John L Bixby

Microtubule Stabilization Reduces Scarring and Causes Axon Regeneration After Spinal Cord Injury

Taxol stimulates the capacity of axons to grow after spinal cord injury. Hypertrophic scarring and poor intrinsic axon growth capacity constitute major obstacles for spinal cord repair. These processes are tightly regulated by microtubule dynamics. Here, moderate microtubule stabilization decreased scar formation after spinal cord injury in rodents through various cellular mechanisms, including dampening of transforming growth factor–β signaling. It prevented accumulation of chondroitin sulfate proteoglycans and rendered the lesion site permissive for axon regeneration of growth-competent sensory neurons. Microtubule stabilization also promoted growth of central nervous system axons of the Raphe-spinal tract and led to functional improvement. Thus, microtubule stabilization reduces fibrotic scarring and enhances the capacity of axons to grow.

Paxillin phosphorylation counteracts proteoglycan-mediated inhibition of axon regeneration

In the adult central nervous system, the tips of axons severed by injury are commonly transformed into dystrophic endballs and cease migration upon encountering a rising concentration gradient of inhibitory proteoglycans. However, intracellular signaling networks mediating endball migration failure remain largely unknown. Here we show that manipulation of protein kinase A (PKA) or its downstream adhesion component paxillin can reactivate the locomotive machinery of endballs in vitro and facilitate axon growth after injury in vivo. In dissociated cultures of adult rat dorsal root ganglion neurons, PKA is activated in endballs formed on gradients of the inhibitory proteoglycan aggrecan, and pharmacological inhibition of PKA promotes axon growth on aggrecan gradients most likely through phosphorylation of paxillin at serine 301. Remarkably, pre-formed endballs on aggrecan gradients resume forward migration in response to PKA inhibition. This resumption of endball migration is associated with increased turnover of adhesive point contacts dependent upon paxillin phosphorylation. Furthermore, expression of phosphomimetic paxillin overcomes aggrecan-mediated growth arrest of endballs, and facilitates axon growth after optic nerve crush in vivo. These results point to the importance of adhesion dynamics in restoring endball migration and suggest a potential therapeutic target for axon tract repair.

Reactive oxygen species regulate axonal regeneration through the release of exosomal NADPH oxidase 2 complexes into injured axons

Reactive oxygen species (ROS) contribute to tissue damage and remodelling mediated by the inflammatory response after injury. Here we show that ROS, which promote axonal dieback and degeneration after injury, are also required for axonal regeneration and functional recovery after spinal injury. We find that ROS production in the injured sciatic nerve and dorsal root ganglia requires CX3CR1-dependent recruitment of inflammatory cells. Next, exosomes containing functional NADPH oxidase 2 complexes are released from macrophages and incorporated into injured axons via endocytosis. Once in axonal endosomes, active NOX2 is retrogradely transported to the cell body through an importin-β1–dynein-dependent mechanism. Endosomal NOX2 oxidizes PTEN, which leads to its inactivation, thus stimulating PI3K–phosporylated (p-)Akt signalling and regenerative outgrowth. Challenging the view that ROS are exclusively involved in nerve degeneration, we propose a previously unrecognized role of ROS in mammalian axonal regeneration through a NOX2–PI3K–p-Akt signalling pathway.Hervera et al. show that extracellular vesicles containing NOX2 complexes are released from macrophages and incorporated into injured axons, leading to axonal regeneration through PI3K–p-Akt signalling.

Treatment with analgesics after mouse sciatic nerve injury does not alter expression of wound healing-associated genes

Animal models of sciatic nerve injury are commonly used to study neuropathic pain as well as axon regeneration. Administration of post-surgical analgesics is an important consideration for animal welfare, but the actions of the analgesic must not interfere with the scientific goals of the experiment. In this study, we show that treatment with either buprenorphine or acetaminophen following a bilateral sciatic nerve crush surgery does not alter the expression in dorsal root ganglion (DRG) sensory neurons of a panel of genes associated with wound healing. These findings indicate that the post-operative use of buprenorphine or acetaminophen at doses commonly suggested by Institutional Animal Care and Use Committees does not change the intrinsic gene expression response of DRG neurons to a sciatic nerve crush injury, for many wound healing-associated genes. Therefore, administration of post-operative analgesics may not confound the results of transcriptomic studies employing this injury model.

Exploiting kinase polypharmacology for nerve regeneration

The human central nervous system (CNS) has a markedly poor capacity for regenerating its axons following injury. This appears to be due to two main causes: 1) a developmentally regulated decline in regenerative capacity within mature CNS neurons, and 2) the presence of biological components that constitute barriers to axon regeneration (e.g., growth-inhibitory molecules). Intrinsic alterations have been elucidated by studies that show causative links between developmental changes in gene expression programs and growth-signaling states on one hand, and changes in regenerative capacity on the other (Moore et al., 2009; Park et al., 2010). In addition to these neuron-intrinsic factors, several molecular species that are native to the CNS microenvironment, such as myelin associated proteins, and others that are secreted by injury-activated astrocytes, such as chondroitin sulfate proteoglycans (CSPGs), exert extrinsic inhibitory influences on growing axons (Young, 2014). Given the various factors that negatively influence axon regrowth in the adult CNS, achieving clinically relevant regeneration to promote recovery from traumatic injury has been difficult. Experimental manipulation of individual inhibitory factors, or their mediators, has resulted in some improvement of regeneration/sprouting (Young, 2014). However, manipulations that simultaneously target intrinsic growth capacity while also blocking inhibition from extrinsic factors appear to produce the most pronounced effects in vivo (Lee et al., 2014). Development of effective therapies thus requires agents that simultaneously modulate multiple sources of regeneration failure. This could be achieved with drugs that engage multiple targets (polypharmacology) within various relevant signaling networks. The use of multi-target drugs to treat complex polygenic disorders is not a new concept; however, the lack of appropriate methodologies has hindered systematic exploitation of polypharmacology (Peters, 2013). Interestingly though, it appears that polypharmacology is important for the therapeutic efficacy of many approved drugs (Peters, 2013). For the past few years, we worked to identify small molecule compounds that can engage multiple targets to promote axon regeneration. We focused our investigations on kinases as targets for several reasons. Kinases are involved in regulating most if not all cellular processes. They are well-established and readily druggable targets, with numerous therapeutic applications ranging from neurology to cancer. Moreover, the homology of kinase catalytic domains gives rise to kinase inhibitor promiscuity (Metz et al., 2011), making it possible to discover drugs with multiple “intended” kinase targets. Numerous kinases have been shown to play a role in regulating axon growth. However, predicting the most effective pharmacological targets (and target combinations) requires detailed knowledge of time-dependent properties of all nodes within relevant signaling networks, which at present is not feasible. Thus, we utilized phenotypic screening, which does not require a priori target hypotheses. Phenotypic screening tests compounds on whole cells rather than individual drug targets, and enables the discovery of compounds with desired biological effects, including those with favourable polypharmacology. We developed an assay that utilizes primary neurons from rat brains and used it to identify kinase inhibitors that strongly promote neurite outgrowth (Al-Ali et al., 2013). This technique by itself, however, does not provide direct information on the identity of the cellular targets through which the compounds exert their biological effects. Without this information, it is not possible to rationally select compounds with the best polypharmacology. To overcome this problem, we made use of another popular drug discovery technique, target-based screening. Target-based screening assays a single functional molecule (in our case a kinase) against a large number of compounds. We assayed several hundred compounds (previously screened in the phenotypic assay) against more than two hundred kinases to obtain the compounds’ inhibition profiles towards those kinases. Using machine learning and information theory to relate the compounds’ kinase inhibition profiles to their influence on neurite outgrowth, we were able to identify kinases that can serve as targets for promoting neurite outgrowth. We also identified kinases whose inhibition represses neurite outgrowth, and thus their targeting should be avoided (anti-targets) (Al-Ali et al., 2015). Typically, a prioritized drug target is a single functional protein. The goal of target-based discovery campaigns is often to identify compounds that potently interact with the assayed target to elicit a desired therapeutic response. However, due to structural similarities arising from homology or parallel evolution, various proteins can show a high degree of similarity in the identity of compounds that they bind (Metz et al., 2011). This means that inhibiting a given target with a compound will frequently also inhibit other proteins that have a topologically similar binding pocket. Thus, as the similarity between binding pockets increases, so does the likelihood for co-inhibition. Proteins with highly similar binding pockets can be said to be pharmacologically linked (i.e., it is difficult to engage one without also engaging the other). This is especially true for catalytically active kinases, given their shared lineage and evolutionarily conserved requirement for ATP binding. To account for this, we extended the concept of drug target from that of a single kinase to a group of pharmacologically linked kinases. Thus, a kinase was considered a robust target only if it was pharmacologically linked to other kinases that also behave as targets, or at least do not behave as anti-targets. Since most small molecule inhibitors will engage all members in a group of pharmacologically linked kinases, this criterion ensures that the overall effect of inhibiting a robust target group will be positive (pharmacologically linked targets and anti-targets will tend to counteract one another). We also identified several pharmacologically linked anti-target groups whose inhibition correlated with strong negative effects on neurite outgrowth. We prioritized the ten most robust target and anti-target groups (based on the outcome of our phenotypic screen), and selected a single member from each group for prioritizing lead compounds. For example, the activity of a compound against rho kinase 2 (ROCK2) was used to represent the activity of that compound against the corresponding group of pharmacologically linked kinases (in this case, ROCK1 and ROCK2). Activated CDC42 kinase 1 (TNK2), rho-associated kinase-II (ROCK2), PI3-kinase δ (PIK3CD), protein kinase C γ (PRKCG), ribosomal protein S6 kinase α-4 (RPS6KA4), cGMP-dependent protein kinase G 1 (PRKG1), and cAMP-dependent protein kinase X (PRKX) were selected as representatives of robust target groups, while p38α MAP kinase (MAPK14), MAP kinase-activated protein kinase 3 (MAPKAPK3), and cyclin-dependent-like kinase 5 (CDK5) were selected as representatives of robust anti-target groups (Figure 1). Using these representative targets and anti-targets, we identified a compound with exemplary polypharmacology, which inhibits 5 out of the 7 robust target groups and does not affect the robust anti-target groups. Amongst its targets, RO0480500-002 inhibits both PKC and ROCK, kinases known to mediate repression of axon growth by myelin and CPSGs in the CNS (Young, 2014). RO0480500-002 also inhibits the growth regulatory S6 kinases, which have been shown to limit intrinsic neuronal capacity for axon growth and regeneration (Hubert et al., 2014). Moreover, RO0480500-002 inhibits PRKG1 and PRKX, two kinases involved in the regulation of cell migration and cytoskeletal rearrangement. Our phenotypic assays showed that this compound promotes neurite outgrowth both from hippocampal neurons and from postnatal cortical neurons (Al-Ali et al., 2015). Importantly, RO0480500-002 has beneficial effects on descending motor axons in vivo following spinal cord injury. Figure 1 First-neighbor protein interaction network for top ranked target and anti-target kinases. This new appreciation for favourable polypharmacology among kinase inhibitors may shed light on earlier results. Previous studies had shown that two kinase inhibitors, Go6976 and Y-27632, promote axon regeneration in vivo. Y-27632 was described as a ROCK inhibitor (Fournier et al., 2003), while Go6976 was described as a PKC inhibitor (Wang et al., 2013). Interestingly, kinase profiling of these two compounds reveals that each of them also inhibits several robust targets identified by our method (Figure 2), raising the possibility that their polypharmacology may have contributed to their in vivo efficacies. Interestingly, the compounds share no chemical similarity despite having somewhat similar target polypharmacology. This underscores the idea that desirable polypharmacology is not necessarily restricted to particular chemical scaffolds. With the expanded view of a polypharmacology profile (as opposed to just a single target), lead compounds with desirable polypharmacology and improved pharmacokinetic/pharmacodynamic properties can be discovered more easily, even if they share no chemical similarity to the original hit compounds. The polypharmacology profile elucidated in this study provides such a platform for discovering and developing effective multi-target drugs for neurodegenerative applications. Figure 2 Kinase inhibitors previously reported to promote axon growth in vivo exhibit poly-pharmacology.

Publisher Correction: Reactive oxygen species regulate axonal regeneration through the release of exosomal NADPH oxidase 2 complexes into injured axons

In the version of this Article originally published, the affiliations for Roland A. Fleck and José Antonio Del Río were incorrect due to a technical error that resulted in affiliations 8 and 9 being switched. The correct affiliations are: Roland A. Fleck: 8Centre for Ultrastructural Imaging, Kings College London, London, UK. José Antonio Del Río: 2Cellular and Molecular Neurobiotechnology, Institute for Bioengineering of Catalonia, Barcelona, Spain; 9Department of Cell Biology, Physiology and Immunology, Facultat de Biologia, Universitat de Barcelona, Barcelona, Spain; 10Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), Barcelona, Spain. This has now been amended in all online versions of the Article.