Talin immunogold density increases in sciatic nerve of …medicina.lsmuni.lt/med/0602/0602-09e.pdftalin-like immunoreactivity in the sciatic nerve of normal, diabetic and NGF treated

147Medicina (Kaunas) 2006; 42(2)

Talin immunogold density increases in sciatic nerve of diabetic ratsafter nerve growth factor treatment

Waleed M. Renno, Farid Saleh, Ivo Klepacek, Farida Al-Awadi1

Department of Anatomy, 1Department of Biochemistry, Faculty of Medicine, Kuwait University, Kuwait

Key words: talin, sciatic nerve, diabetic neuropathy, Schwann cells, endoneurium, perineurium.

Summary. Background. Diabetic neuropathy is a debilitating disorder whose causation ispoorly understood. Recent studies have shown significant reduction in the activity of nerve growthfactor (NGF) and in the amount of talin cytoskeleton protein immunoreactivity in the perineuriumin patients with diabetic neuropathy.

Objective. Since talin is involved in transmembrane connections between extracellular matrixand cytoskeleton, this study investigates the subcellular pattern of talin immunoreactivity andthe effect of NGF treatment of diabetic rats on the distribution of talin in the sciatic nerve.

Materials and methods. Post-embedding immunogold electron microscopy using monoclonalantibody against talin in combination with quantitative procedures was employed to localizetalin-like immunoreactivity in the sciatic nerve of normal, diabetic and NGF treated diabeticrats.

Results. We found the highest densities of gold particles in the Schwann cells (139.6±5.6particles/µm2) and in the fibroblasts (127.4±4.1 particles/µm2). A moderate amount of immu-noreactivity was also present in the endothelial cells of vasa nervosa (32.3±9.1 particles/µm2).The myelinated and unmyelinated nerve fibers and the extracellular matrix profiles were notlabeled (8.7±2.1 particles/µm2, 4.2±2.2 particles/µm2, 6.1±3.2 particles/µm2, 9.5±5.3 partic-les/µm2, respectively). The immunogold localization of talin in diabetic rats was significantly(p<0.001) reduced in Schwann cells (66.3±6.5 particles/µm2) and perineurial and epineurialfibroblasts (56.8±3.9 particles/µm2). Diabetic rats treated with NGF for 12 weeks showedsignificant (p<0.005) increase in talin-like immunogold density in Schwann cells and fibroblasts.Talin immunogold density in Schwann cells and fibroblasts increased approximately 68% and58%, respectively, after NGF treatment. The endothelial cells of endoneurial and epineurialvessel walls showed no significant change in the talin-like immunogold particle density amongcontrol, diabetic and NGF treated diabetic animals.

Conclusions. These results have shown that the administration of exogenous NGF may beessential for inducing functionally significant regenerative mechanisms in diabetic neuropathythrough maintaining the permeability of the barrier properties of the peripheral nerve.

Correspondence to W. M. Renno, Department of Anatomy, Faculty of Medicine, Kuwait University, P. O. Box 24923,13110 Safat, Kuwait. E-mail: [email protected]

IntroductionNerve growth factors (NGF) are required during

development for the survival and growth of peripheralsensory dorsal root ganglia (DRG) and sympatheticneurons (1). The exact function of NGF in adultsensory DRG neurons is unknown. Several studies(2–4) indicated that NGF loss in adult sensory DRGneurons is responsible for the production of at leastthree components of the axon reaction: 1) somatofugalaxonal atrophy, which is the reduction in axonal cali-ber beginning in the proximal axon resulting from adecrease in NGF receptor (pNF) synthesis; 2) nucleareccentricity; and 3) aberrant expression of pNF epito-

pes in neuronal soma. Together, these observationssuggest that NGF influences axonal size while servingin the broader capacity of a more generalized “trophic”factor which regulated the functional status of maturesensory neurons. It is unclear how NGF increases pNFsynthesis in mature DRG neurons. NGF is also pro-duced by Schwann cells in the distal stump followingaxotomy (5) which primarily appears to act locally topromote axonal regeneration. The function of NGFsynthesized by Schwann cells in the distal stump isunknown. Schwann cells in the axotomized distalstump increase both NGF synthesis and the numberof their low-affinity NGF receptors (5). It is known

148

that NGF specifically supports small-fiber sensoryneurons and sympathetic neurons. These neuronsdegenerate early and prominently in the course ofdiabetic polyneuropathy and their degeneration isresponsible for most of the symptoms and signs thatare particularly bothersome to the diabetic patient.Administration of NGF prevented the reduction ofsubstance P and calcitonin gene-related peptide(CGRP) levels in the sciatic nerve in streptozotocin(STZ)-treated rats (6). However, NGF administrationdid significantly improve diabetic nerve conductionabnormality, but this was not unexpected becausenerve conduction measures are dependent primarilyon large fiber function (7).

Talin, a high molecular weight protein of 225–235kDa, is found in a variety of tissues and cell types (8).Talin immunoreactivity is normally found at endo-neurial and epineurial vessel walls, perineurial andepineurial fibroblasts in all the human sural nervesexcept diabetic nerves. In vitro studies have proposedthat talin can interact with at least two other proteinsthat are localized at adhesion plaques, vinculin andintegrin, and that they are involved in transmembraneconnections between the extracellular matrix and thecytoskeleton (9). A. Mazzeo and her co-workers sho-wed a decrease in 235 and 190 kDa bands of talin inthe muscle homogenate of diabetic neuropathy pa-tients by Western immunoblotting. A striking featureis the reduced amount of talin immunoreactivity atinner and outer lamellae of the perineurium only inpatients with diabetic neuropathy (10). Since talin isinvolved in transmembrane connections betweenextracellular matrix and cytoskeletal (8) we proposeto expand this work by assessing the pattern of talinimmunoreactivity in diabetic rats treated with NGF.

The possibility that abnormal availability of neu-rotrophic factors, in particular NGF, contributes to thepathogenesis of diabetic neuropathy has been extensi-vely investigated in recent years. Several studies sug-gest that retrograde transport of endogenous NGF isimpaired in diabetic animals (7, 11, 12). It was shownthat rats’ NGF levels were increased in target tissuesand decreased in sciatic nerve and sympathetic gangliain diabetic rats. These data suggested that retrogradeaxonal transport of NGF was impaired (13). Otherstudies have shown that NGF expression itself is re-duced in experimental and clinical diabetes (14). Ho-wever, a direct connection between abnormal availa-bility on NGF and the pathogenesis of peripheral neu-ropathy has not been demonstrated yet. Likewise, therelation between the NGF and the cytoskeleton pro-teins such as talin in diabetic neuropathy has not been

fully investigated. In order to understand the mecha-nism of NGF action in diabetes and its relation withtalin, the changes in the immunoactivity and localiza-tion of talin protein were assessed in sciatic nerves ofdiabetic rats after treatment with NGF.

Materials and methodsAnimals. All animals were maintained and cared

for as outlined in the guide for the care and used oflaboratory animals (Kuwait University, Faculty ofHealth Publication). Animals were housed in areawhich is well-ventilated, air-conditioned and providedwith independently adjustable light-dark cycle systemand temperature regulation system. The rooms andanimal cages were cleaned daily and the animals wereprovided with fresh food and water on a daily basis.In addition, all animals were inspected daily for anypossible signs of inflammation, respiratory or gastro-intestinal disease. If such signs were present, the in-dividual animal was not used in this protocol.

Experimental protocols and treatments. Adult maleSprague-Dawley rats (Kuwait Animal LaboratoryCenter colony) were group-housed (4–5 animals percage) in standard plastic cages with wood chip bed-ding. Bedding was changed daily for all animals tomaintain sanitary conditions. In individual experi-ments, rats (180–220 g) were injected intraperitoneally(ip) in groups of 5–10 animals with 75 mg/kg (15)streptozotocin (STZ) (Sigma, St. Louis, MO) (40–50mg/kg (16)) dissolved in 0.9% saline. Control animalswere injected with an equal volume of vehicle (0.9%saline). Non-diabetic rats were studied at this age asonset control, to provide a starting value against whichto judge any diabetes induced neuropathic deteriora-tion. STZ solutions were freshly prepared due to thelimited stability of the compound (17). Blood glucoselevels were determined in all animals using an EncoreGlucometer (Bayar, Elkhart, IN) from blood samplesobtained by tail vein bleeds. Since elevated bloodglucose levels appear to stabilize approximately3 weeks post STZ treatment (18), blood glucose levelswere assessed 3–4 weeks following STZ treatment.Rats with blood glucose levels >250 mg/dl (>14 mM)were considered diabetic and used for further studies.

Diabetic rats were verified 24 hours by the pre-sence of hyperglycemia and glucosuria using VisidexII and Diastix kits (Ames Slough, UK). After finalexperiments, plasma glucose was estimated (GOD-Perid method; Boehringer Mannheim, Mannheim,Germany) in samples taken from the tail vein. Whilefew detailed investigations of the time-course for theexpression of STZ-induced alterations in nociceptive

Medicina (Kaunas) 2006; 42(2)

Waleed M. Renno, Farid Saleh, Ivo Klepacek, Farida Al-Awadi

149

responses have been reported, other data also suggestthat there is some variability between studies in thetemporal expression of altered nociception followingSTZ-treatment. For example, the onset of mechanicalhyperalgesia (decreased withdrawal latency to pawpressure) has been reported to range from one (19) to3–4 weeks (18). C. Courteix and his co-workers (18)reported that STZ-induced thermal allodynic respon-ses were present as early as 2 weeks post STZ treat-ment in a subpopulation of diabetic rats, and that thissubpopulation increased in number with time follo-wing STZ treatment. Similarly, STZ induced increasesin nociception in the formalin test were found to de-velop over time requiring at least 4 weeks to fullydevelop (18, 20). After 2 weeks of uncontrolled dia-betes, blood glucose was assessed and the animalswere divided into diabetic control group and experi-mental group for treatment with nerve growth factor(NGF) for 12 weeks.

NGF experimental protocol. Human recombinantNGF (hrNGF: Boehringer Mannheim GmbH, Bio-technology Center Pensberg, FRG) was administeredat a dosage of 0.5 mg/kg according the proceduredescribed by J. W. Unger et al. (21). Briefly, NGFwas stored at –80°C and diluted in phosphate bufferedsaline, pH 7.4, containing cytochrome c (BoehringerMannheim GmbH) at a concentration of 0.1 mg/ml asa carrier protein to minimize loss of NGF due to ad-sorption to plastic ware. Dilutions were prepared fresheach day. A volume of 1 ml/kg body wt was injectedsubcutaneously in the neck region on a three timesweekly dosing schedule for 12 weeks using syringeswith Ultra-Fine gauge needles (Beckton Dickinson,Heidelberg, FRG).

Specimens collection and electron microscopy.Animals were deeply anesthetized (Pentobarbitone,50 mg/kg ip) and perfused transcardially with 200 mlof phosphate buffer saline (PBS) followed by 500 mlof 4% paraformaldehyde / 0.2% glutaraldehyde. Thesciatic nerves were dissected out and post-fixed for48 hours. Then the tissues were processed for EM asdescribed previously by M. W. Renno et al. (22). Brief-ly the tissue sections were rinsed in phosphate bufferand post-fixed in osmium tetroxide and dehydratedin a series of graded ethanol. After infiltration withpropylene oxide and Araldite for 4 h and then with100% Araldite overnight, the tissues were embeddedin fresh Araldite resin. Semi-thin sections were asses-sed first for proper tissue orientation. Ultrathin sec-tions were cut and mounted on nickel grids. Finally,the tissue sections were counterstained with uranylacetate and lead citrate and viewed in the transmission

electron microscope (EM unit at the Health ScienceCenter, Faculty of Medicine).

For immunocytochemical post-embedding labe-ling, the protocol used has been described previously(22). Briefly, the sections were washed in Tris buffersaline (TBS) pH 7.6 buffer (0.05 M Tris pH 7.6 with0.9% NaCl and 0.1% Triton X-100) and drained bytouching the edge of grids to Kimwipe, followed byincubation in anti-talin primary monoclonal antibody(Sigma T3287; clone 8d4 ascites fluid; Sigma-Aldrich,USA) diluted 1:20 in TBS pH 7.6 overnight in a moistchamber at room temperature. The next day, thesections were washed twice in TBST pH 7.6 buffer5 min each followed by a third wash for 30 min. Thenthe sections were conditioned in TBS pH 8.2 buffer(0.05 M Tris pH 8.2 with 0.9% NaCl and 0.1% TritonX-100) for 5 min. Grids were then incubated with anti-mouse IgG (Sigma) conjugated to 10 nm gold particles(diluted 1:20 in TBS pH 8.2 buffer) for 1 h, washedtwice in TBS pH 7.6 for 5 min each, and rinsed indouble distilled water (ddH2O). Finally, the sectionswere counterstained with uranyl acetate and lead cit-rate. Since myelin is a phospholipid, its staining oftenappears weak, whenever low percentage of glutaral-dehyde is used. In our study, we had to use a low per-centage of glutaraldehyde since higher percentagewould mask the talin antigen. Also, we used 0.1%Triton X-100 as an etching agent to remove the os-mium and Araldite plastic in order to improve anti-genic immunoreactivity by exposing antigenic sitesand restoring antigenic affinity to the antibody. Thisin turn may affect the ultrastructural preservation ofmyelin (22–24). However, the usage of low percentageof glutaraldehyde and etching agent does not affectour interpretation of the results since we are comparingand contrasting the specific staining for talin, notmyelin.

For morphometric analysis, three sections fromeach block (two blocks from each rat) were subjectedto quantitative analysis according to procedures des-cribed earlier by many investigators (21–23, 25). Sec-tions from the five rats from each group were quan-titated for sciatic nerve profiles. Random electronmicrographs were taken from the sciatic nerve toassess the immunogold labeling densities in the dif-ferent profiles examined in this study (Schwann cells,endoneurial and perineurial fibroblast, myelinated andunmyelinated nerve fibers, extracellular matrix andendothelium of vasa nervosa).

Quantitation of post-embedding immunostaining.Ultrathin sections were examined under a JEOL1200EX II transmission microscope at 80 kV. Sections


Talin immunogold localization in NGF treated diabetic rats

150

were photographed randomly at a plate magnificationof ×15,000 and printed at a final magnification of×18,000. All gold particles were counted over eachprofile. Aggregates of two or more gold particles werecounted as one. Aggregates could be recognized fromthe uniform small distance among the particles. Oc-casionally, close by, independently attached particlesmay have been counted as one. This would occurmainly in densely labeled structures resulting in theunderestimation of particle density. Since, at the most,only a few percent of the particles were aggregates,this underestimation would not affect the differencesamong cell profiles in the present study. For quan-titative analysis, the number of gold particles per squ-are micrometer for each structure was calculated ac-cording to the morphometric technique of M. Ben-dayan et al. (24) as previously described (26). Briefly,the surface area (Sa) of individual histological struc-tures was first measured using a digitizing tablet (Sig-ma Scan V3.90; Jandel Scientific, Corte Madera; CA)and associated microcomputer (computer model soft-ware WSYE pc+). An electron micrograph was posi-tioned on the tablet, and the perimeter of the profileto be measured was traced. The number of gold par-ticles (Ni) present over individual Schwann cells,endoneurial and perineurial fibroblast, myelin sheath,myelinated and unmyelinated nerve fibers, extracel-lular matrix and endothelium of vasa nervosa was thencounted and the density of labeling (Ns) calculated asNs=Ni/Sa. A background gold particle density persquare micrometer was calculated for each animalfrom immunohistochemically-processed sections inwhich the talin primary antibody was replaced withnormal rabbit serum. This background density of goldparticles was then subtracted from the calculated talingold particles densities to obtain a final density persquare micrometer. A one-way analysis of variance(ANOVA) was applied to assess significant differen-ces between the mean values of gold particles amongthe different neuronal structures quantified in control,diabetic and NGF treated diabetic rat groups. TheScheffe F-test was used to determine individual pro-bability values for multiple comparisons. In addition,the surface density of gold particles was quantitatedover individual neuronal and non-neuronal elementsidentified on the basis of their fine structural featuresand their location in the sciatic nerve in order to de-termine whether the observed differences in talin im-munolabeling were statistically significant. All goldparticles were counted over each profile (Table), andthe results are summarized in Fig. 1 and described inthe text.

ResultsSpecific control. Routine control experiments

(blocking with the talin antigen, omission of primaryantiserum, or substitution of primary monoclonalantibody with preimmune serum) confirmed thespecificity of the immunolabeling. The talin mono-clonal antibody had been shown to possess consi-derable specificity for talin (27). The specificity ofthe staining observed at the electron microscopiclevels is further demonstrated by the following. First,there was no staining when the talin antibody wasabsorbed by preincubation with 100 µM talin. Figure2A is an example of preabsorption of talin antibodywith synthetic talin that significantly reduced the goldlabeling density of talin to background levels. Secondonly background level was observed in the sectionsfollowing the deletion of anti-talin monoclonal anti-body (Fig. 2B). Finally, the labeling was not observedwhen normal mouse serum was substituted for talinmonoclonal antibody during staining procedure (Fig.2C). These control experiments support the conclusionthat the anti-talin monoclonal antibody used in thisstudy recognizes talin. At the ultrastructural level, onlybackground labeling (5.7±4.3 particles/µm2) wasobserved in sections treated with normal mouse serum.This immunogold background labeling was subtractedfrom the values presented in Fig. 1 to provide an indexof specific labeling in the sciatic nerve profiles.

Talin immunolabeling. A low background densityof gold particles was distributed over the entire tissuesection. Within labeled profiles, however, the goldparticles were found at much higher density; thusimmunolabeled profiles were readily distinguishedfrom the backgrounds. Morphological analysis of thedata (Fig. 1) shows two distinctive populations. “Labe-led” or “immunoreactive” profiles contain more that20 particles/µm2 and “unlabeled” profiles contain less

Table. Number of each type of profile measuredin this study

Numbermeasured

Schwann cells 88Myelinated nerve fibers 110Myelin sheath 110Unmyelinated nerve fibers 102Perineurial and epineurial fibroblasts 69Extracellular matrix 78Endothelium of vasa nervosa 44

Cellular profiles



151

than 11 particles/µm2 and are not statistically signi-ficant from gold density labeling of extracellular mat-rix (control group 9.5±5.3 particles/µm2; diabeticgroup 10.9±4.1 particles/µm2; NGF treated diabeticgroup 8.5±3.9 particles/µm2). Using these criteria, 88Schwann cells, 69 fibroblasts, 110 myelinated nervefibers, 102 unmyelinated nerve fibers, 69 extracellularareas, and 44 endothelial cells of vas nervosa wereclassified as talin immunolabeled in the sciatic nerve(Table).

Electron microscopic examination of sectionsthrough the sciatic nerve revealed intense immunogoldlabeling over the Schwann cells (Fig. 3A) and fib-roblasts (Fig. 4A). Analysis of the cellular distributionof talin immunoreactivity in the sciatic nerve showedthat particle density for talin was greatest over labeledprofiles (Fig. 1). These labeled Schwann cells andfibroblasts contained approximately 139.6±5.6 partic-les/µm2 and 127.4±4.1 particles/µm2, respectively, insciatic nerve of control group. The immunogold

localization of talin in diabetic rats was significantly(p<0.001) reduced in Schwann cells (66.3±6.5 par-ticles/µm2) (Fig. 3B) and perineurial and epineurialfibroblasts (56.8±3.9 particles/µm2) (Fig. 4B). Talinimmunogold labeling decrease in Schwann cells andfibroblasts of diabetic group was approximately 47%and 45%, respectively. Diabetic rats treated with NGFfor 12 weeks showed significant (p<0.005) increase(Fig. 1) in talin-like immunogold density in Schwanncells (Fig. 3B) and fibroblasts (Fig. 4C). Talin immu-nogold density in Schwann cells and fibroblasts in-creased approximately 68% and 58%, respectively,after NGF treatment when compared to diabetic rats.

In the five control animals studied, talin-likeimmunogold labeling over the endothelial cells ofcapillaries was significantly (p<0.02) different fromthe background levels (Fig. 1). The density of talingold particles over the endothelial cells of vas nervosawas approximately 77% and 75% less than golddensity over the Schwann cells and fibroblasts of the

Fig. 1. Quantitative analysis of the distribution of talin-like gold particle densities in different cellu-lar profiles of the sciatic nerve of control, diabetic and NGF treated diabetic rat groups

Values represent number of particles/µm2 ± SEM after the background density of gold particles was sub-tracted from the calculated densities of talin gold particles to obtain a final density per µm2. Twelve weeksfollowing the induction of diabetes, talin-like immunogold density in STZ-induced rats was significantlydecreased compared to non-diabetic controls and diabetic rats treated with NGF (ANOVA: *p<0.001). Ad-ministration of NGF, subcutaneously three times a week over the same time course, prevented the decrease intalin-like immunogold staining (**p<0.005). The gold particle density staining over the endothelial cells ofcapillaries was significantly (+p<0.02) different from the background in the control and diabetic rats. Admin-istration of NGF, however, did not change the gold particle density of talin-like immunoreactivity in theendothelial cells of nerve capillaries. A Scheffe F-test of multiple comparisons revealed that no significantdifference existed between the gold particle density overlying the myelinated nerve fibers, unmyelinated

nerve fibers, extracellular matrix and that of background.

Den

sity

of g

old

part

icle

s/µm

2 (m

eans

±SE

M)

controldiabeticdiabetic + NGF

Schwanncells

Myelinatednerve fibers

Myelinsheath

Unmyelinatednerve fibers

Perineurialand epineurial

fibroblasts

Extracellularmatrix

Endothelium(vasa nervosa)



152

Fig. 2. Examples of electron micrographs of control sections from the sciatic nerve of the control anddiabetic rat groups

A – effect of talin pre-absorption on immunolabeling. Note that pre-absorption of talin primary antibodyabolished immunolabeling (arrows) significantly (Ax, axon; asterisks, myelin sheath; U, unmyelinated nervefiber), ×20,000; B – control section in which the first antibody (anti-talin antibody) was deleted during theimmunogold labeling. Note the insignificant amount of gold particles was seen on the different profiles exam-ined (Sc, Schwann cells), ×20,000; C and D – control sections in which the talin antiserum was replaced withnormal mouse serum followed by secondary anti IgG conjugated to 10 nm gold particles. Note that only asmall number of background gold particles (arrows) were seen on different profiles examined in this study

(E, endothelium), ×18,000 and ×18,000.



153

Fig. 3. Electron micrographs of talin-like immunoreactive Schwann cells (Sc) of the rat sciatic nerveA – electron micrograph depicting talin-like immunoreactivity within Schwann cells of the rat sciatic nerve ofcontrol group. Gold particles representing talin immunolabeling are indicated by arrows, ×20,000; B – exam-ple of talin immunolabeling of Schwann cells of diabetic STZ induced rat group. Note the significant decreasein gold labeling in Schwann cells in these diabetic animals, ×20,000; C – electron micrograph of talin-immu-noreactive gold-labeled Schwann cells in NGF treated diabetic rat group. Notice the significant gold particles(arrows) increase in NGF treated diabetic rat group compared to diabetic one, ×20,000 (asterisks, myelin

sheath; Ax, axons).

A – Control

B – Diabetic

C – Diabetic + NGF



154

Fig. 4. Electron micrographs depicting immunogold talin-like immunoreactive perineurialfibroblasts (Fb) of the rat sciatic nerve

A – micrograph depicting talin-like immunoreactivity (arrows) within the fibroblasts of the rat sciatic nerve ofcontrol group, ×18,000; B – example of talin immunolabeling of fibroblast cells of diabetic rat group. Note thesignificant decrease in gold labeling (arrows) in fibroblast in these diabetic animals, ×18,000; C – electronmicrograph of talin-immunoreactive gold-labeled fibroblast in NGF treated diabetic rat group. Note the sig-nificant gold particle (arrows) increase in the fibroblasts of the NGF treated diabetic rat group compared to

diabetic group, ×18,000 (C, collagen fibers).

A – Control

B – Diabetic




155

Fig. 5. Electron micrographs of talin-like immunoreactive endothelium cells of vasa nervosa(perineurial and epineurial vessels) of the rat sciatic nerve

A – micrograph showing talin immunogold labeling of endothelium cell of a capillary in the sciatic nerve ofrat control group, ×20,000; B – electron micrograph of epineurial vessel from a section of sciatic nerve fromdiabetic rat group, ×20,000; C – micrograph showing talin-like immunoreactivity (arrows) of endotheliumcells of epineurial vessel in the sciatic nerve of NGF treated diabetic group. Note that the endothelial cells ofvas nervosa are significantly labeled with talin antibody compared to background gold density level; however,there is no significant difference of gold particle density in these cells among control and treated groups,

×20,000 (E, endothelium; Fb, fibroblast; L, lumen; C, collagen).



A – Control

B – Diabetic


156

control rats, respectively. The endothelial cells of en-doneurial and epineurial vessel walls showed no sig-nificant change (Fig. 1) in talin-like immunogold par-ticle density among control (32.3±9.1 particles/µm2)(Fig. 5A), diabetic (21.9±7.6 particles/µm2) (Fig. 5B)and NGF treated diabetic (26.5±8.1 particles/µm2)(Fig. 5C) animals.

The mean gold particle densities of talin-likeimmunoreactivity over the myelinated (nerve fibers(8.7±2.1 particles/µm2) and myelin sheaths 4.2±2.2particles/µm2)) (Fig. 6) and the unmyelinated nervefibers (6.1±3.2 particles/µm2) (Fig. 7) and the extra-cellular matrix (9.5±5.3 particles/µm2) (Fig. 8) profileswere insignificant from background levels. Talin-likeimmunoreactivity in these profiles did not show anysignificant changes in diabetic or NGF treated animals(Fig. 1).

DiscussionIn order to understand the talin distribution in the

normal sciatic nerve and the effect of NGF on thediabetic rats, we utilized immunogold post-embeddingtechnique at the electron microscopic level to examinethe relationship between talin distribution and NGFtreatment in diabetic animals. Our morphometric ana-lysis was performed on the sciatic nerve because thecommon type of peripheral neuropathy associated withdiabetes in humans is the loss of the distal region oflong and large-diameter axons with survival of mostneuronal somata. These pathological changes arecharacteristically associated with a stocking-and-glovepattern of sensory loss, which is later accompaniedby a similar distribution of motor weakness. Suchchanges are found in a number of vitamin-deficiencystates, metabolic disorders such as diabetes mellitusand advanced age. These conditions are supplementedby a large number of localized mechanical nerveinjuries that lead directly to distal axonal degeneration.The presence of axonal loss, together with the depen-dency of the axon of the neuronal soma, suggests thatdefects or interruptions in the delivery of materialsvia axonal transport are the key to the pathogenesisof axonal degeneration. In the course of this disease,many diabetic patients end up suffering from a varietyof neuropathies, which may include acute painfuldiabetic neuropathy, ataxic and acrodystrophic neuro-pathy (ataxia, foot ulceration, and neuropathic arthro-pathy). Neuropathies are undoubtedly the commonesttype of neurological dysfunction found in diabetes,although the exact prevalence varies according to thediagnostic criteria used (28–30).

In the normal sural nerve, the endoneurial and epi-

neurial vessels are intensely positive for vinculin andtalin. There is spindle-shaped immunoreactivity fortalin which correspond to fibroblasts. Strong and con-centric lines of immunofluorescence for vinculin andtalin are usually seen in the perineurium. Gold immu-noelectron microscopy showed both proteins labelingthe membrane of the perineurial fibroblasts (31, 32).A variety of osmiophilic inclusion bodies (dense os-miophilic staining and paracrystalline laminated ap-pearance) are often associated with disrupted myelinsheath lamellae, lysosomal bodies and periaxonal ex-pansions of Schwann cell cytoplasm in diabetic ani-mals. An unusual Schwann cell immunoreaction isusually evident in nerves of more severely diabeticrats, involving lysosomal digestion of large, pinched-off portions of myelinated nerve fibers. Unmyelinatednerve fibers and associated Schwann cells of both dia-betic and control rats rarely contained inclusion bodiesand secondary lysosomes (25). Experimental (strepto-zotocin-induced) diabetic neuropathy is also associa-ted with a reduced amount of fast anterogradely trans-ported proteins but no overall change in transport ran-ge (13). A defect in fast anterograde transport of spe-cific receptors and neurotransmitter enzymes as mea-sured by their accumulation proximal to nerve ligaturehas been demonstrated in STZ-induced diabetic neuro-pathy (33, 34). Decreased retrograde axonal transportof glycoproteins, measured by accumulation distal toa nerve ligature (13), has also been found in experi-mental STZ-induced diabetic neuropathy. These datamight suggest that an early defect in retrograde axonaltransport underlie the development of axonal degene-ration in these various neuropathies. However, recentevidence indicates that a defect in retrograde axonaltransport leads to a loss of target tissue-derived “tro-phic” support. This hypothesis appears to be consistentwith the decrease in retrograde transport of NGF thathas been observed in diabetic neuropathy (12).

In the present study, we found that intense talin-like immunogold labeling over the Schwann cells andfibroblasts of normal sciatic nerves where as theendothelial cells of capillaries were moderately labe-led. These findings are in agreement with the resultsreported by A. Mazzeo and her co-workers (10). Inaddition, the immunogold localization of talin in dia-betic rats was significantly reduced in Schwann cellsand perineurial and epineurial fibroblasts. Talin im-munogold labeling decrease in Schwann cells andfibroblasts of diabetic group was approximately 47%and 45%, respectively. Previous investigation (10) re-ported that talin immunoreactivity was found at endo-neurial and epineurial vessel walls, perineurial and



157

Fig. 6. Electron micrographs of talin-like immunoreactive myelinated nerve fibers of the rat sciatic nerveA – micrograph showing talin immunogold-labeled myelinated nerve fibers in the sciatic nerve of the controlgroup, ×20,000; B – micrograph depicting talin immunogold labeling in myelinated nerve fibers of the sciaticnerve in diabetic rat group, ×20,000; C – example of a micrograph showing talin-immunogold labeling in themyelinated nerve fibers of the sciatic nerve taken from diabetic rats treated with NGF for 12 weeks after theinduction of diabetes, ×20,000. Note that only a small number of background gold particles (arrows) are

evident in these micrographs (asterisks, myelin sheath; Ax, axon).

A – Control

B – Diabetic




158

Fig. 7. Electron micrographs of unmyelinated nerve fibers of rat sciatic nerve sections stainedwith talin monoclonal antibody

A – micrograph from control group showing a number of unmyelinated nerve fibers (U) enclosed with signifi-cantly talin-like immunoreactive Schwann cell (Sc). Note that the unmyelinated nerve fibers show noimmunogold labeling, ×20,000; B – electron micrograph of a group of unmyelinated nerve fibers of diabeticrat group showing only background immunolabeling (arrows) of talin-like immunoreactivity, ×20,000; C –Micrograph showing sections of sciatic nerve fibers from NGF treated diabetic rat group. The unmyelinatednerve fibers show insignificant immunogold labeling for talin monoclonal antibody. Note the significant labelingof Schwann cell which is enclosing the unmyelinated nerve fibers, ×20,000 (asterisks, myelin sheath; Ax, axon).

A – Control

B – Diabetic




159

Fig. 8. Electron micrographs of talin-like immunogold staining of the extracellular matrixof the rat sciatic nerve

A – electron micrograph depicting talin-like immunoreactivity in the extracellular matrix of the rat sciaticnerve of control group, ×20,000; B – example of talin immunolabeling of extracellular matrix of STZ-induceddiabetic rat group, ×20,000; C – micrograph of talin immunoreactive gold-labeled extracellular matrix inNGF treated diabetic rat group. Note that only a small number of background gold particles (arrows) areevident in control and treated animals, ×20,000 (asterisks, myelin sheath; Ax, axon; Fb, fibroblast; C, collagen).

A – Control

B – Diabetic




160

epineurial fibroblasts in normal human sural nerves;in addition, talin binding was markedly reduced inthe perineurium of the sural nerves from patients withdiabetic neuropathy. This total talin decrease in pa-tients with diabetic neuropathy was confirmed by Wes-tern immunoblotting. However, neither immunofluo-rescence light microscopic study nor western immu-noblotting techniques (10) could show the specificdecrease in Schwann cells or the changes in differentprofiles. Using immunogold localization at electronmicroscopic level in conjunction with quantitativeanalysis, we were able to precisely determine the talin-immunoreactive cellular profiles in the sciatic nerveand determine the approximate changes in the talinimmunogold labeling, quantitatively.

Our results demonstrate that diabetic rats treatedwith NGF for 12 weeks showed significant increasein talin-like immunogold density in Schwann cells andfibroblasts of the diabetic sciatic nerves. Talin immu-nogold density in Schwann cells and fibroblasts in-creased approximately 68% and 58%, respectively,after NGF treatment when compared to diabetic rats.These data support the therapeutic role of the NGF inrestoring the normal functions of the peripheral nervesin diabetic neuropathy as indicated by many recentstudies (6, 7, 35). In addition, our study suggests thatthe NGF mechanism of action in diabetic neuro-pathy involves the modulation of talin protein inSchwann cells and fibroblasts in peripheral neuropa-thic nerve.

The physiologic source of NGF for DRG cells isprovided by local synthesis in Schwann cells or byexpression in the peripheral innervation area. NGFreaches the soma by retrograde transport after intra-cellular uptake (36). However, NGF receptors are notrestricted to DRG neurons, but are also present in theprojection area within the spinal cord (lamina 1 and2), suggesting a modulatory role of NGF in the sensorypathways also at the level of the central nervous system(37). In addition, neurotrophins exert their influencenot only along the central direction of the neuronalpathway of the sensory system, but also within thereceptor field of the primary sensory neuron, i.e. recep-tors in the skin, A delta- for mechano-nociception orC-fibers for thermo-nociception. A previous study hasdemonstrated that collateral sprouting of A delta- andC-fibers in the partially denervated skin is supportedby endogenous NGF and can be blocked by theapplication of NGF antibodies, whereas regrowth ofcrushed nociceptive axons occurred surprisinglyindependent from NGF (38). In light of these findings,J. Diamond and his co-workers (38) suggested that

the cells possess additional properties for the inductionof regenerative mechanisms.

These previous investigations have shown potenteffects of NGF on the structure and function of thesomato- and viscerosensory system. In addition, recentstudies, using animal models of type I diabetesmellitus, have pointed toward a neurotrophic deficitaccompanying the metabolic disease which may playa critical role in the development of neuropathy, i.e.,reduced axonal NGF transport of the DRG in sciaticnerves of diabetic rats with no change in NGF bindingsites (11). Furthermore, NGF protein and NGF mRNAwere found to be reduced in skin and muscle of un-treated diabetic rats four weeks after induction of dia-betes. NGF treatment for 8 weeks was able to correctthese defects (6). Based on the data summarized above,a “neurotrophic hypothesis” of diabetic neuropathywas developed stating that insulin deficit and/orhyperglycemia causes reduced growth factor synthesisin target areas and locally in Schwann cells and dis-turbance of retrograde axonal transport to the perika-ryon. These changes subsequently lead to reducedgene expression of neurotrophin-dependent substan-ces, i.e. transmitters, structural proteins, and neurofi-laments. Thus, lack of neurotrophic support may bepartially responsible for degenerative changes in theneuronal and axonal structure that may contribute tothe impairment in nerve function such as reduced con-ductivity due to axoglial disjunction with shifting ofvoltage-dependent sodium-ion channel (14).

The blood-nerve barrier of the perineurial andendoneurial blood vessels maintains the normal ho-meostasis of the nerve. Perineurium protects the endo-neurium from substances that might leak from per-meable epineurial vessels and it also controls transpor-ting selectivity materials and fluid through the endo-neurial spaces (39). At the level of the tight junctions,the extracellular space between perineurial fibroblastsis obliterated and this cell-to-cell communication ismost likely related to the barrier function of the peri-neurium. The localization of talin at Schwann andperineurial fibroblasts in normal nerve suggests thatthis protein may be implicated in the permeabilitybarrier property of the perineurium. Previous studies(10, 31) also support this role. In diabetic nerves,thickening of the basal laminae of the perineurialfibroblasts (40) and perineurial calcifications whichare mainly localized to the outer lamellae (41) havebeen documented. In addition, disorganized perineu-rial tight junctions in diabetic polyneuropathy (42)and endoneurial edema in experimental galactoseneuropathy which is associated with altered tight junc-



161

tions of the perineurial fibroblasts (43) suggest thatsimilar mechanisms could result in the alterations ofthe perineurial tight junctions in diabetic neuropathyand account for its impaired permeability properties.Our study supports the above investigations and con-clusions; in addition, our data implicate a strong rolefor NGF to restore the normal functions of the tightjunctions by restoring the talin, cytoskeleton protein,in diabetic sciatic nerve.

Although talin-like immunogold labeling over theendothelial cells of capillaries was significantly dif-ferent from the background levels, these cells ofendoneurial and epineurial vessel walls showed nosignificant change in talin-like immunogold particledensity among control, diabetic and NGF treated dia-betic animals. The density of talin gold particles overthe endothelial cells of vas nervosa was approximately77% and 75% less than gold density over the Schwanncells and fibroblasts of the control rats, respectively.Our data indicate that talin may play a minor role inthe vascular abnormalities of the peripheral nerves ofdiabetic animals. Furthermore, this study does notsupport any role for the NGF on talin distribution inthe endothelial cells of nerve capillaries. In spite ofthe fact that the barrier properties in the peripheralnerve are attributed to the presence of extensive tightjunction complexes in the perineurium and the cellsof endoneurial vasculature, A. A. Sima and his co-workers (44) concluded from their study that the en-

dothelial tight junctions abnormalities found in dia-betic neuropathy are unlikely to have any pathogeneticsignificance. Our study, indeed, supports such a con-clusion for the tight junctions in endothelial cells ofvasa nervosa of diabetic sciatic nerve.

In conclusion, these data suggest that in diabeticneuropathy the role of NGF is mainly confined toSchwann cells and fibroblasts of peripheral nerve. Thetalin, cytoskeleton protein, in these two profiles seemsto play a major role in maintaining normal functionin the nerve fibers and its decrease in diabetic animalssuggests a possible role in the pathogenesis of diabeticneuropathy seen in peripheral nerves. In light of thefact that talin is a major component of the tight junc-tions in perineurial and Schwann cells, the alterationof talin in diabetes could lead to abnormal functionof blood-nerve barrier which in turn may expose theaxonal fibers to harmful components. In addition, thisstudy shows that neurotrophic (NGF) support may beessential for maintaining talin protein which in turnpreserves functional significant tight junctions in theperipheral nerves of diabetic neuropathy.

AcknowledgementsWe thank Amal A. Wagdi, Mona Rostum, and

Mohamed Al-Gallaf for their technical assistance.This study was supported by a grant from Research

Administration, Kuwait University, grant numberMA01/00.

Talino, žymėto imunoaukso dalelėmis, tankumo padidėjimas žiurkėms, kuriomssukeltas cukrinis diabetas, sėdmens nerve po gydymo nervų augimo faktoriumi

Waleed M. Renno, Farid Saleh, Ivo Klepacek, Farida Al-Awadi1

Kuveito universiteto Medicinos fakulteto Anatomijos katedra, 1Biochemijos katedra, Kuveitas

Raktažodžiai: talinas, sėdmens nervas, diabetinė neuropatija, Švano ląstelės, endoneuriumas, perineuriumas.

Santrauka. Diabetinė neuropatija – tai sekinanti liga, kurios atsiradimo priežastys kol kas nėra pakankamaiištirtos. Naujausių tyrimų duomenimis, pacientams, sergantiems diabetine neuropatija, rastas sumažėjęs nervųaugimo faktoriaus aktyvumas bei ląstelės griaučių baltymo talino imunoreaktyvumas perineuriume.

Darbo tikslas. Talinas dalyvauja tarpmembraninėje sąveikoje tarp tarpląstelinio užpildo ir ląstelės griaučių,todėl šio tyrimo tikslas – ištirti talino imunoreaktyvumo pobūdį ir nervų augimo faktoriaus poveikį talinopasiskirstymui žiurkių, kurioms sukeltas cukrinis diabetas, sėdmens nerve.

Medžiagos ir metodai. Talino imunoreaktyvumas sveikų žiurkių, žiurkių, kurioms sukeltas cukrinis diabetas,ir nervų augimo faktoriumi gydytų žiurkių, kurioms sukeltas cukrinis diabetas, sėdmens nerve stebėtas naudojantelektroninę mikroskopiją, talinui susijungus su monokloniniu antikūnu, žymėtu aukso dalelėmis.

Rezultatai. Didžiausias aukso dalelių tankumas nustatytas Švano ląstelėse (139,6±5,6 dalelės/µm2) irfibroblastuose (127,4±4,1 dalelės/µm2). Vidutinis imuninis reaktyvumas užfiksuotas nervų kraujagysliųendotelio ląstelėse (32,3±9,1 dalelės/µm2). Mielininės ir nemielininės nervinės skaidulos bei tarpląstelinioužpildo struktūros nebuvo žymėtos (atitinkamai – 8,7±2,1 dalelės/µm2, 4,2±2,2 dalelės/µm2, 6,1±3,2 dalelės/µm2,



162

References1. Johnson EM Jr., Gorin PD, Brandeis LD, Pearson J. Dorsal

root ganglion neurons are destroyed by exposure in utero tomaternal antibody to nerve growth factor. Science 1980;210:916-8.

2. Verge VMK, Tetzlaff W, Bisby MA, Richardson PM. Influenceof nerve growth factor on neurofilament gene expression inmature primary sensory neurons. J Neurosci 1990;10:2018-25.

3. Gold BG, Austin DR, Mobley WC, Storm-Dickerson T.Chromatolytic-like alterations and aberrant neurofilamentphosphorylation in neuronal perikarya following NGF anti-serum injection. Soc Neurosci Abstr 1991;17:1497.

4. Gold BG, Mobley WC, Matheson SF. Regulation of axonalcaliber, neurofilament content and nuclear localization inmature sensory neurons by nerve growth factor. J Neurosci1991;11:943-55.

5. Taniuchi M, Clark MB, Schweitzer JB, Johnson EM Jr.Expression of nerve growth factor receptors by Schwann cellsof axotomized peripheral nerves: ultrastructural location,suppression by axonal contact and binding properties. JNeurosci 1988;8:664-81.

6. Fernyhough P, Diemel LT, Hardy J, Brewster WJ, MohiudddinL, Tomlinson DR. Human recombinant nerve growth factorreplaces deficient neurotrophic support in the diabetic rat.Euro J Neurosci 1995;7:1107-10.

7. Apfel SC, Arezzo JC, Brownlee M, Fereoff H, Kessler JA.Nerve growth factor administration protects against experi-mental diabetic sensory neuropathy. Brain Res 1994;634:7-12.

8. Beckerle MC, Yeh RK. Talin: role at sites of cell-substratumadhesion. Cell Motil Cytoskeleton 1990;16:7-13.

9. Otto JJ. Vinculin. Cell Motil Cytoskeleton 1990;16:1-6.10. Mazzeo A, Rodolico C, Monici MC, Migliorato A, Aguennous

M, Vita G. Perineurium talin immunoreactivity decreases indiabetic neuropathy. J Neurol Sci 1997;146:7-11.

11. Hellweg R, Hartung HD. Endogenous levels of nerve growthfactor (NGF) are altered in experimental diabetes mellitus: apossible role for NGF in the pathogenesis of diabeticneuropathy. J Neurosci Res 1990;26:258-67.

12. Schmidt RE, Modert CW, Yip HK, Johnson Jr EM. Retrogradeaxonal transport of intravenously administered 125I-nervegrowth factor in rats with streptozotocin-induced diabetes.Diabetes 1983;32:654-63.

13. Jakobsen J, Brimijoin S, Skau K, Sidenius P, Wells D. Ret-rograde axonal transport of transmitter enzymes fucose-labeledprotein, and nerve growth factor in streptozotocin-diabeticrats. Diabetes 1981;30:797-803.

14. Dyck PJ, Giannini C. Pathologic alterations in the diabeticneuropathies of humans: a review. J Neuropathol Exp Neurol1996;55:1181-93.

15. Lynch JJ, Jarvis MF, Kowaluk EA. An adenosine kinase in-hibitor attenuates tactile allodynia in a rat model of diabeticneuropathic pain. Eur J Pharmcol 1999;364:141-6.

16. Cameron NE, Cotter MA, Horrobin DH, Tritschler HJ. Effectsof alpha-lipoic acid on neurovascular function in diabetic rats:interaction with essential fatty acids. Diabetologia 1998;41:390-9.

17. Rakieten N, Rakieten ML, Moreshwar VN. Studies on thediabetogenic action of streptozotocin (NSC-37917). CancerChemother Rep 1963;29:91-8.

18. Courteix C, Eschalier A, Lavarenne J. Streptozotocin-induceddiabetic rats: behavioral evidence for a model of chronic pain.Pain 1993;53:81-8.

19. Zhuang HX, Snyder CK, Pu SF, Ishii DN. Insulin-like growthfactors reverse or arrest diabetic neuropathy: effects on hyper-algesia and impaired nerve regeneration in rats. Exp Neurol1996;140:198-205.

20. Calcutt NA, Li L, Yaksh TL, Malmberg AB. Different effectsof two aldose reductase inhibitors on nociception and prosta-glandin. E Eur J Pharmacol 1995;285(2):189-97.

21. Unger JW, Klitzsch T, Pera S, Reiter R. Nerve growth factor(NGF) and diabetic neuropathy in the rat: morphologicalinvestigations of the sural nerve, dorsal root ganglion andspinal cord. Experimental Neurol 1998;153:23-34.

22. Renno WM, Mahmoud MSW, Hamdi A, Beitz AJ. Quan-titative immunoelectron microscopic colocalization of GABAand Enkephalin in the ventrocaudal periaqueductal gray ofthe rat. Synapse 1999;31:216-28.

23. Renno WM. Post-embedding double-gold labeling immu-noelectron microscopic co-localization of neurotransmittersin the rat brain. Med Sci Monit 2001;7:188-200.

24. Bendayan M, Roth J, Pettelet A, Orei L. Quantitative immu-nocytochemical localization of pancreatic proteins in subcel-lular compartments of the rat acinar cells. J Histochem Cyto-chem 1980;28:149-57.

25. Myers SF. Myelin-sheath abnormalities in the vestibular nervesof chronically diabetic rats. Otolaryngol Head Neck Surg

9,5±5,3 dalelės/µm2). Talino, žymėto imunoaukso dalelėmis, tankumas statistiškai reikšmingai (p<0,001)sumažėjo žiurkių, kurioms buvo sukeltas cukrinis diabetas, Švano ląstelėse (66,3±6,5 dalelės/µm2) bei peri-neuriumo ir epineuriumo fibroblastuose (56,8±3,9 dalelės/µm2). Žiurkių, kurioms buvo sukeltas cukrinis diabetasir 12 savaičių duodama nervų augimo faktoriaus, Švano ląstelėse bei fibroblastuose nustatytas statistiškaireikšmingas (p<0,005) talino, žymėto imunoaukso dalelėmis, tankumo padidėjimas. Žiurkėms davus nervųaugimo faktoriaus, talino imunoaukso dalelių tankumas Švano ląstelėse bei fibroblastuose padidėjo atitinkamai –68 ir 58 proc. Endoneuriumo bei epineuriumo kraujagyslių sienelių endotelio ląstelėse statistiškai reikšmingotalino, žymėto imunoaukso dalelėmis, tankumo pokyčio tarp kontrolinių, žiurkių, kurioms sukeltas cukrinisdiabetas, bei nervų augimo faktoriumi gydytų gyvūnų nenustatyta.

Išvados. Tyrimo duomenimis, egzogeninio nervų augimo faktoriaus vartojimas, sergant diabetine neuropatija,palaikant periferinio nervo pralaidumo barjerą, gali būti svarbus sukeliant funkciškai svarbius regeneraciniusprocesus.

Adresas susirašinėti: W. M. Renno, Kuveito universiteto Medicinos fakulteto Anatomijos katedra, P. O. Box 24923,13110 Safatas, Kuveitas. El. paštas: [email protected]



163

1998;119:432-8.26. Lee IS, Renno WM, Beitz AJ. A quantitative light and electron

microscopic analysis of taurine-like immunoreactivity in thedorsal horn of the rat spinal cord. J Comp Neurol 1992;321:65-82.

27. Otey CA, Griffiths W, Burridge K. Characterization of mo-noclonal antibodies to chicken gizzard talin. Hybridoma 1990;9:57-62.

28. Walters D, Gatling W, Mullee MA, Hill RD. The prevalenceof diabetic distal sensory neuropathy in and English com-munity. Diabetic Med 1992;9:349-53.

29. Young MJ, Boulton AJM, Macleod AF, Williams DRR, Sonk-sen PH. A multicentre study of the prevalence of diabeticperipheral neuropathy in the United Kingdom hospital clinicpopulations. Diabetologia 1993;36:150-4.

30. Tesfaye S, Stevens L, Stephesnon J, et al. The prevalence ofdiabetic peripheral neuropathy and its relation to glycaemiccontrol and potential risk factors: the EURODIAB IDDMcomplications study. Diabetologia 1996;39:1377-84.

31. Vita G, Mazzeo A, Rodolico C, Migliorato A, Gagliardi M,Cicciarello R, Muglia U, Messina C. Localization of vinculinand talin at perineurial cells of human sural nerve. NeuroReport 1995;6:2077-80.

32. Massa R, Silvestri G, Sancesario G, Bernardi G. Immuno-cytochemical localization of vinculin in muscle and nerve.Muscle Nerve 1995;18:1277-84.

33. Laduron PM, Jenssen PFM. Impaired axonal transport ofopiate and muscarinic receptors in streptozotocin-diabetic rats.Brain Res 1986;380:359-62.

34. Macioce P, Hassig R, Tavitian B, Di Giamberardino L. Axonaltransport of the molecular forms of acetylcholinesterase inrats at the onset of diabetes induced by streptozotocin. BrainRes 1988;438:291-4.

35. Tomlinson DR, Fernyhough P, Diemel LT. Role of neuro-

trophins in diabetic neuropathy and treatment with nervegrowth factors. Diabetes 1997;46 Suppl 2:S43.

36. Spoerri PE, Perrelli L, Guidolin D, Skaper SD. Co-localizationof low-and high-affinity NGF receptors on PC12 cells, C6glioma cells and dorsal root ganglion neurons. Euro J CellBiol 1993;61:256-63.

37. Molliver DC, Snider WD. Nerve growth factor trkA is downregulated during postnatal development by a subset of dorsalroot ganglion neurons. J Comp Neurol 1997;381:428-38.

38. Diamond J, Foerster A, Holmes M, Coughlin M. Sensorynerves in adult rats regenerate and restore sensory function tothe skin independently of endogenous NGF. J Neurosci 1992;12:1467-76.

39. Thomas PK, Tomlinson DR. Diabetic and hypoglycemicneuropathy. In: Dyck PJ, Thomas PK, Griffin JW, Low PA,Poduslo JF, editors. Peripheral neuropathy. 3rd ed. London,Philadelphia: W. B. Saunders and Co; 1993. p. 1219-50.

40. Johnson PC, Brendel K, Meezan E. Human diabetic perineu-rial cell basement membrane thickening. Lab Invest 1981;44:265-70.

41. King RHM, Lewelyn JG, Thomas PK. Perineurial calci-fication. Neuropathol Appl Neurobiol 1988;14:105-23.

42. Beamish NG, Stolinski C, Thomas PK, King RHM. Freeze-fracture observation on normal and abnormal human peri-neurial tight junctions: alterations in diabetic polyneuropathy.Acta Neuropathol 1991;81:269-79.

43. Beamish NG, Stolinski C, Thomas PK, King RHM, Rud A.A freeze-fracture study of the perineurium in galactose neu-ropathy: morphological changes associated with endoneurialoedema. J Neurocytol 1992;21:67-78.

44. Sima AA, Nathaneil V, Prashar A, et al. Endoneurial mic-rovessels in human diabetic neuropathy: endothelial cell dys-junction and lack of treatment effect by aldose reductase in-hibitor. Diabetes 1991;40:1090-9.

Received 25 May 2005, accepted 12 September 12Straipsnis gautas 2005 05 25, priimtas 2005 09 12



Documents

Talin immunogold density increases in sciatic nerve of …medicina.lsmuni.lt/med/0602/0602-09e.pdftalin-like immunoreactivity in the sciatic nerve of normal, diabetic and NGF treated