ORIGINAL RESEARCH

Dual Effect of Methylglyoxal on the Intracellular Ca2+ Signalingand Neurite Outgrowth in Mouse Sensory Neurons

Beatrice Mihaela Radu • Diana Ionela Dumitrescu •

Cosmin Catalin Mustaciosu • Mihai Radu

Received: 19 January 2012 / Accepted: 21 February 2012 / Published online: 9 March 2012

� Springer Science+Business Media, LLC 2012

Abstract The formation of advanced glycation end prod-

ucts is one of the major factors involved in diabetic neu-

ropathy, aging, and neurodegenerative diseases. Reactive

carbonyl compounds, such as methylglyoxal (MG), play a

key role in cross-linking to various proteins in the extracel-

lular matrix, especially in neurons, which have a high rate of

oxidative metabolism. The MG effect was tested on dorsal

root ganglia primary neurons in cultures from adult male

Balb/c mice. Lower MG doses contribute to an increased

adherence of neurons on their support and an increased glia

proliferation, as proved by MTS assay and bright-field

microscopy. Time-lapse fluorescence microscopy by Fura-2

was performed for monitoring the relative fluorescence ratio

changes (DR/R0) upon depolarization and immunofluores-

cence staining for quantifying the degree of neurites exten-

sion. The relative change in fluorescence ratio modifies

the amplitude and dispersion depending on the subtype of

sensory neurons, the medium-sized neurons are more sen-

sitive to MG treatment when compared to small ones. Low

MG concentrations (0–150 lM) increase neuronal viability,

excitability, and the capacity of neurite extension, while

higher concentrations (250–750 lM) are cytotoxic in a dose-

dependent manner. In our opinion, MG could be metabo-

lized by the glyoxalase system inside sensory neurons up to a

threshold concentration, afterwards disturbing the cell

equilibrium. Our study points out that MG has a dual effect

concentration dependent on the neuronal viability, excit-

ability, and neurite outgrowth, but only the excitability

changes are soma-sized dependent. In conclusion, our data

may partially explain the distinct neuronal modifications in

various neurodegenerative pathologies.

Keywords Methylglyoxal � Viability � Time-lapse

fluorescence microscopy � Neurite outgrowth �Peripheral sensory neurons

Introduction

Dorsal root ganglia (DRG) neurons play a major role in

conveying somatic and visceral sensory information from

peripheral tissues to the spinal cord. A large variety of cells

reside in mammalian DRGs, such as pseudo-unipolar neu-

rons, myelinated, and non-myelinated fibers, macrophages,

fibroblasts, and satellite perineuronal cells (Nascimento et al.

2008). Mouse sensory neurons have a wide heterogeneity,

their cross-sectional area varying between small neurons

(\600 lm2), medium neurons (60071,200 lm2), and large

neurons ([1,200 lm2) (Ren et al. 2012). More than 98% of

the mouse DRG neurons maintained in primary cell culture

for 48 h are either small- or medium-sized neurons that are

Beatrice Mihaela Radu and Diana Ionela Dumitrescu contributed

equally to this article.

B. M. Radu � M. Radu (&)

Department of Neurological, Neuropsychological,

Morphological and Movement Sciences, Section of Anatomy

and Histology, University of Verona, Strada Le Grazie 8,

37134 Verona, Italy

e-mail: mradu@nipne.ro

B. M. Radu � D. I. Dumitrescu

Department of Anatomy, Animal Physiology and Biophysics,

Faculty of Biology, University of Bucharest, Splaiul

Independentei 91-95, 050095 Bucharest, Romania

C. C. Mustaciosu � M. Radu

Department of Life and Environmental Physics, Horia Hulubei

National Institute for Physics and Nuclear Engineering,

Reactorului 30, P.O. Box MG-6, 077125 Bucharest-Magurele,

Romania

123

Cell Mol Neurobiol (2012) 32:1047–1057

DOI 10.1007/s10571-012-9823-5

immunoreactive for substance P and calcitonin gene-related

peptide (CGRP) (Hiruma et al. 2000). Many mouse DRG

neuronal sub-populations have been identified on the basis of

their neurochemical, immunocytochemical, electrical, and

functional properties, or considering different sensitivities to

pharmacological agents (Hiruma et al. 2000; Ren et al.

2012).

Unlike the central nervous system, DRG are not pro-

tected by a blood–nerve barrier, and are consequently

vulnerable to metabolic and toxic injury (Jimenez-Andrade

et al. 2008). Therefore, DRG neurons are one of the main

targets in diabetes, their morphological and functional

changes contributing to diabetic neuropathy (Toth et al.

2004). Multiple abnormalities due to the diabetic condition

have been described in DRGs, such as excessive polyol

flux, microangiopathy, oxidative stress, altered neuron

phenotype, mitochondrial dysfunction, ion channel altera-

tions, and abnormal growth factor signaling (Toth et al.

2004). DRG neurons are sensitive targets in oxidation-

mediated cytotoxic processes (Naziroglu 2007; Ibi et al.

2008; Lupachyk et al. 2011).

The abnormal glucose homeostasis in diabetes due to the

formation of highly reactive dicarbonyl metabolite meth-

ylglyoxal (MG) (Thornalley 2005; Fleming et al. 2011)

may be the key step in triggering the DRG dysfunctions.

By means of Maillard reaction, MG is able to cross-link

with extracellular proteins on targeted amino acids (argi-

nine, lysine), leading to the formation of advanced glyca-

tion end-products (AGEs), and thus contributing to aging

and complications in chronic diseases (Thornalley 2005;

Fleming et al. 2011).

MG may contribute to neurodegeneration as mediator of

oxidative stress (Kikuchi et al. 1999; Amicarelli et al.

2003). High MG levels have been found in the cerebro-

spinal fluid of patients affected by Alzheimer’s disease and

in the plasma of diabetic individuals (Han et al. 2007; Lu

et al. 2011).

MG is detoxified by glyoxalase I (GLO-I) and glyoxa-

lase II (GLO-II) and deficiency of the rate-limiting

GLO-I determines the accumulation of MG-derived AGEs

(Thornalley 2005). MG causes strong weakening of detoxi-

fying capacity and induces the apoptotic cell death in rat

hippocampal neurons, through the impairment of detoxifi-

cation pathway and depletion of reduced glutathione (Di

Loreto et al. 2008). GLO-I abundance varies between dif-

ferent murine strains and within different DRG sensory

neuron populations (Jack et al. 2011). Sensory neurons

exhibit low GLO-I activity compared to astrocytes (Belanger

et al. 2011). Moreover, MG impairs glucose metabolism and

leads to energy depletion in SH-SY5Y neuroblastoma cells

(de Arriba et al. 2007).

Despite the extensive literature describing MG formation

and detoxification, little is known about the MG-induced

changes on viability, excitability, and neurite outgrowth of

different mouse DRG neuron subtypes.

Materials and Methods

Solutions and Chemicals

The IncMix solution for DRG incubation contains (inM):

NaCl 155, K2HPO4 1.5, HEPES 5.6, Na-HEPES 4.8, glu-

cose 5. The standard extracellular solution contains (mM):

NaCl 140, KCl 4, CaCl2 2, MgCl2 1, HEPES 10, NaOH

4.54, glucose 5; pH 7.4 at 25�C. The high-KCl solution

contained (in mM): NaCl 94, KCl 50, CaCl2 2, MgCl2 1,

HEPES 10, NaOH 4.54, glucose 5, pH 7.4 at 25�C. A

methylgyoxal solution (*40% in water) was used, being

diluted in DMEM F120Ham medium (5% horse serum) at

the required final concentration. Laminin is from BD

Biosciences. All others chemicals, unless otherwise speci-

fied, are from Sigma, St. Louis, MO.

Primary Cell Culture

DRG neurons were obtained from all spinal levels from

adult male Balb/c mice (15–20 g), adapted accordingly to

the previous described protocol for rat neuronal primary

culture (Reid et al. 2002). The animals were killed by CO2

inhalation (1 min) followed by decapitation according to

the European Guidelines on Laboratory Animal Care, with

the approval of the institutional Ethics Committee of the

University of Bucharest, Faculty of Biology (approved

protocol 21/04.06.2010). A total number of 20 male mice

have been used, and for each experiment the number of

analyzed images or the repeats is mentioned in ‘‘Results’’

section. DRGs were removed under sterile conditions and

were immediately transferred into IncMix solution (see

‘‘Solutions and Chemicals’’ section). After cleaning the

ganglia from surrounding tissue and counting them, the

DRGs were incubated in a mixture of 1 mg/ml Collagenase

(type XI from Clostridium histolyticum, Sigma) and 1 mg/

ml Dispase (from Bacillus polymyxa; GIBCO, Invitrogen,

Carlsbad, CA, USA) in IncMix solution for 1 h at 37�C, as

previously described Reid et al. (2002). Following enzyme

treatment, the ganglia were washed once in Dulbecco’s

modified Eagle’s medium Ham’s F-12 (DMEM F-12) with

10% horse serum, before mechanical trituration in 0.5 ml

DMEM. The dissociated cells were then washed by cen-

trifugation (at 1,0009g for 10 min, 25�C) followed by

resuspension in fresh DMEM F-12. Following the final

wash, the cell pellet was resuspended in DMEM F-12,

containing 10% horse serum and 50 lg/ml gentamicin.

Following a second trituration, the neurons were seeded

and incubated in sterile culture dishes at 37�C and 5% CO2/

1048 Cell Mol Neurobiol (2012) 32:1047–1057

123

95% air for 2 h. After this incubation period, a further 2 ml

DMEM F-12 was added to each culture dish. Neurons were

kept in the incubator for 1 day after dissection, and then a

long-term (24 h) treatment with MG was done. The via-

bility tests and Ca2? imaging measurements are done on

the third day of neuronal culture, and immunofluorescence

staining was done on the seventh day of culture.

MTS Assay

DRG cells (*5,000 neurons/well) were seeded on 96-wells

plates pre-coated with poly-D-lysine and maintained for 24 h

after isolation in the incubator in DMEM F-12 (10% horse

serum). Afterwards cells were treated with different con-

centrations of MG (0, 50, 100, 150, 200, 250, 300, 400, 500,

600, 750, 800, and 1,000 lM) for 24 h in DMEM F-12 (10%

horse serum). In order to evaluate the cell viability it was

used a CellTiter 96� AQueous One Solution Cell Prolifer-

ation Assay (Promega, USA) which contains a tetrazolium

compound [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymeth-

oxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt;

MTS]. The MTS method has been preferred instead of the

classical MTT, because it is faster and the resolution is better

(www.promega.com). The CellTiter 96� AQueous One

Solution Reagent and electron coupling reagent (phenazine

ethosulfate; PES) contained in the kit have been mixed as

indicated by Promega sheet). Each well was washed upon

MG treatment, 20 ll of mixed kit solution in 100 ll DMEM

F-12 (5% horse serum) were added and the well plate was

incubated for 4 h at 37�C. Absorbance was measured in a

reader-plate (Sunrise-Basic, Tecan GmbH, Austria) at

490 nm. Measurements have been done in five replicates.

Intracellular Ca2? Imaging

The DRG neurons plated on 24-mm coverglass previously

coated with poly-D-lysine were incubated for 45 min at

room temperature in a dark chamber in standard extracel-

lular solution (see ‘‘Solutions and Chemicals’’ section)

containing 10 lM Fura-2 acetoxymethyl ester (Invitrogen,

USA). An increased concentration of Fura-2 acetoxy-

methyl ester compared to the standard one of 4 lM

(Naziroglu et al. 2011) was used, as we did not incubate the

cells with pluronic acid. In this manner, we have avoided

an additional cytotoxical effect that could interfere with

MG effect. After loading with dye, the cells were washed

three times with extracellular solution, mounted on a

holder with 1 ml of extracellular solution and left that

Fura-2 acetoxymethyl ester to be de-esterified by intracel-

lular esterases for another 15 min before recording. For

imaging the Fura-2 acetoxymethyl ester emission an IX 71

Olympus microscope was used. The excitation was per-

formed by a Xe lamp with monochromator Polychrome V

(Till Photonics GmbH, Germany) using wavelengths of

340 ± 5 and 380 ± 5 nm. The emission was collected by a

filter at 510 ± 20 nm. The images were acquired by a

cooled CCD camera iXON? EM DU 897 (Andor, North-

ern Ireland) controlled by iQ 1.8 software package, with a

frequency of one pair of images per 2 s. The Ca2? tran-

sients were induced by a long-pulse of 50 mM high-KCl

solution for 4 min. The perfusion was performed with a

MPS-2 (World Precision Instruments, USA). Quantitative

results were expressed by means of emission ratio R =

I340/I380, I being the emission intensity for excitation at 340

and 380 nm, respectively, which is proportional to the free

cytosolic Ca2? concentration ([Ca2?]i). Each cover glass

covered with neurons was used for a single experimental

variant.

Immunofluorescence

The primary culture of DRG neurons was obtained as

described above, and neurons were seeded on 10-mm

coverslips, previously coated with poly-D-lysine (1 mg/ml,

1 h at 37�C) and laminin (50 lg/ml, 1 h at 37�C). A period

of 24 h the cells have been resting to adhere. For the next

24 h, the MG was added (0. 150, 250, 500, and 750 lM).

After that MG was washed out, neurons were maintained in

the same medium, and neurite outgrowth was allowed. On

the seventh day of culture, cells are fixed in 4% parafor-

maldehyde, permeabilized with 0.1% Triton X-100 and

immunostained. Mouse monoclonal [TU-20] to neuron-

specific beta III Tubulin (ab7751, Abcam, UK; dilution

1:250) and Donkey polyclonal secondary antibody to

mouse IgG-H&L (FITC), pre-adsorbed (ab7057, Abcam,

UK; dilution 1:500) were used. Images were captured with

an inverted epifluorescence microscope AxioObserver D1

(Zeiss, Germany) using 109 air objective.

Data Analysis

For Ca2? imaging recordings, the relative change of

emission ratio (DR/R0) was calculated for each recorded

cell during the time course and the most representative

traces have been represented for each MG concentration

tested. The highest DR/R0 values of individual cells have

been used in a box chart like graph for a comparison among

the cell responses obtained at different MG concentrations.

A standard design of the box chart was used expressing the

following parameters: the mean, the median, the inter-

quartile range (the size of the box) delimited by 25th and

75th percentiles, the position of the 5th and 95th percen-

tiles (delimiting the ends of the bar) and the extreme

values.

Each immunofluorescence image was first transformed in

8 bits image format and background corrected. Afterwards, a

Cell Mol Neurobiol (2012) 32:1047–1057 1049

123

gray level of 4 was used as threshold and the images have

been transformed in a black/white image (white for the pixels

covering the neurons and neurite area). From this image,

using a histogram, the number of the white pixels was

evaluated. A quantitative parameter expressing the neurite

density was calculated as the ratio of white pixels to total

pixels. Since neurites in a captured image may belong to

different neurons, the neurite density was normalized to the

number of neurons in the picture. This parameter is only an

approximation since some of the neurites may belong to

neighboring neurons that are invisible in the captured field.

All the image analysis was done using ImageJ software

package.

All the graphs, the statistical analyses, two sampled

unpaired Student t test (p values are reported in the text)

have been done using OriginPro 8 (OriginLab Corporation,

USA). Average data are presented as mean ± SE.

Results

Long-term (24 h) effect of methylglyoxal was tested on

DRG sensory mouse neurons by quantifying the cell via-

bility, the percentage of neurons and glial cells, the excit-

ability of neurons, and their capacity of neurite outgrowth.

Methylglyoxal Affects the Viability of Sensory Mouse

Neurons

A double method was employed for evaluating the meth-

ylglyoxal effect on the DRG primary cell culture, bright-

field microscopy corroborated with MTS assay.

In the first protocol, bright-field microscopy was

employed for quantifying the percentage of neurons and

glial cells present in DRG cultures. Transmitted light

images, Fig. 1a–e, are collected after 24 h exposure to MG.

Both the neurons (yellow arrows in Fig. 1a) and the glia

(green arrows in Fig. 1a) have been analyzed. These pic-

tures suggest qualitatively not only a modulation of cells

number but also a change in cell morphology. This aspect

is more obvious for the glial cells, in which the prolifera-

tive capacity is strongly impaired. The number of neurons

and glia was quantified in images acquired by bright-filed

microscopy. The evaluation was done separately on neu-

rons and glia. The number of DRG cells (either neurons or

glia) cultured in the presence of MG is presented as per-

centage of cells in control conditions (Table 1). A statis-

tical analysis revealed an increase of the percentage of

neurons at 150 lM MG (140 ± 12%, p \ 0.05, n = 25),

and a decrease at 750 lM MG (28 ± 3%, p \ 0.01, n =

47), compared to control conditions (n = 20). The same

analysis indicated an increase in the percentage of glia

at 150 lM MG (114 ± 9%, p \ 0.05, n = 25), and a

decrease at 250 lM MG (78 ± 17%, p \ 0.05, n = 35),

500 lM MG (57 ± 6%, p \ 0.01, n = 49), and 750 lM

MG (19 ± 1%, p \ 0.01, n = 47), compared to control

conditions (n = 20). No positive correlation was obtained

between the viability percentage and the neuronal soma

diameter.

In a second protocol, by means of the MTS assay

(Fig. 1f), we have quantified the total viability (neurons

and glia) in the DRG primary cell culture after 24 h

exposure to MG. By means of this method the number of

all cells in culture is indirectly evaluated without the pos-

sibility to differentiate among the different cell types

present in cultures (as is the case for DRG primary cul-

tures). In Fig. 1f, an extended scale of MG concentrations

(0–1,000 lM) was used, and the most significant ones have

been chosen to be imaged in Fig. 1a–e.

The total viability of the DRG cells is enhanced for

smaller concentration of MG (07150 lM) and strongly

decreased at higher concentrations (20071,000 lM). A

statistical analysis indicated a significantly viability

decrease at 500 lM MG (p \ 0.01, n = 20) and 750 lM

MG (p \ 0.01, n = 20) compared to control conditions

(n = 20).

A dual-phase effect of MG is revealed by both proto-

cols. For 150 lM MG, the MTS assay indicates a 20%

increase in the total cell viability, while the data from

bright field transmitted light indicates a 40% increase in the

number of neurons, and only a 14% increase in the number

of glia. Thus, neurons seem more ‘‘stimulated’’ by low MG

concentrations than glia. At higher MG concentrations, the

glia number dropped down more rapidly compared to

neurons, and the total viability is dramatically decreased.

Methylglyoxal Induced Changes in Intracellular Ca2?

Signaling

Ca2? imaging protocol based on Fura-2 fluorescent probe

was used to reveal the MG-changes in neuronal excitability

in response to 50 mM KCl exposure. The relative change

of Fura-2 ratio (DR/R0) was used to characterize the vari-

ation of intracellular Ca2? concentration after excitation by

50 mM KCl concentration increasing. The most represen-

tative DR/R0 averages values against time computed from

recordings on sensory neurons for each MG concentration

value are summarized in Fig. 2a–e. The presence of

50 mM KCl in the NaCl-based extracellular medium

sharply elicited the increase DR/R0 in most of the tested

neurons.

Long-term exposure (24 h) to MG (0, 150, 250, 500, and

750 lM) strongly modulated the relative change of cyto-

solic Ca2?. Smaller concentrations of MG (150 lM)

induced a strong increase of DR/R0 compared to control

conditions, while higher MG concentrations (500 and

1050 Cell Mol Neurobiol (2012) 32:1047–1057

123

750 lM) drastically diminished the Ca2? increase. Again a

dual-phase response is revealed, in this case by the

dependence of DR/R0 on MG concentration.

For each trace the maximal DR/R0 was measured, and

box-like chart (Fig. 2f) was drawn for a better character-

ization of the DR/R0 distributions with respect to MG

concentrations. The dispersion of DR/R0 characterized by

the box size changes with the MG concentration in a

similar manner to the mean value of DR/R0. The largest

data distributions occur for the control and 150 lM MG

variants. The ratios for 150 lM MG are more disperse

since 50% of DR/R0 data range in 0.23 units compared to

0.16 units for control conditions. The dispersion decreases

at higher concentrations (750 lM MG) up to 0.05 units.

Several studies have characterized different neuronal

subpopulations in the rodent DRGs, depending on the

transient Ca2? response to various cues (Forster et al. 2009;

Engel et al. 2011) or on the cell body size defined

subpopulations (Lu et al. 2006). Consequently, a distribu-

tion of relative fluorescence change signals (DR/R0 for each

recorded neuron) with respect to the mouse DRG neuronal

subtype may describe in more detail the MG effects. The

diagrams in the Fig. 3b–f show the distribution of cellular

responses in function of cells’ diameters. Each dot from the

diagram is related to a particular sensory neuron. In our

experiments, the soma of neurons ranges between 8 and

35 lm meaning that only small- and medium-sized neu-

rons have been recorded (Hiruma et al. 2000; Ren et al.

2012). Considering the limit size of 25 lm between small

and medium neurons (Hiruma et al. 2000), these two

subpopulations have different patterns of fluorescence ratio

signals for various MG-tested concentrations. In small-

sized neurons, the fluorescence ratio has an upper limit of

0.6 units for control and 150 lM MG, while this limit

significantly decreases to 0.37 units, 0.3 units, and 0.1 units

for 250, 500, and 750 lM MG, respectively. In the small

A B C

D E F

80

100

120

140

ility

(%

)

**

40 m 0 200 400 600 800 10000

20

40

60**

Via

bi

[MG] / M[MG] / μMμ

Fig. 1 DRG primary culture viability in the presence of MG. a, b, c, d,

and e Transmitted light images of representative cell fields of cultures

incubated 24 h with 0, 150, 250, 500, and 750 lM MG, respectively—

arrows mark the neurons (yellow) and glial cells (green); scale bar

40 lm; 109 air objective, f DRG primary culture viability tested by

MTS assay, and the level of statistical significance is **p \ 0.001,

*p\ 0.05. Samples were acquired from 4 mice primary cultures, and

each MG concentration was tested in 5 repeats (Color figure online)

Table 1 Percentage of cells (neurons or glial cells) relative to control in DRG primary cultures exposed to MG

[MG]/lM 0 150 250 500 750

% Sensory neurons 100 ± 5 140 ± 12* 109 ± 7 92 ± 11 28 ± 3**

% Glial cells 100 ± 7 114 ± 9* 78 ± 17* 57 ± 6** 19 ± 1**

n Number of analyzed images 20 25 35 49 47

The results significantly different in comparison to control are labeled by stars, level of statistical significance: ** p \ 0.01, * p \ 0.05

In the third line is indicated the number n of analyzed images

Cell Mol Neurobiol (2012) 32:1047–1057 1051

123

neuronal subpopulation, the excitability is maintained up to

150 lM MG similar to control and it drastically diminishes

towards higher concentrations. In medium-sized neurons,

the upper limit of the fluorescence ratio signal is 0.2 units

in control and significantly increases to 0.6 units for

150 lM MG, while this limit recovers to 0.2 units for

250 lM MG (similar to control), and even lowers below to

0.1 units for 500 and 750 lM MG. In the medium neuronal

subpopulation, the excitability reaches a significant peak at

150 lM MG and then towards higher MG values dimin-

ishes below control. The dispersion analysis (box chart)

was represented for all the recorded cells without any

separation in respect to the neuronal diameter. Based on

cell viability and [Ca2?]i fluctuations upon MG treatment,

one can consider 150 lM MG as a threshold value. In

conclusion, the medium neurons seem to be more sensitive

to this threshold concentration compared to small neurons.

Methylglyoxal Modifies the Neurite Outgrowth

of Sensory Neurons

In order to evaluate a late effect of MG on DRG primary

cultures, the process of neurite outgrowth was analyzed.

The influence of 24 h exposure to MG at the beginning of

the culture period was quantified 5 days after MG was

replaced by fresh culture medium. In this way, the capacity

of neurons recovery after 24 h MG may be observed by

their ability to promote the neurite expansion.

The most representative images for neurite outgrowth are

presented in Fig. 4a–e. As is seen from these images, the

number of neurons is similarly decreased as in Fig. 1a–e. At

low MG concentrations, the neurites are very branched and

dense (orange arrows marked in Fig. 4a, b), while at higher

MG concentrations this pattern disappears and only solid and

bright neurites may be observed (arrows in Fig. 4d, e). The

0 50 100 150 200 250 300

0.0

0.2

0.4

0.6

0.8 Control

0 50 100 150 200 250 300

0.0

0.2

0.4

0.6

0.8 250 μM MG

0 50 100 150 200 250 300

0.0

0.2

0.4

0.6

0.8 500 μM MG

0 50 100 150 200 250 300

0.0

0.2

0.4

0.6

0.8

ΔR

/R0

Time / sec

750 μM MG

0 50 100 150 200 250 300

0.0

0.2

0.4

0.6

0.8 150 μM MGA B

C D

E F

0.0

0.1

0.2

0.3

0.4

0.5

0.6Control 150 μM MG 250 μM MG 500 μM MG 750 μM MG

Max

imal

ΔR

/R0

Time / sec

Time / sec Time / sec

Time / sec

ΔR/R

0

ΔR

/R0

ΔR

/R0

ΔR

/R0

Fig. 2 50 mM KCl stimulated

[Ca2?]i transients in sensory

neurons after 24 h exposure

zto MG. a, b, c, d, and

e Representative traces of DR/

R0 against time for neurons

incubated with 0, 150, 250, 500,

and 750 lM MG, respectively

(the arrow marks the moment

when the 50 mM KCl perfusion

was started); Fura-2 was excited

at 340 ± 5 and 380 ± 5 nm,

and the emission was collected

at 510 ± 20 nm, f box chartpresenting the mean and the

dispersion of maximal DR/R0

data for each MG concentration,

n is indicated in the upper part

of each box

1052 Cell Mol Neurobiol (2012) 32:1047–1057

123

neurite extension was not dependent on the neuronal soma

dimensions.

The quantitative evaluation of the neurons capacity to

produce and extend neurites was done by computing the ratio

between the percent of area covered by neurons and neurites

in a picture and the number of neurons in that picture (for

details see ‘‘Data Analysis’’ section) and is represented in

Fig. 4f. This parameter is quite approximate since a part of

the neurites produced by the neurons with the soma in the

analyzed picture is outside of the field covered by the picture.

On the other hand, some of the neurites in the analyzed image

belong to neurons that are not visible. Considering com-

pensation among these two sources of errors, these param-

eters may give a quantitative evaluation of the MG influence

on the neurite outgrowth process.

The neurites area (%) per number of neurons signifi-

cantly increased at 150 lM MG (1.36 ± 0.05 units,

p \ 0.01, n = 5) and 250 lM MG (0.88 ± 0.02 units,

p \ 0.05, n = 10), compared to control conditions

(0.69 ± 0.05 units, n = 5). On the other hand, neurites

area (%) per number of neurons significantly decreased at

750 lM MG (0.35 ± 0.03, p \ 0.01, n = 12), compared

to control conditions (0.69 ± 0.05 units, n = 5). No sig-

nificant change in the extension capacity was recorded for

500 lM MG (n = 13) compared to control conditions.

Discussions

Unlike axons in the central nervous system, injured

peripheral nerves exhibit the capacity to regenerate, largely

due to the supportive population of Schwann cells within

the peripheral nervous system (Webber et al. 2011). Glial

cells play a critical role in regeneration of adult DRG

sensory axons, and the in vivo growth is conditioned by a

tightly coupled Schwann cells process (Armati 2007).

Different cues that prevent Schwann cells proliferation and

migration, such as hyperglycemia, arrest nerve growth

(Armati 2007; Webber and Zochodne 2010). As MG is an

endogenously produced compound in physiological and

pathological conditions (i.e. hyperglycemia in diabetic

status), it was very important in our study to correctly

evaluate its effect on adult primary peripheral neuronal

cultures, despite the mixture of cells that is present and

taking into account the benefit of the glial cells presence.

MTS assay is becoming a commonly used protocol for

the evaluation of cell viability, and we have employed this

technique for quantifying viability in DRG primary cul-

tures from adult mice. Despite the fact that data are easy to

collect, in primary DRG cultures is difficult to interpret

them, and to attribute the changes in viability to different

DRG subpopulations (Hiruma et al. 2000; Lu et al. 2006;

B C 1.01.0

0.4

0.6

0.8

alΔR

/R0 150 μM MG

0.4

0.6

0.8 Control

lΔR

/R0

n = 24 n = 30

0.0

0.2

Max

ima

0.0

0.2

Max

imal

30 µm

D E F1.0 1.0 1.0

0.4

0.6

0.8

mal

ΔR/R

0

250 μM MG

0.4

0.6

0.8

mal

ΔR/R

0

500 μM MG

0.4

0.6

0.8

mal

ΔR/R

0 750 μM MGn = 14 n = 16n = 29

0 10 20 30 40

0.0

0.2

Max

im

Cell diameter (μm)

0.0

0.2

Max

im

Cell diameter (μm)

0.0

0.2

Max

im

Cell diameter (μm)0 10 20 30 40 0 10 20 30 40

Cell diameter (μm)Cell diameter (μm)0 10 20 30 40 0 10 20 30 40

A

Fig. 3 Maximal values of DR/R0 recorded on neurons after 24 h MG

exposure. a Transmitted light image exemplifying the size of neurons

(small—yellow arrow, medium—green arrow), b–f diagrams of

maximal DR/R0—cell diameter showing the distribution of cells as

excitability with respect to their size, n the number of recorded cells is

indicated in the upper part of each cell distribution (Color figure online)

Cell Mol Neurobiol (2012) 32:1047–1057 1053

123

Forster et al. 2009; Engel et al. 2011; Naziroglu et al. 2011;

Ren et al. 2012). From this point of view, the MTS-assay

was doubled by the bright-field microscopy, and sensory

neurons where discriminated from glia in the quantification

done on the captured images. In these terms, the changes in

the number of sensory neurons in absence/presence of MG

are summarized in Table 1. One possible hypothesis to

explain the increased number of neurons after low-dose

MG treatment (150 lM) is a higher neuronal ability to

maintain adherence to the poly-D-lysine substrate com-

pared to control conditions. As previously described, neu-

ronal aggregation and fasciculation is closely related to the

surface properties of the glial and non-glial cells (Noble

et al. 1984), and by consequence the reactivity of sensory

neurons in primary culture to MG treatment should be

influenced by the presence of glial cells. In addition, the

proliferative capacity of glial cells is slightly increased

(p \ 0.05, Table 1) at 150 lM MG. In other terms, one

should consider the neuroprotective role against MG

treatment played by glial cells in primary DRG culture,

similar to the neuroprotection exerted by astrocytes in

primary cortical mouse cultured neurons (Belanger et al.

2011). At higher MG doses (250–1,000 lM), a dose-

dependent drop in the cell survival occurs.

40 µm

A B

C D

E

0 200 400 600 8000.0

0.3

0.6

0.9

1.2

1.5

Neu

rite

s ar

ea (%

)/ N

o ce

lls

[MG] / μM

F

(5)

(5)

(13)

(12)

(10)

Fig. 4 Neurite outgrowth in the presence of methylglyoxal.

a–e Fluorescence images (excitation: 450–490 nm, beam splitter:

510 nm, emission: 515–565 nm) of arborized neurons (FITC–beta III

tubulin labeled) 5 days after 24 h MG exposure—orange arrowsmark very dense arborization patterns; scale bar 40 lm. 109 air

objective, f MG effect on the efficiency of neurite outgrowth process

quantified by the percentage of area covered by neurons and neurite

normalized to the number of neurons in the image (the results

significantly different in comparison to control are labeled by stars,

level of statistical significance: **p \ 0.001, *p \ 0.05). The number

n of analyzed images is written in the graph above each circle

corresponding to a particular MG concentration (Color figure online)

1054 Cell Mol Neurobiol (2012) 32:1047–1057

123

MG-induced apoptotic death is cell type dependent: it

induces apoptosis in human umbilical vein endothelial cells

(Chan and Wu 2008), Madin–Darby canine kidney renal

tubular cells (Jan et al. 2005), human mononuclear cells

(Hsieh and Chan 2009), Neuro-2A neuroblastoma (Huang

et al. 2008), SH-SY5Y neuroblastoma (Li et al. 2011),

Jurkat (Du et al. 2000) cells, rat hippocampal (Di Loreto

et al. 2008), and cortical neurons(Kikuchi et al. 1999), but

it fails to induce apoptosis in MOLT-4, HeLa, or COS-7

cells (Du et al. 2000). An apoptotic death of the peripheral

sensory neurons in the presence of MG might be very

likely to occur.

Some peripheral DRG neurons are able to survive and to

generate neurites even at higher MG concentrations. For

instance, at 750 lM MG the drop of the total viability

(Fig. 1f) and the decrease in the percentage of neurite area/

number of neurons (Fig. 4f) is around 50% compared to

control conditions. As our data are focused on the periph-

eral nervous system, it is important to discuss the MG

effect compared to neuroblastoma and central nervous

system. At 750 lM MG, the SH-SY5Y human neuroblas-

toma cells survival rate is reduced to 56.85 ± 1.6%

(Li et al. 2011), while at only 124 lM MG, the primary rat

hippocampal neurons have a drop in viability of 50%

(Chen et al. 2010). By consequence, one can emphasize

that the rate survival of peripheral sensory neurons upon

MG treatment is comparable to SH-SY5Y neuroblastoma

cells, but is far more increased than for primary rat hip-

pocampal neurons at the same dose. Moreover, peripheral

sensory neurons seem to be more resistance to methyl-

glyoxal treatment than central nervous system neurons.

These differences in the rate survival might be explained

in terms of MG detoxifying activity by GLO-I and GLO-II

in different neuronal systems. In SH-SY5Y neuroblastoma,

the treatment with bromobenzylglutathione cyclopentyl

diester, an inhibitor of glyoxalase I, leads to reduced cell

viability, strongly retracted neuritis, increase in [Ca2?]i,

and activation of caspase-3 (Kuhla et al. 2006). In

SH-SY5Y neuroblastoma, doses of 100–200 lM MG

induce a significant time- and dose-dependent inhibition of

GLO-I activity, while GLO-II activity does not vary with

respect to controls (Amicarelli et al. 2003). However, the

same MG dose (100 lM) caused a significant decrease in

the specific activities of GLO-I and GLO-II in the rat

hippocampus embryonic neurons (Di Loreto et al. 2008). A

highly efficient glyoxalase system (i.e. GLO-I and GLO-II)

in primary mouse astrocytes is associated with lower

accumulation of AGEs compared with cortical neurons, a

sixfold greater resistance to MG toxicity, and the capacity

to protect neurons against MG in a coculture system

(Belanger et al. 2011).

Further one, our study pointed out that peripheral sen-

sory neurons are well-protected against MG cytotoxicity by

glial cells, if the 24 h treatment is done in the second day of

the DRG culture, and not 1 week after the network for-

mation as for hippocampal neurons (Chen et al. 2010). In

conclusion, it is important to evaluate the role of MG

in various pathologies on peripheral primary neuronal

cultures, and not to extend the conclusions obtained on

neuroblastoma or primary central nervous system cultures.

A distinct point in our study was to correlate the effects

of MG treatment on neuronal viability with the cytosolic

free Ca2? ([Ca2?]i) changes. A few studies have already

proved that short- and long-term MG treatment of different

non-neuronal cells is inducing changes in [Ca2?]i (Jan et al.

2005; Chan and Wu 2008; Hsieh and Chan 2009). High

glucose and long-term MG (24 h) co-treatment of human

mononuclear cells and human umbilical vein endothelial

cells increase [Ca2?]i and nitric oxide levels, activated

nitric oxide synthase, induce the loss of mitochondrial

membrane potential, and lead to activation of caspases-9

and -3, cytochrome c release and finally determines the cell

death (Chan and Wu 2008; Hsieh and Chan 2009). In

addition, brief MG application (40 s) induced a significant

rise of [Ca2?]i in Madin–Darby canine kidney renal tubular

cells by stimulating both extracellular Ca2? influx and

CCCP (carbonylcyanide m-chlorophenylhydrazone)-sensi-

tive intracellular Ca2? release, and causing apoptotic cell

death (Jan et al. 2005).

In vitro Ca2? imaging microscopy and patch-clamp

recordings on rat DRG neurons show sustained activation

of the transient receptor potential ankyrin 1 (TRPA1)

channel by prolonged infusion of MG (300 lM), which is

an endogenous TRPA1 agonist generated in diabetes mel-

litus (Koivisto et al. 2012). Moreover, treatment (1 h) of rat

TRPA1-inducible HEK-293 cells with MG at concentra-

tions ranging from 137 nM to 100 lM and from 1.37 to

1,000 lM, respectively, evoked a biphasic activation of

TRPA1 in fluorometric imaging plate reader measurements

(Koivisto et al. 2012).

Our study constitutes a different approach regarding the

effect of methylglyoxal on the DRG neuronal excitability.

Imaging [Ca2?]i changes by means of time-lapse fluores-

cence microscopy, we have shown that the long-term (24 h)

MG treatment changes the excitability properties of small/

medium-sized DRG neurons. It was proven a dual effect of

MG in mouse DRG sensory neurons depolarized with high

KCl (50 mM) extracellular solution, with an increase in DR/

R0 at low MG doses (0–150 lM) and a decrease in DR/R0 at

higher doses (250–750 lM). This MG dose-specific Ca2?

signaling in peripheral sensory neurons might play an

important role in the early process of MG-cytotoxic action in

peripheral nerves during the pre-diabetic status installation.

The dual MG effect on sensory neurons might indicate a rate-

limiting step in the detoxifying process. A recent in vivo

study indicates that GLO-I activity may affect the way

Cell Mol Neurobiol (2012) 32:1047–1057 1055

123

sensory neurons respond to heightened AGE levels in dia-

betic peripheral neuropathy and that the abundance of GLO-I

varies within different sensory neuron populations (Jack

et al. 2011). Further biochemical studies on sensory neurons

should be done to identify the exact point where the GLO-I

and GLO-II enzymatic cascades are overloaded by the

excess methylglyoxal.

As it concerns the ability of sensory neurons to cope

with their substrate (poly-D-lysine and laminin) in the

presence of MG, our study has proven a dose-dependent

impairment in extending neuritis. Previous studies have

proven that glycation of extracellular matrix proteins by

MG produces failure of sensory nerve regeneration in

streptozotocin-induced diabetes (Duran-Jimenez et al.

2009). Although survival of sensory neurons is not altered

by growth on a methylglyoxal-glycated laminin support,

the neurite outgrowth is dramatically lowered (Duran-

Jimenez et al. 2009). In our experiments the support is pre-

coated, but the MG treatment is done after 24 h from

platting the DRG sensory neurons. It is possible that the

previously described pre-MG-glycation of the support is

less cytotoxic compared to our extracellular MG treatment.

Again, a dual effect of MG treatment is observed (Fig. 4),

at low concentration (150 lM) being increased the exten-

sion area for neuritis in opposite with the high concentra-

tion treatment (500 and 750 lM). Nevertheless, the

extracellular matrix is one of the major targets of MG

effect, but in the same time specific ion channels/receptors

exposed to the extracellular side are affected by the treat-

ment. Therefore, the MG-induced changes in excitability

detected in the presence of high KCl extracellular solution

could be also related to the site-directed cross-linking of

targeted amino acids (i.e. arginine, lysine) (Thornalley

2005; Fleming et al. 2011) in the structure of neuronal

receptors. In this view, an ubiquitary cross-linking of MG

to various active sites of voltage-dependent channels (Best

et al. 1999; WO/2010/136182; Mukohda et al. 2010) or a

specific agonist action against TRPA1 channels (Koivisto

et al. 2012) could be the insight mechanism of early gly-

cosylation effects against peripheral sensory neurons.

In conclusion, peripheral accumulation of methylgyoxal

in the extracellular space could be the trigger for changes

in viability, excitability, and neurite outgrowth. Our study

is the first to point out the dual effect of methylglyoxal in

peripheral sensory neurons. A threshold in metabolizing

methylglyoxal by the neuronal glyoxalase system up to a

certain concentration (150 lM) is very important in

explaining the effect of AGEs in the peripheral nervous

system during various phases of diabetes, aging, or neu-

rodegenerative pathologies.

Acknowledgments This work was supported by the national grant

PNII 41-074/2007 from the Romanian Ministry of Research. A great

thanks to the technicians Cornelia Dragomir, Geanina Haralambie,

and Constantin Radulescu for a constant help during the experiments.

References

Amicarelli F, Colafarina S, Cattani F, Cimini A, Di Ilio C, Ceru MP,

Miranda M (2003) Scavenging system efficiency is crucial for

cell resistance to ROS mediated methylglyoxal injury. Free Rad

Biol Med 35(8):856–871

Armati P (2007) The biology of Schwann cells: development

differentiation and immunomodulation. Cambridge University

Press, Cambridge, pp 2–4

Belanger M, Yang J, Petit JM, Laroche T, Magistretti PJ, Allaman I

(2011) Role of the glyoxalase system in astrocyte-mediated

neuroprotection. J Neurosci 31(50):18338–18352

Best L, Miley HE, Brown PD, Cook LJ (1999) Methylglyoxal causes

swelling and activation of a volume-sensitive anion conductance

in rat pancreatic beta-cells. J Membr Biol 167(1):65–71

Chan WH, Wu HJ (2008) Methylglyoxal and high glucose co-

treatment induces apoptosis or necrosis in human umbilical vein

endothelial cells. J Cell Biochem 103(4):1144–1157

Chen YJ, Huang XB, Li ZX, Yin LL, Chen WQ, Li L (2010) Tenuigenin

protects cultured hippocampal neurons against methylglyoxal-

induced neurotoxicity. Eur J Pharmacol 645(1–3):1–8

de Arriba SG, Stuchbury G, Yarin J, Burnell J, Loske C, Munch G

(2007) Methylglyoxal impairs glucose metabolism and leads to

energy depletion in neuronal cells—protection by carbonyl

scavengers. Neurobiol Aging 28(7):1044–1050

Di Loreto S, Zimmitti V, Sebastiani P, Cervelli C, Falone S,

Amicarelli F (2008) Methylglyoxal causes strong weakening of

detoxifying capacity and apoptotic cell death in rat hippocampal

neurons. Int J Biochem Cell Biol 40(2):245–257

Du J, Suzuki H, Nagase F, Akhand AA, Yokoyama T, Miyata T,

Kurokawa K, Nakashima I (2000) Methylglyoxal induces

apoptosis in Jurkat leukemia T cells by activating c-Jun

N-terminal kinase. J Cell Biochem 77:333–344

Duran-Jimenez B, Dobler D, Moffatt S, Rabbani N, Streuli CH,

Thornalley PJ, Tomlinson DR, Gardiner NJ (2009) Advanced

glycation end products in extracellular matrix proteins contribute

to the failure of sensory nerve regeneration in diabetes. Diabetes

58(12):2893–2903

Engel MA, Leffler A, Niedermirtl F, Babes A, Zimmermann K,

Filipovic MR, Izydorczyk I, Eberhardt M, Kichko TI, Mueller-

Tribbensee SM, Khalil M, Siklosi N, Nau C, Ivanovic-Burmaz-

ovic I, Neuhuber WL, Becker C, Neurath MF, Reeh PW (2011)

TRPA1 and substance P mediate colitis in mice. Gastroenterol-

ogy 141(4):1346–1358

Fleming TH, Humpert PM, Nawroth PP, Bierhaus A (2011) Reactive

metabolites and AGE/RAGE-mediated cellular dysfunction affect

the aging process: a mini-review. Gerontology 57(5):435–443

Forster AB, Reeh PW, Messlinger K, Fischer MJ (2009) High

concentrations of morphine sensitize and activate mouse dorsal

root ganglia via TRPV1 and TRPA1 receptors. Mol Pain 5:17

Han Y, Randell E, Vasdev S, Gill V, Gadag V, Newhook LA, Grant

M, Hagerty D (2007) Plasma methylglyoxal and glyoxal are

elevated and related to early membrane alteration in young,

complication-free patients with Type 1 diabetes. Mol Cell

Biochem 305(1–2):123–131

Hiruma H, Saito A, Ichikawa T, Kiriyama Y, Hoka S, Kusakabe T,

Kobayashi H, Kawakami T (2000) Effects of substance P and

calcitonin gene-related peptide on axonal transport in isolated

and cultured adult mouse dorsal root ganglion neurons. Brain

Res 883(2):184–191

1056 Cell Mol Neurobiol (2012) 32:1047–1057

123

Hsieh MS, Chan WH (2009) Impact of methylglyoxal and high

glucose co-treatment on human mononuclear cells. Int J Mol Sci

10(4):1445–1464

Huang SM, Chuang HC, Wu CH, Yen GC (2008) Cytoprotective

effects of phenolic acids on methylglyoxal-induced apoptosis in

Neuro-2A cells. Mol Nutr Food Res 52(8):940–949

Ibi M, Matsuno K, Shiba D, Katsuyama M, Iwata K, Kakehi T,

Nakagawa T, Sango K, Shirai Y, Yokoyama T, Kaneko S, Saito

N, Yabe-Nishimura C (2008) Reactive oxygen species derived

from NOX1/NADPH oxidase enhance inflammatory pain.

J Neurosci 28(38):9486–9494

Jack MM, Ryals JM, Wright DE (2011) Characterisation of glyox-

alase I in a streptozocin-induced mouse model of diabetes with

painful and insensate neuropathy. Diabetologia 54(8):2174–2182

Jan CR, Chen CH, Wang SC, Kuo SY (2005) Effect of methylglyoxal

on intracellular calcium levels and viability in renal tubular cells.

Cell Signal 17(7):847–855

Jimenez-Andrade JM, Herrera MB, Ghilardi JR, Vardanyan M,

Melemedjian OK, Mantyh PW (2008) Vascularization of the

dorsal root ganglia and peripheral nerve of the mouse: implica-

tions for chemical-induced peripheral sensory neuropathies. Mol

Pain 4:10

Kikuchi S, Shinpo K, Moriwaka F, Makita Z, Miyata T, Tashiro K

(1999) Neurotoxicity of methylglyoxal and 3-deoxyglucosone on

cultured cortical neurons: synergism between glycation and

oxidative stress, possibly involved in neurodegenerative dis-

eases. J Neurosci Res 57(2):280–289

Koivisto A, Hukkanen M, Saarnilehto M, Chapman H, Kuokkanen K,

Wei H, Viisanen H, Akerman KE, Lindstedt K, Pertovaara A

(2012) Inhibiting TRPA1 ion channel reduces loss of cutaneous

nerve fiber function in diabetic animals: Sustained activation of

the TRPA1 channel contributes to the pathogenesis of peripheral

diabetic neuropathy. Pharmacol Res 65(1):149–158

Kuhla B, Luth HJ, Haferburg D, Weick M, Reichenbach A, Arendt T,

Munch G (2006) Pathological effects of glyoxalase I inhibition

in SH-SY5Y neuroblastoma cells. J Neurosci Res 83(8):1591–

1600

Li G, Chang M, Jiang H, Xie H, Dong Z, Hu L (2011) Proteomics

analysis of methylglyoxal-induced neurotoxic effects in SH-

SY5Y cells. Cell Biochem Funct 29(1):30–35

Lu SG, Zhang X, Gold MS (2006) Intracellular calcium regulation

among subpopulations of rat dorsal root ganglion neurons.

J Physiol 577(Pt 1):169–190

Lu J, Randell E, Han Y, Adeli K, Krahn J, Meng QH (2011) Increased

plasma methylglyoxal level, inflammation, and vascular

endothelial dysfunction in diabetic nephropathy. Clin Biochem

44(4):307–311

Lupachyk S, Shevalye H, Maksimchyk Y, Drel VR, Obrosova IG

(2011) PARP inhibition alleviates diabetes-induced systemic

oxidative stress and neural tissue 4-hydroxynonenal adduct

accumulation: correlation with peripheral nerve function. Free

Radic Biol Med 50(10):1400–1409

Mukohda M, Yamawaki H, Okada M, Hara Y (2010) Methylglyoxal

enhances sodium nitroprusside-induced relaxation in rat aorta.

J Pharmacol Sci 112(2):176–183

Nascimento RS, Santiago MF, Marques SA, Allodi S, Martinez AMB

(2008) Diversity among satellite glial cells in dorsal root ganglia

of the rat. Braz J Med Biol Res 41(11):1011–1017

Naziroglu M (2007) New molecular mechanisms on the activation of

TRPM2 channels by oxidative stress and ADP-ribose. Neuro-

chem Res 32(11):1990–2001

Naziroglu M, Ozgul C, Cig B, Dogan S, Uguz AC (2011) Glutathione

modulates Ca(2?) influx and oxidative toxicity through TRPM2

channel in rat dorsal root ganglion neurons. J Membr Biol

242(3):109–118

Noble M, Fok-Seang J, Cohen J (1984) Glia are a unique substrate for

the in vitro growth of central nervous system neurons. J Neurosci

4(7):1892–1903

Reid G, Babes A, Pluteanu F (2002) A cold- and menthol-activated

current in rat dorsal root ganglion neurones: properties and role

in cold transduction. J Physiol 545:595–614

Ren YS, Qian NS, Tang Y, Liao YH, Yang YL, Dou KF, Toi M

(2012) Sodium channel Nav1.6 is up-regulated in the dorsal root

ganglia in a mouse model of type 2 diabetes. Brain Res Bull

87(2–3):244–249

Thornalley PJ (2005) Dicarbonyl intermediates in the Maillard

reaction. Ann N Y Acad Sci 1043:111–117

Toth C, Brussee V, Cheng C, Zochodne DW (2004) Diabetes mellitus

and the sensory neuron. J Neuropathol Exp Neurol 63:561–573

Webber C, Zochodne D (2010) The nerve regenerative microenvi-

ronment: early behavior and partnership of axons and Schwann

cells. Exp Neurol 223(1):51–59

Webber CA, Christie KJ, Cheng C, Martinez JA, Singh B, Singh V,

Thomas D, Zochodne DW (2011) Schwann cells direct periph-

eral nerve regeneration through the Netrin-1 receptors, DCC and

Unc5H2. Glia 59(10):1503–1517

WO/2010/136182 WIPO patent application, Newroth P, Bierhaus A,

Fleming T (2010) Methylglyoxal-scavenging compounds and their

use for the prevention and treatment of pain and/or hyperalgesia

www.promega.com

Cell Mol Neurobiol (2012) 32:1047–1057 1057

123