Results Characterization of MOB1A mutants with selective binding to the

To expand the characterization of our panel of MOB1A versions (see figure 3.7), we characterized the PPIs between MOB1A variants and the Hpo, Warts, and Trc kinases (the Drosophila counterparts of the mammalian MST1/2, LATS1/2, NDR1/2 kinases). Through phylogenetic analysis, it has been shown that Drosophila Mats (dMOB1) and human MOB1A shares 87 % identity, strongly suggesting that the functions of Mats proteins are conserved from Drosophila to humans (Hergovich 2011). Indeed, Lai et al. (2005) reported that loss of function of Mats can be rescued by expression of MOB1A, supporting the notion that the functions of MOB1A are conserved between flies and humans. In addition, it has been shown that the MST1/2 kinases can regulate MOB1A in a similar manner as reported for the regulation of Mats by Hpo in fly cells (Hergovich, 2011; Wei et al, 2007). Furthermore, it has also been reported that MOB1 and Mats can associate with LATS1/2 and NDR1/2, Warts and Trc kinases, respectively. These findings collectively suggest that the functional importance of PPIs between MOB1A and Hippo core kinases are conserved from insects to humans.

First, the PPIs of our selected MOB1A mutants with Hpo were examined. The results are presented in Figure 3.8 below.

129 | P a g e

Figure 3.8. Characterization of the interaction of MOB1A variants with the fly Hpo kinase.

(a-b) Lysates of Drosophila S2R+ cells expressing full-length HA-Hippo(wt) together with

indicated full-length MOB1A versions were subjected to immunoprecipitation (IP) using anti-HA. The KKEE mutations caused loss of binding to Hpo, while the DV mutant bound normally. Complexes were studied by immunoblotting using anti-myc (top) and anti-HA (top middle). Input lysates were probed with anti-myc (bottom middle) and anti-α-tubulin (bottom). Relative molecular weights are shown.

To characterize the interactions of MOB1A(wt) and MOB1A mutants with Hpo, S2R+ cells were co-transfected with N-terminally tagged versions of myc-MOB1A(wt) and myc-MOB1A mutants and HA-Hpo, followed by co- immunoprecipitation using anti-HA antibody. This revealed that Mats and MOB1A(wt) bound similarly to Hpo in S2R+ cells (Figure 3.8), confirming that Mats/Hpo and hMOB1A/Hpo interactions are conserved. As further indicated in Figure 3.8, MOB1A(wt) lane and MOB1A(DV), but not the other MOB1A variants, were able to bind to Hpo. Collectively, our data presented in Figure 3.8 suggest that the MOB1A(KEKE) mutant is deficient in Hpo binding, hence developing a research tool that allow us to understand the role of MOB1A-Hpo complex formation in complex multicellular organism such as Drosophila. Noteworthy, the observed interaction pattern in fly cells (Figure 3.8) is identical with the one observed in mammalian cells (Figure 3.7).

130 | P a g e

Second, the PPIs of our selected hMOB1A mutants with Warts were examined. The results are presented in Figure 3.9 below.

Figure 3.9. Characterization of the interaction of MOB1A variants with the Wts kinase.

(a-b) Lysates of Drosophila S2R+ cells expressing full-length HA-Wts(wt) together with

indicated full-length MOB1A versions were subjected to immunoprecipitation (IP) using anti-HA. Complexes were analyzed by immunoblotting using anti-myc (top) and anti-HA (top middle). Input lysates were analyzed with anti-myc (bottom middle) and anti-α-tubulin (bottom). Relative molecular weights are shown. The DV mutant did not bind to Warts, while the KEKE mutant bound normally.

To characterize the interactions of MOB1A(wt) and MOB1A mutants with Hpo, S2R+ cells were transfected with N-terminally tagged versions of myc-MOB1A(wt) and myc-MOB1A mutants and HA-Hpo, followed by co- immunoprecipitation using anti-HA antibody. As indicated in Figure 3.9a, MOB1A(wt) and Mats were able to bind strongly to Wts. Following that, we characterized the binding of our MOB1A mutants to Wts. This revealed that MOB1A(KEKE) associated with Wts like wild-type MOB1A, while the remaining MOB1A variants did not interact stably with Wts (Figure 3.9b). Collectively, these data suggest that the MOB1A(DV) mutant is deficient in

131 | P a g e

Wts binding, hence establishing a foundation that allow us to study the significance of MOB1A-Wts complex formation in Drosophila. Noteworthy, this interaction pattern is identical with the one observed in mammalian cells (Figure 3.7).

Last, the PPIs of our selected hMOB1A mutants with Trc were examined. The results are illustrated in Figure 3.10 below.

Figure 3.10. Characterization of the interaction of MOB1A variants with the Trc kinase

(a-b) Lysates of Drosophila S2R+ cells expressing full-length HA-Trc(wt) together with

indicated full-length MOB1A versions were subjected to immunoprecipitation (IP) using anti-HA. Complexes were examined by immunoblotting using anti-myc (top) and anti-HA (top middle). Input lysates were investigated with anti-myc (bottom middle) and anti-α- tubulin (bottom). Relative molecular weights are shown.

To characterize the interactions of MOB1A(wt) and MOB1A mutants and Mats with Trc, S2R+ cells were co-transfected with N-terminally tagged versions of myc-hMOB1A(wt) and myc-MOB1A mutants and HA-Trc, followed by co-immunoprecipitation using anti-HA antibody. As indicated in Figure 3.10a, both Mats and MOB1A(wt) were able to associate with Trc. Following that, we characterized the binding of our MOB1A mutants to

132 | P a g e

Trc. This revealed that all mutants of MOB1A normally bound to Trc like wild-type MOB1A.

Taken together, we show here that MOB1A interacts with the Hpo, Warts, and Hpo kinases in fly cells. In addition, we present the characterization of the interaction patterns of our selected MOB1A versions with the fly Hippo core kinases. This revealed that the MOB1A(DV) can interact with Trc and Hpo, while being unable to associate with Wts. In contrast, MOB1A(KEKE) can interact with Wts and Trc, but does not associate with Hpo. Therefore, we are describing here development of research tools that are suitable to investigate the significance of MOB1A-Hpo, MOB1A-Wts, and MOB1A- Trc complex formation in the context of Mats loss-of-function. Furthermore, our results presented here suggest that the interaction patterns of MOB1A variants in fly cells are nearly identical with the patterns observed in mammalian cells (Figure 3.11).

133 | P a g e

Figure 3.11. Generation of MOB1 variants to dissect the importance of MOB1 interactions with Drosophila Hippo core cassette kinases.

(a) Model illustrating two key research questions regarding MOB1 as a hub of Drosophila

Hippo core signalling: (1) Does MOB1 interact differently with Hippo, Warts and Trc in fly cells? (2) Which interactions of MOB1 with Hippo, Warts, or Trc are biologically important in the context of Yki-Hippo signalling? (b) Schematic summary of the biochemical and molecular characterization of the indicated MOB1A variants presented in

Figures 3.8 to 3.10. (c) Model of human MOB1A(33-216) depicting the possibly opposing

binding sites on MOB1 of Trc/Warts and Hpo kinases. Secondary structure elements of MOB1 are highlighted. The locations of Asp63, and Lys104/Lys105 in MOB1 are indicated. (d) Schematic model of how the differential binding of MOB1 to Hippo, Warts and Trc can be experimentally exploited using the D63V and K104E/K105E mutations in order to study the biological importance of these different interactions of MOB1 with Drosophila Hippo core cassette kinases

134 | P a g e

3.2.4. Discussion

Our structural and biochemical studies of human Hippo core components revealed the identity of key residues in MOB1A that are required for the binding of MOB1A to the MST1/2, LATS1/2, and NDR1/2 kinases (see section 3.1 above). Importantly, our structural and biochemical studies presented above revealed that D63 residue of MOB1A is required for LATS1/2-binding but not NDR1/2-binding, while K104E/K105E residues of MOB1A is key binding residues essential of binding to MST1/2, not LATS1/2 and NDR1/2 kinases. Collectively, these results develop an important tool to study the importance of MOB1A-MST1/2, MOB1A- LATS1/2, and MOB1A-NDR1/2 complex formation in the context of the regulation of Hippo signalling in mammalian cells.

After having revealed the key residues that plays crucial roles in the binding of MOB1A to MST1/2, LATS1/2, and NDR1/2, we asked next whether these residues of MOB1A are also critically important for MOB1A binding to the fly Hpo, Warts, and Trc kinases. To do so, we established first that MOB1A(wt) can interact with the Hpo, Wts, and Trc kinases. We next tested whether the important residues that disrupted binding of MOB1A to the MST1/2, LATS1/2, NDR1/2, respectively, are also important for the interactions of MOB1A with the fly Hippo core kinases Hpo, Wts, and Trc. Significantly, these analyses of PPIs in the fly cell extracts revealed that the interaction patterns in Drosophila cells are highly similar with of the patterns observed in mammalian cells. This strongly suggests that the regulation of Hippo core signalling by MOB1A is highly conserved from Drosophila to humans. Thus, our selected mutants of MOB1A will allow us to understand which protein-protein interactions are required for embryogenesis, and tissue growth control in Drosophila, in addition to performing experiments in human cancer cells.

135 | P a g e

3.3. Defining phosphorylation patterns of our MOB1A variants by