WO2016020504A1

WO2016020504A1 - Gap junction blockers for the treatment or prevention of oxaliplatin-induced neuropathy

Info

Publication number: WO2016020504A1
Application number: PCT/EP2015/068212
Authority: WO
Inventors: Kleopas A. KLEOPA; Alexia KAGIAVA; Georgios THEOFILIDIS
Original assignee: The Cyprus Foundation For Muscular Dystrophy Research
Priority date: 2014-08-08
Filing date: 2015-08-06
Publication date: 2016-02-11
Also published as: GB201414122D0

Abstract

A gap junction blocker for use in preventing or treating oxaliplatin-induced neuropathy in a subject who is, has been, or is going to be, administered oxaliplatin. The gap junction blocker may close CX31.3 hemichannels or Cx32 GJ channels. The gap junction blocker may be octanol or a metabolite or prodrug thereof, for example octanoic acid.

Description

GAP JUNCTION BLOCKERS FOR THE TREATMENT OR PREVENTION OF

OXALIPLATIN-INDUCED NEUROPATHY

The present invention relates to combatting oxaliplatin-induced neurotoxicity. The world-wide number of patients receiving chemotherapy with oxaliplatin is estimated to be over a million. Oxaliplatin is used extensively as a first-line drug in gastrointestinal cancer chemotherapy, in particular metastatic colorectal cancer. Colorectal cancer is the third most common cancer in the world, with nearly 1.4 million new cases diagnosed in 2012.

The major dose-limiting toxicity of oxaliplatin is peripheral neuropathy that can manifest in over 60% of treated subjects. Oxaliplatin-induced neurotoxicity (OIN) can present both as an acute or as chronic neuropathy. Acute OIN manifestations are characterized by peripheral nerve hyperexcitability symptoms, including cold-induced perioral or pharyngolaryngeal dysesthesias (3).

Assuming that a million patients will have to have chemotherapy with oxaliplatin yearly, then over 800,000 subjects are likely to experience peripheral neuropathy, preventing the full dose of oxaliplatin from being administered. It is considered that there is a need for therapeutic agents which can either prevent the OIN from occurring in the first place in a subject being treated with oxaliplatin, or alleviate the OIN which may have occurred in the subject being treated with oxaliplatin.

Although oxaliplatin is used extensively in cancer treatment, its mechanism of anti-cancer action remains poorly understood. It is known to be activated by non-enzymatic hydrolysis to oxalate, which forms highly reactive compounds that can bind to proteins, RNA, and nuclear DNA (2). Likewise, the mechanism by which oxaliplatin causes neuropathy is also not understood, contributing to the lack of effective treatments or preventatives for OIN. The major cause for oxaliplatin peripheral neuropathy is thought to be the hyperexcitability in the peripheral nerve fibres. The origin of OIN and hyperexcitability symptoms remains controversial, although a variety of mechanisms have been proposed. Some studies implicated the involvement of voltage-gated potassium channels (VGKCs) (4-10). Peripheral nerve fibre hyperexcitability that develops acutely following oxaliplatin exposure resembles neuromyotonia, a channelopathy linked to dysfunction of juxtaparanodal VGKCs (11-13). Another proposed mechanism for 01 N is the prolongation of the inactivation stage of voltage-gated sodium channels (VGNaCs), supported by some ex vivo (14, 15) and in vivo studies (16-18). Finally, emphasis has been given to the role of oxalate liberated by oxaliplatin, a well-known chelator of both calcium and magnesium (19-22). However, the use of the VGNaC blocker carbamazepine in the clinical setting failed to benefit patients (23, 24) and although VGKC dysfunction could play a role in OIN, patch-clamp studies failed to show any direct effects of oxaliplatin on Shaker-type Kv1.1/1.2 channels (25). Finally, the role of oxalate has been questioned (26) since a recent study failed to demonstrate any protective effect of calcium/magnesium treatment against OIN (27).

Previous ex vivo studies (preclinical studies) on the isolated sciatic nerve fibers of the rat exposed to oxaliplatin suggested that the physiological correlate of this hyperexcitability is in the form of a post-stimulus severe broadening of the repolarising phase of the evoked nerve compound action potential (CAP), called here the "oxaliplatin-effect", which develops in a concentration and time dependent manner (6). Moreover, in the same nerve preparation, intra-axonal recordings from nerve fibres exposed to oxaliplatin have shown that the hyperexcitability occurs as a broadening of the repolarising phase of the evoked action potential while the depolarising phase and the resting membrane potential of the axon was not affected (28). This broadening recorded intra-axonally obviously affects the wave form of the evoked CAP recorded extracellularly producing the oxaliplatin-effect in the whole nerve recordings (28). Such an effect is unique in the physiology of the adult mammalian axons and has been shown only in intra-axonal recordings from nerve fibres of young rats treated with 4-aminopyridine (4-AP) a well-known blocker of VGKCs (29), while 4-AP or TEA had no effect on adult myelinated nerve fibres (6).

Thus, although previous ex vivo studies indicated that VGKCs dysfunction could play a role in OIN, as also suggested by others, the inventors consider it unlikely that VGKCs are the primary target of oxaliplatin because oxaliplatin does not bind to VGKCS (25). Since VGKCs are responsible for the repolarizing phase of the action potential, which is drastically affected by oxaliplatin, the inventors hypothesized that oxaliplatin may affect a functionally related target, leading indirectly to VGKC dysfunction.

The effect of oxaliplatin occurs only in myelinated fibers, since oxaliplatin induces bursts of action potentials in myelinated A-fibers, but not in unmyelinated C-fibers (9). Directly apposing axonal VGKCs, the gap junction (GJ) protein connexin29 (Cx29, the human ortholog is Cx31.3) forms GJ hemichannels in the innermost membrane of myelin surrounding the juxtaparanodal regions (30, 31 ). A gap junction is a specialized intercellular connection between a multitude of animal cell- types. It directly connects the cytoplasm of two cells, which allows various molecules and ions to pass freely between cells. One gap junction channel is composed of two connexons (or hemichannels), which connect across the intercellular space. Unopposed hemichannels may also be found on cell membranes connecting the cytoplasm to the extracellular space. Gap junction proteins are differentially expressed throughout the different tissues and cells types, and can have different functions.

Previous studies suggested possible involvement of gap junctions in OIN, and showed upregulation of Cx43 in spinal astrocytes, as a possible mechanism (Yoon J Pain 2013), but without proof of a specific role in OIN. Cx43 is a connexin of 381 amino acid residues (human isoform) that is widely expressed in several organs and cell types, and is the principal gap junction protein of the heart. Cx43 also forms gap junctions in astrocytes. Upregulation of interastrocytic Cx43 gap junctions is a non-specific reaction to many pathologies, including neuroinflammatory conditions such as multiple sclerosis (Markoullis ef al 2012, Acta Neuropathol), and neurodegenerative conditions such as motor neuron disease (Cui ef al 2014 J Neuroinflamm). The increased expression of Cx43 in the spinal cord may be as a secondary consequence of injury to the peripheral nervous system rather than from direct damage to the central nervous system, as an increased expression of Cx43 is also observed in the spinal cord following chronic constriction of the sciatic nerve (Wu ef al 2011 Pain 152: 2605-2615). Cx43 is also expressed in satellite cells of the dorsal root ganglia and has been shown to be upregulated in these cells following activation of the satellite cells (Warwick and Hanani Eur J Pain 2013), but this again is likely to be as a secondary consequence to nerve injury elsewhere. The gap junction protein connexin29 (mouse Cx29, the human ortholog is Cx31.3) forms GJ hemichannels in the innermost membrane of myelin surrounding the juxtaparanodal regions, directly apposing axonal VGKCs (30, 31). Due to the proximity to Kv1.1/1.2, Cx29 hemichannels have been proposed to open and allow the surplus of K⁺ released during the action potential, from the juxtaparanodal region into Schwann cell cytoplasm (30). Since Cx29 hemichannels are only found in the innermost myelin layers, this functional pathway for the regulation of K⁺ likely also involves Cx32 GJ channels that provide a passage through the entire myelin sheath reaching the abaxonal Schwann cell cytoplasm (Scherer et al., 1995, Balice-Gordon et al., 1998, Kleopa, 2011). Even a modest fluctuation in K+ concentration in the narrow juxtaparanodal space could have an impact on different axonal properties (32) and an increase of extracellular K⁺ in the periaxonal region could lead to hyperexcitation of myelinated nerve fibres with only a minor effect on the resting membrane potential (33, 34).

Oxaliplatin can affect the expression levels of GJs in astrocytes (35) and satellite glial cells (36, 37), while GJ blockage by carbenoxolone results in analgesic-like effects. Moreover, gap junctional communication can counteract the effects of the anti-tumour agent cisplatin (a platinum based compound, as is oxaliplatin) (38) and studies on taxol and oxaliplatin suggested that GJ blockers may have potential in treating chemotherapy-induced neuropathic pain (37). However, carbenoxolone, at the higher doses required to block gap junctions, produces an undesirable anaesthetic-analgesic effect through effects on nodal sodium channels, and may therefore be a less promising candidate for use in counteracting the neurotoxic effects of oxaliplatin in subjects.

EP1465642 discloses the protection against the neurotoxicity of oxaliplatin through the administration of calcium and magnesium.

US20140128461 discloses the use of a food composition in the treatment and/or prevention of neuropathic pain induced by an anticancer agent, for example oxaliplatin.

US20090143464 discloses a method for preventing and/or treating peripheral Neuropathies induced by the administration of an anticancer agent by administering an effective amount of acetyl L-carnitine.

WO 2012145098 discloses the use of adenosine a3 receptor agonists for treatment of neuropathic pain.

CA 2606042 discloses a method of treating an oxaliplatin-sensitive cancer are in which a therapy comprising oxaliplatin is initiated, continued until a predetermined endpoint, stopped, then reintroduced after specific criteria are met. WO 2007044437 discloses octanol formulations and methods of treatment of involuntary tremors. In their preclinical study on mice the inventors of the present invention by examining directly the relevant site of OIN, the peripheral nerve itself, found that rather than being mediated by Cx43 suggested in the prior art, the OIN is actually mediated by Cx31.3 (Cx29 in mice) hemichannels and Cx32 GJ channels in the peripheral nerve, rather than the CNS as previously described. The functional and the morphological changes induced by oxaliplatin on the peripheral nerve fibres of the mice were as a result of the malfunction of Cx29 hemichannels (or GJ) or Cx32 GJ channels and were found to be reversed by the GJ blocker octanol, a known and clinically safe gap junction blocker. Without wishing to be bound by any theory, the inventors believe that oxaliplatin causes a forced opening of the Cx29 gap junction hemichannels and Cx32 GJ channels in the peripheral nerve fibres, creating stagnant K⁺ near the Kv channels. The juxtaparanodal region is filled with excess K⁺ and this extracellular K⁺ affects the function of Kv1.1 and Kv1.2 regulating the repolarising phase. Moreover, oxaliplatin accelerates the opening of Cx31.3 hemichannels and Cx32 GJ channels expressed in cultured cells. Octanol, or other gap junction blockers, close the Cx31.3/Cx29 hemichannels and Cx32 GJ channels in both experimental settings, and compensate for the effect of oxaliplatin, allowing their proper function. It will be appreciated from the data given in the examples that gap junction blockers, for example octanol or a metabolite or prodrug thereof, for example octanoic acid, can be used to treat or prevent oxaliplatin induced acute and chronic neuropathy manifestations, and can be used to protect nerve fibres from the severe electrophysiological and morphological effects of oxaliplatin described above. Accordingly, an aspect of the present invention provides a gap junction blocker, for example octanol or a metabolite or prodrug thereof, for example octanoic acid, for use in preventing or treating oxaliplatin-induced neuropathy in a subject who has been administered or is going to be administered oxaliplatin.

Octanol is considered to have the following formula and structure: 1 -Octanol - CH₃(CH₂)70H

Octanoic acid (also known as Caprylic acid) is an eight-carbon saturated fatty acid, which occurs naturally in palm and coconut oils, as well as in human and bovine milk, and is part of commercially available nutritional supplements (CaprinolH, CaprylH), for example intended to reduce yeast growth in the gut. Octanoic acid is considered to have the following formula and structure:

Octanoic acid is considered to be a metabolite of 1 -octanol. See, for example (49). An open-label, single-dose, crossover study of the pharmacokinetics and metabolism of two oral formulations of 1 -octanol in patients with essential tremor.

The gap junction blocker is considered to act to negatively affect the gap junction or associated proteins. In a preferred embodiment the gap junction blocker acts to negatively affect the gap junction to close the gap junction. In a particularly preferred embodiment the gap junction blocker is octanol or a metabolite or prodrug thereof, such as octanoic acid. In a preferred embodiment, the gap junction blocker, for example octanol, or a metabolite or prodrug thereof, for example octanoic acid, closes the gap junction. The gap junction blocker may close only one type of gap junction, may close two types of gap junctions, or more close three or more types of gap junctions. The gap junction blocker may close the more than one type of gap junction to the same degree, or may close one or more types of gap junctions more than one or more other types of gap junctions.

By closes we include the meaning of directly closes the pre-existing gap junction, and indirectly closing the gap junction, resulting in an overall net effect of closure of the one or more types of gap junctions. For example, a gap junction blocker which acts by, for example, causing downregulation of the expression of a particular connexin, is said to close the gap junction as the net effect of decrease connexin protein production is considered to be an overall decrease in the activity of the gap junctions. Thus in one embodiment, the gap junction blocker is an agent which causes forced closing of a gap junction channel or hemichannel. Preferably the gap junction blocker causes forced closing of the Cx31.3/Cx29 gap junction and/or the Cx32 GJ channel. It will be appreciated that the gap junction blocker, for example octanol or a metabolite or prodrug thereof, for example octanoic acid, may act directly at the site of the gap junction, or may act in such a way upstream as to effect a change at the gap junction, for example the gap junction blocker may positively or negatively affect the expression level of proteins associated with the gap junction or required for the gap junction formation, and affect it in such a way as to result in closure of the gap junction. The gap junction blocker may affect protein modification, for example protein phosphorylation. In one embodiment the gap junction supresses the expression of a connexin or connexins, preferably connexin Cx31.3/Cx29 and/or Cx32. In another embodiment the gap junction blocker, for example octanol or a metabolite or prodrug thereof, for example octanoic acid, acts at the site of the gap junction and affects gap junction function or structure. The gap junction blockermay, for example, bind directly to the connexin proteins forming the gap junction, or may bind to associated or nearby proteins to affect the gap junction structure or function, resulting in gap junction closure. In a preferred embodiment the gap junction blocker affects the activity or function of the Cx31.3/Cx29 hemichannel gap junction and/or Cx32 gap junction. Preferably the gap junction blocker acts to close the Cx31 ,3/Cx29 hemichannel gap junction and/or Cx32 gap junction. Typically the gap junction blocker is not specific, and can affect various gap junctions comprised of different connexins. In one embodiment the gap junction blocker is not specific but does affect the Cx31.3/Cx29 connexin and/or Cx32. In other cases, the gap junction blocker may be specific for a particular connexin gap junction. In one embodiment, the gap junction blocker is specific for the Cx31.3/Cx29 gap junction and/or Cx32 gap junction, for example the gap junction blocker may be an anti-Cx31.3/Cx29 antibody (or fragment thereof) or anti- Cx32 antibody (or fragment thereof) or mimetic peptide which binds to the Cx31.3/Cx29 or Cx32 proteins in such a way as to affect their function, preferably by closing the gap junction. Mimetic peptides and their use in blocking gap junctions is described in Takeuchi ei a/J Biol Chem 2006; 281 :21362-8. In a preferred embodiment, the gap junction blocker acts on the Cx31.3/Cx29 or Cx32 gap junction in the peripheral nerve. Thus it will be understood that the gap junction blocker may be a small molecule, or a protein, for example an antibody, a peptide, a fatty acid, a carbohydrate, a nucleic acid, for example DNA or RNA. Preferably the gap junction blocker is octanol or a metabolite or prodrug thereof, such as octanoic acid.

Agents which act as gap junction blockers are known to those skilled in the art. There may be further agents which are capable of acting as a gap junction blocker in the present invention. An agent which can act as a gap junction blocker of the present invention can be readily determined by those skilled in the art. The examples (Example 2 and Figure 7) provide guidance on, for example, how to use human HeLa cells to determine whether a particular agent is useful in closing a particular gap junction.

Briefly, as an example, HeLa cells are engineered to express a particular connexin, preferably connexin 31.3 or Cx32. Cells are exposed to oxaliplatin with and without the agent under testing (i.e. a potential gap junction blocker of the present invention) and the amount of neurobiotin taken into the cells is assessed. Oxaliplatin opens the gap junctions and allows neurobiotin to enter the cell. A putative gap junction blocker is suitable for the present invention if it reduces the entry of neurobiotin into the cells - preferably reduces entry through the connexin, preferably through the 31 .3 connexin or Cx32 connexin. The gap junction blocker, of the present invention is one which reduces the entry of neurobiotin into the HeLa cells by any significant amount. Preferably the potential gap junction blocker reduces the entry of neurobiotin by at least about 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, 100% compared to cells in the presence of oxaliplatin alone. In a preferred embodiment, the potential gap junction blocker reduces the entry of neurobiotin into the cell by at least about 70%, compared to cells in the presence of oxaliplatin alone, see Example 4 and (Sargiannidou et al Neurobiol Dis 2008).

In a preferred embodiment, the cells, for example HeLa cells, used to determine whether a gap junction blocker is a gap junction blocker of the present invention, are communication incompetent, i.e. they do not naturally express any connexin. The HeLa cells are engineered to express a particular connexin, for example the Cx31.3 connexin and/or Cx32 connexin.

In a preferred embodiment, the gap junction blocker of the present invention, preferably octanol or a metabolite or prodrug thereof, for example octanoic acid, acts in such a way as to prevent the increase in duration of the compound action potential caused by oxaliplatin. Guidance on how the skilled person can monitor the efficacy of a potential gap junction blocker of the present invention in blocking the increase in duration of the compound action potential caused by oxaliplatin is given in Example 1.

The gap junction blocker, for example octanol or a metabolite or prodrug thereof, for example octanoic acid) of the present invention may be a potential therapeutic which has not yet been subjected to clinical safety assessment, or may be a therapeutic or potential therapeutic which has be subjected to clinical safety assessment. Preferably the gap junction blocker is known to be clinically safe, for example safe at the relevant concentrations, for example the gap junction blocker is one that has been used at concentrations that effectively improve neurological symptoms without causing side effects (See, for example 41 , 48 and 49). Preferably the gap junction blocker is known to be clinically safe at high dosages. Preferably the gap junction blocker does not cause anaesthetic-analgesia at high dosages. Preferably the gap junction blocker is not carbenoxolone, which it is understood was tried in the 1970s for peptic ulcer but was considered to give rise to side effects

It will be appreciated that the gap junction blocker can be used in any organism in which oxaliplatin has been, or is going to be administered. The subject may be a mammal, for example a human, a dog, a cat, a mouse, a rat, a horse or a farmed or domesticated animal. Preferably the subject is a human.

Typically, as oxaliplatin is an anti-cancer agent, the subject will be one who has had a positive diagnosis of cancer, for example gastrointestinal cancer, for example metastatic colorectal cancer. It will be appreciated however, that the OIN is not linked to the cancer status of the subject, and therefore the gap junction blocker, preferably octanol or metabolite or prodrug thereof, of the present invention is suitable for use in any subject which has, or is going to have, received oxaliplatin therapy. For example, a new use of oxaliplatin may be discovered for use in treating another disease. This does not render the use of the gap junction blocker of the present invention non-useful. The use of the gap junction blocker of the present invention treat OIN rather than the cancer, or the reason for administration of oxaliplatin itself.

The gap junction blockers of the present invention are for use in treating or preventing oxaliplatin-induced neuropathy (OIN). OIN can present both as an acute or as chronic neuropathy. Acute OIN manifestations are characterized by peripheral nerve hyperexcitability symptoms, including cold-induced perioral or pharyngolaryngeal dysesthesias (motor and sensory dysfunction). Peripheral neuropathy is extremely common in oxaliplatin-treated subjects and may present both as an acute and as a chronic form, believed to result from distinct, but overlapping, pathophysiologic mechanisms. Acute peripheral neuropathy is characterized by paresthesia, dysethesia, or allodynia affecting the extremities, the lips, and the oropharyngolaryngeal area during or shortly after oxaliplatin infusion. It is often triggered by exposure to cold. It usually subsides within a few hours or days (Argyriou et al 2008). (Argyriou AA, Polychronopoulos P, lconomou G, Chroni E, Kalofonos HP. A review on oxaliplatin-induced peripheral nerve damage. Cancer Treat Rev 2008;34:368-77). Chronic oxaliplatin-induced peripheral neuropathy results from cumulative exposure to the drug in about 15% in patients who have received a cumulative dose of about 800 mg/m² (Maindrault-Goebel et al 1999). This chronic neuropathy manifests itself as decreased distal sensations and proprioception, while involvement of motor fibres is rare. It may be irreversible in up to 5% of cases.

By treating OIN we mean the suppression of one, more than one, or all symptoms of OIN, for example those described above, following their presentation subsequent to oxaliplatin administration. It will be standard practice for a clinician to assess the presence and severity of symptoms of OIN, and the clinician will be able to determine whether or not the subject is one which requires treatment with a gap junction blocker of the present invention. By preventing OIN we mean the prevention of one, more than one, or all symptoms of OIN from presenting subsequent to oxaliplatin administration.

The gap junction blockers of the present invention may be administered prior to, concomitantly with, or subsequent to, the administration of oxaliplatin. It will be appreciated that where the aim is to prevent the occurrence of OIN, the gap junction blockers, preferably octanol, will be administered prior to or concomitantly with the oxaliplatin. Typically, administration of the gap junction blocker prior to administration of oxaliplatin can occur at any time preceding administration of oxaliplatin. The gap junction blocker may be administered, for example, 2 days, 1 day, 18 hours, 12 hours, 6 hours, 4 hours, 2 hours, 1 hours, 40 minutes, 20 minutes, 10 minutes, 5 minutes, 1 minute prior to administration of oxaliplatin.

In a preferred embodiment, the gap junction blocker, for example octanol or a metabolite or prodrug thereof, for example octanoic acid, is co-administered with oxaliplatin.

The gap junction blocker may be co-administered as part of the same composition, or may be administered separately but at the same time. Therefore in one embodiment the gap junction blocker, preferably octanol, a metabolite or prodrug thereof, is supplied as part of a composition which also comprises oxaliplatin. The composition may also comprise one or more further therapeutic agents, for example an additional anti-cancer agent or an analgesic agent.

The gap junction blocker may be part of a composition that does not comprise oxaliplatin, but does contain one or more other further therapeutic agents for example an additional anti-cancer agent or an analgesic agent. Alternatively, the gap junction blocker and the oxaliplatin may be two separate agents which are mixed together prior to administration, for example the two agents may be part of a kit of parts. Therefore in another embodiment the gap junction blocker, preferably octanol, a metabolite or prodrug thereof, is supplied as part of a kit of parts which further comprises one or more of oxaliplatin or further therapeutic agents, for example an additional anti-cancer agent or an analgesic agent.

It is preferred that the gap junction blocker and the oxaliplatin are co-administered, i.e. administered at the same time, whether as a single composition or not. Preferably the gap junction blocker is administered each time the subject requires chemotherapy with oxaliplatin. It will be appreciated that between chemotherapy treatments with oxaliplatin, a subject may require an additional dose or doses of the gap junction blocker to prevent or treat the OIN.

In one preferred embodiment, the gap junction blocker is administered before the oxaliplatin, or at the same time as administration of the oxaliplatin. Administration of the gap junction blocker post administration of the oxaliplatin may be less helpful. Therefore in a preferred embodiment the gap junction blocker is not administered after or subsequent to the administration of the oxaliplatin. Typically, the gap junction blocker, preferably octanol or metabolite or prodrug thereof, is administered between (and including) 0 minutes and 7 days before administration of oxaliplatin. For example the gap junction blocker, preferably octanol or metabolite or prodrug thereof, is administered at least 5 minutes, or at least 10 minutes, or at least 30 minutes, or at least 40 minutes, or at least 1 hour, or at least 2 hours, or at least 4 hours, or at least 6 hours, or at least 12 hours, or at least 24 hours, or at least 48 hours, or at least 4 days, before administration of oxaliplatin. The gap junction blocker, preferably octanol or metabolite or prodrug thereof may be administered more than 7 days before administration of the oxaliplatin, In one embodiment repeated doses of the gap junction blocker are administered before oxaliplatin, and one or more of these doses may be administered more than 7 days before administration of the oxaliplatin. Where the aim is to treat OIN wherein the symptoms are already present, the gap junction blocker may be administered subsequent to the administration of oxaliplatin. Typically, the gap junction blocker, preferably octanol or metabolite or prodrug thereof, is administered between (and including) 0 minutes and 7 days after administration of oxaliplatin. For example the gap junction blocker, preferably octanol or metabolite or prodrug thereof, is administered at least 5 minutes, or at least 10 minutes, or at least 30 minutes, or at least 40 minutes, or at least 1 hour, or at least 2 hours, or at least 4 hours, or at least 6 hours, or at least 12 hours, or at least 24 hours, or at least 48 hours, or at least 4 days, after administration of oxaliplatin. The gap junction blocker, preferably octanol or metabolite or prodrug thereof may be administered more than 7 days after administration of the oxaliplatin, In one embodiment repeated doses of the gap junction blocker are administered after oxaliplatin, and one or more of these doses may be administered more than 7 days after administration of the oxaliplatin.

In a further embodiment, the gap junction blocker, preferably octanol or metabolite or prodrug thereof is administered both before administration of the gap junction blocker, and after administration of the gap junction blocker. In this embodiment the gap junction blocker, preferably octanol or metabolite or prodrug thereof id administered between (and including) 0 minutes and 7 days before administration of oxaliplatin and between (and including) 0 minutes and 7 days after administration of oxaliplatin. For example the gap junction blocker, preferably octanol or metabolite or prodrug thereof, is administered at least 5 minutes, or at least 10 minutes, or at least 30 minutes, or at least 40 minutes, or at least 1 hour, or at least 2 hours, or at least 4 hours, or at least 6 hours, or at least 12 hours, or at least 24 hours, or at least 48 hours, or at least 4 days, before administration of oxaliplatin, and is again administered either with the oxaliplatin, for example 0 minutes before and after administration of oxaliplatin, and/or at least 5 minutes, or at least 10 minutes, or at least 30 minutes, or at least 40 minutes, or at least 1 hour, or at least 2 hours, or at least 4 hours, or at least 6 hours, or at least 12 hours, or at least 24 hours, or at least 48 hours, or at least 4 days, after administration of oxaliplatin. The gap junction blocker, preferably octanol or metabolite or prodrug thereof may be administered more than 7 days before and after administration of the oxaliplatin. In one embodiment repeated doses of the gap junction blocker are administered before oxaliplatin with one or more subsequent doses of the gap junction blocker being administered with the oxaliplatin, and/or after administration of the oxaliplatin. In one embodiment repeated doses of the gap junction blocker are administered after oxaliplatin, with one or more previous doses of the gap junction blocker being administered with the oxaliplatin, and/or before administration of the oxaliplatin.

It will be appreciated that administration of any agent described herein is typically administered as part of a pharmaceutical composition together with a pharmaceutically acceptable excipient, diluent, adjuvant, or carrier. Thus, any mention of a gap junction blocker, for example octanol, any mention of oxaliplatin, and any mention of a further therapeutic agent, equally applies to a pharmaceutically acceptable composition comprising that gap junction blocker, oxaliplatin, and/or further therapeutic agent (e.g. a formulation).

Routes of administration will be known to those skilled in the art. For example, the agents (gap junction blocker (preferably octanol or a metabolite or prodrug thereof), oxaliplatin and/or further therapeutic agents) can be administered orally, buccally or sublingually in the form of tablets, capsules, ovules, elixirs, solutions or suspensions, which may contain flavouring or colouring agents, for immediate-, delayed- or controlled-release applications.

The compounds of invention may also be administered via intracavernosal injection. Preferably the gap junction blocker of the present invention is administered orally, more preferably the gap junction blocker is octanol or a metabolite or prodrug thereof, and is administered orally.

The agents may be administered orally or by any parenteral route, in the form of a pharmaceutical formulation comprising the active ingredient, optionally in the form of a non-toxic organic, or inorganic, acid, or base, addition salt, in a pharmaceutically acceptable dosage form. Depending upon the subject to be treated, as well as the route of administration, the agents may be administered at varying doses.

Preferably, the formulation is a unit dosage containing a daily dose or unit, daily sub-dose or an appropriate fraction thereof, a weekly dose, a monthly dose, or a 6 monthly dose of the agent or active ingredient. In human therapy, the agents (e.g. gap junction blocker (preferably octanol or a metabolite or prodrug thereof), oxaliplatin and/or further therapeutic agents) can be administered alone but will generally be administered in admixture with a suitable pharmaceutical excipient diluent or carrier selected with regard to the intended route of administration and standard pharmaceutical practice.

Tablets may contain excipients such as microcrystalline cellulose, lactose, sodium citrate, calcium carbonate, dibasic calcium phosphate and glycine, disintegrants such as starch (preferably corn, potato or tapioca starch), sodium starch glycollate, croscarmellose sodium and certain complex silicates, and granulation binders such as polyvinylpyrrolidone, hydroxypropylmethylcellulose (HPMC), hydroxy-propylcellulose (HPC), sucrose, gelatin and acacia. Additionally, lubricating agents such as magnesium stearate, stearic acid, glyceryl behenate and talc may be included. Capsules or tablets may also be enteric coated to enhance gastric stability.

Solid compositions of a similar type may also be employed as fillers in gelatin capsules. Preferred excipients in this regard include lactose, starch, a cellulose, milk sugar or high molecular weight polyethylene glycols. For aqueous suspensions and/or elixirs, the compounds of the invention may be combined with various sweetening or flavouring agents, colouring matter or dyes, with emulsifying and/or suspending agents and with diluents such as water, ethanol, propylene glycol and glycerin, and combinations thereof.

The agents (e.g. gap junction blocker (preferably octanol or a metabolite or prodrug thereof), oxaliplatin and/or further therapeutic agents) may also be administered parenterally, for example, intravenously, intra-arterially, intraperitoneally, intrathecally, intraventricularly, intrasternally, intracranially, intra-muscularly or subcutaneously, or they may be administered by infusion techniques. They are best used in the form of a sterile aqueous solution which may contain other substances, for example, enough salts or glucose to make the solution isotonic with blood. The aqueous solutions should be suitably buffered (preferably to a pH of from 3 to 9), if necessary. The preparation of suitable parenteral Formulations under sterile conditions is readily accomplished by standard pharmaceutical techniques well-known to those skilled in the art.

Preferably the oxaliplatin is administered intravenously. In a most preferred embodiment, the gap junction blocker is octanol, or metabolite or prodrug thereof and is administered orally, and the oxaliplatin is administered intravenously. Formulations suitable for parenteral administration include aqueous and non-aqueous sterile injection solutions which may contain anti-oxidants, buffers, bacteriostats and solutes which render the Formulation isotonic with the blood of the intended recipient; and aqueous and non-aqueous sterile suspensions which may include suspending agents and thickening agents. The Formulations may be presented in unit-dose or multi-dose containers, for example sealed ampoules and vials, and may be stored in a freeze-dried (lyophilised) condition requiring only the addition of the sterile liquid carrier, for example water for injections, immediately prior to use. Extemporaneous injection solutions and suspensions may be prepared from sterile powders, granules and tablets of the kind previously described.

For oral and parenteral administration to human subjects, the daily dosage level of the agents (e.g. gap junction blocker (preferably octanol or metabolite or prodrug thereof), and/or further therapeutic agents) will usually be from 70 to 5000 mg (for example up to 40, 50, 60, 64, 70, 80 or 100 mg/Kg bodyweight, for example up to 64 mg/Kg bodyweight as used in previous trials with octanol) per adult, administered in single or divided doses.

Thus, for example, the tablets or capsules of the compound of the invention may contain from 1 mg to 1000 mg (i.e. from about 60-120 mg/m²) of active compound for administration singly or two or more at a time, as appropriate. The physician in any event will determine the actual dosage which will be most suitable for any individual subject and it will vary with the age, weight and response of the particular subject. The above dosages are exemplary of the average case. There can, of course, be individual instances where higher or lower dosage ranges are merited and such are within the scope of this invention. The agents (e.g. gap junction blocker (preferably octanol or metabolite or prodrug thereof), oxaliplatin and/or further therapeutic agents) may also be administered intranasally or by inhalation and are conveniently delivered in the form of a dry powder inhaler or an aerosol spray presentation from a pressurised container, pump, spray or nebuliser with the use of a suitable propellant, e.g. dichlorodifluoromethane, trichlorofluoromethane, dichlorotetrafluoro-ethane, a hydrofluoroalkane such as 1 ,1 ,1 ,2-tetrafluoroethane (HFA 134A3 or 1 , 1 ,1 , 2,3,3, 3-heptafluoropropane (HFA 227EA3), carbon dioxide or other suitable gas. In the case of a pressurised aerosol, the dosage unit may be determined by providing a valve to deliver a metered amount. The pressurised container, pump, spray or nebuliser may contain a solution or suspension of the active compound, e.g. using a mixture of ethanol and the propellant as the solvent, which may additionally contain a lubricant, e.g. sorbitan trioleate. Capsules and cartridges (made, for example, from gelatin) for use in an inhaler or insufflator may be formulated to contain a powder mix of a compound of the invention and a suitable powder base such as lactose or starch.

Aerosol or dry powder formulations may preferably be arranged so that each metered dose or "puff contains at least 1 mg of an agent (e.g. gap junction blocker (preferably octanol or metabolite or prodrug thereof), oxaliplatin and/or further therapeutic agents) for delivery to the subject. It will be appreciated that he overall daily dose with an aerosol will vary from subject to subject, and may be administered in a single dose or, more usually, in divided doses throughout the day.

Alternatively, the agents (e g. gap junction blocker (preferably octanol or metabolite or prodrug thereof), oxaliplatin and/or further therapeutic agents) can be administered in the form of a suppository or pessary, or they may be applied topically in the form of a lotion, solution, cream, ointment or dusting powder. The compounds of the invention may also be transdermal^ administered, for example, by the use of a skin patch. They may also be administered by the ocular route, particularly for treating diseases of the eye.

For ophthalmic use, the agents (e.g. gap junction blocker (preferably octanol or metabolite or prodrug thereof), oxaliplatin and/or further therapeutic agents) may be formulated as micronised suspensions in isotonic, pH adjusted, sterile saline, or, preferably, as solutions in isotonic, pH adjusted, sterile saline, optionally in combination with a preservative such as a benzylalkonium chloride. Alternatively, they may be formulated in an ointment such as petrolatum. For application topically to the skin, the agents (e.g. gap junction blocker (preferably octanol or metabolite or prodrug thereof), oxaliplatin and/or further therapeutic agents) may be formulated as a suitable ointment containing the active compound suspended or dissolved in, for example, a mixture with one or more of the following: mineral oil, liquid petrolatum, white petrolatum, propylene glycol, polyoxyethylene polyoxypropylene compound, emulsifying wax and water. Alternatively, they may be formulated as a suitable lotion or cream, suspended or dissolved in, for example, a mixture of one or more of the following: mineral oil, sorbitan monostearate, a polyethylene glycol, liquid paraffin, polysorbate 60, cetyl esters wax, cetearyl alcohol, 2-octyldodecanol, benzyl alcohol and water.

Formulations suitable for topical administration in the mouth include lozenges comprising the active ingredient in a flavoured basis, usually sucrose and acacia or tragacanth; pastilles comprising the active ingredient in an inert basis such as gelatin and glycerin, or sucrose and acacia; and mouth-washes comprising the active ingredient in a suitable liquid carrier.

Generally, in humans, oral or topical administration of the agents (e.g. gapjunction blocker (preferably octanol or metabolite or prodrug thereof), oxaliplatin and/or further therapeutic agents) may be the preferred route, being the most convenient. In circumstances where the recipient suffers from a swallowing disorder or from impairment of drug absorption after oral administration, the drug may be administered parenterally, e.g. sublingually or buccally.

For veterinary use, the agent (e.g. gap junction blocker (preferably octanol or metabolite or prodrug thereof), oxaliplatin and/or further therapeutic agents) is administered as a suitably acceptable formulation in accordance with normal veterinary practice and the veterinary surgeon will determine the dosing regimen and route of administration which will be most appropriate for a particular animal.

Conveniently, the formulation is a pharmaceutical formulation. The formulation may be a veterinary formulation. It will be appreciated that the term administration is not restricted to a one time administration. The term administration is taken to cover all of, but not limited to, a single dose administration, multiple administrations over a period of time, variable dosage administrations over a period of time, variable means of administration over a period of time, administration in conjunction with one or more further therapeutic agents. Administration can be by any means known in the art and includes, but is not limited to, oral, intravenous, topically direct to a tumour, sublingually or suppository.

By a subject which has been administered oxaliplatin we mean any subject which has been administered oxaliplatin in any dosage and via any route, whether in combination with other agents or not. By a subject which is going to be administered oxaliplatin, we mean any subject which is going to be administered oxaliplatin, at any time, for any purpose be it therapeutic or screening or research.

As described earlier, the gap junction blocker of the present invention, preferably octanol or metabolite or prodrug thereof, for example octanoic acid, can be administered at the same time as the oxaliplatin. Therefore a further aspect of the invention provides a composition comprising a gap junction blocker and oxaliplatin for use in treating a subject in need of oxaliplatin therapy. The subject in need of oxaliplatin therapy may be a subject which has cancer, optionally wherein the cancer is gastrointestinal cancer, optionally metastatic colorectal cancer. As gap junction blockers alleviate OIN, it will be appreciated that the present invention also provides a method of treating oxaliplatin-induced neurotoxicity wherein said method comprises administering a gap junction blocker. Preferences for the gap junction blocker, their administration and all other definitions related to this aspect of the invention are as those defined in relation to the first aspect of the invention.

Also provided is a method of preventing oxaliplatin-induced neurotoxicity wherein said method comprises administering a gap junction blocker prior to administration of oxaliplatin. It will be appreciated that oxaliplatin may be used in a subject in need of oxaliplatin therapy, wherein the subject is one which has, or is going to be administered a gap junction blocker. Therefore the present invention also provides oxaliplatin for use in treating a subject in need of oxaliplatin therapy, wherein the subject is, has been, or is going to be, administered a gap junction blocker.

The invention also provides a kit of parts comprising a gap junction blocker and oxaliplatin, optionally wherein the kit of parts further comprises one or more further therapeutic agents.

A gap junction blocker is also provided as an aspect of the invention for use in the manufacture of a medicament for use in treating or preventing oxaliplatin-induced neurotoxicity.

Preferences and definitions for the first aspect of the invention also apply as relevant to the second and all subsequent aspects of the invention, unless otherwise stated.

The listing or discussion of an apparently prior-published document in this specification should not necessarily be taken as an acknowledgement that the document is part of the state of the art or is common general knowledge. Any published document referred to herein is hereby incorporated by reference in its entirety. The invention is now described in more detail by reference to the following, non-limiting, Figures and Examples.

Figure legends

Figure 1

The effects of octanol on mouse sciatic nerve isolated in the three-chamber recording bath.

(A) Recordings of the evoked CAP from a control nerve immersed in normal saline. The time point of each CAP recording is indicated by (t). (B-D) show representative recordings from nerves immersed in octanol 200 (B), 400 (C) and 600 nM (D). (E) Mean time- response curves (average of n=4-6 experiments) showing the percentage decrease in the amplitude of the evoked CAP as a function of incubation time in octanol 200, 400, 600 and 2000 nM, as indicated. The amplitude of the evoked CAP at time 0, before the application of octanol, was considered to be 100%. The curve marked with the open triangles represents the mean time-response curve for the duration (repolarising time (RT) ) of the evoked CAP obtained from nerves exposed to octanol 400 nM. The mean time-response curve for nerves exposed to normal saline is nearly 100% for 20 h (not shown for clarity). The bars represent ±SEM.

Figure 2

The neurotoxic effect of oxaliplatin and neuroprotective effect of octanol on the isolated mouse sciatic nerve: (A) Representative recordings of the evoked CAP from a nerve incubated in 25 μΜ oxaliplatin. The time point of each CAP recording is indicated by (t).

(B) as in (A) but the nerve was incubated simultaneously in 400 nM octanol and 25 μΜ oxaliplatin. (C) as in (A) but the nerve was incubated first in 400 nM octanol (first arrow) and then 25 μΜ oxaliplatin was added (second arrow) in the perfusion chamber. (D) as in (A) but the nerve was first incubated in 25 μΜ oxaliplatin (first arrow) and then 400 nM octanol was added (second arrow) in the perfusion chamber. (E and H (recalculated)) The mean time-response curves represent the percentage decrease in amplitude and duration RT (marked on the curves) of the evoked CAP incubated in 25 μΜ oxaliplatin, as a fraction of incubation time (n=6). The amplitude of the evoked CAP at time 0, before the application of oxaliplatin, was considered to be 100%. (F) as in (E) but the nerves were incubated in a mixture of 400 nM octanol and 25 μΜ oxaliplatin (n=5). (G) as in (E), but the nerves were either first incubated in 400 nM octanol for one hour and then 25 μΜ oxaliplatin was added (n=4; octanol before), or oxaliplatin was applied first and then octanol added (n=3; octanol after, only the plot of the initial CAP amplitude is shown). The bars represent ±SEM.

Figure 3

Concentration-dependent neuroprotection of octanol against oxaliplatin-induced neurotoxicity in the mouse sciatic nerve: (A) Recordings of the evoked CAP from a nerve immersed simultaneously in 200 nM octanol and 25 μΜ oxaliplatin. The time of exposure of the nerve to oxaliplatin for each CAP is indicated (t in min). (B) as in (A) but the nerve was incubated in the mixture of 600 nM octanol and 25 μΜ oxaliplatin. (C) Mean time- response curves (duration RT and amplitude as indicated) obtained from 200 nM octanol/25 μΜ oxaliplatin co-exposure (n=4) as in A. (D) Mean time-response curves obtained from 600 nM octanol/25 μΜ oxaliplatin co-exposure as in B (n= 3). The amplitude and duration RT of the evoked CAP at time 0, before the application of the mixture of octanol and oxaliplatin, was considered to be 100%. The bars represent ± SEM.

Figure 4

The effects of octanol, oxaliplatin and their mixture on the evoked CAP recorded from the isolated sciatic nerve of the Cx32 knockout (KO) mouse. Recordings of the CAP from Cx32 KO nerves incubated in 25 μΜ oxaNplatin (A), in 400 nM octanol (B), or in a mixture of 400 nM octanol and 25 μΜ oxaliplatin. The time of recording of each CAP is indicated (t in min).

Λ

D-F: Diagrams showing the mean time-response curves of the CAP amplitude and duration RT recorded from Cx32 KO nerves during the exposure to 25 μΜ oxaliplatin (n=9) (D), during exposure to 400 nM of octanol (n=4) (E), or during exposure to the mixture of 400 nM of octanol and 25 μΜ oxaliplatin (n=6) (F). The CAP amplitude and duration RT at time 0 before drug exposure was considered to be 100%. The bars represent ± SEM.

Figure 5

Juxtaparanodal abnormalities in myelinated fibers after prolonged oxaliplatin exposure ex vivo. These are images of sciatic nerve teased fibers obtained from control sciatic nerves exposed to normal saline (A, D, G, J, ) and nerves exposed to 25 μΜ oxaliplatin alone (B, E, H, K, N) or combined with 400 nM octanol (C, F, I, L, O). All nerves were placed in the recording bath under continuous 1 Hz stimulation for over 24 h. Both merged images and insets with separate channels underneath are shown. Fibers were immunostained with axonal marker SMI 31 (green) and myelin basic protein (red) in A-C, with Cx29 (red) combined with juxtaparanodal Kv1.1 (D-F) or with paranodal Caspr (G-l) (green), with nodal Nav1.6 (red) combined with Caspr (green) (J-L) as well as for Na+/K+ ATPase a1 subunit along the axon (red open arrowheads) and a3 subunit at the incisures (green open arrowheads) (Wl-O), as indicated. In control fibers nodes are surrounded by paranodal myelin forming loops around the axon, whereas in fibers exposed to oxaliplatin there is a widening of SMI31 immunoreactivity at paranodal and juxtaparanodal areas (arrows) and the surrounding myelin sheath is thinner and appears retracted away from nodes. This abnormality is not seen in fibers treated with oxaliplatin plus octanol. Double labeling with Cx29 and Kv1.1 or Caspr and Nav1.6 with Caspr shows the widening of juxtaparanodes in fibers exposed to oxaliplatin (E), although the organization of axonal domains and localization of Nav1.6 and Kv1.1 channels and ATPase subunits is not altered. Scale bar: 10 μηη.

Figure 6

Ultrastructural changes in sciatic nerve myelinated fibers after prolonged exposure to oxaliplatin and prevention by octanol. A-C: Images of longitudinal semithin sections from ex vivo stimulated sciatic nerves treated with oxaliplatin alone (B) or in combination with octanol (C) or control treated with saline (A). Nodal areas are indicated (arrows), demonstrating in the control nerve the normal architecture of the perinodal areas (open arrowheads), with myelin sheath forming paranodal loops. In contrast, the oxaliplatin treated nerve shows widening of the paranodal and juxtaparanodal area of the axon, causing an obliteration of the paranodal myelin loops, thinning and retraction of the myelin away from the node. Images of ultrathin sections (D-L) confirm the widening of peranodal areas showing a homogeneous axonal area with reduced density consistent with edema and lack of any inclusions in oxaliplatin treated fibers (E, H-L), while nodal structures are preserved (E, K). This axonal swelling is associated with thinning and flattening of perinodal myelin, obliteration of septate junctions and retraction of myelin from paranodal areas (arrowheads). Normal axonal domains, septate junctions, paranodal myelin loops (stars) and normal juxtaparanodal myelin thickness can be seen in the control fibers (D) and in oxaliplatin plus octanol treated fibers (F). Similar axonal swelling and thinning of the myelin sheath in oxaliplatin treated nerves can be seen in axonal areas surrounded by Schmidt-Lantermann incisures (H), as opposed to the normal appearance of incisures in control nerves (G). Scale bars: in A-C 10 μπι; in D-L: 2μηη. Figure 7

Oxaliplatin accelerates Cx31.3 hemichanriel opening and this effect is blocked by octanol in a dose-dependent manner. Expression of Cx31.3 in clonal Hela cells on the cell surface was confirmed by immunostaining with specific anti-Cx31.3 antibody (A), in contrast to control HeLa cells showing no specific Cx31.3 immunoreactivity (B). Representative images of Cx31.3-expressing cells from different neurobiotin incubation times as indicated without (D) or with oxaliplatin preincubation (E) shows increased neurobiotin signal at all time points in oxaliplatin treated cells. Cx31.3 cells without neurobiotin incubation show no specific signal (C). Quantification from 3 independent experiments (G) shows significantly higher neurobiotin uptake at all time points in oxaliplatin treated cells. Accelerated uptake in the presence of oxaliplatin is most pronounced at earlier intervals and reaches a plateau within 20 min, while in the absence of oxaliplatin the plateau is reached after 60 min. F: Exposure of Cx31.3 cells to combination of oxaliplatin and increasing concentrations of octanol for 20 min shows that the effect of oxaliplatin is blocked by octanol in a dose- depended manner. Quantification of neurobiotin signal (H) confirms the significant antagonism of the oxaliplatin effect by increasing octanol concentrations. *:P<0.01 ; **:P<0.01 ; *:P<0.001.

Figure 8: The effects of octanol, oxaliplatin and their mixture on the evoked CAP recorded from the isolated sciatic nerve of the Cx29 KO mouse. Recordings of the CAP from Cx29 KO nerves incubated in 25 μΜ oxaliplatin (A) and in a mixture of 400 nM octanol and 25 μΜ oxaliplatin (B). The time of recording of each CAP is indicated (t in min). C-D: Diagrams showing the mean time-response curves of the CAP amplitude and RT recorded from Cx32 KO nerves during the exposure to 25 μΜ oxaliplatin (n=3) (C), or during exposure to the mixture of 400 nM of octanol and 25 μΜ oxaliplatin (n=3) (D). The CAP amplitude and RT at time 0 before drug exposure was considered to be 100%. The bars represent + SEM.

Figure 9: A-C: Concentration-dependent effects of carbenoxolone (CBX), 18-beta- glycyrrhetinic acid (GRA) and octanoic acid (OA) on the mouse sciatic nerve. CBX is toxic at 50 μΜ rapidly reducing CAP amplitude, while it is only minimally toxic up to 25 μΜ (A). GRA also shows toxicity at 50 μΜ and to a lesser degree at 25 μΜ, that develops more gradually (B). OA starts having mild toxic effects at 300 μΜ (C). When co-administered with 25 μΜ oxaliplatin, carbenoxolone fails to show any protective effect at a non-toxic concentration (D), while GRA at 25 μΜ shows significant neuroprotection ameliorating RT prolongation by 78.8% (E), and OA at 300 μΜ shows a modest protection of 37.6% (F). Figure 10: Oxaliplatin accelerates dye coupling between cells expressing human Cx32. Cx32 expression and formation of GJ-like plaques (open arrowheads in A) was confirmed by immunostaining with a Cx32 antibody (green), while Cx32 was absent from control HeLa cells (B). Scrape loading with neurobiotin (red) resulted after 15 min in dye transfer between cells not directly on the scrape line (C), and this dye transfer was accelerated with more cells coupled when oxaliplatin 25 μΜ is added (D). DAPI staining of cell nuclei is shown under the representative images of scrape loaded cells with neurobiotin signal in C-D. E: Quantification from at least 10 different fields per condition and time point confirms that oxaliplatin significantly accelerates neurobiotin dye coupling at 15 min (p=0.035), while at 30 min baseline dye coupling reaches the same level as in the presence of oxaliplatin (P>0.05). Addition of octanol 400 nM largely blocks dye coupling at 30 min (P<0.001 ), while combined administration of oxaliplatin and octanol inhibits the effect of oxaliplatin at 30 min (P<0.001 ).

References cited

1. Andre T, Boni C, Mounedji-Boudiaf L, et al. Oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment for colon cancer. N Engl J Med 2004;350:2343-51.

2. Kweekel DM, Gelderblom H, Guchelaar HJ. Pharmacology of oxaliplatin and the use of pharmacogenomics to individualize therapy. Cancer Treat Rev 2005;31 :90-105.

3. Argyriou AA, Cavaletti G, Briani C, et al. Clinical pattern and associations of oxaliplatin acute neurotoxicity: a prospective study in 170 patients with colorectal cancer. Cancer 2013;1 19:438-44.

4. Benoit E, Brienza S, Dubois JM. Oxaliplatin, an anticancer agent that affects both Na+ and K+ channels in frog peripheral myelinated axons. Gen Physiol Biophys

2006;25:263-76.

5. Dimitrov AG, Dimitrova NA. A possible link of oxaliplatin-induced neuropathy with potassium channel deficit. Muscle Nerve 2012;45:403-11.

6. Kagiava A, Tsingotjidou A, Emmanouilides C, et al. The effects of oxaliplatin, an anticancer drug, on potassium channels of the peripheral myelinated nerve fibres of the adult rat. Neurotoxicology 2008;29: 1 100-6.

7. Lang PM, Fleckenstein J, Passmore GM, et al. Retigabine reduces the excitability of unmyelinated peripheral human axons. Neuropharmacology 2008;54:1271-8.

8. Nodera H, Spieker A, Sung M, et al. Neuroprotective effects of Kv7 channel agonist, retigabine, for cisplatin-induced peripheral neuropathy. Neurosci Lett

2011 ;505:223-7. 9. R, Carr RW, Fleckenstein J, et al. Enhancement of axonal potassium conductance reduces nerve hyperexcitability in an in vitro model of oxaliplatin-induced acute neuropathy. Neurotoxicology 2010;31 :694-700.

10. Sittl R, Carr RW, Schwarz JR, et al. The Kv7 potassium channel activator flupirtine affects clinical excitability parameters of myelinated axons in isolated rat sural nerve. J

Peripher Nerv Syst 2010;15:63-72.

11. Hart IK, Maddison P, Newsom-Davis J, et al. Phenotypic variants of autoimmune peripheral nerve hyperexcitability. Brain 2002;125:1887-95.

12. Kleopa KA, Elman LB, Lang B, et al. Neuromyotonia and limbic encephalitis sera target mature Shaker-type K+ channels: subunit specificity correlates with clinical manifestations. Brain 2006;129:1570-84.

13. Van Poucke M, Vanhaesebrouck AE, Peelman LJ, et al. Experimental validation of in silico predicted KCNA1 , KCNA2, KCNA6 and KCNQ2 genes for association studies of peripheral nerve hyperexcitability syndrome in Jack Russell Terriers. Neuromuscul Disord 2012;22:558-65.

14. Adelsberger H, Quasthoff S, Grosskreutz J, et al. The chemotherapeutic oxaliplatin alters voltage-gated Na(+) channel kinetics on rat sensory neurons. Eur J Pharmacol 2000;406:25-32.

15. Webster RG, Brain KL, Wilson RH, et al. Oxaliplatin induces hyperexcitability at motor and autonomic neuromuscular junctions through effects on voltage-gated sodium channels. Br J Pharmacol 2005;146:1027-39.

16. Argyriou AA, Cavaletti G, Antonacopoulou A, et al. Voltage-gated sodium channel polymorphisms play a pivotal role in the development of oxaliplatin-induced peripheral neurotoxicity: results from a prospective multicenter study. Cancer 2013;119:3570-7. 17. Krishnan AV, Goldstein D, Friedlander M, et al. Oxaliplatin and axonal Na+ channel function in vivo. Clin Cancer Res 2006;12:4481-4.

18. Park SB, Lin CS, Krishnan AV, et al. Dose effects of oxaliplatin on persistent and transient Na+ conductances and the development of neurotoxicity. PLoS One 2011 ;6:e18469.

19. Grolleau F, Gamelin L, Boisdron-Celle M, et al. A possible explanation for a neurotoxic effect of the anticancer agent oxaliplatin on neuronal voltage-gated sodium channels. J Neurophysiol 2001 ;85:2293-7.

20. Hochster HS, Grothey A, Childs BH. Use of calcium and magnesium salts to reduce oxaliplatin-related neurotoxicity. J Clin Oncol 2007;25:4028-9.

21. Sakurai M, Egashira N, Kawashiri T, et al. Oxaliplatin-induced neuropathy in the rat: involvement of oxalate in cold hyperalgesia but not mechanical allodynia. Pain 2009;147:165-74. 22. Tatsushima Y, Egashira N, Narishige Y, et al. Calcium channel blockers reduce oxaliplatin-induced acute neuropathy: a retrospective study of 69 male patients receiving modified FOLFOX6 therapy. Biomed Pharmacother 2013;67:39-42.

23. von Delius S, Eckel F, Wagenpfeil S, et al. Carbamazepine for prevention of oxaliplatin-related neurotoxicity in patients with advanced colorectal cancer: final results of a randomised, controlled, multicenter phase II study. Invest New Drugs 2007;25:173-80.

24. Wilson RH, Lehky T, Thomas RR, et al. Acute oxaliplatin-induced peripheral nerve hyperexcitability. J Clin Oncol 2002;20:1767-74.

25. Broomand A, Jerremalm E, Yachnin J, et al. Oxaliplatin neurotoxicity-no general ion channel surface-charge effect. J Negat Results Biomed 2009;8:2.

26. Ferrier J, Pereira V, Busserolles J, et al. Emerging trends in understanding chemotherapy-induced peripheral neuropathy. Curr Pain Headache Rep 2013; 17:364.

27. Loprinzi CL, Qin R, Dakhil SR, et al. Phase III randomized, placebo-controlled, double-blind study of intravenous calcium and magnesium to prevent oxaliplatin-induced sensory neurotoxicity (N08CB/Alliance). J Clin Oncol 2014;32:997-1005.

28. Kagiava A, Kosmidis EK, Theophilidis G. Oxaliplatin-induced hyperexcitation of rat sciatic nerve fibers: an intra-axonal study. Anticancer Agents Med Chem 2013;13:373-9.

29. Kocsis JD, Eng DL, Gordon TR, et al. Functional differences between 4- aminopyridine and tetraethylammonium-sensitive potassium channels in myelinated axons. Neurosci Lett 1987;75:193-8.

30. Altevogt BM, Kleopa KA, Postma FR, et al. Connexin29 is uniquely distributed within myelinating glial cells of the central and peripheral nervous systems. J Neurosci 2002;22:6458-70.

31. Sargiannidou I, Ahn M, Enriquez AD, et al. Human oligodendrocytes express Cx31.3: function and interactions with Cx32 mutants. Neurobiol Dis 2008;30:221-33.

32. Kume-Kick J, Mazel T, Vorisek I, et al. Independence of extracellular tortuosity and volume fraction during osmotic challenge in rat neocortex. J Physiol 2002;542:515-27.

33. Chiu SY. Functions and distribution of voltage-gated sodium and potassium channels in mammalian Schwann cells. Glia 1991 ;4:541-58.

34. Lev-Ram V, Makings LR, Keitz PF, et al. Long-term depression in cerebellar Purkinje neurons results from coincidence of nitric oxide and depolarization-induced Ca2+ transients. Neuron 1995;15:407-15.

35. Yoon SY, Robinson CR, Zhang H, et al. Spinal astrocyte gap junctions contribute to oxaliplatin-induced mechanical hypersensitivity. J Pain 2013;14:205-14.

36. Di Cesare Mannelli LP, A; Matera, C; Zanardelli, M; Mello, T; De Amici, M; Dallanoce, C; Ghelardini C. Involvement of alpha7 nAChR subtype in rat oxaliplatin- induced neuropathy: effects of selective activation. Neuropharmacology 2014;79:37- 48.:10.1016/j.neuropharm.2013.10.034. Epub Nov 10.

37. Warwick RA, Hanani M. The contribution of satellite glial cells to chemotherapy- induced neuropathic pain. Eur J Pain 2013;17:571-80.

38. He B, Tong X, Wang L, et al. Tramadol and flurbiprofen depress the cytotoxicity of cisplatin via their effects on gap junctions. Clin Cancer Res 2009;15:5803-10.

39. Kagiava A, Aligizaki K, Katikou P, et al. Assessing the neurotoxic effects of palytoxin and ouabain, both Na(+)/K(+)-ATPase inhibitors, on the myelinated sciatic nerve fibres of the mouse: an ex vivo electrophysiological study. Toxicon 2012;59:416-26. 40. Kern W, Braess J, Bottger B, et al. Oxaliplatin pharmacokinetics during a four-hour infusion. Clin Cancer Res 1999;5:761-5.

41. Haubenberger D, Nahab FB, Voller B, et al. Treatment of Essential Tremor with Long-Chain Alcohols: Still Experimental or Ready for Prime Time? Tremor Other Hyperkinet Mov 2014;doi: 10.7916/D8RX991 R.

42. Ushio S, Egashira N, Sada H, et al. Goshajinkigan reduces oxaliplatin-induced peripheral neuropathy without affecting anti-tumour efficacy in rodents. Eur J Cancer 2012;48:1407-13.

43. Aoki M, Mori A, Nakahara T, et al. Effect of synthetic eel calcitonin, elcatonin, on cold and mechanical allodynia induced by oxaliplatin and paclitaxel in rats. Eur J Pharmacol 2012;696:62-9.

44. Kawashiri T, Egashira N, Itoh Y, et al. Neurotropin reverses paclitaxel-induced neuropathy without affecting anti-tumour efficacy. Eur J Cancer 2009;45:154-63.

45. Di Cesare Mannelli LZ, M; Failli, P; Ghelardini C. Oxaliplatin-induced neuropathy: oxidative stress as pathological mechanism. Protective effect of silibinin. J Pain 2012;13:276-84. doi: 10.1016/j.jpain.2011.11.009. Epub 2 Feb 9.

46. Boyette-Davis J, Dougherty PM. Protection against oxaliplatin-induced mechanical hyperalgesia and intraepidermal nerve fiber loss by minocycline. Exp Neurol 2011 ;229:353-7.

47. Chen XF, Wang R, Yin YM, et al. The effect of monosialotetrahexosylganglioside (GM1) in prevention of oxaliplatin induced neurotoxicity: a retrospective study. Biomed

Pharmacother 2012;66:279-84.

48. Bushara KO, Goldstein SR, Grimes GJ, Jr., Burstein AH, Hallett M. Pilot trial of 1- octanol in essential tremor. Neurology. 2004;62(1 ): 122-4. Epub 2004/01/14

49. Nahab et al (201 1) Neurotherapeutics. 2011 Oct; 8(4): 753-62. doi: 10.1007/s13311- 011-0045-1

Examples Example 1

Octanol prevents oxaliplatin-induced neuropathy in mouse peripheral nerves The major cause for oxaliplatin peripheral neuropathy is the hyperexcitability in the peripheral nerve fibres. This effect has been investigated with various methodologies but there was not a conclusive solution to the problem, as it is mentioned above. In order to study this oxaliplatin-induced neuropathy we performed ex vivo electrophysiological and histological studies using the isolated sciatic nerve of mouse placed in the saline solution and mounted on a recording bath (7). As an indication the vitality of the nerve fibres, under such recording conditions, the electrical properties of the nerve fibres the nerve evoked compound action potential (CAP) was monitored for over 24 h; the CAP is the summation of the potentials of all excitable nerve fibres in the sciatic nerve of the mouse. It is well-known that the voltage-gated potassium channels (VGKCs) and more specific Kv1.1 and Kv1.2 responsible for the repolarising phase of the action potential in myelinated nerve fibres, are located very close to Cx29 hemichannels, responsible for releasing K+ from the juxtaparanodal region (28, 30). Combining the two structural and functional characteristics we came to the conclusion of a possible direct functional interplay between Cx29 and Kv1.1 and Kv1.2 during the oxaliplatin-effect (the broadening of the repolarising phase of the CAP). We assumed that oxaliplatin has an effect (forced opening in this case) on the Cx29 hemichannels creating stagnant K+ near the Kv channels and causing their malfunction and prolongation of the repolarising phase. Thus, in an attempt to reverse the effect of oxaliplatin, we used a common GJ blocker (closer), octanol. There are a number of GJ blockers like carbenoxolone and 18-a- glycyrrhetinic acid, but octanol was chosen because it is known to be safe and has been already used in clinical trials (90). The effects of octanol on the sciatic nerve CAP

The evoked CAPs generated by 1 Hz supramaximal stimulation of the mouse sciatic nerve representing the summation of action potentials of all activated sciatic nerve fibres (100% of myelinated nerve fibres) showed constant amplitudes over the 20-24 h recording period (Figure 1A), followed by a gradual decrease due to nerve fiber inactivation, as in our previous studies (39). Thus, the vitality of the sciatic nerve fibres is not affected by the recording conditions in the first 20 h of the experiment. Although nerve vitality lasts longer than 20 h (Figure 1A), the incubation period with the compounds under investigation lasted for maximum 1200 min (20h). This was the period where the probability to record CAPs of a constant amplitude and duration was 100%. Beyond that time the CAP became unstable in some experiments, therefore we assessed CAPs recorded only up to 20 h exposure times to ensure reliability of the results.

In order to investigate the role of myelin GJs and hemichannels in the generation of action potentials and in extend of the CAP, we first examined the effects of the GJ blocker octanol. Since previous studies suggested that octanol may inhibit VGNaCs expressed in Xenopus oocytes with an IC₅o of 72.1 ±4.5 μΜ within a range of 0.003 (min) - 3 mM (max) (40), we used octanol concentrations of 200, 400, 600 and 2000 nM, far below the minimum concentration required to inhibit VGNaCs. When the nerve was exposed to 200 nM octanol there was a minor, 10-20% decrease in the evoked CAP at the beginning of the 20 h exposure (representative recordings shown in Figure 1 B and mean time-response curve from at least 4 different experiments in Figure 1 E). At a concentration of 400 nM octanol caused a CAP decrease to 54% of its initial value almost immediately after application (Figure 1 C, records t=10 to 60 min) followed by a gradual recovery (records t=300 to 1200 min). The mean time-response curve (400 nM in Figure 1 E) showed a CAP decrease to 46.5±10.69% of its original value (n=6) within 1 h, while by 6 h of exposure the CAP recovered to 65.1 ±10.97% and remained at this level despite continuous exposure to octanol. Application of 600 nM octanol caused a sharp decrease of the CAP amplitude to 42.3±10.28 % within 1 h, gradually complete elimination of the CAP within 6-7 h, and no recovery even after 20 h (600 nM in Figure 1 D-E). Finally, application of 2000 nM octanol eliminated completely the CAP amplitude within 1 h, without any recovery (2000 nM in Figure 1 E). Thus, the drastic, reversible inhibition-recovery effect of octanol on the CAP was concentration-dependent with a minimum inhibitory effect at 200 nM and a maximum at 2000 nM. Octanol caused a reversible decrease in CAP amplitude at 200 and 400 nM, whereas a permanent elimination with loss of active nerve fibres occurred at 600 to 2000 nM. This reversible action of octanol is called here the "octanol-effect". In contrast to the inhibitory effect on CAP amplitude, exposure to octanol had no effect on the duration of the CAP. For example, at 400 nM the duration of the CAP remained constant (duration curve in Figure 1 E).

The effects of oxaliplatin on the sciatic nerve CAP

Incubation of mouse sciatic nerve with oxaliplatin caused a drastic, time-dependent, increase in the duration of the repolarising phase/repolarising time (RT) of the CAP (Figure 2A), similar to our previous findings in the rat sciatic nerve (6, 7). The effect started with a gradual decrease of the negative part of the CAP repolarising phase (Figure 2A, records t=0 to t=40 min). The increase of the duration, measured from the peak of the CAP to the point where the repolarising phase meets the baseline, started from 2.39±0.08 ms under control conditions and reached 13.9±0.16 ms after 300 min of exposure (P<0.01 ; n=6 experiments) (or 26.2±0.16 ms after 300 min of exposure (PO.01 ; n=8 experiments)) remaining at this level for 1200 min. Furthermore, a second peak occurred on the main CAP almost at the beginning of the exposure (Figure 2A, t=60 min) and persisted to the end of continuous recordings (t= 200 min). The mean time-response curve with a per-minute sampling rate showed that the duration of the repolarising phase/RT increased up to 586.9±0.92% of the control values (or 1173.8±0.92% (n=8) of the control values and remained at this level throughout the 1200 min exposure to oxaliplatin (Figure 2E, RT). This broadening of the CAP recorded extracellularly in the presence of oxaliplatin results from broadening in the action potential of individual sciatic nerve fibres, as shown by previous intra-axonal recordings (7). This abrupt increase in the CAP repolarization phase in the presence of oxaliplatin is called here the oxaliplatin-effect. In contrast to duration/RT, we found no significant change in the CAP amplitude at the initial (t=1 h) and at the final stages (t=20 h) of the recording under 25 μΜ oxaliplatin exposure (P>0.05, n=6 or 8) (Figure 2E). Importantly, the duration of the CAP depolarising phase reflected by the rise time of the CAP, during which the VGNaCs are activated, remained unchanged. Co-application of octanol and oxaliplatin abolishes the effects of oxaliplatin

In order to test the possible protective role of the GJ inhibitor octanol against OIN we exposed the sciatic nerve to 25 μΜ oxaliplatin and 400 nM octanol applied in the perfusion chamber simultaneously. The 400 nM octanol concentration was chosen because it had an intermediate effect when applied alone without a drastic neurotoxic effect as seen with higher concentrations (Figure 1). This co-application has an effect on both amplitude and duration of the CAP repolarizing phase. The CAP amplitude decreased within the first hour of exposure with gradual, almost full recovery after 3 to 15 h (Figure 2B), resembling the octanol-effect at 400 nM (Figure 1 B and E). Interestingly, there was a much smaller increase in the CAP duration, in contrast to the marked increase observed when 25 μΜ oxaliplatin was applied alone (Figure 2A). During co-application, the oxaliplatin-effect started with a loss of the negative part of CAP repolarization in the first hour (Figure 2B, records t=0 to t=60 min), but at this point the oxaliplatin-effect stopped, and there was no further increase in the duration of the CAP, which remained stable ranging from 2.33 to 3.92 ms (Figure 2B, record t=120 to t=1200 min). These records clearly indicate that octanol abolishes the effect of oxaliplatin. The RT mean time-response curve (Figure 2F), obtained from n=5 recordings indicates that the octanol-effect dominated over the oxaliplatin-effect. During co-application the effect of octanol developed as a fast decrease of the CAP, to 60.6±8.08% of the control values within 1 h, followed by a relatively slow recovery within 2-12 h. On the other hand, the oxaliplatin-effect, the broadening of the CAP, was almost eliminated and restricted drastically to an increase in the duration of the CAP repolarizing phase up to 146.0±13.29% of the initial value, in contrast to the >550% or >1 173.75% increase in the presence of oxaliplatin alone (P<0.05) (Figure 2E). This is a clear indication of neuroprotection of octanol against oxaliplatin-induced neurotoxicity. Interestingly, oxaliplatin also provided mild "neuroprotection" against the octanol-effect. While the maximum recovery of the CAP amplitude when a nerve was exposed to 400 nM octanol was at the level of 64.0±7.0% after 20h (Figure 1 E), co-administration of octanol (400 nM) and oxaliplatin resulted in higher recovery levels up to 98.2±12.04% (P<0.01) (Figure 2F) .

In contrast to the co-application experiments, when octanol was added 40 min prior to oxaliplatin incubation, it prevented only partially the broadening of the CAP repolarising phase by oxaliplatin, while the octanol-effect, the decrease and recovery of the CAP, persisted (Figure 2C). The mean time-response curve under these conditions (n=4 experiments) showed an increase of the CAP duration by 189.2±4.28% in the first 12 h of exposure and then further to 230.1±13.65% (Figure 2G), significantly higher compared to the corresponding experiment of co-application (P<0.05). Nevertheless, even with pre- application octanol offered an 81.4% protection against OIN while the octanol-effect (decrease and recovery of the CAP amplitude) followed the same pattern as in the previous experiments (Figure 2G). In contrast, the octanol-effect, the decrease and recovery of the CAP amplitude followed the same pattern as in the previous experiments (Figure 2G). Finally, when oxaliplatin was added first followed after 40 min by octanol, the CAP was eliminated completely within less than 1 h, without any recovery after 20 h, indicating that this particular sequence was extremely toxic for the sciatic nerve fibres (Figure 2D, G). Thus, only when octanol (400 nM) was applied simultaneously or prior to oxaliplatin (25 μΜ) was there a significant protection against the oxaliplatin-effect on sciatic nerve fibre function.

Co-application of carbenoxolone and oxaliplatin has no neuroprotective effect

Based on the results of octanol we examined whether another GJ blocker, carbenoxolone, could have a neuroprotective effect against OIN, using the same ex vivo sciatic nerve preparation. In this case 50 μΜ of carboxenolone was found to be neurotoxic eliminating the CAP in less than an hour (n=4). At 25 μΜ carboxenolone had a milder effect causing a 20% decrease of the CAP within 20 h (n=4) and therefore it was tested at this concentration in simultaneous administration with oxaliplatin 25 μΜ. In all four experiments under these conditions there was no neuroprotection against the oxaliplatin effect (data not shown).

Concentration-dependent neuroprotection by octanol against the oxaliplatin-effect To investigate whether the neuroprotection was concentration-dependent 200 and 600 nM octanol were tested against the oxaliplatin-effect. During co-application of 200 nM octanol plus 25 μΜ oxaliplatin there was a neuroprotection with increasing duration of the CAP followed by the weak development of a second peak of the waveform (Figure 3A, record t=120 to 1200 min), which never occurred in the previous treatment (400 nM + 25 μΜ). The mean time-response curves showed that the increase in the duration of the CAP repolarising phase RT reached a maximum of 172.3±10.07 %, significantly higher than with the 400nM + 25 μΜ combination (P<0.05) (Figure 3C). Nevertheless, even at this low concentration (200 nM) octanol offers a significant (85.3%) protection against OIN, very close to the one at 400 nM (88.6%), while the octanol-effect was minor with an amplitude decrease of 16% and a relatively slow recovery (Figure 3C).

The neurotoxic effect of 600 nM octanol, the elimination of the CAP (Figure 1 ; Figure 1 E) was abolished in the presence of oxaliplatin (Figure 3B). This mixture minimized the amplitude of the CAP to 21.0±10.03% within 1 h (n=3; Figures 3B and D), as 600 nM octanol did when applied alone (Figurel E), but after that instead of eliminating the amplitude of the CAP, there was a 86.8±7.80% recovery within 3-5 h (Figure 3B and D). Furthermore, during this unexpected inhibition-recovery of the CAP with 600 nM octanol, octanol protected against the oxaliplatin effect, resulting in an increase up to 144.5±.2.77% (n=3) in the repolarising phase of the CAP, which was far smaller compared to changes caused by oxaliplatin alone (PO.05) (compare Figure 3B and D and Figure 2A, E). Thus, octanol at a higher concentration of 600 nM combined with oxaliplatin (25 μΜ) resulted in a significant (87.7%) neuroprotection against the oxaliplatin-effect and vice versa. Furthermore, neuroprotection offered by octanol against oxaliplatin-induced nerve fiber dysfunction is concentration-dependent, offering overall a wider therapeutic window with a concentration-dependent effect on nerve fibre dysfunction.

Example 2

Oxaliplatin-effect and neuroprotection by octanol persist (but weaker) in Cx32 knockout nerves

Since octanol is considered to be a non-specific GJ blocker, there are two known potential targets to mediate its protective effects against oxaliplatin in myelinated nerve fibres of the mouse sciatic nerve: Connexin32 (Cx32), which forms GJs through the layers of non- compact myelin at paranodes and Schmidt-Lantermann incisures (41), and Cx29, which actually forms hemichannels rather than full GJs at the inner most myelin membrane apposing axonal juxtaparanodes and juxtaincisures (28). In order to distinguish the relevant connexin mediating the protective effects of octanol, we performed the same experiments as above in ex vivo sciatic nerves of Cx32 as well as Cx29 knockout (KO) mice. In both Cx32 KO and Cx29 KO nerves examined under control conditions (n=3 nerves per genotype) using the same recording bath, the CAP generated by 1 Hz stimulation remained constant in amplitude and duration for over 20 h. The recordings were identical to those obtained from WT nerves (P>0.05), indicating that the vitality and physiological properties of sciatic nerves from either connexion KO animals were not affected by the specific stimulating and recording conditions (data not shown).

Application of 25 μΜ oxaliplatin on Cx32 KO nerves resulted in a marked increase in the duration/RT of the CAP repolarising phase (n=9; Figure 4A), while the distinct second peak observed in the recordings from the WT sciatic nerves (Figure 2A) was absent in Cx32 KO nerves examined changing significantly the shape of the CAP waveform. The mean time- response curve from n=9 Cx32 KO nerves (Figure 4D) showed that the CAP/RT broadening reached a maximum of 613.5±49.8% (from 3.1±0.25 to 17.9±1.31 ms, P<0.01 ) within 480 min, but then decreased gradually reaching a value of 476.8±43.56% by the end of the experiment (1200 min); by comparison, the increase in duration in WT nerves was 586.1 ±0.26% at 360 min (no statistical difference between WT and Cx32 KO, P>0.05) and remained constant throughout the 20 h experiment (Figure 2E). In a further experiment, the CAP/RT broadening reached a maximum of or 683.9±36.5% (from 3.1 ±0.25 to 22.7+1.31 ms, P<0.01) within 360 min, but then decreased significantly reaching a value of 430.3±62.08% (P<0.05) near the end of the experiment (960-1080 min); these changes were significantly smaller compared to WT nerves, where RT was prolonged by 1 170.8±1.86% at 360 min (PO.05) and remained constant throughout the 1200 min experiment (Figure 2H). Shorter RT in the Cx32 KO CAP could be due to the absence of the second peak. Furthermore, Cx32 KO nerves exposed to oxaliplatin showed a small decrease in CAP amplitude by 20% (Figure 4A and D), whereas CAP amplitude was never affected during treatment with oxaliplatin in WT nerves (P<0.01 WT vs KO CAP amplitude change). Thus, the oxaliplatin-effect develops also in Cx32 KO nerves, but only partially, showing important similarities, but also some differences compared to nerves from WT animals.

In order to examine whether octanol remains neuroprotective against oxaliplatin also in Cx32 KO nerves, we first clarified the effects of octanol alone. Application of 400 nM octanol caused the same pattern of the inhibitory-recovery effect on CAP amplitude as in WT nerves (Figure 4B). However, the mean time-response curve from Cx32 KO nerves showed a 30% (P<0.05; n=4) increase in CAP duration at the beginning of the exposure which remained at this level throughout the experiment (Figure 4E), in contrast to WT nerves which showed almost constant duration of the repolarising phase when exposed to 400 nM of octanol. Thus, the octanol-effect occurred also in Cx32 KO nerves but in slightly different pattern compared to WT nerves.

Finally, co-application of 25 μΜ oxaliplatin plus 400 nM octanol in Cx32 KO nerves protected against the broadening of the CAP, the oxaliplatin-effect (Figure 4C). The mean time-response curve (Figure 4F, n=6) clearly indicated that the duration of the repolarised phase RT reached 193.5±12.93% of the initial value (P<0.01 comparison to oxaliplatin alone) after treatment for 360 min and then remained at this level for most of the duration of the experiment, before showing a further increase to 237.2±23.0% near the end of the experiment (t=1080 min). However, the octanol neuroprotection against oxaliplatin-effect was significantly weaker in Cx32 KO nerves (71.7%) compared to WT (88.6%) in which RT increased only to 146.0±13.29% (PO.05). Besides a clear neuroprotection of octanol against oxaliplatin in Cx32 KO nerves, there was also an obvious antagonizing action of oxaliplatin against the octanol-effect, as also observed in WT nerves. In Cx32 KO nerves treated with 400 nM octanol the maximum CAP recovery reached 83.7±9.12% after 840 min (Figure 4E), while in nerves treated with octanol combined with oxaliplatin CAP recovery was larger and earlier, 93.5±11.06% in 840 min (n=6; Figure 4F). In conclusion, the results in Cx32 KO nerves indicate that the effects of oxaliplatin as well as the protective action of octanol occurred but were significantly weaker compared to WT, implicating Cx32 GJ channels, at least in part, in OIN. The effects of oxaliplatin as well as the protective action of octanol may be mediated in part though Cx29 hemichannels.

Oxaliplatin-effect and neuroprotection largely persist in Cx29 knockout nerves

Application of 25 μΜ oxaliplatin on Cx29 KO nerves resulted in a marked increase in the RT of CAP (Figure 8A). Also, in all Cx29 KO nerves examined, the distinct second peak appeared as it was observed in the recordings from the WT nerves (compare Figure 2A and 8A). The RT mean time-response curve from n=3 Cx29 KO nerves (Figure 8C) showed that the RT was prolonged to a maximum of 1035.0±44.74% in 400-600 min and then declined to 441.3±62.10% near the end of the experiment (1200 min). This unusually fast increase and unstable reduction of RT, indicated by large variations after 900 min (Figure 8 C) could be due to the gradual decrease of CAP after 600 min of exposure to oxaliplatin (Figure 8A and C). Such drastic changes in these two parameters were not observed in either WT or Cx32 nerves. Thus, the oxaliplatin-effect develops also in Cx29 KO nerves showing important similarities and minor differences compared to WT nerves. Exposure of Cx29 KO nerves to the combination of oxaliplatin (25 μΜ) and octanol (400nM) showed a significant (87.1%) neuroprotection, with a maximum RT increase of only 132.8±4.92% (Figure 8D; n=3). Thus, the degree of neuroprotection offered by octanol against the oxaliplatin effect in Cx29 KO nerves was not different from WT nerves (88.6%; P>0.05), whereas in was clearly stronger than in Cx32 KO nerves (71.7%; P<0.05). Taken together, the findings from Cx32 KO and Cx29 KO nerves suggest that oxaliplatin affects the function of both Cx32 GJ channels and Cx29 hemichannels, but the effect on Cx32 appears to be more prominent compared to Cx29.

Effects of other gap junction blockers against 01 N

Based on the results of octanol we examined whether another three GJ blockers, carbenoxolone, GRA and OA, could have a neuroprotective effect against OIN. We first estimated the concentration with the minimum neurotoxicity on the sciatic nerve. Carbenoxolone at 50 μΜ was found to be extremely neurotoxic eliminating the CAP in 120- 300 min (n=4), while at 10 and 25 μΜ (n=4 each) carbenoxolone had a milder effect causing a 20% decrease at the beginning of exposure and then a recovery within 1200 min (Figure 9A). However, at this non-toxic concentration of 25 μΜ, carbenoxolone failed to provide neuroprotection against the oxaliplatin-effect when tested in simultaneous administration with 25 μΜ oxaliplatin (n=4) (Figure 9D), since RT increased up to 1107.1 ±36.79%. Using the same protocol, 50 μΜ GRA were found to be neurotoxic, decreasing the CAP amplitude to 25% of the control within 720 min (Figure 9B), while 25 μΜ of GRA proved to be less toxic maintaining the CAP amplitude near 50% of its original value (Figure 9 B). Co- application of 25 μΜ GRA and 25 μΜ of oxaliplatin showed a significant neuroprotection since the RT increased only up to 248.5±38.39% (Figure 9E), instead of 1173.8±0.92% when oxaliplatin was applied alone (78.8% neuroprotection, P<0.01). Furthermore, as already shown with octanol (above), we observed a protection of oxaliplatin against the inhibitory effect of GRA on the CAP amplitude. 25 μΜ GRA alone decreased the CAP to 50% (Figure 9B), while in the presence of oxaliplatin only to 75% of baseline (P<0.05). Thus, GRA provides a similar neuroprotection against OIN as seen with octanol, although with a smaller therapeutic window. Finally, three concentrations, 100, 200 and 300 μΜ of OA were tested (n=3 each). The first two had only a minor effect on the vitality of nerve fibres, since the CAP remained at the same level throughout the exposure (1200 min) (Figure 9C), while the concentration of 300 μΜ caused a decrease of the CAP to 65.6±7.17% (Figure 9C). Co-application of the high concentration of 300 μΜ with 25 μΜ oxaliplatin had only a mild protective effect (37.6% protection), since the increase of RT reached 731.6±53.24%, compared to 1 173.8±0.92% when oxaliplatin was applied alone (P<0.05).

Example 3

Morphological changes caused by the prolonged exposure to oxaliplatin

In order to examine whether there is a morphological correlate to the functional changes caused by oxaliplatin in peripheral nerve fibres described above, we also immunostained teased fibers from sciatic nerves following the 20 h ex vivo study. Most fibers were well preserved and were double stained with myelin and axonal markers, as well as markers of axonal domains including Nav1.6/PanNav at nodes, Kv1.1 or Caspr2 at juxtapanodes and Caspr at paranodes, in combination with Cx29. Fibers from sciatic nerves exposed to oxaliplatin for 20 h frequently showed a characteristic enlargement of juxtaparanodal axonal areas, which appeared to cause a thinning of the surrounding myelin and retraction away from paranodal areas (Figure 5B, E). The characteristic myelin loops surrounding the paranodes that are seen in control fibers were thinner or not visible. However, the localization of Cx29, Kv1.1 , Caspr2, Caspr and Nav was not disturbed, indicating that axonal domains were preserved (Figure 5D-L). Juxtaparanodal swelling was not seen in fibers treated with the combination of oxaliplatin and octanol or in untreated fibers that were also stimulated for 20 h with same frequency (Figure 5A, C). These morphological changes were not found in nerves examined after a short exposure to oxaliplatin for 3 hours (data no shown), despite significant broadening in the evoked CAP at this time point (Figure 2A), suggesting that functional changes are initially not accompanied by morphological changes, and that juxtaparanodal swelling occurs only with longer oxaliplatin exposure.

To further clarify the nature of the morphological changes observed by immunostaining, we also examined semithin and ultrathin sections of ex vivo stimulated nerves exposed to each condition and control nerves. In semithin sections of nerves exposed to oxaliplatin, we observed a widening of axonal diameter at juxtaparanodes, and obliteration of surrounding myelin sheath (Figure 6B). These changes were not observed in control nerves or in nerves treated with combination of oxaliplatin and octanol (Figure 6A, C). Further examination of these axonal changes by electron microscopy showed that swollen perinodal areas were edematous, and did not contain any dense or formed material. In addition, the axonal cytoskeletal elements such as microtubules, neurofilaments and associated mitochondria, which are normally dense in perinodal areas, were rarefied. The septate axon-glial paranodal junctions were obliterated and the surrounding myelin sheath appeared retracted and thinner over the edematous paranodal and juxtaparanodal axonal areas. Nodal structures were clearly preserved in most swollen axons (Figure 6E, I, J-L). Similar edematous axonal swellings were observed in axonal areas surrounded by Schmidt-Lantermann incisures (Figure 6H). Thus, prolonged (20-24 h) exposure to oxaliplatin causes profound morphological changes in axons, in addition to the functional changes that occur within the 3 hours of oxaliplatin exposure and are initially not accompanied by any morphological changes.

Example 4

Oxaliplatin causes increased opening of Cx31.3 hemichannels and Cx32 GJ channels in vitro

Our electrophysiological experiments indicated that oxaliplatin effects on the peripheral nerve may be mediated in part through Cx29 hemichannels localized at the inner myelin membrane in close apposition to axonal VGKCs. To further investigate this possibility, we studied the effects of oxaliplatin on clonal HeLa cells expressing the human ortholog of Cx29, Cx31.3, which have been shown to form functional hemichannels on the cell surface (29). Cx31.3 hemichannel function was assessed by the uptake assay. Cx31.3 expressing cells were incubated with or without the addition of oxaliplatin (25 and 50 μΜ concentrations). Neurobiotin uptake increased overtime reaching a plateau after one hour in unexposed Cx31.3 cells (Figure 7D), while in the presence of oxaliplatin, neurobiotin uptake was accelerated, reaching a plateau already after 20 min of exposure, and was significantly increased at all time points examined (P<0.05) (Figure 7E) indicating forced opening of Cx31.3 hemichannels. These results were confirmed by quantitative analysis in at least 3 independent experiments with counts of neurobiotin signal intensity in at least 2000 cells per time point (Figure 7G). This effect of oxaliplatin on neurobiotin uptake was blocked by octanol in a dose-depended manner (P<0.05) (examples in Figure 7F and quantification from 3 independent experiments in H), confirming that it is mediated through opening of Cx31.3 hemichannels. Thus, oxaliplatin appears to open the hemichannels, an effect that can be blocked by octanol. Furthermore, oxaliplatin has similar effects on Cx31.3, the human ortholog of rodent Cx29, assessed in clonal cells, supporting the relevance of this effect for OIN in humans.

Since OIN appears to also involve even more prominently the GJ channels formed by Cx32 in myelinated fibers, we also examined in vitro the effect of oxaliplatin in cultured cells expressing the human Cx32. In order to assess Cx32 GJ channel function, we used the neurobiotin scrape loading assay. The addition of 25 μΜ oxaliplatin resulted in accelerated dye coupling between cells after 15 minutes compared to baseline dye coupling (p<0.05). This effect of oxaliplatin was blocked by co-administration of octanol 400 nM (Figure 10). Thus, oxaliplatin appears to also cause opening of Cx32 GJ channels.

Example 5

Discussion

The clarification of the cellular mechanisms underlying OIN in the peripheral nervous system and identification of a successful neuroprotective strategy could have a major impact on improving patient quality of life and the ability to deliver full doses of oxaliplatin without the peripheral neuropathy as a side effect. Our study implicates both x32 GJ channels and Cx29/Cx31.3 GJ hemichannels in OIN and demonstrates that the coadministration of the GJ blocker octanol can prevent the effects of oxaliplatin on peripheral nerve myelinated fibers. Octanol appears to offer the best therapeutic window compared to three other GJ blockers tested, which showed weaker neuroprotective effects and more toxicity. Furthermore, we provide a model for the origin of hyperexcitability in acute OIN, as well as a structural correlate for the chronic manifestations.

The main advantage of our study is the use of the ex vivo preparation of the isolated mouse sciatic nerve. Under these conditions, we recorded CAPs stable in amplitude and duration from either WT or Cx32 KO animals for at least 20 h. The long vitality of all three types of nerves in saline allowed us to perform long-term (20 h) neuropharmacological experiments using a relatively low concentration of either octanol 200-600nm, or of oxaliplatin 25 μΜ, which may be about 10-fold higher than the concentration used in clinical practice; the plasma oxaliplatin concentration is elevated up to 2.5 μΜ for at least 4 h (42). The concentration of oxaliplatin used in this study may be only slightly above the concentrations used in clinical practice where after a dose of 130 mg/m² infused over 2 h, the plasma platinum mean C_max values were in the range of 2.59-3.22 g/ml corresponding to 6.5- 8.1 1 μΜ (Graham et al., 2000) or 3-6 Mg/ml (8-16 μΜ) (Eckel et al., 2002). Moreover, the same oxaliplatin-effect was demonstrated in our previous ex vivo experiments even with a much lower concentration of 5 μΜ (Kagiava et al., 2008), but with a delay of 15-20 h. Therefore, we used here the concentration of 25 μΜ in order to provide clinical relevance to our ex vivo experiments (Kern et al., 1999), but also to have the oxaliplatin-effect developed within the time frame when the nerve preparation is still in perfect functional condition. In contrast, previous studies using similar nerve exposure models applied oxaliplatin 75 or 250-500 μΜ in order to achieve effects within 10-20 min (11 , 12, 16), an unusually high concentration which may have multiple effects on myelinated nerve fibres that may not be clinically relevant.

The effects of octanol on myelinated nerve fibres sciatic nerve GJs and GJ hemichannels

Octanol is a well-known GJ and hemichannels blocker (50), but in some cases (eg at higher concentrations) it has been reported to also act as a blocker of Nav channels with an IC50 of 75 μΜ (40) or even 455 μΜ (51). Here, octanol was tested against GJ and GJ hemichannels of myelinated nerve fibres at a much lower concentration of 200 and 400 nM causing a concentration-dependent inhibition-recovery effect eg on the CAP amplitude, the octanol-effect, while at higher concentration over 600 nM it had a strong neurotoxic effect, eliminating completely the amplitude of the CAP within 5-7 h, while at 2000 nM in less than 1 h without any recovery. The rapid decrease of the CAP amplitude indicates a fast loss of active nerve fibres. This strong inhibition of the CAP amplitude by octanol at such low concentrations confirms that GJs and GJ hemichannels located on the myelin sheath play a significant role not only in the generation and propagation of the action potential but also in the vitality of the fibres. In order to clarify whether octanol acts on full GJs formed by Cx32 (41) or hemichannels formed by Cx29 (28), the only known connexin channels in peripheral nerves, or both, we further investigated its effects in Cx32 KO mice and Cx29 KO mice. Overall, the response of the Cx32 KO and WT nerves to 400 nM octanol was similar as far as the inhibition-recovery effect indicating that Cx29 plays the main role specifically in the homeostasis of K⁺ in the juxtaparanodal periaxonal space while the role of Cx32 GJs through non-compact myelin is less crucial. However, further analysis and investigation in Cx29 KO mice suggests that both GJ and GJ hemichannels are involved. These effects of octanol indicate that Cx29 and Cx32 play the main role in the homeostasis of K⁺ in the periaxonal space at juxtaparanodes and juxtaincisures. The possible action of octanol against VGNaCs has already been excluded.

Cx29 hemichannels are located in the juxtaparanodal region in close proximity to the VGKCs and they provide a direct communication pathway between the periaxonal juxtaparanodal space and the Schwann cell, suggesting a close functional relationship (Altevogt et al., 2002). On the other hand, Cx32 forms GJs through the layers of non- compact myelin at paranodal loops and Schmidt-Lantermann incisures (Scherer et al., 1995) and plays an even more important role in the homeostasis of myelinated fibers and for preserving homeostasis in the periaxonal space (Anzini et al., 1997, Balice-Gordon et al., 1998, Sargiannidou et al., 2009, Vavlitou et al., 2010, Kleopa, 2011 ). Cx32 is required for normal function in peripheral nerves, as also demonstrated by the progressive neuropathy occurring in Cx32 KO mice (Anzini et al., 1997, Scherer et al., 1998, Sargiannidou et al., 2009, Vavlitou et al., 2010), as well as by the human disorders associated with Cx32 dysfunction (Kleopa, 2011 ). In contrast, Cx29 appears to play a less important role, since Cx29 KO mice show no morphological or basic electrophysiological abnormalities in the PNS or CNS (Eiberger et al., 2006) except for auditory neuropathy in 50% of cases (Tang et al., 2006).

The role of myelin GJ channels and hemichannels in K+ spatial buffering

GJs and GJ hemichannels appear to play a crucial role in extracellular K⁺ buffering at juxtaparanodes and periaxonal space, since blocking by octanol at 600-2000 nM in this study caused cancellation of the action potentials and eventually the loss of CAP, likely resulting from excess K⁺ in the periaxonal space causing the malfunction of VGKCs. Others have similarly shown that the extra K+ can modify membrane excitability (Kume- Kick et al., 2002) and even lead to hyperexcitation of myelinated nerve fibers (Chiu, 1991 , Lev-Ram and Ellisman, 1995). Rapid clearance of K⁺ from the juxtaparanodal periaxonal space is essential because during repetitive action potentials K⁺ released by VGKCs in this very narrow space may reach concentrations of over 60 mM due to a very small distribution volume (Astion et al., 1988).

Based on our results of octanol interaction with GJ and GJ hemichannels and the known location and functional role of VGKCs in myelinated fibers (Konishi, 1990, Chiu, 1991 ), we propose the following siphoning mechanism by which GJ and GJ hemichannels may regulate the outflow of excess K⁺ from the juxtaparanodal region, which is also in keeping with previous studies (Altevogt et al., 2002, Menichella et al., 2006): during an action potential the depolarization caused by the nodal Na⁺ current activates both axonal Kv1.1/Kv1.2 and apposing GJ hemichannels so that they open almost simultaneously, or with a minor delay, the GJ hemichannels preceding the VGKCs. Almost all types of GJ hemichannels are activated (open) upon axon depolarization and are effectively closed by hyperpolarization (Paul, 1991 , Valiunas and Weingart, 2000, Kang et al., 2008, Fasciani et al., 2013) allowing substantial cation fluxes (Trexler et al., 1996). During this synchronised opening Kv1.1/Kv1.2 deliver K⁺ into the narrow juxtaparanodal periaxonal space, Cx29 GJ hemichannels transfer the excess K⁺ into the adaxonal Schwann cell cytoplasm, and from there it is further transported through the radial pathway formed by Cx32 GJs to the abaxonal Schwann cell cytoplasm. This flow results from accumulation of K+ charging positively the juxtaparanodal periaxonal space combined with the negative Schwann cell membrane potential, near -40 mV (Hargittai et al., 1991). This brief influx of K⁺ into the Schwann cell through the open hemichannels and GJ channels following electrical gradients lasts for a few milliseconds and stops at the end of action potential. Cx29 GJ hemichannels are the only known channels apposing juxtaparanodal axonal VGKCs (Altevogt et al., 2002), whereas Cx32 GJs are concentrated in the surrounding non-compact myelin areas (Scherer et al., 1995, Kleopa, 2011). Other channel types such as Kir2.1 and Kir2.3 are located at the microvilli surrounding the nodes of Ranvier (Mi et al., 1996) and may buffer extracellular K⁺ at the node, while Schwann cells may transport excess K⁺ from the cytoplasm to the capillaries through other channels facing the basilar membrane, including Kv1.5 (Horio, 2001).

Tthe vitality of Cx32 KO nerves was not affected by the control recording conditions in saline, although Cx32 GJs are considered vital for the function of the myelinated nerve fibres and for preserving homeostasis in the periaxonal space (52-56). Although no other connexin has been identified so far to participate in the radial diffusion of low molecular weight dyes across the myelin sheath, studies in Cx32 KO nerves showed that this pathway is not interrupted, suggesting that alternative transport mechanisms may compensate at least in part for the loss of Cx32 (57). Under resting or minimal stress conditions caused by 1 Hz low frequency stimulation throughout our experiments, it is possible that nerve fibres are able to functionally recover from disturbances of homeostasis due to luck of Cx32 GJs.

The severe malfunction caused by octanol at 600 and 2000 nM, with a concentration- dependent elimination of the CAP is mainly the result of the closure of GJs and hemichannels since both regulate the concentration of ions and other molecules in the periaxonal microenvironment of myelinated nerve fibres. Given the functional roles of VGKCs in myelinated fibers (27, 30), rapid clearance of K⁺ from the juxtaparanodal periaxonal space is essential because during repetitive action potentials K⁺ released by VGKCs in the very narrow juxtaparanodal periaxonal space may reach concentrations of over 60 mM due to a very small distribution volume (49). Hemichannels are likely key players involved in siphoning of K+ released in the periaxonal space during neuronal activity (28, 58) providing a direct conduit between the extracellular compartment and Schwann cell cytoplasm, allowing in addition to K⁺ the passage of ions and signalling molecules with molecular weight below 1000 Da. Malfunction of the hemichannels during an action potential, when K⁺ is released into the extracellular space between the axon membrane and the surrounding Schwann cell, could lead to significant changes in extracellular [K⁺] that can modify membrane excitability (59) and even lead to hyperexcitation of myelinated nerve fibres (27, 60).

The oxaliplatin effect

The functional correlate of OIN in peripheral nerve fibres as demonstrated here in the mouse sciatic nerve consists of a characteristic, severe, time-dependent broadening of the evoked CAP repolarising phase reaching a plateau at over 5-fold of baseline within 4-5 h, the so called oxaliplatin-effect. This broadening is not caused by a gradual decrease in nerve fibre conduction velocity. Rather, as also shown in our previous intra-axonal recordings from rat nerve fibres, this deformation of the CAP waveform is a result of increased duration, by about 300-400% of control, of the repolarising phase of individual intra-axonally recorded action potential generated by a single stimulus (7; see also Adelsberger et al 2000, Grolleau et al 2001 , Webster et al 2005, Kagiava et al 2008). This type of prolongation indicates a hyperexcitation of the nerve fibers but only during the evoked action potential, since their resting membrane potential remains constant, between -80mV to -90mV, for over 1 h incuation in oxaliplatin (7).

A prolongation of the repolarizing phase of the action potential is closely related to malfunction of Shaker-type Kv1.1/1.2 VGKCs, which are highly enriched in the juxtaparanodal axonal membrane and promote membrane repolarization and maintenance of the internodal resting potential (43, 44). Interestingly, peripheral nerve fiber hyperexcitability that develops acutely following oxaliplatin exposure resembles neuromyotonia, a channelopathy linked to dysfunction of juxtaparanodal VGKCs (45-47). Nevertheless, the strong effect of oxaliplatin on the repolarising phase of either the action potential or the evoked CAP in isolated adult sciatic nerve fibres has not been shown by any other compound, not even 4-AP and TEA, typical VGKC blockers when applied to nerves from adult rats (6). However, when 4-AP was applied on nerve fibres from young rats before completed myelination an evoked action potential broadening occurred (48), having a striking similarity with those recorded from a single fiber of the adult rat exposed to oxaliplatin (7). The studies with VGKC blockers suggested that 4-AP-sensitive channels become masked as they are covered by myelin during maturation in the juxtaparanodal region and remain well-shielded (49), while oxaliplatin may enter the periaxonal space following different routes, like the GJ channels. In contrast to 4-AP, direct binding of oxaliplatin to juxtaparanodal Kv1.1/1.2 VGKCs could not be confirmed (11 , 22). Thus, VGKC involvement in OIN is likely through a functionally related target, which appears to be the Cx29 GJ hemichannels which are located within 30-50 nm from VGKCs in the narrow juxtaparanodal space (28) and are predicted to play a crucial role in siphoning K+ released in the narrow juxtaparanodal periaxonal space during neuronal activity into adaxonal Schwann cell cytoplasm. The severe broadening of RT of the evoked CAP by 25 μΜ oxaliplatin, found in all nerves examined, could be interpreted as a direct distortion of VGKCs, causing increase of K+ in the fiber. However, this is not the case since the whole oxaliplatin-effect was reversed by 88.6% though octanol, a well-known blocker of GJ channels and hemichannels. This a clear indication that the malfunction of VGKCs during the oxaliplatin-effect is secondary to dysfunction of GJ and GJ hemichannels Thus, the conclusion is that oxaliplatin acts as an opener of GJ channels and hemichannels, causing a state of prolonged opening as also reported in other pathological conditions (Contreras et al., 2003, Gomes et al., 2005, Retamal et al., 2007, Orellana et al., 2010). In order to clarify whether octanol acts on full GJs formed by Cx32 (Scherer et al., 1995) or on GJ hemichannels formed by Cx29 (Altevogt et al., 2002), the only known connexin channels in peripheral nerves, or both, we further investigated its effects in Cx32 KO and Cx29 KO mice. The vitality of the KO nerves was not significantly affected by the recording conditions although Cx32 and Cx29 are important for myelinated nerve fibre function (Altevogt et al., 2002, Scherer et al., 1995). This is probably due to low frequency stimulation (1 Hz), which allows the remaining GJs or GJ hemichannels to maintain adequate homeostasis. KO nerves also responded to 400 nM octanol as the WT nerves. A difference was seen in the response to oxaliplatin, which was almost identical for Cx29 KO and WT (RT>1035%), but lower (683.9%) for the Cx32 KO. Shorter RT in the Cx32 KO CAP could be due to the absence of the second peak of the CAP waveform, observed in both WT and Cx29. Also, there was a significant difference in the response to prolonged exposure to oxaliplatin, during which in WT nerves RT increase persisted for over 1200 min while in the KO nerves there was a peak near 240-340 min and then the RT gradually decreased. Finally, the protection by octanol against the oxaliplatin-effect persisted in both connexin KO nerves, although with lower intensity in Cx32 KO than in Cx29 KO. Taken together, these results support our conclusion that oxaliplatin acts as an opener of GJ channels more than hemichannels in peripheral nerve fibres.

The above siphoning model indicates that the voltage-dependent opening and closing of Cx32 channels and Cx29 hemichannels is vital for the function of the myelinated nerve fibres. However, hemichannels typically closed under resting conditions open in response to a variety of physiological and pathological stimuli (67-71). This may cause disruption of the resting membrane potential, release of cytotoxic levels of ATP (63) and glutamate (72) and uptake of water resulting in cell swelling and rupture (73). In our study the characteristic effect of 25 μΜ oxaliplatin on the sciatic nerve is the severe broadening of the CAP repolarising indicating a malfunction of the VGKCs, as it has been previously discussed. However, this is secondary to dysfunction of Cx29 henichannels. Our results show that blocking (closing) of the hemichannels by octanol eliminates the CAP, thus the other possible effect by oxaliplatin could be their complete opening, as it has been suggested for other physiological and pathological stimuli. Prolonged hemichannel opening in the presence of oxaliplatin would permit passive, non-physiological K⁺ transport from the Schwann cell cytoplasm to the juxtaparanodal periaxonal space following concentration gradients, gradually building a surplus of K⁺ in the juxtaparanodal region. Disruption of K⁺ homoeostasis by oxaliplatin with increased extracellular K⁺ is expected to affect VGKC function, explaining the increase of 400% in the repolarising phase of the action potential (7) and eventually the increase in CAP duration shown here.

Prolonged or even permanent GJ channel and hemichannel opening in the presence of oxaliplatin would first prevent the removal of the excess K⁺ caused by the activation of VGKCs and then permit passive, non-physiological K⁺ transport from the Schwann cell cytoplasm to the juxtaparanodal periaxonal space following concentration gradients, gradually building a surplus of K⁺ in this region. Disruption of K⁺ homoeostasis by oxaliplatin with increased extracellular K⁺ in the juxtaparanodal region is expected to affect VGKC function, explaining the increase of 400% in the repolarising phase of the action potential (Kagiava et al., 2013) and eventually the increase in the RT of the CAP shown here. The axon responds to this excess of extracellular K⁺ by post-stimulus repetitive firing, which increases in frequency with time of exposure to oxaliplatin (Kagiava et al., 2013). This repetitive firing starts when repolarizing of the action potential reaches values near - 48 to -50 mV, the activation level of the fast Kv channels (Baker and Ritchie, 1996), and ends near the normal resting membrane potential of -90 mV (Kagiava et al., 2013). This has been suggested to act as a pumping mechanism for the release of the excess K⁺ from the axon to the periaxonal juxtaparanodal region (Kagiava et al., 2013). Such a repetitive firing has also been observed in rat nerve fibres exposed to 4-AP (Astion et al., 1988) or even in the CNS of the locust, where pharmacological blockade of GJs induces repetitive surging of extracellular K⁺. Thus, glial spatial buffering through GJs plays an essential role in the regulation of [K⁺] under normal conditions, contributing to K⁺ clearance following physiologically elevated levels (Rash, 2010, Spong and Robertson, 2013). These phenomena are clinically relevant, as in oxaliplatin treated patients nerve conduction studies reveal hyperexcitability which appears as repetitive motor discharges after single electrical stimulus (Wilson et al., 2002, Lehky et al., 2004), while sensory hyperexcitability findings precede the development of neuropathy (Park et al., 2009). Why cold induces paresthesias in acute OIN (Lehky et al., 2004) remains unclear, but this phenomenon may be related to the fact that cold impairs GJ conductance as well as biosynthesis (Bukauskas and Weingart, 1993, Saitongdee et al., 2000), so that GJ channels may be more vulnerable to the effects of oxaliplatin at lower temperatures. Further experimental insights will be needed to clarify the role of temperature in OIN.

Neuroprotection of octanol against oxaliplatin and Wee versa

The fact that oxaliplatin causes opening of GJ hemichannels was verified using GJ blocker octanol, which eventually reversed the oxaliplatin effect, providing an excellent neuroprotection against OIN as modelled in the ex vivo nerve study. Octanol rescued the prolongation of the CAP repolarising phase to only 145% of baseline after 20 h incubation, down from 586.9% when oxaliplatin was applied alone, a 92% protection, or from 1 173.8±0.92% of baseline when oxaliplatin was applied alone to only 146.0±13.29% even after 1200 min incubation, an 88.6% protection. Indeed, the neuroprotection offered by octanol was slightly concentration-dependent with a minimum at 200 nM (85.3% protection of RT prolongation) and a maximum at 600 nM (87.7% protection). Octanol offers almost the same neuroprotection at the low concentration of 200 nM as at the higher and neurotoxic 600 nM. This is clinically relevant because the minimum effective octanol concentration should be used, to minimize the side effect of decreasing the CAP amplitude and the number of malfunctioning nerve fibers. Our recordings, obtained for over 20 h showed clearly that the octanol inhibition-recovery effect on the evoked CAP dominated over the characteristic oxaliplatin-effect, the severe broadening of the CAP. Furthermore, supporting its opposite action of GJ and hemichannel opening, oxaliplatin appeared to antagonize the octanol-effect of CAP inhibition-recovery. In the presence of oxaliplatin and 400 nM octanol there was a 98% maximum recovery in the amplitude of CAP, instead of 67% when octanol was applied alone. The antagonizing effect of oxaliplatin became even more impressive when applied with 600 nM octanol, preventing the complete elimination of the CAP. This is a clear indication that the opening and fixation of GJ and GJ hemichannels by oxaliplatin is compensated by octanol. These mutual neuroprotective effects occurred only when the two compounds were added together or when the nerve was pre-incubated in octanol. However, when octanol was added first its neuroprotection was drastically decreased, while when the nerve was treaded first with oxaliplatin and then with octanol, oxaliplatin became extremely neurotoxic. There is no obvious explanation for this phenomenon, but a similar effect was reported in other models where connexin hemichannel blockade was neuroprotective after, but not during global cerebral ischemia in near-term fetal sheep (35). The phenomenom may be related to the nature of interaction of each compound with the connexins channels, that could involve common or different binding sites and cause conformational changes of channel forming domains (Spray et al., 1986) affecting the subsequent binding of the other compound. Octanol, similar to other GJ blockers, has been shown to bind only to extracellular connexin sites that differ from the Ca²⁺ binding site (Eskandari et al., 2002), while the biophysical interaction of oxaliplatin with connexins will need to be clarified in future studies. A similar effect was reported in other models where connexin hemichannel blockade was neuroprotective after, but not during global cerebral ischemia in near-term fetal sheep (Davidson et al., 2014). The neuroprotective role of octanol against oxaliplatin found in the sciatic nerve preparation was verified by our in vitro experiments in HeLa cells expressing Cx31.3, the human ortholog of Cx29 (Sargiannidou et al., 2008), where oxaliplatin accelerated the uptake of neurobiotin through hemichannels but this opening effect was reversed by octanol in a concentration-dependent manner. Furthermore, oxaliplatin accelerated dye transfer between Cx32 expressing cells, and this effect was again reversed by octanol.

The results of the sciatic nerve study were verified by our in vitro experiments in HeLa cells expressing Cx31.3, the human ortholog of Cx29 (29), where oxaliplatin accelerated the uptake of neurobiotin through hemichannels and this opening effect was reversed by octanol in a dose-dependent manner. Thus, oxaliplatin appears to uncouple and maintain open Cx32 channels and voltage-gated Cx29 hemichannels that would normally open only briefly during the nodal Na+ current and in response to K⁺ release into the periaxonal space to facilitate siphoning into the Schwann cell. Excess K⁺ in the juxtaparanodal region disrupts VGKC function and this appears in the post-stimulus intra-axonal action potential as either a plateau or as repetitive firing (7). This gradual juxtaparanodal accumulation of K⁺ increasingly affecting VGKC function would explain why the oxaliplatin-effect in this study and in all our previous experiments performed in rodent nerves using similar oxaliplatin concentrations {Kagiava, 2013 #3;Kagiava, 2008 #4} occurs with a delay which is time- and concentration-dependent. In our experiment with 25 μΜ oxaliplatin the development of the maximum oxaliplatin-effect occurred within 300 to 360 min in nerves from either WT or KO animals. In other studies of OIN the oxaliplatin effect on the nerve was very fast, but the concentrations used were much higher and not as clinically relevant (above). At higher concentrations, 150 μΜ oxaliplatin, the same effect occurred within 50- 60 min (Kagiava et al., 2013). The model of oxaliplatin-induced uncoupling of connexin channels proposed here for the peripheral nerve could also explain in part the increased coupling of satellite cells in DRGs induced by oxaliplatin, lowering pain threshold, which can also be antagonized by GJ blockers (Warwick and Hanani, 2013), as well as the increased coupling of spinal cord astrocytes leading to oxaliplatin-induced mechanical hypersensitivity (Yoon et al., 2013). Our results indicate that oxaliplatin counteracts the GJ blocking effect of octanol by maintaining the GJ channels and hemichannels open. The model of oxaliplatin-induced uncoupling of GJ channels and hemichannels we propose here for the peripheral nerve, could also explain in part the increased coupling of satellite cells in DRGs, lowering pain threshold, which can also be antagonized by GJ blockers (32), as well as the increased coupling of spinal cord astrocytes leading to oxaliplatin-induced mechanical hypersensitivity (31).

Morphological changes following prolonged oxaliplatin exposure

Our electrophysiological studies show that nerve fibres incubated in 25 μΜ oxaliplatin develop within the first 3-4 h a severe increase in the duration of the CAP repolarising phase which reaches a maximum after 4 h and persists for the remaining of the 24 h incubation time. We then asked whether there is an ultrastructural correlate to this oxaliplatin effect. Detail morphological studies revealed that after the first 3 h, when the electrophysiological alterations are fully developed, there is no alteration in the fine structure of the nerve fibres, focusing on perinodal areas. Thus, functional alterations in our model precede any morphological changes, as also described in the clinical setting (3). This fact also excludes the possibility that oxaliplatin causes a primary structural alteration involving the VGKCs and associated molecules or Cx29 hemichannels, that could then lead to dysfunction of the peripheral nerve. However, with further oxaliplatin exposure for 20 h we found characteristic axonal edema in the juxtaparanodal and juxtaincisure areas, as shown by teased fiber and electron microscopy examination. These changes occurred only in oxaliplatin treated fibers that exhibited the CAP prolongation, but not in control or in oxaliplatin+octanol treated fibers, again in line with clinical data supporting that the early findings of hyperexcitability are predictive of later development of neuropathy (3). Despite the late morphological changes in the juxtaparanodal areas due to axonal edema, the clustering of VGKCs in oxaliplatin-exposed nerves was not disrupted (78, 79), nor there was any diffusion of VGKCs into the paranodal areas (80) as seen with mutants of axonal domains. Furthermore, none of those mutants show the severe CAP prolongation caused by oxaliplatin. Finally, we could not detect any abnormalities in Na7K⁺ ATPase subunit expression, in agreement with previous electrophysiological studies showing a much different functional effect of ATPase blockers on the sciatic nerve compared to the oxaliplatin effect (Kagiava et al., 2012), indicating that ATPase is not directly involved in OIN. The edematous axonal changes following prolonged oxaliplatin exposure are unique. We speculate that prolonged nerve exposure to oxaliplatin and as a result the uncoupling and persistent opening of Cx29 hemichannels and Cx32 GJs under the myelin sheath leads to reversal of K⁺ flow from Schwann cell cytoplasm to the periaxonal space. Gradual accumulation of K⁺ in this narrow space and cancellation of the activity-dependent siphoning function may also reverse cation flow resulting in osmotic edema in the axon. The collapse of K+ homeostasis may lead to disruption of the resting membrane potential, release of cytotoxic levels of ATP (63) and glutamate (72) and uptake of water resulting in cell swelling and rupture (73). Similar mechanisms have been described in the CNS (76).

Octanol as neuroprotectant against OIN in clinical practice

Numerous attempts to reduce OIN have been reported, using goshajinkigan (81), synthetic eel calcitonin and elcatonin (82), neurotropin (83), silibinin (84), minocycline (85), monosialotetrahexosyl ganglioside (86), and calcium blockers (19). Similarly, ethosuximide (87) and neurotropin (88) were found to reverse paclitaxel-induced peripheral neuropathy without affecting anti-tumour efficacy.

Our studies clearly support the potential for octanol to efficiently prevent OIN, considered to be by blocking Cx32 and Cx29 channels, which are present in rodent peripheral nerves. This protective effect was verified by our in vitro experiments in HeLa cells expressing either of the two connexins. Octanol can be used for medical treatment since it was found to be well-tolerated and safe in a clinical trial for essential tremor (89, 90). The optimal administration protocol will have to be developed, since for maximum protection we had to apply the two compounds, octanol and oxaliplatin simultaneously. When octanol applied before oxaliplatin the neuroprotection was drastically reduced, while when octanol was applied before oxaliplatin it became neurotoxic. Nevertheless, administration of octanol prior to oxaliplatin still provided a marked neuroprotection. Octanol offers the advantage that it can be used at very low concentrations (200 nM) against OIN, while for other blockers such as GRA much higher concentrations (25 μΜ) are required to offer a weaker protection. Although octanol offers almost the same neuroprotection against OIN at low and high concentrations, the lower concentrations appear to have almost no side effects, reflected in CAP preservation. Further studies in in vivo models of OIN will be needed to reproduce the ex vivo findings and inform the best approach for future clinical trials as well. Besides octanol we also studied three more GJ blockers, carbenoxolone, GRA and OA for the potential to offer protection against OIN, but could not see a clear neuroprotective effect. Overall, carbenoxolone, GRA and OA provided weaker or no protection at lower concentrations, over 125-1500 times higher than the corresponding concentration for octanol, with little toxicity, whereas at higher concentrations they were neurotoxic, eliminating the CAP. Thus, carbenoxolone offered no protective effect at non-toxic concentrations (25 μΜ), while OA (300 μΜ) offered only a 37.6% improvement in RT prolongation caused by oxaliplatin. GRA (25 μΜ) showed a better profile, offering a clear neuroprotective effect of 78.8% at relatively non-toxic concentrations. Thus, GRA largely reproduced the effects of octanol, but at much higher concentrations. Furthermore, oxaliplatin antagonized the negative impact of GRA on the CAP, as we observed for octanol. This high concentration may be deleterious when used in in vivo studies or in clinical practice. The lack of protection by carbenoxolone may be due to the fact that it has a number of other effects, including a reduction in excitatory and inhibitory synaptic currents, alteration of intrinsic membrane properties and suppression of action potentials (Rouach et al., 2003, Tovar et al., 2009, Beaumont and Maccaferri, 2011 ). Carbenoxolone has also been reported to block Ca²⁺ channels, pannexin channels and P2X7 receptors at concentrations similar to or lower than those blocking connexin channels (Vessey et al., 2004, Bruzzone et al., 2005, Suadicani et al., 2006). Taken together, the results of testing further GJ blockers against OIN confirm the protective effect of blocking GJ channels. However, octanol remains the best choice with the most favourable therapeutic window.

In conclusion, we provide a mechanistic explanation for OIN based on uncoupling by oxaliplatin of Cx32 GJ channels and Cx29/Cx31.3 GJ hemichannels apposing VGCKs, leading to early functional abnormalities promoting hyperexcitability, and to later morphological changes consisting of axonal edema. Furthermore, the clinically relevant GJ hemichannel blocker octanol effectively prevents oxaliplatin effects ex vivo and in vitro and should be further studied in vivo for the possibility of preventive use in OIN.

Example 6

Methods

Mouse strains and procedures

Two month old wild type (WT) C57BL/6 as well as Gjb1 -null mice/Cx32 knockout (KO) (C57BL/6J29) and G/c3-null/Cx29 KO (129P2/OlaHsd*C57BL/6*SJL) mice obtained from the European Mouse Mutant Archive, Monterotondo, Italy; both originally generated by Prof. Klaus Willecke, University of Bonn, Bonn, Germany) weighing 20-25 g were used. Animals were sacrificed using N₂ and cervical dislocation prior to nerve dissection. The sciatic nerve was exposed and dissected from the spinal cord to the knee. All experimental procedures were conducted in accordance with the animal care protocols outlined by the Aristotle University of Thessaloniki, Greece and the Veterinary Authorities (License No 107241/2013) and the Cyprus Government's Chief Veterinary Officer according to EU guidelines (EC Directive 86/609/EEC).

Ex vivo sciatic nerve recording bath

The sciatic nerves were immersed in a modified Krebs-Ringer solution containing (in mmol/l): 136 NaCI, 4.7 KCl, 2.4 CaC , 1.1 MgCb, 1 NaHC0₃, 11 glucose, and 10 HEPES (all from Sigma, Germany); pH=7.2. All experiments were performed at a constant temperature of 25.0 ± 1.0°C. The epineural sheath was removed to ensure that nerve fibers came into direct contact with the drug under investigation and to maximize drug access to all nerve fibers. To expose the nerve to the drug under investigation and simultaneously monitor its electrophysiological responses to electrical stimuli (whole-nerve extracellular recordings), we used a three-chambered recording bath made of paraffin as described in detail in our previous studies (39). A similar, though slightly larger, recording bath, has been used for a variety of neurotoxicological studies on the rat sciatic nerve (7). Briefly, the recording bath consists of three chambers: the stimulating, the perfusion (middle), and the recording chamber. The chambers with volume of 10-11 ml were placed in a raw 1 mm apart. The nerve was mounted across the three chambers of the recording bath with 2-3 mm of the proximal part of the nerve placed in the stimulating chamber, where it was electrically stimulated (considered to be supramaximal stimuli) at 1 Hz (pulse amplitude: 2.0-3.0 V, duration: 0.01 ms) using an electrode connected to a constant voltage stimulator (Digitimer, England, UK) to evoke the nCAP. The main region of the nerve - about 10-12 mm, over 70.0% of the total nerve length - was bathed in the perfusion chamber and was exposed to the saline (control) or the drugs under investigation (oxaliplatin, octanol and their combination). Finally, the distal part of the nerve was placed in the recording chamber, where the evoked nCAP was recorded using an electrode immersed in the chamber connected to AC amplifier (Neurolog NL822, Digitimer, UK). Immediately after the placement of the nerve, each chamber was filled once with 10.0 ml oxygenated (O2, 100%) saline. The saline solution in the stimulating and recording chambers was stagnant during the 24 h recording period, only the saline in the perfusion chamber was continuously stirred. The electrodes (active and reference) from either amplifier or stimulator were made of 24-carat gold. Finally, the recording bath was air-tight shielded to avoid evaporation of the saline during the long recording period (20 h or more).

Electrophysiological data analysis The amplitude of the CAP (example of baseline recorded evoked CAPs is given in Figure 1 A) was measured from the baseline to the peak (in Volts). The duration of the repolarising phase of the CAP or the repolarizing time (RT) was measured from the peak up to the end of the repolarising phase (in ms), at the point where the repolarizing phase meets the baseline (example in Figure 2A). CAP amplitude and duration, measured in Volts and ms, respectively, were expressed as a percentage of the initial values at t=0, or of the values before the application of the tested compound, which were considered as 100%. Values obtained from repeated experiments were averaged and expressed as means ± SEM. Using these values, the mean time-response curves were plotted (example in Figure 2E). Statistical significance was examined by one-way ANOVA and Student's t- test.

Sciatic nerve drug exposure protocol

The CAP of the mouse sciatic nerve immersed in normal saline solution remains constant for over 24 h (39) indicating that the vitality of the sciatic nerve fibres under these recording conditions is not impaired. Oxaliplatin (Tocris Bioscience) was pre-dissolved in distilled water to make the stock solution (5.0 mg/ml) and stored under light protection at -20.0°C. Oxaliplatin from the stock was diluted in the saline contained in the perfusion (middle) chamber to make the desirable concentration, 25 μΜ, and single evoked CAPs were digitised and stored every 5 min throughout the 24 h experiment, as described above. The same procedure was followed to assess the effect of 200, 400, 600 and 2000 nM octanol, 25 and 50 μΜ carbenoxolone, 25 and 50 μΜ, 18-beta-glycyrrhetinic acid (GRA) and 100- 300 μΜ octanoic acid (OA) (all from Sigma-Aldrich). To assess the action of the combination of oxaliplatin and octanol we used three different conditions: a) octanol (400nM) and oxaliplatin (25 μΜ) were applied simultaneously and the preparation was left for 24h with continuous CAP monitoring as above; b) the nerve was incubated first in octanol (400 nM) for 40 min and then 25 μΜ oxaliplatin was added in the perfusion chamber; and c) the nerve was incubated first in 25 μΜ oxaliplatin for 40 min and then octanol (400 nM) was added in the perfusion chamber. Carbenoxolone, GRA, and OA were applied only simultaneously with oxaliplatin as above.

Teased fiber preparation and immunostaining

Following the ex-vivo electrophysiological studies, sciatic nerves exposed for either 3 or 24 h to oxaliplatin or to the combination of oxaliplatin plus octanol (above) along with control nerves (stimulated for the same time but not exposed to drugs), were fixed for 30 min in fresh 4% paraformaldehyde (PFA) (Sigma, Germany) in 0.1 M PB. Teased nerve fibers were prepared from fixed nerves, dried on SuperFrost Plus glass slides overnight at room temperature (RT) and stored at -20°C. For immunostaining, teased fibers were permeabilized in acetone (-20 °C for 10 min) and incubated at RT with blocking solution (5% bovine serum albumin (BSA), 0.5% Triton-X) for 1 h, followed by primary antibodies diluted in blocking solution overnight at 4°C. Antibodies included mouse monoclonal against Caspr (1 :50; gift of Dr. Elior Peles, Weizmann Institute of Science), MBP (1 :500; Abeam), SMI31 (1 :5000; Abeam), PanNav (1 :50; Sigma), as well as rabbit antisera against Cx29 (1 :300; Invitrogen), MBP (1 :100; Sigma), Kv1.2 (1 :200; Alomone), Caspr2 (1:100; Sigma), Nav1.6 (1 :100; Alomone). Teased fibers were then washed in PBS and incubated with anti-MsFITC and anti-RbTRITC (Jackson ImmunoResearch, 1 :300) for 1 h at RT. Schwann cell nuclei were visualized with 4',6'-diamidino-2-phenylindole (DAPI; Sigma- Aldrich). Slides were mounted with Dako Fluorescent Mounting Medium and images were photographed under a Zeiss fluorescence microscope with a digital camera using the Zeiss Axiovision software (Carl Zeiss Microimaging, Germany). Where appropriate, we obtained images with comparable exposure times to allow better comparison between different experiments.

Electron microscopy

Following ex-vivo electrophysiological studies, sciatic nerves exposed to oxaliplatin or to the combination of oxaliplatin plus octanol (above) along with control nerves without pharmacological exposure, were fixed in glutaraldehyde in 0.1 M PB overnight at 4°C, then osmicated, dehydrated, and embedded in Araldite resin. Transverse semi-thin sections (1 μηι) were obtained and stained with alkaline toluidine blue. Ultrathin sections (80-100 nm) were counterstained with lead citrate and uranyl acetate before being examined in a JEOL JEM-1010 transmission electron microscope (JEOL Ltd, Tokyo, Japan). Neurobiotin uptake assay

Cx31.3-expressing clonal HeLa cells were generated by transfection using the human Cx31.3 gene open reading frame (GenBank accession number AY297109) cloned into plRESpuro3 vector as previously described (29) and maintained in selection media with puromycin (0.5 Mg/ml). Expression of Cx31.3 at the cell membrane of these cells was confirmed by immunostaining and immunoblot analysis (29). For the neurobiotin uptake assay, cells permanently expressing Cx31.3 were grown to about 50-70% confluency in 4-well chamber slides, washed in PBS lacking divalent ions, pre-incubated in Optimen for 30 min, and then incubated with 2% neurobiotin for 5-60 min, washed and fixed for 30 minutes in 4% PFA at 4°C. After blocking in 5% BSA with 0.1% Triton X-100 for 30 min at RT, cells were incubated in streptavidin-Texas red (Vector Laboratories, 1 : 1000) at RT for 10 min, counterstained with DAPI and imaged as above. To determine the effects of oxaliplatin we performed the same experiments in the presence of oxaliplatin alone (25 and 50 μΜ) or oxaliplatin in combination with octanol (200 nM-2000 nM) diluted in Optimem for 30 min before and during incubation with neurobiotin.

For quantitative analysis of neurobiotin uptake with or without addition of oxaliplatin and octanol in untransfected or Cx31.3 expressing cells, the mean red channel pixel intensity was measured in images captured with constant time exposures using the ImageJ software (NIH image). At least 2000 cells in each condition were compared. Statistical significance was examined using the Student's t-test. All experiments were performed in triplicates.

2.8. Neurobiotin scrape loading assay

Cx32 expressing clonal HeLa cells were plated on 6-well plates to reach ~90-100% confluency on the day of the scrape loading. For neurobiotin scrape loading, cells were rinsed 3 times in HBSS without calcium or magnesium and 1 % neurobiotin was added alone, or in combination with either octanol 400 nM, oxaliplatin 25 μΜ, or with both octanol and oxaliplatin. A scalpel blade was used to cut a grid for scraping the cells, and then cells were incubated in neurobiotin for 15 or 30 min, washed 3 times in HBSS, fixed with cold 4% paraformaldehyde (PFA) for 10 min, washed 3 times in PBS, blocked in 5% BSA with 0.1% Triton X-100 for 30 min at RT, incubated in streptavidin-rhodamine (Vector Laboratories, diluted 1 :300) at RT for 1 hour, washed 3 times in PBS, 3 times in H₂0, incubated with DAPI for 5 min at RT, washed, dried, and pictures were taken as above. For quantification of gap junctional connectivity, the number of fluorescent cells in a rectangle on one side of the scrape line but excluding cells that were in the scrape line were counted and the mean ± SEM were calculated for each condition and compared with the Student's t-test.

The ex vivo sciatic nerve preparation for recordings of the CAP

Animals were sacrificed using N2 and cervical dislocation prior to nerve dissection. The sciatic nerve was exposed and dissected from the spinal cord to the knee. The sciatic nerves were immersed in a modified Krebs-Ringer solution containing (in mmol/l): 136 NaCI, 4.7 KCI, 2.4 CaCI2, 1.1 MgCI2, 1 NaHC03, 11 glucose, and 10 HEPES (all from Sigma, Germany); pH=7.2. All experiments were performed at a constant temperature of 26.0 ± 1.0°C. In order to ensure that the nerve fibers came into direct contact with the drug under investigation and to maximize drug access to all nerve fibers, the epineurium (epineural sheath) was removed.

To expose the nerve to the drug under investigation and simultaneously monitor its electrophysiological responses to electrical stimuli (whole-nerve extracellular recordings), we used a three-chambered recording bath made of Plexiglas, as described in detail in our previous studies [6], Briefly, the recording bath consists of three chambers: the stimulating, the perfusion (middle), and the recording chamber. The chambers (24x24x10 mm) have a volume of 10-11 mL and were placed one next to another with a distance of 2 mm. The nerve was mounted across the three chambers of the recording bath. The middle chamber was the perfusion chamber, while the chamber in the left was the stimulating and the chamber on the right the recording. The 2-3 mm of the proximal part of the nerve was placed in the stimulating chamber, where the nerve was electrically stimulated at 1 Hz (pulse amplitude: 2.0-3.0 V, duration: 0.01 ms) using an electrode connected to a constant voltage stimulator (Digitimer, England, UK ) to evoke the nerve CAP. The main region of the nerve - about 10-12 mm, which was over 70.0% of the total nerve length - was bathed in the perfusion chamber, and this part was exposed to the drugs under investigation (oxaliplatin, octanol and their combination). Finally, the distal part of the nerve was placed in the recording chamber, where the evoked CAP was recorded using an electrode immersed in the chamber connected to AC amplifier (Neurolog NL822, Digitimer, UK). Immediately after the placement of the nerve, each chamber was filled once with 10.0 ml oxygenated (02, 100%) saline. The saline solution in the stimulating and recording chambers was stagnant during the 24 h recording period, only the saline in the perfusion chamber was continuously stirred. The electrodes (active and reference) from either amplifier or stimulator were made of 24-carat gold. Finally, the recording bath was air-tight shielded to avoid evaporation of the saline during the long recording period (20-24h).

Example 7

Further studies

We have performed extensive further studies to clarify the cellular and molecular mechanisms of oxaliplatin induced neurotoxicity. First, we used Cx29 knockout (KO) mice to confirm whether Cx29 was indeed the major target of oxaliplatin. Second, we have used additional gap junction (GJ) blockers besides octanol, to examine whether they would offer similar neuroprotective effects. The results of these studies are described in detail below.

1. Studies in Cx29 KO mice

We have performed extensive further studies in Cx29 KO sciatic nerves after establishing this line in our laboratory. To our surprise, the oxaliplatin effect (CAP repolarization phase prolongation) largely persisted on these nerves, indicating that oxaliplatin does not act only on Cx29 hemichannels but also on Cx32. Furthermore, octanol still inhibited the oxaliplatin effect in Cx29 KO nerves. We therefore revisited our results in the Cx32 KO nerves, and realized that the effect of oxaliplatin was decreased in Cx32 KO to a greater degree rather than in Cx29 KO nerves, suggesting that oxaliplatin acts on both connexin channels, but the effect on Cx32 gap junction channels is greater than the effect on Cx29 hemichannels.

The effect of oxaliplatin, apparently forced opening of Cx32 gap junction channels, similar to the opening of Cx29/Cx31.3 hemichannels, has also been verified in vitro using the neurobiotin scrape loading assay in Cx32-expressing HeLa cells. Furthermore, we have revisited the calculation of repolarization phase in our experiments with the CAP measuring for longer post-stimulus periods and found that the prolongation caused by oxaliplatin was actually longer than we initially estimated, therefore we revised the values in all experiments, now termed repolarizing time (RT). Thus we conclude that oxaliplatin affects both Cx32 and Cx29 channels, but more prominently Cx32, based on the residual effect in each of the two types of KO nerves. Loss of Cx32 causes neuropathy in humans and rodent models, clearly suggesting that - at least in Schwann cells- there is no other connexin to compensate for the loss of Cx32 function.

2. Studies with further gap junction blockers

We have studied two further gap junction inhibitors besides octanol, 18-beta-glycyrrhetinic acid (GRA) and octanoic acid (OA), and provide complete results of their effects (along with those of carbenoxolone). While carbenoxolone does not appear to offer any neuroprotection at non-toxic concentrations, GRA offers significant protection at lower while it is toxic at higher concentrations, and OA offers modest protection at higher concentrations at which it starts being toxic. Overall, we reproduced the effects of octanol with these additional GJ blockers, but clearly show that compared to the other blockers, octanol offers the best and widest therapeutic window and we consider is therefore the best candidate for further study to treat OIN.

Furthermore, after revisiting our results, it became clear that not only the simultaneous application of octanol with oxaliplatin results in neuroprotection (by 88.6%), but also the application of octanol prior to oxaliplatin, albeit weaker (81.4%). We have clarified these results in the revised paper. Still the administration of octanol after oxaliplatin appears to be toxic. This may be due to an irreversible effect of oxaliplatin on connexin channels, which becomes even more deleterious with addition of the blockers.

Furthermore, octanol clearly appears to have a wide therapeutic window showing significant neuroprotection (>80% rescue of repolarization time prolongation caused by oxaliplatin) at concentrations from 200-600 nM, with relatively little toxicity measured by amplitude reduction of the evoked nerve CAP. This is in contrast to carbenoxolone (tested in our initial studies) and two additional gap junction blockers tested (GRA and OA), which show partial neuroprotection with more toxicity (amplitude reduction). Thus, octanol appears to offer the optimal efficacy-tolerability ratio, at least in this ex-vivo model. References cited in the Examples

I . Andre T, Boni C, Mounedji-Boudiaf L, Navarro M, Tabernero J, Hickish T, et al. Oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment for colon cancer. N Engl J Med. 2004;350(23):2343-51. Epub 2004/06/04.

2. Kweekel DM, Gelderblom H, Guchelaar HJ. Pharmacology of oxaliplatin and the use of pharmacogenomics to individualize therapy. Cancer Treat Rev. 2005;31 (2):90-105. Epub 2005/04/26.

3. Park SB, Lin CS, Krishnan AV, Goldstein D, Friedlander ML, Kiernan MC. Oxaliplatin-induced neurotoxicity: changes in axonal excitability precede development of neuropathy. Brain. 2009;132:2712-23.

4. Benoit E, Brienza S, Dubois JM. Oxaliplatin, an anticancer agent that affects both Na+ and K+ channels in frog peripheral myelinated axons. Gen Physiol Biophys. 2006;25(3):263-76. Epub 2007/01/02.

5. Lang PM, Fleckenstein J, Passmore GM, Brown DA, Grafe P. Retigabine reduces the excitability of unmyelinated peripheral human axons. Neuropharmacology.

2008;54(8):1271-8. Epub 2008/05/14.

6. Kagiava A, Tsingotjidou A, Emmanouilides C, Theophilidis G. The effects of oxaliplatin, an anticancer drug, on potassium channels of the peripheral myelinated nerve fibres of the adult rat. Neurotoxicology. 2008;29(6): 1100-6. Epub 2008/10/11.

7. Kagiava A, Kosmidis EK, Theophilidis G. Oxaliplatin-induced hyperexcitation of rat sciatic nerve fibers: an intra-axonal study. Anticancer Agents Med Chem. 2013;13(2):373- 9. Epub 2012/06/23.

8. Sittl R, Carr RW, Fleckenstein J, Grafe P. Enhancement of axonal potassium conductance reduces nerve hyperexcitability in an in vitro model of oxaliplatin-induced acute neuropathy. Neurotoxicology. 2010;31(6):694-700. Epub 2010/07/31.

9. Nodera H, Spieker A, Sung M, Rutkove S. Neuroprotective effects of Kv7 channel agonist, retigabine, for cisplatin-induced peripheral neuropathy. Neurosci Lett. 2011 ;505(3):223-7. Epub 2011/09/29.

10. Dimitrov AG, Dimitrova NA. A possible link of oxaliplatin-induced neuropathy with potassium channel deficit. Muscle Nerve. 2012;45(3):403-11. Epub 2012/02/16.

I I . Adelsberger H, Quasthoff S, Grosskreutz J, Lepier A, Eckel F, Lersch C. The chemotherapeutic oxaliplatin alters voltage-gated Na(+) channel kinetics on rat sensory neurons. Eur J Pharmacol. 2000;406(1):25-32. Epub 2000/09/30.

12. Webster RG, Brain KL, Wilson RH, Grem JL, Vincent A. Oxaliplatin induces hyperexcitability at motor and autonomic neuromuscular junctions through effects on voltage-gated sodium channels. Br J Pharmacol. 2005; 146(7): 1027-39. Epub 2005/10/19. 13. Argyriou AA, Cavaletti G, Antonacopoulou A, Genazzani AA, Briani C, Bruna J, et al. Voltage-gated sodium channel polymorphisms play a pivotal role in the development of oxaliplatin-induced peripheral neurotoxicity: results from a prospective multicenter study. Cancer. 2013;119(19):3570-7. Epub 2013/07/04.

14. Krishnan AV, Goldstein D, Friedlander M, Kiernan MC. Oxaliplatin and axonal Na+ channel function in vivo. Clin Cancer Res. 2006;12(15):4481-4. Epub 2006/08/11.

15. Park SB, Lin CS, Krishnan AV, Goldstein D, Friedlander ML, Kiernan MC. Dose effects of oxaliplatin on persistent and transient Na+ conductances and the development of neurotoxicity. PLoS One. 2011 ;6(4):e18469. Epub 2011/04/16.

16. Grolleau F, Gamelin L, Boisdron-Celle M, Lapied B, Pelhate M, Gamelin E. A possible explanation for a neurotoxic effect of the anticancer agent oxaliplatin on neuronal voltage-gated sodium channels. J Neurophysiol. 2001 ;85(5):2293-7. Epub 2001/05/16.

17. Hochster HS, O'Reilly EM, Ajani JA, Venook AP. Section III: Treatment of Advanced Gastrointestinal Cancers. Gastrointest Cancer Res. 2007; 1(6 Suppl):S8-S12. Epub 2007/11/01.

18. Sakurai M, Egashira N, Kawashiri T, Yano T, Ikesue H, Oishi R. Oxaliplatin-induced neuropathy in the rat: involvement of oxalate in cold hyperalgesia but not mechanical allodynia. Pain. 2009; 147(1 -3): 165-74. Epub 2009/09/29.

19. Tatsushima Y, Egashira N, Narishige Y, Fukui S, Kawashiri T, Yamauchi Y, et al. Calcium channel blockers reduce oxaliplatin-induced acute neuropathy: a retrospective study of 69 male patients receiving modified FOLFOX6 therapy. Biomed Pharmacother. 2013;67(1):39-42. Epub 2012/12/05.

20. Wilson RH, Lehky T, Thomas RR, Quinn MG, Floeter MK, Grem JL. Acute oxaliplatin-induced peripheral nerve hyperexcitability. J Clin Oncol. 2002;20(7): 1767-74. Epub 2002/03/29.

21. von Delius S, Eckel F, Wagenpfeil S, Mayr M, Stock K, Kullmann F, et al. Carbamazepine for prevention of oxaliplatin-related neurotoxicity in patients with advanced colorectal cancer: final results of a randomised, controlled, multicenter phase II study. Invest New Drugs. 2007;25(2): 173-80. Epub 2006/09/20.

22. Broomand A, Jerremalm E, Yachnin J, Ehrsson H, Elinder F. Oxaliplatin neurotoxicity-no general ion channel surface-charge effect. J Negat Results Biomed. 2009;8:2. Epub 2009/01/14.

23. Ferrier J, Pereira V, Busserolles J, Authier N, Balayssac D. Emerging trends in understanding chemotherapy-induced peripheral neuropathy. Curr Pain Headache Rep. 2013;17(10):364. Epub 2013/09/03.

24. Loprinzi CL, Qin R, Dakhil SR, Fehrenbacher L, Flynn KA, Atherton P, et al. Phase III randomized, placebo-controlled, double-blind study of intravenous calcium and magnesium to prevent oxaliplatin-induced sensory neurotoxicity (N08CB/Alliance). J Clin Oncol. 2014;32(10):997-1005. Epub 2013/12/04.

25. Kocsis JD, Eng DL, Gordon TR, Waxman SG. Functional differences between 4- aminopyridine and tetraethylammonium-sensitive potassium channels in myelinated axons. Neurosci Lett. 1987;75(2): 193-8. Epub 1987/03/31.

26. Zhou L, Messing A, Chiu SY. Determinants of excitability at transition zones in Kv1.1 -deficient myelinated nerves. J Neurosci. 1999;19(14):5768-81.

27. Chiu SY. Functions and distribution of voltage-gated sodium and potassium channels in mammalian Schwann cells. Glia. 1991 ;4(6):541-58. Epub 1991/01/01.

28. Altevogt BM, Kleopa KA, Postma FR, Scherer SS, Paul DL. Connexin29 is uniquely distributed within myelinating glial cells of the central and peripheral nervous systems. J Neurosci. 2002;22(15):6458-70. Epub 2002/08/02.

29. Sargiannidou I, Ahn M, Enriquez AD, Peinado A, Reynolds R, Abrams C, et al. Human oligodendrocytes express Cx31.3: function and interactions with Cx32 mutants. Neurobiol Dis. 2008;30(2):221-33. Epub 2008/03/21.

30. Konishi T. Voltage-gated potassium currents in myelinating Schwann cells in the mouse. J Physiol. 1990;431 :123-39. Epub 1990/12/01.

31. Yoon SY, Robinson CR, Zhang H, Dougherty PM. Spinal astrocyte gap junctions contribute to oxaliplatin-induced mechanical hypersensitivity. J Pain. 2013;14(2):205-14. Epub 2013/02/05.

32. Warwick RA, Hanani M. The contribution of satellite glial cells to chemotherapy- induced neuropathic pain. Eur J Pain. 2013;17(4):571-80. Epub 2012/10/16.

33. Vessey JP, Lalonde MR, Mizan HA, Welch NC, Kelly ME, Barnes S. Carbenoxolone inhibition of voltage-gated Ca channels and synaptic transmission in the retina. J Neurophysiol. 2004;92(2): 1252-6. Epub 2004/03/19.

34. Rouach N, Segal M, Koulakoff A, Giaume C, Avignone E. Carbenoxolone blockade of neuronal network activity in culture is not mediated by an action on gap junctions. J Physiol. 2003;553(Pt 3):729-45. Epub 2003/09/30.

35. Davidson JO, Drury PP, Green CR, Nicholson LF, Bennet L, Gunn AJ. Connexin Hemichannel Blockade Is Neuroprotective after Asphyxia in Preterm Fetal Sheep. PLoS

ONE. 2014;9(5):e96558.

36. He B, Tong X, Wang L, Wang Q, Ye H, Liu B, et al. Tramadol and flurbiprofen depress the cytotoxicity of cisplatin via their effects on gap junctions. Clin Cancer Res. 2009; 15(18):5803-10. Epub 2009/09/03.

37. Burakgazi AZ, Messersmith W, Vaidya D, Hauer P, Hoke A, Polydefkis M. Longitudinal assessment of oxaliplatin-induced neuropathy. Neurology. 2011 ;77:980-6. 38. Kroigard T, Schroder HD, Qvortrup C, Eckhoff L, Pfeiffer P, Gaist D, et al. Characterization and diagnostic evaluation of chronic polyneuropathies induced by oxaliplatin and docetaxel comparing skin biopsy to quantitative sensory testing and nerve conduction studies. Eur J Neurol. 2014;21 :623-9.

39. Kagiava A, Aligizaki K, Katikou P, Nikolaidis G, Theophilidis G. Assessing the neurotoxic effects of palytoxin and ouabain, both Na(+)/K(+)-ATPase inhibitors, on the myelinated sciatic nerve fibres of the mouse: an ex vivo electrophysiological study. Toxicon. 2012;59(3):416-26. Epub 2011/12/31.

40. Horishita T, Harris RA. n-Alcohols inhibit voltage-gated Na+ channels expressed in Xenopus oocytes. J Pharmacol Exp Ther. 2008;326(1):270-7. Epub 2008/04/25.

41. Scherer SS, Deschenes SM, Xu YT, Grinspan JB, Fischbeck KH, Paul DL. Connexin32 is a myelin-related protein in the PNS and CNS. J Neurosci. 1995;15(12):8281-94. Epub 1995/12/01.

42. Kern W, Braess J, Bottger B, Kaufmann CC, Hiddemann W, Schleyer E. Oxaliplatin pharmacokinetics during a four-hour infusion. Clin Cancer Res. 1999;5(4):761-5. Epub

1999/04/23.

43. Vabnick I, Trimmer JS, Schwarz TL, Levinson SR, Risal D, Shrager P. Dynamic potassium channel distributions during axonal development prevent aberrant firing patterns. J Neurosci. 1999;19(2):747-58. Epub 1999/01/09.

44. Rasband MN, Park EW, Zhen D, Arbuckle Ml, Poliak S, Peles E, et al. Clustering of neuronal potassium channels is independent of their interaction with PSD-95. J Cell Biol. 2002;159(4):663-72. Epub 2002/11/20.

45. Hart IK, Maddison P, Newsom-Davis J, Vincent A, Mills KR. Phenotypic variants of autoimmune peripheral nerve hyperexcitability. Brain. 2002;125(Pt 8): 1887-95. Epub 2002/07/24.

46. Kleopa KA, Elman LB, Lang B, Vincent A, Scherer SS. Neuromyotonia and limbic encephalitis sera target mature Shaker-type K+ channels: subunit specificity correlates with clinical manifestations. Brain. 2006;129(Pt 6): 1570-84. Epub 2006/04/15.

47. Van Poucke M, Vanhaesebrouck AE, Peelman LJ, Van Ham L. Experimental validation of in silico predicted KCNA1 , KCNA2, KCNA6 and KCNQ2 genes for association studies of peripheral nerve hyperexcitability syndrome in Jack Russell Terriers. Neuromuscul Disord. 2012;22(6):558-65.

48. Eng DL, Gordon TR, Kocsis JD, Waxman SG. Development of 4-AP and TEA sensitivities in mammalian myelinated nerve fibers. J Neurophysiol. 1988;60:2168-79. 49. Astion ML, Coles JA, Orkand RK, Abbott NJ. K+ accumulation in the space between giant axon and Schwann cell in the squid Alloteuthis. Effects of changes in osmolarity. Biophys J. 1988;53(2):281-5. Epub 1988/02/01. 50. Verselis VK, Srinivas M. Connexin channel modulators and their mechanisms of action. Neuropharmacology. 2013;75:517-24. Epub 2013/04/20.

51. Poyraz D, Brau ME, Wotka F, Puhlmann B, Scholz AM, Hempelmann G, et al. Lidocaine and octanol have different modes of action at tetrodotoxin-resistant Na(+) channels of peripheral nerves. Anesth Analg. 2003;97:1317-24.

52. Anzini P, Neuberg DH-H, Schachner M, Nelles E, Willecke K, Zielasek J, et al. Structural abnormalities and deficient maintenance of peripheral nerve myelin in mice lacking the gap junction protein connexin32. J Neurosci. 1997;17:4545-61.

53. Scherer SS, Xu Y-T, Nelles E, Fischbeck K, Willecke K, Bone LJ. Connexin32-null mice develop a demyelinating peripheral neuropathy. Glia. 1998;24:8-20.

54. Kleopa KA. The role of gap junctions in Charcot-Marie-Tooth disease. J Neurosci. 2011 ;31 :17753-60.

55. Sargiannidou I, Vavlitou N, Aristodemou S, Hadjisawas A, Kyriacou K, Scherer

55. et al. Connexin32 mutations cause loss of function in Schwann cells and oligodendrocytes leading to PNS and CNS myelination defects. J Neurosci. 2009;29:4748-

61.

56. Vavlitou N, Sargiannidou I, Markoullis K, Kyriacou K, Scherer SS, Kleopa KA. Axonal pathology precedes demyelination in a mouse model of X-linked demyelinating/type I Charcot-Marie Tooth neuropathy. J Neuropathol Exp Neurol. 2010;69:945-58.

57. Balice-Gordon RJ, Bone LJ, Scherer SS. Functional gap junctions in the Schwann cell myelin sheath. J Cell Biol. 1998;142:1095-104.

58. Menichella DM, Majdan M, Awatramani R, Goodenough DA, Sirkowski E, Scherer SS, et al. Genetic and physiological evidence that oligodendrocyte gap junctions contribute to spatial buffering of potassium released during neuronal activity. J Neurosci. 2006;26(43): 10984-91. Epub 2006/10/27.

59. Kume-Kick J, Mazel T, Vorisek I, Hrabetova S, Tao L, Nicholson C. Independence of extracellular tortuosity and volume fraction during osmotic challenge in rat neocortex. J Physiol. 2002;542(Pt 2):515-27. Epub 2002/07/18.

60. Lev-Ram V, Ellisman MH. Axonal activation-induced calcium transients in myelinating Schwann cells, sources, and mechanisms. J Neurosci. 1995;15(4):2628-37. Epub 1995/04/01.

61. Paul DL, Ebihara L, Takemoto LJ, Swenson Kl, Goodenough DA. Voltage gating and permeation in a gap junction hemichannel J Cell Biol. 1991 ;115:1077-89.

62. Valiunas V, Weingart R. Electrical properties of gap junction hemichannels identified in transfected HeLa cells. Pflugers Arch. 2000;440(3): 366-79. Epub 2000/08/23. 63. Kang J, Kang N, Lovatt D, Torres A, Zhao Z, Lin J, et al. Connexin 43 hemichannels are permeable to ATP. J Neurosci. 2008;28(18):4702-11. Epub 2008/05/02.

64. Fasciani I, Temperan A, Perez-Atencio LF, Escudero A, Martinez-Montero P, Molano J, et al. Regulation of connexin hemichannel activity by membrane potential and the extracellular calcium in health and disease. Neuropharmacology. 2013;75:479-90. Epub 2013/04/17.

65. Trexler EB, Bennett MV, Bargiello TA, Verselis VK. Voltage gating and permeation in a gap junction hemichannel. Proc Natl Acad Sci U S A. 1996;93(12):5836-41. Epub 1996/06/11.

66. Hargittai PT, Youmans SJ, Lieberman EM. Determination of the membrane potential of cultured mammalian Schwann cells and its sensitivity to potassium using a thiocarbocyanine fluorescent dye. Glia. 1991 ;4:611-6.

67. Gomes P, Srinivas SP, Van Driessche W, Vereecke J, Himpens B. ATP release through connexin hemichannels in corneal endothelial cells. Invest Ophthalmol Vis Sci. 2005;46(4): 1208-18. Epub 2005/03/26.

68. Contreras JE, Saez JC, Bukauskas FF, Bennett MV. Gating and regulation of connexin 43 (Cx43) hemichannels. Proc Natl Acad Sci U S A. 2003;100:11388-93.

69. Retamal MA, Schalper KA, Shoji KF, Bennett MV, Saez JC. Opening of connexin 43 hemichannels is increased by lowering intracellular redox potential. Proc Natl Acad Sci U S A. 2007; 104(20):8322-7. Epub 2007/05/15.

70. Retamal MA, Froger N, Palacios-Prado N, Ezan P, Saez PJ, Saez JC, et al. Cx43 hemichannels and gap junction channels in astrocytes are regulated oppositely by proinflammatory cytokines released from activated microglia. J Neurosci. 2007;27(50): 13781 -92. Epub 2007/12/14.

71. Orellana JA, Hernandez DE, Ezan P, Velarde V, Bennett MV, Giaume C, et al. Hypoxia in high glucose followed by reoxygenation in normal glucose reduces the viability of cortical astrocytes through increased permeability of connexin 43 hemichannels. Glia. 2010;58:329-43.

72. Ye ZC, Wyeth MS, Baltan-Tekkok S, Ransom BR. Functional hemichannels in astrocytes: a novel mechanism of glutamate release. J Neurosci. 2003;23:3588-96.

73. Quist AP, Rhee SK, Lin H, Lai R. Physiological role of gap-junctional hemichannels. Extracellular calcium-dependent isosmotic volume regulation. J Cell Biol. 2000; 148: 1063- 74.

74. Baker MD, Ritchie JM. Characteristics of type I and type II K⁺ channels in rabbit cultured Schwann cells. J Physiol. 1996;490:79-95. 75. Spong KE, Robertson RM. Pharmacological blockade of gap junctions induces repetitive surging of extracellular potassium within the locust CNS. J Insect Physiol. 2013;59(10):1031-40. Epub 2013/08/07.

76. Rash JE. Molecular disruptions of the panglial syncytium block potassium siphoning and axonal saltatory conduction: pertinence to neuromyelitis optica and other demyelinating diseases of the central nervous system. Neuroscience. 2010;168(4):982- 1008. Epub 2009/10/24.

77. Lehky TJ, Leonard GD, Wilson RH, Grem JL, Floeter MK. Oxaliplatin-induced neurotoxicity: acute hyperexcitability and chronic neuropathy. Muscle Nerve. 2004;29:387-92

78. Poliak S, Salomon D, Elhanany H, Sabanay H, Kiernan B, Pevny L, et al. Juxtaparanodal clustering of Shaker-like K+ channels in myelinated axons depends on Caspr2 and TAG-1. J Cell Biol. 2003; 162(6): 1149-60. Epub 2003/09/10.

79. Traka M, Goutebroze L, Denisenko N, Bessa M, Nifli A, Havaki S, et al. Association of TAG-1 with Caspr2 is essential for the molecular organization of juxtaparanodal regions of myelinated fibers. J Cell Biol. 2003; 162(6): 1161-72. Epub 2003/09/17.

80. Boyle MET, Berglund EO, Murai KK, Weber L, Peles E, Ranscht B. Contactin Orchestrates Assembly of the Septate-like Junctions at the Paranode in Myelinated Peripheral Nerve Neuron. 2001;30:385-97

81. Ushio S, Egashira N, Sada H, Kawashiri T, Shirahama M, Masuguchi K, et al. Goshajinkigan reduces oxaliplatin-induced peripheral neuropathy without affecting anti- tumour efficacy in rodents. Eur J Cancer. 2012;48(9):1407-13. Epub 2011/09/13.

82. Aoki M, Mori A, Nakahara T, Sakamoto K, Ishii K. Effect of synthetic eel calcitonin, elcatonin, on cold and mechanical allodynia induced by oxaliplatin and paclitaxel in rats. Eur J Pharmacol. 2012;696(1 -3):62-9. Epub 2012/09/25.

83. Kawashiri T, Egashira N, Watanabe H, Ikegami Y, Hirakawa S, Mihara Y, et al. Prevention of oxaliplatin-induced mechanical allodynia and neurodegeneration by neurotropin in the rat model. Eur J Pain. 2011 ;15(4):344-50. Epub 2010/09/11.

84. Di Cesare Mannelli LZ, M; Failli, P; Ghelardini C. Oxaliplatin-induced neuropathy: oxidative stress as pathological mechanism. Protective effect of silibinin. J Pain.

2012;13(3):276-84. doi: 10.1016/j.jpain.2011.11.009. Epub 2 Feb 9.

85. Boyette-Davis J, Dougherty PM. Protection against oxaliplatin-induced mechanical hyperalgesia and intraepidermal nerve fiber loss by minocycline. Exp Neurol. 2011 ;229(2):353-7. Epub 2011/03/10.

86. Chen XF, Wang R, Yin YM, Roe OD, Li J, Zhu LJ, et al. The effect of monosialotetrahexosylganglioside (GM1) in prevention of oxaliplatin induced neurotoxicity: a retrospective study. Biomed Pharmacother. 2012;66(4):279-84. Epub 2012/03/09. 87. Flatters SJ, Bennett GJ. Ethosuximide reverses paclitaxel- and vincristine-induced painful peripheral neuropathy. Pain. 2004;109(1-2):150-61. Epub 2004/04/15.

88. Kawashiri T, Egashira N, Itoh Y, Shimazoe T, Ikegami Y, Yano T, et al. Neurotropin reverses paclitaxel-induced neuropathy without affecting anti-tumour efficacy. Eur J Cancer. 2009;45(1): 154-63. Epub 2008/11/26.

89. Bushara KO, Goldstein SR, Grimes GJ, Jr., Burstein AH, Hallett M. Pilot trial of 1- octanol in essential tremor. Neurology. 2004;62(1): 122-4. Epub 2004/01/14.

90. Haubenberger D, Nahab FB, Voller B, Hallet M. Treatment of Essential Tremor with Long-Chain Alcohols: Still Experimental or Ready for Prime Time? Tremor Other Hyperkinet Mov. 2014;doi: 10.7916/D8RX991 R.

91. Beaumont M, Maccaferri G. Is connexin36 critical for GABAergic hypersynchronization in the hippocampus? J Physiol. 2011 ;589(Pt 7):1663-80. Epub 2011/02/09.

92. Tovar KR, Maher BJ, Westbrook GL Direct actions of carbenoxolone on synaptic transmission and neuronal membrane properties. J Neurophysiol. 2009;102(2):974-8.

Epub 2009/06/19.

93. Bruzzone R, Barbe MT, Jakob NJ, Monyer H. Pharmacological properties of homomeric and heteromeric pannexin hemichannels expressed in Xenopus oocytes. J Neurochem. 2005;92(5): 1033-43. Epub 2005/02/18.

94. Suadicani SO, Brosnan CF, Scemes E. P2X7 receptors mediate ATP release and amplification of astrocytic intercellular Ca2+ signaling. J Neurosci. 2006;26(5): 1378-85. Epub 2006/02/03.

Claims

1. A gap junction blocker for use in preventing or treating oxaliplatin-induced neuropathy in a subject who is, has been, or is going to be, administered oxaliplatin.

2. Oxaliplatin for use in treating a subject in need of oxaliplatin therapy, wherein the subject is, has been, or is going to be, administered a gap junction blocker.

3. Use of a gap junction blocker in the manufacture of a medicament for use in treating or preventing oxaliplatin-induced neurotoxicity in a subject.

4. A composition comprising a gap junction blocker and oxaliplatin, and optionally one or more further therapeutic agents.

5. A composition comprising a gap junction blocker and oxaliplatin for use in treating a subject in need of oxaliplatin therapy.

6. A method of treating or preventing oxaliplatin-induced neurotoxicity in a subject wherein said method comprises administering a gap junction blocker.

7. A kit of parts comprising a gap junction blocker and oxaliplatin, optionally wherein the kit of parts further comprises one or more further therapeutic agents.

8. The gap junction blocker of claim 1 , oxaliplatin of claim 2, use of claim 3, composition according to any of claims 4 and 5, method according to claim 6 or the kit according to claim 7, wherein the gap junction blocker closes CX31.3 hemichannels or Cx32 GJ channels, optionally wherein the gap junction blocker specifically closes CX31.3 hemichannels and/or Cx32 GJ channels.

9. The gap junction blocker of claim 1 , oxaliplatin of claim 2, use of claim 3, composition according to any of claims 4 and 5, method according to claim 6 or the kit according to claim 7, wherein the gap junction blocker is a small molecule.

10. The gap junction blocker of claim 1 , oxaliplatin of claim 2, use of claim 3, composition according to any of claims 4 and 5, method according to claim 6 or the kit according to claim 7, wherein the gap junction blocker is a peptide, optionally an antibody or mimetic peptide.

11. The gap junction blocker of claim 1 , oxaliplatin of claim 2, use of claim 3, composition according to any of claims 4 and 5, method according to claim 6 or the kit according to claim 7, where in the gap junction blocker is octanol, or a metabolite or prodrug of octanol, optionally octanoic acid.

12. The gap junction blocker of claim 1 , oxaliplatin of claim 2, use of claim 3, composition according to any of claims 4 and 5, method according to claim 6 or the kit according to claim 7, wherein the gap junction blocker is administered at a dosage of from 1 to 64mg/Kg, optionally wherein the gap junction blocker is co-administered with the oxaliplatin.

13. The gap junction blocker of claim 1 , oxaliplatin of claim 2, use of claim 3, composition according to any of claims 4 and 5, method according to claim 6 or the kit according to claim 7, wherein the subject is a mammal, optionally a human, a dog, a cat, a mouse, rodent, optionally a mouse or rat, a horse, a farmed animal.

14. The gap junction blocker of claim 1 , oxaliplatin of claim 2, use of claim 3, composition according to any of claims 4 and 5, method according to claim 6 or the kit according to claim 7 wherein the subject has cancer, optionally gastrointestinal cancer, optionally metastatic colorectal cancer.

15. The gap junction blocker of claim 1 , oxaliplatin of claim 2, use of claim 3, composition according to any of claims 4 and 5, method according to claim 6 or the kit according to claim 7, wherein the gap junction blocker acts on the peripheral nerve.

16. The gap junction blocker of claim 1, oxaliplatin of claim 2, use of claim 3, composition according to any of claims 4 and 5, method according to claim 6 or the kit according to claim 7 wherein the gap junction blocker prevents the increase in duration of the compound action potential caused by oxaliplatin by increased closure of Cx29/Cx31.3 gap junction hemichannels or Cx32 GJ channels, and also prevents the uptake of gap junction permeant substance neurobiotin into cells through closure of Cx29/Cx31.3 hemichannels or Cx32 GJ channels.