Primer extension analysis?

[LOGAN]

I need advice - having spent ages trying to quantify message induction by 
Northern hybridisation and discovering that it is only ~2 fold for a single  
message type, I am now considering Primer extension analysis using a primer that
would identify 2 other closely related messages hoping that their combined 
induction would be more significant and make studies on things that affect the 
level of induction easier. 
Questions are these:
1. How sensitive is the method (relative to Northerns)?
2. Can I separate out unincorporated primers by some other means than a gel 
without introducing another possible source of errors.i.e. binding to something 
like DEAE-cellulose paper, or TCA precipitation?
3. How reliable is quantitation using a primer that might be slightly 
heterologous to the RNA of interest. i.e. I have a probe from one plant, but the
system is easier to induce and study in another closely related one! I realise 
this is a problem for the Northern analysis too, but when examining the single 
message type I have high stringency conditions that do not allow the detection 
of the closely related ones in the same plant. Is it reasonable to quantify when
that would  suggest that the level of homology for this message between the two 
different plant types is very high?
4. How do you standardise (the equivalent of the "constitutive" probe) the RNA 
in the different reactions, there is a lot more rRNA in leaves than roots?

If this all sounds confusing it probably is because I am feeling confused having
tried looking through various technique books. I would now like someone with 
some experience to give me some straight answers! If you can help please e-mail 
me at the address below, if there seem to be other confused people out there I 
will post a summary.

Thanks in advance

Helen 

Logan at montpellier.inra.fr