Biology Reference
In-Depth Information
In shotgun techniques, the undersampling rules are also
known. However, they apply to peptides, and how they translate in
terms of protein identifi cation and even more in terms of protein
quantifi cation is complex, and linked to the complexity of the pro-
tein inference problem [
7
-
9
].
On top of this strong undersampling problem comes the often
unrecognized fact that 2D gel-based proteomics is a low-yield pro-
cess [
10
,
11
] and the fact that plant samples are among the most
diffi cult samples to prepare for an analysis by 2D gels [
12
], due to
the variety and amounts of interfering, nonprotein compounds
present in many plant cell types and tissues.
All these negative sides explain the trend from 2D gel-based
proteomics to shotgun proteomics, as exemplifi ed by the study of
cadmium response in plant; gel-based in 2006 [
13
] and shotgun-
based in 2011 [
14
].
However, 2D gel-based proteomics is still very popular in plant
proteomics, as shown by recent publications in the fi eld (e.g. [
15
,
16
], but among more than 150 publications in 2011). There are
several built-in reasons for this sustained interest in this technique,
which are named reproducibility, robustness, effi ciency, ability to
analyze intact proteins, adequacy to study posttranslational modi-
fi cations, easy interface with powerful biochemical techniques.
These different points will be developed below.
2
Reproducibility and Robustness
It is not an overstatement to write that 2D gel-based proteomics is
still, and by far, the most reproducible and robust proteomic setup
[
17
]. A further proof of this statement lies in the much higher
expectations from proteomic journals, in terms of sample numbers
and experimental power, applied to 2D gel-based proteomics com-
pared to shotgun proteomics [
18
,
19
]. Still today, what is common
practice in 2D gel-based proteomics, i.e. quantitative comparison
of two or more biological conditions with several individual bio-
logical replicates for each condition, is still rarely found in publica-
tions using shotgun proteomics.
This reproducibility is further testifi ed by the interlaboratory
reproducibility of 2D maps [
20
,
21
], while recent test of shotgun
techniques on the same sample (yeast) revealed a much higher
standard deviation between laboratories [
22
,
23
].
Ironically enough, the high robustness of 2D gel-based pro-
teomics comes in part from the terrible reproducibility of isoelec-
tric focusing in its early days. Put bluntly, only the gels run
simultaneously were comparable. This led the proteomic pioneers
to develop devices to run several gels in parallel [
24
,
25
]. When
immobilized pH gradients were fi nally introduced in 2D gel
electrophoresis with solid protocols [
26
], the reproducibility and
