Background: Breast cancer remains one of the most commonly diagnosed malignancies, yet its underlying mechanisms are incompletely understood. This study aimed to investigate the causal relationships between circulating proteins and breast cancer using Mendelian Randomization (MR) and to identify potential therapeutic targets.

Methods: Genomic data for breast cancer from the FinnGen study and protein quantitative trait loci (pQTL) data from Decode were analyzed. MR was applied to evaluate the effects of specific proteins on breast cancer risk, followed by validation through colocalization analyses and summary-data-based Mendelian Randomization (SMR). Two-step MR was used to assess the mediating role of immune cells in the causal pathway between the target proteins and breast cancer. Moreover, expression patterns and prognostic significance of the identified proteins in breast cancer were examined.

Results: Several proteins [glutathione S-transferase mu (GSTM) 1, N-acylsphingosine amidohydrolase 2 (ASAH2), TNF superfamily member 12 (TNFSF12), hematopoietic prostaglandin D synthase (HPGDS), nidogen 2 (NID2), SEMA4D, snurportin 1 (SNUPN), GSTM3, MYC-associated factor X (MAX), pregnancy zone protein (PZP), GSTM4, and uroporphyrinogen decarboxylase (UROD)] were significantly associated with breast cancer. Colocalization and SMR analyses supported the potential role of GSTM in breast cancer, especially GSTM1 (p_SMR = 0.02 and p_HEIDI = 0.08). Mediation MR implicated CD11c-expressing monocytes (Z = –2.197, p = 0.028) in the GSTM-breast cancer pathway. Notably, lower GSTM1/4 expression levels correlated with poorer clinical outcomes in breast cancer patients.

Conclusion: This study used cis-pQTL MR and colocalization to identify GSTM1/4 as a potential druggable gene and promising therapeutic target for breast cancer, which might be affected by immunity. These findings contributed to understanding the role of metabolic and immune mechanisms in breast cancer.