A

bodawhltrit©

K

1

dutawa

Bf

0d

Neurobiology of Aging 31 (2010) 46–57

Medial septal �-amyloid 1-40 injections alter septo-hippocampalanatomy and function

Luis V. Colom a,b,∗, Maria T. Castaneda a, Cristina Banuelos a, Gustavo Puras a,Antonio Garcıa-Hernandez a, Sofia Hernandez a, Suzanne Mounsey a,

Joy Benavidez a, Claudia Lehker a,b

a Department of Biological Sciences at the University of Texas at Brownsville/Texas Southmost College,80 Fort Brown, Brownsville, TX 78520, USA

b The Center for Biomedical Studies, USA

Received 28 February 2008; received in revised form 22 April 2008; accepted 1 May 2008Available online 10 June 2008

bstract

Degeneration of septal neurons in Alzheimer’s disease (AD) results in abnormal information processing at cortical circuits and consequentrain dysfunction. The septum modulates the activity of hippocampal and cortical circuits and is crucial to the initiation and occurrence ofscillatory activities such as the hippocampal theta rhythm. Previous studies suggest that amyloid � peptide (A�) accumulation may triggeregeneration in AD. This study evaluates the effects of single injections of A� 1-40 into the medial septum. Immunohistochemistry revealeddecrease in septal cholinergic (57%) and glutamatergic (53%) neurons in A� 1-40 treated tissue. Additionally, glutamatergic terminalsere significantly less in A� treated tissue. In contrast, septal GABAergic neurons were spared. Unitary recordings from septal neurons andippocampal field potentials revealed an approximately 50% increase in firing rates of slow firing septal neurons during theta rhythm andarge irregular amplitude (LIA) hippocampal activities and a significantly reduced hippocampal theta rhythm power (49%) in A� 1-40 treatedissue. A� also markedly reduced the proportion of slow firing septal neurons correlated to the hippocampal theta rhythm by 96%. These

esults confirm that A� alters the anatomy and physiology of the medial septum contributing to septo-hippocampal dysfunction. The A�nduced injury of septal cholinergic and glutamatergic networks may contribute to an altered hippocampal theta rhythm which may underliehe memory loss typically observed in AD patients.

2008 Elsevier Inc. All rights reserved.

BAergic

baHesa

eywords: Septo-hippocampal; Amyloid; Glutamatergic; Cholinergic; GA

. Introduction

Alzheimer’s disease (AD) is an age-related progressiveisorder of the brain that leads to memory loss, dementia and,ltimately death. Although AD was described a century ago,he molecular mechanisms underlying neuronal dysfunction

nd degeneration are still unclear. The brain of an individualith AD exhibits extracellular senile plaques of aggregated

myloid � peptide (A�) and a profound loss of basal fore-

∗ Corresponding author at: Department of Biological Sciences, 80 Fortrown, Brownsville, TX 78520, USA. Tel.: +1 956 882 5048;

ax: +1 956 882 5043.E-mail address: luis.colom@utb.edu (L.V. Colom).

io(2attg

197-4580/$ – see front matter © 2008 Elsevier Inc. All rights reserved.oi:10.1016/j.neurobiolaging.2008.05.006

; Septum

rain cholinergic neurons that innervate the hippocampusnd the neocortex (Whitehouse et al., 1982; Selkoe, 2000;ardy and Selkoe, 2002). This loss of basal forebrain cholin-

rgic neurons led to the cholinergic hypothesis of AD whichtates that basal forebrain cholinergic neurons are severelyffected in the course of the disease and that the result-ng cerebral cholinergic deficit leads to memory loss andther cognitive symptoms, which are characteristic of ADDavies and Maloney, 1976; Pearson et al., 1983; Schliebs,005). Recent studies have shown an association between

decline in learning and memory and a deficit in excita-

ory amino acid neurotransmission. This deficit accompanieshe impairment of the cholinergic system which modulateslutamatergic neurotransmission by targeting the neocortical

iology o

aetgg

cctseKtat2hsp

pt1p(CetittClm1wLot2

amb(at(rii

at(e

scPfiarfi2

tnfArektd

tmgdcSpraa1ch2raitaatashua

2

2

L.V. Colom et al. / Neurob

nd hippocampal glutamatergic pyramidal neurons (Francist al., 1999). Thus, AD neurological decline may be attributedo cholinergic hypofunction, as described in the choliner-ic hypothesis, in combination with the loss of excitatorylutamatergic function (Francis et al., 1999).

The septum and the hippocampus are heavily inter-onnected through the fimbria-fornix and are functionallyoupled (Bland and Colom, 1993), often referred to collec-ively as the septo-hippocampal system (Colom, 2006). Theepto-hippocampal projection includes well-known cholin-rgic and GABAergic components (Lynch et al., 1977;ohler et al., 1984; Bland and Colom, 1993). Recent elec-

rophysiological and anatomical studies have shown thatsubpopulation of septal glutamatergic neurons projects

o the hippocampus (Sotty et al., 2003; Hajszan et al.,004; Manseau et al., 2005), constituting 23% of the septo-ippocampal projection (Colom et al., 2005). Hence, theepto-hippocampal projection is a three neurotransmitterathway.

The synchronized depolarization of hippocampal neuronsroduces field potentials in a frequency range of 3–12 Hzypically referred to as theta rhythm (Bland and Colom,993). Several lines of evidence indicate that the septumlays a critical role in hippocampal theta rhythm generationMorales et al., 1971; Colom and Bland, 1991; Bland andolom, 1993; Lee et al., 1994; Vinogradova, 1995; Blandt al., 1999). Furthermore, the occurrence of hippocampalheta rhythm depends on the proportion of septal neuronsnvolved in the rhythmic process while the frequency ofhe theta field activity is determined by the frequency ofhe rhythmical “theta” bursts in septal neurons (Bland andolom, 1993; Vinogradova, 1995; Bland et al., 1999). Septal

esions, in addition to blocking theta, produce severe impair-ents in memory processes (Winson, 1978; Vinogradova,

995). The hippocampal theta rhythm appears to act as aindowing mechanism for synaptic plasticity (Huerta andisman, 1993). Theta also plays a role in the neural codingf place (Winson, 1978; O’Keefe, 1993). Theta also seemso play a role in sensory-motor integration (Bland and Oddie,001).

Human theta oscillations occur during exploratory searchnd goal-seeking behaviors as well as during virtual move-ent when sensory information and motor planning are

oth in flux but not during periods of self-initiated stillnessCaplan et al., 2003). Movement-related theta oscillationsre observed in human hippocampus and cortex, suggestinghat both structures play a role in sensorimotor integrationEkstrom et al., 2005). This suggests that experimental dataegarding A�-induced hippocampal theta rhythm alterationsn rats is important in understanding sensorimotor processingn human patients suffering from AD.

In vitro, medial septal neurons have been classified

ccording to their firing patterns and membrane proper-ies as slow firing, fast firing, regular firing or burst firingJones et al., 1999; Henderson et al., 2001; Garrido-Sanabriat al., 2007). These electrophysiologically characterized

lctI

f Aging 31 (2010) 46–57 47

eptal neurons have been identified in vitro using singleell reverse transcriptase polymerase chain reaction (RT-CR). While cholinergic neurons typically display slowring phenotypes, most GABAergic neurons display fastnd burst firing phenotypes. In contrast, glutamatergic neu-ons display heterogeneous firing properties, including slowring phenotypes (Sotty et al., 2003; Manseau et al.,005).

In animal models of AD, A� peptide 1-42 injec-ions into the medial septum injured neurons. Mosteurons damaged by A� were choline acetyltrans-erase (ChAT) positive, while only minor effects of� were observed on parvalbumin (PV) positive neu-

ons (putative GABAergic) (Harkany et al., 1995). Theffect of A� on glutamatergic septal neurons is notnown due largely to the fact that a significant sep-al glutamatergic neuronal population was only recentlyescribed.

Glutamatergic actions in other regions of the brain showhat glutamate mediates most excitatory synaptic trans-

ission in the brain. In addition, synaptic strength atlutamatergic synapses shows a remarkable degree of use-ependent plasticity which may represent a physiologicalorrelate to learning and memory (McGee and Bredt, 2003).eptal glutamatergic neurons are well posed to control hip-ocampal excitability and rhythmical activities (e.g., thetahythm) that promote synaptic plasticity. Cholinergic mech-nisms modulate glutamatergic synapses and, through thisction, learning and memory formation (Jerusalinsky et al.,997). Glutamatergic neurons, together with basal forebrainholinergic neurons, constitute the two neuronal systemsighly vulnerable to AD (Francis et al., 1999; Bell et al.,006). A�-induced damage of septal glutamatergic neu-ons may play a central role in AD. While glutamatections in various areas of the brain are well character-zed, little is known about septal glutamatergic neurons andheir alterations induced by age-related pathologies suchs AD. Therefore, to investigate the effects of amyloid,n indicator of AD, on all three septal neuronal popula-ions, septal networks and septo-hippocampal function, wedministered single injections of A� 1-40 into the medialeptum. Subsequently, electrophysiological recordings ofipopocampal theta rhythm and immunohistochemistry weresed to assess anatomical and electrophysiological alter-tions.

. Materials and methods

.1. Animals

Adult male Sprague Dawley rats (n = 50; 250–350 g; Har-

an) were housed and maintained on a 12/12-h light/darkycle and provided food and water ad libitum. All animal pro-ocols used in this study were in compliance with the Nationalnstitutes of Health Guide for the Care and Use of Lab-

4 iology o

oCdaa

2

nrioatti

mnFwfpPti

2

Bi(vapmbphupfrd(aamdeD0fe

Hi(a

2

f(pBtmwscSpdpwft

2

Ais1astabuaGapbvapro(cgi

8 L.V. Colom et al. / Neurob

ratory Animals and approved by the Institutional Animalare and Use Committee of UTB/TSC. All surgical proce-ures and perfusions were performed with a ketamine-basednesthetic (ketamine: 42.8 mg/ml, xylazine: 8.6 mg/ml andcepromazine: 1.4 mg/ml in saline; 1 ml/kg).

.2. Aβ septal injections

Synthetic A� peptide 1-40 (Bachem Torrance, Califor-ia) and A� peptide 40-1 (Bachem Torrance, California), aeverse peptide used as a control, were separately dissolvedn 0.01 M phosphate-buffered saline (PBS) at a concentrationf 2 �g/�l and incubated at 37 ◦C for one week before uses described by Gonzalo-Ruiz et al., 2002 to allow aggrega-ion of fibrils. Atomic force microscopy was used to verifyhat incubated A� 1-40 contained soluble oligomers beforenjections (Fig. 1A and B).

Anesthetized animals received a single injection into theedial septum by means of a stereotaxic apparatus (coordi-

ates: AP = 0.5, L = .2, V = −6.5 from the dura) (Fig. 1C).our microliters of A� 1-40, A� 40-1, or PBS were injectedith a 10-�L Hamilton syringe and the needle kept in place

or 15 min before withdrawing. Ten to fifteen animals wererepared for each experimental group (A� 1-40, A� 40-1 andBS). Animals were allowed a one-week recovery period in

heir original housing environment prior to acute electrophys-ological procedures.

.3. Electrophysiology

Animals were initially anesthetized with Isoflurane (Theutler Company, Dublin, OH) while a jugular cannula was

nserted. Isoflurane was then discontinued and UrethaneSigma–Aldrich, St. Louis, MO), 0.8 g/ml was administeredia the jugular cannula to maintain an appropriate level ofnesthesia during the remaining surgical and experimentalrocedures. The rats were placed in a stereotaxic instru-ent (David Kopf Instruments, Tujunga, CA) with the plane

etween bregma and lambda horizontally leveled. Body tem-erature was maintained at 37 ◦C with a self-regulatingeating pad (Fine Science Tools Inc., Foster City, CA). Ann-insulated silver wire (Sigma–Aldrich, St. Louis, MO)laced in the cortex, anterior to bregma served as an indif-erent electrode. Another insulated stainless steel wire forecording hippocampal field activity was placed in the rightorsal hippocampal formation in the dentate molecular layer3.8 mm posterior to bregma, 2 mm lateral to the midlinend 2.5 mm ventral to the dural surface). To show that thetamplitude was not due to the electrode position, the point ofaximum theta amplitude was found in each experiment as

escribed in previous work (Bland and Colom, 1993; Blandt al., 1999). Medial septum diagonal band of Broca (MS-

BB) recordings were made 0.5 mm anterior to bregma,.0–0.5 mm lateral to the midline, and ventral 5.2–7.2 mmrom the dural surface. Cells were recorded with glass micro-lectrodes (15–30 M�) filled with 0.5 M sodium acetate.

N3st

f Aging 31 (2010) 46–57

ippocampal and septal microelectrodes were carried inndependent microdrives, Electrode Manipulator Model 960David Kopf Instruments, Tujunga, CA) and a CMA-12CCctuator (Newport Corporation, Irvine, CA), respectively.

.4. Histology

Animals (n = 35) were deeply anesthetized and per-used intracardially with 0.1 M phosphate-based buffer salinePBS) (pH 7.4) followed by a fixative solution containing 4%araformaldehyde and 0.1% glutaraldehyde in 0.1 M PBS.rains were removed, post-fixed in the same fixative solu-

ion overnight and then cryoprotected in 30% sucrose. Thirtyicrometers slices were cut using a cryostat. Only brainsith confirmed medial septal injection sites were used in this

tudy. Free floating sections were stored at 4 ◦C in well platesontaining PBS until processed for immunohistochemistry.ections surrounding the injection site were mounted androcessed for Thioflavine S to confirm the presence of A�eposits (Fig. 1D). Trajectories of electrodes used in electro-hysiological recordings of theta rhythm in the hippocampusere confirmed with Cresyl violet staining. Only recordings

rom animals with confirmed hippocampal electrode trajec-ories were included in this study (n = 183).

.5. Immunohistochemistry

Free floating immunohistochemistry was performed on� 1-40 treated and control tissue (A� 40-1 and PBS)

n order to visualize neuronal populations in the medialeptum. Briefly, medial septal sections were incubated for0 min in 0.3% H2O2 to inhibit endogenous peroxidasectivity. Non-specific staining was blocked by incubating tis-ue in 10% bovine serum albumin (BSA) for 1 h at roomemperature. Sections were incubated overnight in primaryntibodies at room temperature. A mouse anti-NeuN anti-ody (1:1000, Chemicon) which stains neuronal nuclei wassed to reveal the overall number of neurons. GABAergicnd cholinergic neurons were visualized using a mouse anti-AD67 antibody (1:1000, Chemicon) and a goat anti-ChAT

ntibody (1:200, Chemicon), respectively. Glutamatergicopulations were revealed using a mouse anti-glutamate anti-ody (1:3000, Immunostar). Glutamatergic terminals wereisualized using guinea pig anti-VGLUT1 (1:2000) andnti-VGLUT2 (1:4000) antibodies (Chemicon). Followingrimary antibody incubation, sections were incubated atoom temperature for 2 h in their respective biotinylated sec-ndary antibodies (1:200), NeuN labeling: goat anti-mouseVector), GABAergic labeling: goat anti-mouse (Vector),holinergic labeling: donkey anti-goat (Vector), glutamater-ic labeling: goat anti-mouse (Vector). Finally, sections werencubated in avidin–biotin complex (ABC, Vector) for 1 h.

euronal populations were revealed using the chromogen-3′-diaminobenzidine (DAB, Vector). In each experimentome sections were incubated without the primary antibodyo determine staining specificity. Sections were mounted onto

L.V. Colom et al. / Neurobiology of Aging 31 (2010) 46–57 49

Fig. 1. Atomic force microscope (AFM) measurements showed minimum oligomerization in non-aged A� samples (A) and the presence of abundant oligomersi on site.o at A� ia

gi

2

mcCtbttwoto

2c

aaC

g(AapaItdl(ctayatlbab

n aged samples of A� 1-40 (B). (C) Diagram depicting medial septum injectif A� 1-40. Insert shows a magnified image of the injection, illustrating thdjacent areas were used.

elatin-coated slides, dehydrated in graded ethanol, clearedn xylene and coverslipped for further analysis.

.6. Digital imaging and quantification

Bright field images were captured using an Axiovert 200icroscope (Zeiss) equipped with an Optronics CCD camera

oupled to StereoInvestigator software (MBF BioScience).ontours of the medial septal region were determined using

he corresponding sections of the stereotaxic atlas of the ratrain (Paxinos and Watson, 1998). The area of the contour andhe number of cells within each contour was determined usinghe meander scan function in StereoInvestigator. Cell densityas calculated by dividing the number of cells by the areaf the contour in which the cells were located. Glutamatergicerminals (VGLUT1 and VGLUT2) were estimated using theptical fractionator probe in StereoInvestigator.

.7. Data acquisition, analysis and firing patternlassification

Brain signals were displayed, digitized, and sampled atfrequency of 10 kHz with a 12-bit DT-2839 A/D board

nd SciWorks 3.0 SP1 (DataWave Technologies, Longmont,O), and recorded for off-line analysis. Electroencephalo-

rDpr

(D) Fluorescent image of Thioflavine S in medial septum verifying presences isolated to the medial septum. Only injections with minimal diffusion to

raphic (EEG) signals were amplified and filtered on-linelow-pass at 100 Hz) using an AC/DC amplifier (3000 model,-M Systems, Inc., Carlsborg, WA). Cell recordings were

mplified and filtered on-line (low-pass at 2000 Hz, high-ass at 500 Hz) using a NEURODATA IR-183A recordingmplifier and a FLA-01 filter/amplifier (Cygnus Technology,nc. Delaware Water Gap, PA). Hippocampal field poten-ials and septal cell discharges were simultaneously recordeduring four hippocampal field conditions: (1) large irregu-ar amplitude (LIA) only, (2) transition from LIA to theta,3) theta only, and (4) transition from theta to LIA. Stableell recordings were made for an average of 30 min to insurehat a minimum of 5–10 30-s transitions were acquired fornalysis. Each EEG was subjected to a fast Fourier anal-sis, Clampfit 9.2 (Molecular Devices, CA), and classifieds either theta or LIA by the following criteria: (1) theheta rhythm functional state was defined as a sinusoidal-ike waveform with a peak frequency of 3–8 Hz and a smallandwidth, and (2) the “LIA” functional state was defined aslarge amplitude irregular activity with a broad frequency

and (0.5–25.0 Hz) (Leung et al., 1982). Analysis of cell

ecordings (30 s) using Clampfit 9.2 software (Molecularevices) provided the mean, firing frequency (Hz), actionotential duration (ms), and amplitude (mV). Septal neu-ons were classified as either slow firing or fast firing as

5 iology of Aging 31 (2010) 46–57

dSifi≥oacwps

2

pawflsuC

rtKntpsAs

3

3

4adCsA

Fig. 2. Graph comparing the mean cell density of overall (NeuN), choliner-gic (ChAT), glutamatergic (Gluta) and GABAergic (GABA) populations inthe medial septum of PBS control rats, A� (40-1) control rats and A� 1-40treated rats. A� 1-40 significantly reduced the overall number of medial sep-tal neurons labeled by NeuN (F[2,20] = 4.69, p = 0.02). While the numbers ofboth cholinergic (F[2,32] = 11.2, p < 0.01) and glutamatergic (F[2,32] = 2.32,pts

iiS

3n

gnpoTs(

3g

TM

P

NCGG

0 L.V. Colom et al. / Neurob

escribed by previous studies (Brazhnik and Fox, 1997, 1999;otty et al., 2003; Colom et al., 2006). Septal units hav-

ng a mean firing frequency <12 Hz were considered slowring neurons and units having a mean firing frequency12 Hz were considered fast firing neurons. Firing peri-

dicity (rhythmicity) was examined using autocorrelationnalysis. Rhythmical units showed periodicity in their auto-orrelations (Bland and Colom, 1993). Cross-correlogramsere used to determine whether septal units and hippocam-al theta rhythm recordings were related (SciWorks 3.0 SP1oftware).

.8. Statistical analysis

Mean cell densities of each neuronal population were com-ared amongst A� 1-40, A� 40-1 and PBS treated rats usingone-way analysis of variance (ANOVA). ANOVA analysisas followed by Tukey’s post hoc analysis. Significant dif-

erences for all statistical testing were defined by a p value ofess than 0.05. Numerical data are represented as means andtandard errors (S.E.M.). All statistical tests were performedsing statistical analysis software (SPSS 14.0, SPSS, Inc.,hicago, Illinois).

Mean cell firing frequencies and hippocampal thetahythm power spectrum values were compared amongsthe three groups (A� 1-40, A� 40-1, and PBS) using theruskal–Wallis test. To find if there was a statistically sig-ificant reduction in the numbers of recorded neurons in thehree groups, the Difference in Proportions Test for Two Inde-endent Proportions was used. Significant differences for alltatistical testing were defined by a p value of less than 0.05.ll statistical tests were performed using StatsDirect 2.6.6

tatistical software.

. Results

.1. Histology

A� deposits were detectable at the injection site of A� 1-0 treated tissue as verified by Thioflavine S staining (Fig. 1)nd Congo Red (not shown) while no amyloid deposits were

etectable in A� 40-1 and PBS treated tissue (not shown).ells with glial morphology were visible near the injection

ite in stained tissue of all experimental groups (A� 1-40,� 40-1, PBS). However, glia was visibly more extensive

oca

able 1ean cell densities of neuronal populations within the MS of PBS control rats, A�

henotype PBS (cell/mm2)

euN* 1261.8 ± 164.1 (n = 5)hAT* 287.9 ± 47.7 (n = 11)AD67 273.5 ± 29.3 (n = 11)luta* 770.3 ± 64.1 (n = 11)

* Statistically significant, p < 0.05.

< 0.01) neurons were significantly reduced GABAergic neurons were resis-ant to A� 1-40 (p > 0.55). PBS or the reverse peptide did not produceignificant alterations. (*) Statistically significant, P < 0.05.

n A� 1-40 treated tissue (not shown). This glial reactions consistent with previous studies (Giovannelli et al., 1995;cali et al., 1999).

.2. Aβ 1-40 produces an overall reduction in septaleurons

Analysis of NeuN-positive neuronal cell density in allroups revealed a significant decrease in NeuN positiveeurons in A� 1-40 treated tissue (28%, p = 0.021) com-ared to control groups (PBS treated tissue), indicating anverall reduction of septal neurons (Figs. 2 and 3A–C).here was no significant difference between the cell den-ities of PBS treated tissue and tissue treated with A� 40-1Table 1).

.3. Aβ 1-40 selectively injures cholinergic andlutamatergic septal neurons

Cholinergic neuronal density assessed by ChAT stainingf somas was significantly reduced by 57% compared to PBSontrols (p = .001) (Table 1, Figs. 2 and 3D–F). Addition-lly, glutamatergic neuronal density assessed by glutamate

40-1 control rats and A� 1-40 treated rats using ANOVA

A� 40-1 (cell/mm2) A� 1-40 (cell/mm2)

1240.5 ± 137.7 (n = 6) 910.1 ± 48.5 (n = 12)246.3 ± 24.1 (n = 6) 122.4 ± 9.2 (n = 18)248.5 ± 23.0 (n = 6) 232.2 ± 26.1 (n = 17)710.2 ± 42.0 (n = 6) 362.7 ± 31.0 (n = 18)

L.V. Colom et al. / Neurobiology of Aging 31 (2010) 46–57 51

F I) immE sity in Ar 0 (L) tre

sctiTmGb(

3t

V

ig. 3. Photomicrographs of NeuN (A–C), ChAT (D–F) and glutamate (G–, H) and A� 1-40 (C, F, I) treated rats. Note the reduction in neuronal den

eduction in glutamatergic cell density in PBS (J), A� 40-1 (K) and A� 1-4

taining was significantly decreased by 53% compared toontrols (p = .001) (Table 1, Figs. 2 and 3G–L). In con-rast, populations of GABAergic neurons assessed by GAD67mmunoreactivity were not significantly reduced (Table 1).hus, A� 1-40 selectively injures cholinergic and gluta-

atergic medial septal neurons while sparing medial septumABAergic neurons. No significant differences were foundetween PBS and A� 40-1 in these neuronal populationsTable 1).

rr2V

unoreactive neurons in the medial septum of PBS (A, D, G), A� 40-1 (B,� 1-40 treated tissue compared to control tissue. Panels J–L illustrate theated rats. Scale bar = 50 �m.

.4. Aβ 1-40 significantly reduces glutamatergicerminals in the medial septum

Glutamatergic terminals assessed by VGLUT1 andGLUT2 punctate immunoreactivity were significantly

educed following A� 1-40 injections into the medial septalegion (Figs. 4 and 5). VGLUT1 terminals were reduced by1% compared to PBS controls (p = .001) (Figs. 4 and 5A–C).GLUT2 terminals were reduced by 40% compared to PBS

52 L.V. Colom et al. / Neurobiology o

Fig. 4. Graph comparing the mean density of VGLUT1 and VGLUT2 punctain the medial septum of PBS control rats (n = 10), A� (40-1) control rats(ip

cdcabta

3n

sp1

4cbtd

31

httrtFic

3s

PtwPt3

F1

n = 6) and A� treated rats (n = 10). In A� 40-1 treated tissue, terminal densityn both VGLUT1 (F[2,23] = 52.5, p < 0.01) and VGLUT2 (F[2,32] = 128.4,< 0.01) was significantly reduced. (*) Statistically significant, p < 0.05.

ontrols (p = .001) (Figs. 4 and 5D–F). In control tissue, theensity of VGLUT2 immunoreactive punctate was signifi-antly more abundant than VGLUT1 punctate which is ingreement with previous studies (Hajszan et al., 2004). Whileoth VGLUT1 and VGLUT2 puncta were reduced in A� 1-40issue, VGLUT2 terminals in the medial septum were moreffected than VGLUT1 terminals.

.5. Aβ 1-40 altered the firing pattern of slow firingeurons

The mean firing rate of slow firing septal neurons wasignificantly higher in A� 1-40 treated animals when com-ared to PBS and A� 40-1 control animals (Fig. 6A). In A�-40 treated animals, slow firing neurons fired at a 55% and

pwns

ig. 5. Photomicrograph of VGLUT1 (A–C) and VGLUT2 (D–F) immunoreactive-40 (C, F) treated rats (n = 10).

f Aging 31 (2010) 46–57

6% higher rate during theta rhythm and LIA, respectively,ompared to controls. There was no significant differenceetween the mean firing rates of the PBS treated animals andhe A� 40-1 treated animals (Fig. 6, Table 2). Accumulativeata are depicted in Fig. 6B.

.6. Hippocampal theta rhythm is abnormal in the Aβ

-40 treated rat

Power spectrum analysis from the EEG recordings in theippocampus of the three experimental groups showed thatheta rhythm amplitude at peak frequency was altered inhe A� 1-40 treated group. Theta rhythm amplitude waseduced 49% (54592 mV2/Hz and 48989 mV2/Hz in con-rols to 26514 mV2/Hz in A� 1-40 treated animals, p = 0.001;ig. 6C–D) compared to controls. No significant difference

n power amplitude was found between PBS and A� 40-1ontrols.

.7. Aβ 1-40 reduced the proportion of theta-correlatedlow firing neurons

Examples of theta-correlated slow firing neurons fromBS, A� 40–1 and 1–40 are depicted in Fig. 7A–C, respec-

ively. The percentage of theta-correlated slow firing neuronsas significantly reduced by A� 1-40 septal injections. In theBS and A� 40-1 treated groups, the percentage of recorded

heta-correlated slow firing neurons was 35.6% (21/59) and3.7% (16/49), respectively. In the A� 1-40 treated group, the

ercentage of recorded theta-correlated slow firing neuronsas 04.0% (2/51). In comparison to controls, there was a sig-ificant reduction in the number of recorded theta-correlatedlow firing neurons recorded in the A� 1-40 treated group

puncta in the medial septum of PBS (A, D), A� 40-1 (B, E) (n = 6) and A�

L.V. Colom et al. / Neurobiology of Aging 31 (2010) 46–57 53

Fig. 6. (A) Comparison of hippocampal field potentials (top trace) and firing patterns (bottom trace) of slow firing neurons (unit) from PBS, A� 40-1, andA mean fic uency oo imals. (A cally si

(s

4

(gttao

rJvGhtWe

� 1-40 treated animals. (B) Bar graph depicting a significant increase inontrols. (C) Graph of power spectrum of theta frequencies versus theta freqscillations in controls but lowered the theta amplitude in A� 1-40 treated an� 1-40 treated animals is significantly less than that of control. (*) Statisti

p = 0.001) (Fig. 7D). For all three experimental groups, notatistical differences were found in the fast firing group.

. Discussion

The septo-hippocampal system is heavily affected in ADColom, 2006). Although A� is implicated in AD neurode-enerative processes, little is known about how it affects

he function of specific septo-hippocampal networks. Fur-hermore, there are no studies correlating septo-hippocampalnatomical and functional alterations in AD or animal modelsf AD.

bp2o

ring rates of slow firing neurons in A� 1-40 treated animals compared tof individual neurons illustrating that tail pinch induced robust field potentialD) Graph indicating power spectrum of recorded theta field potentials fromgnificant, p < 0.05.

While A� has been extensively used in vitro as a neu-otoxin (Colom et al., 1998; Butterfield, 2002; Chen, 2005;arvis et al., 2007; Liao et al., 2007), comparative few inivo studies have been performed (Giovannelli et al., 1995;onzalo-Ruiz et al., 2003, 2006; Gonzalez et al., 2007). A�as been applied in vivo using: (a) intraventricular injec-ions and (b) local injections into defined brain structures.

hile the first approach has been used to study generalffects (Nakamura et al., 2001), the second has primarily

een used to determine A� effects on specific neuronalopulations (Giovannelli et al., 1995; Gonzalo-Ruiz et al.,003, 2006; Gonzalez et al., 2007). Since the primary goalf this study was to investigate A� affects on a specific

54 L.V. Colom et al. / Neurobiology of Aging 31 (2010) 46–57

Table 2Mean firing rates of different neuronal firing patterns during theta rhythm and LIA

Firing pattern PBS (n = 69) (spikes/s) A� 40-1 (n = 53) (spikes/s) A� 1-40 (n = 61) (spikes/s)

Slow firingTheta-correlated k = 21 k = 16 k = 2

Firing rate at theta* 2.90 ± 0.54 4.03 ± 0.64 5.00 ± 0.80Firing rate at LIA* 2.57 ± 0.63 2.04 ± 0.42 1.00 ± 0.40

Non-theta-correlated k = 38 k = 33 k = 49Firing rate at theta* 1.15 ± 0.32 1.58 ± 0.39 3.93 ± 0.46Firing rate at LIA* 1.61 ± 0.42 1.41 ± 0.53 3.55 ± 0.42

Fast firingTheta correlated k = 4 k = 1 k = 2

Firing rate at theta 19.55 ± 3.31 21.20 19.00 ± 3.80Firing rate at LIA 11.35 ± 2.00 4.80 15.80 ± 9.80

Non-theta-correlated k = 6 k = 3 k = 8

ru

t

FwPPtS

Firing rate at theta 19.66 ± 2.46Firing rate at LIA 15.87 ± 3.85

* Statistically significant, p < 0.05.

egion of the basal forebrain, local injections of A� weretilized.

Our data demonstrate that local injections of A� intohe medial septum damage both cholinergic and glutamater-

gnse

ig. 7. (A–C) Slow firing unit recordings (second trace) during theta and LIA episith hippocampal field potentials during theta and LIA. (D) Bar graph comparingBS and A� 40-1 treated groups. Slow firing neurons correlated with hippocampalBS and A� 40-1 treated groups, recordings of slow firing neurons correlated to h

o controls, there was a significant reduction in the number of slow firing neuronstatistically significant, p < 0.05.

29.27 ± 7.62 19.11 ± 2.4120.73 ± 5.39 18.97 ± 5.36

ic septal neuronal populations while sparing GABAergiceuronal populations. The use of an antibody against thetructural protein NeuN confirms that the decrease in cholin-rgic and glutamatergic markers is due to a reduction in

odes. Time histograms show neurons cross-correlated and auto-correlatednon-theta-correlated and theta-correlated slow firing neurons in A� 1-40,theta rhythm decreased by 96.0% after septal injections of A� 1-40. In theippocampal theta rhythm were 35.6% and 33.7%, respectively. Comparedcorrelated to hippocampal theta rhythm in the A� 1-40 treated group. (*)

iology o

tThrataamwTnicpa

prlrdsmstoh

drficVpi1or(aha

trdasTtPisti

mttfetKttg2

ttttgioa(f1rt(cagomtAGati

aiioieietsateo

L.V. Colom et al. / Neurob

he population of septal neurons and not transitory injuries.hus, our animal model mimics the pathology observed inumans with AD. While the vulnerability of cholinergic neu-ons to AD and A� has been extensively documented (Daviesnd Maloney, 1976), the vulnerability of glutamatergic sys-ems has only recently begun to be investigated (Francis etl., 1999). Our data show that septal glutamatergic neuronsre affected by A�. Our data also show that septal gluta-atergic terminals (VGLUT2) and glutamatergic terminalsith an extraseptal origin (VGLUT1) are affected by A�.hus, the septal glutamatergic system and the circuits itseurons integrate are highly vulnerable to A�. This A�-nduced glutamatergic damage in addition to the A� inducedholinergic lesion may account for the septo-hippocampalhysiological alterations observed in our analysis of A�ctions.

A� altered septal networks and produced a change in hip-ocampal theta rhythm. Recordings of hippocampal thetahythm in A� 1-40 treated rats displayed a significantlyower power/amplitude (49% reduction) when compared toecordings from rats injected with PBS and A� 40-1. Theecrease in hippocampal theta amplitude suggests that theeptal networks necessary for theta rhythm production andaintenance are altered by A� 1-40 injections in the medial

eptum of rats. This reduction in the theta rhythm ampli-ude may be a consequence of the reduction in the numberf theta-correlated slow firing septal neurons in the septo-ippocampal networks.

The occurrence of hippocampal theta rhythm is depen-ent upon the proportion of septal neurons involved in thehythmic process. Furthermore, the frequency of the thetaeld activity is determined by the frequency of the rhythmi-al “theta” bursts in septal neurons (Bland and Colom, 1993;inogradova, 1995). Septal lesions abolish the hippocam-al theta rhythm in rodents, providing further evidence of themportance of the septum in theta generation (Andersen et al.,979; Rawlins et al., 1979; Winson, 1978). Thus, a reductionf septal cholinergic (57%) and glutamatergic (53%) neu-onal populations plus a reduction of glutamatergic terminalsVGLUT1: 21%, VGLUT2: 40%) is sufficient to significantlylter the amplitude of hippocampal theta rhythm as well as theippocampal functions that require neuronal synchronizationt theta frequencies.

One possible reason for this alteration is that A� sep-al injections are acting at the hippocampal level through aeduction of acetylcholine release (Abe et al., 1994). Theiminished amount of acetylcholine may not be sufficient toppropriately synchronize hippocampal networks. Reducedeptal glutamatergic influences most likely add to this effect.heta frequency and amplitude are affected by lateral sep-

um modulation of NMDA receptors (Puma et al., 1996;uma and Bizot, 1999; Bland et al., 2007). Furthermore, the

njection of NMDA receptor antagonists into the MS-DBignificantly decreases hippocampal theta rhythm ampli-ude (Leung and Shen, 2004). Puma’s and Leung’s resultsn conjunction with our findings suggest that septal gluta-

insr

f Aging 31 (2010) 46–57 55

atergic synapses play an important role in hippocampalheta rhythm generation. Deficit in glutamatergic synapticransmission and synchronous network activity have beenound in hippocampal slices from transgenic mice overxpressing the human form of the amyloid precursor pro-ein (APP) harboring the Swedish mutation (Brown andozlowski, 2004), suggesting that A� increases may lead

o dysfunctional glutamatergic systems. In the medial sep-um bath application of A� (1-40 and 25-35) also depressedlutamatergic synaptic transmission (Santos-Torres et al.,007).

In contrast to the amplitude/power changes observed,he theta frequency was unchanged in recordings from A�-reated rats compared to control rats. This finding suggestshat the surviving neurons and circuits are adequate to main-ain oscillatory activities at theta frequencies and that basicenerator mechanisms continue operating at the same pacen A�-treated rats. The exact mechanism underlying thetascillations is still under debate (Colom, 2006). In A� treatednimals, oscillations at theta frequencies may be produced bya) surviving neurons of A�-altered septal circuits, (b) unaf-ected extra-septal networks or (c) both (Bland and Colom,993; Colom, 2006). Our data suggest that A�-inducedeductions in theta power may be produced by the reduc-ion in rhythmically firing septal neurons. Bassant’s groupSimon et al., 2006) did not find septal cholinergic neuronsorrelated to the hippocampal theta rhythm in anesthetizednd non-anesthetized in vivo preparations. Their work sug-ests that most rhythmical septal neurons are GABAergicr glutamatergic. Reduction of septal cholinergic and gluta-atergic inputs onto GABAergic septal neurons may reduce

he population of rhythmically bursting GABAergic neurons.lthough unaffected in numbers, the population of septalABAergic neurons may be dysfunctional in A�-treated

nimals. Juxtacellular or intracellular recordings with iden-ification of neuronal phenotypes are needed to clarify thisssue.

At the unitary level slow firing neurons from A�-treatednimals significantly increased their firing rates. This find-ng suggests that: (a) surviving slow firing neurons mustncrease their activity to support a reduced amplitude thetar (b) increased firing patterns are a direct result of A�-nduced toxicity. Discerning between these two possibilitiesxceeds the scope of this study and will be the aim of futurenvestigations in our laboratory. Slow firing septal cholin-rgic neurons have typically been considered important forheta generation (Yoder and Pang, 2005; Colom, 2006). Thistudy shows that altered hippocampal theta is associated withltered slow firing patterns in septal neurons. Slow firing sep-al neurons, most likely cholinergic and glutamatergic (Sottyt al., 2003; Manseau et al., 2005), may provide a backgroundf excitation for theta rhythm generation. This activity may be

ncreased in the A�-injured brain to compensate for the septaleuronal loss. Thus, our data further support the notion thatlow firing septal neurons play a role in hippocampal thetahythm generation.

5 iology o

atnVtneriiomc

dsaTcttafoparbmirnamsld

tgTrttt

A

asM

fl

R

A

A

B

B

B

B

B

B

B

B

B

C

C

C

C

C

C

C

D

E

6 L.V. Colom et al. / Neurob

Preliminary experiments not reported in this study usingnti-VGLUT1 and anti-VGLUT2 antibodies in colchicine-reated rats revealed that VGLUT2 neurons constitute a largereuronal population in the medial septum when compared toGLUT1 neurons. Thus, numerous VGLUT2 terminals in

he medial septum may originate from intraseptal VGLUT2eurons, while VGLUT1 terminals might representxtraseptal VGLUT1 projections. These results suggest thateductions in VGLUT2 puncta may correspond to reductionsn septal glutamate-immunoreactive neurons but reductionsn VGLUT1 puncta represent lesions of septal terminals thatriginate in extraseptal glutamatergic populations. Thus, A�ay injure both local circuits and circuits with extraseptal

omponents.Excitotoxic mechanisms have been implicated in cell

eath in AD (Gray and Patel, 1995). This is furtherupported by the neuroprotective effect of the NMDAntagonist memantine (Hynd et al., 2004) in AD patients.he finding that glutamatergic septal neurons are sus-eptible to A� may indicate that A� affects the septumhrough excitotoxic mechanisms. The presence of func-ional NMDA receptors in septal neurons (Kumamotond Murata, 1995) suggests that the machinery neededor excitotoxic processes is present in the septum. Somef the slow firing neurons recorded in this study wererobably glutamatergic (Sotty et al., 2003; Manseau etl., 2005). Slow firing neurons showed increased firingates following A� 1-40 treatment. Those alterations maye part of a process that leads to increases in gluta-ate release and excessive activation of NMDA receptors

n A�-treated rats. This mechanism may lead to neu-odegeneration and explain the reduction in slow firingeuronal numbers in our electrophysiological recordingss well as the reduced number of cholinergic and gluta-atergic neurons in our anatomical experiments. Further

tudies are necessary to determine the mechanisms under-ying A�-induced septal functional alterations and neuronalegeneration.

In conclusion, this study shows that cholinergic and glu-amatergic septal neurons are vulnerable to A� and thatlutamatergic circuits are injured in the presence of amyloid.his alteration in septal neuronal populations and circuitry

esults in abnormal hippocampal theta rhythms suggestinghat these neurons and circuits are important in maintaininghe amplitude of septo-hippocampal rhythmic activities atheta frequencies.

cknowledgements

We thank Alicia Gonzalo-Ruiz for her valuable advicebout A� preparation and administration. This study was

upported by National Institutes of Health: Grants – 1P20

D0001091 and 5S06GM0688550 – Dr. Luis V. Colom.Disclosure statement: There are no actual or potential con-

icts of interest for any of the authors.

F

G

f Aging 31 (2010) 46–57

eferences

be, E., Casamenti, F., Giovannelli, L., Scali, C., Pepeu, G., 1994. Admin-istration of amyloid beta-peptides into the medial septum of ratsdecreases acetylcholine release from hippocampus in vivo. Brain Res.636, 162–164.

ndersen, P., Bland, H.B., Myhrer, T., Schwartzkroin, P.A., 1979. Septo-hippocampal pathway necessary for dentate theta production. Brain Res.165, 13–22.

ell, K.F., Ducatenzeiler, A., Ribeiro-da-Silva, A., Duff, K., Bennett,D.A., Cuello, A.C., 2006. The amyloid pathology progresses in aneurotransmitter-specific manner. Neurobiol. Aging 27, 1644–1657.

land, B.H., Colom, L.V., 1993. Extrinsic and intrinsic properties under-lying oscillation and synchrony in limbic cortex. Prog. Neurobiol. 41,157–208.

land, B.H., Oddie, S.D., 2001. Theta band oscillation and synchrony in thehippocampal formation and associated structures: the case for its role insensorimotor integration. Behav. Brain Res. 127, 119–136.

land, B.H., Oddie, S.D., Colom, L.V., 1999. Mechanisms of neural syn-chrony in the septohippocampal pathways underlying hippocampal thetageneration. J. Neurosci. 19, 3223–3237.

land, B.H., Declerck, S., Jackson, J., Glasgow, S., Oddie, S., 2007. Sep-tohippocampal properties of N-methyl-D-asparate-induced theta-bandoscillation and synchrony. Synapse 61, 185–197.

razhnik, E.S., Fox, S.E., 1997. Intracellular recordings from medial sep-tal neurons during hippocampal theta rhythm. Exp. Brain Res. 114,442–453.

razhnik, E.S., Fox, S.E., 1999. Action potentials and relations to the thetarhythm of medial septal neurons in vivo. Exp. Brain Res. 127, 244–258.

rown, D.R., Kozlowski, H., 2004. Biological inorganic and bioinorganicchemistry of neurodegeneration based on prion and Alzheimer diseases.Dalton Trans., 1907–1917.

utterfield, D.A., 2002. Amyloid beta-peptide (1-42)-induced oxida-tive stress and neurotoxicity: implications for neurodegeneration inAlzheimer’s disease brain. A review. Free Radic. Res. 36, 1307–1313.

aplan, J.B., Madsen, J.R., Schulze-Bonhage, A., Aschenbrenner-Scheibe,R., Newman, E.L., Kahana, M.J., 2003. Human � oscillations relatedto sensorimotor integration and spatial learning. J. Neurosci. 23,4726–4736.

hen, C., 2005. beta-Amyloid increases dendritic Ca2+ influx by inhibitingthe A-type K+ current in hippocampal CA1 pyramidal neurons. Biochem.Biophys. Res. Commun. 338, 1913–1919.

olom, L.V., 2006. Septal networks: relevance to theta rhythm, epilepsy andAlzheimer’s disease. J. Neurochem. 96, 609–623.

olom, L.V., Bland, B.H., 1991. Medial septal cell interactions in relationto hippocampal field activity and the effects of atropine. Hippocampus1, 15–30.

olom, L.V., Castaneda, M.T., Reyna, T., Hernandez, S., Garrido-Sanabria,E., 2005. Characterization of medial septal glutamatergic neurons andtheir projection to the hippocampus. Synapse 58, 151–164.

olom, L.V., Diaz, M.E., Beers, D.R., Neely, A., Xie, W.J., Appel, S.H.,1998. Role of potassium channels in amyloid-induced cell death. J.Neurochem. 70, 1925–1934.

olom, L.V., Garcia-Hernandez, A., Castaneda, M.T., Perez-Cordova, M.G.,Garrido-Sanabria, E.R., 2006. Septo-hippocampal networks in chron-ically epileptic rats: potential antiepileptic effects of theta rhythmgeneration. J. Neurophysiol. 95, 3645–3653.

avies, P., Maloney, A.J., 1976. Selective loss of central cholinergic neuronsin Alzheimer’s disease. Lancet 2, 1403.

kstrom, A.D., Caplan, J.B., Ho, E., Shattuck, K., Fried, I., Kahana, M.J.,2005. Human hippocampal theta activity during virtual navigation. Hip-pocampus 15, 881–889.

rancis, P.T., Palmer, A.M., Snape, M., Wilcock, G.K., 1999. The choliner-gic hypothesis of Alzheimer’s disease: a review of progress. J. Neurol.Neurosurg. Psychiatry 66, 137–147.

arrido-Sanabria, E.R., Perez, M.G., Banuelos, C., Reyna, T., Hernan-dez, S., Castaneda, M.T., Colom, L.V., 2007. Electrophysiological and

iology o

G

G

G

G

G

G

H

H

H

H

H

H

J

J

J

K

K

L

L

L

L

L

M

M

M

N

O

P

P

P

P

R

S

S

S

S

S

S

V

W

L.V. Colom et al. / Neurob

morphological heterogeneity of slow firing neurons in medial sep-tal/diagonal band complex as revealed by cluster analysis. Neuroscience146, 931–945.

iovannelli, L., Casamenti, F., Scali, C., Bartolini, L., Pepeu, G., 1995.Differential effects of amyloid peptides beta-(1-40) and beta-(25-35)injections into the rat nucleus basalis. Neuroscience 66, 781–792.

onzalez, I., Arevalo-Serrano, J., Sanz-Anquela, J.M., Gonzalo-Ruiz, A.,2007. Effects of beta-amyloid protein on M1 and M2 subtypes of mus-carinic acetylcholine receptors in the medial septum-diagonal bandcomplex of the rat: relationship with cholinergic, GABAergic, andcalcium-binding protein perikarya. Acta Neuropathol. 113, 637–651.

onzalo-Ruiz, A., Sanz, J.M., Romero, J.C., 2002. Activation of glial cellsin an animal model of neuronal degeneration and its modulation bycholesterol and aging. Glia 1 (Suppl.), p. S23.

onzalo-Ruiz, A., Gonzalez, I., Sanz-Anquela, J.M., 2003. Effects ofbeta-amyloid protein on serotoninergic, noradrenergic, and cholinergicmarkers in neurons of the pontomesencephalic tegmentum in the rat. J.Chem. Neuroanat. 26, 153–169.

onzalo-Ruiz, A., Perez, J.L., Sanz, J.M., Geula, C., Arevalo, J., 2006.Effects of lipids and aging on the neurotoxicity and neuronal loss causedby intracerebral injections of the amyloid-beta peptide in the rat. Exp.Neurol. 197, 41–55.

ray, C.W., Patel, A.J., 1995. Neurodegeneration mediated by glutamateand beta-amyloid peptide: a comparison and possible interaction. BrainRes. 691, 169–179.

ajszan, T., Alreja, M., Leranth, C., 2004. Intrinsic vesicular glutamatetransporter 2-immunoreactive input to septohippocampal parvalbumin-containing neurons: novel glutamatergic local circuit cells. Hippocampus14, 499–509.

ardy, J., Selkoe, D.J., 2002. The amyloid hypothesis of Alzheimer’s dis-ease: progress and problems on the road to therapeutics. Science 297,353–356.

arkany, T., De Jong, G.I., Soos, K., Penke, B., Luiten, P.G., Gulya, K., 1995.Beta-amyloid (1–42) affects cholinergic but not parvalbumin-containingneurons in the septal complex of the rat. Brain Res. 698, 270–274.

enderson, Z., Morris, N.P., Grimwood, P., Fiddler, G., Yang, H.W.,Appenteng, K., 2001. Morphology of local axon collaterals of electro-physiologically characterised neurons in the rat medial septal/diagonalband complex. J. Comp. Neurol. 430, 410–432.

uerta, P.T., Lisman, J.E., 1993. Heightened synaptic plasticity of hippocam-pal CA1 neurons during a cholinergically induced rhythmic state. Nature364, 723–725.

ynd, M.R., Scott, H.L., Dodd, P.R., 2004. Glutamate-mediated excitotoxi-city and neurodegeneration in Alzheimer’s disease. Neurochem. Int. 45,583–595.

arvis, K., Assis-Nascimento, P., Mudd, L.M., Montague, J.R., 2007. Beta-amyloid toxicity and reversal in embryonic rat septal neurons. Neurosci.Lett. 423, 184–188.

erusalinsky, D., Kornisiuk, E., Izquierdo, I., 1997. Cholinergic neuro-transmission and synaptic plasticity concerning memory processing.Neurochem. Res. 22, 507–515.

ones, G.A., Norris, S.K., Henderson, Z., 1999. Conduction velocities andmembrane properties of different classes of rat septohippocampal neu-rons recorded in vitro. J. Physiol. 517 (Pt 3), 867–877.

ohler, C., Chan-Palay, V., Wu, J.Y., 1984. Septal neurons containing glu-tamic acid decarboxylase immunoreactivity project to the hippocampalregion in the rat brain. Anat. Embryol. (Berl.) 169, 41–44.

umamoto, E., Murata, Y., 1995. Excitatory amino acid-induced currents inrat septal cholinergic neurons in culture. Neuroscience 69, 477–493.

ee, M.G., Chrobak, J.J., Sik, A., Wiley, R.G., Buzsaki, G., 1994. Hippocam-pal theta activity following selective lesion of the septal cholinergicsystem. Neuroscience 62, 1033–1047.

eung, L.S., Shen, B., 2004. Glutamatergic synaptic transmission partici-pates in generating the hippocampal EEG. Hippocampus 14, 510–525.

eung, L.W., Lopes da Silva, F.H., Wadman, W.J., 1982. Spectralcharacteristics of the hippocampal EEG in the freely moving rat. Elec-troencephalogr. Clin. Neurophysiol. 54, 203–219.

W

Y

f Aging 31 (2010) 46–57 57

iao, M.Q., Tzeng, Y.J., Chang, L.Y., Huang, H.B., Lin, T.H., Chyan, C.L.,Chen, Y.C., 2007. The correlation between neurotoxicity, aggregativeability and secondary structure studied by sequence truncated Abetapeptides. FEBS Lett. 581, 1161–1165.

ynch, G., Rose, G., Gall, C., 1977. Anatomical and functional aspects ofthe septo-hippocampal projections. Ciba Found Symp., 5–24.

anseau, F., Danik, M., Williams, S., 2005. A functional glutamatergicneuron network in the medial septum and diagonal band area. J. Physiol.566, 865–884.

cGee, A.W., Bredt, D.S., 2003. Assembly and plasticity of the glutamater-gic postsynaptic specialization. Curr. Opin. Neurobiol. 13, 111–118.

orales, F.R., Roig, J.A., Monti, J.M., Macadar, O., Budelli, R., 1971. Septalunit activity and hippocampal EEG during the sleep-wakefulness cycleof the rat. Physiol. Behav. 6, 563–567.

akamura, S., Murayama, N., Noshita, T., Annoura, H., Ohno, T., 2001.Progressive brain dysfunction following intracerebroventricular infusionof beta(1-42)-amyloid peptide. Brain Res. 912, 128–136.

’Keefe, J., 1993. Hippocampus, theta, and spatial memory. Curr. Opin.Neurobiol. 3, 917–924.

axinos, G., Watson, C., 1998. The Rat Brain in Stereotaxic Coordinates.Elsevier.

earson, R.C., Sofroniew, M.V., Cuello, A.C., Powell, T.P., Eckenstein, F.,Esiri, M.M., Wilcock, G.K., 1983. Persistence of cholinergic neurons inthe basal nucleus in a brain with senile dementia of the Alzheimer’s typedemonstrated by immunohistochemical staining for choline acetyltrans-ferase. Brain Res. 289, 375–379.

uma, C., Bizot, J.C., 1999. Hippocampal theta rhythm in anesthetized rats:role of AMPA glutamate receptors. Neuroreport 10, 2297–2300.

uma, C., Monmaur, V., Sharif, A., Monmaur, P., 1996. Intraseptal infu-sion of selective and competitive glutamate receptor agonist NMDA andantagonist D-2-amino-5-phosphonopentanoic acid spectral implicationsfor the physostigmine-induced hippocampal theta rhythm in urethane-anesthetized rats. Exp. Brain Res. 109, 384–392.

awlins, J.N., Feldon, J., Gray, J.A., 1979. Septo-hippocampal connectionsand the hippocampal theta rhythm. Exp. Brain Res. 37, 49–63.

antos-Torres, J., Fuente, A., Criado, J.M., Riolobos, A.S., Heredia, M.,Yajeya, J., 2007. Glutamatergic synaptic depression by synthetic amyloidbeta-peptide in the medial septum. J. Neurosci. Res. 85, 634–648.

cali, C., Prosperi, C., Giovannelli, L., Bianchi, L., Pepeu, G., Casa-menti, F., 1999. Beta (1-40) amyloid peptide injection into the nucleusbasalis of rats induces microglia reaction and enhances cortical gamma-aminobutyric acid release in vivo. Brain Res. 831, 319–321.

chliebs, R., 2005. Basal forebrain cholinergic dysfunction in Alzheimer’sdisease—interrelationship with beta-amyloid, inflammation and neu-rotrophin signaling. Neurochem. Res. 30, 895–908.

elkoe, D.J., 2000. Toward a comprehensive theory for Alzheimer’s disease.Hypothesis: Alzheimer’s disease is caused by the cerebral accumulationand cytotoxicity of amyloid beta-protein. Ann. N. Y. Acad. Sci. 924,17–25.

imon, A.P., Poindessous-Jazat, F., Dutar, P., Epelbaum, J., Bassant, M.H.,2006. Firing properties of anatomically identified neurons in the medialseptum of anesthetized and unanesthetized restrained rats. J. Neurosci.26, 9038–9046.

otty, F., Danik, M., Manseau, F., Laplante, F., Quirion, R., Williams, S.,2003. Distinct electrophysiological properties of glutamatergic, cholin-ergic and GABAergic rat septohippocampal neurons: novel implicationsfor hippocampal rhythmicity. J. Physiol. 551, 927–943.

inogradova, O.S., 1995. Expression, control, and probable functional sig-nificance of the neuronal theta-rhythm. Prog. Neurobiol. 45, 523–583.

hitehouse, P.J., Price, D.L., Struble, R.G., Clark, A.W., Coyle, J.T., Delon,M.R., 1982. Alzheimer’s disease and senile dementia: loss of neurons inthe basal forebrain. Science 215, 1237–1239.

inson, J., 1978. Loss of hippocampal theta rhythm results in spatial mem-ory deficit in the rat. Science 201, 160–163.

oder, R.M., Pang, K.C., 2005. Involvement of GABAergic and cholinergicmedial septal neurons in hippocampal theta rhythm. Hippocampus 15,381–392.