Association of Structural Forms of 17q21.31 with the Risk of Progressive Supranuclear Palsy and MAPT Sub-haplotypes

ImportanceThe chromosome 17q21.31 region, containing a 900 Kb inversion that defines H1 and H2 haplotypes, represents the strongest genetic risk locus in progressive supranuclear palsy (PSP). In addition to H1 and H2, various structural forms of 17q21.31, characterized by the copy number of , {beta}, and {gamma} duplications, have been identified. However, the specific effect of each structural form on the risk of PSP has never been evaluated in a large cohort study. ObjectiveTo assess the association of different structural forms of 17q.21.31, defined by the copy numbers of , {beta}, and {gamma} duplications, with the risk of PSP and MAPT sub-haplotypes. Design, setting, and participantsUtilizing whole genome sequencing data of 1,684 (1,386 autopsy confirmed) individuals with PSP and 2,392 control subjects, a case-control study was conducted to investigate the association of copy numbers of , {beta}, and {gamma} duplications and structural forms of 17q21.31 with the risk of PSP. All study subjects were selected from the Alzheimers Disease Sequencing Project (ADSP) Umbrella NG00067.v7. Data were analyzed between March 2022 and November 2023. Main outcomes and measuresThe main outcomes were the risk (odds ratios [ORs]) for PSP with 95% CIs. Risks for PSP were evaluated by logistic regression models. ResultsThe copy numbers of and {beta} were associated with the risk of PSP only due to their correlation with H1 and H2, while the copy number of {gamma} was independently associated with the increased risk of PSP. Each additional duplication of {gamma} was associated with 1.10 (95% CI, 1.04-1.17; P = 0.0018) fold of increased risk of PSP when conditioning H1 and H2. For the H1 haplotype, addition {gamma} duplications displayed a higher odds ratio for PSP: the odds ratio increases from 1.21 (95%CI 1.10-1.33, P = 5.47 x 10-5) for H1{beta}1{gamma}1 to 1.29 (95%CI 1.16-1.43, P = 1.35 x 10-6) for H1{beta}1{gamma}2, 1.45 (95%CI 1.27-1.65, P = 3.94 x 10-8) for H1{beta}1{gamma}3, and 1.57 (95%CI 1.10-2.26, P = 1.35 x 10-2) for H1{beta}1{gamma}4. Moreover, H1{beta}1{gamma}3 is in linkage disequilibrium with H1c (R2 = 0.31), a widely recognized MAPT sub-haplotype associated with increased risk of PSP. The proportion of MAPT sub-haplotypes associated with increased risk of PSP (i.e., H1c, H1d, H1g, H1o, and H1h) increased from 34% in H1{beta}1{gamma}1 to 77% in H1{beta}1{gamma}4. Conclusions and relevanceThis study revealed that the copy number of {gamma} was associated with the risk of PSP independently from H1 and H2. The H1 haplotype with more {gamma} duplications showed a higher odds ratio for PSP and were associated with MAPT sub-haplotypes with increased risk of PSP. These findings expand our understanding of how the complex structure at 17q21.31 affect the risk of PSP. Key PointsO_ST_ABSQuestionC_ST_ABSDo large copy number variations (i.e., , {beta}, and {gamma}) inside 17q21.31 contribute to the risk of progressive supranuclear palsy (PSP) independently from the H1 and H2 haplotypes? Do structural forms of 17q21.31, characterized by combinations of , {beta}, and {gamma}, present divergent risk to the development of PSP? Are structural forms of 17q21.31 associated with MAPT sub-haplotypes, such as H1c? FindingsIn this case-control study of 1,684 individuals with PSP and 2,392 control subjects, the copy number of {gamma} duplication was independently associated with the risk of the disease. H1 haplotypes with more {gamma} duplications (H1{beta}1{gamma}2, H1{beta}1{gamma}3, and H1{beta}1{gamma}4) displayed a higher odds ratio for PSP when compared to H1{beta}1{gamma}1. Notably, H1{beta}1{gamma}3 was observed to be in linkage disequilibrium with H1c, a widely recognized MAPT sub-haplotype associated with PSP. MeaningThe association between the H1 and H2 haplotypes and PSP involves multiple contributing factors, including the copy number of {gamma} duplication.