Read on Twitter
Yayy, my first ever preprint is now online. Here we discover an unexpected new regulator for macropinocytosis, CYRI-A. If you like imaging and actin, this is for you :) Tweetorial is followed: https://www.biorxiv.org/content/10.1101/2020.12.04.411645v1
1/n: Our lab and @robinsall lab discovered and characterised a novel regulator for the Scar/WAVE complex called CYRI-B. CYRI-B binds and sequesters active Rac1, thus inhibits actin polymerisation. However, its exact localisation and kinetics was not known.
2/n: I arrived in the lab a few years later and was given whether I would like to work with CYRI-B or its cousin protein CYRI-A. And because I was new and naive (LOL), I chose CYRI-A because its new and nobody has worked on it (and there're many reasons for this).
3/n: CYRI-A is not widely expressed, the protein is extremely difficult to purify, and there was no antibody for even western blotting at all. The literature is empty. So I didn't really know what I've got myself into. But as a young and eager scientist, I took on the challenge.
4/n: CYRI-B knockout (KO) cells have broad lamellipodia. The 1st question I asked: Can CYRI-A rescue this phenotype? And it did. We use Anti-ArpC2 antibody for IF and saw cells expressing CYRI-A lose this localisation.
5/n: Next, we asked can CYRI-A also binds active Rac1? And the answer is yes, the binding is detected with pulldown as well as proved to be direct through SPR.
6/n: So we wanted to look a bit closer into the in vivo biology of CYRI-A, so we obtained these A-673 cells from a collaborator where they express both A and B at a relatively comparable amount.
7/n: We noticed that only when we deleted both A and B (DBKO for double knockout) that we saw a striking phenotype, where these cells adopting this C-shape phenotype. These DBKO cells also migrate faster in both 2D and 3D condition.
8/n: This suggests that CYRI-A and B are able to compensate for each other, at least partially. The next question is where are these proteins in the cells. So a previous PhD student in the lab cloned an internal GFP construct for CYRI-B, while I did the same for CYRI-A.
9/n: When I express my CYRI-A construct in the cells, first in COS-7, i saw a striking localisation of the protein at vesicular structures. At first I was skeptical, every cell type I tried would give the same localisation.
10/n: This was one of those Eureka moments for me since we did not expect the construct to work. We had no reason to, but it did. And this construct will remain my most pride construct of all time.
11/n: Since the size of these vesicles are large, average around 1-2um in diameter. We wonder whether they are macropinosomes. So we performed a dextran incorporation assay and filmed it. And indeed, we could see Dextran present in these CYRI-A-positive vesicles.
12/n: We next asked, what about actin? Since we know actin is an important part of macropinosome formation. And what we found is that CYRI-A signal peaks and is followed sharply by the drop in actin signal specifically at the macropinocytic cup.
13/n: We also found that those vesicles without CYRI-A on tend to have longer actin lifetime. And that cells acutely depleted of CYRIs are slower and less efficient in uptaking Dextran.
14/n: We then also found the same happen for Rac1. The signal of active Rac1 (PDB as the reporter) drops right when CYRI-A appears, suggesting that CYRI-A could dampen down the Rac1-actin signalling axis, specifically at the macropinocytic cup.
15/n: And we found that it is this active Rac1 binding one of the main signals that drive the recruitment of CYRI-A to these cups. When we mutate the Arginine motif important for Rac1 interaction, we no longer see CYRI-A's recruitment here in magenta. Cyan is the WT CYRI-A.
16/n: Going back to the spreading phenotype of the DBKO cells. We know that CYRI-A is involved in the formation and maturation process of macropinosomes. and we know that macropinosomes are also one of the pathways for integrin uptake. So we ask could we localise integrins here?
17/n: And indeed, we saw integrins enriched on these CYRI-A vesicles as they are drawn into the cell. Suggesting that at least in part, CYRI-A does contribute to integrin internalisation.
18/n: When we measured the internalisation capacity of these cells, we saw that CYRI DBKO cells internalise less active integrin a5 than control pLKO cells.
19/n: And we also found that the increased surface expression of these integrins contributes to the enhanced spreading and invasive phenotype in cancer cells, both in organotypic and inverted invasion assay as well as anchorage-independent growth.
20/n: If you want to read the full story, check out the preprint, and also if you're at #cellbio2020, catch me on the 16th Dec. Happy to discuss any question :)
21/n: Overall, this project was a culmination of 2 years of work, and I cannot describe how hard it was. Working with a protein with no tools, and starting with a project that I have absolutely no idea if its gonna lead anywhere was scary.
22/n: I also think luck plays an extremely important role. We didn't know any of our constructs would work, or if we could find the right cell line or the right antibody. I was determined, but without the help along the way of others, this might not be possible. So thank you!
