Over the past decade, advances in bioengineering have led to a new-found appreciation of the effects of mechanical force on stem cells. Micrometre-scale culture systems that can subject cells to highly specific physical deformations have allowed researchers to demonstrate that force can modulate stem-cell behaviours, and even prime stem cells for therapeutic transplantation1,2. However, even the most advanced culture systems merely approximate the complex and dynamic forces that stem cells experience in their native tissues. In a paper in Nature, He et al.3 combine sophisticated genetic approaches and innovative physical manipulations to investigate the role of force on stem cells in vivo. They make the striking discovery that mechanical force drives the differentiation of a specialized population of progenitor cells in the midgut of adult fruit flies (Drosophila melanogaster).
The fruit-fly midgut is equivalent to the stomach and small intestine of vertebrates. All digestive organs experience physical forces that are inherent in their physiological function: ingested food distends the gut, and muscle contractions compress it. These forces continuously deform the gut’s epithelial lining, which includes both mature, differentiated cells (absorptive enterocytes and hormone-secreting enteroendocrine cells) and progenitor cells (stem cells and immature daughters that are committed to, but have not yet adopted, a particular differentiated identity).
He et al. identified a population of progenitor cells in the fly midgut that expresses the stretch-sensitive channel Piezo — a membrane-spanning, trimeric protein complex that opens in response to mechanical stimulation to allow the passage of ions across the membrane4. To trace these cells in vivo, the authors genetically engineered the Piezo-expressing cells such that they, and any cells that arose from them, produced a fluorescent protein. This analysis revealed that the Piezo-expressing cells mature into enteroendocrine cells. He and colleagues therefore named the population enteroendocrine precursors.
The fact that enteroendocrine precursors express Piezo suggested that they might respond to mechanical stimuli. The authors tested this possibility using two approaches. First, they distended the gut tube by feeding flies a diet containing indigestible methylcellulose. Second, they compressed gut tubes ex vivo using a microfluidic device. It has been established4 that mechano-activation of Piezo causes calcium ions (Ca2+) to enter the cell’s cytoplasm, and He et al. found that Ca2+ levels were significantly elevated in enteroendocrine precursors in both distended and compressed midguts. Crucially, Ca2+ levels were not elevated in distended or compressed midguts lacking the Piezo gene. These experiments convincingly demonstrated that the Piezo channel mediates Ca2+ influx in enteroendocrine precursors in response to mechanical stimuli (Fig. 1).
He et al. also observed that the midguts of Piezo-mutant flies failed to maintain normal numbers of enteroendocrine cells. This led the authors to hypothesize that Piezo-mediated Ca2+ influx might promote enteroendocrine differentiation. Consistent with this hypothesis, methylcellulose-distended midguts accumulated an excess of enteroendocrine cells. This effect required both Piezo and Ca2+ influx, and could be replicated in Piezo mutants by various genetic manipulations that increased cytoplasmic Ca2+ levels.
Taken together, these results support a scenario in which the activation of Piezo by mechanical force spurs enteroendocrine precursors to differentiate. Future investigation should reveal the physiological purpose of mechanosensitive enteroendocrine-cell production. Until then, one possibility is that a larger population of enteroendocrine cells can more-efficiently produce the myriad hormones that coordinate local and systemic responses to ingested food.
How exactly do elevated Ca2+ levels promote enteroendocrine differentiation? Such differentiation requires limiting the activity of a membrane-spanning receptor protein called Notch5. He et al. found that elevated Ca2+ levels act to inhibit Notch in enteroendocrine precursors, thus permitting their differentiation.
Interestingly, this sensitivity of Notch to Ca2+ might be specific to enteroendocrine precursors — the authors found that levels of Ca2+ had no effect on Notch activation or differentiation in enterocyte precursor cells, and others have shown6 that the same is true of midgut stem cells. Instead, levels of Ca2+ in stem cells fluctuate in response to extrinsic inputs, such as nutrients, injury and stress, to control a switch between resting and proliferative states. These stark contrasts between enteroendocrine precursors, enterocyte precursors and stem cells raise intriguing questions about how different types of progenitor cell interpret the same chemical signal.
In addition, it seems that enteroendocrine differentiation is not the only cellular behaviour in the midgut to be affected by mechanical force. He et al. observed substantially more cell division in methylcellulose-distended midguts than in controls. However, they also found that enteroendocrine precursors rarely divide, at least under normal circumstances. One explanation that could reconcile these findings is that enteroendocrine precursors divide specifically in response to force. Another is that other, yet-unidentified, mechanosensitive progenitors exist in the midgut, and that they use a different mechanism to sense force and divide.
The identification of enteroendocrine precursors touches on a broader theme in stem-cell biology: the existence of progenitors that are specialized for specific stimuli. Perhaps the best-understood example is injury-inducible ‘reserve’ stem cells, which are normally in a resting state but become activated by damage7. Now, He and colleagues add mechanical force to the list of stimuli that are associated with specialized progenitors. Many other stimuli might also populate this list. Indeed, recent studies8–10 using single-cell RNA sequencing have found disarming diversity among mouse intestinal stem cells, at least at the level of gene expression. This diversity might hint that we have seen just the tip of the iceberg in terms of progenitor specialization.
Are there mechano-responsive progenitors in other organs? Three mammalian organs — the intestine, lungs and skeletal muscle — would be attractive places to look. Like the fly midgut, these organs are both supported by progenitor cells and regularly subject to mechanical forces. A first step could simply be to examine them for progenitors that express Piezo or other mechanosensory channels.
Moving forward, a lack of microscale tools and protocols to manipulate force in adult tissues in vivo is likely to prove a bottleneck to progress. Such customized manipulations, which were crucial to the work of He and colleagues, are not easily translated between different organ systems. Perhaps the growing sophistication of three-dimensional culture systems and organ-on-a-chip devices will, over time, aid the development of technologies for mechanical manipulation of adult organs. More advances in this exciting area of research will no doubt reveal fundamental aspects of adult organ maintenance and improve strategies for tissue engineering.
Nature 555, 34-36 (2018)