Integrins protect sensory neurons in models of paclitaxel-induced peripheral sensory neuropathy

Chemotherapy-induced peripheral neuropathy (CIPN) is a major side effect from cancer treatment with no known method for prevention or cure in clinics. CIPN often affects unmyelinated nociceptive sensory terminals. Despite the high prevalence, molecular and cellular mechanisms that lead to CIPN are still poorly understood. Here, we used a genetically tractable Drosophila model and primary sensory neurons isolated from adult mouse to examine the mechanisms underlying CIPN and identify protective pathways. We found that chronic treatment of Drosophila larvae with paclitaxel caused degeneration and altered the branching pattern of nociceptive neurons, and reduced thermal nociceptive responses. We further found that nociceptive neuron-specific overexpression of integrins, which are known to support neuronal maintenance in several systems, conferred protection from paclitaxel-induced cellular and behavioral phenotypes. Live imaging and superresolution approaches provide evidence that paclitaxel treatment causes cellular changes that are consistent with alterations in endosome-mediated trafficking of integrins. Paclitaxel-induced changes in recycling endosomes precede morphological degeneration of nociceptive neuron arbors, which could be prevented by integrin overexpression. We used primary dorsal root ganglia (DRG) neuron cultures to test conservation of integrin-mediated protection. We show that transduction of a human integrin β-subunit 1 also prevented degeneration following paclitaxel treatment. Furthermore, endogenous levels of surface integrins were decreased in paclitaxel-treated mouse DRG neurons, suggesting that paclitaxel disrupts recycling in vertebrate sensory neurons. Altogether, our study supports conserved mechanisms of paclitaxel-induced perturbation of integrin trafficking and a therapeutic potential of restoring neuronal interactions with the extracellular environment to antagonize paclitaxel-induced toxicity in sensory neurons.

Chemotherapy-induced peripheral neuropathy (CIPN) is a prevalent adverse effect of treatment in cancer patients and survivors (1). CIPN significantly impacts quality of life as damage to sensory nerves may be permanent, and is often a dose-limiting factor during cancer treatment (2–4). Patients with CIPN report pain-related symptoms, including allodynia, hyper- or hypoalgesia, or pain that can be more severe than the pain associated with the original cancer (4). Despite increasing data on agents that protect sensory nerves, our limited understanding of the mechanisms of CIPN impedes effective treatment (5). Studies from model systems may be helpful in identifying molecules that protect sensory neuron morphology and function from the effects of chemotherapeutics.In the present study, we explored the mechanisms of CIPN induced by paclitaxel using two established models: Drosophila larval nociceptive neurons (6, 7) and primary dorsal root ganglia (DRG) neurons isolated from adult mouse (8). Similar to other peripheral neuropathies, CIPN models using paclitaxel, bortezomib, oxaliplatin, and vincristine report changes in unmyelinated intraepidermal nerve fibers (IENFs) that detect painful or noxious stimuli (9–14). These small fibers are embedded in the epidermis, and continuously turn over coincident with the turnover of skin (9, 15). Drosophila class IV nociceptive neurons are a favored model for genetic studies of nociceptive neuron development and signaling mechanisms (16). Prior studies showed that class IV neuron morphology is sensitive to paclitaxel and demonstrated morphological changes of nociceptive neurons at the onset and the end stage of paclitaxel-induced pathology (6, 7). Specifically, chronic treatment of high doses (30 μM) induce fragmentation and simplification of branching of sensory terminals (6). Additionally, acute treatments of moderate doses (10 to 20 μM) induced hyperbranching of sensory arbors without changing the branch patterns or degeneration (7). Nociceptive neurons in Drosophila larvae detect multiple qualities of noxious stimuli (17, 18), and project naked nerve terminals that are partially embedded in the epidermis (19, 20). Larvae have a stereotyped behavioral response toward noxious stimuli that can serve as a readout of nociceptive neuron function (17, 21). Nociceptive neurons in Drosophila larvae may therefore serve as a good in vivo model to study morphological and functional changes to sensory neurons induced by chemotherapeutics.Paclitaxel binds to tubulin and prevents microtubule disassembly. It is a commonly used chemotherapeutic drug for treatment of solid cancers, such as breast, ovarian, and lung cancers, by virtue of its ability to inhibit cell division. Paclitaxel causes chronic sensory neuropathy in patients and animal models (22–24). Several CIPN animal and in vitro models have also revealed acute effects of paclitaxel (7, 8, 24–26). While the mechanisms of acute and chronic neurodegeneration are likely to be distinct (27), how long-term treatment of paclitaxel can affect sensory neuron morphology and function, and how neuronal arbors can be protected against long-term toxicity is not understood.Several studies have shown that nociceptive sensory terminals share a close relationship with specific extracellular structures, most notably epidermal cells and the extracellular matrix (ECM). Thus, in addition to direct effects on neurons, paclitaxel could conceivably destabilize terminals by disrupting relationships with the extracellular environment. Indeed, a study in zebrafish indicates that epidermal cells are directly affected by paclitaxel and that epidermal changes precede neuronal degradation, indicating that degradation of neuronal substrates contributes to degeneration of adjacent arbors (25). For the most part, however, extracellular contributions to neuropathy induced by chemotherapeutics are still poorly characterized. It is therefore important to determine how sensory terminals are maintained in the context of a dynamic extracellular environment that itself may be sensitive to chemotherapeutics. Integrins are a key mediator of the interaction between cells and the ECM, and impact dendrite stabilization and maintenance in both vertebrate and invertebrate systems (20, 28, 29). Prior studies in other systems indicate that integrin levels at the surface are maintained by continuous recycling via tight regulation of the endosomal pathway rather than degradation and de novo synthesis (30). Decreased recycling or increased degradation could lead to depletion of the surface receptors (31, 32) responsible for arbor maintenance and, in turn, degeneration of nociceptive terminals. We therefore explored whether integrin–ECM interactions may impact sensory neuron maintenance upon paclitaxel-induced toxicity and how the endosomal–lysosomal pathway may be linked to the maintenance of sensory neurons.Here, we have used Drosophila and isolated mouse DRG neurons to investigate the pathological effect of paclitaxel in sensory neurons. Morphological changes in Drosophila neurons occurred at paclitaxel doses that also caused changes in thermal nociceptive behaviors. Cell-specific overexpression of integrins protected nociceptive neurons from morphological alterations and prevented the thermal nociceptive behavior deficits caused by paclitaxel in Drosophila. Transduction of integrins also protected adult mouse DRG sensory neurons from paclitaxel-induced toxicity in vitro, indicating that integrin-mediated protection is conserved in a vertebrate model of CIPN. We provide evidence that paclitaxel alters intracellular trafficking in both Drosophila and mouse models of CIPN. Furthermore, our biochemical analysis indicates a reduction of integrin surface availability, suggesting paclitaxel-induced recycling defects in mouse DRG neurons in vitro. Our study suggests that altered interactions between sensory neurons and their extracellular environment are an important contributor to paclitaxel-induced neuronal pathology, and that preventing these changes may offer a therapeutic approach.