Integrative omics analysis of TP53 in multiple myeloma patients

Abstract

Abstract Introduction: Although TP53 gene alterations—such as 17p deletion and TP53 mutations—occur at relatively low frequency in multiple myeloma (MM), they remain among the most powerful adverse prognostic factors. However, the functional status of p53 may be disrupted through mechanisms beyond coding sequence abnormalities. To investigate this possibility, we expanded our analysis to include TP53 intronic variants, the expression of p53-regulated and p53-regulating genes, microRNAs involved in p53 signaling, and p53 protein isoforms, aiming to explore their potential association with disease outcome. Methods: We included bone marrow samples from 150 newly diagnosed MM patients enrolled in the GEM2012 clinical trial (NCT01916252). A DNA capture-based NGS panel was employed to sequence the entire TP53 gene. In parallel, a targeted RNA sequencing panel was used to assess the expression of 64 genes, including p53 targets and regulators. MicroRNA expression was quantified by RT-qPCR, while protein expression of p53 and its isoforms was analyzed using capillary nanoimmunoassay technology (EA Rojas, AJH 2022). The integrative analysis was conducted using all p53-related data from 92 patients, employing the Multi-Omics Factor Analysis (MOFA2) package in R. As input features, all variables analyzed at the four omic levels were included. Results: A total of 10 single nucleotide variants (SNVs) in the TP53 coding DNA sequence were identified in 11 patients with a negative impact on progression free survival (PFS) and overall survival (OS) (p ≤ 0.001), regardless of the VAF. The analysis of non-coding regions showed 3 SNVs in the 3´ UTR in 11 patients, and 53 SNVs within intronic regions in 91 patients. These intronic variants were grouped into six distinct clusters by Louvain method. The third group, which included SNVs in introns 1, 3, and 9, was associated with longer PFS (p = 0.023). Among the 7 miRNAs analyzed, the most expressed were the direct regulators of TP53, miR-125b, miR-25, and miR-30d. High levels of miR-25 and miR-125 were associated with lower levels of p53 protein (p = 0.023 and p = 0.017, respectively). Notably, high expression of miR-30d and miR-34 impacted positively on PFS (p < 0.05). RNA-seq of p53 targets and regulators showed 48 genes differentially expressed compared to normal plasma cells from healthy controls (26 underexpressed and 22 overexpressed genes): GADD45, SEMA3A, and CCND1 were overexpressed in MM patients, while KITLG, APAF1, and TP73 showed a marked underexpression. Interestingly, TP73 and CCND1 were the only genes significantly downregulated in patients harboring del(17p), whereas they were upregulated in those patients with t(11;14). Meanwhile, patients with TP53 mutations or TP53 double hit alterations had lower levels of TNFRSF10B when compared to the rest of patients without this condition (adj p < 0.05). High TNFRSF10B and CCND1 expression were associated with longer OS (p ≤ 0.01). The optimal MOFA2 model was established with 7 latent factors. Clustering based on the 7 latent factors was performed using the PAM algorithm resulting in 3 groups with distinct profiles in terms of PFS. Cluster 1 exhibited the poorest prognosis, whereas Cluster 3 showed the most favorable (p = 0.02). Further characterization of these groups revealed that Cluster 1 had a higher proportion of patients with plasmacytomas (31%) compared to Clusters 2 and 3 (7.5% and 19%, respectively; p = 0.049). In addition, Clusters 1 and 2 were enriched in 1q gain, observed in 72% and 60% of patients, respectively, in contrast to only 11.5% in Cluster 3 (p < 0.001). Conclusions:Patients with TP53 exonic variants exhibited worse outcomes.A group of TP53 intronic variants was associated with a favorable prognostic impact.High levels of miR-34a, which is a direct target of p53, and miR-30d were associated with longer PFS.Differential expression of p53 targets and regulators genes in MM highlights a distinct molecular profile linked to TP53 abnormalities, with high TNFRSF10B and CCND1 expression emerging as potential markers of improved OS.p53 functional status defined three distinct groups with clear differences in PFS. This work has been funded by ISCIII (PI23/00319) (co-financed by FEDER), Junta de Castilla y León (GRS 2331/A/21), AECC (PROYE20047GUTI) and International Myeloma Foundation (BDNovis Research Award).