Sophie G. Martin

To ensure survival, cells need to buffer the effects of environmental stress on their plasma membrane, yet the structural mechanisms by which this is acutely achieved remain largely unknown. Here, we propose atonosomes as a unifying identity for a class of previously observed but enigmatic, tension-responsive, plasma membrane-derived compartments that arise across contexts of acute and chronic membrane tension loss. Leveraging unprecedented high resolution cryo-FIB-ET imaging in yeast, we show that atonosomes are complex, organelle-containing structures bounded by membranes and cell wall material, spanning hundreds of nanometers, and displaying a remarkable morphological diversity. Atonosomes form within seconds in response to reduced plasma membrane tension, and their emergence appears to require no dedicated molecular machinery, arising instead as a direct consequence of membrane biophysics. Upon formation, they recruit key membrane-associated proteins, including TORC2, Slm1, and septins. Under conditions of chronic disruption of PM homeostasis, atonosomes become constitutively present. Their stability and reversibility are further modulated by the cell wall, whose polymerization state influences atonosome dynamics. Structural conservation in fungi and ichthyosporea, demonstrates that atonosomes are a conserved stress-triggered response of cell-wall enclosed organisms. Together, these findings establish atonosomes as a novel compartment that mediates cellular responses to plasma membrane tension variation, coupling membrane remodeling and lipid homeostasis to preserve cellular integrity under stress.

The conserved Cdc42 GTPase is a key driver of symmetry breaking and polarized growth, forming zones of activity that locally recruit effectors to organize the cytoskeleton and polarize secretion. Here, we show that Cdc42 also functions in cell–cell fusion during Schizosaccharomyces pombe sexual reproduction, but concentrates at the fusion site through mechanisms distinct from those proposed in Saccharomyces cerevisiae . Notably, the cdc42-mCherry ^SW allele, which is functional for cell polarization and has been used across organisms for dynamic studies, exhibits a strong fusion defect. These cells block fusion before cell wall digestion but after actin fusion focus formation, indicating that Cdc42 is required to translate the vesicle cluster into polarized cargo delivery. We trace the defect to instability of Cdc42-mCherry ^SW and demonstrate that mating and cell fusion require higher Cdc42 protein levels than mitotic polarized growth. Remarkably, by constructing an allelic series driving Cdc42 expression over a 5-fold range, we discover that mitotic polarized growth responds linearly to Cdc42 protein levels, whereas mating exhibits a sharp switch-like response. We further trace this all-or-none response to pheromone-induced polarized growth. Thus, polarized growth in response to intrinsic or extrinsic cues exhibits distinct requirements to Cdc42 protein levels.

Mitogen-activated protein kinases (MAPK) are ancestral kinases that form essential signalling cascades. However, scaffolds that recruit kinases to subcellular locations and promote signal transduction have only been described in a few species. Notably, no scaffold was thought necessary for the MAPK cascade promoting sexual differentiation in fission yeast. Here, we identify the hemi-arrestin protein Sms1 as a novel scaffold of this MAPK cascade. Interactions with PIP2 and the pheromone receptor–coupled Gα subunit target Sms1 to plasma membrane patches, where it assembles the active cascade by binding each MAP kinase. These interactions are essential for signal transduction and local signal interpretation for polarised growth. Phosphorylation, including by the MAPK itself, antagonises Sms1 membrane translocation, establishing a negative feedback that underlies polarity patch turnover and prevents untimely mating attempts. Thus, Sms1 is a MAPK scaffold with canonical functions despite its distinct structural fold, highlighting convergent evolution of MAPK scaffolds across eukaryotes.

Cell–cell fusion is critical for the sexual life cycle, as it drives the unidirectional transition between haploid and diploid phases. In eukaryotes where fusion occurs between partners with cell wall, fusion requires not only plasma membrane merging but also local removal of the intervening cell wall. ^1–3 . However, the integrity of the cell wall, which resists the strong internal turgor pressure and is essential for survival ⁴ , is normally monitored by signaling pathways that promote repair upon damage ^5,6 . How cell wall integrity pathways allow localized cell wall degradation is not known. In the fission yeast Schizosaccharomyces pombe , we previously identified pak2Δ mutants that exhibit transient fusion: cells briefly fuse and exchange cytoplasmic contents that signal meiosis but then reseal their fusion pore. This reversal of fusion directionality induces meiosis in haploid cells, with catastrophic consequences ⁷ . Here, we show that PAK activity is essential for cell-cell fusion, with major contribution for Pak2 and a more minor role for the essential Pak1. Pak2 accumulates at the fusion site, where its kinase activity is controlled to promote both fusion pore opening and expansion, thereby preventing post-fusion pore closure. Strikingly, osmotic stabilization largely suppresses all PAK-associated fusion defects, pointing to a role in cell-wall remodeling rather than plasma-membrane merging. Indeed, correlative light and cryo–soft-X-ray tomography reveals that resealed pores in PAK mutants rebuilt an intact cell wall. Thus, PAKs enforce fusion directionality by antagonizing cell-wall repair mechanisms that otherwise restore separation between mating partners.

Many processes such as polarized growth and secretion require specific actin networks. In fungi, cell–cell fusion requires cell wall digestion mediated by local secretion of lytic enzymes. In Schizosaccharomyces pombe, the myosin V Myo52 transports enzyme-containing secretory vesicles on the actin fusion focus, an aster-like actin network assembled by the condensate-forming formin Fus1. The fusion focus also concentrates proteins regulating cell polarity, communication, cytoskeleton, exocytosis, and membrane merging. Here, using centroid tracking and averaging, we present a spatiotemporal map of the fusion site with 8-nm precision. We show that a pool of vesicles remains at constant distance from the membrane as the actin structure condenses. Unexpectedly, Myo52 detaches from this pool and colocalizes with Fus1 closer to the membrane. We show that Myo52 binds Fus1 and transports it along actin filaments, and that Myo52 and Fus1 actin assembly activity contribute to focus compaction. Thus, myosin V–driven transport of formin Fus1 along Fus1-nucleated actin filaments underlies positive feedback for actin aster formation.

Summary: Investment in sexual reproduction has long been seen as a fitness cost for asexual proliferation, raising the question of how cells evolve to accommodate cyclical demands of sexual and asexual cycles. We addressed this question by evolving Schizosaccharomyces pombe lacking the critical fusion gene Prm1 through 18 cycles of alternating meiosis and mitosis. Initially, transcriptional tweaks offered a fragile means to fuse despite Prm1’s absence. However, a telomeric 1% genome deletion that epigenetically silenced Clr5 ultimately locked in a latent, Prm1-independent fusion pathway involving the mating regulator Ste11 and the pheromone-MAPK scaffold Sms1. This adaptation, though efficient at restoring sexual reproduction, came at the cost of mitotic growth. By contrast, rationally constructing mutations in a “synthetic evolution” approach bypassed the trade-off, enabling simultaneous high fusion efficiency and near–wild-type proliferation. Our findings illustrate how epigenetic de-repression and genetic assimilation converge to create innovative solutions under cyclical selection pressures with trade-offs.

Summary: Mitogen-activated protein kinases (MAPK) are ancestral kinases that form essential signalling cascades. However, scaffolds that recruit kinases to subcellular locations and promote signal transduction have only been described in few species. Notably, no scaffold was thought necessary for the MAPK cascade promoting sexual differentiation in fission yeast. Here, we identify the hemi-arrestin protein Sms1 as a structurally novel scaffold of this MAPK cascade. Interactions with PIP2 and the pheromone receptor–coupled Gα subunit target Sms1 to plasma membrane patches, where it assembles the active cascade by binding each MAP kinase. These interactions are essential for signal transduction and local signal interpretation for polarised growth. Phosphorylation, including by the MAPK itself, antagonises Sms1 membrane translocation, establishing a negative feedback that underlies polarity patch turnover and prevents untimely mating attempts. Thus, Sms1 is a MAPK scaffold with canonical functions despite its distinct structural fold, highlighting convergent evolution of MAPK scaffolds across eukaryotes.