Abstract
Genetic variations represent major risk factors for Alzheimer's disease (AD). While familial early onset AD is associated with mutations in the amyloid precursor protein and presenilin genes, only the e4 allele of the apolipoprotein E (APOE) gene has so far been established as a genetic risk factor for late onset familial and sporadic AD. It has been suggested that the C→T (224Ala→Val) transition within exon 2 of the cathepsin D gene (CTSD) might represent a risk factor for late onset AD. The objective of this study was to investigate whether possession of the CTSD exon 2 T allele increases the risk of developing AD, and to determine whether this modulates the amyloid pathology of the disease in conjunction with, or independent of, the APOE e4 allele. Blood samples were obtained from 412 patients with possible or probable AD and brain tissues from a further 148 patients with AD confirmed by postmortem examination. CTSD and APOE genotyping were performed by PCR on DNA extracted from blood, or from frontal cortex or cerebellum in the postmortem cases. Pathological measures of amyloid β protein (Aβ), as plaque Aβ40 and Aβ42(3) load and degree of cerebral amyloid angiopathy were made by image analysis or semiquantitative rating, respectively. CTSD genotype frequencies in AD were not significantly different from those in control subjects, nor did these differ between cases of early or late onset AD or between younger and older controls. There was no gene interaction between the CTSD T and APOE e4 alleles. The amount of plaque Aβ40 was greater in patients carrying the CTSD T allele than in non‐carriers, and in patients bearing APOE e4 allele compared with non‐carriers. Possession of both these alleles acted synergistically to increase levels of plaque Aβ40, especially in those individuals who were homozygous for the APOE e4 allele. Possession of the CTSD T allele had no effect on plaque Aβ42(3) load or degree of CAA. Possession of the CTSD T allele does not increase the risk of developing AD per se, but has a modulating effect on the pathogenesis of the disorder by increasing, in concert with the APOE e4 allele, the amount of Aβ deposited as senile plaques in the brain in the form of Aβ40.
Keywords: Alzheimer's disease, cathepsin D gene, apolipoprotein E gene, amyloid β protein, genotype/phenotype correlation
Although initial reports1,2 described a strong association between the C→T (224Ala→Val) transition in exon 2 of the cathepsin D gene (CTSD) and Alzheimer's disease (AD), this finding remains to be replicated.3,4,5,6 Indeed, a recent meta‐analysis7 concluded that the exon 2 CTSD polymorphism is not a major risk factor for AD, although it might, as a disease modifying factor, enhance the effects of the apolipoprotein E (APOE) e4 allele.
Cathepsin D is an aspartyl protease that, in vitro, can cleave amyloid precursor protein into amyloid β protein (Aβ) at both the β‐secretase8,9 and γ‐secretase10 sites. Being functional,11 the exon 2 CTSD C→T polymorphism might modulate the course of AD by increasing Aβ precursor protein (APP) breakdown, thereby promoting Aβ pathology. In this study, we investigated whether the exon 2 CTSD T allele influences the pathological phenotype of AD by modulating brain amyloid plaque load or extent of cerebral amyloid angiopathy (CAA).
MATERIALS AND METHODS
DNA was extracted by routine methods from 2 ml EDTA blood samples taken from 414 patients with AD. These patients were ascertained either through clinical old age psychiatry (OAP) services in Manchester (100 individuals) with diagnosis of AD being made according to DSM III‐R Criteria,13 or through outpatient clinics at the cerebral function unit (CFU) of Greater Manchester Neurosciences Centre, Hope Hospital (314 individuals) in whom diagnosis of AD was consistent with NINCDS‐ADRDA criteria.14 DNA was also extracted from frozen cerebral cortex or cerebellum from a further 146 individuals who had died from AD, verified by postmortem examination according to CERAD Criteria.15 Of these patients, 92 had been identified through the CFU, 45 through OAP services, and 9 through tissue donations under the auspices of the Alzheimer's Society (UK). Twelve of the deceased patients had been investigated in the CFU during their lifetime but were not doubly represented within the postmortem examination group. AD cases were stratified into early onset AD (EOAD; 317 patients, mean (SD) age at onset 56.5 (5.6) years) and late onset AD (LOAD; 243 patients, mean age at onset 72.8 (5.9) years) groups. Control data were derived from blood samples taken from a cohort of 767 mentally normal people aged >50 years resident within the same Greater Manchester region from which the patients with AD were drawn.16 All subjects were white. All blood and brain samples were collected with the approval of the local ethics research committee.
The exon 2 CTSD C→T polymorphism and APOE genotype were determined by PCR as described previously.16,17 Given a CTSD T allele frequency of 0.1, there was at least 80% power to detect an effect size of 1.6 or greater for AD and control cohorts, an effect size of 1.8 or greater for the EOAD cohort and younger controls, and an effect size of 2.1 or greater for the LOAD cohort and the older controls.
Brain levels of Aβ40 or Aβ42(3) were measured by image analysis following immunohistochemical staining of paraffin wax embedded sections of frontal cortex (thickness 6 μm) using BA27 and BC05 antibodies for Aβ40 or Aβ42(3), respectively, and the extent of CAA within the brain was assessed semiquantitatively in Aβ immunostained sections using 4G8 antibody (Signet Laboratories, Waltham MA), as previously described.18,19
Statistical analysis
Logistic regression analysis was performed using Stata (2001) software to determine any synergistic interaction between APOE alleles and the CTSD C→T polymorphism.
RESULTS
CTSD genotype and allele frequencies for AD and control subjects are given in table 1. The frequencies of the CTSD CC, CT, and TT genotypes did not differ significantly between the AD and control groups (χ2 = 3.8; p = 0.148), nor did the frequency of T allele carriers (that is, the total number of individuals with CT or TT genotype) (χ2 = 0.03; p = 0.867). However, there was a trend for the TT genotype to be more common in AD than controls (Fisher's exact test, p = 0.077). There were no significant differences between CTSD genotype frequencies when patients with AD were stratified into EOAD and LOAD groups (χ2 = 0.1; p = 0.743), when patients with EOAD were compared with younger control subjects (χ2 = 2.3; p = 0.321), or when patients with LOAD were compared with older control subjects (χ2 = 1.5; p = 0.463). As expected, the proportion of people bearing at least one APOE e4 allele was significantly higher (χ2 = 34.8; p<0.001) in the AD group (63%) than in the control group (27%). There was no synergistic interaction between the CTSD C→T polymorphism and the APOE e4 allele. CTSD genotype frequencies were not significantly different between APOE e4 allele carriers and non‐carriers in the whole AD cohort, or when stratified into EOAD and LOAD groups (data not shown).
Table 1 CTSD genotype and allele frequencies in 560 AD and 767 control subjects.
| Cohort | n | CTSD genotypes, n (%) | Alleles | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CC | CT | TT | C | T | ||||||||
| All AD | 560 | 475 (84.9) | 79 (14.1) | 6 (1.1) | 1029 (91.9) | 91 (8.1) | ||||||
| EOAD | 317 | 270 (85.1) | 44 (13.6) | 3 (1.0) | 574 (92.2) | 50 (7.8) | ||||||
| LOAD | 243 | 205 (84.3) | 35 (14.4) | 3 (1.3) | 445 (91.6) | 41 (8.4) | ||||||
| All controls | 767 | 648 (84.5) | 117 (15.3) | 2 (0.3) | 1413 (92.0) | 121 (8.0) | ||||||
| Younger controls | 496 | 422 (85.1) | 73 (14.7) | 1 (0.2) | 917 (92.4) | 75 (7.6) | ||||||
| Older controls | 270 | 225 (83.3) | 44 (16.3) | 1 (0.4) | 494 (91.5) | 46 (8.5) | ||||||
The AD group was stratified into EOAD and LOAD cohorts and the control group into younger and older control groups.
In the deceased AD patients, there were no significant differences in the amount of plaque Aβ42(3) or extent of CAA within the brains of CTSD T allele carriers compared with C allele carriers, or according to APOE genotype, or between carriers and non‐carriers of APOE e4 allele (table 2). However, the amount of plaque Aβ40 differed significantly between APOE genotype groups (F5,127 = 11.2; p<0.001), being significantly greater (p<0.001) in APOE e4 allele carriers than in non‐carriers (table 2). Furthermore, plaque Aβ40 was also significantly greater (p = 0.006) in CTSD T allele carriers than in non‐carriers (table 2). Most importantly, levels of plaque Aβ40 levels differed significantly across the 4 APOE e4 allele/CTSD T allele combination groups (F3,127 = 7.6; p<0.001); possession of CTSD T allele significantly augmented the level of Aβ40 in both carriers and non‐carriers of APOE e4 allele (table 2). Indeed, when individuals homozygous for APOE e4 allele were investigated separately, not only was plaque Aβ40 higher in CTSD T allele carriers (10.5 (5.1)) than in non‐carriers (6.8 (4.5)), but those individuals had the highest Aβ40 levels of any of the APOE e4 allele/CTSD T allele combination groups. Because of the relatively small number of patients involved, it was not possible to confirm this finding statistically.
Table 2 Pathological measures stratified according to APOE and CTSD genotypes and alleles, separately and in combination.
| Aβ40 | Aβ42 | CAA | ||||
|---|---|---|---|---|---|---|
| APOE genotype | ||||||
| e2/e2 (1) | 1.2 | 15.2 | 9.0 | |||
| e2/e3 (8) | 2.2 (1.7) | 10.2 (5.4) | 6.3 (1.5) | |||
| e2/e4 (2) | 5.4 (2.4) | 11.2 (0.7) | 6.5 (3.5) | |||
| e3/e3 (48) | 2.2 (2.2) | 9.6 (4.9) | 7.3 (2.5) | |||
| e3/e4 (60) | 3.2 (2.7) | 9.8 (5.3) | 7.1 (3.1) | |||
| e4/e4 (27) | 7.8 (5.0) | 9.8 (3.1) | 8.7 (2.8) | |||
| e4− (57) | 2.1 (2.1) | 9.8 (5.0) | 7.3 (2.4) | |||
| e4+ (89) | 4.5 (4.0) | 9.8 (4.7) | 7.5 (3.1) | |||
| CTSD genotype | ||||||
| CC (127) | 3.3 (3.2) | 9.7 (4.8) | 7.5 (2.9) | |||
| CT (19) | 5.9 (5.2) | 10.5 (4.1) | 7.1 (2.4) | |||
| APOE/CTSD combination | ||||||
| e4− T− (77) | 2.1 (2.0) | 9.4 (4.8) | 7.4 (2.4) | |||
| e4− T+ (6) | 3.8 (2.5) | 12.0 (5.5) | 6.5 (3.1) | |||
| e4+ T− (50) | 4.1 (3.5) | 9.8 (4.9) | 7.6 (3.2) | |||
| e4+ T+ (13) | 7.0 (5.9) | 9.7 (3.3) | 7.3 (2.2) | |||
Data are means (SD).
DISCUSSION
In this study, CTSD genotype and allele frequencies in AD did not differ from controls in either EOAD or LOAD (see also Ntais et al7). Although it has been suggested1,2,6,7 that possession of CTSD T allele might confer a slightly increased risk of AD through enhancement of APOE e4 allelic effects, no (genetic) interaction between CTSD T allele and APOE e4 allele either overall, or in EOAD and LOAD was found (see also Bhoja et al, Crawford et al, Ingegni et al3,4,5).
Consistent with previous reports,20,21 plaque Aβ40 was increased in APOE e4 allele carriers. Importantly, plaque Aβ40 was also increased in CTSD T allele carriers, and there was a synergistic effect between APOE e4 allele and CTSD T allele; carriers of both had a greater deposition of Aβ40 than carriers of either allele alone, or non‐carriers of both. Consequently, the highest levels of Aβ40 were present in patients carrying the CTSD T allele who were also homozygous for the APOE e4 allele. CTSD T allele did not influence plaque Aβ42(3) load or extent of CAA (see above) or phosphorylated tau protein, degree of astrocytosis and microgliosis, or vessel arteriosclerosis (unpublished data).
The mechanism underlying this biological synergy is unclear. The APOE E4 isoform may reduce the threshold for fibrillisation of Aβ, permitting a greater seeding of Aβ40 upon pre‐existing deposits of the more highly fibrillogenic Aβ42. Another possibility, as suggested by studies in transgenic mice,22 is that APOE e4 allele carriers, especially homozygous individuals, have a reduced ability to clear Aβ from the brain compared with homozygous and heterozygous carriers of APOE e3 allele. The T allelic variant of CTSD is functionally more active than the C allelic form,11 and the protein isoform transcribed by T allele may catabolise APP more readily at the β‐secretase and γ‐secretase sites to yield greater amounts of both Aβ40 and Aβ42.8,9,10 This might lead to an increased availability of Aβ40 within the extracellular space, which could enable a greater degree of seeding of this particular isoform of Aβ upon a nidus of pre‐existing Aβ42 than is possible when the APOE E4 isoform alone is present.
One limitation of the present study is that only a single single nucleotide polymorphism (SNP) within CTSD was investigated. There are at least 18 other SNPs in CTSD (see Majores et al12 and www.ncbi.nlm.nih.gov). Two silent mutations occur in exons 3 and 4, and there are two polymorphisms in introns 5 and 8.12 Other SNPs occur in the untranslated reading frame. However, we focused upon the exon 2 SNP for two reasons. Firstly, because this polymorphism is known to be functional in a manner biologically relevant to AD.11 Secondly, the exon 2 coding polymorphism is in linkage disequilibrium with both SNPs in exons 3 and 4, and with the SNP in intron 8, at least, generating two haplotypes.12 Although the SNP in intron 5 generates a further haplotype, this SNP is not considered to be functional, at least in terms of influencing RNA splicing.12 We believe our conclusion relating to the biological effect of CTSD exon 2 polymorphism on amyloid pathology is valid, although we have not examined potential effects of other variations in CTSD.
Future work should assess the effects of CTSD haplotype on amyloid pathology in order to elucidate the genetic elements responsible for exacerbating this facet of AD pathology.
ACKNOWLEDGEMENTS
This study was performed under local ethics agreements (Salford and Trafford 05/Q1405/24, South Manchester SOU/98/132 and University of Manchester Research on Human Beings reference no. 98014).
Abbreviations
Aβ - amyloid beta
AD - Alzheimer's disease
APP - amyloid β precursor protein
CAA - cerebral amyloid angiopathy
CFU - cerebral function unit
CTSD - cathepsin D
EOAD - early onset AD
LOAD - late onset AD
OAP - old age psychiatry
SNP - single nucleotide polymorphism
Footnotes
Competing interests: none
References
- 1.Papassotiropoulos A, Bagli M, Feder O.et al Genetic polymorphism of cathepsin D is strongly associated with the risk for developing sporadic Alzheimer's disease. Neurosci Lett 1999262171–174. [DOI] [PubMed] [Google Scholar]
- 2.Papassotiropoulos A, Bagli M, Kurz A.et al A genetic variation of cathepsin D is a major risk factor for Alzheimer's disease. Ann Neurol 200047399–403. [PubMed] [Google Scholar]
- 3.Bhojak T J, DeKosky S T, Ganguli M.et al Genetic polymorphisms in the cathepsin D and interleukin‐6 genes and the risk of Alzheimer's disease. Neurosci Lett 200028821–24. [DOI] [PubMed] [Google Scholar]
- 4.Crawford F C, Freeman M J, Schinka J.et al The genetic association between cathepsin D and Alzheimer's disease. Neurosci Lett 200028961–65. [DOI] [PubMed] [Google Scholar]
- 5.Ingegni T, Nocentini G, Mariani E.et al Cathepsin D polymorphism in Italian elderly subjects with sporadic late‐onset Alzheimer's disease. Dement Geriatr Cogn Disord 200316151–155. [DOI] [PubMed] [Google Scholar]
- 6.Li X‐Q, Chen D, Zhang Z‐X.et al Association between cathepsin D polymorphism and Alzheimer's disease in a Chinese Han population. Dement Geriatr Cogn Disord 200418115–119. [DOI] [PubMed] [Google Scholar]
- 7.Ntais C, Polycarpou A, Ionnadis J P A. Meta‐analysis of the association of the cathepsin D Ala224Val gene polymorphism with the risk of Alzheimer's disease: A HuGE gene‐disease association review. Am J Epidemiol 2004159527–536. [DOI] [PubMed] [Google Scholar]
- 8.Evin G, Cappai R, Li Q‐X.et al Candidate gamma‐secretases in the generation of the carboxyl terminus of the Alzheimer's disease beta A4 amyloid: possible involvement of cathepsin D. Biochemistry 19953414185–14192. [DOI] [PubMed] [Google Scholar]
- 9.Chevallier N, Vizzavona J, Marambaud P.et al Cathepsin D displays in vitro beta secretase‐like specificity. Brain Res 199775011–19. [DOI] [PubMed] [Google Scholar]
- 10.Sadik G, Kaji H, Takeda K.et al In vitro processing of amyloid precursor protein by cathepsin D. Int J Biochem Cell Biol 1999311327–1337. [DOI] [PubMed] [Google Scholar]
- 11.Touitou I, Capony F, Brouillet J P.et al Missense polymorphism (C/T224) in the human cathepsin D pro‐fragment determined by polymerase chain reaction single strand conformational polymorphism analysis and possible consequences in cancer cells. Eur J Cancer 199430A390–394. [DOI] [PubMed] [Google Scholar]
- 12.Majores M, Kolsch H, Bagli M.et al Screening for new polymorphisms using single‐strand conformation polymorphism analysis. Int J Molec Med 20029185–187. [PubMed] [Google Scholar]
- 13.American Psychiatric Association Diagnostic and statistical manual for mental disorders, 3rd ed, revised. Washington, DC: American Psychiatric Association, 1987
- 14.McKhann G, Drachman D, Folstein M.et al Clinical diagnosis of Alzheimer's disease: Report of the NINCDS‐ADRDA work group under the auspices of the Department of Health and Human Services Task Force on Alzheimer's disease. Neurology 198434939–944. [DOI] [PubMed] [Google Scholar]
- 15.Mirra S S, Heyman A, McKeel D. The consortium to establish a registry for Alzheimer's disease. Part II. Standardization of the neuropathologic assessment of Alzheimer's disease. Neurology 199141479–486. [DOI] [PubMed] [Google Scholar]
- 16.Payton A, Holland F, Diggle P.et al Cathepsin D exon 2 polymorphism associated with general intelligence in a healthy older population. Mol Psychiatry 2003814–18. [DOI] [PubMed] [Google Scholar]
- 17.Wenham P R, Price W H, Blundell G. Apolipoprotein E genotyping by one‐stage PCR. Lancet 19913371158–1159. [DOI] [PubMed] [Google Scholar]
- 18.Iwatsubo T, Odaka N, Suzuki N.et al Visualization of Aβ42(43)‐positive and Aβ40‐positive senile plaques with specific Aβ monoclonal antibodies: evidence that an initially deposited species is Aβ1‐42(43). Neuron 19941345–53. [DOI] [PubMed] [Google Scholar]
- 19.Tian J, Shi J, Bailey K.et al Relationships between arteriosclerosis and cerebral amyloid angiopathy and myelin loss from the cerebral cortical white matter in Alzheimer's disease. Neuropathol Appl Neurobiol 20033046–56. [DOI] [PubMed] [Google Scholar]
- 20.Mann D M A, Iwatsubo T, Pickering‐Brown S M.et al Preferential deposition of amyloid β protein (Aβ) in the form Aβ40 in Alzheimer's disease is associated with a gene dosage effect of the apolipoprotein E E4 allele. Neurosci Lett 199722181–84. [DOI] [PubMed] [Google Scholar]
- 21.Gearing M, Mori H, Mirra S S. Aβ peptide length and apolipoprotein E genotype in Alzheimer's disease. Ann Neurol 199639395–399. [DOI] [PubMed] [Google Scholar]
- 22.Fagan A M, Watson M, Parsadanian M.et al Human and murine Apo E markedly alters A beta metabolism before and after plaque formation in a mouse model of Alzheimer's disease. Neurobiol Dis 20029305–318. [DOI] [PubMed] [Google Scholar]