I have samples of about 100ul EDTA-treated mouse plasma stored at -80 and would like to analyze the differential miRNA content between treated and untreated mice. Probably on affymetrix chips unless there is a better/more current way.
If you are interested in extracellular/circulating miRNAs my group has recently developed a database (http://atlas.dmi.unict.it/mirandola/index.html) submitted for publication and you will find, in the results table, methods to isolate miRNA and experiment description!
I use 200 ul plasma when i isolate small plasma RNA´s from humans (500 ul in total but i transfer 200 ul of the upper phase, to lysis reagent, to remove plasma debris) . Even though i use a RNA carrier my yields are below detection range on a nanodrop. I have never tried to isolate from plasma in mice and i therefore cant tell if the yields will be higher? My guess is that it is around the same. I dont think the microarray approach is sensitive enough for the lower "expresed" miRNA´s. I have used PCR arrays with good results. I can recommend this paper:
I also used phenol-chloroform and glycogen. A good protocol is written by Andrey (check his MM: doi:10.1093/nar/gkr254). If you do LDAs from ABi then I would rather say to median normalize your data. Good luck!
I used to work on reticulocyte mRNA from erythrocyte fraction. Healthy human blood contains about 2% of reticulocyte in blood. I used QIAGEN mini kit and treat material as culture cells with additional on-column digestion. Maybe you should try.