original article receptor for advanced glycation...

Receptor for Advanced Glycation End Products (RAGEs)and Experimental Diabetic NeuropathyCory Toth,

1Ling Ling Rong,

2Christina Yang,

1Jose Martinez,

1Fei Song,

2Noor Ramji,

1

Valentine Brussee,1

Wei Liu,1

Jeff Durand,1

Minh Dang Nguyen,1

Ann Marie Schmidt,2

and Douglas W. Zochodne1

OBJECTIVE—Heightened expression of the receptor for ad-vanced glycation end products (RAGE) contributes to develop-ment of systemic diabetic complications, but its contribution todiabetic neuropathy is uncertain. We studied experimental dia-betic neuropathy and its relationship with RAGE expressionusing streptozotocin-induced diabetic mice including a RAGE�/�

cohort exposed to long-term diabetes compared with littermateswithout diabetes.

RESEARCH DESIGN AND METHODS—Structural indexes ofneuropathy were addressed with serial (1, 3, 5, and 9 months ofexperimental diabetes) electrophysiological and quantitativemorphometric analysis of dorsal root ganglia (DRG), peripheralnerve, and epidermal innervation. RAGE protein and mRNAlevels in DRG, peripheral nerve, and epidermal terminals wereassessed in WT and RAGE�/� mice, with and without diabetes.The correlation of RAGE activation with nuclear factor (NF)-�Band protein kinase C �II (PKC�II) protein and mRNA expressionwas also determined.

RESULTS—Diabetic peripheral epidermal axons, sural axons,Schwann cells, and sensory neurons within ganglia developeddramatic and cumulative rises in RAGE mRNA and protein alongwith progressive electrophysiological and structural abnormali-ties. RAGE�/� mice had attenuated structural features of neu-ropathy after 5 months of diabetes. RAGE-mediated signalingpathway activation for NF-�B and PKC�II pathways was mostevident among Schwann cells in the DRG and peripheral nerve.

CONCLUSIONS—In a long-term model of experimental diabe-tes resembling human diabetic peripheral neuropathy, RAGEexpression in the peripheral nervous system rises cumulativelyand relates to progressive pathological changes. Mice lackingRAGE have attenuated features of neuropathy and limited acti-vation of potentially detrimental signaling pathways. Diabetes57:1002–1017, 2008

Several peripheral nervous system (PNS) abnor-malities complicate diabetes. In particular, theseabnormalities are greatest within the somatic andautonomic nerves, leading to increased morbidity

and mortality within the human diabetic population (1).The most common form of somatic nerve disease in diabeticsubjects is a diabetic symmetric sensorimotor polyneurop-athy (2). Pathological structural changes within the dia-betic PNS include morphological and functional changeswithin peripheral nerve axons, the dorsal root ganglion,and epidermal nerve fibers (3,4). Chronic hyperglycemiahas a robust association with the development of compli-cations in long-term diabetes, as identified during clinicalintervention trials in both type 1 and type 2 diabetes (5,6).Mechanisms believed to be relevant to the pathogenesis ofdiabetic symmetric sensorimotor polyneuropathy (7–9)have included 1) excessive sorbitol-aldose reductase path-way flux, 2) protein kinase C (PKC) isoform(s) overactiv-ity, 3) increased oxidative and nitrative stress, 4) growthfactor deficiency, and 5) microangiopathy. The increasednonenzymatic glycation of proteins, leading to irreversibleformation and deposition of reactive advanced glycationend products (AGEs), may similarly lead to critical abnor-malities within the diabetic PNS. AGEs can be detectedthroughout the central nervous system and PNS in nondi-abetic subjects (10). Most importantly, the receptor forAGEs (RAGE) has been demonstrated on hematopoieticcells and endothelial cells, as well as spinal motor neuronsand cortical neurons (11,12). RAGE has been postulated tocontribute to the development of diabetic complications(13,14), but its role in progressive PNS dysfunction indiabetic peripheral neuropathy is uncertain.

We addressed the pattern and extent of RAGE expres-sion serially in a long-term mouse model of experimentaldiabetes with electrophysiological and structural featuresof neuropathy that parallel human disease (15). We ad-dressed RAGE mRNA and protein expression within threelevels of the PNS and identified the role of downstreamRAGE pathway activation during time periods in whichstructural and electrophysiological features of neuropathywere evolving. In a separate mouse cohort, we then ad-dressed the potential significance of such expression bystudying neuropathy in mice lacking RAGE. Our studiesprovide evidence of substantial RAGE expression and signal-ing involving neurons, axons, and glial cells during the courseof diabetic symmetric sensorimotor polyneuropathy.

RESEARCH DESIGN AND METHODS

Animals. We studied 128 male Swiss Webster wild-type (WT) mice with initialweight of 20–30 g, housed in plastic sawdust-covered cages with a normallight-dark cycle and free access to mouse food and water. All protocols were

From the 1Department of Clinical Neurosciences and the Hotchkiss BrainInstitute, University of Calgary, Calgary, Alberta, Canada; and the 2Depart-ment of Surgery, Columbia University, New York, New York.

Address correspondence and reprint requests to Dr. C. Toth, University ofCalgary, Department of Clinical Neurosciences, Room 155, 3330 Hospital Dr.,N.W., Calgary, Alberta T2N 4N1, Canada. E-mail: [email protected].

Received for publication 28 March 2007 and accepted in revised form 18November 2007.

Published ahead of print at http://diabetes.diabetesjournals.org on 26 No-vember 2007. DOI: 10.2337/db07-0339.

Additional information for this article can be found in an online appendix athttp://dx.doi.org/10.2337/db07-0339.

AGE, advanced glycation end product; CML, NH2-(carboxymethyl)lysine;DRG, dorsal root ganglia; NF, nuclear factor; PBT, PBS with 0.1% Tween-20;PKC, protein kinase C; PNS, peripheral nervous system; RAGE, receptor forAGEs; STZ, streptozotocin.

© 2008 by the American Diabetes Association.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked “advertisement” in accordance

with 18 U.S.C. Section 1734 solely to indicate this fact.

ORIGINAL ARTICLE

1002 DIABETES, VOL. 57, APRIL 2008©

RETRACTED ARTICLE

reviewed and approved by the University of Calgary Animal Care Committeeusing the Canadian Council of Animal Care guidelines. Mice were anesthetizedwith pentobarbital (60 mg/kg) before all procedures. At the age of 1 month, 82mice were injected with streptozotocin (STZ) (Sigma, St. Louis, MO) intra-peritoneally for each of 3 consecutive days, with once-daily doses of 60, 50,and then 40 mg/kg, with the remaining 46 mice injected with carrier (sodiumcitrate) for three consecutive days. Whole blood glucose measurements wereperformed monthly following electrophysiological testing using the tail veinand a blood glucose meter (OneTouch Ultra Meter; LifeScan Canada, Burnaby,BC, Canada). Hyperglycemia was verified 1 week later by sampling from a tailvein. A fasting whole blood glucose level of �16 mmol/l (normal 5–8 mmol/l)was our criterion for experimental diabetes. All animals had whole bloodglucose sampling monthly and animals were weighed monthly as well.Animals were followed for an additional 1, 3, 5, or 8 months of diabetes (up to9 months of life). Male and female RAGE�/� mice were constructed on anSVE129�C57BL/6 background (129/B6) as previously described (16) and werebackcrossed �10 generations into C57BL/6 before enrollment in studies. WTC57BL/6 mice were used as control littermates in these experiments.RAGE�/� mice were also injected with STZ as above (n � 15) or carrier (n �11) and, along with control RAGE�/� nondiabetic littermates and WT diabeticand nondiabetic littermates, were followed for 1, 3, or 5 months of diabetes.In all cases, those mice that did not develop diabetes as defined above afterSTZ injections were excluded from further assessment.

The main cohort of diabetic mice was injected, followed, and harvested atthe University of Calgary. RAGE�/� mice were injected, followed, andharvested at Columbia University. In all cases, mice were raised and studiedin strict pathogen-free environments. Tissues from all mice, irrespective oftheir origin, were amassed, identically processed, and appraised at one site(Calgary), with the measurements carried out by the investigator blinded totheir origin (C.T.).Electrophysiology. Electrophysiological assessment of sciatic nerve conduc-tion was performed as previously described (17) under halothane anesthesia.Initial baseline studies were carrier out before STZ or carrier injection andidentified no significant differences between groups studied. For sensoryconduction studies, the tibial nerve was used with a fixed distance of 30 mmfrom stimulation electrodes to the sciatic notch, where recording electrodeswere placed to measure the sensory nerve action potential amplitude andsensory nerve conduction velocity in orthodromic fashion. All stimulating andrecording electrodes were platinum subdermal needle electrodes (GrassInstruments, Astro-Med, West Warwick, RI), with near-nerve temperature keptconstant at 37 � 0.5°C using a heating lamp. The cohort of WT and RAGE�/�

mice, both diabetic and nondiabetic, underwent electrophysiological testingafter 5 months of diabetes, whereas WT mice were assessed monthly until 8months of diabetes.Tissue harvesting. After 1, 3, 5, or 9 months of diabetes, mice were killed andthe following tissues were harvested: bilateral L3–L6 dorsal root ganglia(DRG), sciatic nerves, sural nerves, and hind and fore footpads. Blood forglycated hemoglobin measurements was taken before death. One-half of alltissues were placed either in Zamboni’s fixative (left-sided tissues) for laterimmunohistochemistry or were fixed in cacodylate-buffered glutaraldehydeand then cacodylate buffer for later epon embedding for morphometricstudies. The remaining tissues (right-sided tissues) were immediately fresh-frozen at 80°C or placed in Trizol fixative (Life Technologies, Rockville, MD)and stored at �80°C. Additional tissues to be used as control samplesincluding brain, spinal cord, liver, and pancreas were removed and fixed insimilar manners.

RAGE�/� mouse tissues and corresponding littermate tissues were har-vested after 1, 3, and 5 months of diabetes and were placed in Zamboni’sfixative. Bilateral L3–L6 DRG, sciatic nerves, sural nerves, and hind and forefootpads were harvested as above.Peripheral nerve, DRG, and epidermal innervation. For peripheral nerveand DRG specimens, samples were fixed in 2.5% glutaraldehyde in 0.025 mol/l

cacodylate buffer overnight, before washing with repeated 0.015 M cacodylateand glucose-buffered cacodylate, progressive dehydration with alcohols andpropylenoxide, staining with osmium tetroxide, and placement in epon.Semi-thin (1-�m) sections of peripheral nerve and L4–L6 DRGs were cut on anultramicrotome (Reichert, Vienna, Austria) and were stained with 0.5%toluidine blue. Image analysis was performed by a single examiner blinded tothe origin of the sections (Zeiss Axioskope at 400� and 1,000� magnificationusing Scion Image v.4.0.2 [Scion, Fredrick, MD]) with measurements of thenumber, axonal area, and myelin thickness of all myelinated fibers within 25nonadjacent transverse nerve sections. Secondary measurements includedmacrophage number, degenerating profile number, and regenerating fibercluster number. For DRGs, neurons with visible nuclei were used for countingwithin a sized area for 25 nonadjacent sections separated by 300 �m foreach L4–L6 DRG for neuronal density and to provide calculations for totalestimated neuronal counts.

For tissues designated for immunohistochemistry, specimens were fixed in2% Zamboni’s fixative overnight at 4°C, washed in PBS, kept overnight in 25%glucose PBS solution, embedded in optimal cutting temperature embeddingsolution, and stored at �80°C until sectioning. Cryostat transverse andlongitudinal nerve sections (10 �m) were placed onto poly-L-lysine– andacetone-coated slides. Sections were stained with anti-human RAGE IgGprimary antibody (supplied by Dr. A.M. Schmidt, Columbia University, NewYork, NY). Antigen retrieval was performed with slides placed in sodiumcitrate in an 80°C water bath, a PBS wash for 5 min, blocking with 10% goatserum for 1 h, and further PBS washing. Slides were incubated with primaryantibody (anti-human RAGE 1:100) overnight at 4°C. After PBS washing,secondary fluorescent antibody (anti-rabbit IgG fluorescein isothiocyanate,1:100; Zymed, San Francisco, CA) was applied with incubation for 1 h at roomtemperature.

Additional immunohistochemistry was performed using antibodies to CML[NH2-(carboxymethyl)lysine] (anti-CML monoclonal antibody, 1:200; Cosmo-Bio, Tokyo, Japan), nuclear factor (NF)-�B (anti–NF-�B p50, 1:200; Santa CruzTechnology, Santa Cruz, CA), PKC�II (anti-PKC�II, 1:200; GeneTex, SanAntonio, TX), HNE (4-hydroxy-29-nonenal) Adduct (anti-HNE Adduct, 1:200;Cedarlane Laboratories, Hornby, ON), and C-Rel (anti–C-Rel, 1:200; CellSignaling Technology, Danvers, MA) to examine RAGE-associated pathways.S100 was used to label Schwann cells (anti-S100, 1:100; Chemicon Interna-tional, Temecula, CA), NF-200 and �-tubulin were used to label peripheralnerve axons (anti-neurofilament, NF-200, 1:100; Chemicon; anti–�-tubulin,1:100; Abcam, Cambridge, MA), and MAP-2 and NeuN were used to labelneurons (anti–MAP-2, 1:100; Sigma-Aldrich, Oakville, Ontario; anti-NeuN,1:100; Chemicon International).

For these immunohistochemistry slides, secondary antibody incubationwas performed as above (anti–fluorescein isothiocyanate or bovine anti-mouse IgG Cy3, 1:100; Zymed), with the exception of the C-Rel immunostain-ing, where a peroxidase-labeled streptavidin biotin antibody (LSAB; Dako,Copenhagen, Denmark) was used for secondary detection. To perform triplelocalization, consecutive samples were collected on alternating slides. Anal-ysis used Image Pro Plus software (Image Pro Plus 5.0; MediaCybernetics,Silver Spring, MD) and Adobe Photoshop software (Adobe Photoshop 7.0,Adobe, San Jose, CA, 2002). Positive profiles were tallied on every 10-�m-thicksection on 20 random sections for each DRG or nerve, and the luminosity ofindividual neurons, glia, or nerve fibers was measured. The numbers ofneurons, axons, or glia with luminosity was classified as none-low (luminosityvalue of 0–150), moderate (150–250), or high (�250) (scale of 0–255 witharbitrary units). Neurons and glia were assessed for nuclear labeling forNF-�B recorded as well using a predetermined luminosity measurementthreshold of 150 (no units). Measurement of the activation of PKC, believed tobe related to translocation from cytosol to neurolemma, and the extent oftranslocation was assessed using a predetermined luminosity measurementthreshold of 150 (no units).

Epidermal foot pads from both hind feet and forefeet were fixed in 2%

TABLE 1Murine weights at induction of diabetes and at harvesting at months 1, 3, 5, and 8 of diabetes

n

Injection ofSTZ/carrier Month 1 Month 3 Month 5 Month 8

WT nondiabetic mice 12 25.3 � 4.1 31.5 � 4.7 39.5 � 6.2* 48.1 � 5.5* 49.6 � 5.2*WT diabetic mice 12 25.1 � 4.8 25.9 � 5.4 28.5 � 5.4 31.8 � 6.3 33.6 � 6.2RAGE null nondiabetic mice 6 26.1 � 6.5 34.1 � 6.9 40.7 � 6.9* 46.9 � 6.2*RAGE null diabetic mice 6 26.2 � 6.2 27.8 � 6.2 30.2 � 7.1 34.8 � 6.9

Data are means � SE. *Significance at P 0.05 using Student’s t testing (��0.05) with WT nondiabetic mice compared with WT diabetic miceand RAGE null nondiabetic mice compared with RAGE null diabetic mice only.

C. TOTH AND ASSOCIATES

DIABETES, VOL. 57, APRIL 2008 1003©

RETRACTED ARTICLE

FIG. 1. Nerve conduction study data for sciatic nerves in both WT and RAGE�/� mice, either with or without diabetes. There were no baselinedifferences between WT and RAGE�/� mice identified. Electrophysiological testing in WT diabetic and nondiabetic mice demonstrated both motor(A and B) and sensory impairment (C and D). Both sciatic compound motor action potential amplitudes (A) and sciatic motor nerve conductionvelocity (B) diminished after 4 months of diabetes when compared with nondiabetic littermates. Meanwhile, sensory nerve action potentialamplitudes were smaller after 3 months (C), and sensory nerve conduction velocity was slower after 2 months (D) in diabetic WT mice comparedwith nondiabetic littermate controls. A time point of 5 months of diabetes (6 months of life) was selected for comparison, and WT diabetic micehad loss of compound motor action potential amplitude for the sciatic-tibial nerve (E), as well as slowing of motor nerve conduction velocity(MNCV) for the sciatic-tibial nerve (F) compared with WT nondiabetic mice and RAGE�/� with or without diabetes. In addition, WT diabetic micehad loss of sensory nerve action potential amplitude for the tibial-sciatic nerve (G), as well as slowing of sensory nerve conduction velocity(SNCV) for the tibial-sciatic nerve (H) compared with WT nondiabetic mice and RAGE�/� with or without diabetes. Significant differences weredetermined by multiple ANOVA tests: *significant difference (P < 0.0125 using Bonferroni corrections) between groups indicated by presence ofa bar over the respective columns (n � 6–8 for WT mice and n � 4 for RAGE�/� mice).

RAGE IN DIABETIC NEUROPATHY


RETRACTED ARTICLE

Zamboni’s fixative overnight at 4°C, washed in PBS, kept overnight in 20%glucose PBS solution, embedded in optimal cutting temperature embeddingsolution, and stored at �80°C until sectioning. The 30-�m sections wereprepared using a cryostat, and samples were placed onto poly-L-lysine– andacetone-coated slides. Immunohistochemistry within foot pads was per-formed in identical fashion as described above, using anti-Human RAGE IgGprimary antibody and the secondary antibody anti-rabbit IgG fluoresceinisothiocyanate and using the panaxonal marker PGP 9.5 to identify allepidermal axons (anti-mouse PGP 9.5 antibody, 1:500; Jackson ImmunoRe-search Laboratories, West Grove, PA; and secondary antibody, anti-goat IgGCy3, 1:100; Chemicon International).

Analysis of epidermal fibers within foot pads was performed as describedpreviously (18). In brief, epidermal nerve fibers extending above the dermalpapillae were quantified per square millimeter of epidermal area. We chose touse this three-dimensional reconstruction to clearly capture all fibers presentwithin the microsection. In addition to this, we also captured linear densitiesby calculating the number of fibers within each microsection expressed as afunction of epidermal length, which has the disadvantage of calculation ofepidermal length over a nonlinear surface, as well as of a possible overcount-ing of undulating fibers identified in multiple sections within three-dimen-sional sectioning. A single Zeiss fluorescent microscope was used in all cases.

Digital photography (Zeiss) of all specimen portions was performed, andAdobe Photoshop 7.0 was used to visualize each specimen for countingpurposes. For each animal, the number of PGP 9.5-immunoreactive profileswas counted by a single observer blinded to the source of each specimen ineach of the 30-�m-thick microsections, which were performed through theentire footpad. A minimum of 200 microsections was examined in each case.In situ hybridization. In situ hybridization for I�B using digoxigenin-labeledRNA probes was performed as described previously (19). DRG cryosections(16 �m) were fixed in 4% paraformaldehyde and washed twice in PBT (PBSwith 0.1% Tween-20). After bleaching with 6% H2O2/PBT and washing, sectionswere treated with 1 mg/ml proteinase K/PBT. Then sections were re-fixed andwashed in PBT before hybridization overnight at 70°C. Sections were washedthree times in 50% formamid/5 � SSC/1% SDS at 70°C, followed by two washesin 50% formamid/2 � SSC at 65°C. Staining was visualized using an Nitro-BlueTetrazolium Chloride (NBT)/5-Bromo-4-Chloro-3�-Indolyphosphate p-Tolu-idine Salt (BCIP) (Roche, Laval, PQ) substrate.Western blot. Peripheral nerve, DRG, liver, and pancreas tissues had proteinquantified from fresh frozen samples that were maintained at �80°C for amaximum of 1 month before protein quantification studies. Tissue portions forprotein studies were homogenized using a RotorStator Homogenizer inice-cold lysis buffer (10% glycerol, 2% SDS, 25 mmol/l Tris-HCl, pH 7.4, Roche

FIG. 2. Morphological assessment of the sural and sciatic nerves from WT and RAGE�/� mice with and without diabetes using semi-thin sectionsstained with toluidine blue. An axonal area histogram from the sural nerve demonstrates a leftward shift indicating axonal atrophy of diabeticWT mice (A) (Tukey’s honestly significant difference test, *P < 0.05, with the horizontal bar indicating which portions of the WT diabeticpopulation are significantly different from the other populations), whereas RAGE�/� mice exposed to 5 months of diabetes were protected fromaxon loss. Images of the sural nerve of a RAGE�/� mouse without diabetes (B) and a RAGE�/� mouse with diabetes (C) after 6 months of age (5months of diabetes) is shown in comparison to a WT mouse without diabetes (D) and a WT mouse with diabetes (E) at the same ages. An axonalarea histogram demonstrates a more subtle leftward shift of diabetic WT mice sciatic nerve axons than was identified in sural nerves of WT micewithout diabetes (F) (Tukey’s honestly significant difference, test, NS). Images of the sciatic nerve of a RAGE�/� mouse without diabetes (G) anda RAGE�/� mouse with diabetes (H) after 6 months of age (5 months of diabetes) is shown in comparison to a WT mouse without diabetes (I)and a WT mouse with diabetes (J) at the same age points. Significant differences between age-comparable groups are demonstrated by horizontalbars after multiple one-way ANOVA tests with samples treated independently were performed (P < 0.0125 using Bonferroni corrections) (n �6–8 for WT mice and n � 4 for RAGE�/� mice). Bar � 10 �m.



RETRACTED ARTICLE

Mini-Complete Protease Inhibitors). Samples were then centrifuged at 10,000g

for 15 min. Supernatant was stored at �20°C before SDS-PAGE and immuno-blotting analysis. Equal amounts (15 �g) of protein were loaded, and sampleswere separated by SDS-PAGE using 10% polyacrylamide gels with 800 V hoursof current applied. Separated proteins were transferred onto nitrocellulosepaper (Bio-Rad) over 16 h at 200 mA in Towbin transfer buffer (25 mmol/l Tris,192 mmol/l glycine, 20% vol/vol methanol, 0.1% vol/vol SDS). Separate blotswere blocked overnight in 7.5% (wt/vol) milk (Nestle, Carnation) in TBS (50mmol/l Tris, 137 mmol/l NaCl, 51 mmol/l KCl, 0.05% [vol/vol] Tween-20).

Anti-human RAGE IgG primary antibody (1:500) anti–NF-�B p50 (1:1,000),anti-PKC�II (1:1,000), and anti–�-actin (1:100, Biogenesis, Poole, U.K.) wereapplied to separate blots. Secondary anti-rabbit or anti-mouse IgG horseradishperoxidase–linked antibody (Cell Signaling) was applied at 1:5,000 in eachcase as appropriate. Signal detection was performed by exposing of the blotto enhanced chemiluminescent reagents ECL (Amersham) for 2 min. The blotswere subsequently exposed and captured on Kodac X-OMAT K film.Quantitative RT-PCR. Total RNA was extracted from peripheral nerve,DRG, brain, spinal cord, liver, and pancreas tissue using Trizol reagent (LifeTechnologies). Total RNA (1 �g) was processed directly to cDNA synthesisusing the Superscript II Reverse Transcriptase system (Invitrogen). RAGEprimers and probe sequences were as follows: forward, 5�-GGACCCTTAGCTGGCACTTAGA-3�; backward, 5�-GAGTCCCGTCTCAGGGTGTCT-3�; andprobe, 5�-ATTCCCGATGGCAAAGAAACACTCGTG-3�. �-Actin primers andprobe sequences were as follows: forward, 5�-CCTGAGCGCAAGTACTCTGTGT-3�; backward, 5�-GCTGATCCACATCTGCTGGAA-3�; and probe, 5�-CGGTGGCTCCATCTTGGCCTCAC-3�. Cyclophyllin primer sequences were asfollows: forward, 5�-TGTGCCAGGGTGGTGACTT-3�, and backward, 5�- TCAAATTTCTCTCCGTAGATGGACTT-3�. PKC�II primers sequences were as fol-lows: forward, 5�-GGTGGCATGTAGAAAGTGCTGC-3�, and backward, 5�-CAAGCATTTTCTCTCCCGTGG-3�. NF-�Bp65 primers sequences were as

follows: forward, 5�-TGTGCGACAAGGTGCAGAAA-3�, and backward, 5�-ACAATGGCCACTTGCCGAT-3�.

RT-PCR was done using SYBR Green dye. All reactions were performed intriplicate in an ABI PRISM 7000 Sequence Detection System. Data werecalculated by the 2- CT method and are presented as the fold induction ofmRNA for RAGE in diabetic tissues normalized to cyclophyllin compared withnondiabetic tissues (defined as 1.0-fold).Analysis. All data were represented as mean � SE. The t testing or ANOVAtesting with multiple comparisons of independently assessed samples andgroups were performed as appropriate in all cases with Bonferroni correc-tions as needed.

RESULTS

Diabetes. Mice injected with STZ developed diabeteswithin 1–4 weeks after injection in �80% of animals, andin each case, diabetes was maintained over the length ofthe study. The time at which diabetes was detected wastaken as the onset of diabetes and was within 2 weeks ofthe last STZ injection in all cases (or otherwise mice werenot included in further studies) and within 1 week of STZinjection in over 90% of mice studied. WT diabetic micewere smaller than WT nondiabetic mice at 1 month afterSTZ injection, and diabetic mice had smaller body weightsthroughout life (Table 1). Although the diabetic mice weresmaller, they nevertheless continued to gain weight overthe course of the study. Mouse glycated hemoglobin wasincreased in WT diabetic mice at 9 months of life. The

TABLE 2Morphological properties of sciatic nerves in nondiabetic and diabetic nerves from WT and RAGE null mice after 1–8 months ofdiabetes

n Month 1 Month 3 Month 5 Month 8

Axonal fiber density (per mm2)WT nondiabetic mice 6 14,975 � 246 14,996 � 218 14,695 � 229 14,732 � 251WT diabetic mice 6 14,943 � 241 14,804 � 217 14,323 � 271 14,701 � 263RAGE null nondiabetic mice 4 15,057 � 291 15,022 � 274 14,911 � 288RAGE null diabetic mice 4 14,992 � 265 14,923 � 276 14,853 � 292

Axonal numbersWT nondiabetic mice 6 979 � 18 977 � 20 974 � 17 975 � 19WT diabetic mice 6 980 � 20 971 � 19 943 � 18 941 � 17RAGE null nondiabetic mice 4 982 � 22 980 � 21 979 � 20RAGE null diabetic mice 4 980 � 23 976 � 21 973 � 20

Axonal area (�m2)WT nondiabetic mice 6 47.2 � 0.5 46.1 � 0.5 42.3 � 0.6 37.5 � 0.8WT diabetic mice 6 46.8 � 0.3 44.1 � 0.4 38.7 � 0.6* 32.1 � 0.5*

RAGE null nondiabetic mice 4 48.6 � 0.5 46.3 � 0.6 43.2 � 0.9RAGE null diabetic mice 4 46.2 � 0.3 45.8 � 0.5 41.9 � 0.6

Myelination thickness (�m)WT nondiabetic mice 6 1.08 � 0.03 1.09 � 0.03 1.05 � 0.02 1.04 � 0.03WT diabetic mice 6 1.07 � 0.03 1.01 � 0.02 0.95 � 0.03* 0.91 � 0.03*

RAGE null nondiabetic mice 4 1.10 � 0.04 1.10 � 0.05 1.06 � 0.05 1.05 � 0.05RAGE null diabetic mice 4 1.10 � 0.04 1.08 � 0.04 1.02 � 0.05 1.00 � 0.04

Number of degenerating profiles(per transverse section)

WT nondiabetic mice 6 0 1 � 1 4 � 1 4 � 1WT diabetic mice 6 2 � 1 7 � 2* 12 � 2* 11 � 2*

RAGE null nondiabetic mice 4 0 0 2 � 1 2 � 1RAGE null diabetic mice 4 1 � 1 1 � 1 4 � 1 4 � 1

Number of regenerating fiber clusters(per transverse section)

WT nondiabetic mice 6 0 0 0 0WT diabetic mice 6 0 0.1 � 0.1 0.1 � 0.1 0RAGE null nondiabetic mice 4 0 0 0 0RAGE null diabetic mice 4 0 0 0.1 � 0.0 0.1 � 0.1

Data are means � SE. *Significance at P 0.05 using multiple ANOVA testing with Bonferroni post-hoc t test comparisons (� � 0.05). Datain bold are statistically significant.



RETRACTED ARTICLE

mortality rate in WT diabetic mice was significantly higherthan in controls, with only 40% of all diabetics survivingthe entire 9-month duration of diabetes. Kaplan-Meiersurvival statistics identified an increased mortality rate indiabetic mice over the duration of the experiment, withsignificant mortality present after 5 months of diabetes.RAGE�/� diabetic mice were viable and were not physi-cally different than their nondiabetic littermates. RAGE�/�

mice displayed levels of hyperglycemia similar to those ofWT diabetic mice.Peripheral nerve trunks. There were no differences inmotor or sensory nerve electrophysiological parametersbetween WT and RAGE�/� nondiabetic or diabetic micepresent at baseline. Electrophysiological recordings iden-tified dysfunction in diabetic WT mice within 5 months ofdiabetes for all sensory and motor measurements (Fig. 1),as previously described in diabetic rodent models (15).For WT diabetic mice, declines in sensory nerve conduc-tion velocities and sensory nerve action potential ampli-tudes were present after 2 months of diabetes, similar topreviously reported results (20–23). In marked contrast,RAGE�/� mice with diabetes had electrophysiologicalresults similar to RAGE�/� mice without diabetes andWT mice without diabetes after 5 months, withoutslowing of nerve conduction velocities or compoundmotor action potential and sensory nerve action poten-tial amplitudes (Fig. 1). Overall, in comparison to WT

diabetic mice, diabetic RAGE�/� mice displayed signif-icant protection against dysfunction in both motor andsensory function.

There were no differences in morphological parametersof sural or sciatic nerves between WT and RAGE�/�

nondiabetic or diabetic mice present at 1 month afterstudy initiation. The sural nerves from WT diabetic micedeveloped a loss of fiber density and axonal area (atrophy)relative to nondiabetic controls (Fig. 2) and sciatic nerveaxonal atrophy was also detected in WT diabetic mice(Fig. 2) at later time points. However, sciatic fiber densitywas not changed over the period of exposure to diabetes.Diabetic sural and sciatic nerve myelin thickness wasreduced after 5 months of diabetes (Tables 2 and 3). Therewere no differences in the number of macrophages ordegenerating fiber clusters seen between WT mice with orwithout diabetes, but an increased number of degenerat-ing axonal profiles were identified in both sciatic and suraldiabetic nerves compared with control nerve (Tables 2 and3). In contrast to WT diabetic mice, RAGE�/� mice withdiabetes were protected from declines in sural axondensity and from axonal atrophy in the sural and sciaticnerves after 5 months of diabetes (Fig. 2 and Table 3).

CML deposition was verified in DRG neurons and withinperipheral nerves, and there was heightened expressionwith diabetes, irrespective of RAGE genetic status of themouse (supplementary Fig. 1, found in an online-only

TABLE 3Morphological properties of sural nerves in nondiabetic and diabetic nerves from WT and RAGE null mice after 1–8 months ofdiabetes


Axonal fiber density (per mm2)WT nondiabetic mice 6 18,756 � 125 18,010 � 132 16,952 � 133 17,015 � 115WT diabetic mice 6 18,436 � 117 17,112 � 122 15,521 � 115* 15,494 � 91*

RAGE null nondiabetic mice 4 19,029 � 127 18,224 � 125 17,811 � 107RAGE null diabetic mice 4 18,881 � 91 18,114 � 98 16,643 � 108

Axonal numbersWT nondiabetic mice 6 323 � 6 319 � 7 313 � 5 314 � 5WT diabetic mice 6 318 � 4 292 � 5 253 � 6* 248 � 4*

RAGE null nondiabetic mice 4 330 � 4 329 � 5 329 � 5RAGE null diabetic mice 4 328 � 3 314 � 6 298 � 6

Axonal area (�m2)WT nondiabetic mice 6 33.2 � 0.8 31.8 � 0.7 28.0 � 1.0 26.1 � 0.7WT diabetic mice 6 31.7 � 0.6 26.3 � 0.6 22.4 � 0.9* 22.3 � 0.6

RAGE null nondiabetic mice 4 33.1 � 0.5 31.6 � 0.7 28.2 � 0.9RAGE null diabetic mice 4 32.9 � 0.6 29.8 � 0.7 26.2 � 0.6

Myelination thickness (�m)WT nondiabetic mice 6 1.03 � 0.02 1.00 � 0.03 0.95 � 0.03 0.92 � 0.03WT diabetic mice 6 1.02 � 0.02 0.93 � 0.03 0.84 � 0.02* 0.76 � 0.02*

RAGE null nondiabetic mice 4 1.04 � 0.03 1.03 � 0.04 1.01 � 0.04 0.95 � 0.04RAGE null diabetic mice 4 1.04 � 0.03 0.97 � 0.03 0.92 � 0.03 0.90 � 0.04

Number of degenerating profiles(per transverse section)

WT nondiabetic mice 6 0 1 � 1 5 � 1 4 � 1WT diabetic mice 6 2 � 1 12 � 1* 16 � 2* 11 � 2*

RAGE null nondiabetic mice 4 0 0 3 � 1 3 � 1RAGE null diabetic mice 4 1 � 1 2 � 1 7 � 1 6 � 1

Number of regenerating fiber clusters(per transverse section)

WT nondiabetic mice 6 0 0 0 0WT diabetic mice 6 0 0.1 � 0.1 0.1 � 0.1 0.1 � 0.1RAGE null nondiabetic mice 4 0 0 0 0RAGE null diabetic mice 4 0 0 0.1 � 0.1 0.1 � 0.1

Data are means � SE. *Significance at P 0.05 using multiple ANOVA testing with Bonferroni post-hoc t test comparisons (� � 0.05). Datain bold are statistically significant.



RETRACTED ARTICLE

appendix at http://doi.org/10.2337/db07–0339). RAGE ex-pression was of greater intensity within axons andSchwann cells of both the sciatic nerve and especially thesural nerve of WT mice with diabetes (Fig. 3 and Table 4).Expression of RAGE was also noted within endothelialcells of vasa nervorum. Identification of Schwann cellsusing the S-100 antibody revealed extensive co-localization

with RAGE within diabetic peripheral nerve and withinDRGs of WT diabetic mice (Fig. 4 and Table 4). RAGE wasalso expressed in peripheral nerves of WT mice withoutdiabetes but at lower levels (Fig. 3 and Table 4). Despitethe long-term diabetes, no apoptosis was detected withinDRG neurons or Schwann cells. Evidence for apoptosiscould be detected within satellite cells of DRGs exposed to

FIG. 3. RAGE immunohistochemistry within transverse sections ofsciatic nerve. RAGE was expressed in control nondiabetic sections(A), but expression in diabetic nerves was more intense (B). Inlongitudinal sections of the sural nerve, RAGE profiles in WT micewith diabetes (C) co-localized particularly with Schwann cells (iden-tified with S-100) (D), as well as with axons (D and E) (bar � 10 �m).RAGE expression in DRG neurons and Schwann cells also was presentin mice without diabetes (F), but labeling was more intense with thepresence of diabetes (G). Schwann cells in DRG of WT mice withdiabetes demonstrated marked upregulation of RAGE, greater thanneurons, exhibited by immunostaining for RAGE (H), �-tubulin (I),S-100 (J), and co-localization (K). Quantitative RT-PCR identifiedmarked upregulation of RAGE transcripts in DRG, sciatic nerve, andbrain from WT mice with diabetes (L). Analysis was performed usingStudent’s t tests, with significance set at � � 0.05 (n � 6–8 for WTmice). (Please see http://dx.doi.org/10.2337/db07-0339 for a high-quality digital representation of this figure.)



RETRACTED ARTICLE

diabetes, with greater predominance found in diabetic WTmice compared with diabetic RAGE null mice (supplemen-tary Fig. 2).DRG. DRG neurons of WT diabetic mice were smaller inarea (neuronal atrophy) than DRG neurons in WT micewithout diabetes beginning after 5 months of diabetes(Table 5). In addition, at later time points available forstudy in the Swiss Webster WT mice (8 months), diabeticDRGs had a decline in profile neuronal density comparedwith nondiabetic control mice, suggesting neuronal drop-out (Fig. 4). In contrast to the WT diabetic mice, RAGE�/�

diabetic mice, available for analysis and studied at 5months of diabetes, were protected from neuronal atro-phy. RAGE expression was greater within neurons andSchwann cells of diabetic WT mouse DRGs compared withnondiabetic controls (Fig. 3 and Table 4).Epidermal innervation. The footpad epidermal nervefiber density of WT mice with diabetes was reducedcompared with control foot pads at and after 3 months ofdiabetes (Fig. 5 and Table 6). RAGE�/� mice with diabeteswere protected from epidermal fiber loss (Fig. 5 and Table6), with no difference identified between epidermal fiberdensities in RAGE�/� diabetic mice compared with thosewithout diabetes after 5 months (Fig. 5 and Table 6).Results of epidermal nerve fiber densities were similarusing two separate methods of quantification, either withepidermal area or epidermal length.

RAGE expression was identified in both control anddiabetic epidermal axons, but there was greater expres-sion within residual diabetic epidermal axons (Fig. 5).RAGE labeling was also increased within both sebaceousand sweat glands, as well as within dermal blood vessels ofdiabetic mice.RAGE mRNA and protein. Diabetes was associated witha relative increase in RAGE mRNA within DRG andperipheral nerve (Fig. 6) compared with the housekeepinggene cyclophyllin. Western blotting similarly confirmedthe presence of increased RAGE protein within diabeticWT mouse DRG and peripheral nerve relative to nondia-

betic WT mouse tissues (Fig. 6), when RAGE protein levelswere normalized for tissue �-actin content.NF-�B and PKC�II expression in PNS tissues andcells. WT mice with diabetes also demonstrated height-ened expression of both protein and mRNA for NF-�B andPKC�II when compared with WT mice without diabetes(Fig. 6). In contrast, RAGE�/� mice had very low levels ofNF-�B and PKC�II for both protein and mRNA in PNStissues, independent of their diabetic status (Fig. 6).

Consistent with these findings, immunohistochemis-try of peripheral nerve and DRGs demonstrated upregu-lation of NF-�B and its increased nuclear translocation(activation) in DRG sensory neurons in WT mice withdiabetes when compared with WT mice without diabe-tes and RAGE�/� mice with or without diabetes (Fig. 7).Both cytoplasmic and nuclear expression of NF-�B wasattenuated in DRG neurons in RAGE�/� mice withdiabetes when compared with WT mice with diabetes(Fig. 7). However, the more striking finding in both DRGand peripheral nerve was the overexpression of NF-�Bin Schwann cells, already exhibiting heightened expres-sion of RAGE (above), from the WT mice with diabetes(Fig. 7). The transcription factor C-Rel was upregulatedin WT diabetic DRG neurons, but not in RAGE nulldiabetic neurons (supplementary Fig. 3). HNE Adduct, amarker of lipid peroxidation and oxidative stress, wasupregulated in diabetic DRG neurons independent ofRAGE genetic status (supplementary Fig. 4). In situhybridization for the inhibitor of NF-�B activation, I�B,identified its significant upregulation in WT diabeticmice (supplementary Fig. 5) compared with othergroups. The prominent expression of NF-�B in Schwanncells was attenuated in RAGE�/� diabetic mice. Further-more, the intensity of NF-�B expression was associatedwith RAGE expression in both WT nondiabetic anddiabetic DRG neurons, especially in diabetic mice (Fig.8). Although peripheral nerve and DRGs also exhibitedupregulation of PKC�II in DRG sensory neurons in WTmice with diabetes when compared with nondiabetic

TABLE 4Quantification of RAGE immunofluorescence and mRNA expression in peripheral nerve tissues

n Low Intermediate High

Graded RAGE immunofluorescence for axons or neuronsSural nerve axons (%)

WT nondiabetic mice 6 79 � 4 18 � 3 3 � 1WT diabetic mice 6 38 � 3 32 � 3 30 � 3*

Sciatic nerve axons (%)WT nondiabetic mice 6 87 � 2 13 � 2 0 � 1WT diabetic mice 6 46 � 2 28 � 2 26 � 2*

DRG neurons (%)WT nondiabetic mice 6 76 � 4 18 � 2 6 � 1WT diabetic mice 6 49 � 3 28 � 2 23 � 2*

Graded RAGE immunofluorescence for Schwann cellsSural nerve axons (%)

WT nondiabetic mice 6 77 � 3 20 � 2 4 � 1WT diabetic mice 6 24 � 2 27 � 2 49 � 2*

Sciatic nerve axons (%)WT nondiabetic mice 6 76 � 3 23 � 2 1 � 1WT diabetic mice 6 34 � 2 29 � 2* 37 � 2*

DRG neurons (%)WT nondiabetic mice 6 71 � 2 22 � 2 7 � 1WT diabetic mice 6 24 � 3 28 � 2* 48 � 2*

Data are means � SE. *Significance at P 0.05 using multiple ANOVA testing with Bonferroni post-hoc t test comparisons (� � 0.0125).Analysis was performed using ANOVA, with significance set at � � 0.05. n � 6–8 for WT mice. Data in bold are statistically significant.



RETRACTED ARTICLE

WT mice and RAGE�/� mice with or without diabetes,these results were much more modest than observedchanges with NF-�B. Taken together, our results dem-onstrate that NF-�B, a potential downstream signal ofRAGE overactivation, dramatically upregulates in ex-perimental diabetic neuropathy dependent on the pres-ence of RAGE.

DISCUSSION

AGEs and RAGE. AGEs are generated by aging andhyperglycemia, but with an accelerated rate in chronicdiabetes (7). On their own, AGEs induce permanent ab-

normalities in extracellular matrix component functionand have also been shown to be mutagenic (24). AGEssuch as CML derived from glyceraldehyde and glycolalde-hyde induce apoptosis and decrease viability in culturedSchwann cells, where they also increase the release oftumor necrosis factor-� and interleukin-6, as well asenhance activation of NF-�B (25). In the murine PNS,diabetes was associated with CML deposition, and itspresence was unrelated to RAGE genetic status. This is anexpected result, based on similar levels of hyperglycemiain WT and RAGE null diabetic mice driving the formationof AGEs such as CML. Although AGEs bind to other

FIG. 4. Morphological assessment of the DRG from WT and RAGE�/� mice with and without diabetes using semi-thin sections and toluidine bluestaining. Diabetes was associated with greater age-dependent atrophy of DRG neurons beginning after 5 months of diabetes. In contrast, theneurons of RAGE�/� mice with diabetes were protected from atrophy after 5 months of diabetes, and declines in neuron profile density could bedemonstrated in WT diabetic mice after 8 months of diabetes. A histogram of neuronal area demonstrated a slight leftward shift indicatingatrophy in diabetic WT mice (A), but without statistical significance (Tukey’s honestly significant difference test, NS). Images of DRG neuronsfrom a RAGE�/� mouse without diabetes (B) and a RAGE�/� mouse with diabetes (C) after 5 months of diabetes in comparison to a WT mousewithout diabetes (D) and a WT mouse with diabetes (E) at the same age points are shown. Significant differences between age-comparable groupsare demonstrated by horizontal bars after multiple one-way ANOVA tests with samples treated independently were performed (P < 0.0125 usingBonferroni corrections) (n � 6–8 for WT mice and n � 4 for RAGE�/� mice). Bar � 10 �m.



RETRACTED ARTICLE

scavenger receptors, the best-characterized receptor forAGE is RAGE, which is a multi-ligand member of theimmunoglobulin superfamily of cell surface molecules.AGE-RAGE ligation leads to cell activation and increasedexpression of extracellular matrix proteins, vascular ad-hesion molecules, cytokines, growth factors, and the gen-eration of reactive oxygen intermediates. In vitro studieshave identified RAGE to be important for mediation ofneurite outgrowth and regulation of gene expressionthrough NF-�B, a transcription factor, through the Ras–mitogen-activated kinase (MAP) kinase pathway (26). Ac-tivation of these signal transduction pathways maycontribute to development of reactive oxygen species,such as with NADPH oxidase activation (27). One of thesepathways includes activation of p21ras, followed by activa-tion of p44/p42 MAP kinases and nuclear translocation ofNF-�B (28,29). NF-�B–regulated gene expression is de-tected in pathological samples where both RAGE andAGEs are present at high levels. RAGE and NF-�B had anassociated upregulation in nondiabetic DRG neurons (Fig.8) and a prominent association in diabetic DRG neurons.In this study, diabetic WT mouse DRG neurons andSchwann cells also had upregulated C-Rel nuclear trans-location. Together with data revealing upregulated I�B,these findings link RAGE to upregulation and expressionof NF-�B.

In vitro studies have also shown that administration ofAGEs in a high-glucose environment potentiates PKCactivation (30,31) in both macrophages and renal tubularcells. AGE-induced modification of signal transductionpathways occur through the induction of PKC activation,but may also occur independent of PKC activation (32). Inmesangial cells, the independent activation of NF-�B andPKC pathways due to AGE presence in vitro is thought tobe an early event contributing to oxidative stress (33). Theinteraction of AGEs with RAGE also leads to perturbationof vascular and inflammatory cells (9,11,12,24,26,34–36).RAGE and AGEs act to diminish vascular barrier functionas demonstrated by the tissue-blood isotope ratio (37);vascular permeability may be reduced because of anenhanced expression of vascular cell adhesion molecule-1(VCAM-1) (37,38). In our model, AGE- and RAGE-medi-

ated changes of signaling pathways for NF-�B and PKCalso target the diabetic PNS in our model.RAGE and diabetic neuropathy. Mice with long-termexperimental diabetes developed changes in both electro-physiology and structure of their nerves and sensoryganglia relative to nondiabetic control littermates. Begin-ning at the DRG, sensory neurons demonstrate atrophy(loss of neuronal area) and mild neuronal loss afterprolonged diabetes (15,39). Within the mixed motor-sen-sory sciatic nerve, diabetic axons are smaller in caliberand have thinner myelin relative to control fibers. Thesefibers had parallel electrophysiological abnormalities withprogressive loss of motor potentials (compound motoraction potentials) and declines in conduction velocity. Inthe sensory-dominant sural nerve, diabetic mice had frankloss of axons with fibers undergoing degeneration, inaddition to atrophy and myelin thinning. Sciatic-tibialsensory nerve conduction studies demonstrated impair-ment of electrophysiological parameters in diabetic nervesrelative to controls. At the most distal site of the PNS,diabetic mice had a loss of epidermal nerve fibers. Thus,this model replicated most of the features of progressivehuman diabetic polyneuropathy.

In this context, we identified robust RAGE expression atall levels of the PNS in experimental diabetes, includingDRG, peripheral nerve, and epidermal fibers. In this modeland in human peripheral nerves, the deposition of AGEs(10) likely contributes in turn to RAGE upregulation (40).In many pathological lesions, the abundance of RAGE-expressing cells has a strict correlation with accumulationof RAGE ligands, such as in diabetic vasculature, wherehighly expressing RAGE cells can be found proximal toAGE-abundant regions (41,42). Although RAGE expres-sion in the peripheral nerve is clearly upregulated withinneurons and axons, Schwann cells within diabetic periph-eral nerve and DRGs have an even greater apparentoverexpression of RAGE in this long-term model of diabe-tes. The presence of RAGE in epidermal axons of skinfootpads is of considerable interest. If, as suggested by ourfindings, its overexpression in abnormal and residualdiabetic epidermal axons reflects ongoing injury, its pres-

TABLE 5Morphological properties of DRG neurons in nondiabetic and diabetic nerves from WT and RAGE null mice after 1–8 months ofdiabetes


Neuronal density (per mm2)WT nondiabetic mice 6 2,568 � 39 2,565 � 31 2,529 � 41 2,521 � 33WT diabetic mice 6 2,582 � 32 2,532 � 36 2,478 � 36 2,456 � 32*

RAGE null nondiabetic mice 4 2,590 � 28 2,574 � 30 2,526 � 32RAGE null diabetic mice 4 2,586 � 22 2,544 � 29 2,521 � 31

Neurons (n)WT nondiabetic mice 6 2,444 � 43 2,436 � 47 2,432 � 41 2,437 � 38WT diabetic mice 6 2,451 � 31 2,376 � 32 2,296 � 27* 2,274 � 26*

RAGE null nondiabetic mice 4 2,422 � 39 2,475 � 37 2,446 � 39 2,439 � 31RAGE null diabetic mice 4 2,468 � 39 2,413 � 38 2,398 � 34 2,401 � 38

Neuronal area (�m2)WT nondiabetic mice 6 621 � 14 618 � 15 609 � 16 602 � 14WT diabetic mice 6 623 � 14 607 � 13 572 � 16* 554 � 14*

RAGE null nondiabetic mice 4 620 � 15 619 � 14 617 � 17RAGE null diabetic mice 4 618 � 14 618 � 15 616 � 18

Data are means � SD. *Significance at P 0.05 using multiple Student’s t testing (� � 0.05) with WT nondiabetic mice compared with WTdiabetic mice and RAGE null nondiabetic mice compared with RAGE null diabetic mice only. Data in bold are statistically significant.



RETRACTED ARTICLE

ence may be relevant to the targeting of small axons bydiabetes, leading to small fiber neuropathy. Epidermalnerve fiber densities are indeed decreased in diabeticpatients, findings that are associated with loss of sensory

function and the development of neuropathic pain (43).Both diabetic and control epidermal axons expressedRAGE, but diabetics had loss of axons with higher expres-sion in residual axons.

FIG. 5. RAGE within epidermal footpads from both WT and RAGE�/� mice, both with and without diabetes (A–P). The basement membrane andvasculature were identified with immunohistochemistry for collagen type IV (A, E, I, and M), whereas Langerhans cells and myelinated dermalfibers were identified with S-100 (B, F, J, and N). Epidermal axons were identified with PGP 9.5 (C, G, K, and O), with co-localization shown aswell (D, H, L, and P). There was evidence of RAGE expression within epidermal nerve axons fibers identified with PGP 9.5 (Q and V), both in WTmice without (R) and with diabetes (U), with RAGE expression accentuated in WT mice with diabetes. Co-localizations are demonstrated for WTdiabetic mice (S) and nondiabetic mice (V). Significant differences between age-comparable groups are demonstrated by horizontal bars(multiple one-way ANOVA tests assuming independent samples, P < 0.0125, using Bonferroni corrections) (n � 8–12 for WT mice and n � 4 forRAGE�/� mice). Bar � 50 �m. (Please see http://dx.doi.org/10.2337/db07-0339 for a high-quality digital representation of this figure.)



RETRACTED ARTICLE

RAGE�/� mice and diabetic neuropathy. RAGE�/�

mice permitted a unique assessment of the potential roleof the RAGE pathway in the development of neuropathy.Although AGE actions independent of RAGE ligation maycontribute to ongoing toxicity within the extracellularmatrix, AGE-RAGE interactions were significantly miti-gated in RAGE null mice. The absence of RAGE attenuatedboth structural and electrophysiological changes withinthe peripheral nerves and the DRG after 5 months ofdiabetes. Previous studies have not evaluated the impactof RAGE deletion during the full repertoire of changesfrom neuropathy in a longer term model. Preliminary workby Bierhaus et al. (13) demonstrated that RAGE�/� mice

did not develop expected hypalgesia of diabetes. In addi-tion, NF-�B activation was heightened in diabetic WT miceand was diminished in RAGE�/� mice or with the inter-vention of soluble RAGE in WT mice (13). In our studies,RAGE�/� mice peripheral neurons and supporting cellswere protected from diabetes-induced upregulation ofboth NF-�B and PKC�II, even in the presence of long-termdiabetes. Long-term upregulation of NF-�B, and possiblyPKC�II, have both been previously linked to the develop-ment of neuropathy in diabetes (44). The upregulation ofPKC in our mouse model was statistically significant butmodest compared with that observed in the case of NF-�B.Previous studies in other species, or in experiments with

FIG. 6. Western blots identified increased RAGE protein withindiabetic DRG and peripheral nerve relative to controls with normal-ization to �-actin content (A). NF-�B and PKC�II identified upregu-lation in WT tissues exposed to long-term (5 months) diabetes (A),whereas RAGE�/� mice were protected from such upregulation (A).Quantitative RT-PCR identified marked age-dependent upregulationfor NF-�B mRNA in DRG, sciatic nerve, and sural nerve from WT micewith diabetes (B). RAGE�/� mice, with or without diabetes, failed todemonstrate any upregulation in NF-�B mRNA with the presence ofdiabetes (B). Similarly, quantitative RT-PCR identified an age-depen-dent upregulation for PKC�II mRNA in DRG, sciatic nerve, and suralnerve from WT mice with diabetes (C). RAGE�/� mice, with or withoutdiabetes, exhibited very low levels of PKC�II mRNA when comparedwith age-matched WT mouse tissues with or without diabetes (C).Analysis was performed using a Student’s t test with significance setat � � 0.05 (n � 4 WT mice and n � 3 for RAGE�/� mice).

TABLE 6Epidermal nerve fiber density within skin of diabetic and nondiabetic WT and RAGE null mice using both measurements related toepidermal area and epidermal linear density


Epidermal nerve fiber density as a measureof epidermal area (per mm2)

WT nondiabetic mice 6 7.8 � 0.8 7.3 � 0.7 7.1 � 0.8 5.1 � 0.8WT diabetic mice 6 6.3 � 0.5 4.3 � 0.5* 3.7 � 0.6* 0.2 � 0.1RAGE null nondiabetic mice 4 7.7 � 0.8 7.5 � 0.9 7.4 � 1.0RAGE null diabetic mice 4 7.5 � 0.9 6.5 � 1.0 5.8 � 1.0

Epidermal nerve fiber density as a measureof epidermal length (per mm)

WT nondiabetic mice 6 27.3 � 2.2 27.1 � 1.8 26.9 � 1.7 20.0 � 2.0WT diabetic mice 6 22.7 � 1.3 15.1 � 1.5* 14.1 � 1.4* 0.7 � 0.1RAGE null nondiabetic mice 4 29.3 � 1.4 27.8 � 1.6 27.4 � 1.8RAGE null diabetic mice 4 27.8 � 1.7 24.7 � 1.6 21.5 � 1.7

Data are means � SD. *Significance at P 0.05 using one-way ANOVA testing with Bonferroni post-hoc t test comparisons (� � 0.0125). Datain bold are statistically significant.



RETRACTED ARTICLE

FIG. 7. Images of NF-�B expression within neurons, identified with �-tubulin (A, E, I, and M), and Schwann cells identified with S-100 (B, F, J,and N) of DRG from both WT and RAGE�/� mice, with and without diabetes. Expression is illustrated for NF-�B (C, G, K, and O) andco-localization is demonstrated for each of these three markers (D, H, L, and P). Although clearly present in neurons, and in activated neurons(within nuclei) of DRG, NF-�B presence is prominent in Schwann cells of WT mice with diabetes (G). Quantification identified significantlygreater expression in neurons of WT mice exposed to diabetes (Q), but less NF-�B expression within RAGE�/� mice with diabetes. The percentageof activated neurons for NF-�B was also significantly elevated in WT mice exposed to 5 months of diabetes (R). Within Schwann cells of suralnerve (S–X), NF-�B was co-localized with S-100 in WT mice with diabetes (S) and in RAGE�/� mice with diabetes (V), NF-�B immunohistochem-ical expression was greater in WT mice with diabetes (T) than in RAGE�/� mice with diabetes (W) (co-localizations demonstrated in U and X).Quantification demonstrated heightened expression in Schwann cells of each of sural and sciatic nerves, as well as in DRG from WT mice withdiabetes (Y). Meanwhile, RAGE�/� mice had an attenuated rise in NF-�B expression. Analysis was performed using ANOVA testing for Q and R,and a Student’s t test was performed to compare activated neuron measurements, with significance set at � � 0.0125 (using Bonferronicorrections) with horizontal lines indicating groups having statistically meaningful differences (all measurements were statistically different)(n � 4–6 for WT mice and n � 3 for RAGE�/� mice). Bar � 5 �m. (Please see http://dx.doi.org/10.2337/db07-0339 for a high-quality digitalrepresentation of this figure.)



RETRACTED ARTICLE

shorter duration of diabetes, have demonstrated eithermild PKC upregulation associated with microvasculatureor absence of definite changes in PKC levels (44–46). It isimportant to note that RAGE�/� diabetic mice did demon-strate evidence of mild morphological and electrophysio-logical deterioration nonetheless. It appears unlikely thatRAGE- and AGE-mediated processes provide an exclusiveexplanation for the abnormalities identified in diabeticneuropathy. It is of interest that RAGE�/� nondiabeticmice also had mild attenuation of age-related changes inmorphology and electrophysiology, suggesting a possiblerole of RAGE in age-related decline.

In addition to its expression in neurons, NF-�B ex-pression was also identified in glial cells. The role ofNF-�B is complex and depends on age and on injurytype and timing (47,48). When diabetic peripheral nerveis exposed to ischemia followed by reperfusion, NF-�Bexpression in Schwann cells of diabetic peripheralnerve rises (49). In human Schwann cell cultures, tumornecrosis factor-� contributes to the transient activationof NF-�B in the absence of apoptosis (50). In humanpatients with chronic inflammatory demyelinating poly-

neuropathy, Schwann cell nuclei demonstrate increasedlevels of NF-�B (50). For Schwann cells exposed tohyperglycemia in vitro, NF-�B expression is similarlyelevated (51). In development, activation of the tran-scription factor NF-�B is an essential differentiationsignal for axon-associated peripheral myelin formation,with its expression progressively declining during adult-hood, presumably to prevent excessive myelination(52). Additionally, NF-�B activation within Schwanncells after angiotensin II’s interaction with its receptormay be important in peripheral nerve regeneration (53).The present data demonstrate a consistent relationshipwith age- and diabetes-dependent upregulated expres-sion of NF-�B in peripheral nerves, attenuated inRAGE�/� mice with diabetes.

In conclusion, our data identify an important role forRAGE in a long-term model of experimental diabetes.Not only is heightened RAGE expression prominent andprogressive at multiple levels (and in multiple cells) ofthe PNS, but its expression strictly parallels structuraland electrophysiological alterations. Deletion of RAGEprevented a number of electrophysiological and mor-

FIG. 7—Continued



RETRACTED ARTICLE

phological changes in the diabetic nervous system.Moreover, we provide evidence that a prominent down-stream target of RAGE (NF-�B) is increased in thediabetic PNS, but not with RAGE deletion. Whereasoverexpression of RAGE, and the consequent aberrantsignaling of neurons and glial cells, is not the onlymechanism of neurological damage in diabetes, our datasuggest that it has an important contribution to neurop-athy in this animal model of diabetes.

ACKNOWLEDGMENTS

This study was supported by an operating grant from theCanadian Institutes of Health Research (CIHR) and theCanadian Diabetes Association (CDA). C.T. is a ClinicalInvestigator of the Alberta Heritage Foundation for Medi-cal Research and D.W.Z. is a Scientist of the AlbertaHeritage Foundation for Medical Research (AHFMR).A.M.S., L.L.R., and F.S. were supported by grants from theUnited State Public Health Service.

REFERENCES

1. Vinik AI, Vinik E: Prevention of the complications of diabetes. Am J Manag

Care 9:S63–S80, 20032. Lozeron P, Nahum L, Lacroix C, Ropert A, Guglielmi JM, Said G: Symp-

tomatic diabetic and non-diabetic neuropathies in a series of 100 diabeticpatients. J Neurol 249:569–575, 2002

3. Smith AG, Ramachandran P, Tripp S, Singleton JR: Epidermal nerveinnervation in impaired glucose tolerance and diabetes-associated neurop-athy. Neurology 57:1701–1704, 2001

4. Toth C, Brussee V, Cheng C, Zochodne DW: Diabetes mellitus and thesensory neuron. J Neuropathol Exp Neurol 63:561–573, 2004

5. Diabetes Control and Complications Trial Research Group: The effect ofintensive treatment of diabetes on the development and progression oflong-term complications in insulin-dependent diabetes mellitus. N Engl

J Med 329:977–986, 19936. UK Prospective Diabetes Study (UKPDS) Group: Intensive blood-glucose

control with sulphonylureas or insulin compared with conventional treat-ment and risk of complications in patients with type 2 diabetes (UKPDS33). Lancet 352:837–853, 1998

7. Schmidt AM, Yan SD, Wautier JL, Stern D: Activation of receptor foradvanced glycation end products: a mechanism for chronic vasculardysfunction in diabetic vasculopathy and atherosclerosis. Circ Res 84:489–497, 1999

8. Wautier JL, Guillausseau PJ: Diabetes, advanced glycation endproductsand vascular disease. Vasc Med 3:131–137, 1998

9. Wautier JL, Wautier MP: [Pathophysiologic aspects of diabetic angiopa-thy]. Diabete Metab 25 (Suppl. 3):21–26, 1999

10. Williams SK, Howarth NL, Devenny JJ, Bitensky MW: Structural andfunctional consequences of increased tubulin glycosylation in diabetesmellitus. Proc Natl Acad Sci U S A 79:6546–6550, 1982

11. Wautier JL, Guillausseau PJ: Advanced glycation end products, theirreceptors and diabetic angiopathy. Diabete Metab 27:535–542, 2001

12. Brett J, Schmidt AM, Yan SD, Zou YS, Weidman E, Pinsky D, Nowygrod R,Neeper M, Przysiecki C, Shaw A: Survey of the distribution of a newlycharacterized receptor for advanced glycation end products in tissues.Am J Pathol 143:1699–1712, 1993

13. Bierhaus A, Haslbeck KM, Humpert PM, Liliensiek B, Dehmer T, Morcos M,Sayed AA, Andrassy M, Schiekofer S, Schneider JG, Schulz JB, Heuss D,Neundorfer B, Dierl S, Huber J, Tritschler H, Schmidt AM, Schwaninger M,Haering HU, Schleicher E, Kasper M, Stern DM, Arnold B, Nawroth PP:Loss of pain perception in diabetes is dependent on a receptor of theimmunoglobulin superfamily. J Clin Invest 114:1741–1751, 2004

14. Haslbeck KM, Schleicher E, Bierhaus A, Nawroth P, Haslbeck M, Neun-dorfer B, Heuss D: The AGE/RAGE/NF-(kappa)B pathway may contributeto the pathogenesis of polyneuropathy in impaired glucose tolerance(IGT). Exp Clin Endocrinol Diabetes 113:288–291, 2005

15. Kennedy JM, Zochodne DW: Experimental diabetic neuropathy withspontaneous recovery: is there irreparable damage? Diabetes 54:830–837,2005

16. Wendt TM, Tanji N, Guo J, Kislinger TR, Qu W, Lu Y, Bucciarelli LG, RongLL, Moser B, Markowitz GS, Stein G, Bierhaus A, Liliensiek B, Arnold B,Nawroth PP, Stern DM, D’Agati VD, Schmidt AM: RAGE drives thedevelopment of glomerulosclerosis and implicates podocyte activation inthe pathogenesis of diabetic nephropathy. Am J Pathol 162:1123–1137,2003

17. Zochodne DW, Ho LT: Normal blood flow but lower oxygen tension indiabetes of young rats: microenvironment and the influence of sympathec-tomy. Can J Physiol Pharmacol 70:651–659, 1992

18. Toth C, Brussee V, Zochodne DW: Remote neurotrophic support ofepidermal nerve fibres in experimental diabetes. Diabetologia 49:1081–1088, 2006

19. Cheng HJ, Nakamoto M, Bergemann AD, Flanagan JG: Complementarygradients in expression and binding of ELF-1 and Mek4 in development ofthe topographic retinotectal projection map. Cell 82:371–381, 1995

20. Song Z, Fu DT, Chan YS, Leung S, Chung SS, Chung SK: Transgenic miceoverexpressing aldose reductase in Schwann cells show more severe nerveconduction velocity deficit and oxidative stress under hyperglycemicstress. Mol Cell Neurosci 23:638–647, 2003

21. Obrosova IG, Mabley JG, Zsengeller Z, Charniauskaya T, Abatan OI, GrovesJT, Szabo C: Role for nitrosative stress in diabetic neuropathy: evidencefrom studies with a peroxynitrite decomposition catalyst. FASEB J 19:401–403, 2005

22. Kellogg AP, Pop-Busui R: Peripheral nerve dysfunction in experimentaldiabetes is mediated by cyclooxygenase-2 and oxidative stress. Antioxid

Redox Signal 7:1521–1529, 200523. Ho EC, Lam KS, Chen YS, Yip JC, Arvindakshan M, Yamagishi S, Yagihashi

S, Oates PJ, Ellery CA, Chung SS, Chung SK: Aldose reductase-deficientmice are protected from delayed motor nerve conduction velocity, in-creased c-Jun NH2-terminal kinase activation, depletion of reduced gluta-thione, increased superoxide accumulation, and DNA damage. Diabetes

55:1946–1953, 200624. Brownlee M, Cerami A, Vlassara H: Advanced glycosylation end products

in tissue and the biochemical basis of diabetic complications. N Engl J Med

318:1315–1321, 198825. Sekido H, Suzuki T, Jomori T, Takeuchi M, Yabe-Nishimura C, Yagihashi S:

Reduced cell replication and induction of apoptosis by advanced glycationend products in rat Schwann cells. Biochem Biophys Res Commun

320:241–248, 2004

FIG. 8. Correlations between RAGE and NF-�B immunohistochemicallabeling were studied with scatter plots in WT nondiabetic mice (A)and WT diabetic mice (B), suggesting a significant logarithmic regres-sion association most robust in diabetic DRG neurons.



RETRACTED ARTICLE

26. Huttunen HJ, Kuja-Panula J, Rauvala H: Receptor for advanced glycationend products (RAGE) signaling induces CREB-dependent chromograninexpression during neuronal differentiation. J Biol Chem 277:38635–38646,2002

27. Wautier MP, Chappey O, Corda S, Stern DM, Schmidt AM, Wautier JL:Activation of NADPH oxidase by AGE links oxidant stress to altered geneexpression via RAGE. Am J Physiol Endocrinol Metab 280:E685–E694,2001

28. Yan SD, Schmidt AM, Anderson GM, Zhang J, Brett J, Zou YS, Pinsky D,Stern D: Enhanced cellular oxidant stress by the interaction of advancedglycation end products with their receptors/binding proteins. J Biol Chem

269:9889–9897, 199429. Lander HM, Tauras JM, Ogiste JS, Hori O, Moss RA, Schmidt AM:

Activation of the receptor for advanced glycation end products triggers ap21(ras)-dependent mitogen-activated protein kinase pathway regulatedby oxidant stress. J Biol Chem 272:17810–17814, 1997

30. Beauchamp MC, Michaud SE, Li L, Sartippour MR, Renier G: Advancedglycation end products potentiate the stimulatory effect of glucose onmacrophage lipoprotein lipase expression. J Lipid Res 45:1749–1757, 2004

31. Heidland A, Sebekova K, Schinzel R: Advanced glycation end products andthe progressive course of renal disease. Am J Kidney Dis 38:S100–S106,2001

32. Iwashima Y, Eto M, Horiuchi S, Sano H: Advanced glycation end product-induced peroxisome proliferator-activated receptor gamma gene expres-sion in the cultured mesangial cells. Biochem Biophys Res Commun

264:441–448, 199933. Lal MA, Brismar H, Eklof AC, Aperia A: Role of oxidative stress in

advanced glycation end product-induced mesangial cell activation. Kidney

Int 61:2006–2014, 200234. Li J, Schmidt A: Characterization and functional analysis of the promoter

of RAGE, the receptor for advanced glycation end products. J Biol Chem

272:16498–16506, 199735. Mamputu JC, Renier G: Advanced glycation end products increase,

through a protein kinase C-dependent pathway, vascular endothelialgrowth factor expression in retinal endothelial cells: inhibitory effect ofgliclazide. J Diabetes Complications 16:284–293, 2002

36. Osicka TM, Yu Y, Lee V, Panagiotopoulos S, Kemp BE, Jerums G:Aminoguanidine and ramipril prevent diabetes-induced increases in pro-tein kinase C activity in glomeruli, retina and mesenteric artery. Clin Sci

(Lond) 100:249–257, 200137. Wautier JL, Zoukourian C, Chappey O, Wautier MP, Guillausseau PJ, Cao

R, Hori O, Stern D, Schmidt AM: Receptor-mediated endothelial celldysfunction in diabetic vasculopathy: soluble receptor for advanced gly-cation end products blocks hyperpermeability in diabetic rats. J Clin

Invest 97:238–243, 199638. Boulanger E, Wautier MP, Wautier JL, Boval B, Panis Y, Wernert N, Danze

PM, Dequiedt P: AGEs bind to mesothelial cells via RAGE and stimulateVCAM-1 expression. Kidney Int 61:148–156, 2002

39. Zochodne DW, Verge VM, Cheng C, Sun H, Johnston J: Does diabetes targetganglion neurones? Progressive sensory neurone involvement in long-termexperimental diabetes. Brain 124:2319–2334, 2001

40. Huttunen HJ, Kuja-Panula J, Sorci G, Agneletti AL, Donato R, Rauvala H:Coregulation of neurite outgrowth and cell survival by amphoterin and

S100 proteins through receptor for advanced glycation end products(RAGE) activation. J Biol Chem 275:40096–40105, 2000

41. Naka Y, Bucciarelli LG, Wendt T, Lee LK, Rong LL, Ramasamy R, Yan SF,Schmidt AM: RAGE axis: animal models and novel insights into thevascular complications of diabetes. Arterioscler Thromb Vasc Biol 24:1342–1349, 2004

42. Ritthaler U, Deng Y, Zhang Y, Greten J, Abel M, Sido B, Allenberg J, OttoG, Roth H, Bierhaus A: Expression of receptors for advanced glycation endproducts in peripheral occlusive vascular disease. Am J Pathol 146:688–694, 1995

43. Hirai A, Yasuda H, Joko M, Maeda T, Kikkawa R: Evaluation of diabeticneuropathy through the quantitation of cutaneous nerves. J Neurol Sci

172:55–62, 200044. Yamagishi S, Uehara K, Otsuki S, Yagihashi S: Differential influence of

increased polyol pathway on protein kinase C expressions betweenendoneurial and epineurial tissues in diabetic mice. J Neurochem 87:497–507, 2003

45. Cameron NE, Cotter MA: Effects of protein kinase Cbeta inhibition onneurovascular dysfunction in diabetic rats: interaction with oxidativestress and essential fatty acid dysmetabolism. Diabete Metab Res Rev

18:315–323, 200246. Nakamura J, Kato K, Hamada Y, Nakayama M, Chaya S, Nakashima E,

Naruse K, Kasuya Y, Mizubayashi R, Miwa K, Yasuda Y, Kamiya H, IenagaK, Sakakibara F, Koh N, Hotta N: A protein kinase C-beta-selectiveinhibitor ameliorates neural dysfunction in streptozotocin-induced dia-betic rats. Diabetes 48:2090–2095, 1999

47. Glazner GW, Camandola S, Geiger JD, Mattson MP: Endoplasmic reticulumD-myo-inositol 1,4,5-trisphosphate-sensitive stores regulate nuclear factor-kappaB binding activity in a calcium-independent manner. J Biol Chem

276:22461–22467, 200148. Fernyhough P, Smith DR, Schapansky J, Van Der PR, Gardiner NJ, Tweed

CW, Kontos A, Freeman L, Purves-Tyson TD, Glazner GW: Activation ofnuclear factor-kappaB via endogenous tumor necrosis factor alpha regu-lates survival of axotomized adult sensory neurons. J Neurosci 25:1682–1690, 2005

49. Wang Y, Schmeichel AM, Iida H, Schmelzer JD, Low PA: Enhancedinflammatory response via activation of NF-kappaB in acute experimentaldiabetic neuropathy subjected to ischemia-reperfusion injury. J Neurol Sci

247:47–52, 200650. Bonetti B, Valdo P, Stegagno C, Tanel R, Zanusso GL, Ramarli D, Fiorini E,

Turazzi S, Carner M, Moretto G: Tumor necrosis factor alpha and humanSchwann cells: signalling and phenotype modulation without cell death.J Neuropathol Exp Neurol 59:74–84, 2000

51. Suzuki T, Sekido H, Kato N, Nakayama Y, Yabe-Nishimura C: Neurotro-phin-3-induced production of nerve growth factor is suppressed inSchwann cells exposed to high glucose: involvement of the polyol path-way. J Neurochem 91:1430–1438, 2004

52. Nickols JC, Valentine W, Kanwal S, Carter BD: Activation of the transcrip-tion factor NF-kappaB in Schwann cells is required for peripheral myelinformation. Nat Neurosci 6:161–167, 2003

53. Reinecke K, Lucius R, Reinecke A, Rickert U, Herdegen T, Unger T:Angiotensin II accelerates functional recovery in the rat sciatic nerve invivo: role of the AT2 receptor and the transcription factor NF-kappaB.FASEB J 17:2094–2096, 2003



RETRACTED ARTICLE

original article receptor for advanced glycation...

Documents