Orally Active Neurotrophin-Enhancing Agent Protects Against Dysfunctions of the Peripheral Nerves in Hyperglycemic Animals

All animal experiments in these studies were performed in accordance with the guidelines of Mitsubishi Pharma and were approved by the animal investigation committee of our institute. In the experiment with rats, diabetes was induced by intravenous injection of STZ (50 mg/kg, dissolved in 0.01 mol/l citrate buffer at pH 5.5) in 12- to 14-week-old male Wister rats (Japan Laboratory Animals, Tokyo, Japan). An additional group of rats was injected with citrate buffer to act as a sham control. Hyperglycemia was verified 4 weeks after STZ injection, and the animals were divided into six groups (sham-vehicle, diabetic-vehicle, and 0.1, 0.3, 1, or 3 mg/kg MCC-257 daily), and then a 4-week treatment with MCC-257 was started. The number of animals in each group was 10, except for 9 in the 0.1-mg/kg group in the nerve conduction study and 8 in the 0.1-mg/kg group in the study of NGF content. In the experiment with mice, male BKS.Cg-+Lepr^db/+Lepr^db/Jcl (db/db) and BKS.Cg-m +/+ Lepr^db/Jcl (db/+m) were used (CLEA Japan, Tokyo, Japan). At 9 weeks of age, body weight, value in a hot plate test (predosing), and plasma glucose level were measured for all animals before the beginning of drug administration. The db/db mice with plasma glucose levels >250 mg/100 ml were used as diabetic animals. Diabetic db/db mice were divided into three experimental groups (diabetic-vehicle and 3 or 10 mg/kg MCC-257 daily) of 20 animals, and then a 7-week treatment with MCC-257 was started.

MCC-257 [N-(5α-cholestan-3α-yl)-5-acetamido-3,5-dideoxy-2-O-methyl-α-d-glycero-d-galacto-2-nonulopyranosonamide, solid dispersion form 1:3 = MCC-257, polymer] was synthesized and manufactured by Mitsubishi Pharma. The compound was suspended in purified water and administered orally. Vehicle-treated diabetic animals were administered purified water. Normoglycemic animals (sham or db/+m) were administered purified water.

Dorsal root ganglion neurons prepared from 14-day-old Wistar-ST rats were grown as dissociated cultures as previously described (29) with minor modification. These cells were plated onto poly-l-lysine–coated culture dishes in serum-free Ham’s F12 medium supplemented with 5 μg/ml insulin, 5 μg/ml transferrin, 20 nmol/l progesterone, 0.1 mmol/l putrescine, 30 nmol/l sodium selenite, 50 units/ml penicillin, and 50 units/ml streptomycin. Cytosine arabionoside (100 μmol/l) was added to the culture medium from day 1 in culture for 24 h. Treatment with MCC-257 in the presence or absence of NGF (1 ng/ml; Toyobo, Osaka, Japan) was started immediately after plating. As a positive control, the other wells were treated with 100 ng/ml NGF.

Pheochromocytoma 12 (PC12) cells (American Type Culture Collection) were cultured in RPMI 1640 medium supplemented with 10% horse serum, 5% fetal bovine serum, 50 unit/ml penicillin G, and 50 μg/ml streptomycin sulfate in culture dishes. After the medium was exchanged for a serum-free RPMI medium, treatment of cultured cells with different doses of MCC-257 without or with NGF (50 ng/ml) was started immediately.

Dorsal root ganglia or PC12 cells were tested at 2 days in culture. 3-(4,5-Dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide solution (5 mg/ml in PBS) was added to each well, and the plates were incubated at 37°C for 4 h in a humidified 5% CO₂ atmosphere. After the incubation, a stop/solubilization solution consisting of 20% SDS in 50% dimethylformamide/50% distilled water was added. At 1 h after the addition of the stop/solubilization solution, the contents of the wells were mixed, and the absorbance at the 570-nm wavelength was measured.

When PC12 cells were near confluent, the cells were preincubated with serum-free Dulbecco’s modified Eagle’s medium overnight and stimulated for 5 min at 37°C with MCC-257 in the presence or the absence of NGF (50 ng/ml). The cells were lysed with lysis buffer (20 mmol/l Tris-HCl, pH 7.5, 150 mmol/l NaCl, 10 mmol/l NaF, 5 mmol/l EDTA, 2 mmol/l Na₃VO₄, 80 units/ml aprotinin, and 1 mmol/l phenylmethylsulfonyl fluoride). The protein levels in the lysates were measured and equalized accordingly. Immunoprecipitation was performed with a Trk antibody (C-14; Santa Cruz Biotechnology, Santa Cruz, CA). The immunoprecipitates were washed with lysis buffer and then resuspended in Laemmli-SDS sample buffer (30) and boiled for 5 min. The samples and several doses of the positive control (epidermal growth factor–stimulated A431 cell lysate; Invitrogen, Charlottesville, VA) were subjected to SDS-PAGE and transferred to polyvinylidine difluoride membrane (Millipore, Billerica, MA). The membranes were probed with anti-phosphotyrosine antibody (4G10; Upstate Biotechnology). TrkA tyrosine phosphorylation on the membrane was detected by an enhanced chemiluminescent detection system (Amersham Biosciences, Piscataway, NJ). The densities of blots for the samples were measured with a densitometer (Personal Densitometer; Molecular Dynamics, Sunnyvale, CA). To compare the measured value of the samples among the trials, the amount of tyrosine phosphorylation of TrkA for an individual sample (namely, the density of each blot) was converted into the sample buffer volume of the positive control (μl) that was loaded on the same electrophoretic gel.

At the end of the treatment period, the rats were anesthetized with pentobarbital sodium (50 mg/kg i.p.; Abbott Laboratories). Rectal temperature was monitored during the procedure, and the rats were placed on a heating mat to diminish the hypothermic effects of anesthesia. The left sciatic nerve was stimulated at the sciatic notch and then at the Achilles tendon. Stimulations were delivered via concentric bipolar needle electrodes with a Synax 1100 (NEC San-ei Instruments, Tokyo, Japan). The M waves and H-reflexes to each stimulus were recorded via a needle electrode from the second interosseous muscle using Synax 1100. The temporal separation of the peaks of the M waves and H-reflexes was calculated with a Synax 1100. Nerve length from the sciatic notch to the Achilles tendon was measured and used to calculate motor nerve conduction velocity (MCV) and sensory nerve conduction velocity (SCV).

At the end of the treatment period, all of the rats were killed, and the sciatic nerves (both sides of each animal) were dissected. NGF was extracted by repetitive freezing-thawing cycles and soaked in washing buffer (0.1 mol/l Tris-HCl buffer containing 0.4 mol/l NaCl, 0.02% NaN₃, 0.1% BSA, and 1 mmol/l MgCl₂). NGF was measured by the enzyme immunoassay system as follows. First, 96-well immunoplates were coated with anti-mouse NGF monoclonal antibody or anti-mouse digoxigenin antibody (Roche Diagnostics, Penzberg, Germany) and then washed and blocked. After a 5-h incubation of the extracted samples or NGF standards, the plates were washed, and β-galactosidase–conjugated anti-NGF antibody (0.1 unit/ml in washing buffer; Roche Diagnostics) was added. After overnight incubation and washing, a substrate for the enzyme (4-methyumbelliferyl-β-d-galactoside, 20 μg/ml in washing buffer; Sigma) was then added to the plates, which was incubated for 5 h at room temperature in the dark. The reaction was stopped with 0.1 mol/l glycine-NaOH buffer (pH 10.3). Fluorescence intensity was measured at two wavelengths of 355 and 460 nm.

A hot plate test was performed before and after drug treatment. Each animal was habituated to the test apparatus (MK-350A; Muromachi Kikai, Tokyo) before the nociception assay. The mice were placed on the hot plate maintained at 53°C, and the latency to response (licking of front or hind paw) was recorded. If the response did not occur within 45 s, the animal was taken out, and the result was recorded as 45 s.

On the day after the hot plate test, which followed the treatment period, all of the mice were anesthetized with diethyl ether and decapitated, and the plantar skins were dissected bilaterally. Left planter skins were stored at −80°C as the samples for biochemical evaluation. Right plantar skins were fixed in 10% neutral buffered formalin. The sectioned skin was embedded in paraffin and cut into ∼5-μm-thick sections for histopathological evaluation.

The tissues were boiled for 15 min in 0.5 ml of boiling 2 mol/l acetic acid containing 10 mmol/l hydrochloride, 1 mmol/l EDTA, and 1 mmol/l dithiothreitol. After boiling, the tissues were homogenized, the remaining solution was centrifuged (9,000g for 5 min), and the supernatant was freeze dried overnight. Substance P was determined by enzyme-linked immunosorbent assay, using commercially available antisera (Peninsula Laboratories, San Carlos, CA).

Paraffin blocks of plantar skins were stained for protein gene product 9.5 (PGP9.5) by immunohistochemistry. After deparaffinization and rehydration, each section was blocked with 1% BSA. The sections were incubated with the polyclonal primary antibody PGP9.5 (1:1,000; Ultraclone, Isle of Wight, U.K.) for 60 min at 37°C. The sections were then incubated with the biotinylated link secondary antibody (Vector Laboratories, Burlingame, CA) for 30 min and peroxidase-labeled streptavidin (Vector Laboratories) according to the manufacturer’s instructions. 3′5′ diaminobenzidine was used as the substrate-chromogen solution. Counterstaining was performed with hematoxylin.

The tissue sections stained with PGP9.5 were observed by optical microscopy (AX-80; Olympus, Tokyo, Japan). Digital images were captured with a charged-coupled device camera (DP70, Olympus). Image acquisition was controlled through DP control (Olympus) running on a computer. The digital images were saved as TIFF files. The length of the epidermal/dermal junction was measured as skin length, using image analysis software (Image-Pro Plus, version 4.5; Media Cybernetics, Silver Spring, MD). Every PGP9.5-positive nerve fiber in a target field (dermis within a 50-μm radius from the epidermal/dermal junction) was traced, and its area was measured using Image-Pro plus. The number of PGP9.5-immunoreactive nerve fibers with a cross-sectional area of ≤20 μm² was counted, and it was normalized with skin length (epidermal/dermal junction), using the following: nerve fibers (≤20 μm²) per unit skin length (counts/mm) = PGP9.5-immunoreactive nerve fibers (≤20 μm²) (counts)/skin length (μm) × 1,000.

Unless otherwise noted, all data are the means ± SE. Differences among experimental group were detected by ANOVA, and the significance of differences between groups was assessed by paired or unpaired t test or Dunnett’s multiple comparison. Significance was defined as P < 0.05.

A high concentration of NGF (100 ng/ml) had a protective effect on the survival of dorsal root ganglion neurons compared with nontreated cells (P < 0.01) (Fig. 1A). MCC-257 (0.1, 0.3 μmol/l; P < 0.01, respectively) significantly increased the cell viability of cultured dorsal root ganglion neurons in the presence of a low concentration of NGF (1 ng/ml) (Fig. 1B), whereas in the absence of NGF in the media, MCC-257 had no effect on the viability of dorsal root ganglion neurons up to 0.3 μmol/l. MCC-257 significantly deceased the viability of dorsal root ganglion neurons at a concentration of 1 μmol/l because of its insolubility (P < 0.05) (Fig. 1A).

Exposure to NGF (50 ng/ml) for 5 min significantly increased the tyrosine phosphorylation of TrkA in PC12 cells compared with nontreated cells (P < 0.01) (Fig. 2A). The exposure of PC12 cells to MCC-257 in the presence of NGF (50 ng/ml) significantly enhanced TrkA autophosphorylation (0.1 μmol/l, P < 0.01) (Fig. 2A), whereas in the absence of NGF, treatment with MCC-257 had no effect on the tyrosine phosphorylation of TrkA (Fig. 2A). MCC-257 (0.1 μmol/l, P < 0.01) significantly increased the cell viability of serum-deprived PC12 cells in the presence of NGF (50 ng/ml). In the absence of NGF in the media, MCC-257 had no effect on PC12 cell viabilities up to 0.1 μmol/l (Fig. 2B).

The MCV and SCV of 8-week hyperglycemic rats were reduced compared with those of animals in the sham group by 7.2 and 10.9 m/s, respectively (MCV P < 0.05, SCV P < 0.05) (Table 1). These impairments were significantly improved by 4 weeks of treatment with MCC-257 (MCV: 3 mg/kg orally, P < 0.05; SCV: 0.3, 1, and 3 mg/kg orally, P < 0.05, respectively) (Table 1).

The effect of MCC-257 on NGF content in the sciatic nerves was also examined in diabetic rats. Compared with sham-treated rats, the NGF content in the sciatic nerves of hyperglycemic rats (vehicle group) was significantly decreased (P < 0.01) (Table 1). MCC-257 (1 and 3 mg/kg orally, P < 0.05) significantly increased NGF content in the sciatic nerves in STZ-induced hyperglycemic rats compared with untreated hyperglycemic rats (Table 1).

Compared with nondiabetic db/+m mice, the licking latency of db/db mice was significantly longer after 7 weeks of drug administration (10.26 s in db/+m vs. 18.13 s in db/db vehicle, P < 0.01) (Fig. 3), suggesting that the db/db mice showed thermal hypoalgesia. Compared with the db/db vehicle-treated group, the licking latency of the db/db MCC-257–treated group (10 mg/kg orally) was significantly shorter after 7 weeks of administration (18.13 s in db/db vehicle vs. 14.77 s in db/db MCC-257 10 mg/kg orally, P < 0.05) (Fig. 3), which suggests that MCC-257 has a protective effect on thermal nociceptive function.

The substance P level was measured in plantar skins obtained from the animals used in the behavioral test. The substance P level was significantly reduced in the skin from db/db mice compared with that of db/+m mice (P < 0.01) (Fig. 4). Treatment with MCC-257 for 7 weeks significantly elevated substance P level at doses of 3 and 10 mg/kg orally (P < 0.01, respectively) (Fig. 4). This result suggests that MCC-257 protects small fiber innervations in diabetic plantar skins.

The nerve fiber density (number of PGP9.5-positive nerve fibers that have an area <20 μm² per unit length of skin) for db/+m mice was 52.8 ± 2.1 counts per skin length. Reduced cutaneous innervation was observed in the plantar skin of db/db mice compared with that of db/+m mice (38.7 ± 1.9 counts per skin length in db/db vehicle, P < 0.01) (Fig. 5D). This reduction was attenuated by a 7-week treatment with MCC-257 at doses of 3 and 10 mg/kg orally (P < 0.05 and P < 0.01, respectively) (Fig. 5D). This result is consistent with results obtained from behavioral and biochemical evaluations.

We have developed a novel compound, MCC-257, that potentiates the action of neurotrophins on central and peripheral neurons, which leads to the protection of neuronal cells from some fatal conditions. Furthermore, the compound, a sialic acid derivative, exerts protective effects on both the sciatic nerve and the primary afferent nerve fibers in diabetic animals. The compound prevented neuronal cell death only in the presence of a small amount of NGF, whereas this effect was not elicited when NGF was not provided to the neuronal cells. These findings suggest that the compound does not in itself have a neurotrophic effect but enhances NGF signal transduction. Mutoh et al. (21) reported that GM1 preferentially activates the TrkA receptor. In that report GM1 was tightly associated with TrkA and increased the NGF-induced autophosphorylation of TrkA. Because MCC-257 was originally synthesized as a GM1 mimetic, direct or indirect association of MCC-257 with Trk proteins may result in the stimulation of NGF-induced TrkA phosphorylation and enhancement of neurotrophin activity. Moreover, because the compound activated the autophosphorylation of TrkB in a primary culture of embryonic rat cerebral cortex neurons (data not shown), it may have broad activity on Trk proteins. It is hypothesized that the cells that can produce neurotrophins and the cells expressing TrkA or TrkB may be the main targets of this compound. The compound also increased NGF content in the sciatic nerve of diabetic rats. This result suggests that the compound facilitated NGF production in the peripheral tissues and/or Schwann cells and improved the availability of NGF.

In addition, the compound protected against impairment of nerve conduction velocity in STZ-induced diabetic rats. Nerve conduction velocity is an indicator of large fiber function. Because neurotrophin-3 is trophic for large-caliber sensory neurons and motor neurons (31,32), the effect of this compound on nerve conduction velocity may be partially attributable to the enhancement of neurotrophin-3 production or autophosphorylation of TrkC. The effect of MCC-257 on TrkC is also under investigation.

From a clinical point of view, it is especially noteworthy that the compound ameliorated small-caliber sensory nerve function and its cutaneous innervation. This is the first report that a nonpeptide small molecular compound was able to improve sensory dysfunction and cutaneous innervation in diabetic animals. These results of behavioral, biochemical, and histopathological studies were consistent with each other. Because thermal nociceptive stimuli are mainly transmitted through unmyelinated C-fiber afferents and NGF has trophic effects on small fibers, MCC-257 acted on small fibers in experimental thermal hypoalgesia. Substance P content in peripheral skins could be a good marker of the presence of active small fibers or cutaneous innervation because a population of small fibers expresses substance P (33). The elevation of substance P level in diabetic skins seems to be responsible for the amelioration of thermal sensory functions. In the biochemical study, however, we could not deny the possibility that MCC-257–induced NGF increased substance P expression independently of the cutaneous innervation. Hence, we conducted a histopathological quantification of cutaneous nerves with PGP9.5 antibody, which is generally used for the evaluation of clinical and experimental cutaneous innervation and is well characterized as a pan-neuronal marker. Taking this into account, for quantitative analysis, we focused on nerve fibers with a cross-sectional area <20 μm². Although nerve fibers whose cross-sectional area is <20 μm² are not always small fibers, this compound increased smaller-caliber nerve fibers, including small fibers.

Recent studies have suggested that reduction in availability of neurotrophic factors also contributes to diabetic peripheral neuropathy (34). However, the efficacy of NGF or GM1 was mixed in clinical trials for the treatment of diabetic peripheral neuropathy. One of the reasons that NGF or GM1 did not show efficacy in the latest clinical trails is their pharmacokinetic properties. Because NGF is a polypeptide, the route of administration was restricted to subcutaneous injection. Therefore, the side effects associated with direct action on mast cells (35) in the skin or potentiation of bradykinin and prostaglandin E₂ activities (36) were inevitable. Mixed gangliosides, including GM1, were also given parenterally for clinical trials because of their poor absorption. Because MCC-257 is a nonpeptide small molecule potentiating the action of neurotrophins, we can adopt oral administration, which may prevent local adverse effects related to the injection. Once this pharmacokinetic problem has been overcome, supplemental therapy with neurotrophins may be a realistic option in treating diabetic peripheral neuropathy. The pathogenesis of diabetic peripheral neuropathy seems to be composed of multiple factors, including a disorder of the polyol pathway, oxidative stress, and the abnormal activation of PKC, and therefore aldose reductase inhibitors alone or PKC inhibitors alone may be less effective against diabetic peripheral neuropathy. For this reason, supplemental therapy is a meaningful option for the treatment of diabetic peripheral neuropathy.

Our results showed that a new sialic acid derivative, MCC-257, enhanced signal transduction of NGF. Moreover, this compound improved both nerve conduction velocity and hot plate response accompanied by cutaneous innervation in diabetic animals. These results indicate that MCC-257 may have a beneficial effect on the clinical treatment of diabetic peripheral neuropathy. Further investigation of this compound as an agent for diabetic peripheral neuropathy is underway.