Loss of muscarinic M4 receptors in hippocampus of Alzheimer patients

Brain Research 960 (2003) 259–262 www.elsevier.com / locate / brainres

Short communication

Loss of muscarinic M 4 receptors in hippocampus of Alzheimer patients Ezra Mulugeta a , Evert Karlsson a , Atiqul Islam a , Raj Kalaria b , Halinder Mangat a , Bengt Winblad a , Abdu Adem a,c , * a

Department of Clinical Neuroscience ( NEUROTEC), Section of Experimental Geriatrics, Karolinska Institute, Novum, 141 86 Huddinge, Sweden b Wolfson Research Centre, Institute for Health of the Elderly, Newcastle General Hospital, Westgate Road, Newcastle upon Tyne, UK c Department of Pharmacology, Faculty of Medicine and Health Sciences, University of United Arab Emirates, Al Ain, United Arab Emirates Accepted 28 August 2002

Abstract We assessed muscarinic M 1 , M 2 and M 4 receptor subtypes in the hippocampus of Alzheimer’s and control brains by receptor autoradiography using ligands such as [ 125 I]muscarinic toxin-1 ([ 125 I]MT-1, M 1 selective), [ 3 H]AFDX-384 (M 2 partially selective) and [ 125 I]muscarinic toxin 4 ([ 125 I]M 4 toxin-1, M 4 selective). Our results revealed a significant decrease in muscarinic M 4 receptor binding in the dentate gyrus and CA4 regions of brain sections from Alzheimer’s patients compared to controls. No changes in the density of M 1 or M 2 receptor binding were observed. Our findings suggest that, relative to other muscarinic receptor subtypes, the M 4 receptor could be the subtype which is selectively compromised in Alzeheimer’s disease (AD).  2002 Published by Elsevier Science B.V. Keywords: Alzheimer’s disease; Hippocampus; Muscarinic receptor; Mamba toxin

Alzheimer’s disease (AD) is a neurodegenerative disorder characterized by various mental dysfunctions including memory deficits, impaired intellectual abilities, emotional instability and changes in personality. Neuropathological characteristics include the presence of senile plaques and neurofibrillary tangles [4]. The symptoms of AD are highly correlated with the loss of cholinergic functions in cortical and hippocampal brain areas [8] and a cholinergic hypothesis for AD is widely accepted. Several reports have described decreases in several cholinergic parameters, such as changes in choline acetyltransferase and acetylcholinesterase activity, decreased acetylcholine levels, and changes in nicotinic and muscarinic receptors levels in specific brain areas [14,15]. Muscarinic acetylcholine receptors (mAChR) play a central role in the regulation of memory, learning and cognitive processes. A number of mAChR ligands such as muscarine and scopolamine from natural sources and synthetic compounds such as pirenzepine and AFDX-384 have been *Corresponding author. Tel.: 146-8-585-83885; fax: 146-8-58583880. E-mail address: [email protected] (A. Adem).

characterized [7]. The use of these ligands, although not very selective, has provided valuable information on mAChR localization, distribution and pharmacology [9,10,14,15]. However, the lack of high affinity ligands with a high degree of selectivity has resulted in conflicting reports on mAChR changes in AD. During the last 15 years, new very selective ligands have been isolated from mamba (Dendroaspis angusticeps) venom [2,1]. Three of these toxins are selective for M 1 , M 2 and M 4 receptors [1,6,12,13]. Others such as brucine and its analogues [5] are selective for the M 3 receptor and monoclonal antibodies have been developed against all five subtypes [10,17]. In the present study, mAChR changes were investigated in postmortem hippocampal brain sections from AD patients and controls using [ 125 I]MT-1 (M 1 selective), [ 125 I]M 4 toxin-1 (M 4 selective) and [ 3 H]AFDX384 (M 2 partial selective). Brains (seven Alzheimer and four controls) were obtained from the brain bank at the Department of Neuropathology, Wolfson Research Centre, Newcastle, UK. The age, gender and postmortem delay of the subjects is shown in Table 1. The diagnosis of AD was confirmed by histopathological examination. The presence of neuritic

0006-8993 / 02 / $ – see front matter  2002 Published by Elsevier Science B.V. PII: S0006-8993( 02 )03542-4

E. Mulugeta et al. / Brain Research 960 (2003) 259–262

260 Table 1 Control and Alzheimer subjects Groups

Control Control Control Control

1 2 3 4

Sex

Age (years)

Post mortem delay (h)

F M M M

70 79 72 76

6 5 8 6

74.363.3

6.360.9

60 87 78 78 78 71 62

4 6 4 7 6 3 6

73.468.1

5.161.3

Mean6S.E.M. AD AD AD AD AD AD AD

1 2 3 4 5 6 7

Mean6S.E.M.

F M M F M F M

Relative degree of histopathological findings (CERAD criteria)

11 11 111 111 111 111 111

F, female; M, male; 1 to 111 relative degree of pathological changes.

Fig. 1. Levels of M 4 receptors in dentate gyrus (DG) and CA4 regions of hippocampus from Alzheimer’s patients (n57) and controls (n54). Values are mean6S.E.M. Significant decrease was seen in dentate gyrus (*P,0.05) and in CA4 (*P,0.05).

plaques in the cortex and neurofibrillary tangles in hippocampus of AD brains met the criteria set by the consortium to establish a registry for AD (CERAD, 1992). None of the patients had been treated with acetylcholinesterase inhibitors or other drugs that affect the cholinergic system (e.g., anti-cholinergics). Controls had no evidence of dementia and were also without neurological or psychiatric disorders. The majority of controls and AD had died from bronchopneumonia. The muscarinic toxin MT-1 and M 4 toxin-1 were isolated essentially as described previously [1,2,12] and iodinated according to the chloramine-T method. Receptor autoradiography was performed as described previously [3]. Briefly, 10-mm slide-mounted sections were incubated with 10 nM [ 125 I]MT-1, M 1 selective (specific activity 37.6 Ci / mmol, Ki 522–49 nM), or with 2 nM [ 125 I]M 4 toxin-1, M 4 selective (specific activity 21.8 Ci / mmol, 1.4–2 nM) in 50 mM Na,K phosphate buffer containing 1% BSA. Since we had no toxin for M 2 receptors, sections were incubated with low concentration (2.4 nM) [ 3 H]AFDX-384 (NEN specific activity 120 Ci / mmol, Ki 56.0 nM for M 2 and 10 nM for M 4 ) in 50 mM Na,K phosphate buffer, pH 7.4. At this concentration, [ 3 H]AF-DX-384 displays a 1.5-fold selectivity for M 2 versus M 4 receptors. After incubation the sections were exposed to Hyperfilm (Amersham, UK) for 7 days along with calibrated radioac-

tive standards ( 125 I-Microscales, Amersham). Specific binding was obtained by subtracting binding in the presence of 10 24 M atropine (nonspecific) from the binding in the absence of atropine (total binding). Statistical analyses were performed according to Fisher’s least significant difference method for multiple comparisons after a oneway analysis of variance (ANOVA). Statistical significance was set at P,0. 05. Binding to specific muscarinic M 1 , M 2 and M 4 receptors was measured in several hippocampal regions and the results are shown in Table 2. Fig. 1 shows a significant decrease in [ 125 I]M 4 toxin-1 binding in the dentate gyrus (P,0.05) by 32614% and to CA4 (P,0.05) by 31616% in AD patients compared to controls. No difference in M 4 receptor binding was observed in CA3 and CA1 regions of AD hippocampi compared to controls. Furthermore, no significant differences in the level of M 1 and M 2 receptors between AD and control brains were observed in the hippocampal regions (Table 2). The main finding of this study was a significant decrease in M 4 receptor binding in the dentate gyrus and CA4

Table 2 Binding of ligands to hippocampus from Alzheimer patients (n57) and age matched controls (n54) Ligand

[ 125 I]MT-1 [ 125 I]M 4 -toxin 1 [ 3 H]AFDX-384

Dentate

CA1

CA3

CA4

Control

AD

Control

AD

Control

AD

Control

AD

14864 13767 141614

162612 93613* 136627

8266 3563 191615

8266 3767 157625

10462 7566 103614

138617 6669 100626

12065 9666 113616

11865 66610* 108629

Values are mean6S.E.M. *P,0.05.

E. Mulugeta et al. / Brain Research 960 (2003) 259–262

regions of the hippocampus of AD patients compared to controls. Similar results were obtained in a preliminary report [1]. However, in that report [1], the binding of the M 4 -toxin to dentate gyrus was only about one-third of that obtained in the present study. These differences in binding are due to differences in the incubation concentration of the toxin (0.8 nM in the previous report versus 2 nM in the present study). Our findings are in contrast to those obtained by immunoprecipitation using monoclonal antibodies in which a 25% increase in M 4 receptors was reported [10,14]. Flynn et al. [10] determined by immunoprecipitation the content of the mAChR in various parts of human brain. We estimated total receptor content from a histogram reported in Flynn et al. [10]. The amount of total precipitated receptor was regarded as 100%, although it was slightly higher than the total content based on [ 3 H]Nmethyl scopolamine binding. In hippocampus M 1 , M 2 , M 3 and M 4 accounted for 59, 19, 4 and 18%, respectively, of the total receptor content. M 5 receptors were not detected. The relative amounts are 3.3, 1.1, 0.2 and 1.0 (M 4 taken as 1). The relative 1 /Ki :s for M 4 toxin is 1 / 40 (alternatively 1 / 200) for M 1 and 1 for M 4. In control samples, the relative bindings 3.331 / 4050.08 and 1. Thus, the binding to M 4 51 /(0.0811)593% of the total. In dentate gyrus and CA4 of AD brains the M 4 receptors were reduced by about 30% and the relative amounts are 3.3 and 0.7, respectively. The relative bindings are 0.08 to M1 and 0.7 to M 4 and 90% binds to M 4 . In the control dentate gyrus, the binding of [ 125 I]M 4 toxin-1 was 137 fmol / mg wet tissue (Table 2) and 0.9331375127 fmol was to M 4 receptors. The binding to M 4 in the AD samples was 0.90393584 fmol and the decrease in M 4 receptors5 (129–84) / 129535%. With relative 1 /Ki :s51 / 200 (0.005) and 1 (Ki :s51.4 and 300 nM) the binding to M 4 receptors in AD dentate gyrus50.7 /(0.710.005)599%. Thus, the decrease in the binding of [ 125 I]M 4 -toxin-1 gives a good estimate of the reduction in M 4 receptors. Our results showed no change in the levels of muscarinic M 1 receptors in the hippocampus of AD patients compared to controls indicating that hippocampal M 1 receptors are not affected in AD. These results are in agreement with previous findings which also reported no changes in M 1 receptors in AD [10,14]. However, our results differ from those findings obtained by immunoprecipitation using monoclonal antibodies [10,14]. In the latter studies of whole hippocampus M 1 and M 2 receptors decreased by about 50% [10]. A possible explanation for the decrease of M 1 receptors could be that this receptor have lost epitopes recognized by antibodies but retain full ligand binding [17]. In contrast to previous reports, which showed decreases in M 2 receptors [10,14], we found no significant differences in the level of M 2 receptor binding between control and AD brains. The relative 1 /Ki :s of AF-DX 384 for M 1 to M 5 are 0.3, 1.7, 0.2, 1 and 0.01, respectively (M 4 51) and the relative binding 0.99, 1.87, 0.04, 1.00 and 0.00

261

(sum53.90). Binding to the four subtypes M 1 to M 4 accounts for 25, 48, 1 and 26%, respectively of the total. Thus, only about 50% of the AF-DX 384 was bound to M 2 indicating that AF-DX 384 even at such a low concentration is not a selective ligand to assess M 2 receptor changes. Hippocampal function is modulated in two different ways via mAChR. One way involves a presynaptic action which inhibits the release of various transmitter substances such as glutamate, aspartate, g-aminobutyric acid and acetylcholine itself [14,16,17]. The second way is via postsynaptic excitation of central synapses (Ref. [14] and Refs. therein). Although the functional role of the M 4 receptor is not yet clear, it has been suggested that one effect of acetylcholine on this receptor subtype is to modulate the release of neurotransmitters [16]. Consistent with this function is the demonstration of enhanced glycine release from human cerebral synaptosomes via the M 4 receptor subtype [16]. Glycine is required for the activation of the N-methyl-D-aspartate glutamate receptor. Thus, the ACh-evoked release of glycine may represent a link between cholinergic and glutamatergic transmission, two systems powerfully involved in cognitive processes [16]. Injection of the M 4 toxin-1 into hippocampus of rats causes amnesia. If M 4 receptors also have a similar role in humans, the decrease in M 4 receptors may be responsible for memory deficits in Alzheimer’s disease [11]. In conclusion, our finding of decreased M 4 receptor levels in the dentate gyrus and CA4 of AD subjects suggests that, relative to other subtypes, the M 4 subtype could be the cholinergic muscarinic receptor subtype which is important for cognitive function, and selectively compromised in Alzheimer’s disease.

Acknowledgements We are grateful to Dr. Adlan Elhasan and Dr. Gunnar Johansson for their helpful discussions and to Dr. Allan E. Johnson for valuable methodological discussions and language correction. These experiments were supported by grants from the European Union (QLK6-CT-1999-02112), the Swedish Medical Council and Alzheimerfonden to A. Adem and E. Karlsson. Preliminary results of these studies were reported previously [1].

References [1] A. Adem, E. Karlsson, Muscarinic receptor subtype selective toxins, Life Sci. 60 (1997) 1069–1076. ˚ [2] A. Adem, A. Asblom, G. Johansson, P.M. Mbugua, E. Karlsson, Toxins from the venom of the green mamba Dendroaspis angusticeps inhibit the binding of quinuclidinyl benzilate to muscarinic acetylcholine receptors, Biochim. Biophys. Acta 968 (1988) 340– 345. [3] A. Adem, M. Jolkkonen, N. Bogdanovic, A. Islam, E. Karlsson,

262

[4]

[5]

[6] [7]

[8]

[9]

[10]

E. Mulugeta et al. / Brain Research 960 (2003) 259–262 Localization of M1 muscarinic receptors in rat brain by using selective muscarinic toxin-1, Brain Res. Bull. 44 (1997) 597–601. R.T. Bierer, D.P. Hof, D.P. Purohit, L. Carlin, J. Schmeidler, K.L. Davis, D.P. Perl, Neocortical neurofibrillary tangles correlate with dementia severity in Alzheimer’s disease, Arch. Neurol. 52 (1992) 82–88. N.J.M. Birdsall, T. Parries, P. Gharagozlo, S. Kobayashi, S. Lazareno, Subtype selective positive cooperative action between brucine analogs and acetylcholine at muscarinic receptors: functional studies, Mol. Pharmacol. 55 (1999) 778–786. J.M. Carsi, H.H. Valentine, L.T. Potter, m2-toxin: a selective ligand for m2 muscarinic receptors, Mol. Pharmacol. 56 (1999) 933–937. M.P. Caulfield, N.J.M. Birdsall, International union of pharmacology. XVII. Classification of muscarinic acetylcholine receptors, Phrmacol. Rev. 50 (2) (1998) 279–290. J.T. Coyle, D.L. Price, M.R. DeLong, Alzheimer’s disease: a disorder of cortical cholinergic innervation, Science 219 (1983) 1184–1190. ¨ F. Dorje, J. Wess, G. Lambrecht, R. Tacke, E. Mutschler, M.J. Brann, Antagonist binding profiles of five cloned human muscarinic receptor subtypes, J. Pharmacol. Exp. Ther. 256 (1991) 727–733. D. Flynn, G. Ferrari-DiLeo, D.C. Mash, A.L. Levey, Differential regulation of molecular subtypes of muscarinic acetylcholine receptors in Alzheimer’s disease, J. Neurochem. 64 (1995) 1888– 1891.

[11] D. Jerusalinsky, E. Kornisiuk, P. Alfaro, J. Quillfeldt, M. Alonso, E. ˜ Rial Verde, C. Cervenansky, A.L. Harvey, Muscarinic toxin selective for m4 receptors impairs memory in rats, Neuroreport 9 (1998) 1401–1411. [12] M. Jolkkonen, P.L. van Giersbergen, U. Hellman, C. Wernstedt, E. Karlsson, A toxin from the green mamba Dendroaspis angusticeps: amino acid sequence and selectivity for muscarinic m4 receptors, FEBS Lett. 352 (1994) 91–94. [13] E. Karlsson, M. Jolkkonen, E. Mulugeta, P. Onali, A. Adem, Snake toxins with high selectivity for subtypes of muscarinic acetylcholine receptors, Biochimie 82 (2000) 793–806. [14] A.L. Levey, Muscarinic acetylcholine receptor expression in memory circuits: implications for treatment of Alzheimer disease, Proc. Natl. Acad. Sci. USA 93 (1996) 13541–13546. ´ [15] R. Rodrıguez-Puertas, J. Pascual, T. Vilaro, P. Angel, Autoradiographic distribution of M1, M2, M3, and M4 muscarinic receptor subtypes in Alzheimer’s disease, Synapse 26 (1997) 341–350. [16] C. Russo, M. Marchi, G.C. Andrioli, P. Cavazzani, M. Raiteri, Enhancement of glycine release from human brain synaptosomes by acetylcholine action at M4 muscarinic receptors, J. Exp. Ther. 266 (1993) 142–146. [17] E.A. Van der Zee, P.G.M. Luiten, Muscarinic acetylcholine receptors in hippocampus, neocortex and amygdala: a review on immunocytochemical localization in relation to learning and memory, Prog. Neurobiol. 58 (1999) 409–471.