To identify DUBs involved in the regulation of β-‐catenin at steady state, I
have used the large-‐scale DUB library screen A549 lysate (described in section 3.2.4) for SDS-‐PAGE and Western Blot analysis. 2 96-‐well plates of lysates were run and Figure 4.1 shows a representative blot image acquired for this set of experiments. Across the gels, there were multiple samples with higher levels of β-‐catenin relative
to control, such as the knockdown samples of UCHL3, USP6, USP27X, USP54 and TNFAIP3. On the other hand, significant loss of β-‐catenin following DUBs knockdown
was less obvious as judged by eye.
Densitometric analysis of the blot images was performed to give a quantitative measurement of the β-‐catenin level (Figure 4.2). The DUBs, whose
knockdown resulted in at least a 2-‐fold increase in β-‐catenin level were USP27X,
TNFAIP3, USP54 and USP32. Among these, the effect of USP27X and TNFAIP3 depletion was the most dramatic, resulting in about a 4-‐fold increase in β-‐catenin
level. On the other hand, siRNA depletion of BAP1, USP9Y, DUB4, PSMD7, STAMBPL1 and MPND resulted in at least a 2-‐fold decrease in β-‐catenin level.
Figure 4.1. siRNA DUB library screen to identify DUBs regulating β–catenin in A549 cells. NP40 samples from a large scale A549 DUB screen (Figure 3.15) were resolved on 10% SDS-‐PAGE gels and transferred to nitrocellulose membrane before immunoblotting with antibody against β-‐catenin and tubulin. Blot images were
acquired by infra-‐red scanner (Odyssey, LICOR).
Figure 4.2. Change in β–catenin level following knockdown of DUBs. Densitometric
analysis of blot images (Figure 4.1) was performed using ImageJ to determine relative amount of β–catenin level, normalised to the level of tubulin and median.
The normalised β–catenin level was then log-‐transformed and ranked in ascending
order. Columns exceeding the dotted line represent samples with more than two-‐ fold change in β–catenin protein level.
Figure 4.3. Deconvolution of targets which altered β–catenin level. (A) A549 cells were transfected with single oligos against candidate DUBs at a final concentration of 40nM. Cell were lysed using NP40 lysis buffer 72 hours later. Lysates were resolved by SDS-‐PAGE and immunoblotted for β–catenin and tubulin. Gel images
were acquired by infra-‐red scanner (Odyssey, LICOR). Note: The siControl and knockdown samples of USP27X were run on the same gel, and part of the gel was cropped as indicated by the dotted lines. (B) Densitometric analysis was performed using Image J to determine the level of β–catenin following normalisation to tubulin
level. The normalised β–catenin of each sample was then again normalised to that of
the siControl sample and was log transformed. Data shown represents average of 3 technical repeats, and the 5 columns each represents oligo 1, 2, 3, 4 and pool respectively (See Figure 3.18 for knockdown efficiency of TNFAIP3).
A
From both ends, only the top 3 candidates were chosen for subsequent deconvolution experiments (Figure 4.3). For BAP1, the knockdown using pool oligos reproduced the decrease in β-‐catenin level as was observed in the screen, and oligos
3 and 4 recapitulated the pool knockdown effect. Immunoblotting with BAP1 antibody confirmed that BAP1 was significantly depleted in all knockdown samples. Among the knockdown performed using single oligos, oligo 3 gave the most dramatic depletion in β-‐catenin level. Both the USP9Y and DUB4 knockdown using pool oligos
did not reproduce the decreased β-‐catenin level observed in the screen and the β-‐
catenin level was similar to that of the non-‐targeting siRNA (siControl) sample. The knockdown performed using single oligos against these 2 DUBs also did not result in significant change in β-‐catenin level compared to the siControl sample.
The knockdown of USP54 using pool oligos reproduced the increased β-‐
catenin level that was observed in the screen. However, this effect of USP54 depletion on β-‐catenin level was only recapitulated by oligo 4, but not the other 3
oligos, suggesting the increase in β-‐catenin level was an off-‐target effect. The
knockdown using pool oligos against TNFAIP3 did not reproduce the increased β-‐
catenin level observed in the screen. Among the single oligos against TNFAIP3, only oligo 1 resulted in a slight increase in β-‐catenin level relative to the siControl (Figure
4.3A) while knockdown using oligos 3 and 4 resulted in depletion of β-‐catenin. Since
all 4 oligos resulted in depletion of TNFAIP3 (Figure 3.18) and that the knockdown effect of TNFAIP3 on β-‐catenin level was not consistent, the observed increase in β-‐
catenin level following TNFAIP3 knockdown in the screen was very likely an off-‐ target effect. The knockdown of USP27X using pool oligos again resulted in about a 4-‐fold increase in β-‐catenin level. However, none of the 4 single oligos recapitulated
such an extent of increase: where oligos 1 and 2 resulted in a marginal increase in β-‐
catenin level while oligo 3 and 4 resulted in a marginal decrease in β-‐catenin level.
Knockdown efficiency of USP27X was not checked since the antibody for USP27X was not available in our laboratory. Hence, among the targets that were deconvoluted, only BAP1 emerged as a potential regulator of β-‐catenin.
4.2.2 Characterisation of functional relationship between BAP1 and β–catenin in