Preeclampsia (PE) is a major cause of maternal and perinatal morbidity, and early prediction is critical for timely intervention. This study aimed to identify predictive biomarkers for PE through transcriptomic analysis of second-trimester amniotic fluid supernatant (AFS) collected prior to clinical symptom onset. AFS samples from women who later developed PE (n = 7) and matched controls (n = 7) underwent RNA sequencing to identify differentially expressed genes (DEGs). Candidate genes were validated by real-time PCR in HTR-8/SVneo cells exposed to fluid shear stress at 3, 10, and 20 dyn/cm2 for 24 h, mimicking the hemodynamic environment of PE, and siRNA-mediated knockdown was used to assess effects on trophoblast migration and invasion. RNA sequencing revealed 19 DEGs, with 3 upregulated and 16 downregulated genes in the PE group. HOOK2 emerged as the most significantly upregulated gene. Four candidate genes, including HOOK2, CCDC160, CKB, and PARP15, were selected for further validation. HOOK2 mRNA expression significantly increased with higher shear stress levels, consistent with sequencing data. Knockdown of HOOK2 led to a significant increase in trophoblast invasion, while migration showed no significant change. These findings suggest that HOOK2 may serve as a promising early biomarker for PE by modulating trophoblast invasiveness under altered hemodynamic conditions, with potential to improve risk stratification and personalized monitoring in pregnancy.