Accepted Manuscript Experimental diabetes in neonatal mice induces early peripheral sensorimotor neuropathy Lorena Ariza, Gemma Pagès, Belén García-Lareu, Stefano Cobianchi, Pedro José Otaegui, Jesús Ruberte, Miguel Chillón, Xavier Navarro, Assumpció Bosch PII: DOI: Reference:
S0306-4522(14)00407-2 http://dx.doi.org/10.1016/j.neuroscience.2014.05.015 NSC 15417
To appear in:
Neuroscience
Accepted Date:
9 May 2014
Please cite this article as: L. Ariza, G. Pagès, B. García-Lareu, S. Cobianchi, P.J. Otaegui, J. Ruberte, M. Chillón, X. Navarro, A. Bosch, Experimental diabetes in neonatal mice induces early peripheral sensorimotor neuropathy, Neuroscience (2014), doi: http://dx.doi.org/10.1016/j.neuroscience.2014.05.015
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
EXPERIMENTAL DIABETES IN NEONATAL MICE INDUCES EARLY PERIPHERAL SENSORIMOTOR NEUROPATHY
Lorena Ariza1, Gemma Pagès1, Belén García-Lareu1, Stefano Cobianchi2,3, Pedro José Otaegui1, Jesús Ruberte1,4, Miguel Chillón1,5, Xavier Navarro2,3, Assumpció Bosch1
1
Center of Animal Biotechnology and Gene Therapy (CBATEG) and Department of
Biochemistry and Molecular Biology, Universitat Autònoma de Barcelona, Bellaterra, Barcelona, Spain. 2
Department of
Cell
Biology,
Physiology and Immunology and Institute of
Neurosciences, Universitat Autònoma de Barcelona, Bellaterra, Spain. 3
Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas
(CIBERNED), Instituto de Salud Carlos III, Spain. 4
Department of Animal Health and Anatomy, Veterinary School, Universitat Autònoma
de Barcelona, Bellaterra, Barcelona, Spain. 5
Institut Català de Recerca i Estudis Avançats (ICREA), Barcelona, Spain.
Correspondence should be addressed to A.B. ([email protected]): CBATEG, Edifici H, Campus Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain. Telephone: 34 93 5814203; FAX: 34 93 5814205
Short title: Neonatal diabetes induces early peripheral neuropathy
1
ABSTRACT
Animal models of diabetes do not reach the severity of human diabetic neuropathy but relatively mild neurophysiological deficits and minor morphometric changes. The lack of degenerative neuropathy in diabetic rodent models seems to be a consequence of the shorter length of the axons or the shorter animal life span. Myelin proteins have a median half-life of 200 days, thus, diabetes-induced demyelination needs many weeks or even months before it can be evident by morphometrical analysis. In mice myelination of the peripheral nervous system starts at the prenatal period and it is complete several days after birth. Here we induced experimental diabetes to neonatal mice and we evaluated its effect on the peripheral nerve 4 and 8 weeks after diabetes induction. Neurophysiological values showed a decline in sensory nerve conduction velocity at both time-points. Morphometrical analysis of the tibial nerve demonstrated a decrease in fiber size and myelin thickness, and myelinated fiber loss at both timepoints studied. Moreover, aldose reductase and poly(ADP-ribose) polymerase activities were increased even if the amount of the enzyme was not affected. Thus, type 1 diabetes in newborn mice induces early peripheral neuropathy and may be a good model to assay pharmacological or gene therapy strategies to treat diabetic neuropathy.
Key words: diabetic neuropathy, neonatal mice, myelination, aldose reductase activity, PARP activity Abbreviations AF, amyelinic fibers; AR, aldose reductase; CMAP, compound muscle action potential; CNAP, compound nerve action potential; DPN, peripheral polyneuropathy; MF, myelinated
fibers;
NCV,
nerve
conduction
velocity;
PARP,
poly-ADP-ribose
polymerase; SC, Schwann cells; STZ, streptozotocin.
2
INTRODUCTION
Peripheral polyneuropathy (DPN) is one of the most common long-term secondary complications of diabetes mellitus and affects both somatic sensory and autonomic nerves caused by hyperglycemia/hypoinsulinemia. Between the characteristic pathological features of human DPN are axonal degeneration, nerve fiber loss, segmental demyelination and problems in remyelination. The prevalence in humans is about 60% of the patients but electrophysiological alterations can be demonstrated in almost 100% of diabetic patients even if the pathology is subclinical (AguilarRebolledo, 2005). The molecular mechanisms involved in the pathogenesis of diabetic complications are multiple. Among them, acceleration of the polyol pathway due to the high levels of glucose was one the first factors suggested. This pathway is composed of two enzymes, aldose reductase (AR) that is highly expressed in Schwann cells (SC) and converts glucose to sorbitol using NADPH as a cofactor, and sorbitol dehydrogenase that converts sorbitol to fructose using NAD+. High levels of sorbitol induce damage in nervous tissue and, if NAPDH is not available in SC, the amount of reduced glutathione decreases and there is an increased susceptibility to intracellular oxidative stress (Suzuki et al., 1999; Oka & Kato, 2001; Brownlee, 2005). Slowed conduction of action potentials is triggered by both, polyol accumulation which decreases Na+,K+-ATPase activity and leads to intra-axonal sodium accumulation (Mandersloot et al., 1978), and by paranodal and segmental demyelination (Sima & Sugimoto, 1999). On the other hand, under hyperglycemic conditions there is an increase of cell metabolism and of reactive oxygen species production that may alter the mitochondrial membrane potential. Under these conditions the enzyme poly-ADP-ribose polymerase (PARP) is activated, a fundamental mechanism in the development of diabetic complications, including endothelial dysfunction, cardiomyopathy, retinopathy, nephropathy and
3
neuropathy (Garcia Soriano et al., 2001; Pacher et al., 2002; Obrosova et al., 2005a; Drel et al., 2006). Understanding the biochemical mechanisms underlying DPN requires the use of appropriate animal models that reproduce the main features of the human disease (Sullivan et al., 2008). Among them, streptozotocin (STZ) administration to induce type 1 diabetes is one of the most used models for diabetic neuropathy (Kimura et al., 2005; Hong & Kang, 2008; Jang et al., 2010; Sanchez-Zamora et al., 2010). STZ is a drug with diabetogenic properties due to its specific toxicity to pancreatic beta cells. Two procedures have been described to induce diabetes with this antibiotic, using either a single high dose or multiple low doses of STZ. The latter treatment induces subtoxic effects on beta cells and resembles type 1 diabetes due to slow progressive hyperglycemia and lymphocytic infiltration in Langerhans islets. Animals with STZinduced diabetes present termal hyperalgesia, reduced intraepidermal inervation and hypoalgesia at the long-term as neuropathic signs (Cameron et al., 2001; Calcutt, 2004; Li et al., 2005). They also show decreased nerve conduction velocity (NCV) (Calcutt, 2004; Obrosova et al., 2005a; Obrosova, 2009). Among the genetic animal models that have been characterized as appropriate models for diabetic neuropathy, the B6Ins2Akita mice and the BB/Wor (bio-breeding Worcester) mice for diabetes type 1 present deficits in conduction velocity (Choeiri et al., 2002; Stevens et al., 2004; Kamiya et al., 2005; Sullivan et al., 2007), and the most used db/db and ob/ob mice show decreased thermal sensibility and conduction velocity as models for type 2 diabetes (Sullivan et al., 2007; Vareniuk et al., 2007; Vincent et al., 2007). Unfortunately, animal models of diabetes do not commonly reach the severity of human DPN. Animal nerves usually show relatively mild neurophysiological deficits and minor morphological changes, thus limiting the relevance of experimental studies (Sharma & Thomas, 1987; Wright & Nukada, 1994). The lack of degenerative neuropathy in diabetic rodent models seems to be a consequence of the short life span of rodents or the physically shorter axons. Degenerative neuropathy is minimal even in
4
larger animals like dogs or primates with the exception of cats (Mizisin et al., 2007). Demyelination-induced by diabetic neuropathy is hardly detected in animal models as myelin proteins have a half-life longer than 200 days. As an alternative, combination of diabetes and injury has been used to exacerbate the sensorymotor neuropathy since diabetic patients and animal models show impaired nerve regeneration (Kennedy & Zochodne, 2000; Homs et al., 2011). However, in this case it is difficult to distinguish the contribution of diabetic neuropathy from the influence due to nerve injury, and therapeutic strategies that may work in DPN may not necessarily be effective in nerve regeneration and vice-versa. Type 1 diabetes in humans can develop as early as in infanthood and often affects young children and adolescents. Here, by inducing experimental diabetes to neonatal mice, we aimed to produce hyperglycemia during the period when the PNS is most actively myelinating and to study whether experimental diabetes in newborn mice was able to cause severe DPN. We analyzed diabetes induced during the first week of age affectes the peripheral nerve by biochemical, electrophysiological and morphological studies at 4 and 8 weeks after diabetes induction.
5
MATERIALS AND METHODS
Animals Male newborn Hsd:ICR(CD-1) mice received 3 intraperitoneal injections of 40mg/Kg of STZ (dissolved in 0.1 mol/l citrate buffer, pH 4.5, immediately before administration) at post-natal day 3 (P3), P4 and P8. Control animals were injected with citrate buffer at the same ages. Body weight and blood glucose were measured weekly after weaning with a Glucometer Elite (Bayer, Leverkusen, Germany). Breading and weaned mice were fed ad libitum with a standard diet (2018S Teklad Global, Harlan Laboratories; 17% calories from fat) and kept under a light–dark cycle of 12 hr (lights on at 8:00 am). Mice were euthanized at 4 and 8 weeks after the STZ administration. Animal care and experimental procedures were approved by the Biosafety and the Animal and Human Experimentation Ethical Committees of the Universitat Autònoma de Barcelona.
AR activity in sciatic nerve protein extracts AR activity was spectrophotometrically measured following the decrease of NADPH during the reduction of glyceraldehydes (both from Sigma, St Louis, MO) as published (Sato, 1992). Briefly, the reaction was started by adding 100 mM NADPH to 100 µg of sciatic nerve protein extract in 100 mM sodium phosphate buffer containing 1mM D-Lglyceraldehyde and 400 mM lithium sulfate. Absorbance of NADPH at 340 nm was measured at 37ºC before and 5, 10 and 15 min after addition of NADPH. Results are expressed as nanomoles of NADP+ generated per minute per milligram of protein.
Western blot analysis Sciatic nerves were dissected from non-diabetic and 4- or 8-week diabetic ICR. All the animals were above 250 mg/dl of blood glucose level. Samples were sonicated and homogenized in lysis buffer (50 mM Tris-Cl ph 7.4, 150 mM NaCl, 1mM EDTA, 1% NP40, 0.25% sodiumdeoxycholate, 50mM sodium floride, 1mM sodium orthovanadate,
6
10mM sodium glycerophosphate, 5 mM pirophosphate and protease inhibitor cocktail (Roche Diagnostics, Basel, Germany)). Protein concentration was determined by binchoninic acid (BCA) Protein Assay (Pierce, Rockford, IL, USA), and 15 µg was separated on 10% acrylamide gels. Proteins were transferred in polyvinylidene fluoride (PVDF) membranes and incubated with anti-AR (1:500, kindly provided by T.G.Flynn, Queen’s University, Ontario, Canada), anti-poly-ADP-ribose modified proteins (1:1,000; Biomol International, Plymouth Meeting, PA, USA), anti-tubulin (1:1,000, Sigma, St
Louis, MO). Anti-rabbit (DakoCytomation, Glostrup, Denmark) or anti-mouse (GE Healthcare) conjugated to horseradich peroxidase (HRP) was used as secondary antibody (1:10,000). Band intensities were quantified by GeneSnap software for Gene Genius Bio Imaging System (Syngene, Cambridge, UK) and normalized by anti-tubulin levels in each line.
Functional tests Nerve conduction tests were performed in the sciatic nerve at 4 and 8 weeks after STZ administration. With animals under anesthesia (pentobarbital 40mg/kg i.p.) the nerve was stimulated percutaneously through a pair of small needle electrodes placed first at the sciatic notch and then at the ankle. Rectangular electrical pulses (Grass S88) of 0.01 ms duration were applied up to 25% above the voltage that gave a maximal response. The compound muscle action potentials (CMAPs), elicited by orthodromic conduction (M wave) and by the monosynaptic reflex arc (H wave), were recorded from the third interosseus plantar muscle and from the tibial anterior muscle with microneedle electrodes. Similarly, compound nerve action potentials (CNAP) were recorded using electrodes placed at the fourth toe, near the digital nerves. All evoked action potentials were amplified and displayed on a storage oscilloscope (Tektronix 420, Baaverton, OR) at settings appropriate to measure the amplitude from baseline to peak and the latency to the onset of the CMAP and the CNAP and to the peak of the H wave. The NCV was calculated for each segment tested: motor NCV for the sciatic
7
notch-ankle segment, sensory NCV for the proximal segment sciatic notch-ankle (SNCVp), and for the distal segment ankle-digit (SNCVd). During electrophysiological tests, the animals were placed over a warm flat steamer controlled by a hot water circulating pump, and the hind-paw skin temperature was maintained above 32ºC.
Morphological methods Tibial nerve samples were fixed in glutaraldehyde-paraformaldehyde (2.5%:2%) in 0.1 M phosphate buffer (pH 7.4) for 4-6 h at 4ºC, then post-fixed with 1% osmium tetroxide and 0.8% potassium hexacyanoferrate, washed in distilled water, dehydrated in graded series of acetone and embedded in spurr resin. Light microscopy observations were performed on 1-µm-thick sections stained with methylene blue. Morphometrical evaluation including nerve cross-sectional area, axonal counts and myelinated axon and fiber perimeters and diameters, was made from images processed at 200x (for measurement fo the cross-sectional area of the entire nerve) and at 1,000x (for myelinated fiber measurements). Printed images of systematic randomly selected fields, covering at least 300 random axons were chosen. The analysis was performed with a digital tablet (Wacom Co., Ltd., Otone, Japan) and ImageJ software (NIH, USA), drawing the inner and outer perimeter of each fiber. Area and diameter of the fiber and axon, g ratio, myelin thickness and density of myelinated fibers (MF) were calculated. Ultrathin sections (65 nm) were subsequently cut, collected on grids, stained with uranyl acetate and lead citrate, and photographed with a JEOL 1400 transmission electron microscope (Tokyo, Japan). Plantar skin pads were harvested from mice at 8 weeks of age, fixed in Zamboni’s solution overnight and cryoprotected. Cryotome pad sections 40-µm-thick were washed free-floating in PBS with 0.3% Triton-X100 and 1% normal goat serum for 1 h, then incubated in primary rabbit antisera against protein gene product 9.5 (PGP, 1:800; Ultraclone). After washes, sections were incubated in secondary antibody conjugated to Cyanine 3 and processed as described previously (Navarro et al., 1995; Verdu et al.,
8
1999). Samples were viewed under an epifluorescence microscope using appropriate filters. Five sections from each pad were used to quantify the number and density of intraepidermal nerve fibers.
Data analysis and statistical comparisons The results are shown as mean and SEM. Statistical comparisons between groups were made by one-way ANOVA with post-hoc Bonferroni test for multiple comparisons.
9
RESULTS
Metabolic results To analyze the development of diabetic neuropathy and the effect of long-term hyperglycemia on the nerve, diabetes was induced in ICR mice with three doses of STZ at post-natal days 3, 4 and 8. Males with blood glucose levels above 250 mg/ml 4 weeks after STZ injections were considered diabetic and selected for the study. Two groups of animals, four and eight weeks after the induction of diabetes, were used for electrophysiological studies and euthanized as described in the experimental design (Fig. 1a). The optimization of diabetes induction in neonatal mice is described elsewhere (Ariza et al., 2011). Body weight and blood glucose levels for control and diabetic mice are represented in Figures 1b and c, respectively. Animals in both groups grew according to the standard curves, however control mice gained up to 32% more weight than diabetic mice after weanling (maximum difference at postnatal day 33) although this difference was transient as it was gradually reduced to 15% at 2 months of age, the last time point studied. Blood glucose levels from 2-month old diabetic animals reached 600 mg/dl, the maximum glucometer capacity of glucose detection.
Aldose Reductase analysis It has been described that as a consequence of hyperglycemia, AR activity is increased in diabetic neuropathy in animal models and in human patients. Enzyme activity was measured in sciatic nerve extracts from mice of four and eight weeks of diabetes as the nanomoles of NADP+ generated per minute per milligram of protein on sciatic nerve extracts. We have previously shown that there are no changes in AR activity in DRG at one- or 8-weeks of diabetes (Homs et al., 2011). As represented in Fig. 2a and b a significant 6-fold increase compared to control mice at both ages was observed, in the range of what have been described for adult diabetic animals (Homs
10
et al., 2011). However, levels of AR protein in sciatic nerves showed no changes at both time points studied compared with control mice (Fig. 2c-f), consistent with previous studies.
Poly(ADP-ribosyl)ated proteins results Poly(ADP-ribosyl)ated proteins are an index of PARP activation in response to free radicals and oxidative stress due to the high levels of glucose. Western blot quantification indicated a 1.5-fold significant increase in diabetic mice compared with controls after 4 weeks of diabetes (Fig. 3a-c). Eight weeks after induction of diabetes with STZ, levels of poly(ADP-ribosyl)ated proteins continued increased in diabetic animals, up to 2.7 times compared with control mice (Fig. 3d), both numbers are also in the range of poly(ADP-ribosyl)ated protein increase in the same strain of adult diabetic animals (Homs et al., 2011).
Neurophysiological evaluation The electrophysiological tests were performed in control and diabetic mice 4 and 8 weeks after the induction of diabetes. In STZ-induced diabetic mice at 4 weeks sensory NCVs was 15-20% decreased in the proximal and distal nerve segments compared to age-matched control mice (Fig. 4). At eight weeks the NCV was faster, according to the progression of postnatal maturation (Verdu et al., 2000); nevertheless, diabetic mice had still about 15% decrease in sensory NCV (Fig. 4). The amplitude of the muscle and nerve compound action potentials did not show significant changes in diabetic compared with control mice at the two times evaluated. No significant changes in motor NCV were observed even if there was a tendency to decrease, more evident eight weeks after STZ injection (Fig. 4).
11
Morphological results Diabetic rats and mice induced with STZ at adulthood do not usually show significant morphometrical alterations at short or midterm of diabetes. Morphometric results of tibial nerves of diabetic and control animals are shown in Table 1. Four weeks after induction of diabetes, there was a significant decrease (4.69 ± 0.08 µm) in the MF diameter compared with the controls (5.22 ± 0.13 µm). Significant reduction of about 20% and 8% was also observed in the myelin thickness and in the number of MF, respectively (Table 1). At 8 weeks after STZ administration, diabetic mice presented a 10% reduction in MF diameter and myelin thickness. The average number of MF was 12% lower in diabetic than in control mice at that age (Table 1). The mean fiber diameter was reduced in diabetic animals while axonal diameter did not change at 4 weeks and it was reduced by 9% at 8 weeks, although numbers were not statistically significant. The histogram distribution of fiber diameters showed significant differences between control and diabetic groups, with a higher percentage of fibers smaller than 4 µm and lower percentages of fibers larger than 5 µm in the diabetic mice (Fig. 5). No presence of ischemic areas or endoneurial edema was detected during the pathological survey of the samples. No changes in the general ultrastructure, crosssectional nerve area, MF density or amyelinic fibers were observed in control and diabetic mice (Fig. 6).
Cutaneous innervation Anatomical evidence of distal nerve fiber loss is a key feature of human diabetic neuropathy (Smith et al., 2005). Plantar pads of diabetic mice showed a mild decrease of the epidermis thickness and a clear reduction in the density of PGP9.5 immunolabeling of the subepidermal nerve plexus. The density of intraepidermal nerve fibers (IENF/mm) showed a significant 42% decrease in diabetic samples (12.72 ± 1.55; p=0.003) with respect to the control animals (21.85 ± 1.17) (Fig. 7).
12
DISCUSSION
Patients with diabetic neuropathy commonly display alterations of sensory and autonomic functions, electrophysiological evidence of nerve conduction impairment and anatomical signals of nerve fiber loss, which are not frequently reproduced in animal models of diabetes (Calcutt, 2004; Sullivan et al., 2007). Herein we show that ICR mice injected with low doses of STZ during the first week of their life exhibit these abnormalities. Even if chronic hyperglycemia and duration of diabetes mellitus are the main determinants for the development of neuropathy in human patients, the pathogenesis of diabetic neuropathy is multi-factorial and a high number of disturbances appear to influence in the progression of this complication of diabetes. Different mechanisms as increased polyol flux through the AR pathway, functional and structural alterations of nerve microvessels, nerve and ganglia hypoxia, oxidative stress, endoplasmic reticulum stress, non-enzymatic glycosylation and impairment in neurotrophic support for peripheral nerves and neurons have been involved in the development of diabetic neuropathy (Yagihashi, 1995; Sima & Sugimoto, 1999; Tomlinson, 1999; Zochodne, 1999; Simmons & Feldman, 2002; Obrosova, 2009; Lupachyk et al., 2013a; Lupachyk et al., 2013b). In this study we have evaluated AR and PARP activation four weeks after induction of diabetes when hyperglycemia was just above 300 mg/dl, and eight weeks after the induction, when hyperglycemia was well established. The two enzymes activities were significantly increased at the first time point analyzed and were even higher one month later. These results correlate with our previous experiments of diabetes induced in adult ICR and NOD mice, in which the activity was increased already after 1 week of hyperglycemia (Homs et al., 2011), and suggest that probably the molecular mechanisms involved in the development of diabetic neuropathy were well established already at the initial stages of diabetes. These events have also been described in mouse models of type 2 diabetes, as pre-diabetic obese mice presented
13
alterations in the peripheral nerve sorbitol pathway and also accumulation of poly(ADPribosyl)ated proteins (Obrosova et al., 2007). Neurophysiological tests show evidence of abnormalities in nerve conduction velocity in a majority of diabetic patients even if they do not present symptoms on physical examination (Dyck et al., 1993). In most type 1 diabetic patients neurophysiological results worsen in a time-dependent manner, with decrease of NCV and later of amplitude of CMAPs and CNAPs, as well as electromyographic signs of denervation (Dyck et al., 1993; Kennedy et al., 1995). Diabetic neuropathy in animal models is characterized by peripheral nerve dysfunction that correlates with a decrease in NCV, even though the magnitude varies depending on multiple factors as the animal strain, time of diabetes, diet and methodology used. Sensory and motor NCV are reduced in STZ-induced diabetic rats (Kalichman et al., 1998; Biessels et al., 1999; Andriambeloson et al., 2006). Previous studies in diabetic mice induced with multiple low-dose STZ showed a decrease in sensory NCV in several mouse strains (Obrosova et al., 2005b; Kellogg et al., 2007), but no in C57BL6 mice (Sullivan et al., 2007) where only a single administration of a high dose of STZ induced MNCV and SNCV reduction (Christianson et al., 2003; Obrosova et al., 2005b). Diabetes induced with STZ in Swiss Wistar mice exhibit slow motor and sensory NCV and reduced amplitude of CMAPs and CNAPs that evolved with time (Kennedy & Zochodne, 2000; 2005). Our results show reduction of sensory NCV between 10 and 30% and a trend to reduction of motor NCV although not significant at least until eight weeks of the induction of diabetes, similar to what we have previously described in ICR mice rendered diabetic when adults (Homs et al., 2011). This may indicate that sensory neurons are more susceptible to hyperglycemia than motor neurons, as has been suggested by the lack of recovery of sensory but not motor neurons after long-term diabetes (9 months) in a mouse model of spontaneous recovery form diabetes (Kennedy & Zochodne, 2005).
14
Morphometrical analysis performed in our mice showed that high glucose levels induced a significant decrease in myelinated fiber but not axonal diameter, in the myelin thickness and in the number of myelinated fibers. Most previous studies in STZinduced diabetic rats and mice did not show such significant morphometrical alterations, making thus difficult to extrapolate from these experimental models to human diabetic neuropathy (Wright & Nukada, 1994; Weis et al., 1995; Malone et al., 1996; Kennedy & Zochodne, 2000; Homs et al., 2011). Here we show that decreased NCV correlated with and the decrease in fiber size and myelin thickness, as it was previously reported in human DPN (Pittenger et al., 2004; Lauria et al., 2005). Axonal loss was also evident by a marked reduction of the number of intraepidermal terminal axons in mice with diabetes from neonatal age, significantly more severe (27% vs 42%) than in the same strain of mice with diabetes induced in adulthood (Homs et al., 2011). Further studies may be done in the future using this mouse model in order to determine the long-term evolution of the neuropathy, as well as to ascertain the regeneration potential in these animals, as it has been described that additionally on the deficits of polyneuropathy, diabetic nerves of mice and rats present failures to regenerate (Ekstrom et al., 1989; Ekstrom & Tomlinson, 1989; Kennedy & Zochodne, 2005). In conclusion, we showed here that diabetes induction by STZ at the beginning of life causes peripheral neuropathy in ICR mice as early as four weeks of life. Important neurophysiological and morphological changes were observed in the nerves of these diabetic mice exhibiting susceptibility for developing neuropathy by STZ-induced diabetes. It is also remarkable that AR and PARP activities were significantly increased at four weeks of diabetes and continued increasing with time. These observations suggest that this could be a suitable model for diabetic neuropathy to test pharmacological or gene therapies, as well as to study early signs of diabetic neuropathy.
15
ACKNOWLEDGMENTS We thank Meritxell Puig, David Ramos, Angel Vázquez and Veronica Melgarejo (CBATEG, UAB), and Marta Morell (UAB) for technical assistance. LA, GP and BGL were recipients of predoctoral fellowships (GP and LA from the Generalitat de Catalunya: 2009FI_B00219 and 2006FI00762, respectively, and BGL from the Ministerio de Educación: EDU/3445/2011). This work was supported by the Instituto de Salud Carlos III (PS/09720), the Cell Therapy Network (TERCEL), the Generalitat de Catalunya (SGR 2009-1300) and the UAB (EME04-07). The authors declare that they have no conflict of interest on the subject.
16
FIGURE LEGENDS
Figure 1. Experimental design and metabolic parameters. (a) Experimental design used in this study, (b) Body weight curves for diabetic and control male mice, (c) Blood glucose levels of diabetic and control males. Values are expressed as mean ± SEM; n=18 for diabetic and n=11 for control mice, ***p
S0306-4522(14)00407-2 http://dx.doi.org/10.1016/j.neuroscience.2014.05.015 NSC 15417
To appear in:
Neuroscience
Accepted Date:
9 May 2014
Please cite this article as: L. Ariza, G. Pagès, B. García-Lareu, S. Cobianchi, P.J. Otaegui, J. Ruberte, M. Chillón, X. Navarro, A. Bosch, Experimental diabetes in neonatal mice induces early peripheral sensorimotor neuropathy, Neuroscience (2014), doi: http://dx.doi.org/10.1016/j.neuroscience.2014.05.015
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
EXPERIMENTAL DIABETES IN NEONATAL MICE INDUCES EARLY PERIPHERAL SENSORIMOTOR NEUROPATHY
Lorena Ariza1, Gemma Pagès1, Belén García-Lareu1, Stefano Cobianchi2,3, Pedro José Otaegui1, Jesús Ruberte1,4, Miguel Chillón1,5, Xavier Navarro2,3, Assumpció Bosch1
1
Center of Animal Biotechnology and Gene Therapy (CBATEG) and Department of
Biochemistry and Molecular Biology, Universitat Autònoma de Barcelona, Bellaterra, Barcelona, Spain. 2
Department of
Cell
Biology,
Physiology and Immunology and Institute of
Neurosciences, Universitat Autònoma de Barcelona, Bellaterra, Spain. 3
Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas
(CIBERNED), Instituto de Salud Carlos III, Spain. 4
Department of Animal Health and Anatomy, Veterinary School, Universitat Autònoma
de Barcelona, Bellaterra, Barcelona, Spain. 5
Institut Català de Recerca i Estudis Avançats (ICREA), Barcelona, Spain.
Correspondence should be addressed to A.B. ([email protected]): CBATEG, Edifici H, Campus Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain. Telephone: 34 93 5814203; FAX: 34 93 5814205
Short title: Neonatal diabetes induces early peripheral neuropathy
1
ABSTRACT
Animal models of diabetes do not reach the severity of human diabetic neuropathy but relatively mild neurophysiological deficits and minor morphometric changes. The lack of degenerative neuropathy in diabetic rodent models seems to be a consequence of the shorter length of the axons or the shorter animal life span. Myelin proteins have a median half-life of 200 days, thus, diabetes-induced demyelination needs many weeks or even months before it can be evident by morphometrical analysis. In mice myelination of the peripheral nervous system starts at the prenatal period and it is complete several days after birth. Here we induced experimental diabetes to neonatal mice and we evaluated its effect on the peripheral nerve 4 and 8 weeks after diabetes induction. Neurophysiological values showed a decline in sensory nerve conduction velocity at both time-points. Morphometrical analysis of the tibial nerve demonstrated a decrease in fiber size and myelin thickness, and myelinated fiber loss at both timepoints studied. Moreover, aldose reductase and poly(ADP-ribose) polymerase activities were increased even if the amount of the enzyme was not affected. Thus, type 1 diabetes in newborn mice induces early peripheral neuropathy and may be a good model to assay pharmacological or gene therapy strategies to treat diabetic neuropathy.
Key words: diabetic neuropathy, neonatal mice, myelination, aldose reductase activity, PARP activity Abbreviations AF, amyelinic fibers; AR, aldose reductase; CMAP, compound muscle action potential; CNAP, compound nerve action potential; DPN, peripheral polyneuropathy; MF, myelinated
fibers;
NCV,
nerve
conduction
velocity;
PARP,
poly-ADP-ribose
polymerase; SC, Schwann cells; STZ, streptozotocin.
2
INTRODUCTION
Peripheral polyneuropathy (DPN) is one of the most common long-term secondary complications of diabetes mellitus and affects both somatic sensory and autonomic nerves caused by hyperglycemia/hypoinsulinemia. Between the characteristic pathological features of human DPN are axonal degeneration, nerve fiber loss, segmental demyelination and problems in remyelination. The prevalence in humans is about 60% of the patients but electrophysiological alterations can be demonstrated in almost 100% of diabetic patients even if the pathology is subclinical (AguilarRebolledo, 2005). The molecular mechanisms involved in the pathogenesis of diabetic complications are multiple. Among them, acceleration of the polyol pathway due to the high levels of glucose was one the first factors suggested. This pathway is composed of two enzymes, aldose reductase (AR) that is highly expressed in Schwann cells (SC) and converts glucose to sorbitol using NADPH as a cofactor, and sorbitol dehydrogenase that converts sorbitol to fructose using NAD+. High levels of sorbitol induce damage in nervous tissue and, if NAPDH is not available in SC, the amount of reduced glutathione decreases and there is an increased susceptibility to intracellular oxidative stress (Suzuki et al., 1999; Oka & Kato, 2001; Brownlee, 2005). Slowed conduction of action potentials is triggered by both, polyol accumulation which decreases Na+,K+-ATPase activity and leads to intra-axonal sodium accumulation (Mandersloot et al., 1978), and by paranodal and segmental demyelination (Sima & Sugimoto, 1999). On the other hand, under hyperglycemic conditions there is an increase of cell metabolism and of reactive oxygen species production that may alter the mitochondrial membrane potential. Under these conditions the enzyme poly-ADP-ribose polymerase (PARP) is activated, a fundamental mechanism in the development of diabetic complications, including endothelial dysfunction, cardiomyopathy, retinopathy, nephropathy and
3
neuropathy (Garcia Soriano et al., 2001; Pacher et al., 2002; Obrosova et al., 2005a; Drel et al., 2006). Understanding the biochemical mechanisms underlying DPN requires the use of appropriate animal models that reproduce the main features of the human disease (Sullivan et al., 2008). Among them, streptozotocin (STZ) administration to induce type 1 diabetes is one of the most used models for diabetic neuropathy (Kimura et al., 2005; Hong & Kang, 2008; Jang et al., 2010; Sanchez-Zamora et al., 2010). STZ is a drug with diabetogenic properties due to its specific toxicity to pancreatic beta cells. Two procedures have been described to induce diabetes with this antibiotic, using either a single high dose or multiple low doses of STZ. The latter treatment induces subtoxic effects on beta cells and resembles type 1 diabetes due to slow progressive hyperglycemia and lymphocytic infiltration in Langerhans islets. Animals with STZinduced diabetes present termal hyperalgesia, reduced intraepidermal inervation and hypoalgesia at the long-term as neuropathic signs (Cameron et al., 2001; Calcutt, 2004; Li et al., 2005). They also show decreased nerve conduction velocity (NCV) (Calcutt, 2004; Obrosova et al., 2005a; Obrosova, 2009). Among the genetic animal models that have been characterized as appropriate models for diabetic neuropathy, the B6Ins2Akita mice and the BB/Wor (bio-breeding Worcester) mice for diabetes type 1 present deficits in conduction velocity (Choeiri et al., 2002; Stevens et al., 2004; Kamiya et al., 2005; Sullivan et al., 2007), and the most used db/db and ob/ob mice show decreased thermal sensibility and conduction velocity as models for type 2 diabetes (Sullivan et al., 2007; Vareniuk et al., 2007; Vincent et al., 2007). Unfortunately, animal models of diabetes do not commonly reach the severity of human DPN. Animal nerves usually show relatively mild neurophysiological deficits and minor morphological changes, thus limiting the relevance of experimental studies (Sharma & Thomas, 1987; Wright & Nukada, 1994). The lack of degenerative neuropathy in diabetic rodent models seems to be a consequence of the short life span of rodents or the physically shorter axons. Degenerative neuropathy is minimal even in
4
larger animals like dogs or primates with the exception of cats (Mizisin et al., 2007). Demyelination-induced by diabetic neuropathy is hardly detected in animal models as myelin proteins have a half-life longer than 200 days. As an alternative, combination of diabetes and injury has been used to exacerbate the sensorymotor neuropathy since diabetic patients and animal models show impaired nerve regeneration (Kennedy & Zochodne, 2000; Homs et al., 2011). However, in this case it is difficult to distinguish the contribution of diabetic neuropathy from the influence due to nerve injury, and therapeutic strategies that may work in DPN may not necessarily be effective in nerve regeneration and vice-versa. Type 1 diabetes in humans can develop as early as in infanthood and often affects young children and adolescents. Here, by inducing experimental diabetes to neonatal mice, we aimed to produce hyperglycemia during the period when the PNS is most actively myelinating and to study whether experimental diabetes in newborn mice was able to cause severe DPN. We analyzed diabetes induced during the first week of age affectes the peripheral nerve by biochemical, electrophysiological and morphological studies at 4 and 8 weeks after diabetes induction.
5
MATERIALS AND METHODS
Animals Male newborn Hsd:ICR(CD-1) mice received 3 intraperitoneal injections of 40mg/Kg of STZ (dissolved in 0.1 mol/l citrate buffer, pH 4.5, immediately before administration) at post-natal day 3 (P3), P4 and P8. Control animals were injected with citrate buffer at the same ages. Body weight and blood glucose were measured weekly after weaning with a Glucometer Elite (Bayer, Leverkusen, Germany). Breading and weaned mice were fed ad libitum with a standard diet (2018S Teklad Global, Harlan Laboratories; 17% calories from fat) and kept under a light–dark cycle of 12 hr (lights on at 8:00 am). Mice were euthanized at 4 and 8 weeks after the STZ administration. Animal care and experimental procedures were approved by the Biosafety and the Animal and Human Experimentation Ethical Committees of the Universitat Autònoma de Barcelona.
AR activity in sciatic nerve protein extracts AR activity was spectrophotometrically measured following the decrease of NADPH during the reduction of glyceraldehydes (both from Sigma, St Louis, MO) as published (Sato, 1992). Briefly, the reaction was started by adding 100 mM NADPH to 100 µg of sciatic nerve protein extract in 100 mM sodium phosphate buffer containing 1mM D-Lglyceraldehyde and 400 mM lithium sulfate. Absorbance of NADPH at 340 nm was measured at 37ºC before and 5, 10 and 15 min after addition of NADPH. Results are expressed as nanomoles of NADP+ generated per minute per milligram of protein.
Western blot analysis Sciatic nerves were dissected from non-diabetic and 4- or 8-week diabetic ICR. All the animals were above 250 mg/dl of blood glucose level. Samples were sonicated and homogenized in lysis buffer (50 mM Tris-Cl ph 7.4, 150 mM NaCl, 1mM EDTA, 1% NP40, 0.25% sodiumdeoxycholate, 50mM sodium floride, 1mM sodium orthovanadate,
6
10mM sodium glycerophosphate, 5 mM pirophosphate and protease inhibitor cocktail (Roche Diagnostics, Basel, Germany)). Protein concentration was determined by binchoninic acid (BCA) Protein Assay (Pierce, Rockford, IL, USA), and 15 µg was separated on 10% acrylamide gels. Proteins were transferred in polyvinylidene fluoride (PVDF) membranes and incubated with anti-AR (1:500, kindly provided by T.G.Flynn, Queen’s University, Ontario, Canada), anti-poly-ADP-ribose modified proteins (1:1,000; Biomol International, Plymouth Meeting, PA, USA), anti-tubulin (1:1,000, Sigma, St
Louis, MO). Anti-rabbit (DakoCytomation, Glostrup, Denmark) or anti-mouse (GE Healthcare) conjugated to horseradich peroxidase (HRP) was used as secondary antibody (1:10,000). Band intensities were quantified by GeneSnap software for Gene Genius Bio Imaging System (Syngene, Cambridge, UK) and normalized by anti-tubulin levels in each line.
Functional tests Nerve conduction tests were performed in the sciatic nerve at 4 and 8 weeks after STZ administration. With animals under anesthesia (pentobarbital 40mg/kg i.p.) the nerve was stimulated percutaneously through a pair of small needle electrodes placed first at the sciatic notch and then at the ankle. Rectangular electrical pulses (Grass S88) of 0.01 ms duration were applied up to 25% above the voltage that gave a maximal response. The compound muscle action potentials (CMAPs), elicited by orthodromic conduction (M wave) and by the monosynaptic reflex arc (H wave), were recorded from the third interosseus plantar muscle and from the tibial anterior muscle with microneedle electrodes. Similarly, compound nerve action potentials (CNAP) were recorded using electrodes placed at the fourth toe, near the digital nerves. All evoked action potentials were amplified and displayed on a storage oscilloscope (Tektronix 420, Baaverton, OR) at settings appropriate to measure the amplitude from baseline to peak and the latency to the onset of the CMAP and the CNAP and to the peak of the H wave. The NCV was calculated for each segment tested: motor NCV for the sciatic
7
notch-ankle segment, sensory NCV for the proximal segment sciatic notch-ankle (SNCVp), and for the distal segment ankle-digit (SNCVd). During electrophysiological tests, the animals were placed over a warm flat steamer controlled by a hot water circulating pump, and the hind-paw skin temperature was maintained above 32ºC.
Morphological methods Tibial nerve samples were fixed in glutaraldehyde-paraformaldehyde (2.5%:2%) in 0.1 M phosphate buffer (pH 7.4) for 4-6 h at 4ºC, then post-fixed with 1% osmium tetroxide and 0.8% potassium hexacyanoferrate, washed in distilled water, dehydrated in graded series of acetone and embedded in spurr resin. Light microscopy observations were performed on 1-µm-thick sections stained with methylene blue. Morphometrical evaluation including nerve cross-sectional area, axonal counts and myelinated axon and fiber perimeters and diameters, was made from images processed at 200x (for measurement fo the cross-sectional area of the entire nerve) and at 1,000x (for myelinated fiber measurements). Printed images of systematic randomly selected fields, covering at least 300 random axons were chosen. The analysis was performed with a digital tablet (Wacom Co., Ltd., Otone, Japan) and ImageJ software (NIH, USA), drawing the inner and outer perimeter of each fiber. Area and diameter of the fiber and axon, g ratio, myelin thickness and density of myelinated fibers (MF) were calculated. Ultrathin sections (65 nm) were subsequently cut, collected on grids, stained with uranyl acetate and lead citrate, and photographed with a JEOL 1400 transmission electron microscope (Tokyo, Japan). Plantar skin pads were harvested from mice at 8 weeks of age, fixed in Zamboni’s solution overnight and cryoprotected. Cryotome pad sections 40-µm-thick were washed free-floating in PBS with 0.3% Triton-X100 and 1% normal goat serum for 1 h, then incubated in primary rabbit antisera against protein gene product 9.5 (PGP, 1:800; Ultraclone). After washes, sections were incubated in secondary antibody conjugated to Cyanine 3 and processed as described previously (Navarro et al., 1995; Verdu et al.,
8
1999). Samples were viewed under an epifluorescence microscope using appropriate filters. Five sections from each pad were used to quantify the number and density of intraepidermal nerve fibers.
Data analysis and statistical comparisons The results are shown as mean and SEM. Statistical comparisons between groups were made by one-way ANOVA with post-hoc Bonferroni test for multiple comparisons.
9
RESULTS
Metabolic results To analyze the development of diabetic neuropathy and the effect of long-term hyperglycemia on the nerve, diabetes was induced in ICR mice with three doses of STZ at post-natal days 3, 4 and 8. Males with blood glucose levels above 250 mg/ml 4 weeks after STZ injections were considered diabetic and selected for the study. Two groups of animals, four and eight weeks after the induction of diabetes, were used for electrophysiological studies and euthanized as described in the experimental design (Fig. 1a). The optimization of diabetes induction in neonatal mice is described elsewhere (Ariza et al., 2011). Body weight and blood glucose levels for control and diabetic mice are represented in Figures 1b and c, respectively. Animals in both groups grew according to the standard curves, however control mice gained up to 32% more weight than diabetic mice after weanling (maximum difference at postnatal day 33) although this difference was transient as it was gradually reduced to 15% at 2 months of age, the last time point studied. Blood glucose levels from 2-month old diabetic animals reached 600 mg/dl, the maximum glucometer capacity of glucose detection.
Aldose Reductase analysis It has been described that as a consequence of hyperglycemia, AR activity is increased in diabetic neuropathy in animal models and in human patients. Enzyme activity was measured in sciatic nerve extracts from mice of four and eight weeks of diabetes as the nanomoles of NADP+ generated per minute per milligram of protein on sciatic nerve extracts. We have previously shown that there are no changes in AR activity in DRG at one- or 8-weeks of diabetes (Homs et al., 2011). As represented in Fig. 2a and b a significant 6-fold increase compared to control mice at both ages was observed, in the range of what have been described for adult diabetic animals (Homs
10
et al., 2011). However, levels of AR protein in sciatic nerves showed no changes at both time points studied compared with control mice (Fig. 2c-f), consistent with previous studies.
Poly(ADP-ribosyl)ated proteins results Poly(ADP-ribosyl)ated proteins are an index of PARP activation in response to free radicals and oxidative stress due to the high levels of glucose. Western blot quantification indicated a 1.5-fold significant increase in diabetic mice compared with controls after 4 weeks of diabetes (Fig. 3a-c). Eight weeks after induction of diabetes with STZ, levels of poly(ADP-ribosyl)ated proteins continued increased in diabetic animals, up to 2.7 times compared with control mice (Fig. 3d), both numbers are also in the range of poly(ADP-ribosyl)ated protein increase in the same strain of adult diabetic animals (Homs et al., 2011).
Neurophysiological evaluation The electrophysiological tests were performed in control and diabetic mice 4 and 8 weeks after the induction of diabetes. In STZ-induced diabetic mice at 4 weeks sensory NCVs was 15-20% decreased in the proximal and distal nerve segments compared to age-matched control mice (Fig. 4). At eight weeks the NCV was faster, according to the progression of postnatal maturation (Verdu et al., 2000); nevertheless, diabetic mice had still about 15% decrease in sensory NCV (Fig. 4). The amplitude of the muscle and nerve compound action potentials did not show significant changes in diabetic compared with control mice at the two times evaluated. No significant changes in motor NCV were observed even if there was a tendency to decrease, more evident eight weeks after STZ injection (Fig. 4).
11
Morphological results Diabetic rats and mice induced with STZ at adulthood do not usually show significant morphometrical alterations at short or midterm of diabetes. Morphometric results of tibial nerves of diabetic and control animals are shown in Table 1. Four weeks after induction of diabetes, there was a significant decrease (4.69 ± 0.08 µm) in the MF diameter compared with the controls (5.22 ± 0.13 µm). Significant reduction of about 20% and 8% was also observed in the myelin thickness and in the number of MF, respectively (Table 1). At 8 weeks after STZ administration, diabetic mice presented a 10% reduction in MF diameter and myelin thickness. The average number of MF was 12% lower in diabetic than in control mice at that age (Table 1). The mean fiber diameter was reduced in diabetic animals while axonal diameter did not change at 4 weeks and it was reduced by 9% at 8 weeks, although numbers were not statistically significant. The histogram distribution of fiber diameters showed significant differences between control and diabetic groups, with a higher percentage of fibers smaller than 4 µm and lower percentages of fibers larger than 5 µm in the diabetic mice (Fig. 5). No presence of ischemic areas or endoneurial edema was detected during the pathological survey of the samples. No changes in the general ultrastructure, crosssectional nerve area, MF density or amyelinic fibers were observed in control and diabetic mice (Fig. 6).
Cutaneous innervation Anatomical evidence of distal nerve fiber loss is a key feature of human diabetic neuropathy (Smith et al., 2005). Plantar pads of diabetic mice showed a mild decrease of the epidermis thickness and a clear reduction in the density of PGP9.5 immunolabeling of the subepidermal nerve plexus. The density of intraepidermal nerve fibers (IENF/mm) showed a significant 42% decrease in diabetic samples (12.72 ± 1.55; p=0.003) with respect to the control animals (21.85 ± 1.17) (Fig. 7).
12
DISCUSSION
Patients with diabetic neuropathy commonly display alterations of sensory and autonomic functions, electrophysiological evidence of nerve conduction impairment and anatomical signals of nerve fiber loss, which are not frequently reproduced in animal models of diabetes (Calcutt, 2004; Sullivan et al., 2007). Herein we show that ICR mice injected with low doses of STZ during the first week of their life exhibit these abnormalities. Even if chronic hyperglycemia and duration of diabetes mellitus are the main determinants for the development of neuropathy in human patients, the pathogenesis of diabetic neuropathy is multi-factorial and a high number of disturbances appear to influence in the progression of this complication of diabetes. Different mechanisms as increased polyol flux through the AR pathway, functional and structural alterations of nerve microvessels, nerve and ganglia hypoxia, oxidative stress, endoplasmic reticulum stress, non-enzymatic glycosylation and impairment in neurotrophic support for peripheral nerves and neurons have been involved in the development of diabetic neuropathy (Yagihashi, 1995; Sima & Sugimoto, 1999; Tomlinson, 1999; Zochodne, 1999; Simmons & Feldman, 2002; Obrosova, 2009; Lupachyk et al., 2013a; Lupachyk et al., 2013b). In this study we have evaluated AR and PARP activation four weeks after induction of diabetes when hyperglycemia was just above 300 mg/dl, and eight weeks after the induction, when hyperglycemia was well established. The two enzymes activities were significantly increased at the first time point analyzed and were even higher one month later. These results correlate with our previous experiments of diabetes induced in adult ICR and NOD mice, in which the activity was increased already after 1 week of hyperglycemia (Homs et al., 2011), and suggest that probably the molecular mechanisms involved in the development of diabetic neuropathy were well established already at the initial stages of diabetes. These events have also been described in mouse models of type 2 diabetes, as pre-diabetic obese mice presented
13
alterations in the peripheral nerve sorbitol pathway and also accumulation of poly(ADPribosyl)ated proteins (Obrosova et al., 2007). Neurophysiological tests show evidence of abnormalities in nerve conduction velocity in a majority of diabetic patients even if they do not present symptoms on physical examination (Dyck et al., 1993). In most type 1 diabetic patients neurophysiological results worsen in a time-dependent manner, with decrease of NCV and later of amplitude of CMAPs and CNAPs, as well as electromyographic signs of denervation (Dyck et al., 1993; Kennedy et al., 1995). Diabetic neuropathy in animal models is characterized by peripheral nerve dysfunction that correlates with a decrease in NCV, even though the magnitude varies depending on multiple factors as the animal strain, time of diabetes, diet and methodology used. Sensory and motor NCV are reduced in STZ-induced diabetic rats (Kalichman et al., 1998; Biessels et al., 1999; Andriambeloson et al., 2006). Previous studies in diabetic mice induced with multiple low-dose STZ showed a decrease in sensory NCV in several mouse strains (Obrosova et al., 2005b; Kellogg et al., 2007), but no in C57BL6 mice (Sullivan et al., 2007) where only a single administration of a high dose of STZ induced MNCV and SNCV reduction (Christianson et al., 2003; Obrosova et al., 2005b). Diabetes induced with STZ in Swiss Wistar mice exhibit slow motor and sensory NCV and reduced amplitude of CMAPs and CNAPs that evolved with time (Kennedy & Zochodne, 2000; 2005). Our results show reduction of sensory NCV between 10 and 30% and a trend to reduction of motor NCV although not significant at least until eight weeks of the induction of diabetes, similar to what we have previously described in ICR mice rendered diabetic when adults (Homs et al., 2011). This may indicate that sensory neurons are more susceptible to hyperglycemia than motor neurons, as has been suggested by the lack of recovery of sensory but not motor neurons after long-term diabetes (9 months) in a mouse model of spontaneous recovery form diabetes (Kennedy & Zochodne, 2005).
14
Morphometrical analysis performed in our mice showed that high glucose levels induced a significant decrease in myelinated fiber but not axonal diameter, in the myelin thickness and in the number of myelinated fibers. Most previous studies in STZinduced diabetic rats and mice did not show such significant morphometrical alterations, making thus difficult to extrapolate from these experimental models to human diabetic neuropathy (Wright & Nukada, 1994; Weis et al., 1995; Malone et al., 1996; Kennedy & Zochodne, 2000; Homs et al., 2011). Here we show that decreased NCV correlated with and the decrease in fiber size and myelin thickness, as it was previously reported in human DPN (Pittenger et al., 2004; Lauria et al., 2005). Axonal loss was also evident by a marked reduction of the number of intraepidermal terminal axons in mice with diabetes from neonatal age, significantly more severe (27% vs 42%) than in the same strain of mice with diabetes induced in adulthood (Homs et al., 2011). Further studies may be done in the future using this mouse model in order to determine the long-term evolution of the neuropathy, as well as to ascertain the regeneration potential in these animals, as it has been described that additionally on the deficits of polyneuropathy, diabetic nerves of mice and rats present failures to regenerate (Ekstrom et al., 1989; Ekstrom & Tomlinson, 1989; Kennedy & Zochodne, 2005). In conclusion, we showed here that diabetes induction by STZ at the beginning of life causes peripheral neuropathy in ICR mice as early as four weeks of life. Important neurophysiological and morphological changes were observed in the nerves of these diabetic mice exhibiting susceptibility for developing neuropathy by STZ-induced diabetes. It is also remarkable that AR and PARP activities were significantly increased at four weeks of diabetes and continued increasing with time. These observations suggest that this could be a suitable model for diabetic neuropathy to test pharmacological or gene therapies, as well as to study early signs of diabetic neuropathy.
15
ACKNOWLEDGMENTS We thank Meritxell Puig, David Ramos, Angel Vázquez and Veronica Melgarejo (CBATEG, UAB), and Marta Morell (UAB) for technical assistance. LA, GP and BGL were recipients of predoctoral fellowships (GP and LA from the Generalitat de Catalunya: 2009FI_B00219 and 2006FI00762, respectively, and BGL from the Ministerio de Educación: EDU/3445/2011). This work was supported by the Instituto de Salud Carlos III (PS/09720), the Cell Therapy Network (TERCEL), the Generalitat de Catalunya (SGR 2009-1300) and the UAB (EME04-07). The authors declare that they have no conflict of interest on the subject.
16
FIGURE LEGENDS
Figure 1. Experimental design and metabolic parameters. (a) Experimental design used in this study, (b) Body weight curves for diabetic and control male mice, (c) Blood glucose levels of diabetic and control males. Values are expressed as mean ± SEM; n=18 for diabetic and n=11 for control mice, ***p
