Centrosome Loss Alters Axonal and Muscle Development

Beatriz González, Júlia Sellés-Altés, Jèrica Pla-Parron, Judith Castro-Ribera, Sofia J. Araújo

Posted on: 20 October 2026

Preprint posted on 14 September 2026

They said flies could develop without centrosomes. The embryos may beg to differ.

Selected by Jeny Jose

Categories: cell biology, developmental biology, neuroscience

When we train at the gym, we think about building muscle. Yet a muscle does nothing on its own. Every contraction starts with a signal from a neuron, a nerve cell made up of a cell body and long projections that carry signals. Its main projection, the axon, carries signals away from the cell body to other cells. Motor neurons are the neurons whose axons reach muscles, and each of those axons had to find its way to its exact muscle while the body was still forming.

For an axon to grow long, straight and in the right direction, the neuron relies on an internal scaffold of microtubules. And in many cells, a tiny structure called the centrosome organizes those microtubules. The centrosome is best known for its role in cell division. What it contributes once a cell has stopped dividing, in a neuron extending its first axon towards a muscle, has been far less clear.

In fruit flies, the question seemed largely settled. Flies lacking the centrosome protein Sas-4 can develop to adulthood. Yet these adults are far from healthy (Figure 1A). Turns out, the survivors weren’t telling the whole story. The answer starts in the embryo. Mutant embryos start out with Sas-4 supplied by their mother but lose Sas-4 – and therefore their centrosomes – gradually as development proceeds. By late embryonic stages, centrosomes are thought to be undetectable in 50-80% of cells, just as nerves and muscles are taking shape. González and colleagues went back to these mutants and looked at the stages when motor axons first grow out and find their muscles. They found that without centrosomes, axons lose their way and grow wiggly, and the muscles that guide them develop abnormally too.

Key Findings

How to study the effects of removing centrosomes from a cell?

In their preprinted study, González and colleagues used two tools to achieve this goal. They looked at mutants lacking Sas-4 in every cell. Also, they used RNAi knockdown reducing Sas-4 only in chosen cells, using “drivers” active in a specific cell type (the Gal4/UAS system). Here, they targeted pioneer motor neurons, sensory neurons or muscle cells. Comparing the two showed the authors where each defect comes from. Restoring Sas-4 throughout the embryo then tested whether the defects could be rescued

Do mutant embryos survive?

As previously reported, Sas-4 mutants could reach adulthood. But only half of them did. The other half died as embryos. After that point, the proportion held steady through the larval, pupal and adult stages. The authors concluded that once embryonic stages are passed, centrosomes are no longer needed to reach adulthood.

Do their nerves reach the right muscles?

Bundles of motor axons form nerves. In the fly embryo, the intersegmental nerve (ISN) targets the dorsal muscles and the segmental nerve (SN) the ventral ones (Figure 1C).

About half of Sas-4 mutant embryos (52.8%) had ISN defects. Their axons innervated the wrong targets or none at all, strayed from their normal routes, or crossed into neighbouring segments, and severity varied widely between embryos (Figure 1D-G). Sensory axons were misrouted in 55-60% of embryos. The authors describe these defects as mild. Axons were still present, and the defects appeared in only a proportion of embryos, at varying severity (incomplete penetrance and variable expressivity).

Restoring Sas-4 partially rescued both motor and sensory defects. Embryos lacking Sas-6 showed motor axon defects at similar rates (54-59%), suggesting the defects come from losing centrosomes in general.

Figure 1. Figure 1 from the preprint – (A) A Sas-4 homozygous adult fly 6 hours after emerging as an adult; surviving mutant adults barely moved and failed to extend their wings. (C) The three main motor nerves in two embryonic segments: transverse nerve (TN, green), segmental nerve (SN, blue) and intersegmental nerve (ISN, magenta). (D-G) Motor nerves in control and Sas-4 loss-of-function embryos, with defects marked. Preprint figure 1 made available under a CC-BY-NC-ND 4.0 International license

Is the fault in the neuron or in the muscle?

The mutants used in this study lack Sas-4 everywhere, so the defects could arise in the neurons, in the tissues around them, or both. Reducing Sas-4 only in pioneer motor neurons caused defects in 33.3% of embryos. These included misrouting, shorter nerves and problems with how nerves bundle together (fasciculation/defasciculation) (Figure 2). Reducing Sas-4 only in sensory neurons also caused misrouting. Part of the problem, then, lies within the neurons.

Only part of it, though. The pioneer neuron knockdown did not reproduce the full strength of the mutant defects (33.3% of embryos, compared with 52.8%). So the authors turned to the muscles, which provide the cues that guide motor axons to their targets. In Sas-4 mutants, the regular pattern of 30 muscle fibres per body segment broke down, with fibres misplaced or missing. Laid side by side, the two maps lined up. Misrouted axons often sat exactly where muscles were disrupted, and every mutant embryo with muscle defects also had ISN defects.

Reducing Sas-4 only in muscle caused muscle defects in 80% of embryos, and in 63% of cases the nerve defects matched the location of the muscle defects. Reducing it only in pioneer neurons left the muscles almost untouched. Since innervation does not shape muscle development in fly embryos, the authors concluded that axons go astray for two reasons: centrosomes are missing from the neurons and from the muscles that guide them.

Figure 2. Figure 2 from the preprint – Stage 16 embryos stained for muscle (myosin, green) and motor axons (FasII, magenta) in control and Sas-4 loss-of-function conditions (D″-G″). Asterisks mark misrouted axons at sites of muscle disruption. Made available under a CC-BY-NC-ND 4.0 International license

Do young neurons have centrosomes and, if so, where?

If centrosome loss acts within neurons, these neurons should have centrosomes to lose. Earlier studies concluding that fly neurons organize microtubules without centrosomes mostly looked at later stages, after axons had formed. So the authors examined pioneer neurons at stages 12-13, at which point their axons first grow out. Centrosomes were there, and they weren’t randomly placed. In one of these neurons, the centrosome sat in the same quarter of the cell as the emerging axon in 72% of cases, and never in the opposite quarter. In Sas–4 mutants, centrosome numbers in these neurons fell roughly threefold.

Sensory neurons told a similar story. Most carried centrosomes, which overlapped with microtubule markers, while Sas-4 mutant embryos had only about one-third of the normal number of centrioles. Mutant embryos without visible defects had slightly more centrosomes than those with defects, though still less than half the normal number, suggesting that centrosomes are not strictly required for neuronal development here, but may contribute to its robustness.

Are those centrosomes working?

Finding a centrosome does not show that it is doing anything. In pioneer neurons, centrosomes overlapped with stable microtubules in around 70% of cases. In about half, they also overlapped with γ-tubulin, the protein that starts new microtubules. The more direct test was to break the scaffold and watch it rebuild. After the embryos were chilled to dismantle their microtubules, new ones radiated from the centrosome within 90 seconds in 61% of control neurons, rising to 77% after 5 minutes. Mutant neurons still rebuilt their microtubules, but more slowly: 12.5% at 90 seconds and 50% at 5 minutes. The authors concluded that these centrosomes are active. Neurons can build microtubules without them, but the centrosome may help that happen efficiently and on time.

What happens to axon shape without them?

Axons without centrosomes still grow. But normally?

Beyond taking wrong turns, axons in Sas-4 mutants also changed shape. Axonal tortuosity, or “wiggliness,” describes axons that stray from straight paths, which has been associated with neurodegeneration. In controls, 6-8% of pioneer axons were wiggly (Figure 3). In Sas-4 mutants, and in embryos with Sas-4 reduced only in these neurons, 23-30% were. The clearest link came from scoring individual neurons. 54% of those without a centrosome had wiggly axons, compared with 12% of those that still had one. JNK signalling, a known cause of wiggly axons, was not elevated.

Figure 3. Figure 7 from the preprint – Pioneer motor neuron axons in control, Sas-4 mutant and Sas-4 RNAi embryos (C-E), with asterisks marking wiggly axons. Panels I-K show single neurons: a control neuron with a centrosome (arrowhead), and mutant and RNAi neurons without one, with wiggly axons (asterisks). Made available under a CC-BY-NC-ND 4.0 International license.

Putting it together

Figure 4. Figure 8 from the preprint – Centrosome loss disrupts axon development by (A) impairing muscle development and (B) the neuron’s microtubule cytoskeleton. Made available under a CC-BY-NC-ND 4.0 International license.

The authors link the variability to how unevenly centrosomes are likely to be shared out as cells divide. Nervous system cells keep dividing after many others have stopped, and muscle cells divide until embryonic stage 10, so the maternal Sas-4 supply may run lower with each division. They also note that wiggly axons are, to their knowledge, the first axon shape defect linked to centrosome loss in flies. Since tortuosity is a hallmark of neurodegeneration, they propose it as an entry point for studying disease progression. Centrosome loss in humans is linked to conditions including microcephaly and hereditary spastic paraplegia, and the authors suggest the fly could help uncover the mechanisms behind them.

Why this preprint is worth your attention

What makes this preprint exciting is that it revisits a result many considered settled. By looking earlier and counting carefully, the authors show that half of the flies once thought to manage without centrosomes never make it past the embryo stage.

It also changes what “dispensable” means. Neurons without centrosomes still grow axons and rebuild their microtubules, but they rebuild more slowly, and their axons are more often misrouted and wiggly (Figure 4). The authors suggest that centrosomes matter most in the narrow window when axons first grow out. And the wiring problem reaches beyond the neuron, to the muscles that guide it.

What’s still missing is a clean separation of cause and effect. Because centrosomes disappear gradually and unevenly, each mutant embryo is a different mix of cells with and without them. The neuron-by-neuron link to wiggly axons, while strong, remains an association.

The takeaway:

Neurons can grow axons without centrosomes. With them, microtubules rebuild faster, and axons grow straighter and to the right place.

Glossary

Pioneer neurons: neurons that extend growth cones along a future fibre tract, interpreting environmental cues to define the nerve’s path.
Penetrance: the proportion of individuals carrying a genotype who show the associated phenotype.
Expressivity: how severe the phenotype is in individuals that show it.

Tags: axon guidance, centrosome, drosophila, muscle development, neurons

Read preprint 1 Star (No Ratings Yet)

Authors' response

Dr. Sofia J. Araújo & Dr. Beatriz Gonzalez shared

Could an acute, cell-specific way of removing centrosomes at stages 12-13 separate the effect of timing from gradual depletion?

Yes, acute, cell-specific centrosome removal at stages 12–13 could help distinguish the effects of the timing of centrosome loss from those of gradual depletion. Laser ablation could be an option, although it is technically challenging and would require tracking the same neurons and their axons throughout subsequent embryonic development. An alternative approach would be the inducible degradation of centrosomal proteins. However, this strategy may not fully abolish all centrosome functions and would need to act rapidly and efficiently to disrupt centrosome activity during this specific developmental window. Importantly, acute centrosome removal differs from inhibiting centrosomal activity, which would test specific functions rather than the role of the organelle as a whole. However, this approach may not be fundamentally different from analyzing individual neurons that lack centrosomes, as we already do, and its added value would depend on whether it allows us to isolate a specific developmental window.

Do the wiggly axons persist into larval and adult stages, and could they contribute to the uncoordinated movement of surviving adults?

It would be very interesting to determine whether these aberrant axonal trajectories persist into the larval and adult stages and contribute to the locomotor defects observed in surviving adults. It would also be important to assess whether surviving larvae exhibit locomotor abnormalities. Integrating axonal analyses across developmental stages with behavioural assessments could help establish whether axonal abnormalities correlate with locomotor dysfunction and how these relationships evolve during development. This possibility is also relevant to human neurological disease: mutations affecting microtubule organization and axonal maintenance, such as SPAST mutations associated with hereditary spastic paraplegia, can cause motor dysfunction. Although this does not establish a direct link between centrosome loss and human ataxia, it supports the broader idea that defects in axonal organization may have lasting functional consequences.

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 cell biology category:

The Tensile Expansion Microscopy (TExM) cell stretcher: an iris expansion device integrated with automated real-time autofocus and tracking to super-resolve cells

Ramita Arampongpun, Tejasvin Shrikanth, Vignesh Venkataramani, et al.

Selected by 19 October 2026

Felipe Del Valle Batalla

HiExM Enables Scalable Mapping of Organelle Morphology and Spatial Heterogeneity

John H. Day, John Farrell, Dandan Yang, et al.

Selected by 19 October 2026

Dina Kabbara

Discussion

Fast single molecule localization microscopy in 3D

Edward S. Allgeyer, Jennifer H. Richens, Jeanne Lefévère-Laoide, et al.

Selected by 14 October 2026

Kanishka Parashar

Also in the developmental biology category: