Radiotranscriptomics of non-small cell lung carcinoma for assessing high-level clinical outcomes using a machine learning-derived multi-modal signature

BACKGROUND

Multi-omics research has the potential to holistically capture intra-tumor variability, thereby improving therapeutic decisions by incorporating the key principles of precision medicine. The purpose of this study is to identify a robust method of integrating features from different sources, such as imaging, transcriptomics, and clinical data, to predict the survival and therapy response of non-small cell lung cancer patients.

METHODS

2996 radiomics, 5268 transcriptomics, and 8 clinical features were extracted from the NSCLC Radiogenomics dataset. Radiomics and deep features were calculated based on the volume of interest in pre-treatment, routine CT examinations, and then combined with RNA-seq and clinical data. Several machine learning classifiers were used to perform survival analysis and assess the patient's response to adjuvant chemotherapy. The proposed analysis was evaluated on an unseen testing set in a k-fold cross-validation scheme. Score- and concatenation-based multi-omics were used as feature integration techniques.

RESULTS

Six radiomics (elongation, cluster shade, entropy, variance, gray-level non-uniformity, and maximal correlation coefficient), six deep features (NasNet-based activations), and three transcriptomics (OTUD3, SUCGL2, and RQCD1) were found to be significant for therapy response. The examined score-based multi-omic improved the AUC up to 0.10 on the unseen testing set (0.74 ± 0.06) and the balance between sensitivity and specificity for predicting therapy response for 106 patients, resulting in less biased models and improving upon the either highly sensitive or highly specific single-source models. Six radiomics (kurtosis, GLRLM- and GLSZM-based non-uniformity from images with no filtering, biorthogonal, and daubechies wavelets), seven deep features (ResNet-based activations), and seven transcriptomics (ELP3, ZZZ3, PGRMC2, TRAK1, ATIC, USP7, and PNPLA2) were found to be significant for the survival analysis. Accordingly, the survival analysis for 115 patients was also enhanced up to 0.20 by the proposed score-based multi-omics in terms of the C-index (0.79 ± 0.03).

CONCLUSIONS

Compared to single-source models, multi-omics integration has the potential to improve prediction performance, increase model stability, and reduce bias for both treatment response and survival analysis.

Co-expression Analysis of Genes and Tumor-Infiltrating Immune Cells in Metastatic Uterine Carcinosarcoma

Uterine carcinosarcoma (UCS) is a malignant tumor with a high tendency to invasion and metastasis. However, the underlying invasion and metastasis mechanisms of UCS remain poorly understood. Genetic alteration and tumor-infiltrating immune cells play important roles in tumorigenesis, progression, and metastasis. To better understand the underlying mechanisms of UCS, we screened tumor-infiltrating immune cells by applying CIBERSORT algorithm and constructed nomograms to predict the prognosis of UCS patients based on metastasis-specific tumor-infiltrating immune cells and genes, and demonstrated their utility by the high AUC values. Combining gene co-expression and experimental validation results, we propose a potential mechanism of AK8, MPZ, and mast cells activated might play important parts in UCS metastasis.

The Biological Significance of Targeting Acetylation-Mediated Gene Regulation for Designing New Mechanistic Tools and Potential Therapeutics

The molecular interplay between nucleosomal packaging and the chromatin landscape regulates the transcriptional programming and biological outcomes of downstream genes. An array of epigenetic modifications plays a pivotal role in shaping the chromatin architecture, which controls DNA access to the transcriptional machinery. Acetylation of the amino acid lysine is a widespread epigenetic modification that serves as a marker for gene activation, which intertwines the maintenance of cellular homeostasis and the regulation of signaling during stress. The biochemical horizon of acetylation ranges from orchestrating the stability and cellular localization of proteins that engage in the cell cycle to DNA repair and metabolism. Furthermore, lysine acetyltransferases (KATs) modulate the functions of transcription factors that govern cellular response to microbial infections, genotoxic stress, and inflammation. Due to their central role in many biological processes, mutations in KATs cause developmental and intellectual challenges and metabolic disorders. Despite the availability of tools for detecting acetylation, the mechanistic knowledge of acetylation-mediated cellular processes remains limited. This review aims to integrate molecular and structural bases of KAT functions, which would help design highly selective tools for understanding the biology of KATs toward developing new disease treatments.

Same but different - Molecular comparison of human KTI12 and PSTK

Kti12 and PSTK are closely related and highly similar proteins implicated in different aspects of tRNA metabolism. Kti12 has been identified as an essential regulatory factor of the Elongator complex, involved in the modification of uridine bases in eukaryotic tRNAs. PSTK phosphorylates the tRNA^Sec-bound amino acid serine, which is required to synthesize selenocysteine. Kti12 and PSTK have previously been studied independently in various organisms, but only appear simultaneously in some animalia, including humans. As Kti12- and PSTK-related pathways are clinically relevant, it is of prime importance to understand their biological functions and mutual relationship in humans. Here, we use different tRNA substrates to directly compare the enzymatic activities of purified human KTI12 and human PSTK proteins. Our complementary Co-IP and BioID2 approaches in human cells confirm that Elongator is the main interaction partner of KTI12 but additionally indicate potential links to proteins involved in vesicular transport, RNA metabolism and deubiquitination. Moreover, we identify and validate a yet uncharacterized interaction between PSTK and γ-taxilin. Foremost, we demonstrate that human KTI12 and PSTK do not share interactors or influence their respective biological functions. Our data provide a comprehensive analysis of the regulatory networks controlling the activity of the human Elongator complex and selenocysteine biosynthesis.

How Elongator Acetylates tRNA Bases

Elp3, the catalytic subunit of the eukaryotic Elongator complex, is a lysine acetyltransferase that acetylates the C5 position of wobble-base uridines (U34) in transfer RNAs (tRNAs). This Elongator-dependent RNA acetylation of anticodon bases affects the ribosomal translation elongation rates and directly links acetyl-CoA metabolism to both protein synthesis rates and the proteome integrity. Of note, several human diseases, including various cancers and neurodegenerative disorders, correlate with the dysregulation of Elongator's tRNA modification activity. In this review, we focus on recent findings regarding the structure of Elp3 and the role of acetyl-CoA during its unique modification reaction.

Allende M. Elongator Subunit 3 (Elp3) Is Required for Zebrafish Trunk Development

Transfer RNAs (tRNAs) are the most post-transcriptionally modified RNA species. Some of these modifications, especially the ones located in the anti-codon loop, are required for decoding capabilities of tRNAs. Such is the case for 5-methoxy-carbonyl-methyl-2-thio-uridine (mcm5s2U), synthetized by the Elongator complex. Mutants for its sub-units display pleiotropic phenotypes. In this paper, we analyze the role of elp3 (Elongator catalytic sub-unit) in zebrafish development. We found that it is required for trunk development; elp3 knock-down animals presented diminished levels of mcm5s2U and sonic hedgehog (Shh) signaling activity. Activation of this pathway was sufficient to revert the phenotype caused by elp3 knockdown, indicating a functional relationship between Elongator and Shh through a yet unknown molecular mechanism.