Neural membrane repair at the core of regeneration

Loss of plasma membrane integrity can compromise cell functioning and viability. To counteract this eminent threat, eukaryotic cells have developed efficient repair mechanisms, which seem to have co-evolved with the emergence of vital membrane processes (Cooper and McNeil, 2015). This relationship between basic cellular functioning and membrane repair highlights the fundamental significance of preserving membrane integrity for cellular life.

Membrane injuries are common, resulting from various mechanical or biochemical stresses, and can affect different cell types under both normal and pathological conditions. Adequate membrane repair responses play a crucial role in preventing the development and progression of diseases (Dias and Nylandsted, 2021). For instance, invasive cancer cells exhibit excessive membrane repair activity, while in other conditions (such as muscular dystrophies, heart failure, neurodegenerative diseases, and bacterial infections) the endogenous membrane repair capacity that each cell is equipped with may become overwhelmed.

The plasma membrane of neural cells can become compromised during physiological aging due to metabolic stress, including oxidative stress, which becomes more pronounced in conditions of acute and chronic neurodegeneration. The oxidation of lipids (lipid peroxidation) compromises the rigidity of membranes and alters their net charge. At a cellular level, these biophysical changes result in excessive permeability, compromising ion homeostasis and signaling pathways. In acute neurodegenerative conditions, such as in brain traumas, the insult to membranes is not only chemical in nature, but also mechanical. Within seconds, the rapid increase in axonal strain causes transient membrane tearing. Cells exposed to sublethal forces activate a “pathological cascade” of events, which are driven by disruptions of the intracellular environment and are characterized by the appearance of focal axonal swellings. As these swellings increase in size, axolemmal disconnections occur (termed secondary axotomy). Membrane lesions that occur in chronic neurodegenerative diseases, on the other hand, are predominantly biochemical in nature. Disease proteins associated with proteinopathies (such as Alzheimer's disease and Parkinson's disease) interact deleteriously with the plasma membrane. The role of membrane integrity in neurons under physiological and pathological conditions has been reviewed (Dias and Nylandsted, 2021).

Given their post-mitotic state and limited regeneration capacity, neural cells are particularly vulnerable to lesions. One would expect that, at the expense of proliferative activity, neural cells are equipped with repair mechanisms. However, these mechanisms are still being discovered. Although mechanisms of membrane sealing are heavily conserved across cell types (reviewed in Hendricks and Shi, 2014), the vast majority of these have not been verified in neuronal models. Rather, only potential repair proteins have been associated with neuronal injury based on their expression pattern (summarized in Hendricks and Shi, 2014). Undoubtedly, neural membrane repair plays an integral role in regeneration following injury, although the complex process of neural regeneration is still under discovery (Figure 1).

Figure 1.

Neural regeneration in the aftermath of injury has different phases: stabilization, recovery, and re-growth.

Firstly, neural injury (e.g. axotomy, highlighted in yellow around the torn axon) compromises membrane integrity (represented by the dashed outline of the neuron in the schematic), resulting in the uncontrolled influx of calcium (depicted in red). This injury state is harmful and compromises neural viability and potential regeneration. To halt this influx, calcium-activated repair mechanisms reseal the plasma membrane to restore its integrity, with regards to barrier functions (reverting the “dashed outline”). Adequate membrane repair is key to initiating neural regeneration, bringing the abrupt decay of cellular homeostasis to a halt (as part of the stabilization phase) and allowing its recovery and potential re-growth, achieving “healthier” paradigms. Other than the fundamental restoration of its ionic balance, the recovery phase is marked by the injury-driven pro-regenerative transcriptional profile that is mounted and by the subsequent membrane repair that occurs (membrane restructuring), events that are likely to promote axonal re-growth. Increased ability to adequately restructure the membrane towards one that fosters axonal re-growth (e.g. capable of forming growth cones and extend) may, to some extent, promote the transition from the recovery phase to the re-growth phase. However, given the multifactorial nature of neural regeneration, the environment of the central nervous system, for example, may prevent such re-growth. If axonal growth is possible, membrane repair mechanisms are likely to become activated during the extension of the growth cone into the dense extracellular matrix, an event characterized by high membrane stress. Restoration of membrane homeostasis is therefore an integral part of neural regeneration, important for its initiation, progression, and maintenance. Created with Adobe Illustrator 2023.

It is well appreciated that upon trauma, some neurons survive (due to exposure to sub-lethal forces and, to some degree, intrinsic resilience towards membrane stress and endogenous repair capacity), while others die (Hendricks and Shi, 2014). What happens to those that are injured and survive? They must undergo a period of stabilization and recovery to restore homeostasis. This starts with resealing of the plasma membrane upon trauma to halt the deleterious influx of calcium and prevent the worsening of ion homeostasis to a point of “no return” (surpassing the threshold for cell death). Membrane disruptions have been reported to be resolved within 60 minutes post-injury (Shi et al., 2000), despite variability cross injury type and axon calibre (Hendricks and Shi, 2014). Of note, the proposed underlying mechanism is spontaneous resealing, a passive event that given its slow kinetics may not be physiologically relevant (relative to the rapid calcium influx upon injury). It is far more feasible that membrane resealing occurs through an active process, mediated by an orchestra of repair proteins, including the family of annexin proteins (Dias and Nylandsted, 2021). In vitro, we have observed neuronal repair by bleb formation following axotomy, and indeed trauma often leads to the formation of a swollen dystrophic end bulb (Bradke et al., 2012). This is the canonical repair mechanism upon exposure to pore-forming toxins and has been shown to be efficient both in isolating the wound and confining calcium ions. Others have proposed exocytosis-mediated plasma membrane repair (patch repair) due to the anterograde transport of vesicles to the lesioned axon end (Bradke et al., 2012). Upon membrane resealing (stabilization), a recovery phase can occur (restoring ion homeostasis and membrane polarisation).

Therefore, although neural regeneration has been almost exclusively described by axonal re-grow, active stabilization and recovery is part of regeneration and key for subsequent events to unravel. This is not only explained by the simple fact that neurons that fail to reseal their membrane (being exposed to too much calcium) are not able to regenerate and re-grow, but also because the repair process per se may promote regeneration and re-grow in a direct manner. Membrane repair includes two phases – a resealing phase (occurring within minutes) and a restructuring phase (over minutes to days post-injury). Membrane resealing by blebbing could promote the formation of the growth cone, which tends to form within a bleb. A functional growth cone requires actin and microtubule re-polymerization at the tip of the axon end (after the initial calcium-driven depolymerization) (Bradke et al., 2012). Key repair proteins involved in blebbing stabilize cytoskeletal components. In fact, most repair proteins function by modulating cytoskeletal dynamics. We hypothesize that this activity favors axonal regrowth, since cytoskeletal pro-stabilization strategies have been successful in experimental preparation (Bradke et al., 2012). During the restructuring phase, the profile of membrane proteins and lipids at the wounded area and adjacent regions changes to meet the cellular activity, replacing molecules that previously were introduced to favor rapid membrane repair. In breast cancer cells, we have shown that such membrane restructuring is mediated by endocytosis- and exocytosis-mediated mechanisms (Sønder et al., 2021). Interestingly, Golgi-derived vesicles were found to be transported to and fuse with the distal axon tip (Bradke et al., 2012). It has been hypothesized that these simply supply the membrane with lipids and proteins to support growth (Bradke et al., 2012), although they may also reflect active restoration of membrane homeostasis. Taken together, membrane repair (both resealing and restructuring mechanisms) can promote axonal regeneration.

The transcriptomic profile that arises after membrane resealing, during the stabilization/recovery period, also promotes neuronal regeneration. Upon injury, there are three consecutive transcriptomic events: (1) induction of immediate-early response genes, (2) activation of specific kinase-dependent pathways, and (3) activation of inflammatory and immune responses (Häger et al., 2021). Interestingly, this dynamic response seems to be conserved and is not dependent on injury per se, but rather on an increase of cytosolic calcium (Häger et al., 2021). Therefore, regulation of calcium influx in the aftermath of injury (via membrane resealing) is of great importance for an adequate transcriptomic response to unravel. Within such response mitogen-activated protein kinase signaling cascades become activated (Häger et al., 2021) and regulate apoptosis, axonal regeneration (Watkins et al., 2013) and the pro-inflammatory milieu that is created. Activation of interleukin-6 cascade (Häger et al., 2021) is also important, as this cytokine possesses growth-promoting (Niemi et al., 2022) and pro-inflammatory properties – inflammation is believed to be necessary for axon regrowth (Cooke et al., 2022). Furthermore, single-cell expression profiling has shown that, following injury, both central and peripheral neurons can revert to an “embryonic-like state” that favors axon regeneration (Poplawski et al., 2020). This reversion to a developing phenotype may be part of the common neuronal injury response (Cooke et al., 2022). Taken together, the calcium-dependent, transcriptome response that follows injury is pro-regenerative (Cooke et al., 2022) and is likely to be important for the transition from neuronal recovery to axonal re-growth upon axotomy. Concomitantly, this highlights the importance of adequate stabilization and recovery responses in shaping regeneration, which are heavily dependent on membrane repair.

Having explored potential links between membrane regeneration and neural regeneration in the dynamic and multifaceted response to injury, many questions now open. Do differences in membrane repair responses (resealing and restructuring) influence the axonal cytoarchitecture and the regenerative transcriptome elicited, ultimately influencing neural regeneration? Can this explain differences in the regenerative capacity of neural subpopulations? Can neural regeneration be enhanced by manipulating membrane regeneration?

Therapeutic strategies to modulate neural regeneration have focused solely on the enhancement of re-growth – being somewhat a reflection of the scope of research within the field. Efforts include modulation of regenerative cytoskeletal dynamics and prevention of interactions that destabilize filamentous actin (LaPlaca et al., 2019). Earlier stages within the broader process of neural regeneration (such as stabilization and recovery) have been out shadowed, although they “set-the-scene” for regrowth and overall regeneration capacity. Neurons exposed to sub-lethal stress must be first made fitter to then re-grow. Furthermore, within the scope of axonal re-grow, the primordial focus has been to identify “axon regeneration inhibitors”. Over the last two decades, this has shifted to active pro-regenerative strategies, such as “nerve graft and transplantation” (Chou et al., 2022). However, therapeutic success is hindered by our limited understanding of mechanisms underlying regeneration, including axonal stabilization, recovery, and regrowth. Modulating these early events may favor regeneration.

Some groups are investigating activators/inhibitors of regeneration through a comparative approach. This includes assessing potential explanations for differences in regeneration capacity in the peripheral nervous system vs. the central nervous system (CNS), and within the CNS (“regeneration-competent” vs. “regeneration-incompetent” neurons) (Akram et al., 2022; Cooke et al., 2022). In contrast to most neurons in the CNS that display limited regenerative capacity, serotonin neurons and norepinephrine neurons are robust in unaided regeneration following injury (Cooke et al., 2022). This unpredicted phenomenon questions the status-quo: that the CNS environment is non-permissive to all axon regeneration. However, the CNS remains much more limited in regenerative capacity than the peripheral nervous system, causing defects at the circuitry level to persist following physical injury, increasing the risk for chronic neurodegenerative conditions (Dias and Nylandsted, 2021).

In addition to other potential explanations (reviewed in Cooke et al., 2022), it has been proposed that the ability for neurons to penetrate through the glial scar could contribute to explaining the different regenerative capacities. In addition to glial cells, the scar is composed of invading connective tissue elements, including immediate filaments and multiple extracellular matrix proteins (particularly proteoglycans) that inhibit growth and regeneration. One would therefore expect that penetration into the glial scar exposes neurons to high membrane stress. Is the ability of neurons to regenerate through the scar correlated with a better membrane repair capacity? This concept has parallels with cancer biology: invasive cancer cells enhance their membrane repair capacity to withstand the membrane stress imposed when migrating through dense extracellular tissue (Dias and Nylandsted, 2021). Furthermore, the idea that axons may be exposed to microlesions during the re-growth phase adds complexity to the regeneration process. In addition to the initial stabilization (wound closure) and recovery, membrane repair mechanisms might also be activated during the re-growth phase. Despite its pertinence, it remains to be investigated whether differences in membrane repair capacities could help to explain the increased ability of serotonin and norepinephrine neurons to regenerate, compared to other neuron types.

It is well established that axonal regeneration is multifactorial, dependent on proteomic profile, microtubule stability, and molecular signaling (Akram et al., 2022). However, regeneration tends to be simplistically characterized as axonal re-growth (Cooke et al., 2022). Neuronal stabilization and recovery (i.e., repair) are essential for subsequent axonal regrowth. Upon axotomy, for example, membrane resealing is required to halt the influx of calcium and efflux of intracellular content. A recovery stage follows to restore overall cellular homeostasis and restructure the membrane (also a form of regeneration). The ability for an injured neuron to repair influences the next steps of the regeneration process and may even be important throughout the whole regeneration window, including during axonal growth. However, regeneration has not been investigated from a membrane-centric point of view. Efforts to explore this perspective could unravel novel strategies to enhance the regeneration potential of neurons. A better understanding of how different neural cells repair membrane lesions and how membrane dynamics foster the process of regeneration could shine light on key mechanisms that could have promising targets.

This work was supported by the Novo Nordisk Foundation (NNF18OC0034936) and the Lundbeck Foundation (R380-2021-1262) (to CD and JN).