Bleomycin hydrolase (BH), a cysteine protease from the papain superfamily, is considered to be a candidate for the β-secretase, which is presumably involved in the production of β-amyloid peptide. The G/G genotype of BH was identified as a significant risk factor for the development of Alzheimer's disease (AD) in subjects not carrying the apolipoprotein E ε4 allele (apoE-ε4). However, this finding was recently challenged. We studied this polymorphism in a homogenous sample of German AD patients and controls. The over-representation of the G/G genotype in AD patients could be confirmed, however it was more pronounced in apoE-ε4 carriers. Additional studies should be undertaken to increase the confidence that the BH polymorphism is associated with AD and to explore the relationship between BH and apoE.
The mode of inheritance of Alzheimer's disease (AD) is complex. Numerous studies have shown that interactions between genetic and environmental factors modify the risk for the development of AD.1 The presence of at least one apolipoprotein E (apoE) ε4 allele (apoE-ε4) is hitherto the only definite genetic risk factor for sporadic AD.2
The aggregation of β-amyloid peptide (Aβ) is considered to be a key event in AD pathogenesis. Consequently, genes coding for proteins involved in the metabolism of Aβ are interesting candidates when studying possible associations with the risk of AD.
Bleomycin hydrolase (BH), a cysteine protease from the papain superfamily, is considered to be a candidate for the β-secretase, which is presumably involved in the production of Aβ. The gene coding for BH consists of 12 exons and is located at 17q11.1–11.2. An A1450G polymorphism results in an I443V amino acid substitution in the C-terminus of the protein.3 Recently, Montoya and colleagues identified the G/G genotype as a significant risk factor for the development of AD.4 The increased risk for AD in individuals homozygous for the G allele was confined to apoE-ε4 non-carriers. Using a large, pooled sample of Caucasian AD patients and controls, Farrer and colleagues failed to confirm the association between the G/G genotype of the BH polymorphism and AD.5 Surprisingly, the authors found a protective effect of the G/G genotype among apoE-ε4 carriers. Both studies used large samples with high statistical power, however subjects were recruited in many different institutions and then pooled for statistical testing. This could have resulted in a selection and stratification bias6 which could partially explain the controversial results. Therefore, we investigated whether the G/G genotype of the BH polymorphism is increased in AD in a homogenous German Caucasian population.
BH genotypes and allele frequencies were determined for 102 AD patients and 351 control subjects (ie 191 healthy subjects and 160 depressed patients) (Table 1). BH genotypes were in Hardy–Weinberg equilibrium in the control group (P > 0.2), whereas this was not the case in the patient group (P = 0.04) because of an excess of A/G heterozygotes. Based on the genotype frequencies reported by Montoya et al,4 our study had a power of 70% for α = 0.05. The frequency of the G/G genotype in AD patients (12.7%) was increased when compared to the controls (7.4%, P = 0.05). The G allele was over-represented in AD patients (40.2%) compared to controls (32.9%) (P = 0.033). The genotypic distribution in our sample was very similar to that reported by Montoya and colleagues (G/G frequency in AD patients 12.7%, in controls 6.6%).4
When stratified for the apoE genotype, the increase in the frequency of the G/G genotype in AD patients was only observed in carriers of the apoE-ε4 allele (19.4% vs 3.1%, P = 0.001). The odds ratio (OR) for developing AD in apoE-ε4 positive subjects homozygous for the G allele was 7.5 (95% confidence interval (CI): 2.0–27.9). Among non-carriers of the apoE-ε4 allele there was an under-representation of the G/G genotype in AD patients (2.5%) compared to controls (9.1%), without however reaching statistical significance (P = 0.06).
Age- and gender-adjusted logistic regression confirmed the positive association between the G/G genotype and AD in apoE-ε4 carriers (P = 0.019, adjusted OR = 6.1, 95% CI = 1.3–27.9). The G/G genotype could not be identified as a risk factor for AD in apoE-ε4 non-carriers (P > 0.3). There was no interaction between BH and apoE genotype on AD risk (P = 0.19 for the interaction term).
It should be noted that the two independent control samples (ie healthy subjects and depressed patients) had almost identical BH genotypic distribution (G/G frequency in healthy controls 7.8%, in depressed patients 6.2%, P > 0.2).
Our study confirms the role of the G/G genotype of BH as a risk factor for AD as originally reported by Montoya et al.4 However, the role of the G/G genotype as a risk factor for AD was pronounced in carriers of the apoE-ε4 allele which is in contrast to the findings of Montoya et al, where the increased risk was observed in non-carriers of the apoE-ε4 allele. Our results suggest that there is no interaction between BH and apoE genotype on the risk for AD. Consequently, possible differences of the BH genotypic distribution in subjects with different apoE genotypes can be attributed to stratification effects. This is also supported by the lack of pathophysiologic evidence showing an interaction between apoE and BH.
Our results are also in discordance with the study of Farrer et al who failed to replicate the observed association in a larger sample, reporting a possible protective effect of the G/G genotype among apoE-ε4 carriers.5 Large, pooled samples from different diagnostic centers are prone to stratification bias.6 The failure to confirm the association between BH genotype and AD could be a consequence of stratification bias, as also mentioned by the authors of the latter study. Assuming a rather small effect of the G/G genotype on AD risk, different observations may be a result of even smaller biases in the sample. This could also explain the inconsistencies concerning the relationship between BH and apoE because of the necessary additional stratification in apoE-ε4 carriers and non-carriers. Pathologically confirmed cases had a higher G/G frequency compared to clinically diagnosed cases,4 indicating that misclassification might give rise to false negative results because of the reduction of the effect size. Although smaller in size, our sample had the advantage of being geographically and ethnically homogenous, and having been recruited in one institution. Moreover, the possibility of false-positive results was minimized by the inclusion of two independent control samples (ie healthy controls and depressed inpatients).
In conclusion, this is the first study which confirms the positive association between BH genotype and risk of sporadic AD. Additional studies should be undertaken to increase the confidence that the BH polymorphism is associated with AD and to explore the relationship between BH and apoE.
A total of 102 German Caucasian AD patients with a mean age of 74.4 ± 10.3 years (range 51–101 years) were recruited from the outpatient memory disorders clinic of the Psychiatric Department of the University of Bonn. Sixty-eight (66.7%) of the patients were females. The diagnosis of AD followed NINCDS-ADRDA criteria.7
The control group comprised 191 healthy subjects (51.8% females) randomly selected from the elderly general population of the city of Bonn. The mean age was 70.6 ± 11.4 years (range 50–100 years).
Association studies are vulnerable for false positive associations which can derive from the recruitment of ‘super healthy’ control samples. Consequently, a second and independent sample of 160 non-demented, depressed hospitalized patients was recruited in the Department of Psychiatry, University of Bonn (65.6% females). Mean age was 68.0 ± 7.7 years (range: 50–88 years). Like AD patients, healthy subjects and depressed patients were German Caucasians. All patients and control subjects gave informed consent.
Standard protocols were used for the isolation of leukocyte DNA. ApoE genotyping was performed as described by Hixson and Vernier.8 BH genotyping was performed using PCR-SSCP as previously described.9 The primers used were 5′-GGAAGCATGTCCCTGAA GAGGTGC-3′ (forward) and 5′-CCTGGATCTGTCC TTTGCAGCTACG-3′ (reverse). Each amplification reaction contained 500 ng DNA, 240 nM of each primer, 200 μM of each NTP, 1.9 mM MgCl2, 0.5 units of Taq polymerase (Boehringer Mannheim, Germany), and 7% DMSO in a final volume of 25 μl. The reaction mixture was subjected to 30 cycles of 1 min at 94°C, 0.5 min at 58°C and 1 min at 72°C. The PCR products were screened by SSCP analysis.9
The Pearson's χ2 test was used for the comparison of BH genotype frequencies between patients and controls. Allele frequencies were compared by the Fisher's exact test. Since mean age and gender distribution were different in cases and controls, logistic regression analysis was used for the simultaneous assessment of the influence of age, gender, and BH genotype on the risk for developing AD. Statistical significance was established at P ≤ 0.05.
Lendon CL, Ashall F, Goate AM . Exploring the etiology of Alzheimer disease using molecular genetics JAMA 1997; 277: 825–831
Cummings JL, Vinters HV, Cole GM, Khachaturian ZS . Alzheimer's disease: etiologies, pathophysiology, cognitive reserve, and treatment opportunities Neurology 1998; 51: (Suppl 1) 2–17
Bromme D, Rossi AB, Smeekens SP, Anderson DC, Payan DG . Human bleomycin hydrolase: molecular cloning, sequencing, functional expression, and enzymatic characterization Biochemistry 1996; 35: 6706–6714
Montoya SE, Aston CE, DeKosky ST, Kamboh MI, Lazo JS, Ferrell RE . Bleomycin hydrolase is associated with risk of sporadic Alzheimer's disease Nat Genet 1998; 18: 211–212
Farrer LA, Abraham CR, Haines JL et al. Association between bleomycin hydrolase and Alzheimer's disease in Caucasians Ann Neurol 1998; 44: 808–811
Sackett DL . Bias in analytic research J Chronic Dis 1979; 32: 51–63
McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM . 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 1984; 34: 939–944
Hixson JE, Vernier DT . Restriction isotyping of human apolipoprotein E by gene amplification and cleavage with HhaI J Lipid Res 1990; 31: 545–548
Montoya SE, Ferrell RE, Lazo JS . Genomic structure and genetic mapping of the human neutral cysteine protease bleomycin hydrolase Cancer Res 1997; 57: 4191–4195
Supported by the Deutsche Forschungsgemeinschaft (He 2318) and the BONFOR grants program of the Medical Faculty of the University of Bonn (AP 111/34).
About this article
Physiological Reviews (2019)
A Commensal Dipeptidyl Aminopeptidase with Specificity for N-Terminal Glycine Degrades Human-Produced Antimicrobial Peptides in Vitro
ACS Chemical Biology (2018)
Chemistry & Biology (2015)
Metabolism and neurotoxicity of homocysteine thiolactone in mice: protective role of bleomycin hydrolase
Amino Acids (2012)
Analysis of association between bleomycin hydrolase and apolipoprotein E polymorphism in Alzheimer’s disease
Neurological Sciences (2010)