Huge thanks to all my lab members, and our funders for their support, including @brotmanbaty.bsky.social , @alleninstitute.org , @hhmi.org , UW Ob&Gyn, @uwgenome.bsky.social , @uwiscrm.bsky.social , @uwmedicine.bsky.social , @cziscience.bsky.social , etc. (9/n)

It has been an incredible journey with all co-authors, Connor Kubo, Stella Xu, Riza M. Daza, Eva Nichols, Wei Yang, Anh Vo, Mary B. O’Neill (8/n)

Finally, we tested if the AP-gastruloid model can model the neural tube defects by treating Rock inhibitor and Folic acid antagonist. Interestingly, those two treatments demonstrated distinct patterns of defects. (7/n)

Importantly, this “pattern-and-mix” strategy could robustly induce these self-organization processes with >90% efficiency. (6/n)

scRNA-seq revealed that the AP-gastruloid contained 26 cell types (CS 11 stage), including regionalized neural crest (cranial, vagal, trunk), head and paraxial mesoderm, cardiomyocytes, renal epithelium, and early neurons. (5/n)

The two aggregates immediately fused and began to self-organize as a single structure. Anterior (WNT-free) cells intercalated to the center of the posterior aggregate to form hindbrain, a continuous neural tube, and bilaterally segmented somites. (4/n)

We reasoned that homogeneous supplementation of the WNT signal, which activates the gastrulation programme, would simultaneously activate posterior embryonic fate. Therefore, we induced two populations (WNT-free and WNT-treated) and combined them, and we were surprised (3/n).

Stem-cell-based embryo models (SCBEMs), including Gastruloid models, have pioneered a new avenue in the field of experimental developmental biology. However, the existing human gastruloid models primarily capture posterior embryonic fates. (2/n)

Apologies for the oversight, we will fix it soon! Thank you for catching it!