Help with cDNA

MUNGO at UVVM.UVIC.CA MUNGO at UVVM.UVIC.CA
Tue Feb 22 13:05:37 EST 1994


I am looking for some help, opinions, or comments on a number of issues
relating to RNA isolation and the construction of cDNA libraries.
1) I wish to isolate polyA+ RNA from a large sample of small (but not
precious) tissue samples for RTPCR.  I would like a method whereby I cause an
oligo dT matrix to select straight out of the guanidinium lysis buffer. I
would also like to wash and elute from the matrix using a small spin column
(similar to InVitrogens micro fast track kit).  Does anyone have suggestions
or better yet detailed protocols.  Any suggestions on the oligo dT matrix ?
I have tried a number of oligo dT cellulose matrices in the spin columns but
they tend to plug (I have removed the fines from the matrix).

2) I wish to make some random primed cDNA libraries that have a GOOD
representation of 5' ends (within 150-200 bp of start) of long (7-8 Kb)
transcripts.  We are presently using a library that is oligo dT primed yet
surprisingly contains some, but very few, 5' sequence.  There are a number of
kits available, are they worth it if I make only a few libraries ?  Any
recommendations ?  The same with vectors.  I like gt 10 as it is easy and works
very well.  However some of the "newer" vectors are much easier in allowing
in vivo excision of plasmids.  The problems I have had are that most clones are
in puc 18 derivatives (as are the plasmids in these vectors) and that when I
make probes from gel purified inserts I always get some contamination with
vector and the screening gives high background.  This may be a unique problem
related to my abilities...Any suggestions appreciated.

3) Finally does anyone have a procedure for using PCR to make high SA P32
labelled probes that can be used to screen libraries ?  We presently use a
random primed method but the DNA we label is short (150-200 bp) and the

logistics of cutting large quantities of plasmid to get small amounts of insert
that then label poorly is annoying.  If we could label by PCR it would give a
lot of probe easily
Mungo



More information about the Methods mailing list