Mechanically-induced Septin Networks Protect Nuclear Integrity

Margaret E. Utgaard, Alexia Caillier, Shreya Chandrasekar, Joseph J. Tidei, Harikrushnan Balasubramanian, Rachel M. Lee, Owen Puls, Satya Khuon, Jesse Aaron, Teng-Leong Chew, Jordan R. Beach, Patrick W. Oakes

Posted on: 25 May 2026

Preprint posted on 21 January 2026

Septins come to the rescue to form mechanical barriers at the cortex and protect the nucleus from the chaos around the cell.

Selected by Filipe Nunes Vicente

Categories: biophysics, cell biology

Background

Imagine for a second that you are attending a very busy, crowded concert. If you want to move down to the front rows to get closer to your favourite artist, you’ll have to squeeze through gaps of angry (or kind) people. If you stay in place or dance like a maniac, you will either get trapped by large groups or crash into absent-minded concertgoers. Depending on the concert, you might even find yourself in the middle of a mosh pit, an extreme form of crowding.

Such crowded, chaotic environments also exist at the cellular level during migration, apical constriction, or stretching. Cells need to navigate and adapt to these environments, while withstanding continuous pressure or confinement forces. Cell migration is particularly similar to a very crowded concert, as cells must squeeze through narrow constrictions between other cells or the matrix to reach their desired target.

The nucleus, due to its size and stiffness, is particularly sensitive to mechanical forces in such environments. Moreover, the nucleus itself can mediate “cell proprioception” and shape sensing through mechanotransduction pathways (1); however, excessive deformation can lead to DNA damage and apoptosis. As the main regulator of cell shape and mechanics, the cytoskeleton is essential to buffer the nucleus from mechanical stress and retain its physiological mechanosensitive abilities. This will likely require specialized, interconnected networks with special mechanical properties, resulting from the “crosstalk” between different elements of the cytoskeleton.

In this regard, septins could provide a missing link between cytoskeleton crosstalk and mechanical protection of the nucleus. These proteins form crosslinked networks which bind both actin and microtubules, and display high affinity for intracellular forces, leading to their accumulation in the contractile ring or stress fibers (2).

In this preprint, the authors have combined high-resolution cell imaging with genetic perturbations and confinement assays to unveil the role of septin networks in nuclear mechanoprotection.

Key findings

Figure 1. Dynamics and organization of septin networks. A) Structured Illumination Microscopy (SIM) super-resolution images showing how septin network (magenta) forms both on the ventral and dorsal side of the nucleus (blue) in conjunction with the actin cortex (green). Adapted from Fig. 2A. B) Representation of the spatial coupling between septin (magenta) and the nucleus (blue). Adapted from Fig. 1C. C) Mechanosensitive assembly of septin under the nucleus after dynamic confinement at 5 µm. Adapted from Fig. 3A. D) When septin is depleted (shRNA), nucleus undergoes substantial rupture upon harsh 2.5 µm confinement (lower panels) compared to WT condition (upper panels). Adapted from Fig. 6E. Figure under aCC-BY-NC-ND 4.0 International license. With permission of the authors.

Septins form a perinuclear ventral and dorsal dynamic network at the cortex:

The authors started by investigating the organisation and subcellular distribution of septin networks by imaging endogenously tagged septin with confocal microscopy. Similar to previous observations, septin accumulated primarily under the nucleus in alignment with stress fibres, although a dorsal network could also be observed. The ventral septin network was spatially coupled to the nucleus and dynamically organised during migration.

Structured Illumination Microscopy (SIM) confirmed that the septin filaments are located below and above the nucleus, but not around it, and extend towards the cortex (Fig.1A). Super-resolution Interferometric Photoactivation and Localization Microscopy (iPALM), which allows 3D nanometric single-molecule localisation, confirmed that septins are consistently positioned below actin filaments. Therefore, septin networks are perinuclear, localise primarily at the cortex, and are spatially separated from actin.

Septin network formation is induced by external and internal forces

The existence of a dynamic perinuclear and subcortical septin network, given the high affinity of septin to forces, suggests that this network could also be mechanosensitive. The authors addressed this hypothesis from two perspectives: external forces and internal forces. For the external forces, dynamic confinement, which activates nuclear mechanosensing pathways (1), led to delayed septin accumulation under the nucleus (Fig. 1C). To test the role of internal forces, the authors imaged septin dynamics during migration through microfluidic channels, with septin accumulating along the membrane at both squeezing points and device walls. In both scenarios, septin networks exhibited mechanosensitive behaviour rather than simply becoming deformed by forces.

Septin network assembly is driven by mechanical signals

Despite the mechanosensitive responses of the septin network, the nucleus could still be the primary driver of septin accumulation, since enucleated cells displayed less septin alignment in the cell centre. However, in a very creative experiment, the authors added large glass beads to the cells to create a “fake” nucleus, creating a spatial coupling without activating nuclear signalling. In enucleated cells, the “fake” nucleus partially rescued central septin assembly, thereby showing the importance of mechanical signals in the formation of perinuclear septin networks.

Septin network buffers the nucleus from mechanical stress

Since the septin network is both mechanosensitive and spatially coupled to the nucleus, an important question is: could it also protect the nucleus from mechanical strain? To assess this, the authors harnessed the power of dynamic confinement at narrower spaces, which triggered nuclear damage. When septin was depleted, cells exhibited a much higher percentage of nuclear rupture events compared to controls (Fig. 1D), confirming that septins act as a buffer to protect the nucleus from mechanical strain.

What I like about the preprint

This preprint highlights the importance of cytoskeleton “crosstalkers” in nuclear protection within the context of a crowded microenvironment. Moreover, it provides a novel connection between septin organisation and mechanical function, which could explain previously observed septin-driven reinforcement of dendritic spines (3) or the actin cortex at bacterial infection sites (4). The notion of spatial coupling is particularly exciting because it highlights a mechanical role of the nucleus in guiding septin accumulation. Given the role of the nucleus in cell proprioception and shape sensing, the septin perinuclear network could be a further player in such processes.

Although this work was carried out in fibroblasts, septins are also found on other highly motile cell types, such as T-cells or spermatozoa, that are likely to encounter mechanical challenges in physiological conditions. The results in this study could thus help understand how these cells maintain their nuclear integrity during migration in confined environments. Another preprint on septins was released concurrently with this work, showing how septins protect against genomic instability during confined migration and that overexpression of septins could be mechanoprotective for cancer cells (5). The dynamics, structure, and spatial coupling of the septin network could also be important elements in ensuring mechanoprotection during metastasis.

References

Venturini, V. et al. The nucleus measures shape changes for cellular proprioception to control dynamic cell behavior. Science 370, (2020).
Lam, M. & Calvo, F. Regulation of mechanotransduction: Emerging roles for septins. Cytoskeleton (Hoboken) 76, 115–122 (2019).
Ewers, H. et al. A Septin-Dependent Diffusion Barrier at Dendritic Spine Necks. PLOS ONE 9, e113916 (2014).
Aigal, S. et al. Septin barriers protect mammalian host cells against Pseudomonas aeruginosa invasion. Cell Reports 41, (2022).
Mayca-Pozo, F., Butts, S. M., Schaefer, A. W., Montagna, C. & Spiliotis, E. T. Septins buffer actomyosin forces to protect the nucleus from genotoxic mechanical stress. 2026.01.21.700789 Preprint at https://doi.org/10.64898/2026.01.21.700789 (2026).

Read preprint

(No Ratings Yet)

Author's response

Patrick W. Oakes shared

1.The iPALM experiments reveal a consistent gap between actin and septin filaments with a median offset of 45 nm. What could drive the formation and maintenance of this gap? Moreover, this offset is measured in normally spread cells, hence it could be interesting to assess whether it would change depending on mechanical stimulation, such as confinement.

These are great questions. First, we note the noise in that measurement, as we observe a wide spread of distances (about 0 – 100 nm) between septin and actin, so we don’t think there is a 45 nm long binding partner. Also, even in the example in the figure, you can see areas in XZ where septin signal is encroaching into the actin signal. So it’s pretty clear that this varies within the network itself. While it’s possible/probable there are bridge proteins, which is something we’d like to explore in future work, it’s also possible that direct actin-septin binding maintains the connection. Septins have known membrane binding capacity, so it makes sense that they are closer to the membrane. It’s possible this sets the septin orientation to place the actin binding or bridge binding domain away from the membrane, establishing the observed PM-septin-actin arrangement.

While we applaud the creativity, doing iPALM within the dynamic confiner setup is unfortunately impossible as it requires two objectives positioned opposite from each other and is only possible on very thin samples.

2.The mechanosensitive accumulation of septin appears robust and can be triggered both by external/internal forces and spatial coupling to the nucleus. However, it is unclear how this accumulation happens: is this dependent on actomyosin networks or does it arise from structural and mechanical properties of the septin network itself?

This is again something that we are excited to pursue in the future. We don’t know exactly what signaling is occurring, though suspect that the mechanosensation is occurring upstream of septin accumulation instead of a direct mechanosensation by the septins themselves. The actomyosin network is probably also playing some role, as treatment with actin destabilizing drugs causes the subnuclear network to disassemble and the septins to relocate to the peripheral membrane.

3. The septin network and its respective mechanical properties were extensively characterised in fibroblasts. However, there are other cell types which undergo faster and more frequent migration through confined, narrow spaces, such as T cells. Therefore, would the spatial organisation and mechanosensitivity of septin networks vary across cell types depending on their deformability and motility? This logic could be carried over to metastatic cancer cells where septins appear to be mechanoprotective; could the network be even more robust in such cases?

We think you’re making some excellent connections here. While we did all our work in fibroblasts, we know these septin networks form in other cell types. Previous works from Matthew Krummel’s lab (1) and Erdem Tabdanov’s lab (2) have showed that septins play crucial roles in T cell motility, and if you look at the images in their papers you see that they localize in networks right around the nucleus. Therefore we think this is a universal mechanism, and that would fit with it being a mechanical response to force on the cortex. As you mention, the concurrent work from Elias’s Spiliotis’s lab (3) suggests that over-expression of SEPT9 in cancer cells could actually contribute to their metastasis. It really seems to be about striking the right balance in the cell.

Gilden JK, Peck S, Chen YC, Krummel MF. The septin cytoskeleton facilitates membrane retraction during motility and blebbing. J Cell Biol. 2012 Jan 9;196(1):103-14. doi: 10.1083/jcb.201105127. PMID: 22232702; PMCID: PMC3255977.
Zhovmer AS, Manning A, Smith C, Nguyen A, Prince O, Sáez PJ, Ma X, Tsygankov D, Cartagena-Rivera AX, Singh NA, Singh RK, Tabdanov ED. Septins provide microenvironment sensing and cortical actomyosin partitioning in motile amoeboid T lymphocytes. Sci Adv. 2024 Jan 5;10(1):eadi1788. doi: 10.1126/sciadv.adi1788. Epub 2024 Jan 3. PMID: 38170778; PMCID: PMC13155490.
Mayca-Pozo, F., Butts, S. M., Schaefer, A. W., Montagna, C. & Spiliotis, E. T. Septins buffer actomyosin forces to protect the nucleus from genotoxic mechanical stress. 2026.01.21.700789 Preprint at https://doi.org/10.64898/2026.01.21.700789 (2026).

Have your say Cancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Sign up to customise the site to your preferences and to receive alerts

Also in the biophysics category:

Loss of Sun2 ablates nuclear mechanosensing-driven extracellular matrix production and mitigates lung fibrosis

Emma Carley, Sandra Sandria, Xueyan Peng, et al.

Selected by 13 May 2026

Beth Chopak

Shape independent fluidisation in epithelial monolayers

Pradip K Bera, Anh Q. Nguyen, Molly McCord, et al.

Selected by 29 April 2026

Sindhu Muthukrishnan

Discussion

Exogenous Amyloid Sequences: Their Role in Amyloid-Beta Heterotypic Aggregation

Jofre Seira Curto, Sandra Villegas, Maria Rosario Fernandez, et al.

Selected by 15 April 2026

Cyntia Alves Conceição, Marcus Oliveira

Discussion

Also in the cell biology category: