Hippo mechanosignaling emerges as a gatekeeper of endothelial identity
Human pluripotent stem cells do not become endothelial cells by following a single straight path. They face branching decisions at each step, and those decisions are enforced by competing gene programs that either preserve one fate or push the cell toward another. In this preprint, investigators built a framework to chart those choices at scale, combining genome-wide CRISPR perturbation with single-cell RNA sequencing and gene program inference to identify regulators of differentiation as they act in real time.
The work focused on human arterial endothelial differentiation, where the transition from progenitor to vascular identity depends on precise timing and transcriptional coordination. Using circular capture Perturb-seq, or CC-Perturb-seq, the team improved guide RNA detection in split-pool combinatorial indexing workflows and used that gain to connect individual perturbations to downstream gene programs. The result was a regulatory map that did more than list candidate genes. It placed each regulator in the developmental sequence and showed when a perturbation redirected cells toward a different fate.
One result stood out immediately. Hippo mechanosignaling proved necessary for endothelial specification.
A single-cell perturbation map of human vascular fate choice
The framework integrated genome-wide CRISPR screens, single-cell readouts, and computational inference to identify regulators that promote or inhibit endothelial cell differentiation. Across the differentiation process, the investigators tracked gene programs through three stages and identified 174 regulators linked to endothelial fate acquisition. Some pushed cells toward endothelial identity, while others suppressed it or diverted progenitors into alternate developmental trajectories.
This design matters because differentiation is not only about whether a marker is turned on. It is about when a program starts, how strongly it rises, and whether another lineage program is suppressed in parallel. Single-cell resolution exposes those decisions far better than bulk assays, where opposing cell states can blur into the average. By layering perturbations onto lineage progression, the study captured the regulatory wiring behind fate restriction rather than merely cataloging expression changes after the fact.
The endothelial lineage in this model depends on ETV2, a master transcriptional regulator of endothelial identity. The screens placed Hippo pathway activity upstream of that node. During lateral mesoderm differentiation, Hippo mechanosignaling was required to activate ETV2, positioning the pathway at a biologically consequential checkpoint before vascular identity becomes fixed.
ETV2 activation failed when Hippo signaling was blocked
The mechanistic implication is direct: without Hippo signaling, endothelial fate did not hold. Blocking the pathway prevented ETV2 induction even in the presence of VEGF, the classic pro-endothelial cue used in differentiation systems. Cells that should have entered the endothelial program instead shifted toward a mesenchymal fate in vitro. In developmental terms, the pathway did not merely tune efficiency; it controlled lineage choice.
That divergence is especially important because VEGF often functions as a dominant signaling input in vascular differentiation models. Here, it was not enough on its own. The data suggest that the vascular program requires a permissive mechanosensitive state in addition to soluble growth factor signaling. The biologic logic is plausible: endothelial cells arise in a physical environment where tissue tension, substrate properties, and cell shape can shape transcriptional competence. Hippo signaling is well suited to integrate those mechanical cues and convert them into nuclear transcriptional responses.
The study also extended beyond human stem-cell culture. In developing zebrafish embryos, Hippo signaling was required for maximal etv2 expression. Cross-species concordance strengthens the argument that this is not a quirk of one in vitro system but a conserved developmental requirement. A mechanism that appears in both a mammalian stem-cell model and a vertebrate embryo carries more weight than a single-system observation alone.
Pharmacologic Hippo activation improved endothelial purity
The reverse experiment was equally informative. Pharmacologic activation of the Hippo pathway increased the purity of both arterial and venous endothelial differentiation in vitro. Instead of simply asking whether cells could become endothelial, the investigators asked whether the pathway could sharpen lineage output. It could. The practical consequence for stem-cell biology is straightforward: manipulating mechanosignaling may improve the yield and fidelity of vascular differentiation protocols.
For translational work, that kind of gain is not cosmetic. Differentiation protocols often struggle with mixed populations, incomplete specification, and residual mesenchymal or off-target cells. A pathway that improves purity while reinforcing endothelial identity may reduce downstream noise in disease modeling, drug testing, and potentially cell-based therapeutic manufacturing. The preprint suggests that lineage quality can be improved not only by adding growth factors, but by controlling the mechanical signaling state that permits the endothelial program to run.
The broader regulatory map is equally valuable. Identifying 174 regulators across the differentiation sequence provides a catalog of intervention points, but the more important advance is temporal resolution. Regulators were not just found; they were placed into a developmental order that reveals whether they act early to open competence, midstream to reinforce identity, or later to stabilize the emerging endothelial state.
What this means for developmental biology and stem-cell engineering
This work reframes endothelial differentiation as a process shaped by mechanical signaling as much as by canonical vascular cues. Hippo mechanosignaling sits at the intersection of tissue mechanics and transcriptional control, and here it appears to function as an upstream license for ETV2 activation. Without that license, progenitors do not simply differentiate inefficiently; they take a different route altogether.
The framework itself is also a methodological advance. CC-Perturb-seq addresses a technical bottleneck in split-pool combinatorial indexing by improving gRNA detection, which makes large perturbation studies more tractable. That matters because cell-fate decisions are rarely governed by one gene or one pathway. Mapping the regulatory network requires enough scale to test many candidates while still reading out their effect at single-cell resolution.
Several limitations are implicit in the preprint format. The findings have not yet been certified by peer review, and the most decisive evidence still comes from experimental models rather than human in vivo vascular development. The mechanistic link between Hippo signaling, ETV2 activation, and mesenchymal diversion is compelling, but the study leaves open how these signals are integrated molecularly at chromatin and transcription-factor levels. It also remains unclear which upstream mechanical inputs most strongly tune the pathway during differentiation and whether the same dependency applies across broader endothelial subtypes beyond arterial and venous contexts.
Even so, the central signal is hard to miss: endothelial identity depends on more than VEGF and lineage markers. It requires a mechanosensitive transcriptional state that allows ETV2 to turn on at the right moment, in the right cells, with the right strength. When that state is disrupted, progenitors veer away from the endothelial program and into another fate entirely.