100 5 0 - 2 5 - ■ % nu H % n u /c y t □ % c y t
■B
REV(1.4)-GFP % 100 7 5 - 5 0 - 2 5 - nu □ “inu /cyt □ %cyt REV(1.4>-GFP HT, REV(1.4>-GFP ActD NT ActD LM B Figure 28Nuclear export assay (I) Controls.
Cells transfected with Rev-GFP (A and B) or Rev(1 4)-GFP (C and D) were treated with cycloheximide at 15 ^ig/ml for 3 hours, either untreated (NT), in the presence Actinomycin D (5 iig/ml) (ActD) or LMB (10 ng/ml). (A and C) Quantification of the cellular localisation of each construct after drug treatment. Each experiment was done at least three times and
approximately 500 cells were counted for each condition. Error bars indicate standard deviations (B and D) Representative images of the cells under each indicated treatment.
Treatm ent of cells w ith Leptomycin B (LMB) causes a dram atic
change in the distribution of Rev-GFP, and 3 hours after treatm ent, Rev-
GFP is exclusively nuclear in the majority of the cells. Representative
images of cells expressing Rev-GFP after each treatm ent are show n in
Figure 28b. The drugs do not affect the localisation of Rev(1.4)-GFP
plasmid, which rem ains mainly nuclear in the transfected cells (Figure 28 c
and d).
W hen NESl was tested in the nuclear export assay (Rev-a-NESl) it
was exclusively nuclear in 35% of the cells (Figure 29a), in the rem aining
cells, it was found both in the nucleus and cytoplasm. T reatm ent w ith
Actinomycin D resulted in an increase in the num ber of cells w here Rev-a-
NESl was exclusively cytoplasmic. This suggests that a-N ESl has w eak
export activity. Consistent w ith this, LMB d id not affect the localisation of
Rev-a-NESl, which rem ained predom inantly nuclear. Representative
images of ceUs expressing Rev-a-NESl after each treatm ent are show n in
Figure 29b.
W hen NES2 was tested is the nuclear export assay (Rev-a-NES2), its
distribution was similar to that of Rev-GFP. In untreated cells, Rev-a-NES2
was found either both in the nucleus and cytoplasm (-75% of cells) or
exclusively in the cytoplasm (Figure 29c). Treatm ent w ith Actinomycin D
resulted in a shift to the cytoplasm in m any of the cells. In addition,
treatm ent w ith LMB resulted in a shift of Rev-a-NES2 to the nucleus in the
majority of the cells (Figure 29c). These results indicate that the a-NES2 is
a strong NES. Images of cells expressing Rev-a-NES2 after each treatm ent
are show n in Figure 29d.
In order to confirm that a-NES2 is a canonical NES, I generated
residues in the predicted NES, have been m utated to alanine. These
changes completely abrogated nuclear export activity NES2 (Figure 30a).
Moreover, treatm ent w ith Actinomycin D or LMB did n o t affect the
localisation of Rev-a-NES2mut. Indeed, Rev-a-NES2mut behaved
similarly to Rev(1.4)-GFP, which contains a non-functional NES. Images of
cells expressing Rev-a-NES2mut after each treatm ent are show n in Figure
30b.
The relative nuclear export activities of various NES from different
proteins have been m easured by Flenderson and Eleftheriou (2000). In
the scoring system described by these authors, the NES present in HIV-
Rev protein receives a score of +7 (where the maximal score is +9). The
score is based on the percentages of ceUs containing the Rev protein
exclusively in the nucleus or in the nucleus an d cytoplasm, and on the
changes in these percentages after treatm ent w ith Actinomycin D. W hen I
com pared the export activities of a-NESl, a-NES2 and Rev-NES in my
experiments, a-NESl scores +3, while a-NES2 scores +7. As expected, Rev-
a-NES2mut, similar to Rev(1.4)-GFP, scores 0.
V.3- Effects o f LMB on the localisation o f endogenous and exogenous a- catenin
The results presented above suggest that at least one of the NES
sequences found in a-catenin is a strong NES w hen assayed in the context
of Rev-GFP. To investigate the possibility th at a-catenin localisation is
regulated by nuclear export, I examined the effect of LMB on the
localisation of endogenous a-catenin in colon cancer cells.
First, DLD-1 cells were either untreated or treated w ith 10 n g /m l
LMB for 16 hours, fixed and then stained for endogenous a-catenin and |3-
levels of either a-catenin (compare Figures 31a and 31c) or p-catenin
(compare Figures 31b and 31d) after LMB treatm ent. As a control, parallel
DLD-1 cultures were stained for NFkB p65. NFKB/LcBa complexes shuttle
betw een the cytoplasm and the nucleus owing to the presence of both a
NLS and a CR M -l-dependent NES (Fluang et al., 2000). A ddition of LMB
caused a complete shift of NFkB from the cytoplasm (Figure 31 e) to the
nucleus (Figure 31f), indicating that DLD-1 cells do respond to LMB. The
lack of effect of LMB on a-catenin and p-catenin localisation may be
because these proteins are already nuclear in these cells, m aking it difficult
to detect an increase in nuclear localisation.
As show n in Chapter IV, HCT116 cells do not norm ally contain
endogenous nuclear a-catenin (Figure 17d). Therefore I examined the
effects of LMB on endogenous catenin localisation in HCT116 cells (Figure
32). How ever, LMB did not affect the localisation of a-catenin (compare
Figures 32a and 32c), or p-catenin (compare Figures 32b and 32d) in
HCT116 cells. This suggests that there are processes, other than nuclear
export, controlling the localisation of the catenins in these cells. One
possibility is that the catenins are sequestered either in the cytoplasm by
B