Background The clinical manifestation of Immune thrombocytopenia (ITP) is heterogeneous, and its unclear pathogenesis makes treatment challenging. Traditionally, ITP has been considered an autoantibody-mediated platelet destruction disorder. Thus, identifying suitable plasma proteins for novel therapeutic regimens is crucial.

Methods This study employed Proteome-Wide colocalization, which included data from deCODE Genetics and the UK Biobank Pharma Proteomics Project (UKB-PPP) to screen potential proteins. To determine causality, Mendelian Randomization (MR) was conducted, treating candidate proteins as exposures and using ITP's Genome-wide association studies (GWAS) summary statistics from FinnGen R11 as outcome. The inverse variance weighted (IVW) method was primarily relied upon when all methods indicated a consistent direction of effect, while the Wald ratio method was employed if only one instrumental variable (IV) remained. Multiple sensitivity analyses and false-discovery rate adjustments (FDR-adjusted) were conducted to enhance robustness and reliability. Transcriptome-wide colocalization from the eQTLGen consortium was also performed to verify if gene levels corresponded with previously confirmed proteins. If consistent, further MR analysis was continued as described.

Results Through Proteome-Wide colocalization, proteins with a Posterior Probability of Hypotheses 4 (PP.H4.abf) ≥ 0.3 were included for further MR analysis. After screening, 17 proteins from deCODE and 16 from UKB-PPP were identified, with 4 proteins overlapping between datasets. Eventually, 19 proteins from both datasets were confirmed to have significant effects on ITP. Here, we report proteins with PP.H4.abf exceeding 0.7, indicating a strong association with ITP. After FDR correction, DNAJC21 and RNF5, which only appeared in the UKB-PPP dataset and not in deCODE, showed to be risk (OR=4.07, 95% CI=2.26-7.31, P=8.91e-06) and protective (OR=0.02, 95% CI=0.01-0.05, P=6.51e-13) factors, respectively. Further Transcriptome-wide colocalization and MR results confirm that the gene levels of DNAJC21 and RNF5 have similar effects to their protein levels, with DNAJC21's gene positively associated with ITP (OR=1.99, 95% CI=1.40-2.85, P=4.04e-04) and RNF5's gene negatively associated with ITP (OR=0.70, 95% CI=0.57-0.85, P=1.32e-03).

Conclusions

  • In this study, protein and gene levels of DNAJC21 were shown to increase the risk of ITP, while RNF5's protein and gene levels demonstrated a protective effect.

  • Mutations in DNAJC21 had been linked to Bone Marrow Failure Syndrome (BMF), while RNF5 was primarily associated with immune reactions. Therefore, the discovery of the roles of DNAJC21 and RNF5 helps to elucidate the underlying mechanisms and develop novel drug targets for ITP.

Disclosures

No relevant conflicts of interest to declare.

This content is only available as a PDF.
2024
Sign in via your Institution