APBB2

Abstract: Abnormal processing of the membrane-spanning amyloid precursor protein (APP), resulting in the production of increased amounts of fibrillogenic b-amyloid peptide (Ab), is considered to be one of the key metabolic events underlying Alzheimer’s disease (AD; Selkoe 1994). The function of APP is not fully understood, and the precise cellular mechanisms that lead to A b production are not clearly defined. However, one pathway for A b production involves the re-internalization of membrane-bound APP into lysosomes where fragments of APP containing intact A b are generated (Selkoe 1994). In common with a number of cell surface receptors, the carboxy terminal cytoplasmic domain of APP contains an AsnPro-Thr-Tyr (NPTY) motif which mediates re-internalization via clathrin-coated pits (Chen et al. 1990). This motif has also been demonstrated to be a consensus sequence for binding to phosphotyrosine binding/interacting domain (PTB)-bearing proteins (van der Geer and Pawson 1995). We and others have recently reported that the cytoplasmic domain of APP binds to four human PTB proteins: X11, X11-like, Fe65, and Fe65-like (Borg et al. 1996; Bressler et al. 1996; Fiore et al. 1995; Gue ́nette et al. 1996; McLoughlin and Miller 1996). It has been confirmed that the YENPTY sequence in the cytoplasmic domain of APP is responsible for mediating the interactions between the PTB domain in X11 and the second of two PTB domains in Fe65 (Borg et al. 1996; Fiore et al. 1995). PTB domain proteins are believed to be involved in signal transduction processes (van der Geer and Pawson 1995), and the interaction of APP with X11, X11-like, Fe65, and Fe65-like suggests a role for APP in such signal transduction mechanisms. Furthermore, as they interact with the YENPTY motif in APP, these PTB proteins may modulate processing of APP and hence formation of A b. Therefore, mapping of the genes coding for these proteins is important as they represent new candidate susceptibility genes for AD. The approved gene symbols for the members of these APP binding protein (APB) families are presented in Table 1. The gene for human X11 (APBA1) is already known to be on Chromosome (Chr) 9 close to marker D9S411E (Duclos et al. 1993), and the gene for human Fe65 (APBB1) has been localized to Chr 11 at 11p15 (Bressler et al. 1996). The existence of murine X11 and murine Fe65-like has not yet been reported. Here we report the chromosomal assignment of human APBA2 and APBB2 plus the chromosomal mapping of the murine homologs of X11-like (Apba2) and Fe65 ( Apbb1). In order to map the human APBA2 and APBB2 genes, we selected PCR primers from the previously identified cDNA clones (McLoughlin and Miller 1996) and overlapping sequences deposited in the databases (accession numbers R89683, R13010, R18654, and T16098 for APBA2 and accession number HSU62325 for APBB2). For APBA2 the following primer pair: forward, 58-TTACAAGTCGTGTCCTGGGAG-38, and reverse, 58-GACGTCTGGGGTCCTGTG-3 8, generated a small PCR product of 103 bp. For APBB2 the following primer pair: forward, 58-CACAGAGAAGAGTCTGGCCC-38 and reverse, 5 8-AGGTTGCTTGTGACAGGTCC-38, generated a PCR product of 114 bp. These PCR products were sequenced to confirm they originated from the correct genes. Both human APBA2 and APBB2 genes were mapped using the Genebridge 4 radiation hybrid panel (HGMP Resource Centre, Cambridge, UK) consisting of 94 hamster-derived cell lines. PCR amplification of human DNA with PCR primers designed for these genes resulted in products of the expected size, while no amplification products were obtained from the hamster DNA control sample. Scores for individual cell lines were submitted at the WICGR mapping service at http:// www.genome.wi.mit.edu. APBA2 was assigned to human Chr 15 between the markers WI-5590 (10.31 cR) and D15S144 (21.7 cR). APBB2 was assigned to human Chr 4 between the markers D4S405 (4.6 cR) and D4S496 (10.1 cR). To map theApba2andApbb1loci in the mouse, we used the EUCIB resource which comprises 982 interspecific backcross progeny for high-resolution genetic mapping across the mouse genome (Breen et al. 1994). It is clear from sequence alignments that the mouse sequence L34676 available in the Genbank database corresponds to the mouse homolog of APBA2 ( pba2) rather than to the mouse homolog of APBA1 (McLoughlin and Miller 1996). The following primer pair was selected for mouse Apba2 PCR amplifications: forward, 5 8-GCGCTCTGATCTCAATGG38; reverse, 58-GGAAATGATGCCACCTTC-38. This generated an approximately 1000-bp PCR product. Primers for mouse Apbb1 were designed from the published rat sequence (accession number X60468). The following primer pair was designed for mouse Apbb1 PCR amplifications: forward, 5 8-CTGGCACATCCCAACAGG-38; reverse, 58-AGCAAAGCCAGTCCAGGT-38. The PCR product was 202 bp. Both of these murine PCR products were sequenced to confirmed their origin. The mouseApba2andApbb1PCR products did not show any allelic size difference between C57BL/6 and Mus spretus, the two parental strains of the EUCIB interspecific backcross. However, in both cases, SSCP analysis (Chang et al. 1993) did show a clear polymorphism between C57BL/6 and Mus spretus.In the case of Apba2the large 1-kb PCR product was Sau3AI digested prior to loading on the SSCP gel. 92 random samples from the EUCIB backcross were analyzed for the segregation of C57BL/6 and Mus Correspondence to: D.M. McLoughlin at Dept. of Neuroscience Mammalian Genome 9, 473–475 (1998).

Abstract: Glyocogen synthase kinase 3 (GSK3) plays an important role in the pathophysiology of Alzheimer’s disease (AD) through the phosphorylation of tau. Recent work has suggested that GSK3β also plays a role in the amyloid pathway of AD through genetic interactions with APP and APBB2 on in vivo measures of amyloid. This project extends the previously identified genotype interactions to an autopsy measure of amyloid, while also testing the same interactions leveraging gene expression data quantified in the prefrontal cortex. 797 participants (251 cognitively normal, 196 mild cognitive impairment, and 350 Alzheimer’s disease) were drawn from the Religious Orders Study and Rush Memory and Aging Project. A mean score of amyloid load was calculated across eight brain regions, gene expression levels from frozen sections of the dorsolateral prefrontal cortex were quantified using RNA amplification, and expression signals were generated using Beadstudio. Three SNPs previously identified in genetic interactions were genotyped using the Illumina 1M genotyping chip. Covariates included age, sex, education, and diagnosis. We were able to evaluate 2 of the 3 previously identified interactions, of which the interaction between GSK3β (rs334543) and APBB2 (rs2585590) was found in this autopsy sample (p = 0.04). We observed a comparable interaction between GSK3β and APBB2 when comparing the highest tertile of gene expression to the lowest tertile, t(1) = −2.03, p = 0.043. These results provide additional evidence of a genetic interaction between GSK3β and APBB2 and further suggest that GSK3β is involved in the pathophysiology of both of the primary neuropathologies of Alzheimer’s disease.