replies to "housekeeping genes"

robert winz winz at
Thu Jan 20 13:32:09 EST 2000

Dear Arabnet,

These are the responses I received as of now to the following query:

"I was wondering what other plant researchers were using for "housekeeping
genes" in their Taqman quantitiative PCR reactions (or Northern blots).
It's my impression that in many instances genes such as GAPDH and actin
aren't suitable to be used as housekeeping genes. This leaves 18S rRNA and
the like. Any further suggestions?"

I use 18S as a reference in TaqMan assays. In my experience, 18S has a
genomic copy number of about 500, and is expressed at extremely high
levels. This leads to a problem: small differences in amplification
efficiency between 2 amplicons lead to greater errors in the final results,
which become greater as the difference between the Ct values of the 2
amplicons increases. Recently, I've done a few simple calculations, which
can then be displayed in an Excel Graph, to show the error threshold for
when e.g. incorrect copy numbers first start to arise. Unfortunately, I
know of no single-copy gene, whose sequence is conserved enough so that one
can find a suitable amplicon that will work for all commonly used plant

We have been using successfully the Ran gene of Arabidopsis to standardize
Northern blots. It has a moderate expression level (good detection in 5 ug
total RNA after overnight exposure), yields a single sharp band and is
reliably reflecting RNA loading. The reference is:
Haizel T, Merkle T, Pay A, Fejes E, Nagy F (1997) Characterization of
proteins that interact with the GTP-bound form of the regulatory GTPase Ran
in Arabidopsis.
Plant J 11(1):93-103

I have been using ACTIN2 in Arabidopsis as an
internal (housekeeping)control in my TaqMan assays. This is one member of
the Arabidopsis actin family that is highly and ubiquitously expressed.
I still have a niggling doubt though whether I should check the expression
of ACTIN2 relative to 18S ribosomal in the various treatments that I use.The
trouble with 18s ribosomal is that it has no introns hence your RNA has to
be DNased completely, especially if your other TaqMan probes are cDNA

I use EtBr staining on Northerns to ensure equal loading. I would highly
discourage the use of rRNA probes for Northerns (as loading ctl) as one
can't possibly label up enough probe to make the RNA on the blot the
limiting factor. I have done tests loading 1x,2x, 4x RNA on blots and
probed with 18S rRNA. The results were unimpressive in that (for ex.) the
1x couldn't be readily distinguished from the 2x.

Some people have success with ubiquitin genes.

There has been a lot of discussion on the plant array newsgroup about which
"housekeeping" genes to use as control for constitutive expression and here
is a list I just received.  From my own experience I have used both the 18S
and Ef1 alpha as controls for loading and expression studies with
Here's a summary of the contribution of Philippe Reymond, who is working
with Arabidopsis.  He has selected a set of candidate control clones that
cover the range of absolute intensities and tested them with 8
independent replicates of a wounding experiment.  The error was quite
modest from test to test.
Ranking the responding from least to most change in ratio of expression
(ignoring the sign of change) are:
Phosphoglycerate kinase
FPS1 (farnesyl pyrophosphase synthase)
Histone H4
beta tubulin
alpha tubulin
TUFA (elongation factor Tu)
PRK (phophoribulokinase)
OEC (Oxygen evolving complex)
Expression ratios are listed  on a log10 scale.
Only OEC had a log10 ratio greater than .5

Robert Winz
Max-Planck-Institut für Chemische Ökologie,
Abteilung Molekulare Ökologie,
Carl-Zeiss-Promenade 10, D-07745 Jena/Thüringen, Germany
FAX (work): -49/0-3641-643653, TEL (work): -49/0-3641-643655,


More information about the Arab-gen mailing list