Biomedical Engineering Reference
In-Depth Information
to confi rm that the increase in lectin staining is indeed attributable
to the suspected glycosylation pathway. For example, knockdown
of signaling genes such as ERK8 can result in a dramatic relocation
of GALNTs from the Golgi to the endoplasmic reticulum (ER)
[
26
,
27
]. This relocation induces a marked increase in cellular Tn
levels [
27
,
28
]. A knockdown of GALNT1 and 2 together with
ERK8 abrogates the dramatic increase in Tn staining observed
with ERK8 depletion alone, confi rming that the ERK8-dependent
signal is due to GALNT activity (Fig.
1a
).
It should be noted that as in all RNAi screens, hit validation to
discard or reduce off-target results can involve various previously
described methods, including using individual deconvoluted
siRNA sequences to the gene, using a completely different siRNA
library with different siRNA sequences, using shRNA depletion, or
using overexpression of candidate hit genes [
29
]. Whatever the
reagents chosen, the experimental and analysis methods are gener-
ally similar to the primary screen. We have thus focused here on
validation of glycophenotypes as detected by lectin or antibody
staining. Another consideration particular to RNAi screening in
general is that knockdown penetrance and effi ciency varies with
siRNA sequence and many experimental factors; thus there will
likely always be a fraction of false-negatives present in a screen.
To summarize, we provide a simple but robust and informative
method using high-throughput RNAi downregulation of gene
activities in 384-well plates to systematically and quantitatively
investigate glycosylation using fl uorescent lectin staining (Fig.
2
).
We begin with 384-well plates preprinted with siRNA libraries.
Transfection complexes are made by adding transfection reagent to
these siRNAs, after which the cells to be transfected are added and
assayed 3-5 days later. In this case, the cells are fi xed, stained, and
imaged. An automated HCS imaging microscope is employed to
ensure good and consistent image quality. We also provide a work-
fl ow for downstream analysis of the image data generated (Fig.
3
).
For secondary validation, we suggest a double knockdown of each
candidate hit gene with the relevant glycosylation enzyme neces-
sary for synthesis of the glycan structure being probed. This proto-
col can be adapted for other non-lectin screening projects. As is the
case for a major experimental campaign, there are many variations
possible on both the experimental and analysis methods. A simple
and typical workfl ow is suggested in our protocol, with some varia-
tions mentioned in the Notes; others can further be explored in
the References. Downstream analysis of the hits usually involves
bioinformatics (e.g., protein-protein interactions, gene ontology
analysis) to make conclusions or form further questions that are
biologically meaningful. This is beyond the scope of this chapter
and will not be covered here.
