


I also have used Northerns and RPA and find RPA to be more sensitive. High
specific activity riboprobes (800 Ci/mmole) are as hot as random primed DNA
probes and since probes are single-stranded, the hybridization can be driven
to saturation.  Variability in RPA is reduced if you use an internal standard
(I use 18S or ribosomal protein L19 RNA).  If you plan ahead, probe and
protected fragment lengths can be adjusted to allow multiplexing.  Thus
multiple transcripts and an internal standard can be assayed simultaneously.
One distinct disadvantage, and this may be a problem for Melissa, is that RPA
requires 100% identity between probe and target.  In order to assay
transcripts from cotton, you will need to use regions of cotton cDNAs that are
not polymorphic between strains.

My usual modus operandi is to just do RPA.

Good luck
Rich Rohan  

ikekim at merle.acns.nwu.edu (Isaac Y Kim) wrote:

>In article <mkhill-140995141518 at mg-a12-1-81.bio.usyd.edu.au>, 
>mkhill at bio.usyd.edu.au (Melissa Hill) says:
>>
>>
>>I have been planning to compare the expression of particular genes in a
>>number of different varieties of cotton, and was wondering what peoples
>>opinions and experiences were of RNase Protection Assays as opposed to
>>Northern analysis.
>>
>>Mel
>
>I have used both Northern and RPA in the past. In my experience, RPA 
>appears to be easier to use and quantitate. It has also been claimed that 
>RPA has higher sensitivity than Northern. With this claim, though, my 
>experience does not agree all the time. I think, in some cases, RPA has 
>higher sensitivity than Northern because RPA can still detect partially 
>degraded messages. Regarding the increased sensitivity of RPA, I will 
>appreciate any other responses. With Northern, the obvious advantage is 
>that the blot can be reprobed multiple times and that you can tell the 
>size of the native message. My usual pattern of work is that I start with 
>Northern, then try RPA. Good luck.
>
>Isaac.





More information about the Methods mailing list