Studies on Memo, an important ErbB2 receptor-mediated component of the cellular migratory machinery

The ErbB2 receptor tyrosine kinase has been shown to play an important role in cancer cell
motility and metastases formation. This receptor is often overexpressed in human tumors of
diverse origins, including breast and ovarian cancer. Individuals with ErbB2 over expressing
tumors have shown poor clinical outcome.
Our studies are focused on signaling molecules that interact with autophosphorylated tyrosine
residues of the cytoplasmic tail of the receptor. Two of the sites, Tyr 1201 (YC) and Tyr 1227
(YD) are fully able to restore the migratory phenotype of breast carcinoma cells. Studies of
the functional role of ErbB2 phosphorylation sites identified PLCγ1 as an interacting partner
of the YC autophosphorylation site, and Memo (Mediator of ErbB2-driven cell Motility) as a
binding partner of the YD site that is required for ErbB2 induced cell motility.
Memo is encoded by a unique gene that is found in all branches of life, from bacteria to
humans. Memo has no characterized domains, nor does it have obvious catalytic activity.
Various approaches were used to position Memo in a signaling pathway and to uncover its
biochemical function. Memo was initially detected based on its important role in ErbB2-
induced cell motility. In fact, tumor cells with a specific knock-down (KD) of Memo failed to
grow microtubules in response to Heregulin (HRG)-induced ErbB2 activation and were
impaired in their migration.
Cell migration proceeds in distinct steps. In response to a chemotactic stimulus, cells extend
protrusions at the front that help in attachment. This is followed by contraction of the cell
body and tail detachment at the rear allowing movement in the direction of the ligand. The
initial event in the process is sensing of the ligand in response to activation of cellular
receptors like EGFR or ErbB2. Their activation initiates signaling pathways that lead to
polymerization of new actin at the leading edge, which is necessary for generating the
protrusive force allowing migration. An important goal of my thesis work has been to
investigate the step(s) of the migratory process that require Memo.
In the first study, we explored migration using Dunn chambers and analyzed the chemotactic
response of tumor cells in a shallow gradient of ligand. By tracing HRG-stimulated cell
migration in time-lapse video microscopy, we found that Memo or PLCγ1 KD strongly
impairs cell directionality, reflecting an important role for Memo and PLCγ1 in orchestrating
directional cell migration. We also demonstrated that depletion of Memo or PLCγ1 resulted in
very similar phenotypes, with a strong impairment of HRG-induced cytoskeletal organization.
To gain more insight into Memo’s function, we carried out a Yeast-2-Hybrid (YTH) analysis
and found a number of interesting new partners of interaction for Memo. Of particular interest
is the small protein cofilin, one of the major cellular actin severing and depolymerizing
factors that is known to have an essential role in directional sensing during chemotaxis. This
interaction was confirmed in vitro using recombinant proteins and in vivo in coimmunoprecipitation
experiments where Memo was detected in complexes with cofilin,
ErbB2 and PLCγ1. Interestingly, we also found that HRG-induced PLCγ1 phosphorylation
was decreased in Memo KD cells, suggesting that Memo regulates PLCγ1 activation.
Furthermore, by introducing GFP-tagged cofilin into control, Memo or PLCγ1 siRNA
transfected breast tumor cells, we showed that HRG-induced recruitment of GFP-cofilin to
lamellipodia is impaired in Memo- and in PLCγ1 KD cells, suggesting that both proteins lie
upstream of cofilin in models of ErbB2-driven tumor cell migration. Finally, we examined the
effect of Memo on cofilin binding and severing/depolymerizing properties. In vitro F-actin
binding assays showed that Memo does not impair cofilin binding to F-actin, and revealed
that Memo is a novel F-actin binding protein. In vitro F-actin depolymerization assays
indicated that Memo promotes cofilin depolymerizing/severing activity. Altogether, these
data suggest a novel role for Memo during the migratory process and its implication in the
regulation of actin dynamics through cofilin binding.
In the second study, we used two different Memo-defective cellular models to examine
Memo’s function in more detail. We demonstrated that inhibition of Memo impairs activation
of a number of signaling molecules including Src, Shc, ERK and PLCγ1. We also provide
evidence that Memo interacts with the three Shc isoforms, p46shc, p52shc, and p66shc, and
showed that Shc is required for Memo binding to the ErbB2 receptor. Control and Memodeficient
cells were also scored for their migration and adhesion properties. These assays
indicated that Memo is important in both cell migration and adhesion processes. Also,
morphological and biochemical analyses of control and Memo-deficient cells suggested that
Memo is involved in focal adhesion organization and rear cell deadhesion during the
migratory process.
Altogether, these two studies revealed important roles for Memo at different steps of cell
migration and metastasis, making it a potential interesting target for cancer therapy.
Genetic approaches in model organisms have been important for gaining insight into the
function of evolutionarily conserved proteins. To position Memo within a genetic network,
experiments in the model organism S. cerevisae that lends itself to rapid genetic screening
were performed. We investigated cellular localization of Memo in yeast and found that Memo
is located in the nucleus and cytoplasm of the cell. A S. cerevisae memo Δ strain has been
generated and is viable. Considering the role of Memo in the microtubule and actin networks
that we described in mammalian cells, we examined the memo Δ strain for defects in different
cytoskeletal dynamics. No significant effect was observed. We also performed a Synthetic
Lethal Screen of genetic interactions between a memo Δ strain and an ordered array of 4700
Yeast strains containing non-essential gene deletions. This analysis revealed a limited number
of synthetic interactions. Lethality was observed in combination with the plc1Δ strain. PLC1
encodes for the unique isoform of phosphatidylinositol-specific phospholipase C of S.
cerevisiae. The results are intriguing and exciting considering the data obtained in the
mammalian models; in fact, we demonstrated that Memo and PLCγ1 interact with ErbB2
autophosphorylation sites and are essential for directional migration. We also showed that
Memo is found in a complex with PLCγ1 and ErbB2 and that Memo is likely contributing to
PLCγ1 activation. We hypothesize that in Yeast, Memo and PLC1 act in the same or in
distinct but related pathways, and suggest that the connection between PLC and Memo
induced-pathways is also conserved through evolution.