Spatiotemporally Controlled Matrix Softening Facilitates Deterministic Crypt Formation in iPSC-Derived Intestinal Organoids

Introduction:: Introduction: Intestinal organoids hold unique promise as a benchtop model, owing to their capacity to mimic key architectural and compositional metrics of their in vivo counterparts, including a biomimetic crypt-villus axis. In vitro, crypts emerge from spherical colonies as a result of local matrix remodeling. However, state-of-the-art methods to study crypt morphogenesis rely on uncontrolled cell-mediated matrix remodeling, stress relaxation, dynamic photosoftening, or dynamic hydrolytic softening, all of which result in spatially uncontrolled budding events and highly heterogeneous crypt structure, size, and frequency. This inherent stochasticity severely limits the capacity to study and control symmetry breaking and crypt bud initiation, thereby hindering downstream, broader use applications of organoids that require a high degree of reproducibility, including for drug testing applications and for use as cellular therapeutics. Further complicating matters, induced pluripotent stem cell (iPSC)-derived organoids do not readily form crypt structures, owing to their relatively immature (i.e., fetal-like) genetic signatures and insufficient cellular diversity. Taking into account these limitations, we sought to photopattern changes in matrix mechanics adjacent to growing organoid colonies at precise points in space and time to (1) mimic extracellular matrix (ECM)-associated changes in the developing intestine, (2) direct symmetry breaking of the organoid, and (3) enable user-defined formation of fully differentiated and functional crypts that match the spatial and compositional organization of human intestinal tissue.


Materials and Methods:: Materials and Methods: Human epithelium-only intestinal organoids (HDEs) were generated from iPSCs following established protocols¹. After differentiation and sorting for epithelial cells, organoid colonies were cultured in PEG-based hydrogels comprised of crosslinked networks formed by reacting dibenzylcycloocytne (DBCO)- and nitrobenzyl azide (NBA)-functionalized multi-arm PEG macromers (Fig. 1a, b)². Here, we took advantage of the use of the photolabile azide crosslinker, NBA, to photosoften regions adjacent to live organoid colonies using a laser scanning confocal microscope, which resulted in spatiotemporal variations in matrix mechanics (Fig. 1c).


Results, Conclusions, and Discussions:: Results: We first optimized PEG hydrogel initial stiffness and RGD concentration to facilitate iPSC-derived organoid colony growth (Fig. 1b). We then tuned the light dose, and associated degree of NBA cleavage and matrix softening, to reduce cytotoxicity and maximize crypt forming efficiency of HDEs. Using the optimal longwave UV light dose (405 nm; 1.58 µs pixel dwell, 16x averaging, 100% laser power), which resulted in a reduction to approximately 50% of the initial modulus (~0.5G’₀), we photosoftened regions of precise, user-defined dimensions to match the dimensions of human intestinal crypts (50 x 150 µm). We altered the spacing (i.e., the distance from the edge of the epithelium to the patterned regions) and the density of intestinal crypts (1, 2, 3, or 4 crypts per organoid), and determined that a (-)10 µm spacing with 4 crypts per organoid resulted in the highest crypt forming efficiency (Fig. 1d). Following local matrix softening, epithelial shape changes and symmetry breaking rapidly occurred, followed by the emergence of crypt-like structures populated by proliferative (Ki67+) cells. These engineered crypts matched the dimensions and proliferative cell regionalization of native human crypts (Fig. 1e).

Conclusions: Human epithelium-only intestinal organoids can grow as colonies in PEG-based hydrogels with optimized initial matrix stiffness and RGD concentration. After photopatterning and local matrix softening, crypt-like structures emerged, which were populated by proliferative cells, mimicking the cell regionalization that occurs in vivo. Moving forward, cell populations will be examined in detail, and maturation state of the organoids will be assessed.


Acknowledgements (Optional): : Acknowledgements: NIH R01 DK120921, NSF RECODE 2033723, NIH K99 DK135907

References (Optional): : References: ¹Spence, J., et al., Nature, 2010. ²Kloxin, A., et al., Science, 2009