Destabilization of IL-8 mRNA by Anthrax Lethal Toxin ... · elements confer destabilization of transcripts by binding to AU-binding proteins (AUBPs) that ... and is more common in
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Destabilization of IL-8 mRNA by Anthrax Lethal Toxin:
Demonstration of the Requirement for TTP and Examination of its Cellular Interactions
by
M.C. Edith Chow
A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy
Laboratory Medicine and Pathobiology University of Toronto
Contributions: I performed all the experiments in this chapter with the exception of Fig. 3.3B, which was performed by Sarah Batty.
3.1 Summary
Anthrax lethal toxin (LeTx) is a bipartite toxin composed of protective antigen (PA) and
lethal factor (LF) – PA is the receptor-binding moiety and LF is the enzymatic moiety that
specifically cleaves and inactivates mitogen-activated protein kinase kinases (MAPKKs). LeTx
subverts the immune response to B. anthracis in a variety of ways, such as downregulating
interleukin-8 (IL-8) by increasing the rate of IL-8 mRNA degradation. Many transcripts are
regulated through cis-acting elements that bind proteins that either impede or promote
degradation. Some of these RNA binding proteins are regulated by MAPKs and previous work
has demonstrated that interfering with MAPK signaling decreases the half-life of IL-8 mRNA.
Here, a segment within the IL-8 3’ untranslated region responsible for LeTx-induced transcript
destabilization is localized and it is shown that this is caused by inhibition of the p38, ERK, and
JNK pathways. TTP, an RNA binding protein involved in IL-8 mRNA decay, became
hypophosphorylated in LeTx-treated cells and exhibited increased localization to Processing-
bodies, which are structures that accumulate transcripts targeted for degradation. Furthermore, it
42
is observed that LeTx promoted the formation of Processing-bodies, revealing a link between the
toxin and a major mRNA decay pathway.
3.2 Results
3.2.1 LeTx accelerates IL-8 mRNA decay through the 3’ UTR
The effect of LeTx on IL-8 mRNA stability in HT1080 fibroblasts was assessed in cells
pretreated with TNF-α to increase the level of endogenous IL-8 mRNA, and then with
actinomycin D, to halt de novo mRNA synthesis. Total RNA was extracted at various times and
IL-8 transcript levels were quantified using quantitative real-time PCR (qPCR) and standardized
to β-actin mRNA levels. The half-life (t1/2) of IL-8 mRNA in LeTx-treated cells was ~52 min
compared to ~140 min in unintoxicated cells (Fig. 3.1A), indicating that LeTx destabilizes IL-8
mRNA in this human fibroblast cell line.
While it was demonstrated previously that the 3’ UTR of the IL-8 transcript confers
LeTx-dependent destabilization to a reporter transcript (Batty et al., 2006), it has been shown
that stability of other transcripts can be influenced by regions in the 5’ UTR and coding region
and by whether or not the transcript has undergone splicing (Zhao and Hamilton, 2007; Chen et
al., 2000). Therefore, to determine if regions outside of the 3’UTR affect LeTx-mediated
transcript destabilization, the IL-8 genomic sequence, the cDNA sequence, and the 3’ UTR was
cloned into reporter constructs containing tags downstream of the transcription start site that
allowed for detection by qPCR. The reporter constructs were transiently transfected into
HT1080 cells for 2 h and the cells were either left untreated or were treated with LeTx for 24 h.
RNA was isolated from untreated and toxin-treated cells and the ratios of the levels of reporter
transcripts were calculated (Fig. 3.1B). The levels of the transcripts containing IL-8 sequences
were reduced by LeTx to similar extents, suggesting that the element(s) responsible for toxin-
43
mediated destabilization are confined to the 3’ UTR and that splicing does not affect the stability
of the transcript.
Figure 3.1: LeTx accelerates IL-8 mRNA decay through the 3’ UTR.
A. Endogenous IL-8 mRNA was induced by incubating HT1080 cells for 2 h with TNF-α (10 ng/ml), followed by treatment with LeTx (10-8 M PA and 10-9 M LF) for 1 h. Transcription was inhibited by addition of actinomycin D (1 μg/ml) and total RNA was isolated at the indicated times and transcript levels was measured using qPCR. Error bars indicate SEM of 3 independent experiments. B. HT1080 cells were transiently transfected with the indicated plasmids containing IL-8 sequences and treated with LeTx (10-8 M PA and 10-9 M LF). Total RNA was isolated from untreated and intoxicated cells and reporter transcript levels were measured using qPCR. Error bars indicate SEM of 3 independent experiments.
3.2.2 IL-8 3’UTR contains AU-rich element that confers mRNA instability
In order to identify cis-acting elements that destabilize IL-8 mRNA, various truncations
of the IL-8 3’ UTR were cloned behind the EYFP coding region (Fig. 3.2A). The constructs
44
were transiently transfected into HT1080 cells and reporter mRNA was quantified using qPCR
and standardized to β-actin mRNA. mRNA containing the entire IL-8 3’ UTR (EYFP-IL-8401-
1648) was detected at an ~10-fold lower level than that of the EYFP control mRNA, confirming
the presence of destabilization elements in the 3’ UTR. A segment comprising nucleotides 1000-
1100 (EYFP-IL-81000-1100), containing 4 clustered AUUUA pentamers, was found to exert
significant destabilization to the reporter transcript, whereas other ARE-containing regions did
not confer significant destabilization to the EYFP coding region.
To test if EYFP-IL-81000-1100 mRNA is responsive to LeTx-mediated destabilization,
stable transfectants were treated with toxin and half-lives were assessed after the addition of
actinomycin D (Fig. 3.2C). EYFP mRNA that lacks IL-8 sequences had a half-life of more than
200 min and LeTx had no effect on the stability of this transcript. Transcripts containing the
EYFP coding region fused to the IL-8 3’ UTR (EYFP-IL-8401-1648) had a half-life of ~130 min in
untreated cells, and ~95 min in LeTx-treated cells. Similarly, transcripts containing the EYFP
coding region fused to nucleotides 1000-1100 of the IL-8 3’ UTR had a half-life of ~74 min in
untreated cells, and ~42 min in LeTx-treated cells. These results indicate that LeTx causes
destabilization of IL-8 mRNA through nucleotides 1000-1100 within the 3’ UTR.
45
Figure 3.2: Identification of a region within the IL-8 3’UTR that confers mRNA instability. A. Scheme depicting the reporter genes containing various truncations of the IL-8 3’UTR. Vertical lines indicate the positions of AUUUA sequences. The thicker vertical lines indicate two adjacent AUUUA sequences. B. IL-8 3’UTR constructs were transiently expressed in HT1080 cells. Total RNA was isolated and reporter mRNA levels were measured using qPCR. Error bars indicate SEM of 3 independent experiments and asterisks indicate significant differences (p < 0.05). C. Cells stably expressing EYFP, EYFP-IL-8401-1648, and EYFP-IL-81000-1100 were treated with LeTx (10-8 M PA and 10-9 M LF) for 1 h followed by addition of actinomycin D (1 μg/ml). RNA was then isolated at the indicated times and measured using qPCR. Error bars indicate SEM of three independent experiments.
46
3.2.3 Inhibition of ERK1/2, p38, and JNK MAPK pathways are required for IL-8 mRNA
destabilization
LeTx downregulates the p38, ERK, and JNK MAPK pathways and it was previously
reported that disruption of each MAPK pathway using pharmacological inhibitors destabilizes
IL-8 mRNA in HUVECs (Batty et al., 2006). Here, the relative importance of the MAPK
pathways on IL-8 mRNA stability in fibroblasts was assessed. HT1080 cells were treated with
actinomycin D and either SB202190 (p38 inhibitor), SP600125 (JNK inhibitor), U0126
(MEK1/2 inhibitor), or a combination of all three inhibitors. There was no appreciable
difference between the decay rates of IL-8 mRNA in cells treated with any one of the inhibitors
individually compared to mock treatment, whereas co-treatment with all three inhibitors
destabilized IL-8 mRNA by more than two-fold (Fig. 3.3A).
Pair-wise combinations of the three inhibitors were tested and it was found that each
of the JNK pathway, in combination with either the ERK or the p38 pathways, was least potent
at destabilizing IL-8 mRNA, accelerating the decay by ~1.6 and ~1.7-fold respectively compared
to mock treatment. The combination of inhibiting the p38 and ERK pathway caused a ~2.3 fold
change, nearly as much as inhibiting all three inhibitors together, which caused a ~2.8-fold
change. These results suggest that LeTx-mediated destabilization of IL-8 mRNA is largely due
to inhibition of the p38 and ERK pathways.
47
Figure 3.3: Decay analysis of IL-8 mRNA in response to pharmacological inhibitors. A. HT1080 cells were treated with 1 μg/ml actinomycin D in combination with either DMSO, 10 μM SB202190, 20 μM SP600125, 10 μM U0126, or all three inhibitors. Total RNA was isolated at the indicated times and IL-8 transcript levels were assessed by qPCR. Error bars indicate SEM of three independent experiments. B. HT1080 cells were treated with 1 μg/ml actinomycin D and either DMSO, pair-wise combinations of inhibitors, or all three inhibitors as indicated. Total RNA was isolated at the indicated times and IL-8 transcript levels assessed by qPCR. Error bars indicate SEM of three independent experiments.
3.2.4 TTP is required for LeTx-mediated IL-8 destabilization
Since TTP, TIAR, and KSRP have been demonstrated previously to bind IL-8 mRNA in
vitro (Winzen et al., 2007; Suswam et al., 2005a; Suswam et al., 2005b), the effect of
48
overexpression of these AUBPs on the level of IL-8 mRNA in HT1080 cells was assessed.
FLAG-tagged forms of these proteins were over-expressed and IL-8 mRNA levels were
quantified by qPCR. IL-8 mRNA expression was lowered in cells over-expressing TTP or
TIAR, but not KSRP (Fig. 3.4A).
To ascertain whether TTP or TIAR is required for LeTx-mediated IL-8 transcript
destabilization, RNA interference was used to downregulate TTP and TIAR levels. The TTP
protein level in cells transfected with TTP siRNA was reduced to ~7% of that detected in cells
transfected with negative control siRNA (Fig. 3.4C). Knock-down of TTP increased the stability
of IL-8 mRNA and this stability was not diminished by LeTx treatment (Fig. 3.4D). In contrast,
knock-down of TIAR to ~10% of the control level increased the half-life of IL-8 mRNA from
~81 min to ~116 min, but did not prevent LeTx treatment from increasing the decay rate by 1.6-
fold (Fig. 3.4E and F). These results indicate that TTP, but not TIAR, mediates LeTx-stimulated
IL-8 mRNA decay.
49
50
Fig. 3.4: Involvement of AUBPs in IL-8 mRNA stability. A. Indicated AUBPs were overexpressed in HT1080 cells and total RNA was isolated. Endogenous IL-8 mRNA was measured and normalized to β-actin mRNA levels. Error bars indicate SEM of three independent experiments and the asterisk indicates significant difference (P < 0.05). B. Cytoplasmic extracts were prepared from cells transfected as in (A) and AUBPs were detected by the FLAG-tag. β-Actin protein levels were measured as a loading control. The blot is representative of three independent experiments. C. Extracts from cells transfected with negative control siRNA or siRNA directed against TTP were prepared and immunoblotted for TTP. β-Actin expression was used as loading control. The blot is representative of three independent experiments. D. HT1080 cells from (C) were treated with 1 mg ml-1 actinomycin D in the absence or presence of LeTx. Total RNA was isolated at the indicated times and IL-8 transcript levels were assessed by qPCR. Error bars indicate SEM of three independent experiments. E. Extracts from cells transfected with negative control siRNA or siRNA directed against TIAR were prepared and immunoblotted for TIAR. β-Actin expression was used as loading control. The blot is representative of three independent experiments. F. HT1080 cells from (E) were treated with 1 mg ml-1 actinomycin D in the absence or presence of LeTx. Total RNA was isolated at the indicated times and IL-8 transcript levels were assessed by qPCR. Error bars indicate SEM of three independent experiments. G. HT1080 cells were left untreated or were treated with LeTx for 2 h. Cytoplasmic and nuclear proteins were isolated and equivalent amounts of extract were subjected to Western blotting using antibodies against TTP, TIAR, tubulin and p53. A sample of untreated cytoplasmic extract was treated with lambda protein phosphatase (ppase). The blot is representative of three independent experiments. H. HT1080 cells were treated with LeTx or with indicated pharmacological inhibitors for 2 h. Cytoplasmic proteins were extracted and probed for TTP and tubulin. The blot is representative of seven independent experiments.
3.2.5 Treatment of LeTx or MAPK inhibitors leads to dephosphorylation of TTP
The possibility that LeTx alters the endogenous expression level or localization of TTP
and TIAR was also addressed. Cytoplasmic and nuclear fractions were prepared and probed for
tubulin (a cytoplasmic marker) and p53 (a nuclear marker) (Fig. 3.4G). Most of the TIAR
protein was found in the nuclear fraction in untreated cells (compare lane 1 and 3); toxin
treatment did not alter its expression level or localization. The doublet observed likely
represents two isoforms that resulted from alternative splicing (Taupin et al., 1995).
51
TTP was found to be localized predominantly to the cytoplasm in both untreated and
intoxicated cells (Fig. 3.4G). Multiple bands corresponding to TTP were observed in both cell
lysates, but the uppermost bands were not apparent in lysates prepared from intoxicated cells and
the lower band was more prominent (Fig. 3.4G, compare lanes 1 and 2). That these bands
represented differentially phosphorylated forms of TTP was shown by incubating untreated cell
lysates with Lambda protein phosphatase (lane 5), which led to the loss of the upper forms and
an increase in the amount of the lower form. Thus, LeTx activity induces the dephosphorylation
of TTP.
Pharmacological inhibitors were used to determine which of the MAPK pathways
contribute to TTP phosphorylation. Cells were treated for 2 h with LeTx or with
pharmacological inhibitors before cytoplasmic proteins were prepared and subjected to Western
blotting (Fig. 3.4H). More of the hypophosphorylated (lower) form of TTP was detected in
lysates from cells treated with either LeTx or the MEK1/2 inhibitor (U0126) alone compared
with lysates from untreated cells (compare lane 1 with lanes 2 and 4). Only a slight, yet
reproducible decrease in the amount of the upper TTP band was observed upon treatment with
the p38 or JNK inhibitors alone (compare lane 1 with lanes 3 and 5). Combinations of the
inhibitors did not substantially increase the levels of hypophosphorylated TTP compared with the
MEK1/2 inhibitor treatment, but the upper bands representing the hyperphosphorylated forms of
TTP were clearly less prominent. Together, these results suggest that each of the three MAPK
pathways contributes to the phosphorylation of TTP.
3.2.6 Increase of visible P-bodies in cells treated with LeTx
Processing bodies are dynamic cytoplasmic loci that are enriched in enzymes involved in
5′ to 3′ mRNA decay. Previous studies have shown that TTP can target ARE containing
52
transcripts to P-bodies (Fenger-Gron et al., 2005; Kedersha et al., 2005). This finding, together
with the data demonstrating that overexpression of TTP lowered IL-8 mRNA levels and that
LeTx activity caused dephosphorylation of TTP, led us to investigate the effects of LeTx on P-
body formation and TTP localization. The effects of LeTx on the assembly of P-bodies were
examined by subjecting untreated or intoxicated HT1080 cells to immunofluorescence analysis.
The helicase DDX6 (p54/RCK) was used as a marker to identify P-bodies (red) and Hoescht dye
was used for nuclear staining. The absence (Fig. 3.5A) or the presence (Fig. 3.5B) of P-bodies
can easily be distinguished using this marker. P-bodies were observed in ~26% of untreated cells
and ~52% of intoxicated cells (Fig. 3.5C). Similarly, cells treated with a combination of
pharmacological inhibitors against p38, MEK1/2 and JNK also exhibited a significant increase in
P-body formation (data not shown). These foci corresponded to P-bodies and not stress granules
as they did not colocalize with the stress granule marker TIA-1 (data not shown).
53
54
Fig. 3.5: Effect of LeTx on P-body formation and TTP localization. A–C. HT1080 cells were left untreated or treated with LeTx for 1 h. DDX6 was used to visualize formation of P-bodies. Representative immunofluorescence micrographs of HT1080 cells exhibiting diffuse (A) or punctate (B) staining of endogenous DDX6 (red). Hoechst dye (blue) was used for nuclear staining. A P-body at higher magnification is shown in the insert. The fraction of cells exhibiting P-bodies from untreated or intoxicated cells was quantified from a minimum of 100 cells per sample (C). Values indicate mean and SEM of three independent experiments. The asterisk indicates significant difference (P < 0.05). D–J. HT1080 cells transiently transfected with FLAG-TTP were left untreated or were treated with LeTx for 1 h. Representative immunofluorescence micrographs show cells exhibiting diffuse (D) or punctate (G) staining of FLAG-TTP. Localization of FLAG-TTP to P-bodies was examined by co-staining with DDX6 (E and H). Merged images are shown (F and I). The fraction of cells exhibiting punctate FLAG-TTP staining was quantified (J). Values indicate mean and SEM of three independent experiments. The asterisk indicates significant difference (P < 0.05).
3.2.7 Increased localization of TTP to P-bodies in cells treated with LeTx
TTP localization was compared between untreated and toxin-treated HT1080 cells. Since
endogenous TTP could not be visualized in these cells by immunohistochemistry, FLAG-tagged
TTP was transiently transfected and cells were stained with anti-FLAG and anti-DDX6
antibodies (Fig. 3.5D–I). In some cells, FLAG-TTP was diffusely distributed in the cytoplasm
(Fig. 5D), whereas in others it concentrated at cytoplasmic foci that always colocalized with the
P-body marker DDX6 (Fig. 5H and I). FLAG-TTP colocalized with P-bodies in ~3% of
untreated cells and in ~15% of LeTx-treated cells (Fig. 3.5J).
Next, the effect of LeTx on TTP localization was examined in HeLa cells, which
constitutively display P-bodies and are used by numerous groups to study mRNA decay (Franks
and Lykke-Andersen, 2007; Fenger-Gron et al., 2005; Stoecklin et al., 2004). HeLa cells
were transfected with FLAG-TTP and stained for FLAG and DDX6. In contrast to HT1080 cells,
P-bodies were visible in almost all of the HeLa cells, and treatment with LeTx did not affect P-
body size or number in these cells. Diffuse staining of FLAG-TTP (Fig. 3.6A) that does not
colocalize with P-bodies (Fig. 3.6B and C) is largely seen in untreated cells. A majority of cells
55
treated with LeTx exhibited punctate staining of FLAG-TTP in the cytoplasm (Fig. 3.6D) that
colocalized to P-bodies (Fig. 3.6E and F). FLAG-TTP accumulated at P-bodies in ~22% of the
untreated cells and in ~54% of the toxin-treated cells (Fig. 3.6G). These data suggest therefore
that LeTx causes TTP to be recruited to P-bodies in both HT1080 and HeLa cells.
Fig. 3.6: LeTx increases recruitment of TTP to P-bodies in HeLa cells. HeLa cells transiently transfected with FLAG-TTP were left untreated or were treated with LeTx for 1 h. Representative immunofluorescence micrographs show cells exhibiting diffuse (A) or punctate (D) staining of FLAG-TTP. Localization of FLAG-TTP to P-bodies was examined by co-staining with DDX6 (B and E). Merged images are shown (C and F). The fraction of cells exhibiting punctate FLAG-TTP staining was quantified (G). Values indicate mean and SEM of three independent experiments. The asterisk indicates significant difference (P < 0.05).
56
3.3 Discussion
Interfering with host gene expression is an effective means for a bacterial pathogen to
evade the immune response. Not surprisingly then, bacteria and their toxins have developed
various ways to downregulate gene expression. Anthrax LeTx downregulates expression of the
neutrophil attractant IL-8 both transcriptionally and post-transcriptionally. A recent study
demonstrated that through the inhibition of histone phosphorylation, LeTx decreased chromatin
accessibility to NF-kB, leading to lowered IL-8 transcription. This group further implicated this
mechanism in reducing neutrophil recruitment during a B. anthracis infection (Raymond et al.,
2009). It was previously demonstrated that LeTx post-transcriptionally regulates IL-8 expression
by increasing the rate of IL-8 transcript decay (Batty et al., 2006). In the current study, the cis-
acting and trans-acting elements involved in this process is characterized.
The element in the IL-8 transcript that confers sensitivity to LeTx is confined to the 3′
UTR. This region, encompassing nucleotides 1000–1100, has an AU content of 82% and
contains four AUUUA motifs; this ARE has been identified previously as a potent
destabilization element (Winzen et al., 2004b). Surprisingly, the EYFP–IL-81000–1100 reporter
transcript had a shorter half-life than that of the EYFP–IL-8401–1648 transcript. Thus, there may
be a stabilizing element in the 3′ UTR located outside of this ARE, or alternatively, the distance
between the stop codon and destabilizing element might affect the efficiency of decay.
Inhibition of the MAPK pathways using pharmacological inhibitors was found to have similar
destabilizing effects as LeTx on EYFP–IL-81000–1100 mRNA (data not shown), suggesting that it
is the inactivation of these pathways by LeTx that causes IL-8 mRNA decay.
Investigation into the possible involvement of the ARE binding proteins TTP, TIAR and
KSRP in LeTx-mediated IL-8 mRNA destabilization was motivated by past studies that
demonstrated their participation in IL-8 mRNA decay (Suswam et al., 2008; Winzen et al., 2007;
57
Suswam et al., 2005b). Overexpression of KSRP did not alter the level of IL-8 mRNA in
HT1080 cells, whereas overexpression of either TTP or TIAR lowered the level of IL-8 mRNA.
When TIAR expression was knocked down, the half-life of IL-8 mRNA increased from ~81 min
in control cells to ~116 min. However, as was observed in control cells, LeTx accelerated IL-8
mRNA decay by 1.6-fold in TIAR knock-down cells. This observation suggests that while TIAR
influences IL-8 transcript stability, its activity is not regulated by LeTx or by the MAPK
pathways. Not surprisingly then, treatment with LeTx did not cause the redistribution of TIAR
from the nucleus to the cytoplasm or affect its expression level.
Treatment of cells with LeTx decreased the level of phosphorylation of TTP. Knocking
down TTP using siRNA caused increased stability of IL-8 mRNA – no appreciable decay was
measured in these cells and importantly, LeTx treatment did not destabilize the transcript (Fig.
4D). These results correlated with the observation that IL-8 mRNA decay increased upon
pharmacological inhibition of p38, ERK and JNK, and that these inhibitors also caused
dephosphorylation of TTP. Past studies identified TTP as a substrate of the p38/MAPKAP K2
pathway (Chrestensen et al., 2004; Stoecklin et al., 2004), although inhibition of MEK1/2 was
found to have minimal effect on TTP phosphorylation (Suswam et al., 2008), suggesting that the
involvement of the ERK pathway may differ between cell types. Phosphorylation of TTP
through the JNK pathway has not been reported previously.
TTP activity is regulated by phosphorylation in several ways. It has been reported that
phosphorylated TTP has a lower affinity for mRNA, which would reduce its ability to mediate
transcript degradation (Carballo et al., 2001). Furthermore, phosphorylated TTP binds 14-3-3, a
ubiquitously expressed phosphoserine- and phosphothreonine-binding protein that exerts a
variety of effects on its binding partners (Bridges and Moorhead, 2005). The binding of 14-3-3
58
to TTP affects its localization and potentially its ability to interact with components of the
mRNA degradation machinery.
It was found that hypophosphorylation of TTP in intoxicated cells was associated with an
increase in the number of cells exhibiting P-bodies and an increase in the localization of TTP to
P-bodies. P-bodies are cytoplasmic granules that contain translationally repressed mRNA. The
protein composition of P-bodies has not been fully defined, but the known components include
decapping enzymes, activators of decapping enzymes and exonucleases (Parker and Sheth, 2007;
Eystathioy et al., 2002). P-bodies are dynamic structures that vary in size and number depending
on the availability of non-translating mRNA pools. TTP can nucleate the formation of P-bodies
or deliver ARE-containing transcripts to pre-existing P-bodies for storage or degradation. This
work suggests that LeTx might affect mRNA turnover by altering the number and mRNA
content of P-bodies.
59
Chapter 4
Identification of TTP associated proteins and their role in
IL-8 mRNA destabilization
Contributions: I performed all experiments presented in this chapter.
4.1 Summary
Through the inactivation of the MAPK pathways, anthrax lethal toxin (LeTx) causes the
accelerated decay of IL-8 mRNA through the AU-rich element (ARE) in its 3’ untranslated
region (UTR). TTP is an AU-binding protein (AUBP) that was found to play a key role in LeTx-
mediated destabilization of IL-8 transcripts. Here, the interactions of select TTP-associated
proteins were characterized and examined for their role in IL-8 transcript decay. Non-muscle
myosin heavy chain 9 (myosin-9) and HSC-70 were shown to interact with TTP. Their binding
to TTP was not affected by LeTx treatment. RNase treatment did not affect HSC-70 binding to
TTP, but increased the binding between myosin-9 and TTP. A mutant TTP defective in RNA
binding interacted with HSC-70 at a slightly higher level compared to wild-type TTP, but its
interaction with myosin-9 was significantly diminished. Both myosin-9 and HSC-70 expression
play a role in IL-8 transcript stability, as knock-down of each protein led to a lower rate of IL-8
mRNA destabilization. However, treatment of LeTx continued to mediate the accelerated decay
of IL-8 mRNA in these siRNA-transfected cells, indicating that LeTx may not be exerting its
destabilization effects through myosin-9 or HSC-70. In addition, knock-down of myosin-9 led to
a decrease in TTP expression, and but this could not be attributed to the rate of transcription, or
to transcript or protein stability.
60
4.2 Results
4.2.1 Tandem affinity purification of TTP associated proteins
In the previous chapter, it was demonstrated that IL-8 transcripts were stabilized in cells
transfected with siRNA directed against TTP. In addition, the accelerated decay of IL-8 mRNA
that occurs with LeTx treatment was abolished when TTP was knocked down. Other studies
have shown that TTP binds to IL-8 mRNA in vitro (Suswam et al., 2005b), and various protein
components of P-bodies, including the decapping enzyme hDcp2 and the 5’-3’ exonuclease
hXrn1 (Lykke-Andersen and Wagner, 2005). Here, the possibility of TTP associating with
different proteins between untreated and LeTx-treated cells was assessed. The coding region of
TTP was cloned into a vector containing two tags with affinities for streptavidin and for
calmodulin (pNTAP-TTP). This vector was developed for tandem affinity purification (TAP,
Stratagene), where two consecutive purification steps using each of the two tags mentioned
above allows for a clean isolation of interacting proteins. pNTAP-TTP was transiently
transfected into HT1080 cells and then treated with LeTx for 1 h. Then, the TAP-tagged TTP
was isolated with streptavidin resin and eluted. Co-precipitated proteins were separated by SDS-
PAGE and visualized by Coomassie staining (Fig. 4.1). Select proteins that were precipitated
from cells transfected with pNTAP-TTP but not from untransfected cells were identified by LC
MS/MS. Two proteins between 25 and 35 kDa were identified to be isoforms of 14-3-3. Heat
shock cognate protein (HSC)-70 and myosin-9 were also identified as proteins that associated
with TAP-tagged TTP.
61
Figure 4.1: Isolation of TTP-associated proteins. Cells were either left untransfected or transfected with pNTAP-TTP. Lysates were precipitated with streptavidin resin, subjected to gel electrophoresis, and visualized by Coomassie staining. Arrows indicate protein bands that were isolated for LC-MS/MS identification. Gel is representative of 3 independent experiments.
4.2.2 LeTx treatment does not affect myosin-9 and HSC-70 binding to Flag-TTP
To assess whether the treatment of LeTx would affect TTP association with myosin-9 or
HSC-70, the pNTAP-TTP vector was transfected and protein complexes were isolated using
tandem affinity purification and levels of binding were visualized by Western blot assays. The
level of myosin-9 binding to TTP was unchanged between untreated cells or cells treated with
LeTx for 2h or 4h (Fig. 4.2A). LeTx treatment also did not affect the association between HSC-
70 and TTP (Fig. 4.2B).
62
Figure 4.2: Dependence of LeTx and RNA on TTP binding to myosin-9 and HSC-70 A. Cells were untransfected or transfected with pNTAP-TTP. Untransfected cells were left untreated and only transfected cells were treated with LeTx for 2 or 4 h. Lysates undergo affinity purification according to the TAP protocol, and then along with 1% input lysates, were probed for endogenous myosin-9 by immunoblotting (IB). Probe for calmodulin binding protein (CBP) was used as IP control. Blots are representative of 3 independent experiments. B. Cells were transfected and lysates were purified as in Fig. 4.2A, and levels of precipitated HSC-70 were probed by IB. Blots are representative of 3 independent experiments.
63
C. Cells were either transfected with pNTAP-TTP or pNTAP-TTP F126N. Where indicated, lysates were treated with 0.1μg/μl of RNaseA (Sigma) for 30 min. followed by precipitation with streptavidin resin. 1% input lysate and precipitated proteins were subjected to western blot and probed for HSC-70 and CBP. Blots are representative of 3 independent experiments. D. pcDNA-HA-myosin-9 was co-transfected with either pcDNA-FLAG-TTP or pcDNA-FLAG-TTP-F126N. Lysates were immunoprecipitated with 1μg of HIS antibody or HA antibody and then was probed for FLAG-TTP or HA-myosin-9 by western blotting. Blots are representative of 3 independent experiments.
4.2.3 Binding of myosin-9 and HSC-70 to Flag-TTP is not RNA-dependent
The possibility that the interaction between myosin-9 or HSC-70 with TTP is not simply
through simultaneous binding to the same transcript was examined. Two methods were used to
test the RNA dependence of the interaction between the proteins. First, lysates were treated with
RNase A prior to isolation of the target protein complex. Second, wild-type TTP was compared
to TTP-F126N, which contains a mutation in the zinc finger domain that renders it deficient in
RNA-binding. Streptavidin resin was used to purify target protein complexes, and the level of
bound HSC-70 was examined by Western blotting. RNase treatment did not affect the binding
between HSC-70 and TTP, but an increase in binding was exhibited between HSC-70 and TTP
F126N (Fig. 4.2B). To test the RNA-dependence of the interaction between TTP and myosin-9,
pcDNA-HA-HMM, which is myosin-9 truncated at the rod domain but retains the regulatory
behavior of the whole heavy chain, was expressed with either pcDNA-FLAG-TTP or pcDNA-
FLAG-TTP-F126N where indicated. Cells were lysed by EBC buffer and lysates were
immunoprecipitated with anti-HA antibody and proteins were detected by Western blotting. The
level of FLAG-TTP precipitated was increased in lysates treated with RNase compared to
untreated lysates (Fig. 4.2C). The RNA-binding deficient form of TTP did not co-
immunoprecipitate with myosin-9.
64
4.2.4 Knock-down of myosin-9 or HSC-70 stabilizes IL-8 mRNA
It was previously demonstrated that knock-down of TTP led to stabilization of IL-8
mRNA (Chow et al., 2010). Here, the role of the TTP-associated proteins, myosin-9 and HSC-
70, on IL-8 mRNA stability was assessed. Myosin-9 was knocked down by siRNA, and the
destabilization of IL-8 transcripts was measured and compared to negative control cells. The
protein level of myosin-9 was reduced to ~25% in cells transfected with myosin-9 siRNA
compared to cells transfected with negative control siRNA (Fig. 4.3A). Knock-down of myosin-
9 increased the stability of IL-8 mRNA by 2.8-fold, increasing the half-life from ~63 min to
~173 min (Fig. 4.3B). LeTx caused a decrease in IL-8 transcript destabilization in myosin-9
knock-down cells, from ~173 min to ~53 min. The levels IRF1 mRNA were also measured as a
control as it contains an ARE in its 3’ UTR and its stability is known not to be altered in cells
treated with LeTx (data not shown). The stability of IRF1 mRNA exhibited no significant
changes between control cells or myosin-9 siRNA-transfected cells (Fig. 4.3C).
siRNA-mediated knock-down of HSC-70 caused an ~90% decrease in HSC-70 protein
levels compared to cells transfected with negative control siRNA (Fig. 4.4A). Knock-down of
HSC-70 increased the half-life of IL-8 mRNA from ~63 min to ~116 min, and treatment with
LeTx decreased the half-life to ~53 min (Fig. 4.4B). Again, levels of IRF1 mRNA were not
significantly altered between control cells or myosin-9 siRNA-transfected cells (Fig. 4.4C).
65
Figure 4.3: Effect of myosin-9 siRNA knock-down on IL-8 mRNA destabilization A. Extracts from cells transfected with negative control siRNA or siRNA directed against myosin-9 were immunoblotted for myosin-9. β-actin expression was used as loading control. Blots are representative of 3 independent experiments. B. Cells from (A) were treated with 1 μg/ml actinomycin D in the absence or prescence of LeTx. Total RNA was isolated at the indicated times and IL-8 transcript levels were assessed by qPCR. Error bars indicate SEM of three independent experiments. C. Identical to (B) but IRF1 transcripts levels were detected by qPCR.
66
Figure 4.4: Effect of HSC-70 siRNA knock-down on IL-8 mRNA destabilization A. Extracts from cells transfected with negative control siRNA or siRNA directed against HSC-70 were immunoblotted for HSC-70. β -actin expression was used as loading control. Blots are representative of 3 independent experiments. B. Cells from (A) were treated with 1 μg/ml actinomycin D in the absence or prescence of LeTx. Total RNA was isolated at the indicated times and IL-8 transcript levels were assessed by qPCR. Error bars indicate SEM of three independent experiments. C. Identical to (B) but IRF1 transcripts levels were detected by qPCR.
67
4.2.5 Decreased expression of TTP protein upon myosin-9 knock-down
TTP expression in myosin-9 knock-down cells was examined by Western blotting.
Expression of TTP was diminished to ~45% of control cells, while expression of another AUBP,
TIAR, was unchanged (Fig. 4.5A). It was first examined whether this decrease in TTP
expression was due to an altered level in TTP transcription or transcript stability. Control and
myosin-9 knock-down cells were left untreated or treated with LeTx, and actinomycin D was
added to halt de novo transcription. TTP mRNA levels at time=0 exhibited no discernable
difference, and decay of TTP mRNA was also similar between samples (Fig. 4.5B). TTP protein
stability was next compared between control and myosin-9 knock-down cells. Cytoplasmic
proteins were extracted at various time points after the addition of cycloheximide, and the level
of TTP expression was detected by Western blotting (Fig. 4.5C and D). The level of TTP
expression at 0 min was set to 100% for each respective sample and expression at subsequent
time points was calculated as a percentage. Both control and myosin-9 knock-down cells had
~47% of TTP remaining after 480 min.
68
Figure 4.5: Effect of myosin-9 siRNA knock-down on TTP mRNA and protein expression A. Endogenous TTP, TIAR, myosin-9, and β-actin were detected by western blotting of lysates from cells transfected with either negative control siRNA or myosin-9 specific siRNA. B. Cells from (A) were treated with 1 μg/ml actinomycin D in the absence or presence of LeTx. Total RNA was isolated at the indicated times and TTP transcript levels were assessed by qPCR. Error bars indicate SEM of three independent experiments. C. 16 h after the 2nd round of siRNA transfection, cells were treated with 50μM cycloheximide. Cytoplasmic lysates were extracted at the indicated times and probed for TTP, myosin-9, and β-actin by IB. Blot is representative of 3 independent experiments. D. The intensity of TTP signals at each time point were quantified by densitometry. Error bars indicate SEM of three independent experiments.
69
4.2.6 The head domain of myosin-9 is sufficient for TTP binding
To localize the binding region of TTP on myosin-9, different regions of myosin-9 were
cloned in an HA-tagged vector. The head domain, along with the HMM-like fragment
containing the head domain and a truncated rod domain that is sufficient for dimerization was
overexpressed with Flag-TTP. Cells were lysed and immunoprecipitated with anti-HA antibody
and then immunoblotted with anti-FLAG antibody. Both the myosin-9 head domain and the
HMM-like fragment were able to immunoprecipitate Flag-TTP (Fig 4.6A).
Figure 4.6: Characterization of the region that mediates myosin-9 binding to TTP FLAG-TTP was co-transfected with either pcDNA-HA-Head or pcDNA-HA-HMM. Cells were lysed and IP with antibody directed to HA. One percent of input lysates and immunoprecipitated proteins were subjected to western blot analysis using anti-FLAG or anti-HA antibody.
4.3 Discussion
TTP is a regulator of transcript stability of many important inflammatory cytokines. The
critical role of TTP in regulating the expression of cytokines, such as TNFα, is reflected by TTP
knock-out mice, which suffer from a complex inflammatory phenotype that can be prevented by
injection of antibodies directed against TNFα (Taylor et al., 1996). The last decade and a half
70
have yielded many important findings on the mechanism of TTP mediated transcript decay.
Many components of the mammalian mRNA degradation machinery can interact with TTP
(Tiedje et al., 2010), which is thought to facilitate the degradation of TTP-associated mRNA. I
previously discovered that TTP is required for LeTx-mediated destabilization of IL-8 mRNA
(Chow et al., 2010). Based on this, I screened for molecular interactions of TTP that may be
modulated with the treatment of LeTx. Here, novel interacting partners for TTP, HSC-70 and
myosin-9, were identified, and their roles in IL-8 transcript stability were examined.
In addition to HSC-70 and myosin-9, 14-3-3 adaptor proteins were precipitated with
TAP-tagged TTP. The interaction between TTP and 14-3-3 proteins is thought to be regulated
by the phosphorylation of specific serines on TTP, and this interaction prevents TTP from
associating with the cellular mRNA decay machinery (Chrestensen et al., 2004; Stoecklin et al.,
2004; Johnson et al., 2002). Since the mechanism and downstream effects of this interaction
have been established by numerous groups elsewhere, this study was directed to two novel
protein interactions of TTP.
The interactions of myosin-9 and HSC-70 with TAP-tagged TTP were not significantly
affected by treatment of LeTx (Fig. 4.2A and B). With respect to LeTx-mediated IL-8 transcript
destabilization, I previously demonstrated that knock-down of TTP abrogates this effect.
However, in this study, knock-down of either myosin-9 or HSC-70 did not interfere with LeTx-
mediated decay of IL-8 mRNA, although it stabilized IL-8 transcripts in untreated cells.
Therefore, it is likely that HSC-70, myosin-9, and LeTx each affect the function of TTP
independently and through different mechanisms.
Previous reports demonstrated the ARE-binding potential of HSC-70 (Matsui et al.,
2007; Zimmer et al., 2001). This protein belongs to the heat shock protein (HSP)-70 family, and
both HSC-70 and HSP-70 have the ability to interact directly to sequence specific RNAs through
71
their ATPase domain and a C-terminal substrate binding domain (Zimmer et al., 2001). Both
proteins have been described to play a role in ARE-directed transcript stability (Knapinska et al.,
2011; Matsui et al., 2007; Laroia et al., 1999). The two studies propose different mechanisms
for regulation of mRNA decay: one demonstrated the direct binding of HSC-70 to the ARE on
the 3’UTR of BIM mRNA, preventing the association of AUBPs, and the other demonstrated the
ubiqitination of AUF1 by the HSP-70-HSC-70 complex, leading to proteasomal degradation of
the AUBP and its associated mRNA. RNase A treatment did not disrupt the interaction between
HSC-70 and TTP, excluding the possibility that this co-precipitation was due to independent
binding of the same transcript (Fig. 4.2C). Therefore, it is possible that TTP expression may be
regulated by HSC-70 through ubiqintination, and this may be an additional decay pathway for
TTP-associated mRNAs. Another possibility is that through the chaperone functions of HSC-70,
it promotes the folding of TTP required for binding of decay enzymes. Compared to wt-TTP, the
RNA-binding deficient form of TTP, TTP-F126N (Lai et al., 2002), exhibits a slight increase in
binding to HSC-70; whereas RNase treatment did not affect the binding between wt-TTP and
HSC-70. This result is somewhat confounding, as both strategies were used to determine the
RNA-dependency of the interaction. It is possible that loss of myosin-9 binding to the mutant
TTP (Fig. 4.2D) allowed the binding site for HSC-70 to be more accessible, but further studies
will be required to test this model.
The actin cytoskeleton has been previously suggested to play a role in lymphokine ARE-
mediated mRNA decay (Henics et al., 1997). The authors demonstrated that disruption of actin
filaments by cytochalasin resulted in stabilization of IL-2 and TNFα mRNAs (Henics, 1999;
Henics et al., 1997) - both transcripts are also TTP substrates (Tiedje et al., 2010; Brooks et al.,
2004). These studies imply that actin filaments play a role in post-transcriptional regulation of
certain ARE-containing mRNAs. In this current study, I demonstrate that the head domain of
72
myosin-9 can interact with TTP (Fig. 4.6). This specific isoform of heavy chain binds to myosin
light chains to form non-muscle myosin IIA (NM IIA) (Vicente-Manzanares et al., 2009). The
globular head domain encoded by the heavy chain binds to actin and propels actin filaments by
movement of the head domain caused by ATP hydrolysis. It is conceivable that TTP is in a
complex that contains both NM IIA and actin filaments, and these interactions enhance the
destabilizing activities of TTP by translocating TTP and its mRNA substrate to areas of the
cytoplasm that is populated with RNA decay enzymes.
In addition, knock-down of the myosin-9 led to a specific and significant reduction of
TTP protein expression (Fig. 4.5A). The observed decrease in TTP expression cannot be
attributed to a diminished rate of transcription or an accelerated decay of TTP transcript (Fig.
4.5B). TTP protein stability was also similar between control and myosin-9 knock-down cells
(Fig. 4.5C), which suggest that myosin-9, through its function as a motor protein or by an
unknown mechanism, can modulate the rate of TTP translation. Efforts were made to measure
the rate of translation through radiolabelling endogenous TTP proteins, but the lack of a good
quality TTP antibody prevented specific precipitation of TTP (results not shown).
The interaction between TTP and various protein components of the exosome and P-body
was thought to be a molecular link between mRNAs targeted for decay and the cellular decay
machinery. Here, the interaction between HSC-70 and TTP was demonstrated, which may
provide a reason to explore the possibility that the proteasome degradation pathway may be in
play for TTP-associated transcripts. The additional regulation of TTP by myosin-9 is also
fascinating, and provides for another avenue of TTP research. The regulation and mechanism of
TTP-mediated mRNA decay is far from conclusive, and the current study opens new directions
for future studies of TTP.
73
Chapter 5
Discussion
5.1 Summary of thesis findings
The ability of LeTx to modulate various aspects of the host immune response is well
documented. Past studies have described the inhibition of dendritic cell maturation (Agrawal et
al., 2003), or suppression of T-cell and B-cell proliferation (Fang et al., 2006; Fang et al., 2005;
Paccani et al., 2005) by LeTx. The work in this thesis describes a novel mechanism in which
LeTx dampens the host immune response, which is by altering the stability of a transcript that
encodes for a gene important for innate immunity. Associated proteins of the trans-element
important for mediating LeTx-induced destabilization were also examined for their role in
activating transcript destabilization.
The cis- and trans-elements important for LeTx mediated destabilization of IL-8
transcripts were elucidated and presented in chapter 3. Using deletion analysis, I localized a
region of 100 nucleotides in the 3’UTR that is responsive to LeTx-mediated destabilization of
IL-8 mRNA. The treatment of cells with pharmacological MAPK inhibitors paralleled the
destabilization effects on IL-8 mRNA by LeTx, suggesting that it is through the inhibitions of
these pathways by LeTx that causes IL-8 mRNA decay. TTP was identified to be the trans-
acting element important for LeTx-mediated IL-8 mRNA destabilization. Knock-down of TTP
expression by siRNA rendered stabilization of IL-8 mRNA, and prevented LeTx from exerting
its destabilization effect on IL-8 transcripts. Knock-down of TTP-associated proteins, myosin-9
and HSC-70, also stabilized IL-8 transcripts, but LeTx treatment was still able to cause an
accelerated decay of IL-8 in both types of siRNA transfected cells. Collectively, these data
suggest LeTx, myosin-9, and HSC-70 regulate the function of TTP independently.
74
Fluorescent microscopy revealed that LeTx induces the formation of visible P-bodies,
and promotes the co-localization of TTP to P-bodies. The latter observation suggests the
involvement of cellular motor proteins in translocation of TTP to cytoplasmic sites of mRNA
decay under LeTx stimuli. Indeed, myosin-9 is the heavy chain component of NM IIA, which
propels actin filaments. Therefore, it is conceivable that TTP utilizes the motor function of
myosin-9 to transport targeted mRNAs to sites of decay. HSC-70 was also identified in the TTP
precipitated complex, which may affect the folding of TTP that promotes the binding of mRNA
decay enzymes.
Binding of myosin-9 and HSC-70 to TTP are both insensitive to LeTx treatment. RNase
treatment did not affect HSC-70 binding to TTP, but was slightly increased between HSC-70 and
TTP F126N. The binding between myosin-9 and TTP was increased in lysates treated with
RNase compared to no treatment, and myosin-9 binding to the mutant TTP was significantly
diminished. This suggests that myosin-9 may be competing with RNA substrates that bind to the
same site on TTP. Conversely, data indicates the head domain of myosin-9 is sufficient to bind
to TTP.
5.2 Future directions
5.2.1 Involvement of P-bodies in LeTx mediated decay of IL-8 mRNA
Data in chapter 3 demonstrated that treatment of LeTx promotes the formation of visible
P-bodies and the localization of TTP to P-bodies, suggesting the involvement of P-bodies in
LeTx mediated IL-8 mRNA decay. In addition, the possibility that IL-8 mRNA decay may
occur in the exosome must be examined, as evidence exists for both P-bodies or exosomes as the
site of ARE-mediated transcript decay (Lin et al., 2007; Stoecklin et al., 2006; Fenger-Gron et
al., 2005; Lykke-Andersen and Wagner, 2005). Co-immunoprecipitation assays indicate that
75
TTP can interact with core components of P-bodies, which include DDX6, hDcp2, hDcp1a,
hEdc3, and Hedls in HeLa cells (Fenger-Gron et al., 2005; Lykke-Andersen and Wagner, 2005).
Decay of reporter mRNA containing the ARE of GM-CSF was slowed when protein members of
the P-body were knocked-down in HT1080 cells (Stoecklin et al., 2006). However, visualization
of reporter mRNA with the same type of ARE by immunofluorescence indicates that they are
localized to exosomes in HeLa cells (Lin et al., 2007). Therefore, it must be empirically
determined the pathway IL-8 mRNA undertakes for decay under the stimulus of LeTx, and
whether the same decay pathway is utilized for decay at steady state.
To localize IL-8 mRNA and monitor its decay in a spatial and temporal manner, multiple
bacteriophage MS2 binding sites that have specific affinity for bacteriophage MS2 coat proteins
(Peabody, 1993) can be inserted into the 3’ UTR of IL-8. Together with the expression of
fusion proteins comprised of GFP and MS2 coat protein, the exogenously expressed transcripts
containing MS2 binding sites can be tracked by immunofluoresence (Lin et al., 2007; Kedersha
et al., 2005; Sheth and Parker, 2003). Investigation of the co-localization of the GFP-labelled
IL-8 mRNA with distinct cellular mRNA decay machinery can be achieved by dual-labelling
with antibodies directed against DDX6 or PM-Scl75, biological markers for P-bodies and
exosomes respectively. Results may show a lack of co-localization between IL- 8 transcripts
with either P-bodies or exosomes. It is possible that IL-8 transcript decay occurs in the diffuse
cytoplasm, as most P-body components are distributed diffusely throughout the cytoplasm as
well as being localized to P-bodies (Yu et al., 2005; Cougot et al., 2004a; Eystathioy et al., 2003;
Eystathioy et al., 2002). To examine this possibility, siRNA-mediated knock-down of P-body
components described below would be expected to decrease the speed of IL-8 mRNA
degradation.
76
A 100-nucleotide long AU-rich region was shown to be sufficient in mediating transcript
decay in chapter 3. By attaching the MS2-binding site 3’ of this 100-nucleotide region on a
reporter construct, the potential of this region in recruiting the mRNA degradation machinery can
be compared to that of the full length IL-8 3’ UTR. The results would indicate whether this
region is promoting transcript destabilization by actively recruiting the cellular decay machinery.
Knock-down of key protein components of the P-body or exosome can be used to verify
their role in IL-8 mRNA decay. The half-life of IL-8 mRNA can be measured in both untreated
and LeTx-treated cells that have been transfected with siRNA directed against various
components of the decay machinery. There are studies that indicate mRNAs can enter and be
released from P-bodies in a reversible manner (Brengues et al., 2005; Kedersha et al., 2005;
Teixeira et al., 2005). Therefore, time-lapse microscopy should be used to test whether IL-8
mRNA targeted for decay by LeTx can be shuttled back into the translational pool by treatment
of LPS. Results can either confirm that the fate of IL-8 transcripts targeted for decay cannot be
reversed, or that extracellular conditions may be able to stimulate IL-8 transcripts to exit P-
bodies and re-enter the translating pool. The latter scenario may provide the cell with a robust
response time to quickly adapt to new environmental conditions.
5.2.2 Mapping the interaction between TTP and myosin-9
The central domain of TTP consists of two zinc finger motifs that are important for RNA
binding (Brewer et al., 2004; Lai and Blackshear, 2001; Lai et al., 2000). Unlike the N-terminal
domain, the C-terminal domain has not been demonstrated to interact with any RNA decay
enzymes (Lykke-Andersen and Wagner, 2005). However, deletion of the N-terminal domain
does not render TTP completely inactive, and over-expression of just the C-terminal domain of
TTP impairs ARE-mediated decay (Lykke-Andersen and Wagner, 2005). Together, these
77
observations suggest that a limiting factor in ARE-mediated decay interacts with the C-terminal
region. Therefore, to map the interface TTP requires for myosin-9 binding will further
contribute to the current understanding of the decay mechanism of TTP.
Purified myosin-9 can be incubated alone, or with GST-purified proteins containing
either the N-terminal, the C-terminal, or the central domain of TTP. By separation with native
PAGE, the mixture that contains an interaction between myosin-9 and GST fused proteins can be
deciphered by its altered mobility when compared to the control that contains purified myosin-9
alone. The results from this experiment can be verified by repeat use of these purified proteins
and utilizing cell lysates to test for myosin-9 binding. The GST-tagged proteins described above
can be used to coat glutathione agarose beads by incubation in EBC buffer. Agarose beads
coated with different domains of TTP can then be incubated with HT1080 cell lysates. After
incubation, the standard pull-down protocol can be followed. The presence or absence of
myosin-9 binding to different domains of TTP can be visualized by western blotting.
5.2.3 Importance of NM IIA motility function in its regulation on TTP function
The cellular function of NM IIA is central to cell adhesion and migration. It binds to
actin and propels the sliding or contracting of actin filaments by utilizing the energy from ATP
hydrolysis (Vicente-Manzanares et al., 2009). Out of the three isoforms of NM II, NM IIA
exhibits the highest rate of ATP hydrolysis, and therefore is most suitable for propelling actin
filaments more readily than the other isoforms. A previous study demonstrated that cytochalasin
B treatment stabilized ARE-containing transcripts, IL-2 and TNFα, and that an unidentified
AUBP can associate strongly with actin (Henics et al., 1997). Cytochalasin inhibits actin
polymerization by binding to the growing end of the actin filament, preventing assembly or
78
disassembly of individual actin monomers at the end of the filament. Therefore, this study
provides a possible link between ARE-mediated decay and the actin cytoskeleton.
The findings in chapter 4 revealed that myosin-9, the heavy chain isoform in NM IIA,
interacts with TTP and plays a role in IL-8 mRNA stability. Due to the established function of
NM IIA in actin binding and motility, and the possible role of actin filaments in ARE-mediated
decay, it would be logical to examine the destabilizing function of TTP under conditions of actin
disruption. Latrunculin A binds to monomeric actin at a 1:1 ratio and prevents assembly of
filamentous actin (Spector et al., 1989). Blebbistatin inhibits myosin by preferentially binding to
the myosin-ADP-Pi, which interferes with phosphate release and leaves myosin in an actin-
detached state (Kovacs et al., 2004). I compared the rate of IL-8 mRNA decay in cells that were
untreated, treated with LeTx, with 10μM blebbistatin, or with 1μM latruculin A. As I previously
saw in chapter 3, LeTx activity led to an accelerated decay of IL-8 mRNA compared to untreated
cells, but treatment of blebbistatin or latruculin A did not affect IL-8 stability (Fig. 5.1). This
suggests that F-actin assembly and the motor function of myosin are not involved in regulation
of IL-8 transcripts. However, it may be worthwhile to test the effect of cytochalasin B on IL-8
transcript stability as it disables assembly of filamentous actin by a different mechanism
compared to latrunculin A, and treatment of this agent led to the stabilization of other transcripts
(Henics et al., 1997).
79
Figure 5.1: Effect of blebbistatin and latrunculin treatment on IL-8 mRNA destabilization HT1080 cells were either treated with LeTx for 1 h followed by addition of actinomycin D, or were treated with DMSO, blebbistatin, or latrunculin together with actinomycin D. Total RNA was isolated at the indicated times and transcript levels were measured using qPCR. Error bars indicate SEM of three independent experiments.
5.2.4 Role of TTP in MYH9 diseases
Various mutations in the MYH9 gene have been identified to be responsible for human
disorders such as May-Hegglin anomaly, Fechtner syndrome, and Sebastian syndrome (Heath et
al., 2001; Seri et al., 2000). These conditions are now collectively called MYH9-related
disorders, and are characterized by large platelets, thrombocytopenia, and other non-
hematological manifestations that include the appearance of inclusion bodies in neutrophils
(Kunishima et al., 2001). The inclusion bodies are composed of aggregates of myosin-9 protein,
and are visible by immunofluorescence (Kunishima et al., 2003). Due to the physical connection
between TTP and myosin-9 established in chapter 4, it would be of great interest to examine the
TTP binding potential of mutant myosin-9, and the downstream affects of TTP function in cells
expressing the mutations seen in MYH9 disorders.
80
There are at least 14 amino acid mutations or deletions that have been identified from
patients suffering from MYH9-related disorders (Kunishima et al., 2001). Select mutations can
be engineered into a HA-tagged plasmid that would otherwise contain the coding region of
myosin-9. The binding potential of these mutants to pcDNA-FLAG-TTP can be determined
using standard IP protocols.
The destabilizing functions of TTP should also be examined in cells that exhibit inclusion
bodies in the cytoplasm. Mutant myosin-9 are aggregated and do not retain their wt functions,
and could well affect TTP activity. Neutrophils and other leukocytes from clinical samples of
patients stricken with MYH9-related disorders can be isolated by percoll and ficoll density
gradients. mRNAs can be isolated from normal cells and MYH9-mutated cells, and the levels of
TTP-regulated mRNAs, such as GM-CSF, TNFα, or IL-8 can be examined by real-time PCR. If
the destabilizing activity of TTP is modulated by defective myosin-9, then the levels of these
transcripts should be stabilized. These results can potentially contribute to the field of MYH9
diseases as the influence of mutant MYH9 on leukocytes is still poorly understood.
5.3 Conclusions
Destabilization of mRNA reduces protein synthesis, and also irreversibly inhibits gene
expression. This highly regulated process is receptive to cell stimuli, and stabilizes or
destabilizes transcripts that are appropriate to respond accordingly to the stimuli. AUBPs act as
a link between extracellular signals and mRNAs to be stabilized or targeted for degradation. The
results demonstrated here provide evidence that under the stimuli of LeTx, TTP is modulated
post-translationally and its cellular localization shifts to P-bodies. TTP was found to be able to
bind to IL-8 transcripts in vitro (Suswam et al., 2005b). Further studies are required to directly
implicate the role of P-bodies in IL-8 mRNA decay in cells treated with LeTx. The complete
81
picture of the mechanism of TTP mediated mRNA decay is far from conclusive and the
observation that TTP binds to two novel interactors in myosin-9 and HSC-70 adds to the enigma.
While it is clear TTP can interact with various components of the mRNA decay machinery, there
may be additional factors that may be necessary for TTP to cause mRNA decay, such as factors
involved in mRNA-protein remodeling or in localization of mRNA to P-bodies. More studies
will be required to examine the possibility of TTP-associated proteins, including myosin-9 and
HSC-70, which act as co-factors to TTP.
82
References
Agrawal, A., Lingappa, J., Leppla, S.H., Agrawal, S., Jabbar, A., Quinn, C. and Pulendran, B.
(2003) Impairment of dendritic cells and adaptive immunity by anthrax lethal toxin.
Nature. 424: 329-334.
Aihara, M., Tsuchimoto, D., Takizawa, H., Azuma, A., Wakebe, H., Ohmoto, Y., et al (1997)
Mechanisms involved in Helicobacter pylori-induced interleukin-8 production by a
gastric cancer cell line, MKN45. Infect Immun. 65: 3218-3224.
Al-Souhibani, N., Al-Ahmadi, W., Hesketh, J.E., Blackshear, P.J. and Khabar, K.S. (2010) The
RNA-binding zinc-finger protein tristetraprolin regulates AU-rich mRNAs involved in
breast cancer-related processes. Oncogene. 29: 4205-4215.
Amrani, N., Sachs, M.S. and Jacobson, A. (2006) Early nonsense: mRNA decay solves a