European Journal of Neurology 2009, 16: 870–873
doi:10.1111/j.1468-1331.2009.02611.x
DCUN1D1 is a risk factor for frontotemporal lobar degeneration C. Villaa, E. Venturellia, C. Fenoglioa, F. Clericib, A. Marconec, L. Benussid, S. Gallonee, D. Scalabrinia, F. Cortinia, M. Serpentea, F. Martinelli Boneschif, S. Cappac,g, G. Binettid, C. Marianib, I. Raineroe, M. T. Giordanae, N. Bresolina, E. Scarpinia and D. Galimbertia a
Department of Neurological Sciences, ‘‘Dino Ferrari’’ Center, University of Milan, IRCCS Fondazione Ospedale Maggiore Policlinico,
Milan, Italy; bCenter for Research and Treatment on Cognitive Dysfunctions, Chair of Neurology, University of Milan, ‘‘Luigi Sacco’’ Hospital, Milan, Italy; cDivision of Neurology, San Raffaele Turro Hospital, San Raffaele Scientific Institute, Milan, Italy; d
NeuroBioGen-Lab-Memory Clinic, IRCCS Centro S.Giovanni di Dio-Fatebenefratelli, Brescia, Italy; eDepartment of Neuroscience,
University of Turin, Turin, Italy; fDepartment of Neurology and INSPE, Scientific Institute San Raffaele, Milan, Italy; and gVita-Salute San Raffaele University, Milan, Italy
Keywords:
DCUN1D1, frontotemporal lobar degeneration, polymorphism, risk factor Received 25 July 2008 Accepted 18 February 2009
Background and purpose: Frontotemporal lobar degeneration (FTLD) is considered as a proteinopathy; therefore, it is conceivable that genes encoding for factors involved in protein misfolding and/or degradation could play a role in its pathogenesis. Methods: An association study of defective in cullin neddylation 1 (DCN-1)-domain containing 1 (DCUN1D1), which is involved in protein degradation, was carried out in a population of 220 patients with FTLD as compared with 229 age-matched controls. Results: A statistically significant increased frequency of the GG genotype of the DCUN1D1 rs4859146 single nucleotide polymorphism (SNP) was observed in patients compared with controls (6.9 vs. 1.7%, P = 0.011, adjusted OR: 4.39, 95% CI: 1.40– 13.78). Stratifying according to the clinical syndrome, significant differences were observed between the behavioral variant of frontotemporal dementia and controls (GG frequency: 6.3 vs. 1.7%, P = 0.02, OR:4.0, 95%, CI = 1.24–12.92), as well as between patients with progressive aphasia compared with controls (15.4 vs. 1.7%, P = 0.014, OR = 11.30, 95%, CI = 1.63–78.45), but not in patients with SD versus controls (8.3 vs. 1.7%, P = 0.18, OR = 5.24, 95% C.I. = 0.45–60.63). No significant differences in allelic and genotypic frequencies of the DCUN1D1 rs4859147 SNP were found. Conclusions: The GG genotype of the DCUN1D1 rs4859147 SNP represents a risk factor for the development of FTLD, increasing the risk of about fourfold.
Introduction The term frontotemporal lobar degeneration (FTLD) is an example of a disease group characterized by atrophy of the prefrontal and anterior temporal lobes. It represents the second most common type of early-onset neurodegenerative dementia after AlzheimerÕs disease (AD), accounting for 5–10% of cases of dementia [1]. The presenting clinical features consist of behavioral dysfunction and personality changes, often with language impairment, loss of social awareness, overeating, and impulsiveness, whilst memory is relatively spared
Correspondence: Daniela Galimberti, Department of Neurological Sciences, ‘‘Dino Ferrari’’ Center, University of Milan, IRCCS Fondazione Ospedale Maggiore Policlinico, Milan, Italy (tel.: ++39 2 55033858; fax: ++39 2 50320430; e-mail: daniela.
[email protected]).
870
[2]. The current consensus criteria [3] identify three clinical syndromes: behavioral variant of frontotemporal dementia (bvFTD), progressive non-fluent aphasia (PA) and semantic dementia (SD), which reflect the clinical heterogeneity of FTLD. bvFTD is characterized by behavioral abnormalities, whereas PA is associated with progressive loss of speech, with hesitant, non-fluent speech output [4], and SD is associated with loss of knowledge about words and objects. This variability is determined by the relative involvement of the frontal and temporal lobes, as well as by the involvement of right and left hemispheres [2]. Mutations in MAPT and PGRN have been repeatedly demonstrated to be the cause of familial FTD with, respectively, tau or ubiquitin pathology [5,6]. Whereas there is agreement that MAPT mutations lead to a very early onset of the disease, PGRN mutations have been found also in apparently sporadic cases with a late onset of the disease [7].
2009 The Author(s) Journal compilation 2009 EFNS
DCUN1D1 is a risk factor for FTLD
Genes encoding for factors involved in protein misfolding and/or degradation could play a role in FTLD pathogenesis. Protein polyubiquitination is required for subsequent degradation by the 26S proteasome [8]. E3 ubiquitin ligases are necessary for polyubiquitination, and their catalytic core, made of cullins, needs neddylation to be functional. Defective in cullin neddylation 1 (DCN-1) is a critical factor required for cullin neddylation [8]. Studies in Caenorhabditis elegans and in Saccharomyces cerevisiae demonstrated that DCN-1 directly bind Nedd8 and is physically associated with cullins in both species. Neddylation and ubiquitination use parallel mechanisms; each involves their own E1, E2, and E3 proteins, and both follow a sequential modification process. Cullin-based E3 ligase complexes are thought to provide specificity to the degradation reaction [8]. Thus, DCN-1 could be a good candidate for diseases caused by accumulation of misfolded protein, including FTLD. Given these premises, we carried out an association study of DCN-1-domain containing 1 (DCUN1D1), a 259-aminoacid protein expressed in the brain, in a population of 220 patients with FTLD as compared with 229 age-matched controls in order to determine whether it could influence the susceptibility to develop the disease.
Patients and methods Two-hundred twenty Italian patients with FTLD (102 males and 118 females, mean age at onset ±SEM: 65.9 ± 0.69 years) were consecutively recruited at the Alzheimer Units of Ospedale Maggiore Policlinico (Milan), Ospedale L. Sacco (Milan), Ospedale S. Raffaele Turro (Milan), IRCCS S. Giovanni di Dio (Brescia) and University of Turin, including 194 patients with FTD (87 males and 107 females, mean age at onset ±SEM: 65.8 ± 0.74 years), 14 with PA (seven males and seven females, mean age at onset ± SEM: 68.3 ± 2.48 years) and 12 with SD, (eight males and four females, mean age at onset ±SEM: 66.3 ± 3.24 years). Ten patients reported a documented family history for dementia; whereas, 56 patients had a reported, but not documented, family history. Cognitive dysfunctions were assessed by the clinical dementia rating, the Mini Mental State Examination (MMSE), the frontal assessment battery, the Wisconsin Card Sorting Test, and the Tower of London test. Diagnosis of bvFTD, PA and SD met current criteria [3]. The control group consisted of 229 non-demented volunteers matched for ethnic background and age (78 males and 151 females, mean age ± SEM: 66.8 ± 0.74 years) without memory and psycho-
871
behavioral dysfunctions (MMSE ‡28). The age of controls didnÕt significantly differ from that of patients (P > 0.05). The study has been approved by the Institutional Review board of the Department of Neurological Sciences, University of Milan, Fondazione Ospedale Maggiore Policlinico, Mangiagalli e Regina Elena. Informed consent to participate in this study was given by all subjects or their caregivers. The entire open reading frame of PGRN exons 1-12, including for each exon at least 80 base-pairs of exonintron boundaries, was sequenced using specific primers, as previously described [5]. Furthermore a total of 638 base-pairs in the 5¢ untranslated region, including the non-coding exon 0, were additionally sequenced. In addition, MAPT exons 9–13 were sequenced as previously described [9]. One patient with PA and two with bvFTD were carriers of a deletion in PGRN previously reported [10] and were therefore excluded by the study. No carriers of known MAPT mutations were found. Mutation scanning of DCUN1D1 exons and flanking regions was carried out in 20 patients by direct sequencing [11], using specific primers (available upon request). rs4859147 SNP was analyzed using Taqman methodology [11], by using a specific Assay-on-demand (ABI assay ID: C_25764476 _10). rs4859146 SNP was detected by using a novel PCRRestriction Fragment Length Polymorphisms protocol. Two primers were used to amplify a region of 188 bp, containing the SNP (Fwd: 5¢-GAT GGC ATA CAG CAG TTC TG-3¢; Rev: 5¢-CAG TTC CAC AGT AAC ATA TC-3¢). The fragment was then digested with HpyCH4V (New England Biolabs, Beverly, MA, USA). The unique restriction site, present in wild-type allele only, was determined using the Webcutter 2.0 software (http://www.firstmarket.com/cutter). PCR product (10 ll) was digested with 3 U of HpyCH4V overnight at 37C. Digested products yielded two fragments (68 and 120 bp) visualized on a 3.0% agarose gel stained with ethidium bromide. This new protocol produced clearcut genotyping results (Fig. 1). In few doubtful cases (n < 10), sequencing was carried out to confirm RFLP genotyping. Allelic and genotypic frequencies were obtained by direct counting. Chi square test was used to test for Hardy Weinberg Equilibrium (HWE). Haploview 3.2 software (Broad Institute of Harvard and Mit, Cambridge, MA, USA) was used to test for linkage disequilibrium and for differences in haplotype distribution between cases and controls. Statistical significance was estimated empirically using the bootstrap function in Haploview. Bootstrap P-values are calculated using 10 000 bootstrap samples. Calculation of DÕ is based on block definition by Gabriel et al. [12].
2009 The Author(s) Journal compilation 2009 EFNS European Journal of Neurology 16, 870–873
872
C. Villa et al.
1
2
3
Table 1 Allelic and genotypic frequencies of DCUN1D1 rs4859146 SNP, given as n(%), in patients and controls
4
DCUN1D1 rs4859146 Allele A G Genotype A/A A/G G/G
188 bp 120 bp
Patients (n = 217)
Controls (n = 229)
306 (70.5) 128 (29.5)
345 (75.3) 113 (24.7)
104 (47.9) 98 (45.2) 15 (6.9)*
120 (52.4) 105 (45.9) 4 (1.7)
*P = 0.011, OR: 4.39, 95% CI: 1.40–13.78 versus controls; three carriers of a known PGRN deletion not included.
68 bp
Table 2 Allelic and genotypic frequencies of DCUN1D1 rs4859147 SNP, given as n(%), in patients and controls
Figure 1 Polymerase chain reaction-based genotyping of the HpyCH4V polymorphism of A/G in exon 2 of DCUN1D1 gene. Lane 1: A/A; lane 2: A/G; lane 3: G/G; lane 4: DNA weight marker VIII (Boehriringer-Roche, Mannheim, Germany).
A multivariate logistic regression analysis, adjusted for age and gender, was carried out with the SPSS software version 13.0 (SPSS for Windows, Rel. 13.0; SPSS Inc., Chicago, IL, USA), and adjusted odds ratios (OR) with related 95% confidence interval (CI). According to the BonferroniÕs correction, the threshold for significance was set at P < 0.025. Sigmastat 3.1 (Systat Software GmbH, Erkrath, Germany) was used for an estimation of the power of study.
Results One patient with PA and two with bvFTD were carriers of a known deletion in PGRN [10] and were therefore excluded by the analysis. All remaining patients were sporadic and non-carriers of mutations in PGRN. DCUN1D1 mutation scanning in a sub-population of 20 patients demonstrated the presence of two SNPs: rs4859146 in exon 3 and rs4859147 in exon 3-5¢-flanking region. Subsequently, the frequency of these SNPs was determined in the overall population of 217 patients and 229 controls. The control population was not in HWE for rs4859146 (P = 0.035); whereas, the HWE was respected in patients for the same SNP (P = 0.506) as well as in patients and controls for rs4859147 (P = 0.586 and P = 0.998, respectively). No statistically significant differences in DCUN1D1 rs4859146 and rs4859147 allelic frequencies between cases and controls were found (29.5 vs. 24.7, P = 0.1268, adjusted OR = 1.26, 95%, CI = 0.94–1.69, Table 1
DCUN1D1 rs4859147 Allele G A Genotype G/G G/A A/A
Patients (n = 217)*
Controls (n = 229)
281 (64.7) 153 (35.3)
295 (64.4) 163 (35.6)
87 (40.0) 107 (49.3) 23 (10.7)
95 (41.5) 105 (45.9) 29 (12.6)
*Three carriers of a known PGRN deletion not included.
and 35.3 vs. 35.6%, P = 0.92, OR = 1.02, 95% CI = 0.77–1.33, Table 2). Nevertheless, a statistically significant increased frequency of the DCUN1D1 rs4859146 SNP GG genotype was observed in patients as compared with controls (6.9 vs. 1.7%, P = 0.011, OR: 4.39, 95%, CI: 1.40–13.78, power = 0.782, Table 1). Stratifying patients according to the clinical syndrome, significant differences were found between patients with bvFTD and controls (GG frequency: 6.3 vs. 1.7%, P = 0.02, OR:4.0, 95%, CI = 1.24–12.92, power = 0.675), as well as between patients with PA compared with controls (15.4 vs. 1.7%, P = 0.014, OR = 11.30, 95%, CI = 1.63–78.45, power = 0.886), but not in patients with SD versus controls (8.3 vs. 1.7%, P = 0.18, OR = 5.24, 95% C.I. = 0.45–60.63, power = 0.327) possibly due to the limited sample size of the subgroup. Stratifying for gender, no differences were observed for both polymorphisms (data not shown). As regards haplotype analysis, the two SNPs of DCUN1D1 had a DÕ of 0.86 (95% C.I.: 0.81–0.92), and, according to the method implemented in Haploview, they were not recognized as a block. However, when the two SNPs were analysed together as a block, the combination AA were found more frequent in controls than cases (13.4% vs. 7.9%; P = 0.02).
2009 The Author(s) Journal compilation 2009 EFNS European Journal of Neurology 16, 870–873
DCUN1D1 is a risk factor for FTLD
Discussion According to these results, the GG genotype of the DCUN1D1 rs4859146 SNP represents a susceptibility factor for FTLD, increasing the risk to develop the disease of about fourfold. Considering each syndrome separately, the significance was maintained in patients with bvFTD as well as in patients with PA. The frequency of the G allele was particularly increased also in patients with SD, although the significance threshold was not reached. Nevertheless, the numbers for PA (n = 14) and SD (n = 12) were very small and therefore lack any statistical power to make definitive statements on the role of DCUN1D1 rs4859146 SNP in these diseases. Our control population was not in HWE for the rs4859146 SNP. A possible explanation for this is that controls have been selected in a very small region of Northern Italy (all by the same Center), whereas patients have been recruited in a widen region by five centers. Thus, patients are far more representative of the global Italian population than controls. However, to get definitive results, a replication study on a larger and independent population is needed. The mechanism by which this SNP determines an increased risk is still unknown. However, it should be taken into account that this polymorphism does not lead to an aminoacidic change; therefore, theoretically the function of the protein is not affected. Thus, the association observed could be due to other allelic variants located in proximity to the one studied. Regarding the presence of other potential functional variants, to date only a polymorphism in exon 5, named rs36086481, has been reported, although no population frequency data are available and there is no validation (http://www.ensembl.org). We didnÕt find this SNP in 20 patients sequenced, thus its frequency is likely very low. Notably, all patients included in this study were negative for the presence of causal mutations in PGRN gene. In our clinical FTLD series, we found only three patients carrying a causal PGRN deletion [10], in accordance with very recent data demonstrating that the frequency of PGRN mutations is very rare (1.64%) in Italian population [13]. Although patients included in this study have a clinical diagnosis of FTLD, none of them underwent post-mortem pathological analysis; therefore, a diagnosis of frontal variant AD canÕt be excluded [14]. In conclusion, according to these findings, the GG genotype of the DCUN1D1 rs4859146 SNP likely acts as risk factor for sporadic FTLD. Nevertheless, a proper analysis of the haplotype block structure of DCN-1 is required to illuminate the nature of this
873
possible association, with addition genotyping of appropriate tagging SNPs, together with the neuropathology confirmation of clinical diagnoses.
Acknowledgements This work was supported by grants from Associazione ‘‘Amici del Centro Dino Ferrari’’, Monzino Foundations, IRCCS Ospedale Maggiore Milano, ‘‘Associazione per la Ricerca sulle Demenze (ARD)’’, and Ing. Cesare Cusan.
References 1. Graff-Radford NR, Woodruff BK. Frontotemporal dementia. Semin Neurol 2007; 27: 48–57. 2. Hou CE, Carlin D, Miller BL. Non-AlzheimerÕs disease dementias: anatomic, clinical, and molecular correlates. Can J Psychiatry 2004; 49: 164–171. 3. Neary D, Snowden JS, Gustafson L, et al. Frontotemporal lobar degeneration: a consensus on clinical diagnostic criteria. Neurology 1998; 51: 1546–1554. 4. Scarpini E, Galimberti D, Guidi I, Bresolin N, Scheltens P. Progressive, isolated language disturbance: its significance in a 65-year-old-man. A case report with implications for treatment and review of literature. J Neurol Sci 2006; 240: 45–51. 5. Gass J, Cannon A, Mackenzie IR, et al. Mutations in progranulin are a major cause of ubiquitin-positive frontotemporal lobar degeneration. Hum Mol Genet 2006; 15: 2988–3001. 6. Rademakers R, Cruts M, van Broekhoven C. The role of tau (MAPT) in frontotemporal dementia and related tauopathies. Hum Mutat 2004; 24: 277–295. 7. Boeve BF, Hutton M. Refining frontotemporal dementia with parkinsonism linked to chromosome 17. Arch Neurol 2008; 65: 460–464. 8. Kurz T, O¨zlu¨ N, Rudolf F, et al. The conserved protein DCN-1/Dcn1p is required for cullin neddylation in C. elegans and S. cerevisiae. Nature 2005; 435: 1257–1261. 9. Rizzu P, Van Swieten JC, Joosse M, et al. High prevalence of mutations in the microtubule-associated protein tau in a population study of frontotemporal dementia in the Netherlands. Am J Hum Genet 1999; 64: 414–421. 10. Benussi L, Binetti G, Sina E, et al. A novel deletion in progranulin gene is associated with FTDP-17 and CBS. Neurobiol Aging 2008; 29: 427–435. 11. Cortini F, Fenoglio C, Guidi I, et al. Novel exon 1 progranulin gene variant in AlzheimerÕs disease. Eur J Neurol 2008; 15: 1111–1117. 12. Gabriel SB, Schaffner SF, Nguyen H, et al. The structure of haplotype blocks in the human genome. Science 2002; 296: 2225–2229. 13. Borroni B, Archetti S, Alberici A, et al. Progranulin genetic variations in frontotemporal lobar degeneration: evidence for low mutation frequency in Italian clinical series. Neurogenetics 2008; 9: 197–205. 14. Johnson JK, Head E, Kim R, Starr A, Cotman CW. Clinical and pathological evidence for a frontal variant of Alzheimer disease. Arch Neurol 1999; 56: 1233–1239.
2009 The Author(s) Journal compilation 2009 EFNS European Journal of Neurology 16, 870–873
