RT-PCR from a single cell?

Andre Hamel hamel at ccu.umanitoba.ca
Thu Mar 4 11:16:45 EST 1993

In article <1n2nrt$nd2 at rigel.fccc.edu> hardy at mighty.fccc.edu writes:
> Have been thinking recently how nice it would be to detect gene
>expression (for things like cytokines) or even amplify sufficient DNA to
>sequence Ig rearrangements from SINGLE CELLS.  We've done both from
>populations of cells (say 10^4 order), but I wonder if anyone has done or
>has any idea how one might scale this down to the single cell level. 
>Preferably I guess doing all reactions in a single tube.  Perhaps ligated
>PCR?  Anyone has any experience with 3SR?  Both have been mentioned as
>approaches that might work...  Thanks in advance for any advice!
>R. Hardy

I'm currently playing around with FoLT PCR (based upon article appearing
in BIOTECHNIQUES, Vol.14, No.2 (Feb.1993), pages 238- 243
"FoLT PCR: A simple PCR protocol for amplifying DNA directly from whole
blood", Panaccio, M. et.al.

I first tired it on whole blood as in their article, then on clots, then
various tissues (small thin shavings from -70oC frozen tissue done on dry
ice) ... all worked great! Works well with Perkin Elmers Tth pol. (up to
18% formamide {final conc. v/v} and New England Biolab's Vent exo- pol.
{up to 18% formamide but needed 4 mM final conc. MgSO4}. 

Now I've tried on various samples (cultured cells, tissues, blood {whole
and clots} for an RNA virus and lo and behold!  This involved PE's rTth
pol. which has RT activity in presence of MnCl2 ... up to 15% formamide
{final conc. v/v} in 20uL RT reaction (ie., 3 uL of formamide "lysate"
wherein tissue/blot was added to at least 2 volumes formamide {I used 100
uL formamide with up to 50 uL sample ... overlayed with oil, heated 95oC
for 10 min, then 0oC hold ... 0.5- 3 uL of this was added to rTth pol. RT
reaction mixture}). After RT reaction (30oC/1 min then ramp to 50oC over
10 min period, hold 50oC for 3 min then hold 0oC), 80 uL Tth pcr buffer
(called "chelating buffer" by PE ... has EGTA to chelate out the Mn, has
Mg for the pcr reaction), for 40- 50 cycles at 90oC for 15 sec, 45oC for
15 sec, 65oC for 15 sec. 

While I've gotten this to work, compared to standard Chomczynski
(GITC-acid phenol) method of RNA extraction, the RT-PCR efficiency that I've
observed to date for this formamide extraction procedure is noticeably
less. For single cells I strongly suspect this formamide "extraction/
lysis" to be ideal. At present I don't have my hands on any primers that I
could use to further test this method on other RNA's. One potential
problem is that genomic DNA will no doubt be present. But if expression
quantitation (via number of means such as competitive pcr before/after the RT
reaction) and comparison to "benchmark" RNAs such as rRNA, perhaps the DNA
aspect would be less of a factor(?).

I'd be interested to hear from you if you give it a try. So far seems to
be promising (for sure for pcr of DNA, maybe for RNA as well). The virus
I'm studying does not make a DNA copy, the only nucleic acids
associated with it is its full 12.5 kb length positive ss RNA (not even
any shorter transcripts ... thus "background" DNA is at least in theory
not a factor ... even if there were ANY DNA, I'm only ionterested in
detecting this virus' nucleic acid (RNA or DNA).


Andre Hamel
Manitoba Vet. Virol.
Winnipeg, Manitoba

email: hamel at ccu.umanitoba.ca

More information about the Methods mailing list