Whole genome analysis in APOE4 homozygotes identifies the DAB1-RELN pathway in Alzheimer's disease pathogenesis

The APOE-{epsilon}4 allele is known to predispose to amyloid deposition and consequently is strongly associated with Alzheimer's disease (AD) risk. There is debate as to whether the APOE gene accounts for all genetic variation of the APOE locus. Another question which remains is whether APOE-{epsilon}4 carriers have other genetic factors influencing the progression of amyloid positive individuals to AD. We conducted a genome-wide association study in a sample of 5,390 APOE-{epsilon}4 homozygous ({epsilon}4{epsilon}4) individuals (288 cases and 5,102 controls) aged 65 or over in the UK Biobank. We found no significant associations of SNPs in the APOE locus with AD in the sample of {epsilon}4{epsilon}4 individuals. However, we identified a novel genome-wide significant locus associated to AD, mapping to DAB1 (rs112437613, OR=2.28, CI=1.73-3.01, p=5.4x10-9). This identification of DAB1 led us to investigate other components of the DAB1-RELN pathway for association. Analysis of the DAB1-RELN pathway indicated that the pathway itself was associated with AD, therefore suggesting an epistatic interaction between the APOE locus and the DAB1-RELN pathway.


Introduction 41
Genome wide association studies (GWAS) have led to the identification of many genetic loci 42 influencing the risk of dementia [1]. However, none of these approach the importance of 43 the APOE locus [2] where the APOE-e4 allele has a frequency of ~15% in controls and has a 44 risk ratio of >3 in cases. Other loci with allele frequencies of >1% have risk ratios of <1.4. 45 Recent studies have shown that the APOE genotype is almost solely responsible for amyloid 46 deposition whereas other components of Alzheimer's disease (AD) genetic risk contribute to 47 the occurrence of dementia in the context of amyloid deposition [3]. Furthermore, 48 neuropathologic studies have shown that clinical diagnoses in Alzheimer series had a 49 diagnostic accuracy of around 80%: this accuracy is implied by analyses comparing the large 50 clinical GWAS with the smaller neuropathologic GWAS, leading to the concern that these 51 larger GWAS are contaminated by other diagnoses. This concern is heightened by the 52 reports of loci for frontotemporal dementia in case series labelled as Alzheimer's disease in 53 the most recent GWAS for the disorder [4]. 54 55 With this background, we have undertaken an AD GWAS in individuals who are APOE-e4 56 homozygotes for three reasons. First, because in this group diagnostic accuracy is very high; 57 second, to assess whether in this context there is additional genetic risk at the APOE locus; 58 and third, to assess which previously reported loci are replicated in these cases and whether 59 there are any novel loci we can identify which are dependent on APOE-e4 homozygosity. 60 This study was possible in the UK Biobank [5] because it has a very large cohort, with a 61 sufficient number (for statistical analyses) of APOE-e4 homozygotes, where many 62 participants are now reaching the age where they are at risk. 63 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.28.22274418 doi: medRxiv preprint 6 (corresponding to the SNP OR), and the sample size as in our study (N cases = 288, N 109 controls = 5,102). Plots of regional associations were created using LocusZoom [8]. 110 111 Epistasis was defined as deviation from joint two SNPs linear effects in the logistic 112 regression model (known as statistical interaction). Significance of the interaction term was 113 assessed using --epistasis option PLINK [7], accounting for the same covariates as above. The 114 interaction plots were produced using matplotlib in python [9]. 115 116 Gene-based analysis was run by MAGMA using FUMA v1.3.7 [10, 11]. Competitive setting of 117 MAGMA was also used to test the candidate pathways for the enrichment of AD significant 118 genes as compared to the rest of the genome. 119 120

DAB1-RELN pathway analysis 121
The canonical Reelin-Dab1 signalling pathway has been studied extensively in mouse 122 neurons and brain [12]. For analysis, we divided the pathway into three sections: a) the 123 receptor complex, (Reelin, the receptors ApoER2, VLDLR, the adaptor protein DAB1, and the 124 tyrosine kinases SRC, FYN and YES) [13-16], b) branch 1 that regulates N-cadherin (CRK, 125 CRKL, C3G, RAP1, P120 catenin, N-cadherin) [17-19] and c) branch 2 that is involved in 126 microtubule-associated protein tau (MAPT) phosphorylation (PI3K, PDK, AKT, GSK3, STK25) 127 [20-22]. We converted these mouse proteins to the homologous human genes with the 128 BioConductor function in R and the NCBI database (www.ncbi.nlm.nih.gov/) yielding: a) 129 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 28, 2022.  RELN, LRP8, VLDLR, DAB1, SRC, FYN, YES1, b) CRK, CRKL, RAPGEF1, RAP1A, CTNND, CDH2, c)  Loci previously reported as GWAS for association with Alzheimer's disease status did not 145 show a strong replication in the current analysis of APOE-e4 homozygotes only 146 (Supplemental Table S1). Though the power to detect the GWAS-reported effect sizes in this 147 sample is not sufficient (see last column of Supplemental Table S1), four loci in CD33 148 (p=0.004), IQCK (p=0.009), LILRB2 (p=0.005) and SORL1 (p=0.007, MAF=0.04) had the 149 strongest evidence for association in the current analysis and a consistent direction of effect 150 between the current and previous GWAS. Weaker but nominally significant associations 151 with the consistent direction of the effect were also observed in the APH1B (p = 0.024), BIN1 152 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.28.22274418 doi: medRxiv preprint (p=0.011), SEC61G (p=0.015) and SNX1 (p=0.048) genes. In total, eight out of 77 SNPs 153 (previously reported as genome-wide significant and available in our study), replicated at 154 0.05 significance level with the same direction of association, which is statistically greater 155 than chance (p=0.038). In addition, 53/77 (69%) SNPs have same direction of effect in the 156 current analysis and previous GWAS which is greater than expected by chance (p = 0.001). 157 158

Identification of DAB1 as a locus 159
Multiple novel genome-wide significant intronic SNPs were present in DAB1 (lead SNP: 160 rs112437613, OR=2.28, CI=1.73-3.01, p=5.36x10 -9 ; Figure 1 and Supplemental Figure S2, 161 Table 1). The minor allele T was associated with disease risk (MAF=6% in non-AD and 12% in 162 AD e4e4-participants of the UK Biobank). To allow for the fact that some individuals might 163 not have reached the age at onset, we fit a survival regression model (adjusting for PCs and 164 sex). The result remained highly significant (Hazard Ratio=2.27, CI=1.75-2.95, p= 7.8x10 -10 ). 165 The Kaplan-Meier graph ( Figure 2) demonstrates that probability of getting the disease (y-166 axis) earlier (x-axis) is higher as the number of the risk alleles of rs112437613 SNP increases. 167

168
The frequency of this allele is reported 4%-7% in European population cohorts 169 (1000Genomes, TOPMED, GnomAD, Estonian, ALSPAC-UK, TWINSUK, Northern Sweden, see 170 https://genome.ucsc.edu). However, this SNP (and others in LD with it) did not show even a 171 nominal association to AD in recent GWAS that did not preselect for the e4e4 genotype: e.g. 172 a study of 21,982 cases and 41,944 controls the p-values were p~0.5 (see Table 1) [24]. 173 Indeed, in a case/control sample (without screening for the APOE-e4 status), the effect size 174 of this SNP would be OR=1.016, as the proportion of cases, with both T allele of 175 rs112437613 and e4e4, is 0.016 (=MAF(e4) 2 *MAF(rs112437613 in e4e4) = 0.36 2 *0.12, where 176 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.28.22274418 doi: medRxiv preprint 0.36 is the e4 allele frequency in cases [25], and, similarly, of controls is 0.001. Therefore, 177 the frequencies of the T allele in the overall sample are expected to be 0.061 in cases and 178 0.06 in controls, and consequently, the power to detect it with the sample size of the [24] 179 study is close to 0 (~3x10 -7 ). 180

181
This observation led us to test for an epistatic effect in the whole sample of the UK Biobank 182 aged 65+ (N=229,748). There was indeed significant epistasis between the two loci 183 (p=1.5x10 -5 ), whereas the effect of the T allele (rs112437613) was positive (OR=1.16, 184 SE=0.11), but only nominally significant (p=0.021), providing evidence for cooperation 185 between these two loci. The risk allele frequencies in this locus depending on APOE and AD 186 status are shown in Table S2 and Figure S3). Similar to DAB1, we tested this SNP for 199 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.28.22274418 doi: medRxiv preprint interaction with APOE-e4 in the whole UK Biobank sample. The interaction term was not 200 significant (p=0.24), however the pattern of AD risk based on the pair of these markers was 201 similar to DAB1 (Supplemental Figure S4). Using the results of the gene-based analyses described above, we tested whether the 220 receptor complex and the two pathway branches contained significantly more AD 221 associated genes as compared with the rest of the genome. We found that they were 222 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.28.22274418 doi: medRxiv preprint significantly (or almost significantly) enriched for genes associated to AD in the APOE-e4 223 homozygotes (p-values 0.06, 0.009, 0.075, for the receptor complex and branches 1 and 2, 224 respectively). The strongest significance was achieved when we combined the receptor 225 complex and the two branches of the pathway (p=0.0055). Table 2  be determined, based on previous studies it would seem likely that they cause a partial loss-288 of-function that is potentially age dependent or cell-type specific in nature and would result 289 in altered expression (eQTL) or splicing (sQTL). More than partial disruption of activity would 290 likely lead to a developmental disorder in the homozygous individuals similar to loss-of-291 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 28, 2022. In conclusion, we find a novel genome-wide significant hit in DAB1 in an APOE-e4 306 homozygote AD GWAS. This seems to be a hit only in APOE-e4 homozygotes. Furthermore, 307 it seems that this association marks a more general importance of the DAB1-RELN pathway 308 in disease pathogenesis. It is not clear why this pathway should be of such importance in 309 APOE-e4 homozygotes only, but a clue may be that such individuals have particularly dense 310 Ab pathology and one can imagine that this pathway either has a role in modulating APP 311 processing or in driving tau-phosphorylation in a manner that is dependent on high Aβ 312 levels. This work suggests that DAB1 has a protective role in late onset AD and highlights the 313 importance of resolving the mechanism that likely involves the REELIN-DAB1 pathway for 314 therapeutic development. 315 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Funding 317
This work was largely funded by the UK DRI, which receives its funding from the DRI Ltd,

Competing interests 326
The authors report no competing interests. 327 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 28, 2022. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.28.22274418 doi: medRxiv preprint CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.28.22274418 doi: medRxiv preprint Figure 2: The cumulative risk of AD among APOE-e4 homozygous of the UK Biobank 517 participants, who carry 0, 1 or 2 risk alleles T the lead SNP (rs112437613) in DAB1. 518 519 520 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.28.22274418 doi: medRxiv preprint CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.28.22274418 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.28.22274418 doi: medRxiv preprint

534
CHR -chromosome; BP -base-pair position in build37; SNP -single nucleotide polymorphism, closest gene -genes were annotated with 535 assembly hg19; effect/non-effect -effect and non-effect alleles; freq-frequency of reference allele in the UK Biobank APOE-e4 homozygotes 536 individuals; OR, SE, p-value -odds ratio, standard error and p-value of the current and previous reported AD GWAS association studies; GWAS 537 -reference of the corresponding GWAS study 538