Benoît P. Nicolet, Anouk P. Jurgens, Kaspar Bresser, Antonia Bradarić, Aurélie Guislain et al.
Accurate protein expression in human immune cells is essential for appropriate cellular function. The mechanisms that define protein abundance are complex and are executed on transcriptional, posttranscriptional, and posttranslational levels. Here, we present SONAR, a machine learning pipeline that learns the endogenous sequence code and that defines protein abundance in human cells. SONAR uses thousands of sequence features (SFs) to predict up to 63% of the protein abundance independently of promoter or enhancer information. SONAR uncovered the cell type–specific and activation-dependent usage of SFs. The deep knowledge of SONAR provides a map of potentially biologically active SFs, which can be leveraged to manipulate the amplitude, timing, and cell type specificity of protein expression. SONAR informed on the design of enhancer sequences to boost T cell receptor expression and to potentiate T cell function. Beyond providing fundamental insights into the regulation of protein expression, our study thus offers innovative means to improve therapeutic and biotechnology applications.
Benoît P. Nicolet, Anouk P. Jurgens, Kaspar Bresser, Antonia Bradarić, Aurélie Guislain, Monika C. Wolkers. Learning the sequence code of protein expression in human immune cells. 2025-07-23. https://doi.org/10.1126/sciadv.ads0510
A 467-member synthetic 3′UTR library, with six repeats of each designed motif in a GFP reporter, was delivered by an integrating retroviral vector to primary CD8+ T cells from three healthy donors. The sorted GFP-high, GFP-low, and total GFP-positive fractions were quantified by genomic-DNA amplicon sequencing.
A 467-member synthetic 3′UTR library, with six repeats of each designed motif in a GFP reporter, was delivered by an integrating retroviral vector to primary CD4+ T cells from three healthy donors. The sorted GFP-high, GFP-low, and total GFP-positive fractions were quantified by genomic-DNA amplicon sequencing.
A 467-member synthetic 3′UTR library, with six repeats of each designed motif in a GFP reporter, was delivered by an integrating retroviral vector to HEK293 cells. Three HEK batches were sorted into GFP-high, GFP-low, and total GFP-positive fractions, and the integrated library elements were quantified by genomic-DNA amplicon sequencing.
Original supplemental and deposited inputs retained for this study. Download files individually or together as a ZIP; nested folders are preserved. Source reuse terms apply, and sequencing reads may be omitted.
Download all 7 files (ZIP)GSE291719_family.soft.gzGSE291719_SONAR_MPRA_readcounts.tsv.gzGSE291719_SONAR_MPRA_readcounts_norm.tsv.gzsource_manifest.txtsupplementary_Data_S3.xlsxsupplementary_Data_S4.xlsxsupplementary_materials.pdf