B R A I N R E S E A R C H 1 2 1 7 ( 2 0 0 8 ) 1 8 5 – 1 9 4

ava i l ab l e a t www.sc i enced i rec t . com

www.e l sev i e r. com/ l oca te /b ra in res

Research Report

Dystrophic serotonergic axons in neurodegenerative diseases

Efrain C. Azmitiaa,b,c,⁎, Ralph Nixonb,c,d

aDepartment of Biology and Center for Neural Science, 100 Washington Square East, New York, New York 10003, USAbDepartment of Psychiatry, New York School of Medicine, New York, New York 10016, USAcDepartment of Cell Biology, New York School of Medicine, New York, New York 10016, USAdNathan Kline Institute, Orangeburg, New York 10962, USA

A R T I C L E I N F O

⁎ Corresponding author. 10-09 Silver BuildingE-mail address: eca1@nyu.edu (E.C. Azmi

0006-8993/$ – see front matter © 2008 Elsevidoi:10.1016/j.brainres.2008.03.060

A B S T R A C T

Article history:Accepted 14 March 2008Available online 7 April 2008

Neurodegenerative diseases such as Parkinson's disease (PD), frontal lobe dementia (FLD) anddiffuse Lewy-body dementia (DLBD) have diverse neuropathologic features. Here we reportthat serotonin fibers are dystrophic in the brains of individuals with these three diseases. Inneuropathologically normal (control) brains (n=3), serotonin axons immunoreactive (IR) withantibodies against the serotonin transporter (5-HTT) protein were widely distributed in cortex(entorhinal and dorsolateral prefrontal), hippocampus and rostral brainstem. 5-HTT-IR fibers-of-passage appeared thick, smooth, and unbranched in medial forebrain bundle, mediallemniscus and cortex white matter. The terminal branches were fine, highly branched andvaricose in substantia nigra, hippocampus and cortical gray matter. In the diseased brains,however, 5-HTT-IR fibers in the forebrain were reduced in number and were frequentlybulbous, splayed, tightly clustered and enlarged. Morphometric analysis revealed significantdifferences in the size distribution of the 5-HTT-IR profiles in dorsolateral prefrontal areabetween neurodegenerative diseases and controls. Our observations provide directmorphologic evidence for degeneration of human serotonergic axons in the brains ofpatients with neurodegenerative diseases despite the limited size (n=3 slices for each region(3) from each brain (4), total sliceswas n=36) and the lack of extensive clinical characterizationof the analyzed cohort. This is the first report of dystrophic 5-HTT-IR axons in postmortemhuman tissue.

© 2008 Elsevier B.V. All rights reserved.

Keywords:DegenerationNeuropathologyParkinson'sLewy-bodyFrontal lobe dementiaAgingPostmortemHippocampusPrefrontalEntorhinalSubstantia nigraHuman

1. Introduction

The anatomy and plasticity of 5-HT projecting axons inexperimental animals has been extensively studied. The fine5-HT axons were first visualized in the rat brain withhistochemical fluorescence (Fuxe, 1965). Immunocytochem-ical analyses with antibodies raised against 5-HT laterrevealed the full global projections of the raphe neurons,showing also that 5-HT fibers-of-passage are thick and un-branched while the terminal fibers are fine, highly branched

, 100 Washington Squaretia).

er B.V. All rights reserved

and varicose (Steinbusch, 1981; Azmitia and Gannon, 1983).Dense innervation is apparent throughout the neuroaxis andit has been suggested that every cell in the rat cortex is near a5-HT containing axon (Molliver, 1987).

It has been difficult to study the morphology of the 5-HTaxonal system in the adult human brain. Postmortem auto-lysis and the usual methods of postmortem brain fixationresult in the loss of 5-HT from axonal storage sites, whichmakes histochemical fluorescence and 5-HT immunocyto-chemistry ineffective. Antibodies raised against monoamine

East, New York, New York 10003-6688, USA. Fax: +1 212 995 4015.

.

mailto:eca1@nyu.eduhttp://dx.doi.org/10.1016/j.brainres.2008.03.060

Fig. 1 – 5-HTT-IR axons are seen in brains from control individuals. A. In the midbrain/substantia nigra sections heavy fiberlabeling is seen in the medial lemniscal with the fibers throughout the entire pathway. These fibers-of-passage have smallvaricosities and often form tight bundles (arrows). Scale bar is 200μm.B. In the hippocampus and entorhinal areas the 5-HTT-IRaxons are forming terminal boutons. In these sections larger boutons can be seen (arrow). Scale bar is 50μm. C. In the prefrontalcortex, the 5-HTT-IR fibers form clusters around cell bodies in pyramidal layers (arrows). Scale bar is 200 μm.

186 B R A I N R E S E A R C H 1 2 1 7 ( 2 0 0 8 ) 1 8 5 – 1 9 4

hydroxylase enzymes label 5-HT cell bodies and processes inthe human brainstem, but fail to reveal the distal terminalfibers (Craven et al., 2005; Underwood et al., 1999; Baker et al.,1991; Haan et al., 1987).

A better approach may be to focus on the 5-HT transporterprotein. The 5-HT transporter (5-HTT) protein is specific to 5-HTaxons in the adult brain, although occasional astrocytic stainingcan be observed. The axonal sitewas visualized in experimentalanimals by autoradiography after uptake 3H-5-HT by thetransporter protein in situ (Descarries et al., 1975) or in vitro infresh brain slices (Azmitia, 1981). Antibodies to 5-HTT allowedthemorphology of 5-HT terminal axons to be studied immuno-cytochemically in brains of non-human animals (Qian et al.,1995; Zhou et al., 1996). In humans, radioactive ligands specificfor the5-HT transporter permit thegross regional distributionof5-HT axons to be revealed in vivo by PET (Abi-Dargham et al.,1966; Szabo et al., 1996) or in postmortem tissue by film auto-radiography (Varnas et al., 2004; Chinaglia et al., 1993). 5-HTTimmunoreactive (IR) axons of normalmorphology were seen inthe postmortem brainstem and prefrontal cortex of individualswithout known neurological disorders (Qian et al., 1995; Austinet al., 2002). The 5-HTT-IR axons were described as fine, highlybranchedandvaricose, similar inappearance to thosedescribedin non-human animals.

In animal studies, serotonin axons can be damaged under avariety of conditions. Immunocytochemical analysis or immu-nofluoresence labeling with antibodies raised against 5-HTdetects dystrophic immunoreactive fibers in rats after neuro-toxic injections (Wiklund and Bjorklund, 1980; Frankfurt andAzmitia, 1983); ingestion of designer-drugs of abuse (Wilson

Fig. 2 – Dystrophic 5-HTT-IR axons are seen, but infrequently, inentorhinal cortex is heavily innervated with 5-HTT-IR axons. Amdense clusters, which can condense into a larger degenerating prolarge varicosities can be found (arrow). C. In the dendritic regionScale bar is 50 μm in all panels.

and Molliver, 1994); and in aged animals (van Luijtelaar et al.,1989). Immunocytochemistry with antibodies to 5-HTT alsoshows dystrophic 5-HTT immunoreactive axons after neuro-toxic injections (Zhou et al., 1996). These dystrophic axonsare enlarged, bulbous or splayed after acute damage to 5-HTfibers, suggesting an active degeneration process. In aginganimals, 5-HT fibersmay appear abnormally fine, varicose andtightly clustered suggesting atrophy and withdrawal fromtheir forebrain targets (van Luijtelaar et al., 1980).

Despite a preponderance of pharmacological, neurochem-ical, andmolecular evidence that the 5-HT system is disruptedin many clinical disorders, dystrophic serotonergic axons inthe human brain have never been described. There is indirectevidence that 5-HT axons are damaged in neurological dis-orders such as FLD (Sparks andMarkesbery, 1991; Menza et al.,1999); PD (Halliday et al., 1990; Marksteiner et al., 2003); Al-zheimer's disease and ischemic heart disease (Stout et al,2003); and DLBD (Ballard et al., 2002). For example in DLBD,Lewy bodies occur in the dorsal raphe nucleus, and serotoninlevels are markedly reduced in the striatum, neocortex andfrontal cortex (Langlais et al., 1993; Ohara et al., 1998; Perryet al., 1993).

In this study, we obtained well-characterized postmor-tem neurodegenerative brains from patients with PD, FLDand DLBD, as well as from neuropathologically normal (con-trol) individuals. In neuropathologically normal brains, 5-HTTimmunocytochemistry showed typical 5-HT axonal morphol-ogy in abundant axons, including both projecting fibersand terminating regions, and only rare dystrophic fibers. Bycontrast, dystrophic 5-HTT-IR fibers were frequent and wide-

the brain of a control 79-year-old-male. A. Layer III ofong the normal fibers, abnormal fibers can be seen formingfile (arrows). B. In the polymorphic area of the DG abnormallyof CA3 degenerating profiles are seen among normal fibers.

Fig. 3 – 5-HTT-IR axons in brains from patients diagnosed with diffuse Lewy-body dementia. A. This picture shows normalappearing 5-HTT-IR axonal bundles in the dorsal aspect of the section just above the inferior colliculus (IC). There is evidence ofdystrophic 5-HTT-IR axons in the fiber bundles (arrows). Scale bar is 200μm. B. Normal and abnormal (arrows) terminals in thesubstantia nigra. Two pigmented nigra neurons can be seen. Scale bar is 50 μm. C. Splayed 5-HTT-IR terminals seen in CA3dendritic region. Scale bar is 50 μm. D. Tight clusters of 5-HTT-IR axons in layer III of entorhinal cortex. The number of labeledfibers is reduced in this region. Scale bar is 50 μm. E. Tight cluster of 5-HTT-IR fibers in layer III of prefrontal cortex. Scale bar is50 μm. F. In deep layers of cortex, aggregates of fibers are found among dystrophic 5-HTT-IR axons with enlarged varicosities.Scale bar is 200 μm.

187B R A I N R E S E A R C H 1 2 1 7 ( 2 0 0 8 ) 1 8 5 – 1 9 4

spread in all brains from patients with degenerative disease.Using amorphometric software system, we identified significantalterations in thenumber and sizes of 5-HTT-IR axonal structuresin the prefrontal terminal regions of all three neurodegenerativediseases. To our knowledge, this study provides the first directevidence that 5-HT axons in the human brain are vulnerable todegeneration in neurodegenerative states.

2. Results

2.1. 5-HTT patterns in neuropathologically normal brains

5-HTT-IR fibers were seen in all examined regions of neuro-pathologically normal brains. The 5-HTT-IR axons formedlarge, dense bundles in the rostral midbrain especially withinthemedial lemniscus (arrows, Fig. 1A). The 5-HTT-IR fibers-of-passage were thick, unbranched and either long and straight(possibly myelinated fibers) or wavy in appearance (Fig. 1A).

2.1.1. Ascending projectionsWe observed fine and varicose, or coarse and non-varicose 5-HTT-IR axons in all regions examined. The non-varicose fiberswere particularly abundant in brainstem areas associatedwith major ascending pathways as well as in the forebrain incorpus callosum, fornix, and perforant path fibers. The 5-HTT-IR fibers in these forebrain pathways ascended into the grayterminal areas of the cortex where the fibers became thin-ner and more branched. In addition, a dense fiber plexus oftangential projecting fibers was found in layer I of entorhinal(Fig. 1B) and prefrontal cortical regions (Fig. 1C). These fiberswere in close proximity to the pia layer and could be seenextending into layers II and III.

2.1.2. Terminal distributionIn the terminal areas within the midbrain (region of the Sub-stantia Nigra), the 5-HTT antibody labeling resolved thin, va-ricose, and highly branched 5-HTT-IR axonal fibers. Distinct 5-HTT-IR boutons were visualized at the terminals of thesefibers. In entorhinal and prefrontal cortices, the distribution of5-HTT-IR fibers extended throughout all cortical layers, withextensive branching seen in deeper layers near large pyra-midal neurons. The 5-HTT-IR axons were fine, highly bran-ched and formed irregularly spaced varicosities (Figs. 1B-C).The terminals were frequently seen closely surrounding largepyramidal neurons in what can be described as pericellularplexuses.

Occasional dystrophic 5-HTT-IR axons (Fig. 2) were seen intheparahippocampal terminal regions, especially obvious in theoldest brain examined (B3573; male, 79 years of age, postmor-tem interval of 15.3 h). The abnormal 5-HTT axons – tightgrouping of small varicosities (clustering) (Fig. 2A, arrows) andenlarged varicosities (Figs. 2B–C, arrows) – were scatteredamong typical fibers, especially in the hilus of the dentategyrus (Fig. 2B) and the deep entorhinal layers (layers V–VI)(Fig. 2C) and in the CA layers of the hippocampus (not shown).However, therewasnoevidence in controlmaterial of advancedneurodegeneration (splayed endings or dark aggregates ofstained material) or a marked reduction in fiber density.

2.2. 5-HTT patterns in the brain in neurodegenerativediseases

5-HTT-IR axons were found in all regions of the brains fromindividuals with any of the three neurodegenerative diseasesexamined. In the brainstem from these cases, 5-HTIR axonsappeared normal, although occasional abnormal fibers were

Fig. 4 – 5-HTT-IR axons in brains frompatients diagnosedwith Parkinson's disease. A. 5-HTT-IR axons in the red nucleus showsome abnormal fibers (arrows) among many normal fibers. Scale bar is 200 μm. B. In midline areas 5-HTT-IR fibers withenlarged varicosities are found among normal fibers. Scale bar is 50μm. C. Dystrophic 5-HTT-IR axonswith enlarged varicosityin the dendritic regions of CA1. Scale bar is 50 μm. D. Dystrophic 5-HTT-IR axons are splayed and degenerating in the upperlayers of the entorhinal cortex. Scale bar is 50 μm. E. Fine degenerating 5-HTT-IR axons in the upper layer of prefrontal cortex.Scale bar is 50 μm. F. Fine, dense clusters of 5-HTT-IR axons in the deeper layers of prefrontal cortex. Scale bar is 50 μm.

188 B R A I N R E S E A R C H 1 2 1 7 ( 2 0 0 8 ) 1 8 5 – 1 9 4

seen. Similarly, in the corpus callosum of the forebrain, long 5-HTT-IR axons could be followed with no apparent evidence ofpathology, although in the gray matter from this brain regionand other cortical gray areas the number of 5-HTT-IR axonsappeared markedly reduced. Significant numbers of 5-HTT-IRaxons were dystrophic in shape in prefrontal and temporalcortical areas. The hippocampus and entorhinal and prefron-tal cortices showed four main types of abnormal profiles: (1)enlarged, twisted and swollen varicosities, sometimes appear-ing as ballooned profiles; (2) fine fibers forming tight clus-

Fig. 5 – 5-HTT-IR axons in brains from patients diagnosed with Frois seen in the substantia nigra nucleus. No dystrophic 5-HTT-IR fibethe medial lemniscal tract. No dystrophic 5-HTT-IR fibers seen. Scaldentategyrus.Scalebar is50μm.D.Several splayedanddegeneratinNote reduced appearance of normal 5-HTT-IR fibers. Scale bar is 50layers II–III of the prefrontal cortex. Note normal tangential fibers in lthe deeper layers of prefrontal cortex. 5-HTT-IR axonal innervation

ters; (3) isolated splayed fibers with an irregular shape; and(4) densely labeled aggregates and degenerating profiles.

2.3. Diffuse Lewy-body dementia

In brainstem sections from cases of DLBD, the 5-HTT-IR fibersappeared less dense than normal and occasionally the pro-cesses were abnormal in appearance and darkened (arrows,Figs. 3A–B). In both the entorhinal cortex (Figs. 3B–C) andprefrontal (Figs. 3E–F), 5-HTT-IR axons in layer I, close to the pia,

ntal Lobe Dementia. A. Dense innervation by 5-HTT-IR fibersrs seen. Scale bar is 200 μm. B. Heavy labeling of the fibers ine bar is 200 μm. C. Enlarged varicosities seen in hilus area ofg5-HTT-IR terminals seen in theupper layerof entorhinal cortex.μm. E. Dense clustering and aggregation of 5-HTT-IR axons inayer I. Scale bar is 50μm. F. Enlarged 5-HTT-IR axons are seen inappears reduced from normal. Scale bar is 50 μm.

Fig. 6 – This figure shows the results from morphometricanalysis of the labeled objects found in prefrontal cortex ofcontrol and diseased brains using threshold setting ofimmunoreactive density. A. The total number of particles(varicosities, fibers and degenerating axons) was counted inan area of 0.5mm2 in layers III–V of the cortex. Each bar is themean of three brains from each group (see Experimentalprocedures for details). The number of particles selected waslower for all diseased-groups compared to normal prefrontalcortex, and was significant for the DLBD and PD.B. A histogram of all particle areas was made form theparticles selected for part A. The percentage of the totalnumber of particles which had the smallest area (

Table 1 – The diagnosis, demographic, postmortem interval (PMI) history of depression and antidepressant treatment

Brain # Diagnosis Gender Age PMI Depression Antidepressant treatment

B2374 None M 72 8.5 – –B2469 None F 54 3.5 – –B3573 None M 79 15.3 – –B4426 DLBD M 75 23.5 + –B4866 DLBD M 67 14.85 + –B5085 DLBD M 68 5.5 + +B4927 FLD F 68 14.33 + +B5007 FLD F 72 6.58 N/A N/AB5035 FLD F 75 20 + +B4934 PD M 68 17.16 N/A N/AB5207 PD M 64 18.58 N/A N/AB5227 PD M 77 26 + +

190 B R A I N R E S E A R C H 1 2 1 7 ( 2 0 0 8 ) 1 8 5 – 1 9 4

axons in postmortem brain sections from individuals withouta diagnosed neurological or psychiatric diseasewere similar tothose described previously in humans and animal studies. Wedetected extensive neuropathology of 5-HTT-IR axons, how-ever, in brains of patientswith Parkinson's disease, frontal lobedementia or diffuse Lewy-body dementia. This is the firstreport of 5-HT axonal pathology in humans and shows that the5-HT system in the human brain is vulnerable to degenerationin various neurodegenerative disorders, as shown previouslyin animal models of disease.

3.1. Serotonin system and neurodegeneration

Although numerous pathologies in late-age onset neurodegen-erative diseases have been identified with markers of specificpathologies (eg.β-amyloid, ubiquitin, hyperphosphorylated-tauor α-synuclein reactive), there are relatively few examples of atransmitter-specific neuronal degeneration in neurodegenera-tive diseases (see reports on acetylcholine: Bossy-Wetzel et al.(2004), Perry et al. (1993) and norepinephrine: Haglund et al.(2006)). Even in Parkinson's diseases, we could not find a des-cription of dystrophic dopaminergic fibers (see Dickson et al.(1994)) although the normal distribution of projections fromdopaminergic neurons to the human caudate has been des-cribed (Kung et al., 1998).

Wenowreport that serotonin fibers, immunocytochemicallylabeled with antibody against the 5-HTT, show extensive andwide-spread pathology in the brains of patients with PD, FLD orDLBD. In animal studies, serotonin fibers degenerate whenexposed to a variety of environmental and neurotoxic factors(e.g. van Luijtelaar et al. (1989)—aging; O'Hearn et al. (1988)—MDMA and related drugs; Zhou et al. (1994)—alcohol; Liu andNakamura (2006), Aucoin et al. (2005)—amyloid; Baumgartenand Bjorklund (1976), Frankfurt and Azmitia (1984)—5,7-DHT).Our findings demonstrate that human serotonin axons are alsovulnerable to degeneration in pathological states.

3.2. Comparison of axonal pathology

Depletion of normal fibers and appearance of degeneratingprofileswere evident in all three groups of brains frompatientswith degenerative diseases. There was no evidence that gen-der or postmortem interval contributed to these observations.Recent unpublished work using 5-HTT immunostaining (n=22

brains from neurologically typical normal subjects, range 32–85 years, average age 57.2 years and n=3 female) showed noevidence of the frequent severe 5-HTT-specific axonal pathol-ogy reported here in the brains from patients with neurode-generative diseases. The 5-HTT-IR dystrophic axonal profilesin this report are similar in appearance to those describedafter systemic administration of neurotoxin (Baumgarten andBjorklund, 1976) and designer-drug (Molliver and Molliver,1990) induced degeneration of 5-HT fibers in rat brain. Thedystrophic axons seen in the deep layers of FLD prefrontalcortex exhibited increased caliber, reduced branching, andswollen varicosities and resembled those seen after 5,7-DHTneurotoxin intracerebral injections into the 5-HT fibers-of-passage in MFB (Frankfurt and Azmitia, 1984) or cingulumbundle (Zhou andAzmitia, 1986). The 5-HTT fibers inDLBDandPD brains show degenerating fibers characterized by varicoseswelling and clustering of fine terminals. This pattern ofdegeneration is seen inaged rats (vanLuijtelaar et al., 1989) andS100B knockout animals (unpublished observation). Thedystrophic axonal pattern suggests that these terminal fibersmaybe retracting froma region inwhich levels of trophic factorare reduced.

3.3. Possible relevance of 5-HTT axon degeneration tosymptoms of neurodegenerative diseases

Dystrophic degeneration of serotonergic axonsmay contributeto the development of many of the symptoms of neurodegen-erative diseases such as mood, motor, sensory, autonomic,cognitive, and sleepdisorders (see SandykandFisher (1988)). InPD patients, selective 5-HT reuptake inhibitors (SSRIs) improvebradykinesia treatedwith L-dopa (Rampello et al., 2002). In ourstudy, six of nine patientswith neurodegenerative disease hadhistories of late-life depressive symptoms. Depression isbelieved to be a risk factor for the onset of Alzheimer's disease(Green et al., 2003; Chen et al., 1999; Kokmen et al., 1991; Kraland Emery, 1989). Affective disorders are often comorbid withneurodegenerative disorders that are associated with demen-tia (Allen and Burns, 1995; Schreinzer et al., 2005; Lauterbachet al., 2004). Affective symptoms are frequently part of theinitial presentation in neurodegenerative diseases (Ishiharaand Brayne, 2006; Kessing and Andersen, 2004), and, in somecases, may be the dominant presenting symptoms (Ballard etal., 2002). A history of major depression, without specification

Fig. 7 – Dark field montage.

191B R A I N R E S E A R C H 1 2 1 7 ( 2 0 0 8 ) 1 8 5 – 1 9 4

of episode-related cognitive impairment, appears to be arisk factor for subsequent onset of dementia (Kessing andAndersen, 2004). This is supported by early work with twins,which found that depression and psychiatric illness were riskfactors for developing dementia (Wetherell et al., 1999). Moodmay be improved in depressed cognitively impaired older pa-tients by treatment with antidepressants including selectiveserotonin reuptake inhibitors (SSRI) even in the absence ofimprovement in cognitive or motor domains (Schrag, 2006;Ishihara and Brayne, 2006; Nebes et al., 2003). The actions ofserotonin in the development and treatment of impairmentsseen in neurodegenerative diseases require further investiga-tion (Gruden et al., 2007; Schmitt et al., 2006; Truchot et al.,2007).

3.4. Study limitations

Although our findings demonstrated consistent differences inthe extent and character of 5-HT axonal pathology in the nineneurodegenerative disease cases compared to the threecontrol cases analyzed, the small size of the sample precludeda precise matching of the cases for age and postmorteminterval. Recent unpublishedwork using 5-HTT immunostain-ing now includes 22 brains from neurologically typicalnormals, three of which are female, age range is 32–85 andaverage age is 57.2. No evidence of the frequent 5-HTT-specificaxonal pathology reported here in the brains from neurode-generative diseases is seen. The occasional 5-HTT-specificaxonal pathology in the temporal cortical regions of neurolo-gically normal subjects is confirmed (unpublished observa-tion). Moreover, in contrast to the extensive neuropathologicalevaluation, the limited clinical characterization of the subjectsin the study precluded a complete assessment of psychiatrichistory andmedications in all cases. Nevertheless, these initialfindings on 5HT fiber pathology in neurodegenerative diseases

provide a strong rationale for more extensive analyses ofthe relationship between cognitive or affective parameters and5-HT pathology in a larger cohort of clinically well-character-ized subjects.

4. Experimental procedures

4.1. Brains

Brains were obtained from the Harvard Brain Tissue ResourceCenter at McLean Hospital in Belmont, Massachusetts. Weobtained 36 tissue blocks for our study from 12 brains, whichmet diagnostic criteria for the neurodegenerative diseases ofParkinson's disease (PD, n=3), Frontal Lobe Dementia (FLD,n=3), and Lewy-body dementia (DLBD, n=3). The patients allmet pre-mortem and postmortem criteria for the diagnosismade at Harvard Brain Tissue Resource Center. The durationof the illness or specific treatments were not available fromthe records for all cases in this study. A majority of thepatients with a neurodegenerative disorder also had apsychiatric history of depression (6/9), and in three cases,the information was unavailable and cannot be considerednegative for depressive disorder. Neuropathologically normalbrains (n=3) from donors with no medical or psychiatric his-tory of illness were studied for comparison (Table 1).

4.2. Selection of regions for study

We used anterior brainstem blocks, which contained thesubstantia nigra and red nucleus, and the fiber tracts ofthe medial forebrain bundle, dorsal raphe cortical tract andthe medial lemniscus. The temporal block we used containedthehippocampus (subiculum, CA fields anddentate gyrus) andthe entorhinal cortex and associated whitematter. The frontalblock we used contained the dorsolateral prefrontal cortex(Brodmann layer 9) with all six cortical layers and a significantamount of the corpus callosum. We examined all layers ofentorhinal and prefrontal cortices. In the hippocampus, welooked at the hilus of the dentate gyrus and the stratum ra-diatum of CA3 (paying particular note to themossy fibers) andstratum radiatum of CA1. We also examined the fibers-of-theperforant path connecting the entorhinal cortex to CA1 andmolecular layer of the dentate gyrus.

4.3. Orientation

We developed a method to produce a full section dark-fieldmontage (Fig. 7) using a prototype condenser (Leitz Micro-scopic Company, Germany) for viewing with a 1.6× objective.Pictures were taken with a Kodak-6.0 million pixel camera(DCL 760). The montage of the various brain regions requiredbetween 6–25 images to assemble and was used to establishprecise locations within the brain section.

4.4. Immunocytochemistry

Immunocytochemical studies were performed as previouslydescribed (Shiurba et al., 1998) using rabbit polyclonal anti-bodies raised against serotonin transporter (synthetic peptide

192 B R A I N R E S E A R C H 1 2 1 7 ( 2 0 0 8 ) 1 8 5 – 1 9 4

from the carboxy-terminus (aa 602–622: GTLKERIIKSITPETP-TEIPC) of the cloned rat serotonin transporter (AB177L,Calbiochem.). This site-specific antibody has been demon-strated to be highly selective for the 5-HTT proteins inimmunocytochemistry and western analysis (Zhou et al.,1996; Qian et al., 1995). Dilutions of antibody were calculatedto be in the linear range of staining intensity (1/1000–1/10,000ab5-HTT). Brain tissue used for immunocytochemical analyseswas immersion-fixed in cold 10%phosphate-buffered formalin(0.15 mol/L), pH 7.4. Immunocytochemistry was performedby the avidin-biotin complex (ABC) method using Vectastainkits (Vector Labs, Inc., Burlingame, CA). Free-floating vibra-tome 50 μm sections were treated for 30 min with 3%methanolic hydrogen peroxide to block non-specific endogen-ous peroxidase activity and rinsed in 0.05 M Tris-buffered (pH7.4) saline (TBS) containing 0.4% Triton X-100. The sectionswere treated for 30minwith 20% normal rabbit serum (NRS) toreduce non-specific background staining. Sections were thenincubated in TBS with 1% NRS and 0.4% Trition X-100 withappropriate dilutions of primary antisera (AB-1, 1/5000) over-night at room temperature. The tissue was incubated first inbiotinylated-secondary antibody (1:200 dilution) and subse-quently in preformed ABC (90/d/10 ml avidin and 90;d/mlbiotin). The final reactionwas achievedby treating the sectionswith 0.02% hydrogen peroxide and DAB (0.5 mg/ml) in 0.1 MTBS, pH 7.4, for 5 min. Vibratome sections were mounted ongel-coated slides and air-dried following any immunocyto-chemical or histological procedures. All sections were dehy-drated in a series of ethanol to xylene and coverslipped withPermount. Immunocytochemical controls consisted of eitherincubating tissue in non-immune sera or omitting incubationin primary antisera.

4.5. Morphometric methods

Non-stereologic, morphometric measures of particle numberand area were performed on each brain (n=3) for statisticalanalysis. Stereologic methods were not used because ofthe limited number of brain sections available, and the highdensity of 5-HTT-IR axons. The sections were viewed with aLeitz orthoplan microscope with Kohler illumination andphotographed with a Nikon DCL760 digital camera for theanalysis of 5-HTT-IR axonal number and area. Morphometricanalysis was performed with the UTHSCSA Image Tool forwindows (version 3.00) using pictures taken with a 25×PlFluotar Objective with 8× column magnification. At leastthree pictures were randomly taken in the prefrontal corticallayers (II–VI) for each section from all groups. The photographswere untouched except for conversion into Grayscale beforeperforming a threshold selection for illumination densitymeasures of 5-HTT-IR labeling (setting of 150 on a 1–255scale). All 5-HTT-IR labeled structures were automatically se-lected by the computer (10 pixel minimum cut-off) for count-ing, and measures of area. In the morphometric analysis ofparticles, normal varicosities that were assumed from theimages obtained to have an area of

193B R A I N R E S E A R C H 1 2 1 7 ( 2 0 0 8 ) 1 8 5 – 1 9 4

Baumgarten, H.G., Bjorklund, A., 1976. Neurotoxic indoleaminesand monoamine neurons. Annu. Rev. Pharmacol. Toxicol. 16,101–111.

Bossy-Wetzel, E., Schwarzenbacher, R., Lipton, S.A., 2004. Molecularpathways to neurodegeneration. Nat. Med. 10, S2–S9 Suppl.

Chen, P., Ganguli, M., Mulsant, B.H., DeKosky, S.T., 1999.The temporal relationship between depressive symptoms anddementia: a community-based prospective study. Arch. Gen.Psychiatry 56, 261–266.

Chinaglia, G., Landwehrmeyer, B., Probst, A., Palacios, J.M., 1993.Serotoninergic terminal transporters are differentially affectedin Parkinson's disease and progressive supranuclear palsy:an autoradiographic study with [3H]citalopram. Neuroscience54, 691–699.

Craven, R.M., Priddle, T.H., Cooper, S.J., Crow, T.J., Esiri, M.M., 2005.The dorsal raphe nucleus in schizophrenia: a post mortemstudy of 5-hydroxytryptamine neurones. Neuropathol.Appl. Neurobiol. 31, 258–269.

Descarries, L., Beaudet, A., Watkins, K.C., 1975. Serotonin nerveterminals in adult rat neocortex. Brain Res. 100, 563–588.

Dickson, D.W., Schmidt, M.L., Lee, V.M., Zhao, M.L., Yen, S.H.,Trojanowski, J.Q., 1994. Immunoreactivity profile ofhippocampal CA2/3 neurites in diffuse Lewy body disease.Acta Neuropathol. (Berl) 87, 269–276.

Frankfurt, M., Azmitia, E.C., 1983. The effect of intracerebralinjections of 5,7-dihydroxytryptamine and 6-hydroxydopamineon the serotonin-immunoreactive cell bodies and fibers in theadult rat hypothalamus. Brain Res. 261, 91–99.

Frankfurt, M., Azmitia, E.C., 1984. Regeneration of serotonergicfibers in the rat hypothalamus following unilateral5,7-dihydroxytryptamine injection. Brain Res. 298, 273–282.

Fuxe, K., 1965. Evidence for the existence of monoamine neuronsin the central nervous system. IV. The distribution ofmonoamine terminals in the central nervous systemActa Physiol. Stand. 64 (Suppl. 247), 41–85.

Green, R.C., Cupples, L.A., Kurz, A., Auerbach, S., Go, R., Sadovnick,D., Duara, R., Kukull, W.A., Chui, H., Edeki, T., Griffith, P.A.,Friedland, R.P., Bachman, D., Farrer, L., 2003. Depression as arisk factor for Alzheimer disease: the MIRAGE Study.Arch Neurol. 60, 753–759.

Gruden, M.A., Davidova, T.B., Malisauskas, M., Sewell, R.D.,Voskresenskaya, N.I.,Wilhelm, K., Elistratova, E.I., Sherstnev, V.V.,Morozova-Roche, L.A., 2007.Differential neuroimmunemarkers tothe onset of Alzheimer's disease neurodegeneration anddementia: autoantibodies to Abeta((25–35)) oligomers,S100b and neurotransmitters. J. Neuroimmunol. 186, 181–192.

Halliday, G.M., Blumbergs, P.C., Cotton, R.G., Blessing, W.W.,Geffen, L.B., 1990. Loss of brainstem serotonin- and substanceP-containing neurons in Parkinson's disease. Brain Res. 510,104–107.

Haan, E.A., Jennings, I.G., Cuello, A.C., Nakata, H., Fujisawa, H.,Chow, C.W., Kushinsky, R., Brittingham, J., Cotton, R.G., 1987.Identification of serotonergic neurons in human brain by amonoclonal antibody binding to all three aromatic amino acidhydroxylases. Brain Res. 426, 19–27.

Haglund, M., Sjobeck, M., Englund, E., 2006. Locus ceruleusdegeneration is ubiquitous in Alzheimer's disease: possibleimplications for diagnosis and treatment. Neuropathology 26,528–532.

Ishihara, L., Brayne, C., 2006. What is the evidence for a premorbidparkinsonian personality: a systematic review. Mov. Disord. 21,1066–1072.

Kessing, L.V., Andersen, P.K., 2004. Does the risk of developingdementia increase with the number of episodes in patientswith depressive disorder and in patients with bipolar disorder?J Neurol. Neurosurg. Psychiatry 75, 1662–1666.

Kokmen, E., Beard, C.M., Chandra, V., Offord, K.P., Schoenberg, B.S.,Ballard, D.J., 1991. Clinical risk factors for Alzheimer's disease: apopulation-based case-control study. Neurology 41, 1393–1397.

Kral, V.A., Emery, O.B., 1989. Long-term follow-up of depressivepseudodementia of the aged. Can. J. Psychiatry 34, 445–446.

Kung, L., Force, M., Chute, D.J., Roberts, R.C., 1998.Immunocytochemical localization of tyrosine hydroxylase inthe human striatum: a postmortem ultrastructural study.J. Comp. Neurol. 390, 52–62.

Langlais, P.J., Thal, L., Hansen, L., Galasko, D., Alford, M., Masliah,E., 1993. Neurotransmitters in basal ganglia and cortex ofAlzheimer's disease with and without Lewy bodies. Neurology43, 1927–1934.

Lauterbach, E.C., Freeman, A., Vogel, R.L., 2004. Differential DSM-IIIpsychiatric disorder prevalence profiles in dystonia andParkinson's disease. J. Neuropsychiatry Clin. Neurosci. 16,29–36.

Liu, Y., Nakamura, S., 2006. Stress-induced plasticity ofmonoamine axons. Front. Biosci. 11, 1794–1801.

Marksteiner, J., Walch, T., Bodner, T., Gurka, P., Donnemiller, E.,2003. Fluoxetine in Alzheimer's disease with severe obsessivecompulsive symptoms and a low density of serotonintransporter sites. Pharmacopsychiatry 36, 207–209.

Menza, M.A., Palermo, B., DiPaola, R., Sage, J.I., Ricketts, M.H., 1999.Depression and anxiety in Parkinson's disease: possible effectof genetic variation in the serotonin transporter. J. Geriatr.Psychiatry Neurol. 12, 49–52.

Molliver, M.E., 1987. Serotonergic neuronal systems: what theiranatomic organization tells us about function. J. Clin.Psychopharmacol. 7 (6 Suppl), 3S–23S.

Molliver, D.C., Molliver, M.E., 1990. Anatomic evidence for aneurotoxic effect of (+/−)-fenfluramine upon serotonergicprojections in the rat. Brain Res. 511, 165–168.

Nebes, R.D., Pollock, B.G., Houck, P.R., Butters, M.A., Mulsant, B.H.,Zmuda, M.D., Reynolds III, C.F., 2003. Persistence of cognitiveimpairment in geriatric patients following antidepressanttreatment: a randomized, double-blind clinical trial withnortriptyline and paroxetine. J. Psychiatr. Res. 37, 99–108.

Ohara, K., Kondo, N., Ohara, K., 1998. Changes of monoamines inpost-mortem brains from patients with diffuse Lewy bodydisease. Prog. Neuro-psychopharmacol. Biol. Psychiatry 22,311–317.

O'Hearn, E., Battaglia, G., De Souza, E.B., Kuhar, M.J., Molliver, M.,1988. Methylenedioxy amphetamine (MDA) andmethylenedioxymethamphetamine (MDMA) cause selectiveablation of serotonergic axon terminals in forebrain:immunocytochemical evidence for neurotoxicity.J. Neurosci. 8, 2788–2803.

Perry, E.K., Irving, D., Kerwin, J.M., McKeith, I.G., Thompson, P.,Collerton, D., Fairbairn, A.F., Ince, P.G., Morris, C.M., Cheng, A.V.,et al., 1993. Cholinergic transmitter and neurotrophic activitiesin Lewy body dementia: similarity to Parkinson's and distinctionfrom Alzheimer disease. Alzheimer Dis. Assoc. Disord. 7,69–79.

Qian, Y., Melikian, H.E., Rye, D.B., Levey, A.I., Blakely, R.D., 1995.Identification and characterization of antidepressant-sensitiveserotonin transporter proteins using site-specific antibodies.J. Neurosci. 15 (2), 1261–1274.

Rampello, L., Chiechio, S., Raffaele, R., Vecchio, I., Nicoletti, F.,2002. The SSRI, citalopram, improves bradykinesia in patientswith Parkinson's disease treated with L-dopa.Clin. Neuropharmacol. 25 (1), 21–24.

Sandyk, R., Fisher, H., 1988. Serotonin in involuntary movementdisorders. Int. J. Neurosci. 42, 185–208.

Schmitt, J.A., Wingen, M., Ramaekers, J.G., Evers, E.A., Riedel, W.J.,2006. Serotonin and human cognitive performance.Curr. Pharm. Des. 12, 2473–2486.

Schreinzer, D., Ballaban, T., Brannath, W., Lang, T., Hilger, E.,Fasching, P., Fischer, P., 2005. Components of behavioralpathology in dementia. Int. J. Geriatr. Psychiatry 20, 137–145.

Schrag, A., 2006. Quality of life and depression in Parkinson'sdisease. J. Neurol. Sci. 248, 151–157.

194 B R A I N R E S E A R C H 1 2 1 7 ( 2 0 0 8 ) 1 8 5 – 1 9 4

Shiurba, R.A., Spooner, E.T., Ishiguro, K., Takahashi, M., Yoshida,R., Wheelock, T.R., Imahori, K., Cataldo, A.M., Nixon, R.A., 1998.Immunocytochemistry of formalin-fixed human brain tissues:microwave irradiation of free-floating sections. Brain Res.Brain Res. Protoc. 2, 109–119.

Sparks, D.L., Markesbery, W.R., 1991. Altered serotonergic andcholinergic synaptic markers in Pick's disease. Arch. Neurol.48, 796–799.

Steinbusch, H.W., 1981. Distribution of serotonin-immunoreactivityin the central nervous system of the rat-cell bodies andterminals. Neuroscience 6, 557–618.

Stout, S.S., Somerset, W.I., Miller, A., Musselman, D.L., 2003.Paroxetine use in medically ill patients. Psychopharmacol.Bull. 37 (Suppl 1), 108–122.

Szabo, Z., Kao, P.F., Mathews, W.B., Ravert, H.T., Musachio, J.L.,Scheffel, U., Dannals, R.F., 1996. Positron emission tomographyof 5-HT reuptake sites in the human brain with C-11 McN5652extraction of characteristic images by artificial neural networkanalysis. Behav. Brain Res. 73, 221–224.

Truchot, L., Costes, S.N., Zimmer, L., Laurent, B., Le Bars, D.,Thomas-Anterion, C., Croisile, B., Mercier, B., Hermier, M.,Vighetto, A., Krolak-Salmon, P., 2007. Up-regulation ofhippocampal serotonin metabolism in mild cognitiveimpairment. Neurology 69, 1012–1017.

Underwood, M.D., Khaibulina, A.A., Ellis, S.P., Moran, A., Rice, P.M.,Mann, J.J., Arango, V., 1999. Morphometry of the dorsal raphenucleus serotonergic neurons in suicide victims.Biol. Psychiatry 46, 473–483.

van Luijtelaar, M.G., Steinbusch, H.W., Tonnaer, J.A., 1989.Similarities between aberrant serotonergic fibers in the agedand 5,7-DHT denervated young adult rat brain. Exp. Brain Res.78, 81–89.

Varnas, K., Halldin, C., Hall, H., 2004. Autoradiographic distributionof serotonin transporters and receptor subtypes in humanbrain. Hum. Brain Mapp. 22, 246–260.

Wetherell, J.L., Gatz, M., Johansson, B., Pedersen, N.L., 1999. Historyof depression and other psychiatric illness as risk factors forAlzheimer disease in a twin sample. Alzheimer Dis. Assoc.Disord. 13, 47–52.

Wiklund, L., Bjorklund, A., 1980. Mechanisms of regrowth in thebulbospinal serotonin system following 5,6-dihydroxytryptamineinduced axotomy. II. Fluorescence histochemical observations.Brain Res. 191, 109–127.

Wilson,M.A., Molliver, M.E., 1994.Microglial response to degenerationof serotonergic axon terminals. Glia 11, 18–34.

Zhou, F.C., Azmitia, E.C., 1986. Induced homotypic sprouting ofserotonergic fibers in hippocampus. II. An immunocytochemistrystudy. Brain Res. 373, 337–348.

Zhou, F.C., Bledsoe, S., Lumeng, L., Li, T.K., 1994. Reducedserotonergic immunoreactive fibers in the forebrain ofalcohol-preferring rats. Alcohol Clin. Exp. Res. 18,571–579.

Zhou, F.C., Xu, Y., Bledsoe, S., Lin, R., Kelley, M.R., 1996. Serotonintransporter antibodies: production, characterization, andlocalization in the brain. Brain Res. Mol. Brain Res. 43,267–278.

Dystrophic serotonergic axons in neurodegenerative diseasesIntroductionResults5-HTT patterns in neuropathologically normal brainsAscending projectionsTerminal distribution

5-HTT patterns in the brain in neurodegenerative diseasesDiffuse Lewy-body dementiaParkinson's diseaseFrontal lobe dementiaMorphometric analysis of pathologic changes

DiscussionSerotonin system and neurodegenerationComparison of axonal pathologyPossible relevance of 5-HTT axon degeneration to symptoms of neurodegenerative diseasesStudy limitations

Experimental proceduresBrainsSelection of regions for studyOrientationImmunocytochemistryMorphometric methodsStatistical analysis

AcknowledgmentsReferences