j ourna l homepage: www.e lsev ie r .com/ locate /expgero
Changes in cardiac innervation during maturation in long-term diabetes
Marija Bakovic a, Marina Juric Paic a, Elena Zdrilic a, Katarina Vukojevic b, Lejla Ferhatovic c,Ana Marin a, Natalija Filipovic a, Ivica Grkovic a, Livia Puljak c,⁎a Laboratory for Neurocardiology, University of Split School of Medicine, Soltanska 2, 21000 Split, Croatiab Laboratory for Early Human Development, University of Split School of Medicine, Soltanska 2, 21000 Split, Croatiac Laboratory for Pain Research, University of Split School of Medicine, Soltanska 2, 21000 Split, Croatia
⁎ Corresponding author at: University of Split School oSplit, Croatia. Tel.: +385 21 557 807; fax: +385 21 557 8
Diabetic autonomic neuropathy being a common complication of diabetes mellitus (DM) is related to anincreased risk of cardiovascular mortality. However, mechanisms underlying changes of innervation density inaffected hearts remain insufficiently understood. Hence, the aim of this study was to describe quantitativechanges of intra-myocardial nerve terminals in hearts of diabetic rats of various ages.Male Sprague–Dawley rats were injected with 55mg/kg streptozotocin (STZ) (DM group) or with citrate buffer(control). After 2weeks, 2months, 6months and 12months, sections of their hearts were analyzed in five areas—left ventricle, interventricular septum, right ventricle, anterior and posterior wall. Nerve fibers were visualizedimmunohistochemically, using antibody against a general neuronal marker, protein gene product 9.5 (PGP 9.5).Significant increase in total nerve fibers from all heart areas was observed 2weeks and 2months after diabetesinduction, followed by a decrease at 6 months and again increase at 12 months was observed in both controland diabetic rats. Significant difference between control and diabetic rats was visible after 2weeks and 2months,with diabetic rats exhibiting significantly more nerve fibers. There were no consistent differences in quantity ofnerve fibers in different areas of the heart within a particular age-related group of animals.In conclusion, cardiac innervation undergoes dynamic changes both in control and in diabetic rats, with a time-dependent significant increase in neuronal fiber density in diabetic animals. This novel information maycontribute to our understanding of pathophysiological changes associated with diabetic cardiac neuropathy.
Diabetesmellitus (DM) is one of themost common chronic diseases.Hyperglycemia is a hallmark sign of DM and is considered to be one ofthe factors causing diabetic neuropathy (Faerman et al., 1977). Cardiacsensory denervation seen in DM patients presents special clinicalconcern because it can lead to diabetic sensory neuropathy and silentmyocardial ischemia. Autonomic neuropathy is a common feature ofDM, and abnormalities of autonomic nerve fibers have been observedin DM patients who died after painless myocardial infarction (Goochand Podwall, 2004). Diabetic sensory neuropathy is characterized byloss of pain perception during myocardial ischemia and it frequentlyleads to sudden cardiac death in diabetic patients. However, despitethe severity of this complication of DM, mechanisms underlyingchanges in innervations density in diabetic hearts remain poorlyunderstood, and it requires a better understanding of anatomical distri-bution of cardiac nerves (Ieda and Fukuda, 2009).
f Medicine, Soltanska 2, 2100011.
ghts reserved.
Specific neuronal markers can be used to visualize different neuronsand their fibers. Protein gene product 9.5 (PGP 9.5) is a general neuralmarker specific for neurons and neuroendocrine cells of the peripheraland central nervous system in mammals (Schofield et al., 1995). Dueto its specific expression in neurons, PGP 9.5 has been used as a markerof innervation in numerous immunohistochemistry studies (Lauwerynsand Van Ranst, 1988; Ramieri et al., 1990; Terenghi et al., 1993;Wirnsberger et al., 1992). Although regulation of its expression andfunction of PGP 9.5 throughout the lifespan of neurons is not sufficientlyexplored, a high level of expression suggests its important role inphysiology of neurons (Kent and Clarke, 1991). There are severalreports concerning the expression of PGP 9.5 in cardiac neurons andtheir fibers, but there are no studies that quantify expression of thatprotein in hearts of diabetic rats in long-term studies, during aging.
2. Material and methods
2.1. Ethics
The experimental protocolwas approved by Ethics Committee of theUniversity of Split School of Medicine.
1474 M. Bakovic et al. / Experimental Gerontology 48 (2013) 1473–1478
2.2. Experimental animals
Male Sprague–Dawley rats, weighing 160 to 180 g, were obtainedfrom the animal husbandry of theUniversity of Split. All ratswere raisedunder controlled conditions (temperature 22±1°C, light schedule: 12hof light and 12h of dark).
2.3. Induction and validation of diabetes
Experiments were conducted on a rat model of diabetes mellitustype 1, induced by intraperitoneal injection of 55mg/kg streptozotocin(STZ) freshly dissolved in citrate buffer (pH 4.5) after an overnightfasting. Rats were fed ad libitumwith normal laboratory chow consistingof 27% proteins, 9% fat and 64% carbohydrates (4RF21 GLP, Mucedola,Settimo Milanese, Italy).
Induction of diabetes was validated by measuring plasma glucoselevel and body weight. Plasma glucose was measured with One TouchVita instrument (LifeScan, High Wycombe, UK) using blood from tailvein. Rats with glucose level above 16.5 mmol/L were considereddiabetic, and they were included in further experiments. Body weightwas measured with a scale.
Rats were divided into 4 groups, based on duration of diabetes frominjection of STZ until the end of experiment (2 weeks, 2 months,6months and 12months). Each diabetic group was matched with thecontrol group, which was kept in experiment for the same respectiveperiod. Control rats were injected intraperitoneally with citrate bufferonly. There were 6 animals in each diabetic and control group forevery required age.
2.4. Tissue collection and immunohistochemistry
Animals were anesthetized with isoflurane (Forane, Abbott Labo-ratories, Queenborough, UK) and perfused through the ascendingaorta via the left ventricle with saline followed by 300mL of Zamboni'sfixative (4% paraformaldehyde and0.19%picric acid in 0.1Mphosphate-buffered saline) at pH7.4. Heart tissuewas removed for further analysesand post-fixed in the same fixative. Hearts were cut transverselythrough ventricles and tissue was embedded in paraffin blocks, cut insections 5 µm thick and analyzed using immunofluorescence. Afterdeparaffinization, sections were rehydrated in ethanol and water.Sections were briefly rinsed with distilled water, followed by heatingin sodium citrate buffer (pH 6.0) for 12 min on 95 °C in microwaveoven. After being cooled to the room temperature, sections wereincubated with primary antibody.
Mouse antiserum PGP 9.5 (ab8189, Abcam, Cambridge, MA, USA),was diluted 1:100 in Dako REAL antibody diluent (Dako Denmark A\S,Glostrup, Denmark), and then applied. Primary antibody was leftovernight on room temperature in a humidified chamber. After rinsingsections in PBS, theywere incubated for 1h in humidified chamberwithsecondary antibody goat anti mouse IgG Texas Red T-862 (LifeTechnologies Corporation, Carlsbad, CA, USA). Sectionswere then rinsedin PBS. Staining of neural fibers was viewed and photographed usingBX51 microscope (Olympus, Tokyo, Japan) equipped with DP71 digitalcamera (Olympus, Tokyo, Japan) and processed with Cell A ImagingSoftware for Life Sciences Microscopy (Olympus Tokyo, Japan).
Sections were analyzed within five areas: left ventricle, interven-tricular septum, right ventricle, anterior and posterior wall (Fig. 1c). Ineach of the listed areas, 6 non-overlapping fields were captured forthe further analysis, under 40× objective. Each field was one image.Microphotographswere analyzed usingMetamorph software (MolecularDevices, Sunnyvale, CA, USA), measuring threshold area percent (%),which was determined by the intensity of fluorescence of neural fibers.In each picture surface of every fiber was marked, and then measuredshare of fibers in the total surface of the picture. The number andpercentage of the surface covered in marked neural fibers (thresholdarea %) were compared between control and diabetic group. After
separate analyses were conducted for different areas of the heart foreach group and period, data were pooled for all areas of the control ordiabetic rats, and analyzed again.
2.5. Statistics
For statistical analysis, Mann–Whitney test was used to examinedifferences between two groups, while multiple groups were analyzedwith Kruskal–Wallis test, followed by Dunn's post-hoc test (GraphPadSoftware, La Jolla, CA, USA) after testing distribution of the data. Thedata are expressed as mean ± SD. Statistical significance was set atpb0.05.
3. Results
3.1. Validation of diabetes
Already oneweek after induction of diabetes, glucose level in plasmaof DM animals was significantly higher compared to controls, and thesame trend was observed throughout the experiment (Fig. 1a). At thebeginning of the experiment, there was no difference in weightbetween DM1 animals and controls, however control rats kept gainingweight right to the end of the experiment, unlike DM animals whoseweight increase was much smaller (Fig. 1b).
3.2. Total neuronal fibers in diabetic and control hearts
Rats were 40–50 days old at the beginning of the experiments,before the STZ injection. Immunoreactivity for the general neuronalmarker PGP 9.5 was used to detect all axonal profiles in young andmature diabetic and control rats. PGP 9.5 positive profileswere detectedas an intense red staining within different areas of heart ventricles,mostly arranged around blood vessels (Fig. 2). When data from allanalyzed areas of the heart were combined, a significant difference inneuronal fibers was observed between control and diabetic animals at2weeks (p b 0.001) and 2months (p b 0.05), and at both times highervalues of neuronal fibers were found in hearts from diabetic animals.There was no significant difference in area covered with nerve fibersbetween control and diabetic hearts at 6 months (p = 0.457) and/or12 months (p = 0.224) when data from different areas were pooled(Fig. 1d).
3.3. Neuronal fibers in different areas of diabetic and control hearts
In ‘two weeks animals’, the proportion of the surface covered inmarked neural fibers in the left ventricle only was higher in diabeticanimals than in controls (p=0.009, Mann–Whitney test), while in theother four analyzed areas there was no difference between the twogroups (Fig. 3a).
Two months following the induction of diabetes, only significantdifference between diabetic and control rats was observed in the areaof the anterior wall, where a higher proportion of the surface coveredin marked neuronal fibers was observed in diabetic rats, compared tocontrols (p= 0.010) (Fig. 3b). Six months after the onset of diabetes,there was no difference in the surface covered in marked neuronalfibers in any of the analyzed areas of neither diabetic nor control hearts(Fig. 3c). After 12 months, significant increase in the proportion ofsurface covered in neuronal fibers was observed only in the rightventricle of diabetic rats, compared to controls (p=0.008) (Fig. 3d).
3.4. Pattern of changes in neuronal fibers of diabetic and control rats duringmaturation
Pattern of changes was the same for all areas of the heart in controland diabetic animals. The percentage of neuronal fibers was increasingfrom 2 weeks to 2 months. Subsequently a decrease was observed at
Fig. 1. Plasma glucose of control and diabetic rats throughout the experiment (a). Bodyweight of control and diabetic rats throughout the experiment (b). Transversal section of the heartthrough ventricles showing areas of interest: 1 = left ventricle, 2 = interventricular septum, 3 = right ventricle, A = anterior wall, P = posterior wall, RV = right ventricle, LV = leftventricle. Hematoxylin&Eosin staining, Scale bar 1 mm (c). Quantity of nerve fibers in all areas of the heart of control and diabetic rats at 2 weeks (2w), 2 months (2 m), 6 months(6m) and 12months (12m). Asterisk denotes significant difference: *p≤ 0.05, ***p≤ 0.001. DM=diabetes mellitus, N=6 animals per group (d). Data presented as M± SD.
1475M. Bakovic et al. / Experimental Gerontology 48 (2013) 1473–1478
6months and then an increase at 12months (Fig. 1d). All values in bothcontrol animals between different time periods were significantlydifferent (p b 0.001 for all), as well as for diabetic animals (p b 0.001for all).
3.5. Neuronal fibers of control rats during maturation
At 2weeks, significantly lower percentagewas found in left ventriclecompared to the right ventricle (p b 0.001), while there was nodifference between other areas of control hearts at that time (p N 0.05for all). At 2months, left ventricle had significantly higher surface areaoccupied by neuronal fibers compared to the interventricular septum(p b 0.010), anterior wall (p b 0.010) and posterior wall (p b 0.010). At6 months, significantly lower values were found in interventricularseptum compared to right ventricle (p b 0.05) while there was nodifference between other areas. Twelve months after the onset ofdiabetes, significant difference was observed between all pairs ofanalyzed heart areas in control rats (all pb0.05).
3.6. Neuronal fibers of diabetic rats during maturation
At 2 weeks, 2 months and 6 months there was no differencebetween the five analyzed areas of the diabetic hearts (p N 0.05for all). However, at 12 months, larger area occupied by neuronalfibers was found in the right ventricle compared to interventricularseptum (p b 0.001), anterior wall (p b 0.001) and posterior wall(p b 0.001) of diabetic hearts.
4. Discussion
In this study, dynamic changes in nerve fibers appearance in heartsof control and diabetic animals were observed in experiments involvinganimals belonging to different age groups. To our best knowledge, this isthe first study aiming to quantify distribution of neuronal profiles indiabetic heart during maturation in such a long time span.
Very early in diabetes mellitus, after two weeks following itsinduction, a significant increase in the area of nerve fiberswas observed,compared to control animals, and this trend was also detected after2 months (Fig. 1d). This increase of area occupied by nerve fiberscould represent a compensatory mechanism by which the peripheralnervous system tries to produce more nerve fibers in order tocompensate lack of nerve functionality, implying that even in the earlystages of DM heart is suffering from impaired heart innervations. Byobserving all four time periods that were studied, a wave-like patternwas observed, as the quantity of all fibers significantly increased after2months, decreased after 6months and again increased at 12months(Fig. 1d). Decrease in neuronal fibers that was found after 6 monthscould imply that the compensating mechanism has been consumed.Another increase of innervation density appears again at 12months ofDM (Fig. 1d), and could represent a final stage of recovery of previousdamage, indicating that changes in heart areas covered with nervesare probably not the only important factor in diabetic heart neuropathybut also changes in autonomic innervation. However, in this study wemeasured all nerve fibers, not the specific subsets.
Namely, heart tissue is extensively innervated by autonomic nervesand alterations in cardiac innervation are evident in various conditions,especially disturbances in sympathetic innervations that may trigger
Fig. 2. PGP 9.5 positive cells were seen as red staining of axonswithin different areas of heart ventricles andmostly arranged around blood vessels. Nerve fiber at 2weeks in control hearts(a) and diabetic hearts (b); at 2months in control hearts (c) and diabetic hearts (d); at 6months in control hearts (e) and diabetic hearts (f); and at 12months in control hearts (g) anddiabetic hearts (h). Scale bar 100 μm, applies to all. Legend: bv=blood vessels.
1476 M. Bakovic et al. / Experimental Gerontology 48 (2013) 1473–1478
lethal arrhythmia via modulation of ion channels in cardiac myocytes(Chen et al., 2007).
Studies in patients who have diabetic autonomic neuropathy haveconsistently demonstrated presence of heterogeneous left ventriclesympathetic denervation, with preservation of areas of innervations inthe proximal myocardium (Stevens et al., 1998). It has recently beendemonstrated that the STZ-induced diabetes in a rat model can beused as a model for human cardiac sympathetic denervation that iscomplicating diabetes. Namely, distal left ventricular denervationobserved in STZ-diabetic rats was associated with a proximal-to-distal
gradient of myocardial depletion of nerve growth factor (Schmid et al.,1999).
In this study we did not find a consistent pattern of regionaldifferences between control and diabetic rats at each of the examinedtime periods. As shown in Fig. 3, significant differences between controland diabetic rats were found at three time points, and in all of them itwas for different heart areas—nerve fiber quantity was higher indiabetic animals in left ventricle at 2 weeks, in the anterior wall at2 months and in the right ventricle at 12 months. However, we didfind significant differences between control and diabetic rats in total
Fig. 3. Quantity of nerve fibers in five areas of the heart of control and diabetic rats at 2weeks (a), 2months (b), 6months (c) and 12months (d). Asterisk denotes significant difference:*pb0.05, **pb0.01, ***pb0.001. Legend: 1=left ventricle, 2=interventricular septum, 3=right ventricle, A=anteriorwall, P=posteriorwall, DM=diabetesmellitus. N=6 animals pergroup. Data presented as M±SD.
1477M. Bakovic et al. / Experimental Gerontology 48 (2013) 1473–1478
surface area covered with neuronal fibers at 2 weeks and 2 months,when data from all analyzed areas of the heart were pooled. Comparedto findings in diabetic patients (Stevens et al., 1998), indicating leftventricle denervation, our findings showed that long-term diabetes, asobserved in periods 2–12months, is characterizedwith less nerve fibersin the left ventricle of diabetic, compared to control rats, although thisdifference was not statistically significant. This trend was not seen inthe right ventricle, where the nerve fiber quantity was consistentlyhigher in all studied periods, in diabetic rats compared to controls(Fig. 3). Therefore, our findings indicate that rat studies are consistentwith findings on human participants.
The increase of nerve fiber quantity in the right ventricle couldbe a consequence of left ventricle denervation. While left ventriculardysfunction has been recognized to be a common complication ofdiabetes mellitus, information about right ventricular performancein patients with diabetes is incomplete and conflicting (Kosmalaet al., 2004, 2007; Mittal, 2007; Movahed and Milne, 2007). Ourfindings indicate that diabetes affects the left and right ventriclesin a different temporal and quantitative sequence.
We analyzed differences in our variables over time to confirmwhether significant differences in cardiac innervation occur duringaging, and not onlywhether there are differences in control and diabeticrats in 4 analyzed time points. Studying changes of heart innervationduring aging allows us insights into heart physiology of healthy ratsand diabetes pathophysiology in diabetic rats.
It has been previously reported that general observation of theyoung andmaturemyocardium reveals that the condition of the cardiac
musculature of the young and mature animals differs markedly, andchanges include limited muscle degeneration, increases in the amountof connective tissue and muscle mass of the left ventricle, thinning ofthe right ventricular wall and occurrence of the large spaces withinthe myocardium (Hewa-Geeganage, 2003). It has also been shownthat there are significantly more terminals in the ventricular myocar-dium of the young animals compared to mature ventricles. In thisstudy young animals were rats aged 3–5weeks, and mature rats wereaged 6 months (Hewa-Geeganage, 2003). In our study, animals wereincluded in the experiments at the age of about 8 weeks, and theirneuronal terminals were analyzed at the age of 10 weeks, 4 months,8months and 14months.
We observed depletion of nerve fibers at 6 months, compared toprevious measurements at 2 weeks and 2 months, however anotherincrease was observed at 12 months, when animals were aged14months. Therefore, our study, with four time points of measurementup to 12months gives a new insight into dynamic changes of healthyand diabetic hearts during aging. The conclusion of our data is thatnerve depletion is not uniform and not straightforward, but that itoccurs in cycles, and that there is a window of opportunity to preventdecline of dennervation.
Schmid et al. have reported that 6months of STZ-induced diabetes ina rat already results in heterogeneous cardiac sympathetic denervation,with maximal denervation occurring distally. In their study, myocardialsympathetic denervation was analyzed scintigraphically using thesympathetic neurotransmitter analog C-11 hydroxyephedrine andcompared with regional changes in myocardial nerve growth factor
1478 M. Bakovic et al. / Experimental Gerontology 48 (2013) 1473–1478
(NGF) protein and norepinephrine content after 6 and 9 months incontrol and STZ-induced diabetic rats (Schmid et al., 1999). Theyused Wistar rats of similar age at the beginning of the experimentas in our study. The dose of STZ that was used was similar as inour study (50 mg/kg vs. 55 mg/kg). Although initial changes wereobserved after 6 months, after 9 months this change was dramaticin both proximal and distal segment of the heart (Schmid et al.,1999). In this study, decrease was seen after 6 months, but also afurther increase in density at 12months.
The effect of age of diabetes induction and duration of STZ diabeteson sympathetic noradrenergic innervations of the rat heart was alsopreviously examined. The results showed that norepinephrine levelswere increased in the hearts of diabetic rats, compared to control rats1 month after diabetes induction, regardless when the diabeteswas induced—at the age of 1, 2, 3 or 4 months. Ventricular levels ofnorepinephrine remained elevated after 2 months of diabetes, with asubsequent decrease after 4 months of diabetes (Felten et al., 1982).The conclusion of this study was that early phases of STZ-induceddiabetes are characterized with intact noradrenergic nerves, and thelater fall of norepinephrine may signal actual neuronal damage, whichsuggests that early intervention may be necessary to protect thesenerves from degeneration (Felten et al., 1982). The results of our studyare consistent with this conclusion.
In our study, we have observed that first 2 months after diabetesinduction there is mostly no variation in the density of nerve terminalsin different areas of the heart between diabetic and control animals,although significantly higher values were observed in diabetic animalscompared to controls when all analyzed heart areas were combined,but only after 2 weeks and 2 months. This could indicate that earlystages of diabetes present a window of opportunity for the preventionof neuronal damage. This issue is potentially important in the view ofhigh incidence of death from myocardial infarction in diabetics withautonomic neuropathy (Felten et al., 1982).
Autonomic neuropathy in diabetic subjects is gradual in its onset,withsigns often hidden for many years by reflex compensatory mechanisms.The ability to detect early complications of diabetes is important, becausesuch evidencemaymodify subsequentmanagement (Malpas andMaling,1990). Our results could reflect those compensatory mechanisms.
Some measures of autonomic function change with age. In thestudy about relationship between aging and autonomic nervoussystem function, evidence was found of an age-related increase ofcardiovascular sympathetic nervous system activity, in addition tothe reduction of cardiac parasympathetic nervous system activity.These findings may indicate that there is sympathetic and parasym-pathetic nervous system compensation of cardiovascular function inresponse to an age-related decrease in baroreceptor sensitivity(Pfeifer et al., 1983).
Animal models have been used extensively in the study of DM.Nowadays, most animal models of diabetes are induced in rodents,using substances such as streptozotocin (STZ) (Rees and Alcolado,2005). Previous studies have shown that aged human heart is similarin many ways to that of rodents, and rats in particular (Lakatta, 2003;Lakatta and Sollott, 2002). Therefore, findings of this study could berelevant for humanpathophysiology studies of diabetes-induced changes.
It is likely that the sole quantity of neuronal fibers is not the onlycontributing factor to the cardiac neuropathy in diabetes, but also thetype of fibers. Therefore, future long-term studies of cardiac innervationsshould explore different neuronal subtypes as well. In the future studies,insulin-treated group should be included aswell, to test whether changesobserved in diabetic rats can be reversed.
In conclusion, cardiac innervation undergoes dynamic changes both incontrol and in diabetic rats, with a time-dependent significant increase inneuronal fibers in diabetes. This novel information may contribute to our
understanding of pathophysiological changes associated with diabeticcardiac neuropathy.
Conflict of interest
The authors have no conflicts of interests.
Acknowledgments
The study was funded by the Croatian Foundation for Science (HRZZ)grant no. 02.05./28 awarded to Livia Puljak and the Ministry of Science,Education and Sports, Republic of Croatia grant no. 216-2160528-0067awarded to Ivica Grkovic.
References
Chen, L.S., Zhou, S., Fishbein, M.C., Chen, P.S., 2007. New perspectives on the role ofautonomic nervous system in the genesis of arrhythmias. J. Cardiovasc. Electrophysiol.18, 123–127.
Faerman, I., Faccio, E., Milei, J., Nunez, R., Jadzinsky, M., Fox, D., Rapaport, M., 1977.Autonomic neuropathy and painless myocardial infarction in diabetic patients.Histologic evidence of their relationship. Diabetes 26, 1147–1158.
Felten, S.Y., Peterson, R.G., Shea, P.A., Besch Jr., H.R., Felten, D.L., 1982. Effects ofstreptozotocin diabetes on the noradrenergic innervation of the rat heart: alongitudinal histofluorescence and neurochemical study. Brain Res. Bull. 8, 593–607.
Gooch, C., Podwall, D., 2004. The diabetic neuropathies. Neurologist 10, 311–322.Hewa-Geeganage, S., 2003. Innervation of the rat ventricular myocardium: patterns of
sympathetic, parasympathetic and sensory supply related to ageing. Department ofAnatomy and Cell Biology. The University of Melbourne, Parkvile, Victoria, Australia.
Ieda, M., Fukuda, K., 2009. New aspects for the treatment of cardiac diseases based on thediversity of functional controls on cardiac muscles: the regulatory mechanisms ofcardiac innervation and their critical roles in cardiac performance. J. Pharmacol. Sci.109, 348–353.
Kent, C., Clarke, P.J., 1991. The immunolocalisation of the neuroendocrine specific proteinPGP9.5 during neurogenesis in the rat. Brain Res. Dev. Brain Res. 58, 147–150.
Kosmala, W., Colonna, P., Przewlocka-Kosmala, M., Mazurek, W., 2004. Right ventriculardysfunction in asymptomatic diabetic patients. Diabetes Care 27, 2736–2738.
Kosmala, W., Przewlocka-Kosmala, M., Mazurek, W., 2007. Subclinical right ventriculardysfunction in diabetes mellitus—an ultrasonic strain/strain rate study. Diabet. Med.24, 656–663.
Lakatta, E.G., 2003. Arterial and cardiac aging: major shareholders in cardiovasculardisease enterprises: part III: cellular and molecular clues to heart and arterial aging.Circulation 107, 490–497.
Lauweryns, J.M., Van Ranst, L., 1988. Protein gene product 9.5 expression in the lungs ofhumans and other mammals. Immunocytochemical detection in neuroepithelialbodies, neuroendocrine cells and nerves. Neurosci. Lett. 85, 311–316.
Malpas, S.C., Maling, T.J., 1990. Heart-rate variability and cardiac autonomic function indiabetes. Diabetes 39, 1177–1181.
Mittal, S.R., 2007. Right ventricular functions in patients with type 2 diabetes below50 years. J. Assoc. Physicians India 55, 599–600.
Movahed, M.R., Milne, N., 2007. Presence of biventricular dysfunction in patients withtype II diabetes mellitus. Congest. Heart Fail. 13, 78–80.
Pfeifer, M.A., Weinberg, C.R., Cook, D., Best, J.D., Reenan, A., Halter, J.B., 1983. Differentialchanges of autonomic nervous system function with age in man. Am. J. Med. 75,249–258.
Ramieri, G., Anselmetti, G.C., Baracchi, F., Panzica, G.C., Viglietti-Panzica, C., Modica, R., Polak,J.M., 1990. The innervation of human teeth and gingival epithelium as revealed bymeans of an antiserum for protein gene product 9.5 (PGP 9.5). Am. J. Anat. 189,146–154.
Schofield, J.N., Day, I.N., Thompson, R.J., Edwards, Y.H., 1995. PGP9.5, a ubiquitinC-terminal hydrolase; pattern of mRNA and protein expression during neuraldevelopment in the mouse. Brain Res. Dev. Brain Res. 85, 229–238.
Stevens, M.J., Dayanikli, F., Raffel, D.M., Allman, K.C., Sandford, T., Feldman, E.L., Wieland,D.M., Corbett, J., Schwaiger, M., 1998. Scintigraphic assessment of regionalized defectsin myocardial sympathetic innervation and blood flow regulation in diabetic patientswith autonomic neuropathy. J. Am. Coll. Cardiol. 31, 1575–1584.
Terenghi, G., Sundaresan,M., Moscoso, G., Polak, J.M., 1993. Neuropeptides and a neuronalmarker in cutaneous innervation during human foetal development. J. Comp. Neurol.328, 595–603.
Wirnsberger, G.H., Becker, H., Ziervogel, K., Hofler, H., 1992. Diagnostic immunohis-tochemistry of neuroblastic tumors. Am. J. Surg. Pathol. 16, 49–57.
