- Email: [email protected]
Brain-derived neurotrophic factor, apolipoprotein E genetic variants and cognitive performance in Alzheimer’s disease
Neuroscience Letters 367 (2004) 379–383
Brain-derived neurotrophic factor, apolipoprotein E genetic variants and cognitive performance in Alzheimer’s disease Benedetta Nacmias∗ , Carolina Piccini, Silvia Bagnoli, Andrea Tedde, Elena Cellini, Laura Bracco, Sandro Sorbi Department of Neurological and Psychiatric Sciences, University of Florence, Viale Pieraccini 6, 50129 Florence, Italy Received 23 April 2004; received in revised form 31 May 2004; accepted 12 June 2004
Abstract Since greater attention has been paid to the direct link of genetic variation to cognition and memory performance, apolipoprotein E (ApoE) and brain-derived neurotrophic factor (BDNF) have been the two most frequently studied genes. To investigate the effect of BDNF and ApoE polymorphisms on the cognitive profile of mild–moderate Alzheimer’s disease (AD) cases, AD patients, genotyped for ApoE and BDNF polymorphisms, underwent extensive neuropsychological investigation. The effect of either ApoE ε4 allele and BDNF genetic variant on the neuropsychological pattern of mental impairment was examined both in terms of group differences in performance on the neuropsychological tests between carriers and non-carriers of each variant and by selecting the best predictor of cognitive performance among demographic and genetic factors by means of a multiple regression analysis. Our data confirm a specific effect caused by the presence and amount of ApoE ε4 allele, while they suggest that BDNF genetic variants are not a susceptibility factor to AD. © 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Alzheimer’s disease; Brain derived neurotrophic factor; Apolipoprotein E; Polymorphism
Alzheimer’s disease (AD) is a devastating neurodegenerative disorder that typically occurs after age 65, with incidence increasing concurrently with age. The disease is characterized by pronounced memory loss, due to neuropathological changes in the mesial temporal lobes, as the pathology spreads throughout the cerebral cortex lobes. However, it is still unknown why some areas are more affected than others, with subsequent heterogeneity of phenotype and variability in the clinical course. AD is a genetically complex multi-factorial disease with the involvement of several candidate genes. The apolipoprotein E (ApoE) ε4 allele is the only genetic factor consistently implicated in AD, neither necessary nor sufficient for the disease [24]. A number of studies have proposed brain-derived neurotrophic factor (BDNF, 11p13-p14) [10,17], one of the most important neurotrophins that also exerts a modulatory action at hippocampal synapses involved in learning and memory, among the candidate genes. Since 2000, in vivo and in vitro animal studies have shown that the BDNF protein ∗ Corresponding author. Tel.: +39-055-4271-379; fax: +39-055-4271-380. E-mail address: [email protected] (B. Nacmias).
plays an important role in hippocampal long-term potentiation (LTP), thus suggesting that the BDNF genetic variant may affect such a predominant model of hippocampal learning mechanisms [8,12,19]. A common single nucleotide polymorphism [13], consisting of a missense change (196G/A) producing a non-conservative amino acid substitution (Valine to Methionine) in the coding exon of the BDNF gene at position 66 (Val66Met) was recently described as associated with an increased risk of AD susceptibility [25]. According to this study, the frequency of individuals carrying two copies of the Val allele was significantly increased in patients in comparison to the controls. A recent study has shown that the minor genetic variation (Met variant) in the growth factor gene may influence how effectively the hippocampus does function and may impair memory in patients with schizophrenia [6]. Moreover, in 2003, Hariri et al. [11] showed that the Val66Met polymorphism affects human memory-related hippocampal activity and predicts memory performance in healthy individuals. To the best of our knowledge, no study has analyzed the effect of BDNF genetic variant on memory performance in AD patients.
0304-3940/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2004.06.039
380
B. Nacmias et al. / Neuroscience Letters 367 (2004) 379–383
In light of these contrasting findings, we investigated the segregation of BDNF and ApoE polymorphisms in patients with sporadic AD and in healthy controls and analyzed their relative effect, if any, on the cognitive profile of mild–moderate AD cases. As part of a longitudinal survey on primary degenerative dementia, we studied 83 consecutive patients (58 females and 25 males, mean age at entry 72.2 ± 7 years, education 7.0 ± 3.9 years, age at onset 68.5 ± 7.3 years, Mini Mental State Examination score 21.2 ± 4.4 points) with a diagnosis of probable AD who were referred to the Department of Neurology, University of Florence. The local ethical committee approved the protocol of the study and written consent for genetic screening was obtained from all patients or, where appropriate, their relative or legal representative. We based the clinical diagnosis on data from the patient’s clinical history, neurological examination, laboratory test and CT/MRI scans, collected in a standardized protocol [3] that does not substantially differ from the NINCDS-ADRDA guidelines [15]. As a control for the genetic analysis, we also examined a sample of 97 healthy individuals (36 males and 61 females, mean age 72.9 ± 24 years). All controls were carefully assessed using a rigorous clinical history and general/neurological examination in order to exclude the presence of any neurological disorder. DNA from affected and control individuals was extracted from peripheral blood samples using the phenol–chloroform procedure. We analyzed BDNF and ApoE gene polymorphisms using standard PCR and RFLP methods as previously described [22,25]. In particular the Val66Met polymorphism changes a PmlI restriction site. Primer sequences were 5 -ACTCTGGAGAGCGTGAATGG-3 and 5 -ACTACTGAG CATCACCCTGGA-3 . In the presence of Val66 allele digestion with 10 U of PmlI at 37 ◦ C for 4 h produces two fragments of 99 and 72, whereas the Met66 allele resulted uncut (171 bp). The frequencies of ApoE and BDNF alleles and genotypes were estimated by gene counting. Comparisons of genotype and allele frequencies were made using the χ2 -test. Cognitive functions of AD cases were evaluated using an extensive neuropsychological battery [2], including scales examining daily living activities, as well as tasks exploring verbal and spatial memory, orientation, calculation, language, reading–writing capacities, and visuo-motor functions. The battery, standardized on a group of 146 normal subjects, consisted of the Blessed Dementia Scale for Daily Living Activity, Information-Memory-Concentration Test (IMCT), Digit Span, Corsi Tapping Test, Five Words and Paired Words Acquisition and Recall, Babcock Story, Set Test (Category Fluency), Token Test, Clifton Assessment Schedule and Copying Drawing. Furthermore, we assessed executive functions by means of the Letter Fluency Test (a 60-s verbal fluency test for each of the letters P, F and L)
and the Trail Making A–B Test [9]; the attentive/executive component of this task was computed by subtracting the time of execution of part A from that of part B of the test (Trail B–A). All tests scores were adjusted for each subject’s age and education level. The data were analyzed using SPSS software (Statistical Package for Social Sciences 11 for Windows Chicago, SPSS Inc. 2002). The cases were dichotomized by means of the presence or absence of either the BDNF met allele or the ApoE ε4 allele. We used the Student’s t-test to compare group differences in continuous variables and the chi-square for non-continuous variables. We repeated the group comparisons after dividing the study population into mild (MMSE score >20) and moderate (MMSE score ≤20) cases. We applied the multiple linear regressions analysis to select the best predictors of the degree of cognitive and especially memory impairment considering the demographic and genetic factors (ApoE, BDNF). Distribution of the BDNF and ApoE genotypes followed Hardy–Weinberg equilibrium in all groups and did not significantly differ from that of controls (P > 0.1) (Table 1). BDNF genotypes and alleles distributions were similar in the AD cases (57.8% were homozygous for GG, 34.9% heterozygous GA and 7.2% homozygous for AA) and the controls (56.7% GG, 39.1% GA, 4.2% AA). We found a strong association between AD and the ApoE ε4 allele (P < 0.0001, OR > 5.00), confirming earlier reports [7] (Table 1). In addition, we examined the distribution of the Val/Met allele after stratification of the data in terms of ApoE ε4-allele status, but we failed to reveal any significant difference (data not shown). In not one of our AD patients did we find the combination of the BDNF Met66 Met and ApoE ε4/ε4 genotypes; three patients homozygous for the Met allele resulted ApoE ε3/ε4 and three ε3/ε3. We compared carriers (22 females, 13 males) with non-carriers (36 females, 12 males) of the Met allele: no statistically significant differences both in clinico-demographic features and in the mean scores on all the neuropsychological tests were found, even when considering mild (34 females and 17 males, mean MMSE score 23.9 ± 2.4) and moderate (24 females and 8 males, mean MMSE score 16.7 ± 3.0) cases separately. On the contrary, however, carriers (28 females, 9 males) of the ApoE ε4 allele, when compared to non-carriers (30 females, 16 males), displayed a statistically significant inferior performance on five out of the eight tasks exploring the cognitive domain of episodic memory, while no differences emerged on the other tests, as listed in Table 2. We obtained the same results when comparing carriers and non-carriers in the mild group, while we failed to find any significant differences between carriers and non-carriers in the moderate group. In order to establish the best predictor of severity of cognitive impairment in different neuropsychological tasks,
B. Nacmias et al. / Neuroscience Letters 367 (2004) 379–383
381
Table 1 BDNF 196 A/G and ApoE polymorphisms: genotypes and corresponding allele frequencies in AD and controls Samples
Number
BDNF 196 A/G AD 83 Controls 97
BDNF genotypea
Allele frequencyb
AA (%) Met/Met
AG (%) Val/Met
GG (%) Val/Val
A (%) Met
G (%) Val
6 (7.2) 4 (4.2)
29 (34.9) 38 (39.1)
48 (57.8) 55 (56.7)
41 (24.6) 46 (23.7)
125 (75.3) 148 (76.2)
APO-E genotypec
APOE AD Controls a b c d
83 97
Allele frequencyd
ε2/ε3 (%)
ε3/ε3 (%)
ε3/ε4 (%)
ε4/ε4 (%)
ε2 (%)
ε3 (%)
ε4 (%)
7 (8.4) 7 (7.2)
39 (47) 77 (79.3)
28 (33.7) 13 (13.4)
9 (10.8) –
7 (4.2) 7 (3.6)
113 (68) 124 (89.6)
46 (27.7) 13 (6.7)
χ2 = 1; d.f. = 2; P = 0.60. χ2 = 0.01; d.f. = 1; P = 0.92. ε4+ carrier (ε3/ε4 + ε4/ε4) vs. ε4− carrier (ε2/ε3 + ε3/ε3) = χ2 = 21.6; d.f. = 1; P < 0.0001; OR = 5.2; 95% CI = 2.38–11.45. χ2 = 27.3; d.f. = 1; P < 0.0001; OR = 5.34; 95% CI = 2.66–10.9.
Table 2 Comparison of cognitive performances between carriers and non-carriers either of the BDNF Met variant or of the APOE ε4 allele BDNF
ApoE
Met− (n = 48) (5/8%) MMSE IMCT Digit Span Corsi Block Tapping Test Short Story Short Story 10 recall Token Test Five Words Acquisition Five Words 10 recall Five Words 24 h recall Paired Words Acquisition Paired Words 10 recall Paired Words 24 h recall Category Fluency Letter Fluency TMT: B–A difference Copying Drawings
21.7 25.9 4.0 2.9 2.8 2.6 28.5 5.2 9.6 5.3 6.7 12.9 10.7 34.2 18.6 95.2 13.1
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
4.5 5.4 1.2 1.6 2.7 4.0 4.4 4.6 6.3 5.1 4.2 7.0 7.0 6.4 10.5 67.4.7 3.2
Met+ (n = 35) (42%) 20.5 25.1 4.1 2.4 2.6 1.6 27.7 4.7 8.0 4.3 6.4 10.8 9.1 31.7 15.2 101.0 13.0
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
4.3 5.5 1.2 1.4 2.7 2.3 4.1 4.0 6.2 4.1 5.1 7.8 6.6 8.5 7.6 64.0 3.9
P
ε4− (n = 46) (55%)
ε4+ (n = 37) (45%)
20.8 ± 4.4 26.1+5.9 3.9 ± 1.3 2.6 ± 1.6 3.5 ± 2.6 2.6 ± 3.5 27.6 ± 4.8 5.5 ± 4.1 10.4 ± 6.2 5.6 ± 5.0 8.0 ± 4.6 13.8 ± 7.6 11.7 ± 6.7 32.6 ± 8.5 16.1 ± 8.3 99.4 ± 66.0 12.7 ± 3.6
21.6 ± 4.2 24.9 ± 4.7 4.2 ± 1.1 2.8 ± 1.5 1.7 ± 2.4 1.5 ± 0.0 28.9+3.4 4.3 ± 4.2 7.1 ± 6.0 4.0 ± 4.2 4.9 ± 3.9 9.9 ± 6.7 8.0 ± 6.0 33.7 ± 6.0 18.4 ± 10.7 95.5 ± 66.3 13.4 ± 3.5
n.s. n.s. n.s. n.s. 0.004∗ n.s. n.s. n.s. 0.02∗ n.s. 0.002∗ 0.02∗ 0.01∗ n.s. n.s. n.s. n.s.
Values are means ± S.D.; Met− = Val/Val; Met+ = Val/Met, Met/Met; E4− = ε3/ε3, ε2/ε3; E4+ = ε3/ε4, ε4/ε4; MMSE: Mini Mental State Examination; IMCT: Information Memory Concentration Test TMT: Trail Making Test. ∗ Comparing ApoE ε4 carriers vs. non-carriers (Student’s t-test).
we carried out a multiple regression analysis using age, gender, education, presence/amount of the ε4 allele and presence/amount of the BDNF met allele as independent variables. The amount of ApoE ε4 allele was confirmed to be the best predictor of performance on different memory tests either alone (Short Story immediate recall: r = −0.40, P = 0.001; Short Story 10 recall: r = −0.26, P = 0.03; Five Words 10 : r = −0.26, P = 0.03; Paired Words 10 recall: r = −0.30, P = 0.01; Paired Words 24 h recall: r = −0.33, P = 0.005) or with age (Paired Words Acquisition: r = −0.36, P = 0.002). Performance on tests of global cognitive impairment such as the IMCT or on tasks exploring cognitive domains other than episodic memory such as the Token Test, the Corsi
Block Tapping Test, Category and Letter Fluency, Copying Drawings and the Trail Making B–A difference were not predicted by the amount of either the ApoE ε4 or the Met allele. Over the last decade, there has been an increasing interest in the effect of genetic variation on cognition with the assumption that several sets of genes may differentially contribute to individual differences in how cognitive systems operate [18]. In particular, greater attention has been paid to the direct link of genetic variation to memory performance, hippocampal neurochemistry, and hippocampal function [8]. The analysis of genotype–phenotype correlations has revealed that the presence of the ApoE ε4 allele in AD, in conjunction with other loci distributed across the genome,
382
B. Nacmias et al. / Neuroscience Letters 367 (2004) 379–383
influences disease onset, brain atrophy, cerebrovascular perfusion, blood pressure, beta-amyloid deposition, ApoE secretion, lipid metabolism, brain bioelectrical activity, cognition, apoptosis and treatment outcome [4]. The differential impact of ApoE alleles on neuronal plasticity [16], as well as the well-established association between the presence of the ApoE ε4 allele and an increased risk of AD [5,7], have provided the background to investigate the effect of the ApoE genotype on cognition: a relatively selective effect of the ε4 allele on episodic memory [18,23,26] and attentional functions [20] has been reported both in healthy subjects and in AD cases, especially in the early stage of the disease [27]. Likewise, an implication of BDNF gene polymorphisms in AD [25] and cognition [6,11] was recently reported, although with conflicting results. The aim of our study was to investigate the segregation of the BDNF and ApoE polymorphisms in Italian AD patients with mild–moderate disease severity, extensively assessed from a neuropsychological point of view, who were also free of other neurological or internistic diseases. Our findings confirm the well-known increased risk of AD for carriers of the ApoE ε4 allele, while they reveal no association for the BDNF polymorphisms. In particular, our data do not support previous data that found an association between the Val allele and AD [25]. Stratification of the BDNF data, based on the presence or absence of the ApoE ε4 allele, (patients with one or more ε4 alleles were rated as ε4-positive and the others were rated as ε4-negative) did not significantly modify the results that showed no correlation between ApoE and BDNF genotypes and failed to show any epistatic effect between BDNF and ApoE genotypes. With regards to the relationship between the ApoE and Val66Met variant and memory performance in AD, our data confirm a specific effect caused by the presence and amount of the ApoE ε4 allele on the performance of tasks exploring episodic memory, both in terms of group differences between carriers and non-carriers and in terms of the best predictor of performance. Such an effect appears to be greater in the mild cases, as already reported [27], although interpretation of this finding requires caution due to the small sample size of the moderate group. With regards to the Met allele, our data do not confirm previous observations [6,14], as no statistically significant differences both in clinico-demographic features and in the mean scores on all the neuropsychological tests were found in AD patient Met allele carriers with respect to Met allele non-carriers. Neither do our recent data [1] confirm a possible correlation of another recent BDNF gene polymorphism (C270T in the 5’UTR) with late onset AD [21]. To the best of our knowledge, all the findings linking BDNF genetic variant to episodic memory functions have been reported in normal [11] or pathological conditions [6] that are neutral in respect to the presence of memory deficits, while no data are available in cases affected by AD where memory impairment is crucial in making a clinical diagnosis.
In conclusion, our data suggest that BDNF genetic variants are not an AD susceptibility factor, nor do they mitigate the effect of the ApoE ε4 allele in the risk of developing AD.
Acknowledgements This study was supported by Compagnia San Paolo, and by ISS (Istituto Superiore Di Sanità).
References [1] S. Bagnoli, B. Nacmias, A. Tedde, B.M. Guarnieri, E. Cellini, C. Petruzzi, A. Bartoli, L. Ortenzi, S. Sorbi, Brain derived neutrophic factor genetic variants are not susceptibility factors to Alzheimer’s disease in Italy, Ann. Neurol. 55 (2004) 447–448. [2] L. Bracco, L. Amaducci, D. Pedone, G. Bino, M.P. Lazzaro, F. Carella, R. D’Antona, R. Gallato, G. Denes, Italian Multicentre Study on Dementia (SMID): a neuropsychological test battery for assessing Alzheimer’s disease, J. Psychiatr. Res. 24 (1990) 213–226. [3] L. Bracco, L. Amaducci, Italian Multicenter Study on Dementia: a protocol for data collection and clinical diagnosis of Alzheimer’s disease. The SMID Group, Neuroepidemiology 11 (1992) 39–45. [4] R. Cacabelos, The application of functional genomics to Alzheimer’s disease, Pharmacogenomics 4 (2003) 597–621. [5] E.H. Corder, A.M. Saunders, W.J. Strittmatter, D.E. Schmechel, P.C. Gaskell, G.W. Small, A.D. Roses, J.L. Haines, M.A. Pericak-Vance, Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families, Science 261 (1993) 921– 923. [6] M.F. Egan, M. Kojima, J.H. Callicott, T.E. Goldberg, B.S. Kolachana, A. Bertolino, E. Zaitsev, B. Gold, D. Goldman, M. Dean, B. Lu, D.R. Weinberger, The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function, Cell 112 (2003) 257–269. [7] L.A. Farrer, L.A. Cupples, J.L. Haines, B. Hyman, W.A. Kukull, R. Mayeux, R.H. Myers, M.A. Pericak-Vance, N. Risch, C.M. van Duijn, Effects of age, sex, and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease. A meta-analysis. APOE and Alzheimer Disease Meta. Analysis Consortium, JAMA 278 (1997) 1349–1356. [8] J.D. Gabrieli, A.R. Preston, Visualizing genetic influences on human brain functions, Cell 112 (2003) 144–145. [9] A.R. Giovagnoli, M. Del Pesce, S. Mascheroni, M. Simoncelli, M. Laiacona, E. Capitani, Trail making test: normative values from 287 normal adult controls, Ital. J. Neurol. Sci. 17 (1996) 305–309. [10] I.M. Hanson, A. Seawright, V. van Heyningen, The human BDNF gene maps between FSHB and HVBS1 at the boundary of 11p13-p14, Genomics 13 (1992) 1331–1333. [11] A.R. Hariri, T.E. Goldberg, V.S. Mattay, B.S. Kolachana, J.H. Callicott, M.F. Egan, D.R. Weinberger, Brain-derived neurotrophic factor val66met polymorphism affects human memory-related hippocampal activity and predicts memory performance, J. Neurosci. 23 (2003) 6690–6694. [12] B. Lu, W. Gottschalk, Modulation of hippocampal synaptic transmission and plasticity by neurotrophins, Prog. Brain Res. 128 (2000) 231–241. [13] P.C. Maisonpierre, M.M. Le Beau, R. Espinosa, N.Y. Ip, L. Belluscio, S.M. de la Monte, S. Squinto, M.E. Furth, G.D. Yancopoulos, Human and rat brain-derived neurotrophic factor and neurotrophin-3: gene structures, distributions, and chromosomal localizations, Genomics 10 (1991) 558–568.
B. Nacmias et al. / Neuroscience Letters 367 (2004) 379–383 [14] J. Marx, Neuroscience. Minor variation in growth-factor gene impairs human memory, Science 299 (2003) 639–640. [15] G. McKhann, D. Drachman, M. Folstein, R. Katzman, D. Price, E.M. Stadlan, Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease, Neurology 34 (1984) 939–944. [16] M.M. Mesulam, Neuroplasticity failure in Alzheimer’s disease: bridging the gap between plaques and tangles, Neuron 24 (1999) 521–529. [17] M.G. Murer, Q. Yan, R. Raisman-Vozari, Brain-derived neurotrophic factor in the control human brain, Prog. Neurobiol. 63 (2001) 71– 124. [18] L.G. Nilsson, L. Nyberg, L. Backman, Genetic variation in memory functioning, Neurosci. Biobehav. Rev. 26 (2002) 841–848. [19] M.M. Poo, Neurotrophins as synaptic modulators, Nat. Rev. Neurosci. 2 (2001) 24–32. [20] R. Parasuraman, P.M. Greenwood, T. Sunderland, The apolipoprotein E gene, attention, and brain function, Neuropsychology 16 (2002) 254–274. [21] M. Riemenschneider, S. Schwarz, S. Wagenpfeil, J. Diehl, U. Muller, H. Forstl, A. Kurz, A polymorphism of the brain-derived neurotrophic factor (BDNF) is associated with Alzheimer’s disease in patients lacking the Apolipoprotein E epsilon4 allele, Mol. Psychiatry 7 (2002) 782–785.
383
[22] S. Sorbi, B. Nacmias, P. Forleo, S. Latorraca, I. Gobbini, L. Bracco, S. Piacentini, L. Amaducci, ApoE allele frequencies in Italian sporadic and familial Alzheimer’s disease, Neurosi. Lett. 177 (1994) 100– 4102. [23] G.E. Smith, D.L. Bohac, S.C. Waring, E. Kokmen, E.G. Tangalos, R.J. Ivnik, R.C. Petersen, Alzheimer’s disease but not in healthy control subjects, Neurology 50 (1998) 355–362. [24] W.J. Strittmatter, A.D. Roses, Apolipoprotein E and Alzheimer disease, Proc. Natl. Acad. Sci. U.S.A. 92 (1995) 4725– 4727. [25] M. Ventriglia, L. Bocchio Chiavetto, L. Benussi, G. Binetti, O. Zanetti, M.A. Riva, M. Pennarelli, Association between the BDNF 196 A/G polymorphism and sporadic Alzheimer’s disease, Mol. Psychiatry 7 (2002) 136–137. [26] R.S. Wilson, J.A. Schneider, L.L. Barnes, L.A. Beckett, N.T. Aggarwal, E.J. Cochran, E. Berry-Kravis, J. Bach, J.H. Fox, D.A. Evans, D.A. Bennett, The apolipoprotein E epsilon 4 allele and decline in different cognitive systems during a 6-year period, Arch. Neurol. 59 (2002) 1154–1160. [27] K.W. Kim, J.H. Jhoo, J.H. Lee, D.Y. Lee, K.U. Lee, J.C. Youn, J.Y. Youn, J.I. Woo, The domain-specific, stage-limited impact of the Apolipoprotein E epsilon-4 allele on cognitive functions in Alzheimer’s disease, Dement. Geriatr. Cogn. Disord. 13 (2002) 125– 129.
Brain-derived neurotrophic factor, apolipoprotein E genetic variants and cognitive performance in Alzheimer’s disease Benedetta Nacmias∗ , Carolina Piccini, Silvia Bagnoli, Andrea Tedde, Elena Cellini, Laura Bracco, Sandro Sorbi Department of Neurological and Psychiatric Sciences, University of Florence, Viale Pieraccini 6, 50129 Florence, Italy Received 23 April 2004; received in revised form 31 May 2004; accepted 12 June 2004
Abstract Since greater attention has been paid to the direct link of genetic variation to cognition and memory performance, apolipoprotein E (ApoE) and brain-derived neurotrophic factor (BDNF) have been the two most frequently studied genes. To investigate the effect of BDNF and ApoE polymorphisms on the cognitive profile of mild–moderate Alzheimer’s disease (AD) cases, AD patients, genotyped for ApoE and BDNF polymorphisms, underwent extensive neuropsychological investigation. The effect of either ApoE ε4 allele and BDNF genetic variant on the neuropsychological pattern of mental impairment was examined both in terms of group differences in performance on the neuropsychological tests between carriers and non-carriers of each variant and by selecting the best predictor of cognitive performance among demographic and genetic factors by means of a multiple regression analysis. Our data confirm a specific effect caused by the presence and amount of ApoE ε4 allele, while they suggest that BDNF genetic variants are not a susceptibility factor to AD. © 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Alzheimer’s disease; Brain derived neurotrophic factor; Apolipoprotein E; Polymorphism
Alzheimer’s disease (AD) is a devastating neurodegenerative disorder that typically occurs after age 65, with incidence increasing concurrently with age. The disease is characterized by pronounced memory loss, due to neuropathological changes in the mesial temporal lobes, as the pathology spreads throughout the cerebral cortex lobes. However, it is still unknown why some areas are more affected than others, with subsequent heterogeneity of phenotype and variability in the clinical course. AD is a genetically complex multi-factorial disease with the involvement of several candidate genes. The apolipoprotein E (ApoE) ε4 allele is the only genetic factor consistently implicated in AD, neither necessary nor sufficient for the disease [24]. A number of studies have proposed brain-derived neurotrophic factor (BDNF, 11p13-p14) [10,17], one of the most important neurotrophins that also exerts a modulatory action at hippocampal synapses involved in learning and memory, among the candidate genes. Since 2000, in vivo and in vitro animal studies have shown that the BDNF protein ∗ Corresponding author. Tel.: +39-055-4271-379; fax: +39-055-4271-380. E-mail address: [email protected] (B. Nacmias).
plays an important role in hippocampal long-term potentiation (LTP), thus suggesting that the BDNF genetic variant may affect such a predominant model of hippocampal learning mechanisms [8,12,19]. A common single nucleotide polymorphism [13], consisting of a missense change (196G/A) producing a non-conservative amino acid substitution (Valine to Methionine) in the coding exon of the BDNF gene at position 66 (Val66Met) was recently described as associated with an increased risk of AD susceptibility [25]. According to this study, the frequency of individuals carrying two copies of the Val allele was significantly increased in patients in comparison to the controls. A recent study has shown that the minor genetic variation (Met variant) in the growth factor gene may influence how effectively the hippocampus does function and may impair memory in patients with schizophrenia [6]. Moreover, in 2003, Hariri et al. [11] showed that the Val66Met polymorphism affects human memory-related hippocampal activity and predicts memory performance in healthy individuals. To the best of our knowledge, no study has analyzed the effect of BDNF genetic variant on memory performance in AD patients.
0304-3940/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2004.06.039
380
B. Nacmias et al. / Neuroscience Letters 367 (2004) 379–383
In light of these contrasting findings, we investigated the segregation of BDNF and ApoE polymorphisms in patients with sporadic AD and in healthy controls and analyzed their relative effect, if any, on the cognitive profile of mild–moderate AD cases. As part of a longitudinal survey on primary degenerative dementia, we studied 83 consecutive patients (58 females and 25 males, mean age at entry 72.2 ± 7 years, education 7.0 ± 3.9 years, age at onset 68.5 ± 7.3 years, Mini Mental State Examination score 21.2 ± 4.4 points) with a diagnosis of probable AD who were referred to the Department of Neurology, University of Florence. The local ethical committee approved the protocol of the study and written consent for genetic screening was obtained from all patients or, where appropriate, their relative or legal representative. We based the clinical diagnosis on data from the patient’s clinical history, neurological examination, laboratory test and CT/MRI scans, collected in a standardized protocol [3] that does not substantially differ from the NINCDS-ADRDA guidelines [15]. As a control for the genetic analysis, we also examined a sample of 97 healthy individuals (36 males and 61 females, mean age 72.9 ± 24 years). All controls were carefully assessed using a rigorous clinical history and general/neurological examination in order to exclude the presence of any neurological disorder. DNA from affected and control individuals was extracted from peripheral blood samples using the phenol–chloroform procedure. We analyzed BDNF and ApoE gene polymorphisms using standard PCR and RFLP methods as previously described [22,25]. In particular the Val66Met polymorphism changes a PmlI restriction site. Primer sequences were 5 -ACTCTGGAGAGCGTGAATGG-3 and 5 -ACTACTGAG CATCACCCTGGA-3 . In the presence of Val66 allele digestion with 10 U of PmlI at 37 ◦ C for 4 h produces two fragments of 99 and 72, whereas the Met66 allele resulted uncut (171 bp). The frequencies of ApoE and BDNF alleles and genotypes were estimated by gene counting. Comparisons of genotype and allele frequencies were made using the χ2 -test. Cognitive functions of AD cases were evaluated using an extensive neuropsychological battery [2], including scales examining daily living activities, as well as tasks exploring verbal and spatial memory, orientation, calculation, language, reading–writing capacities, and visuo-motor functions. The battery, standardized on a group of 146 normal subjects, consisted of the Blessed Dementia Scale for Daily Living Activity, Information-Memory-Concentration Test (IMCT), Digit Span, Corsi Tapping Test, Five Words and Paired Words Acquisition and Recall, Babcock Story, Set Test (Category Fluency), Token Test, Clifton Assessment Schedule and Copying Drawing. Furthermore, we assessed executive functions by means of the Letter Fluency Test (a 60-s verbal fluency test for each of the letters P, F and L)
and the Trail Making A–B Test [9]; the attentive/executive component of this task was computed by subtracting the time of execution of part A from that of part B of the test (Trail B–A). All tests scores were adjusted for each subject’s age and education level. The data were analyzed using SPSS software (Statistical Package for Social Sciences 11 for Windows Chicago, SPSS Inc. 2002). The cases were dichotomized by means of the presence or absence of either the BDNF met allele or the ApoE ε4 allele. We used the Student’s t-test to compare group differences in continuous variables and the chi-square for non-continuous variables. We repeated the group comparisons after dividing the study population into mild (MMSE score >20) and moderate (MMSE score ≤20) cases. We applied the multiple linear regressions analysis to select the best predictors of the degree of cognitive and especially memory impairment considering the demographic and genetic factors (ApoE, BDNF). Distribution of the BDNF and ApoE genotypes followed Hardy–Weinberg equilibrium in all groups and did not significantly differ from that of controls (P > 0.1) (Table 1). BDNF genotypes and alleles distributions were similar in the AD cases (57.8% were homozygous for GG, 34.9% heterozygous GA and 7.2% homozygous for AA) and the controls (56.7% GG, 39.1% GA, 4.2% AA). We found a strong association between AD and the ApoE ε4 allele (P < 0.0001, OR > 5.00), confirming earlier reports [7] (Table 1). In addition, we examined the distribution of the Val/Met allele after stratification of the data in terms of ApoE ε4-allele status, but we failed to reveal any significant difference (data not shown). In not one of our AD patients did we find the combination of the BDNF Met66 Met and ApoE ε4/ε4 genotypes; three patients homozygous for the Met allele resulted ApoE ε3/ε4 and three ε3/ε3. We compared carriers (22 females, 13 males) with non-carriers (36 females, 12 males) of the Met allele: no statistically significant differences both in clinico-demographic features and in the mean scores on all the neuropsychological tests were found, even when considering mild (34 females and 17 males, mean MMSE score 23.9 ± 2.4) and moderate (24 females and 8 males, mean MMSE score 16.7 ± 3.0) cases separately. On the contrary, however, carriers (28 females, 9 males) of the ApoE ε4 allele, when compared to non-carriers (30 females, 16 males), displayed a statistically significant inferior performance on five out of the eight tasks exploring the cognitive domain of episodic memory, while no differences emerged on the other tests, as listed in Table 2. We obtained the same results when comparing carriers and non-carriers in the mild group, while we failed to find any significant differences between carriers and non-carriers in the moderate group. In order to establish the best predictor of severity of cognitive impairment in different neuropsychological tasks,
B. Nacmias et al. / Neuroscience Letters 367 (2004) 379–383
381
Table 1 BDNF 196 A/G and ApoE polymorphisms: genotypes and corresponding allele frequencies in AD and controls Samples
Number
BDNF 196 A/G AD 83 Controls 97
BDNF genotypea
Allele frequencyb
AA (%) Met/Met
AG (%) Val/Met
GG (%) Val/Val
A (%) Met
G (%) Val
6 (7.2) 4 (4.2)
29 (34.9) 38 (39.1)
48 (57.8) 55 (56.7)
41 (24.6) 46 (23.7)
125 (75.3) 148 (76.2)
APO-E genotypec
APOE AD Controls a b c d
83 97
Allele frequencyd
ε2/ε3 (%)
ε3/ε3 (%)
ε3/ε4 (%)
ε4/ε4 (%)
ε2 (%)
ε3 (%)
ε4 (%)
7 (8.4) 7 (7.2)
39 (47) 77 (79.3)
28 (33.7) 13 (13.4)
9 (10.8) –
7 (4.2) 7 (3.6)
113 (68) 124 (89.6)
46 (27.7) 13 (6.7)
χ2 = 1; d.f. = 2; P = 0.60. χ2 = 0.01; d.f. = 1; P = 0.92. ε4+ carrier (ε3/ε4 + ε4/ε4) vs. ε4− carrier (ε2/ε3 + ε3/ε3) = χ2 = 21.6; d.f. = 1; P < 0.0001; OR = 5.2; 95% CI = 2.38–11.45. χ2 = 27.3; d.f. = 1; P < 0.0001; OR = 5.34; 95% CI = 2.66–10.9.
Table 2 Comparison of cognitive performances between carriers and non-carriers either of the BDNF Met variant or of the APOE ε4 allele BDNF
ApoE
Met− (n = 48) (5/8%) MMSE IMCT Digit Span Corsi Block Tapping Test Short Story Short Story 10 recall Token Test Five Words Acquisition Five Words 10 recall Five Words 24 h recall Paired Words Acquisition Paired Words 10 recall Paired Words 24 h recall Category Fluency Letter Fluency TMT: B–A difference Copying Drawings
21.7 25.9 4.0 2.9 2.8 2.6 28.5 5.2 9.6 5.3 6.7 12.9 10.7 34.2 18.6 95.2 13.1
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
4.5 5.4 1.2 1.6 2.7 4.0 4.4 4.6 6.3 5.1 4.2 7.0 7.0 6.4 10.5 67.4.7 3.2
Met+ (n = 35) (42%) 20.5 25.1 4.1 2.4 2.6 1.6 27.7 4.7 8.0 4.3 6.4 10.8 9.1 31.7 15.2 101.0 13.0
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
4.3 5.5 1.2 1.4 2.7 2.3 4.1 4.0 6.2 4.1 5.1 7.8 6.6 8.5 7.6 64.0 3.9
P
ε4− (n = 46) (55%)
ε4+ (n = 37) (45%)
20.8 ± 4.4 26.1+5.9 3.9 ± 1.3 2.6 ± 1.6 3.5 ± 2.6 2.6 ± 3.5 27.6 ± 4.8 5.5 ± 4.1 10.4 ± 6.2 5.6 ± 5.0 8.0 ± 4.6 13.8 ± 7.6 11.7 ± 6.7 32.6 ± 8.5 16.1 ± 8.3 99.4 ± 66.0 12.7 ± 3.6
21.6 ± 4.2 24.9 ± 4.7 4.2 ± 1.1 2.8 ± 1.5 1.7 ± 2.4 1.5 ± 0.0 28.9+3.4 4.3 ± 4.2 7.1 ± 6.0 4.0 ± 4.2 4.9 ± 3.9 9.9 ± 6.7 8.0 ± 6.0 33.7 ± 6.0 18.4 ± 10.7 95.5 ± 66.3 13.4 ± 3.5
n.s. n.s. n.s. n.s. 0.004∗ n.s. n.s. n.s. 0.02∗ n.s. 0.002∗ 0.02∗ 0.01∗ n.s. n.s. n.s. n.s.
Values are means ± S.D.; Met− = Val/Val; Met+ = Val/Met, Met/Met; E4− = ε3/ε3, ε2/ε3; E4+ = ε3/ε4, ε4/ε4; MMSE: Mini Mental State Examination; IMCT: Information Memory Concentration Test TMT: Trail Making Test. ∗ Comparing ApoE ε4 carriers vs. non-carriers (Student’s t-test).
we carried out a multiple regression analysis using age, gender, education, presence/amount of the ε4 allele and presence/amount of the BDNF met allele as independent variables. The amount of ApoE ε4 allele was confirmed to be the best predictor of performance on different memory tests either alone (Short Story immediate recall: r = −0.40, P = 0.001; Short Story 10 recall: r = −0.26, P = 0.03; Five Words 10 : r = −0.26, P = 0.03; Paired Words 10 recall: r = −0.30, P = 0.01; Paired Words 24 h recall: r = −0.33, P = 0.005) or with age (Paired Words Acquisition: r = −0.36, P = 0.002). Performance on tests of global cognitive impairment such as the IMCT or on tasks exploring cognitive domains other than episodic memory such as the Token Test, the Corsi
Block Tapping Test, Category and Letter Fluency, Copying Drawings and the Trail Making B–A difference were not predicted by the amount of either the ApoE ε4 or the Met allele. Over the last decade, there has been an increasing interest in the effect of genetic variation on cognition with the assumption that several sets of genes may differentially contribute to individual differences in how cognitive systems operate [18]. In particular, greater attention has been paid to the direct link of genetic variation to memory performance, hippocampal neurochemistry, and hippocampal function [8]. The analysis of genotype–phenotype correlations has revealed that the presence of the ApoE ε4 allele in AD, in conjunction with other loci distributed across the genome,
382
B. Nacmias et al. / Neuroscience Letters 367 (2004) 379–383
influences disease onset, brain atrophy, cerebrovascular perfusion, blood pressure, beta-amyloid deposition, ApoE secretion, lipid metabolism, brain bioelectrical activity, cognition, apoptosis and treatment outcome [4]. The differential impact of ApoE alleles on neuronal plasticity [16], as well as the well-established association between the presence of the ApoE ε4 allele and an increased risk of AD [5,7], have provided the background to investigate the effect of the ApoE genotype on cognition: a relatively selective effect of the ε4 allele on episodic memory [18,23,26] and attentional functions [20] has been reported both in healthy subjects and in AD cases, especially in the early stage of the disease [27]. Likewise, an implication of BDNF gene polymorphisms in AD [25] and cognition [6,11] was recently reported, although with conflicting results. The aim of our study was to investigate the segregation of the BDNF and ApoE polymorphisms in Italian AD patients with mild–moderate disease severity, extensively assessed from a neuropsychological point of view, who were also free of other neurological or internistic diseases. Our findings confirm the well-known increased risk of AD for carriers of the ApoE ε4 allele, while they reveal no association for the BDNF polymorphisms. In particular, our data do not support previous data that found an association between the Val allele and AD [25]. Stratification of the BDNF data, based on the presence or absence of the ApoE ε4 allele, (patients with one or more ε4 alleles were rated as ε4-positive and the others were rated as ε4-negative) did not significantly modify the results that showed no correlation between ApoE and BDNF genotypes and failed to show any epistatic effect between BDNF and ApoE genotypes. With regards to the relationship between the ApoE and Val66Met variant and memory performance in AD, our data confirm a specific effect caused by the presence and amount of the ApoE ε4 allele on the performance of tasks exploring episodic memory, both in terms of group differences between carriers and non-carriers and in terms of the best predictor of performance. Such an effect appears to be greater in the mild cases, as already reported [27], although interpretation of this finding requires caution due to the small sample size of the moderate group. With regards to the Met allele, our data do not confirm previous observations [6,14], as no statistically significant differences both in clinico-demographic features and in the mean scores on all the neuropsychological tests were found in AD patient Met allele carriers with respect to Met allele non-carriers. Neither do our recent data [1] confirm a possible correlation of another recent BDNF gene polymorphism (C270T in the 5’UTR) with late onset AD [21]. To the best of our knowledge, all the findings linking BDNF genetic variant to episodic memory functions have been reported in normal [11] or pathological conditions [6] that are neutral in respect to the presence of memory deficits, while no data are available in cases affected by AD where memory impairment is crucial in making a clinical diagnosis.
In conclusion, our data suggest that BDNF genetic variants are not an AD susceptibility factor, nor do they mitigate the effect of the ApoE ε4 allele in the risk of developing AD.
Acknowledgements This study was supported by Compagnia San Paolo, and by ISS (Istituto Superiore Di Sanità).
References [1] S. Bagnoli, B. Nacmias, A. Tedde, B.M. Guarnieri, E. Cellini, C. Petruzzi, A. Bartoli, L. Ortenzi, S. Sorbi, Brain derived neutrophic factor genetic variants are not susceptibility factors to Alzheimer’s disease in Italy, Ann. Neurol. 55 (2004) 447–448. [2] L. Bracco, L. Amaducci, D. Pedone, G. Bino, M.P. Lazzaro, F. Carella, R. D’Antona, R. Gallato, G. Denes, Italian Multicentre Study on Dementia (SMID): a neuropsychological test battery for assessing Alzheimer’s disease, J. Psychiatr. Res. 24 (1990) 213–226. [3] L. Bracco, L. Amaducci, Italian Multicenter Study on Dementia: a protocol for data collection and clinical diagnosis of Alzheimer’s disease. The SMID Group, Neuroepidemiology 11 (1992) 39–45. [4] R. Cacabelos, The application of functional genomics to Alzheimer’s disease, Pharmacogenomics 4 (2003) 597–621. [5] E.H. Corder, A.M. Saunders, W.J. Strittmatter, D.E. Schmechel, P.C. Gaskell, G.W. Small, A.D. Roses, J.L. Haines, M.A. Pericak-Vance, Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families, Science 261 (1993) 921– 923. [6] M.F. Egan, M. Kojima, J.H. Callicott, T.E. Goldberg, B.S. Kolachana, A. Bertolino, E. Zaitsev, B. Gold, D. Goldman, M. Dean, B. Lu, D.R. Weinberger, The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function, Cell 112 (2003) 257–269. [7] L.A. Farrer, L.A. Cupples, J.L. Haines, B. Hyman, W.A. Kukull, R. Mayeux, R.H. Myers, M.A. Pericak-Vance, N. Risch, C.M. van Duijn, Effects of age, sex, and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease. A meta-analysis. APOE and Alzheimer Disease Meta. Analysis Consortium, JAMA 278 (1997) 1349–1356. [8] J.D. Gabrieli, A.R. Preston, Visualizing genetic influences on human brain functions, Cell 112 (2003) 144–145. [9] A.R. Giovagnoli, M. Del Pesce, S. Mascheroni, M. Simoncelli, M. Laiacona, E. Capitani, Trail making test: normative values from 287 normal adult controls, Ital. J. Neurol. Sci. 17 (1996) 305–309. [10] I.M. Hanson, A. Seawright, V. van Heyningen, The human BDNF gene maps between FSHB and HVBS1 at the boundary of 11p13-p14, Genomics 13 (1992) 1331–1333. [11] A.R. Hariri, T.E. Goldberg, V.S. Mattay, B.S. Kolachana, J.H. Callicott, M.F. Egan, D.R. Weinberger, Brain-derived neurotrophic factor val66met polymorphism affects human memory-related hippocampal activity and predicts memory performance, J. Neurosci. 23 (2003) 6690–6694. [12] B. Lu, W. Gottschalk, Modulation of hippocampal synaptic transmission and plasticity by neurotrophins, Prog. Brain Res. 128 (2000) 231–241. [13] P.C. Maisonpierre, M.M. Le Beau, R. Espinosa, N.Y. Ip, L. Belluscio, S.M. de la Monte, S. Squinto, M.E. Furth, G.D. Yancopoulos, Human and rat brain-derived neurotrophic factor and neurotrophin-3: gene structures, distributions, and chromosomal localizations, Genomics 10 (1991) 558–568.
B. Nacmias et al. / Neuroscience Letters 367 (2004) 379–383 [14] J. Marx, Neuroscience. Minor variation in growth-factor gene impairs human memory, Science 299 (2003) 639–640. [15] G. McKhann, D. Drachman, M. Folstein, R. Katzman, D. Price, E.M. Stadlan, Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease, Neurology 34 (1984) 939–944. [16] M.M. Mesulam, Neuroplasticity failure in Alzheimer’s disease: bridging the gap between plaques and tangles, Neuron 24 (1999) 521–529. [17] M.G. Murer, Q. Yan, R. Raisman-Vozari, Brain-derived neurotrophic factor in the control human brain, Prog. Neurobiol. 63 (2001) 71– 124. [18] L.G. Nilsson, L. Nyberg, L. Backman, Genetic variation in memory functioning, Neurosci. Biobehav. Rev. 26 (2002) 841–848. [19] M.M. Poo, Neurotrophins as synaptic modulators, Nat. Rev. Neurosci. 2 (2001) 24–32. [20] R. Parasuraman, P.M. Greenwood, T. Sunderland, The apolipoprotein E gene, attention, and brain function, Neuropsychology 16 (2002) 254–274. [21] M. Riemenschneider, S. Schwarz, S. Wagenpfeil, J. Diehl, U. Muller, H. Forstl, A. Kurz, A polymorphism of the brain-derived neurotrophic factor (BDNF) is associated with Alzheimer’s disease in patients lacking the Apolipoprotein E epsilon4 allele, Mol. Psychiatry 7 (2002) 782–785.
383
[22] S. Sorbi, B. Nacmias, P. Forleo, S. Latorraca, I. Gobbini, L. Bracco, S. Piacentini, L. Amaducci, ApoE allele frequencies in Italian sporadic and familial Alzheimer’s disease, Neurosi. Lett. 177 (1994) 100– 4102. [23] G.E. Smith, D.L. Bohac, S.C. Waring, E. Kokmen, E.G. Tangalos, R.J. Ivnik, R.C. Petersen, Alzheimer’s disease but not in healthy control subjects, Neurology 50 (1998) 355–362. [24] W.J. Strittmatter, A.D. Roses, Apolipoprotein E and Alzheimer disease, Proc. Natl. Acad. Sci. U.S.A. 92 (1995) 4725– 4727. [25] M. Ventriglia, L. Bocchio Chiavetto, L. Benussi, G. Binetti, O. Zanetti, M.A. Riva, M. Pennarelli, Association between the BDNF 196 A/G polymorphism and sporadic Alzheimer’s disease, Mol. Psychiatry 7 (2002) 136–137. [26] R.S. Wilson, J.A. Schneider, L.L. Barnes, L.A. Beckett, N.T. Aggarwal, E.J. Cochran, E. Berry-Kravis, J. Bach, J.H. Fox, D.A. Evans, D.A. Bennett, The apolipoprotein E epsilon 4 allele and decline in different cognitive systems during a 6-year period, Arch. Neurol. 59 (2002) 1154–1160. [27] K.W. Kim, J.H. Jhoo, J.H. Lee, D.Y. Lee, K.U. Lee, J.C. Youn, J.Y. Youn, J.I. Woo, The domain-specific, stage-limited impact of the Apolipoprotein E epsilon-4 allele on cognitive functions in Alzheimer’s disease, Dement. Geriatr. Cogn. Disord. 13 (2002) 125– 129.