Eco

Ya

b

c

h

••••

a

ARRA

KACWCHR

ApchaaA

sT

c

h0

Neuroscience Letters 568 (2014) 72–76

Contents lists available at ScienceDirect

Neuroscience Letters

jo ur nal ho me p age: www.elsev ier .com/ locate /neule t

ffects of amyloid �-peptide fragment 31–35 on the BKhannel-mediated K+ current and intracellular free Ca2+ concentrationf hippocampal CA1 neurons

u Zhanga,∗, Zhi-Gang Shia, Zhi-Hua Wangb, Jian-Guo Lia, Jin-Yuan Chenc, Ce Zhanga,∗

Department of Neurobiology, Shanxi Key Laboratory of Cell Physiology, Shanxi Medical University, Taiyuan, Shanxi 030001, PR ChinaDepartment of Pathology, Shanxi Medical University, Taiyuan, Shanxi 030001, PR ChinaFunctional Laboratory, Shanxi Medical University, Taiyuan, Shanxi 030001, PR China

i g h l i g h t s

The early-transient currents are from large conductance Ca2+-activated K+ channels.A�31–35 is the shortest active fragment of full A� sequence.A�31–35 could suppress the early-transient part of BK currents.A�31–35 elevated the intracellular load of Ca2+, as is the larger fragment A�25–35.

r t i c l e i n f o

rticle history:eceived 29 March 2013eceived in revised form 12 March 2014ccepted 19 March 2014

eywords:�31–35a2+-activated big K+ currents (BK currents)

a b s t r a c t

The present study characterizes the effects of A�31–35, a short active fragment of amyloid �-peptide (A�),upon the BK channel-mediated K+ current and intracellular free Ca2+ concentration ([Ca2+]i) of freshlydissociated pyramidal cells from rat CA1 hippocampus by using whole-cell patch-clamp recording andsingle cell Ca2+ imaging techniques. The results show that: (1) in the presence of voltage- and ATP-gated K+ channel blockers application of 5.0 �M A�31–35 significantly diminished transient outwardK+ current amplitudes at clamped voltages between 0 and 45 mV; (2) under the same conditions [Ca2+]i

was minimally affected by 5.0 �M but significantly increased by 12.5 �M and 25 �M A�31–35; and (3)

hole-cell patch recordinga2+ imagingippocampal CAl neuronsats

when 25 �M of a larger fragment of the amyloid �-peptide, A�25–35, was applied, the results weresimilar to those obtained with the same concentration of A�31–35. These results indicate that A�31–35is likely to be the shortest active fragment of the full A� sequence, and can be as effectively as the full-length A� peptide in suppressing BK-channel mediated K+ currents and significantly elevating [Ca2+]i inhippocampal CA1 neurons.

© 2014 Elsevier Ireland Ltd. All rights reserved.

lzheimer’s disease (AD) has been mainly characterized by theresence of extensive extracellular senile plaques in the cerebralortex and hippocampus, while the deposition of these plaquesas been associated with progressive deterioration of cognitive

nd mental functions. A major constituent of senile plaques ismyloid-� peptide (A�) containing 39–43 amino acid residues.lthough A� is well-known to be toxic to many types of neurons,

∗ Corresponding authors at: Department of Neurobiology, Shanxi Medical Univer-ity, 56 South Xinjian Road, Taiyuan, Shanxi 030001, PR China.el.: +86 0351 413 5560.

E-mail addresses: zhyucnm@gmail.com, zhyucnm@163.com (Y. Zhang),ezh2002@yahoo.com (C. Zhang).

ttp://dx.doi.org/10.1016/j.neulet.2014.03.028304-3940/© 2014 Elsevier Ireland Ltd. All rights reserved.

the mechanisms underlying A�-mediated neurodegeneration areas yet not well understood [22–25,29].

Our previous electrophysiological [17,19,31] and morphological[30] studies have shown that a smaller fragment of A�, A�31–35,can produce nearly all of the degenerative neurotoxic effects gen-erated by larger A� fragments [6,20,28]. Our hypothesis is thatA�31–35 is the shortest active sequence of A�, and is hence likelyto be useful for characterizing the mechanisms of neurotoxic actionby full-length A�.

Previous studies have also demonstrated that A� destabilizes

intracellular Ca2+ ([Ca2+]i) homeostasis, resulting in an elevation ofintracellular free Ca2+ concentration [8,14], while other work hasshown that the A� fragments A�1–40 and A�25–35 can increase[Ca2+]i and induce neuronal apoptosis in the rat brain tissue

ence L

[tapcrawwiacAm

hsCotTw[

rma0Os2ca5cncMttrcHtwn

ccppstfifbTtlesta(w

experimental conditions.

Y. Zhang et al. / Neurosci

6,20,28]. We have found that K+ channels [17–19], especiallyhe big conductance mediated by Ca2+-activated K+ channel (BK),s shown by single-channel recordings from excised membraneatches, are significantly affected by A�31–35 application. As BKhannels are generally considered to be essential for membraneepolarization in excitable cells and hence important for manyspects of neuronal physiology [10,26], the present experimentsere designed to explore their interaction with A�31–35 by usinghole-cell patch-clamp recording to measure pharmacologically

solated, BK channel-mediated K+ currents. We have also char-cterized the changes in [Ca2+]i evoked by A�31–35 by usingalcium fluorescence imaging techniques. The effects of the larger� fragment A�25–35 on the K+ current and [Ca2+]i were alsoonitored as a positive control.All experiments were carried out with acutely disassociated

ippocampal CA1 neurons of Sprague–Dawley rats of eitherex, aged 9–12 days old (supplied by the Research Animalenter of Shanxi, China, with approval of Shanxi Committeen Ethics of Animal Research). Animals were deeply anes-hetized with isoflurane and the brain was quickly removed.he hippocampus was carefully dissected, and CA1 neuronsere dissociated by a procedure as described by Kay and Wong

9].The following solutions were used in this study: For slice prepa-

ation, artificial cerebro-spinal fluid (ACSF) was used, containing (inM): 126 NaCl, 5 KCl, 2 MgSO4, 2 CaCl2, 10 glucose, 25 NaHCO3,

nd 1.5 NaH2PO4. For low calcium ACSF, [Ca2+] was reduced to.1 mM. ACSF was typically equilibrated with a mixture of 95%2 and 5% CO2. Solution for disassociation of neurons (incubation

olution) contained (in mM): 130 NaCl, 5.4 KCl, 1 CaCl2, 1 MgCl2,5 glucose, 10 HEPES. For whole-cell recording, recording solutionontained (in mM): 140 NaCl, 3 KCl, 1 MgCl2, 2 CaCl2, 10 HEPESnd 10 glucose, which typically includes 4-aminopyridine (4-AP,

mM) and glibenclamide (5 �M), to block the voltage-activated K+

hannels to unmask the early transient Ca2+-activated BK chan-els, and to block ATP-sensitive K+ channels, respectively. In someases, CaCl2 of recording solution was replaced by equal molargCl2 to make a calcium free recording solution for examining

he Ca2+ dependence of K+ currents. All solutions were adjustedo pH 7.4 and osmolarity 300 ± 5 mOsm. Recording pipettes hadesistances of 4–8 M� after being filled with internal solutionontaining (in mM): 140 KCl, 10 NaCl, 2 MgCl2, 0.1 CaCl2, 10EPES, 1.1 EGTA, 5 Na2-ATP, 0.5 Na2-GTP and 15 phosphocrea-

ine, with pH 7.4 and osmolarity 285 ± 5 mOsm. All experimentsere conducted at room temperature (22–25 ◦C) unless otherwiseoticed.

Hippocampal CA1 neuronal suspension was added to a glassoverslip to allow the cells attach on the glass. Coverslip withells were then placed in a recording chamber and continuouslyerfused with recording solution. Whole-cell recordings [6] wereerformed under visual control, and current responses to voltageteps were recorded under voltage-clamp mode, with a hold poten-ial of −70 mV. BK currents were elicited by a protocol modifiedrom others. [3,28]: cells were held at −70 mV and hyperpolar-zed to −110 mV for 1 s, then depolarized to potential rangingrom −60 to +45 mV with 15 mV increments for 1.8 s, followedy repolarization back to holding potential (Fig. 1Bd). TypicallyTX was not included in recording solution because of its celloxicity and the minimal effects of fast inward Na+ current onong-lasting outward K+ currents, as observed in our preliminaryxperiments (not shown). We used a multichannel fast perfusionystem for drug delivery, routinely added to the recording solu-ion. The end opening of perfusion tube to the recorded cell was

bout 100 �m. Current signals were recorded with an amplifierMulticlamp 700B, Molecular Devices, USA), and P-clamp 10 soft-are was used for data acquisition and analysis. Electric signals

etters 568 (2014) 72–76 73

were digitized and recorded to a computer hard drive for offlineanalysis.

[Ca2+]i was measured at room temperature, using protocolsmodified from others [5,13]. Briefly, neurons settled down on glasscoverslip were washed twice with incubation solution and incu-bated at 34 ◦C for 15 min in the same solution with the addition of5 �M Fura-2/AM, a membrane permeable, high affinity, intracellu-lar ratiometric calcium indicator dye that is excited by UV light. Toremove extracellular Fura-2/AM, the loaded cells were rinsed threetimes with incubation solution. Then neurons were incubated for30 min at room temperature to allow de-esterification of the Fura-2 dye. Ca2+ transients in single cells were imaged with an invertedepifluorescence microscope (Zeiss Axiovert S100). A 12-bit slowscan cooled charge-coupled digital (CCD) camera system (Photon-ics, Germany) was used to acquire digitized images of Fura-2 asdescribed elsewhere [2,15,16]; namely, excitation wavelengths 340and 380 nm were given alternatively and emission at 510 nm wasacquired from the Fura-2 loaded neurons at a rate of 1 pair every10 s. The fluorescence ratio was obtained by dividing the fluores-cence excited at 340 nm by that excited at 380 nm (F 340/380).Data captured from each cell were analyzed off-line using VisionSoftware (TILL Photonics, Germany) independently for each timepoint in the captured sequence. For individual cells, [Ca2+]i at eachtime point was calculated from the values of fluorescence ratios asdescribed previously [5]. These two parameters, the fluorescenceratios and [Ca2+]i, are linearly correlated with each other and canbe automatically computed by our imaging systems before, duringand after application of test agents for offline analysis.

Test agents 4-AP, TEA, glibenclamide, Na2-ATP, Na2-GTP, phos-phocreatine, A�25–35 (GSNKGAIIGLM) and Fura-2 fluorescencewere purchased from Sigma (USA); A�31–35 (IIGLM) was fromSangon Instruments (Shanghai, China). All agents were stored at−20 ◦C until use. All test agents were bath applied at concentrationsas indicated.

All electrophysiological recording data were expressed asmean ± SEM. Statistical analysis was performed by Independent-Sample t-test and one-way ANOVA. Statistical analysis of the dataof calcium imaging was performed by Paired t-test. P < 0.05 wasconsidered statistically significant.

Freshly disassociated CA1 cells can be easily identified underour visualized patch recording set up. BK-channel mediated K+

current was isolated and elicited by a specific protocol (Fig. 1Bd)with recording solution including 4-AP (5 mM) to block transientvoltage-activated K+ currents, and glibenclamide (5 �M) to blockATP-sensitive K+ currents, as described earlier. Current signals wereroutinely recorded in such fashion unless otherwise noticed.

Under these conditions, evoked currents response typically con-sist of a brief inward current (1–2 ms) and a long-lasting outwardcurrent following the voltage steps. The brief inward current wassensitive to voltage-gated Na+ channel blocker TTX (not shown).The outward current can be further divided into a larger earlytransient component and a smaller delayed steady-state compo-nent (Fig. 1Aa). According to Sun et al. [27], the outward currents,particularly the early transient component is primarily mediatedby BK channels. In our experiments, the outward, but not the briefinward current, can be reversibly attenuated by Ca2+-free recordingsolution, especially the early transient component (Fig. 1Ab and c),indicating their Ca2+ dependence. Moreover, when TEA (20 mM),an non-selective K+ channel blocker [4], was bath applied, the out-ward current nearly disappears (Fig. 1Ad). These results indicatethat voltage-dependent Ca2+-sensitive K+ current (BK current) canbe reliably recorded in our disassociated CA1 neurons under our

We then examined the effects of short A� fragments on BKchannel mediated K+ currents. As shown in Fig. 1B, bath appli-cation of A�31–35 (5 �M) clearly reduced the outward currents,

74 Y. Zhang et al. / Neuroscience Letters 568 (2014) 72–76

Fig. 1. Recordings of outward K+ current recorded in acutely disassociated hippocampal CA1 neurons under voltage-clamp mode, while cells were perfused with recordingsolution containing 4-aminopyridine (4-AP, 5 mM) and glibenclamide (5 �M), to block voltage-activated K+ channels to unmask the early transient Ca2+-activated BK channels,and to block ATP-sensitive K+ channels, respectively. (A) Isolation of BK-channel mediated K+ currents evoked by voltage steps shown in Bd. (a) Control; (b) responses underC M) ina e outwc comes

prSow22tFnv

aweari(ptr23AoStiii

cc5csvd

a2+-free condition; (c) responses 3 min after washout; (d) responses under TEA (20 mddition of 5 �M A�31–35 to the bath; (c) 3 min after washout; (d) protocol to evokontrol conditions and after addition of 5 �M A�31–35. Note that the difference be

articularly the early transient component (Fig. 1Bb), and theeduction was largely recovered 3 min after washout (Fig. 1Bc).imilar tests results were observed in 8 cells. The peak valuesf earlier transient currents evoked by voltage steps to +45 mVere remarkably reduced, from 3096 ± 257.4 pA (control) to

443 ± 279.0 (A�31–35 pA bath application), a reduction of0.8% ± 9.7% (n = 8; P < 0.001). Results from these cells are pooledogether and plotted as a function of voltage steps applied inig. 1C, indicating that A�31–35 at concentration of 5 �M sig-ificantly suppresses the BK channel mediated K+ current in aoltage-dependent fashion.

We next measured intracellular Ca2+ transients in disassoci-ted CA1 cells. As described earlier, individual cells were preloadedith the fluorescent Ca2+ indicator Fura-2. Single cells of inter-

sts were excited alternately by UV light at wavelengths of 340 nmnd 380 nm, and images of emission at 510 nm were acquired andecorded by a computer. We found that intracellular fluorescencentensity was not affected by bath application of A�31–35 at 5 �MFig. 2A, top row). However, when the concentration of the sameeptide was increased to 12.5 �M, the fluorescence intensity of theested cell was moderately, but significantly enhanced (Fig. 2A, 2ndows). When the concentration of the peptide was increased to5 �M, the fluorescence intensity was robustly increased (Fig. 2A,rd rows). We then treated the cells with the larger fragment�25–35 at 25 �M, it was found that the fluorescence intensityf tested cells were remarkably enhanced (Fig. 2A, bottom row).imilar tests at each dosage were repeated in at least 5 cells, andhe results are summarized as a bar graph in Fig. 2B. These resultsndicate that the short A� peptide A�31–35 at a high dose (25 �M)s as effective as its longer precursor A�25–35 for elevating [Ca2+]in freshly isolated CA1 cells.

Finally we characterized the [Ca2+]i dynamics in single CA1ells under treatments of short and long A� peptides at the con-entrations tested above. As shown in Fig. 2C, 12.5 �M, but not

�M A�31–35 induced a clearly increase in intracellular fluores-

ent intensity; at 25 �M, both A�31–35 and A�25–35 inducedimilar kinetics of increase in fluorescence intensities. The ele-ation of the fluorescence intensity typically occurred with aelay of 3–5 min and lasted for the length of the recording.

the bath. (B) Effects of 5 �M A�31–35 on K+ current. (a) Control; (b) responses withard K+ currents. (C) I–V plot, showing average transient K+ current intensity under

statistically significant for voltages >0 mV (n = 8, *P < 0.05, **P < 0.01, ***P < 0.001).

Quantitatively, fluorescence intensities were significantly elevatedby both peptides at concentration of 25 �M: from 823.58 ± 7.78to 1195.4 ± 51.51 for A�31–35, increased by 32.7% ± 5.4% (n = 5,P < 0.01), and from 862.21 ± 13.10 to 1213.20 ± 47.88 for A�25–35,increased by 34.1% ± 5.4% (n = 5, P < 0.01). These results indicatethat at 25 �M level, both forms of A� peptides significantly ele-vate intracellular fluorescence intensities, thereby [Ca2+]i in freshlydisassociated CA1 cells.

It is generally accepted that K+ channels, especially largeconductance (100–300 pS) mediated by Ca2+-activated K+ (BK)channels, play a pivotal role in setting the resting membranepotential, shaping action potentials, maintaining fast after hyper-polarization, and timing interspike intervals during repetitivefiring, thereby preventing the potentially toxic effects of excessiveneuronal excitability [4,10,11,27].

In the first of these experiments, we used whole-cell patch-clamping techniques to record outward currents from freshlydissociated hippocampal CA1 pyramidal cells in the presence of4-AP and glibenclamide to block voltage- and ATP-gated K+ chan-nels. Previous investigations have shown that the early transientphase and a component of the delayed sustained phase of suchcurrents are mediated by BK channels [4,7,11,27]. We chose thepeak of the transient phase evoked by voltage steps of increasingdepolarization as the parameter for measuring the effect of the A�fragments A�31–35 and A�25–35 on BK channel activity. Applica-tion of 5 �M A�31–35, an effective dose used frequently in this lab[17,19], significantly suppressed the transient phase of the outwardcurrents by 20.8% ± 9.7%. This is consistent with the results of a pre-vious investigation [17], in which unitary BK channel currents weremeasured in inside-out membrane patches from rat hippocampalCA1 neurons. In that study, application of 5 �M A�31–35 had noeffect on the mean current amplitude of single BK channels butsignificantly decreased the channel’s mean open probability (Po)and open frequency, as well its mean open time [17]. Moreover,our previous experiments [17–19] have showed that application of

A�25–35, a larger fragment of A�, also significantly reduced themean Po of BK channel and its mean open time within 1–3 min.These results suggest that A�31–35 (or A�25–35) functionally sup-presses BK channels by altering single channel gating properties. Of

Y. Zhang et al. / Neuroscience Letters 568 (2014) 72–76 75

Fig. 2. Ca2+ imaging of freshly isolated hippocampal CA1 neurons. Cells were pre-loaded with Fura-2/AM Ca2+ indicator and test agents were bath applied. (A) Examples ofc depenH owing* same

cotb

s5oewA

tedecamhoBgatii

ells, showing that A�31–35 enhances Ca2+ related fluorescent intensity in a dose

istogram of maximum F340/F380 fluorescence ratios from at least 20 neurons, shP < 0.05, **P < 0.01). (C) Time series of F340/F380 for control conditions (a) and the

ourse, we cannot exclude the possibility that A�31–35, A�25–35,r the full length A� molecule may suppress other components ofhe recorded outward current, although this possibility remains toe explored further.

Secondly, we observed changes in single cell fluorescence inten-ity or [Ca2+]i following A�31–35 application. The presence of

�M A�31–35 had no visible effect, while higher concentrationsf the peptide (12.5 and 25 �M) elicited significant and prolongedlevations in [Ca2+]i in a dose-dependent manner. These effectsere also duplicated by the application of the larger A� fragment,�25–35.

Mattson et al. [12] and Barger et al. [1] have hypothesized thathe full length �-amyloid peptide or its shorter fragments maynhance glutamate neurotoxicity in cortical cell cultures and ren-er these neurons more vulnerable to environmental stimuli thatlevate [Ca2+]i levels. Sun et al. [27] noted that a fast transient BKurrent is activated by Ca2+ entry through high-threshold, voltage-ctivated L- and N-type Ca2+ channels, although this influx aloneay not be identical to the prolonged intracellular Ca2+ loading we

ave observed following A�31–35 or A�25–35 application. More-ver, although 5 �M A�31–35 could primarily suppress transientK currents, only larger dosages significantly elevated [Ca2+]i, sug-esting that A� or its fragments interfere with BK channel gating

nd increase [Ca2+]i through different mechanisms. It is possiblehat the A� peptide-mediated block of BK current (perhaps bynterfering with the opening of these channels by interfering withnternal Ca2+ binding) may produce an increase in cell excitability,

dent manner (top three rows), and A�25–35 has similar effects (bottom row). (B) that intracellular free Ca2+ levels increase with either type of A� fragment (n = 5,

concentrations of A�31–35 (b–d) and A�25–35 (e) as in A and B.

leading to enhanced Ca2+ influx and its intracellular accumula-tion. Another possibility is that these peptides are incorporatedinto the lipid bilayer of the plasma membrane and form new Ca2+-permeable channels. At least two recent reports support the latterhypothesis. Rhee et al. [21] provides biochemical and structuralevidence that the A� fragment A�1–42 forms calcium-permeablechannels and thus may induce cellular toxicity in Alzheimer’s dis-ease by interfering with calcium homeostasis. Also, in our lab Qi andQiao [18] have shown that A�31–35 forms cation-selective, Zn2+-and Cd2+-sensitive channels in membrane patches excised from rathippocampal neurons. Further investigation at the molecular levelwill be necessary to resolve these issues.

Acknowledgements

We thank Prof. Jian-Tian Qiao for his useful discussions andmany helpful suggestions on the project. We also thank Dr. GerhardMagnus for his editing help on the final version of the manuscript.This work was supported by Scientific and Technological Innova-tion Programs of Higher Education Institutions in Shanxi (STIP), (No.20081011), China.

References

[1] S.W. Barger, V.L. Smith-Swintosky, R.E. Rydel, M.P. Mattson, Beta-amyloid pre-cursor protein mismetabolism and loss of calcium homeostasis in Alzheimer’sdisease, Ann. N. Y. Acad. Sci. 695 (1993) 158–164.

7 ence L

[

[

[

[

[

[

[

[

[

[

[

[

[[

[

[

[

[

[

[

[

6 Y. Zhang et al. / Neurosci

[2] J.A. Connor, Digital imaging of free calcium changes and of spatial gradientsin growing processes in single, mammalian central nervous system cells, Proc.Natl. Acad. Sci. U.S.A. 83 (1986) 6179–6183.

[3] S.P. Fraser, Y.H. Suh, M.B. Djamgoz, Ionic effects of the Alzheimer’s disease beta-amyloid precursor protein and its metabolic fragments, Trends Neurosci. 20(1997) 67–72.

[4] V.K. Gribkoff, J.E. Starrett Jr., S.I. Dworetzky, Maxi-K potassium channels: form,function, and modulation of a class of endogenous regulators of intracellularcalcium, Neuroscientist 7 (2001) 166–177.

[5] G. Grynkiewicz, M. Poenie, R.Y. Tsien, A new generation of Ca2+ indicatorswith greatly improved fluorescence properties, J. Biol. Chem. 260 (1985)3440–3450.

[6] L.M. He, L.Y. Chen, X.L. Lou, A.L. Qu, Z. Zhou, T. Xu, Evaluation of beta-amyloidpeptide 25–35 on calcium homeostasis in cultured rat dorsal root ganglionneurons, Brain Res. 939 (2002) 65–75.

[7] G.A. Hicks, N.V. Marrion, Ca2+-dependent inactivation of large conductanceCa2+-activated K+ (BK) channels in rat hippocampal neurones produced bypore block from an associated particle, J. Physiol. (Lond.) 508 (Pt 3) (1998)721–734.

[8] B.L. Kagan, Y. Hirakura, R. Azimov, R. Azimova, M.C. Lin, The channel hypothesisof Alzheimer’s disease: current status, Peptides 23 (2002) 1311–1315.

[9] A.R. Kay, R.K. Wong, Isolation of neurons suitable for patch-clamping from adultmammalian central nervous systems, J. Neurosci. Methods 16 (1986) 227–238.

10] R. Latorre, A. Oberhauser, P. Labarca, O. Alvarez, Varieties of calcium-activatedpotassium channels, Annu. Rev. Physiol. 51 (1989) 385–399.

11] U.S. Lee, J. Cui, BK channel activation: structural and functional insights, TrendsNeurosci. 33 (2010) 415–423.

12] M.P. Mattson, B. Cheng, D. Davis, K. Bryant, I. Lieberburg, R.E. Rydel, Beta-amyloid peptides destabilize calcium homeostasis and render human corticalneurons vulnerable to excitotoxicity, J. Neurosci. 12 (1992) 376–389.

13] O. Meucci, R.J. Miller, gp120-induced neurotoxicity in hippocampal pyrami-dal neuron cultures: protective action of TGF-beta1, J. Neurosci. 16 (1996)4080–4088.

14] H.S. Mogensen, D.M. Beatty, S.J. Morris, O.S. Jorgensen, Amyloid beta-peptide(25–35) changes [Ca2+] in hippocampal neurons, Neuroreport 9 (1998)1553–1558.

15] W. Muller, J.A. Connor, Cholinergic input uncouples Ca2+ changes from K+ con-ductance activation and amplifies intradendritic Ca2+ changes in hippocampalneurons, Neuron 6 (1991) 901–905.

16] W. Muller, J.A. Connor, Dendritic spines as individual neuronal compartmentsfor synaptic Ca2+ responses, Nature 354 (1991) 73–76.

[

etters 568 (2014) 72–76

17] J.S. Qi, J.T. Qiao, Amyloid beta-protein fragment 31–35 forms ion channels inmembrane patches excised from rat hippocampal neurons, Neuroscience 105(2001) 845–852.

18] J.S. Qi, J.T. Qiao, Suppression of large conductance Ca2+-activated K+ channelsby amyloid beta-protein fragment 31–35 in membrane patches excised fromhippocampal neurons, Sheng Li Xue Bao 53 (2001) 198–204.

19] J.S. Qi, L. Ye, J.T. Qiao, Amyloid beta-protein fragment 31–35 suppresses delayedrectifying potassium channels in membrane patches excised from hippocampalneurons in rats, Synapse 51 (2004) 165–172.

20] M. Ramsden, Z. Henderson, H.A. Pearson, Modulation of Ca2+ channel currentsin primary cultures of rat cortical neurones by amyloid beta protein (1–40) isdependent on solubility status, Brain Res. 956 (2002) 254–261.

21] S.K. Rhee, A.P. Quist, R. Lal, Amyloid beta protein-(1–42) forms calcium-permeable, Zn2+-sensitive channel, J. Biol. Chem. 273 (1998) 13379–13382.

22] D.J. Selkoe, Alzheimer’s disease, Cold Spring Harb. Perspect. Biol. 3 (2011) 1–6.23] D.J. Selkoe, The molecular pathology of Alzheimer’s disease, Neuron 6 (1991)

487–498.24] D.J. Selkoe, D. Schenk, Alzheimer’s disease: molecular understanding pre-

dicts amyloid-based therapeutics, Annu. Rev. Pharmacol. Toxicol. 43 (2003)545–584.

25] R.A. Sperling, B.C. Dickerson, M. Pihlajamaki, P. Vannini, P.S. LaViolette, O.V.Vitolo, T. Hedden, J.A. Becker, D.M. Rentz, D.J. Selkoe, K.A. Johnson, Functionalalterations in memory networks in early Alzheimer’s disease, NeuromolecularMed. 12 (2010) 27–43.

26] J.F. Storm, Potassium currents in hippocampal pyramidal cells, Prog. Brain Res.83 (1990) 161–187.

27] X. Sun, X.Q. Gu, G.G. Haddad, Calcium influx via L- and N-type calcium chan-nels activates a transient large-conductance Ca2+-activated K+ current in mouseneocortical pyramidal neurons, J. Neurosci. 23 (2003) 3639–3648.

28] T. Yagami, K. Ueda, K. Asakura, T. Kuroda, S. Hata, T. Sakaeda, Y. Kambayashi,M. Fujimoto, Effects of endothelin B receptor agonists on amyloid beta protein(25–35)-induced neuronal cell death, Brain Res. 948 (2002) 72–81.

29] K. Yamada, T. Nabeshima, Animal models of Alzheimer’s disease and evaluationof anti-dementia drugs, Pharm. Ther. 88 (2000) 93–113.

30] X.Z. Yan, J.T. Qiao, Y. Dou, Z.D. Qiao, Beta-amyloid peptide fragment 31–35induces apoptosis in cultured cortical neurons, Neuroscience 92 (1999)

177–184.

31] L. Ye, J.T. Qiao, Suppressive action produced by beta-amyloid peptide fragment31–35 on long-term potentiation in rat hippocampus is N-methyl-d-aspartatereceptor-independent: it’s offset by (−)huperzine A, Neurosci. Lett. 275 (1999)187–190.