My RPA protocol
brett at BORCIM.WUSTL.EDU
Fri Oct 17 10:49:34 EST 1997
Okay, okay. After mentioning the sensitivity of my RPA in this forum, I
have received numerous requests for my RPA protocol, so I am posting it
here. I have used this protocol to measure down to 106 copies of viral
RNAs from an equivalent number of cells. A bit of math suggests this is
only 10-100x more sensitive than Northern analysis. I insist there is
nothing magic here. I based my protocol on the one in Current Protocols
in Molecular Biology and a two-cycle method of Novak and Kirkegaard (J.Virol
65:3384-7). I also found the Ambion web page to be chock full o'advice
(although I eschew their kits). I simply optimized things for my
application. Parameters which were optimized include specific activity
of riboprobes (maximize), probe purity (gel purify!), RNA extraction
procedures (Trizol as good or better than SDS/ProtK), hybridization kinetics
(temp/time), and digestion conditions (choice of RNase, quantity). I
imagine these things should be optimized for sensitivity with any RPA. BTW
(insert self-serving plug here), this work should make the December J.Virol
if anyone is interested. I should mention I was interested in detecting viral
minus and plus strands of an RNA virus, thus the use of a 2-cycle method.
Therefore I used a probe cloned in both orientations into a transcription
vector. This is also useful for generating synthetic RNA standards for
optimization and quantitating the final RPA. Synthetic standards were
"cold" in that they were transcribed with only trace amounts of 3H-UTP
for the purpose of monitoring NTP utilization and yield. They were gel
purified as per below (except they were visualized by EtBr), extracted,
and stored as single-use aliquots at -80C. Since most people will be using
this protocol for mRNAs (and therefore the strand-specificity issue is
moot), I have simplified my actual protocol by expunging the two-cycle
method. In practice I also include an internal control probe for a house-
keeping gene (GAPDH) to control for RNA quantity and quality between
samples (in this case use a lower specific activity probe for the
housekeeping gene in order to give a similar intensity as the rare
transcript). If you use this, it would be nice to hear from you, receive
acknowledgement, etc. Here it is:
1) Hot probe construction:
2ul 5x trxn buffer [200mM Tris-HCl (pH7.9), 30mM MgCl2,
0.25ul 400mM DTT
1.2ul 3NTP (4mM ea. UTP,ATP, GTP)
3ul [alpha-32P] CTP (75pmoles; Amersham PB20382)
1ul linearized template (0.5ug)
0.2ul RNasin (Promega or similar)
0.2ul SP6 or T7 RNAP (Epicentre, 50U/ul)
cook 37C, 1-1.5h
Add 0.5ul RQ1 RNase (Promega), cook 15' 37C
Add 80ul TE, 10ul 3M NaoAc, and 1ul (10ug) carrier yeast tRNA
Spot 2x0.5ul on DE-81 paper
Phenol extract 2x, ppt. with 0.3ml EtOH
Resuspend in 3ul H2O, add 7ul SSB. Denature 95C, 5'; snap cool. Load on 5%
mini-sequencing gel. Run gel 15mA, ~20-30'. Wrap gel in plastic and
write on it with black marker. In dark room, place gel over 5"x7" film
and flash 3x with photo flash. Expose additional 15". Align writing on
film and excise bands. Check activity with Geiger. Add 0.3ml elution buffer
and elute 55C, 1-6h. During this time wash one DE-81 of each transcript 5x5'
in 0.5M Na2HPO4, 1x H2O, 1x EtOH. Air dry. Phenol extract probe from elution
2x and ppt with 10ug carrier tRNA and 1ml EtOH. Resuspend in 50ul 1x hyb
buffer, spot 0.5ul on a new DE-81. Count all samples. Calculate yield and
specific activities (see below). Store probes at -20C until use and use
within 1-2 days!
2) Extract RNA from whole, unwashed monolayer using Trizol, as per Gibco.
To really avoid junk, spin 20', top speed, 4C. Collect all of top phase,
including some interphase. Respin and collect, this time avoiding junk at
bottom. After precipitating Spin hard at RT, wash once with 80%EtOH, and
remove all EtOH using drawn pipette tip (ie gel-loading tip), but DO NOT
3) For quantitative standards, dilute (10-fold, serially) cold synthetic
RNA to 105 molecules/ul in TE + 1ug/ul yeast tRNA. Add 105-108 standard to
cellular RNA lacking message (extracted from an equivalent number of
uninfected cells as above). Precipitate (1ul NaoAc, 20ul EtOH) for >10',
-20C. Pellet as above.
4) Resuspend samples in 30ul 1x hyb buffer containing 5-10 pmoles of hot
probe and vortex well. Heat to 55C for a few minutes, and re-vortex. Add
clamps to tubes to prevent popping/evaporation. Denature at 85C for 5'.
Anneal (12-16h) at 55C.
5) Spin down samples, and add 0.3ml RNaseT2 cocktail (0.5ul of T2
(23U/ul; Boerhinger)/ml of T2 buffer). Digest 1hr, 37C.
6) Add 20ul 10%SDS, 2.5ul ProteinaseK (20mg/ml). Vortex, spin, and cook
30', 37C. Add 0.3ml phenol and 1ul (10ug) of tRNA carrier, vortex. Spin
30". Remove all of upper layer (SDS helps phase separation) and place in
new tube. Add 1ml EtOH and precipitate >10', -20C. Spin 10' and remove
7) Resuspend in 3ul H20, vortex. Add 7ul SSB. Boil and snap cool, load
on your favorite sequencing gel. Relax.
5% Mini seqn gel
(for BioRad mini-Protean or similar): gel elution buffer:
2.4g urea 0.5M NH3oAc
0.5ml 10x TBE 0.1M Tris, pH 8.0
0.625ml 40% (19:1) acrylamide 0.2M NaCl
2.2ml H2O 0.1% SDS
40ul 10% APS 10mM EDTA
1x hyb buffer
0.8ml deionized formamide (formamide + mixed bed resin (BioRad
AG-501-X8), allow to settle. Recrystallize formamide away from any H2O
at 4C. aliquot formamide. Store -20oC).
0.2ml 5x hyb
5x hyb buffer
200mM PIPES, pH 6.4
50ml T2 buffer:
0.83ml 3M NaoAc
1.25ml 4M NaCl
1ml 0.5M EDTA
SSB (sequencing sample buffer):
95% deionized formamide, 20mM EDTA, bromophenol blue and xylene
cyanol to taste.
1) calculate % incorporation (cpm washed/cpm unwashed).
2) calculate yield of RNA:
pmoles synthesized = %incorporation x pmoles input CTP/#C's in transcript
3) calculate [RNA] = pmole RNA/initial rxn volume
4) calculate specific activity = cpm washed (in 0.5ul)/pmoles (in 0.5ul)
It is also good to calculate the theorteical S.A. of your probe based on
label activity, amount of input label, and the maximal yield based on total
incorporation. This is how Ambion does it (my method is the experimentally
determined S.A., perhaps more relevant, although the two should agree well).
For sensitive detection, you need probes => 109 cpm/ug (~105 cpm/fmol).
Washington University - St Louis
brett at borcim.wustl.edu
More information about the Methods