Godwin Agbeka

I'd encourage other faculty to explore this program. Afrisnet now has @HHMI and Burroughs-Wellcome funding to support these internships, and the students are among the best in their nations

This is a wonderful partnership. Afrisnet mentors dozens of youhg STEM professionals across the continent of Africa & has helped us and other Universities identify promising students for summer internships. Our four previous scholars are in PhD programs in the US or will start them in the Fall 2/n

Joseph Dordor and Lawrencia Semireka in front of UNC's Belltower

Joseph Dordor and Lawrencia Semireka in front of UNC's Belltower

I want to give a belated Bluesky welcome to our two summer science interns from KNUST in Ghana, Joseph Dordor and Lawrencia Semireka. They are spending 10 weeks in our lab doing research, in cooperation with our Afrisnet colleagues. 1/n
afrisnet.org

Cell Contact GRC--here we come. Step two--get coffee in Heathrow and wait for the flight to Geneva

Man with blue sky and clouds behind

RDU-London flight

Cell Contact GRC--here we come. Step one--fly to London
www.grc.org/cell-contact...

Of course, new answers pose new questions. What additional functions does RA1 have in addition to Rap1 binding? How does Rap1 binding “activate Canoe”? We have a lot to learn! 15/n

Images of pupal eyes in wild type and mutants revealing defects

Why has evolution retained an RA2 domain over the 600 million years separating us from flies? More subtle tests of function revealed that Canoe∆RA2 is not fully functional. For example, the complex cell shape changes shaping each ommatidium in the compound eye have defects in that mutant. 14/n

Diagrams of gaps at cell-cell junctions and quantification

We dug in and did a lot of cell biology, revealing that the RA1 domain is key for the ability to strengthen cell-cell junctions and maintain their connections to the cytoskeleton when force is exerted on them, while RA2 is almost completely unnecessary 13/n

Even more surprising, replacing RA1 with RA2 did NOT restore function, while an animal with two RA1 domains was also viable and fertile. Thus despite similar abilities to bind Rap1, they are functionally quite different 12/n

Pictures of adult cno∆RA2 and cnoRA1RA1 mutants

It was here we got our next surprise. Deleting RA1 almost completely eliminated Canoe function—mutants died as embryos with dramatic defects in cell shape change, as in mutants lacking both RA domains. However, mutants lacking RA2 survived to adulthood & we could establish a mutant stock! 11/n

Diagram of our mutant proteins from Fig 3 in the link

This suggested the two RA domains might have redundant functions. We tested this in the fruit fly. CRISPR-tools allow us to replace the wildtype canoe gene with versions with either RA1 or RA2 cleanly deleted, or, in a further test, replace one with the other, to see if they are interchangeable 10/n

Fig 2 from the link showing ITC binding studies

Leah then revisited the biochemical interaction experiments, now using Canoe’s RA domains and including the N-terminal extensions. To our surprise she found both RA domains bind Rap1 with similar micromolar affinities! 9/n

Intriguingly, the predictions suggested both RA1 and RA2 should bind Rap1, and also revealed a key feature—a surprising N-terminal extension on each RA domain that was present in the RA1 domain used for binding studies and truncated in the version of RA2 used in those experiments 8/n

Fig 1 of the link showing predicted structures of RA1 and RA2 bound to Rap1

Biochemical work on Afadin 20 years ago suggested they differed, with RA1 binding Rap1 tightly and RA2 not binding it at all. We re-examined this question. AlphaFold (awarded this year’s Nobel Prize in Chemistry!) allow us to predict protein structures, and we did so for RA1 or RA2 and Rap1 7/n

Title authors and abstract from the link

In work now in press @jcellsci.bsky.social , we explored this, combining bioinformatic, biochemical, genetic and cell biological tools. This was powered by a remarkable team of postbac scholars including @emily-mcparland.bsky.social
journals.biologists.com/jcs/article/...

Title, authors and abstract from the link

In earlier work we found that deleting both domains inactivates Canoe, confirming the importance of Rap1 interaction. However, why are there TWO RA domains? Do they have similar or different functions? 5/n
rupress.org/jcb/article-...

Protein domains in Canoe/Afadin

Most proteins that are regulated by small GTPases have conserved binding sites for them called “Ras-association” or RA domains. Canoe/Afadin is unusual in having two RA domains instead of one. 6/n

Wikipedia--GTPase cycle https://en.wikipedia.org/wiki/Guanine_nucleotide_exchange_factor

One key question is how adhesion & cell shape change are regulated. Cell-cell signaling via small GTPases of the Ras superfamily is key. In their active state they bind proteins like Canoe/Afadin and “turn them on”. We’re seeking to understand how the small GTPase Rap1 regulates Canoe. 5/n

Diagrams showing complex multi domain architecture and conservation of Canoe/Afadin

We study a key part of this protein complex, called Canoe in fruit flies & Afadin in us. It's key for cell shape changes throughout development & continues to play roles in tissue homeostasis. Mutations in it are found in some breast cancer subtypes. We want to understand how it works as machine 4/n

Diagram of proteins at cell-cell adherens junctions

In cell-cell junctions proteins called cadherins link cells to one another and a complex of proteins binds their cytoplasmic tails to link to actin filaments 3/n

These events require cells to generate force, which often involves the actin & myosin cytoskeleton, the same machines driving muscle movement. This force-generating machine is then linked to the plasma membrane at cell junctions that link cells together or link them to the extracellular matrix 2/n

While our bodies seem static, our cells are constantly changing shape or on the move during tissue homeostasis or wound repair & even more dramatically during embryonic development, during events like gastrulation. Defining how cellular machines mediate these events is a key task for our field 1/n

Email accepting me to the meeting I am organizing....

I am excited to report that the organizers of the 2025 Cell Contact and Adhesion Gordon Conference accepted my application. It's one of my favorite meetings! Oh wait, I am the organizer.......
PS Hope folks join us in the Swiss Alps for amazing science
www.grc.org/cell-contact...

Thanks @peiferlabunc.bsky.social for this mention

Title and abstract of the memoriam in the Journal of Cell Biology in the link

In September the world lost one of the great biologists of our time: Joe Gall. He bridged the transition from the histology era of cell biology to our molecular present, contributing a remarkable set of insights in his 9 decades 1/n 🧪
rupress.org/jcb/article-...