Sodium-Calcium Exchanger 2: A Pivotal Role in Oxaliplatin Induced Peripheral Neurotoxicity and Axonal Damage?

Oxaliplatin (OHP)-induced peripheral neurotoxicity (OIPN) is a frequent adverse event of colorectal cancer treatment. OIPN encompasses a chronic and an acute syndrome. The latter consists of transient axonal hyperexcitability, due to unbalance in Na+ voltage-operated channels (Na+VOC). This leads to sustained depolarisation which can activate the reverse mode of the Na+/Ca2+ exchanger 2 (NCX2), resulting in toxic Ca2+ accumulation and axonal damage (ADa). We explored the role of NCX2 in in vitro and in vivo settings. Embryonic rat Dorsal Root Ganglia (DRG) organotypic cultures treated with SEA0400 (SEA), a NCX inhibitor, were used to assess neuroprotection in a proof-of-concept and pilot study to exploit NCX modulation to prevent ADa. In vivo, OHP treated mice (7 mg/Kg, i.v., once a week for 8 weeks) were compared with a vehicle-treated group (n = 12 each). Neurophysiological and behavioural testing were performed to characterise acute and chronic OIPN, and morphological analyses were performed to detect ADa. Immunohistochemistry, immunofluorescence, and western blotting (WB) analyses were also performed to demonstrate changes in NCX2 immunoreactivity and protein expression. In vitro, NCX inhibition was matched by ADa mitigation. In the in vivo part, after verifyingboth acute and chronic OIPN had ensued, we confirmed via immunohistochemistry, immunofluorescence, and WB that a significant NCX2 alteration had ensued in the OHP group. Our data suggest NCX2 involvement in ADa development, paving the way to a new line of research to prevent OIPN.


Introduction
Oxaliplatin (OHP)-induced peripheral neurotoxicity (OIPN) is a frequent toxicity, experienced by a growing population of colorectal cancer survivors that can be long-lasting, or even permanent [1,2], altering patients' quality of life [3,4]. At the state of the art, there is no efficacious preventive or curative treatment for OIPN [5]. One of the reasons for this unmet clinical need is the incomplete knowledge on axonal damage (ADa) mechanisms [6,7]. Therefore, robust experimental evidence is still required to devise novel treatments. OIPN is characterised by two different conditions. A chronic sensory neuropathy characterised by ADa, known as chemotherapy-induced peripheral neurotoxicity (CIPN), which is one of the commonest late toxicities of several anticancer drugs (platinum drugs, taxanes, vinca alkaloids, proteasome inhibitors, epothilones, and thalidomide [8][9][10][11]). OIPN, however, is also characterised by a specific acute neurotoxicity syndrome, as soon as after the first leading to NCX reverse mode activation and subsequent Ca 2+ -related toxicity in axons [42]; similar findings were then replicated in peripheral nervous system myelinated axons exposed to anoxia [43,44]. Ca 2+ toxicity due to the reverse mode activation was related to several neurological diseases, and the possible beneficial role of its modulation has been described in different conditions such as brain ischemia [31], Alzheimer disease [45], and amyotrophic lateral sclerosis [46]. However, further observations are still required. Therefore, we performed in vitro and in vivo experiments to shed light into the potential role of NCX2 in ADa development. In the in vivo study, we provide evidence of potential NCX2 involvement in OHP-related ADa, and in the in vitro part we propose a proof-ofconcept and pilot study of NCX modulation effects.

In Vivo Observations
Study design for the characterization of an in vivo comprehensive OIPN mouse model was as follows. Nerve conduction studies (NCS) and behavioural test parameters were used to verify homogeneity between the experimental groups before 1st chemotherapy administration (no statistically significant difference for all parameters). Acute OPIN was verified after 1st OHP injection monitoring NET changes in the subsequent 72 h. Chronic OIPN and, therefore, ADa development was verified via NCS, neuropathology (caudal nerve morphometry/morphology and intraepidermal nerve fiber shows neurite outgrowth as a percentage respect to the CTRL group (statistical significance of one-way ANOVA (followed by Tukey's Multiple Comparison Test) is also provided: * p < 0.05 vs. CTRL; *** p < 0.001 vs. CTRL; • p < 0.05 vs. OHP, •• p < 0.01 vs. OHP).

In Vivo Observations
Study design for the characterization of an in vivo comprehensive OIPN mouse model was as follows. Nerve conduction studies (NCS) and behavioural test parameters were used to verify homogeneity between the experimental groups before 1st chemotherapy administration (no statistically significant difference for all parameters). Acute OPIN was verified after 1st OHP injection monitoring NET changes in the subsequent 72 h. Chronic OIPN and, therefore, ADa development was verified via NCS, neuropathology (caudal nerve morphometry/morphology and intraepidermal nerve fiber density [IENFD]), and behavioural tests after 8 weeks of treatment. Immunohistochemistry, immunofluorescence, and western blotting were used to assess if OIPN was associated with alterations in immunoreactivity and protein expression in DRG of animals treated with OHP for 8 weeks.

NET after the 1st Administration
NET verified acute OIPN ensued after the 1st chemotherapy cycle. Over 72 h of monitoring, the most notable change was related to an upward shift in the recovery cycle curve (p-value < 0.01 at 2 ms and at 2.5 ms refractoriness assessment, Mann-Whitney Utest). These findings are consistent with our previous study and suggest the onset of OHP-related axonal hyperexcitability that alters Na + VOC functioning, leading to an aberrant depolarization [25]. Figure 2 gives an overview of the monitoring over 72 h showing statistically significant parameters. Individual traces of each recording are provided in supplementary Figure S1. treated with OHP for 8 weeks.
2.2.1. NET after the 1st Administration NET verified acute OIPN ensued after the 1st chemotherapy cycle. Over 72 h of monitoring, the most notable change was related to an upward shift in the recovery cycle curve (p-value < 0.01 at 2 ms and at 2.5 ms refractoriness assessment, Mann-Whitney U-test). These findings are consistent with our previous study and suggest the onset of OHP-related axonal hyperexcitability that alters Na + VOC functioning, leading to an aberrant depolarization [25]. Figure 2 gives an overview of the monitoring over 72 h showing statistically significant parameters. Individual traces of each recording are provided in supplementary Figure S1.

NCS and Behavioral Tests at the End of Treatment
A pattern compatible with polyneuropathy was observed via NCS: significant decrease in both caudal and digital nerve sensory action potential (SAP) amplitude and digital nerve conduction velocity was observed in OHP-treated animals. These findings demonstrated that relevant OHP-related ADa ensued, able to cause a functional impairment of large myelinated fibers, which are the ones tested with NCS [49]. Similarly, mechanical allodynia (Dynamic test) had ensued in OHP-treated animals (p < 0.001), mirroring a functional impairment ensued in small fibers (that convey painful information) [26]. Data are summarised in Figure 3.

NCS and Behavioral Tests at the End of Treatment
A pattern compatible with polyneuropathy was observed via NCS: significant decrease in both caudal and digital nerve sensory action potential (SAP) amplitude and digital nerve conduction velocity was observed in OHP-treated animals. These findings demonstrated that relevant OHP-related ADa ensued, able to cause a functional impairment of large myelinated fibers, which are the ones tested with NCS [49]. Similarly, mechanical allodynia (Dynamic test) had ensued in OHP-treated animals (p < 0.001), mirroring a functional impairment ensued in small fibers (that convey painful information) [26]. Data are summarised in Figure 3.

Caudal Nerve Morphology and Morphometry
Morphological examination of caudal nerves, harvested after sacrifice at the end of 8 weeks of treatment, demonstrated mild ADa in large myelinated fibers, matching observations obtained via NCS. Morphometrical analyses showed a statistically significant decrease in the fiber density of large myelinated fibers in the OHP treated group.

Intraepidermal Nerve Fiber Density (IENFD)
Analysis of IENFD allows to formally count small nerve fibers (and eventually assess their loss), relying on cutaneous terminals. At the end of treatment, IENFD demonstrated a statistically significant decrease in treated animals compared to controls, mirroring the functional impairment that was observed at Dynamic test for mechanical allodynia. Mann-Whitney U-test significance and representative images are shown in Figure 5.

Intraepidermal Nerve Fiber Density (IENFD)
Analysis of IENFD allows to formally count small nerve fibers (and eventually assess their loss), relying on cutaneous terminals. At the end of treatment, IENFD demonstrated a statistically significant decrease in treated animals compared to controls, mirroring the functional impairment that was observed at Dynamic test for mechanical allodynia. Mann-Whitney U-test significance and representative images are shown in Figure 5. DRG harvested at sacrifice after 8 weeks of treatment were used to assess whether difference can be found between control and OHP animals. We conducted analyses wi three techniques to verify the robustness of our observations. IHC and IF showed a simil pattern of mainly cytosolic NCX2 immunoreactivity (shown in red in IF images and dark brown in DAB-stained IHC images); a statistically significant reduction in the OH group with respect to control was observed with both techniques, exploiting a quantit tive measurement of NCX2 immunoreactivity ( Figure 1C,D, respectively), and a simil reduction was demonstrated via WB analysis on the DRG pool ( Figure 6E,F). This is a indirect confirmation that NCX2 reverse mode was activated. If reverse mode was activate due to an aberrant depolarization, as already demonstrated by Boscia et al. [50], a repe tive spreading of depolarisation current resulted in a downregulation of NCX2 in neuro (this is an endogenous autoprotective mechanism to prevent Ca 2+ overload). DRG harvested at sacrifice after 8 weeks of treatment were used to assess whether a difference can be found between control and OHP animals. We conducted analyses with three techniques to verify the robustness of our observations. IHC and IF showed a similar pattern of mainly cytosolic NCX2 immunoreactivity (shown in red in IF images and in dark brown in DAB-stained IHC images); a statistically significant reduction in the OHP group with respect to control was observed with both techniques, exploiting a quantitative measurement of NCX2 immunoreactivity ( Figure 1C,D, respectively), and a similar reduction was demonstrated via WB analysis on the DRG pool ( Figure 6E,F). This is an indirect confirmation that NCX2 reverse mode was activated. If reverse mode was activated due to an aberrant depolarization, as already demonstrated by Boscia et al. [50], a repetitive spreading of depolarisation current resulted in a downregulation of NCX2 in neurons (this is an endogenous autoprotective mechanism to prevent Ca 2+ overload).

Discussion
So far, no mouse model has been described reproducing both acute and chronic OIPN; therefore, we first validated the robustness of the proposed model. Recently, Makker et al., in fact, characterised a mouse model showing only acute OIPN features, exploiting NET [18]. As performed in a previous study from our group in a rat model [25], we relied on a multimodal approach to demonstrated both acute and chronic OIPN had ensued [6,26]. Recovery cycle parameters at NET after the first cycle showed the typical acute OIPN pattern: OHP group showed an upward shift of the recovery cycle curve, compatible with altered Na + VOC functioning as already demonstrated [15,18,25]. We relied on multiple outcome measures to verify the full expression of chronic OIPN (i.e., ADa) at the end of treatment [26]. NCS were used to demonstrate the presence of a sensory axonal neuropathy; consistent with clinical observations, we observed mainly a significant drop in SAP amplitude compatible with sensory ADa [20,51,52]. Morphology of caudal nerves matched neurophysiological results, and morphometry demonstrated axonal loss of myelinated fibers. Larger diameter fibers were the ones with the most evident damage, matching clinical experience; patients mainly show sensory large fibers impairment, up to a disabling sensory ataxia [1,11,53]. Even if large fibers are the main target of OHP ADa, impairment in small fibers was clearly demonstrated in animal models [25,54] and was also observed in a small clinical cohort [55]; our model also reproduced this feature as demonstrated via IENFD reduction. Finally, allodynia was tested to demonstrate a functional impairment in small fibers; mechanical allodynia was tested, instead of temperature-triggered allodynia, to avoid possible overlap between acute OIPN and ADa, given the former is cold triggered [1,11]. Thus, the full OIPN spectrum was reproduced in our animals allowing us to perform NCX2-related observations in a setting mirroring the clinical condition.
We verified NCX2 immunolocalisation in DRG via immunohistochemistry/ immunofluorescence. NCX2, quite notably, is known for being well represented in DRG [33], which are the first target in case of ADa due to OHP [11]. Even if indirectly, our data were compatible with NCX2 reverse mode activation. First of all, we demonstrated that the prerequisite of reverse mode activation ensued (i.e., Na + dysfunction) via NET; this is the trigger to switch the mode of NCX functioning. If reverse mode was activated via an aberrant depolarization, neurons downregulated NCX to avoid a Ca 2+ overload [50], which is the final effect of a prolonged reverse mode activation. Therefore, the 2 months of treatment and repetitive oxaliplatin administration (each injection is a trigger for reverse mode activation) were able to determine a reverse mode activation relevant enough that the downregulation mechanism was established; this is an endogenous and autoprotective mechanism that it is not ultimately strong enough to prevent axonal damage though, as evidenced by ADa development in OHP animals as seen at NCS, Dynamic test for allodynia, and histopathological observations of large and small nerve fibers as discussed above.
Preliminary, mainly in vitro findings suggested that NCX2 modulation might mitigate toxic neuropathies [56,57], pointing out that a pharmacological modulation of this axis can be more efficacious than the endogenous mechanisms. To further investigate this, we conducted a proof-of-concept and pilot in vitro study to test beneficial effects of NCX modulation. Among possible inhibitors, we selected SEA0400 since it is the most potent and selective NCX inhibitor [58]. Other options, such as the less strong NCX inhibitor KB-R7943, showed a rather low specificity for NCX family even at low dosages, with larger impact on off-targets with respect to SEA0400 [58][59][60]. Moreover, SEA0400 was already used in studies that investigated nervous system injury due to NCX2 reverse mode activation/Ca 2+ toxicity, caused by alterations of Na + currents: Matsuda et al. [58] showed that SEA0400 attenuated, dose-dependently, damage in models of brain ischemia. Koyama et al. [61] also obtained promising neuroprotection with SEA0400 in a model of brain ischemia. Notably, in a different model of peripheral neuropathy, Yilmaz et al. showed that some neuronal populations of the peripheral nervous system can be resistant to the effect of KB-R7943, but responsive to SEA0400 [57]. Therefore, SE0400 profile was the most appropriate drug for a proof-of-concept and pilot experiment to test our hypothesis. However, the specific model of disease and system we are dealing with was carefully considered to select the range of dosages to be tested. Literature data reported above, concerning the central nervous system, mostly used nM concentrations [62], but we also considered other published works to better mirror the peripheral nervous system setting; to obtain effects on axons (squid giant axon), a higher-despite still selective-dosage was require in the range of µM [63]. Moreover, in the work by Yilmaz et al., in a different model of peripheral neuropathy [57], SEA0400 was used in a microM concentrations, as well as in neuroprotection experiments relying on SH-SY5Y cells [64]. We also carefully evaluated the most appropriate OHP dose to be used, relying on the extensive review of Calls et al. [65], which presented an overview of the in vitro and in vivo models of OIPN; starting from these, we selected a dosage that was similar to the mean plasma levels in patients treated with OHP [66,67] and able to induce neurite alterations.
To ensure SEA0400 was active against down-stream events, secondary to Na + VOC alteration, a pre-treatment exposure to SEA0400 was performed (3 h); it was necessary, in fact, to administer SEA0400 before OHP exposure to ensure the block of NCX activity before the reverse mode was activated, as a consequence of Na + VOC unbalance, similarly to what we previously performed in rats with TPM [25]. Exploiting a consolidated approach to assess neurotoxicity in vitro [68], we observed promising data suggesting the neuroprotective role of NCX inhibition against OHP detrimental effects in neurons. This confirms, indirectly, that NCX2 reverse mode activation plays a role in ADa. However, some limitations should be acknowledged. Ours is a proof-of-concept and pilot study that would require further investigation. Despite the fact that our data are promising, of course, it could be argued that a combined administration of Na + current and Ca 2+ current distinct modulators might be effective too in preventing calcium overload as described in models of cardiac ischemiareperfusion injury and cardiac glycoside toxicity [69]. However, targeting NCX with novel drugs/approaches, instead of these ion channels, is nevertheless a promising strategy since, differently from these ion channels, it is a relatively simpler family; Na + and Ca 2+ ion channels are characterised by several subtypes with relevant functional differences, which makes off-target effects quite easier [70,71]. Of course another point to be considered is that SEA0400 too, despite its specificity, could exert eventual off-target actions, in particular at higher dosages [60]; but, as stated above, the dosages we selected are expected to be quite selective in the peripheral nervous system. Overall, it can be suggested that our data pointed out that NCX2 is potentially a pivotal element in ADa and, therefore, a druggable target to be further investigated. This paves the way to a new line of research to further explore the intriguing possibility that NCX2 modulation might prevent ADa, relying on highly selective approaches to avoid off-target effects (e.g., NCX2 modulation via siRNA in in vivo OIPN models).

Drugs
OHP was purchased from Sigma Chemical Co. (Milan, Italy). It was dissolved in physiological solution to make a stock solution of 1 mg/mL, which was diluted with medium to obtain the working concentration of 7.5 µM. This concentration has been selected in a similar range to the mean plasma concentrations in patients treated with OHP [64,70] and able to induce a relevant reduction in neurite elongation after exposure. The strong and selective NCX inhibitor SEA0400 [58] was purchased from Sigma Chemical Co. (Milan, Italy). It was dissolved in dimethyl sulfoxide (DMSO) to make a stock solution of 1 mM, which was diluted with medium to obtain the 20 µM concentration; this concentration was selected after proof-of-concept experiments testing a scaling dose of SEA0400 (1, 10, 20, and 30 µM) and selecting the most promising dosage in terms of neuroprotection. The inhibition effect of SEA0400 was evaluated by a 3 h pre-treatment of SEA0400 then followed by the OHP exposure. Pre-treatment was mandatory on the basis of our previously published data, demonstrating that neuroprotection against OIPN is obtained modulating ion channels/currents before OHP exposure [25].

Dorsal Root Ganglia (DRG) Explants
DRG from 15-day-old embryonic Sprague Dawley rats were aseptically removed and cultured onto a single layer of rat-tail collagen surfaces in 35 mm dishes (4 ganglia/dish). DRG were incubated in media (MEM plus 10% calf bovine serum, 50 µg/mL ascorbic acid, 1.4 mM L-glutamine, 0.6% glucose) with 5 ng/mL NGF for 2 h in a 5% CO 2 humidified incubator at 37 • C. DRG were pre-treated with SEA0400 (20 µM) and then also exposed to OHP 7.5 µM for 24 h and 48 h. Phase contrast micrographs of all DRG were made after 24 h and 48 h of OHP exposure. For each DRG, the longest neurite was measured. These magnified measurements were compared with a calibration grating photographed under identical conditions. Each experiment was performed three times to validate the results. Animals were maintained under standard animal housing conditions, thus with a 12 h light-dark cycle and a room temperature and relative humidity at 20 ± 2 • C and 55 ± 10%, respectively. Drug-and vehicle-treated mice were housed separately with free access to water and food.

Study Design
Animals were divided into 2 groups (n = 12 each) and treated as follows: control (CTRL) group was composed of vehicle-treated animals (5% glucose solution); OHP group was treated with OHP 7 mg/kg i.v., once a week per 8 weeks (1 qw 8 ws). Sample size for each experiment was calculated on the basis of nerve conduction velocity (NCV) reference values of our laboratory [72], assuming that the relevant difference between CTRL and OHP groups is 5 m/s (standard deviation = 7); thus, if a 2-sided 5% alpha and a 80% power is set, the sample size is 7 animals/group (www.dssresearch.com/KnowledgeCenter/ toolkitcalculators/samplesizecalculators.aspx (accessed on 12 December 2019)). In each experiment the sample size was increased above the defined number (12 animals/group) in order to have enough animals to be tested at each time point, taking into account the different duration of treatment between experiments and avoiding underpowered statistical analysis in case of animal loss due to treatment. To ensure groups homogeneity, animals were randomised based on nerve conduction studies (NCS) values at baseline. Dynamic test and NCS were performed at baseline to ensure homogeneity among groups and then repeated and the end of treatment to ensure neuropathy induction on all animals/group. Nerve excitability testing was performed at 24, 48, 72 h after the 1st OHP administration to monitor changes in axonal excitability on all animals/group. At the end of the observational period (after the execution of NCS and behavioural tests at the end of treatment) animals were sacrificed and specimen for histological analysis and western blotting were harvested (3 animals/group).

Drug Administration
OHP (Oxaliplatin 5 mg/mL solution, Accord Healthcare Limited, Durham, NC, USA), 7 mg/kg or the vehicle solution (5% glucose solution) was administered i.v. in the tail vein once a week for a period of 8 weeks.

Dynamic Test
The mechanical nociceptive threshold was assessed using a Dynamic Aesthesiometer Test (model 37450, Ugo Basile Biological Instruments, Comerio, Italy), which generated a linearly increasing mechanical force. At each time point, after the acclimatization period, a pointed metallic filament (0.5-mm diameter) was applied to the plantar surface of the hind paw, which exerted a progressively increasing punctuate pressure, reaching up to 15 g within 15 s. The pressure evoking a clear voluntary hind-paw withdrawal response was recorded automatically and taken as representing the mechanical nociceptive threshold index. Results represented the maximal pressure (expressed in grams) tolerated by animals. There was an upper limit cut-off of 15 s, after which the mechanical stimulus was automatically terminated.

Caudal Nerve Morphology and Morphometry
Caudal nerves were isolated and dissected out to avoid stretching. Nerves were immediately fixed by immersion in 3% glutaraldehyde in 0.12 M phosphate buffer solutions pH 7.4, post-fixed in OsO4 and embedded in epoxy resin. Then, 1.5-µm semithin sections were obtained and stained with toluidine blue. Morphometric analysis was performed on a 60x stitched image of a single nerve section per animal (3 animals/group) with Nexcope NE 920 light microscope (TIEsseLab S.r.l., Milan, Italy). The images were acquired in stitching mode using Capture V2.0 Software (Revolutionary Computational Imaging Software, TIEsseLab S.r.l., Milan, Italy) and processed by an automatic image analyser (Image-Pro Plus Software, Immagini e Computer SNC, Milan, Italy). The fiber diameter and the frequency distribution of myelinated fibers were determined, and data were analysed with GraphPad Prism Software (GraphPad Software, San Diego, CA, USA).

IENFD
Hind paws were collected after sacrifice, and 3-mm round shapes were taken and immediately fixed in PLP 2% (paraphormaldehyde-lysine-periodate sodium). Three sections (20 µm thick) were randomly obtained from each footpad and immunostained with rabbit polyclonal antiprotein gene product 9.5 (UCHL1/PGP 9.5 (protein gene product 9.5) Rabbit Polyclonal antibody, Proteintech, Illinois, Rosemont, IL, USA) using a free-floating protocol. The total number of PGP-positive fibers IENFs were counted under a light microscope (Nexcope NE 920 light microscopel TIEsseLab S.r.L., Milan, Italy) at 40X magnification with the assistance of a microscope-mounted video camera. Individual fibers were counted (blind examiner) as they crossed the dermal-epidermal junction, and secondary branching within the epidermis will be excluded. A computerised system measured epidermidis length to calculate the linear density of IENF/mm.

Immunohistochemistry for NCX2
DRG of three animals per group were dissected, fixed in 10% formalin for 3 h at RT and paraffin embedded. Three-µm-thick slices were cut with a Leica RM2265 microtome (Microsystems GmbH, Wetzlar, Germany). Immunohistochemistry was performed using a rabbit polyclonal anti-Na + /Ca 2+ Exchanger 2 antibody (NCX-2, TA328916, OriGene, Rockville, MD, USA). Paraffin sections were deparaffinised with xylene, rehydrated, and heated in a steamer for 20 min (1 mM EDTA pH 7.4) to retrieve antigens. Endogenous peroxidase activity was quenched by incubation in 3% H 2 O 2 for 10 min at RT. The slides were washed in PBS and incubated in 5% NGS for 1 h at RT. Then, the sections were incubated with anti-NCX-2 antibody (1:200 in 1% NGS) o/n at 4 • C. The following day, the slides were washed and incubated with a biotinylated secondary antibody to rabbit IgG for 1 h at RT (1:100, Vector Laboratories, Peterborough, UK) followed by incubation with streptavidin-conjugated horseradish peroxidase for 1 h at RT (1:100, ABC kit Vectastain, Vector Laboratories, Peterborough, UK). The antigen-antibody complex was visualised by incubating the sections with 3,3-diaminobenzidine hydrochloride (DAB) (Sigma, St. Louis, MO, USA) dissolved in PBS with 10 µL of 3% H 2 O 2 . Negative controls were incubated only with the secondary antibody. DAB intensity of images was then quantified using ImageJ software and Colour Deconvolution plugin.

Immunofluorescence for NCX2
Three-µm-thick paraffin DRG sections were deparaffinised with xylene and rehydrated and non-specific tissue binding was blocked with 5% NGS and 5% BSA in PBS for 1 h. Samples were then incubated overnight at 4°C with NCX2 primary rabbit antibody (1:200, OriGene, Rockville, MD, USA) and with a goat anti-rabbit secondary antibody Alexa red 546 (1:200, Invitrogen Waltham, MA, USA) for 1 h at room temperature. Sections were then examined, and stitched images were acquired using an epifluorescence microscope (Cell observer. Zeiss, Germany). Fluorescence intensity of images was then quantified using ImageJ software.

NCS
Nerve conduction studies were performed with the electromyography apparatus Matrix Light (Micromed, Mogliano Veneto, Italy) and performed as previously described [72]. Briefly, subdermal needle electrodes were used (Ambu Neuroline (Ambu, Ballerup, Denmark)). Recordings were performed under deep isofluorane anesthesia and animal body temperature was monitored and kept constant at 37 ± 0.5 • C with a thermal pad (Harvard Apparatus, Holliston, MA, USA). Caudal nerve and digital nerves were assessed. Caudal nerve conduction study was performed positioning a pair of recording needle electrodes at the base of the tail (inter-electrode distance: 0.5 cm) and a pair of stimulating needle electrodes (inter-electrode distance: 0.5 cm) was placed 3.5 cm distally with respect to the active recording electrode; the ground electrode was placed 1 cm distally to the active recording electrode. Digital nerve conduction study was obtained positioning the positive recording electrode in front of the patellar bone, the negative recording electrode close to ankle bone, the positive and negative stimulating electrodes close to the fourth toe near the digital nerve and under the paw, respectively; the ground electrode was placed in the sole. Intensity, duration, and frequency of stimulation were set up in order to obtain optimal results. Averaging technique was applied carefully. Recordings filters were kept between 20 Hz and 3 KHz. Sweep was kept at 0.5 ms.

NET
During recordings, mice were under deep isoflurane anaesthesia and body temperature was kept constant at 37 ± 0.5 • C, as stated above. The montage to perform caudal nerve recordings was adapted slightly, modifying the previously devised protocol by Boërio et al. [73]. Briefly, disposable, non-polarizable, subdermal platinum iridium needle electrodes (MedCat Supplies, Klazienaveen, The Netherlands) were used for stimulation; as ground and recording electrodes, subdermal needle electrodes were instead used (Ambu Neuroline (Ambu, Ballerup, Denmark)). The stimulating cathode was placed at the base of the tail and the anode was placed in the base of the rear foot, on the same side of the caudal nerve tested (left). Recording needle electrodes were inserted subcutaneously in the left side of the tail (interelectrode distance: 1 cm), 3.5 cm distally respect to the cathode. Ground electrode was inserted in the right side of the tail, in mid between the cathode and the active recording electrode. As a stimulator, an isolated linear bipolar constant current stimulant (maximal output ±10 mA, DS4, Digitimer, Welwyn Garden City, UK); the Xcell3 Microelectrode Amplifier (FHC, Bowdoin, ME) was connected to the recording electrodes via a customised probe/adapter specifically designed by FHC for our needs; National Instrument USB-6221-BNC Acquisition Device (National Instrument Italy, Assago, Italy) was used to connect all these instruments. For NET recordings, Qtrac© software (Institute of Neurology, Queen Square, London, UK) and TROND protocol were used. For the purposes of this study, Recovery Cycle is the curve of interest. Recovery Cycle was performed using paired pulse stimuli: a supramaximal stimulus followed by a conditioning one at different interstimulus intervals (2-200 ms) [74]. Threshold changes at 2 and 2.5 ms interstimulus intervals were used to determine refractoriness. The minimum mean threshold change of three adjacent points defined superexcitability; instead, subexcitability was determined as the minimum mean threshold change obtained after a 10 ms interstimulus intervals.

Statistical Analyses
Descriptive statistics were generated for all variables. Normally distributed data were analyzed with parametric tests (t-test, one-way Anova followed by Dunnet test) and nonnormally distributed with non-parametric tests (Mann-Whitney U-test, Kruskal-Wallis test, followed by pairwise comparison, adjusted by the Bonferrroni correction for multiple tests). Two-sided tests were used. A p-value < 0.05 was set as significant. All analyses were conducted in GraphPad environment (v4.0), apart from NET recordings. The latter were analyzed with QtraS© (Institute of Neurology, Queen Square, London, UK), specifically designed to dialogue with the recording software QtraC© (Institute of Neurology, Queen Square, London, UK).

Conclusions
Our journey from the bench side to bed side and vice versa allowed us to pave the way to a new line of research exploiting NCX2 modulation to prevent OIPN-related ADa, highlighting a novel druggable target to cure this condition and, potentially, other conditions in which ADa ensues, as consequences of Na + VOC/NCX family alterations.  Data Availability Statement: Data will be made available upon request to the corresponding author.