Empowering multiplexed ultra-throughput RiboWich to move towards single-cell ribosome profiling

Cell-to-cell heterogeneity is an intrinsic characteristic of all biological tissues, driving both physiological and pathological processes. This cellular variation emerges across the multiple layers of gene expression regulation, including the genome, transcriptome, translatome, and proteome. In the last 20 years, a multitude of single-cell technologies have been developed to study this cellular diversity, with particular effort dedicated to the investigation of the transcriptome via single-cell RNA sequencing (scRNASeq). However, mRNA levels and protein abundance frequently decouple, showing only a modest correlation. Consequently, to gain insights into gene expression regulation, inspecting the transcriptome alone is not sufficient. Translation is an interesting regulatory hub controlling gene expression, but major technical limitations have left a significant gap in our understanding of how it actively regulates cellular heterogeneity at single-cell resolution. The gold standard method for studying translation at genome-wide scale and single-nucleotide resolution is ribosome profiling (RiboSeq). RiboSeq is a next-generation sequencing (NGS) technique that allows the identification of transcripts undergoing translation and the precise mapping of ribosomes along these transcripts at single-nucleotide resolution. The method is based on a digestion step in which a nuclease chops free RNA into its nucleotide constituents, leaving only the mRNA fragments protected by ribosomes, the so-called ribosome-protected fragments (RPFs). These 30-nucleotide-long RPFs are isolated and converted into a dsDNA library suitable for sequencing. Traditional RiboSeq is characterized by several limitations, including high input requirements, low throughput, and the absence of multiplexing approaches. Moreover, the protocol remains complex, labor-intensive, and is still characterized by challenging steps, especially when working with minimal material, such as ribosome isolation, RPF size selection, and purification. Finally, the library preparations that are currently in use are known to suffer from multiple biases. Overall, these technical hurdles are the reason as to why very few works have been published performing low-input RiboSeq, and only two attempts have been made to profile translatomes at single-cell resolution. Despite these pioneering efforts, both single-cell protocols possess critical technical characteristics that present several limitations. In this context, the aim of my thesis is to overcome the above-mentioned challenges and to develop novel technological solutions for studying translational regulation in low-input samples and single-cell contexts. To address these points, my work proposes RiboWich (Ribosome sandWich), a novel technological solution based on direct, in-lysate ligation between the sequencing adaptor and the RPFs while they are still embedded within ribosomes. With this simple reaction, RiboWich streamlines and accelerates the entire protocol, avoiding any purification step prior to ligation, such as ribosome isolation, RNA extraction, and RPF size selection and purification. In head-to-head comparisons with a standard ribosome profiling method, RiboWich proved to capture bona fide RPFs with the same efficiency and accuracy as the reference protocol. Furthermore, to successfully handle low-input samples, RiboWich integrates a one-step multiplexing strategy, exploiting barcoded adaptors during ligation. Indeed, different samples can be tagged with different barcodes, allowing them to be pooled soon after the ligation step so that subsequent reactions occur in a single tube. I demonstrated that RiboWich allows the profiling of as few as 2,000 HEK293T cells. This strategy improves sample handling and minimizes material loss, while reducing time and costs. Moreover, RiboWich implements a dual-step multiplexing strategy that enables high-throughput ribosome profiling. This approach leverages not only a barcoded ligation but also a second barcoded reaction during reverse transcription (RT). By utilizing barcoded RT primers, soon after cDNA synthesis, samples can be further pooled together to maximize the throughput. Ultimately, the unique combination of barcodes from both the ligation and RT steps allows for unambiguous sample identification. The high-throughput RiboWich protocol was adapted for and validated in a 96-well plate format using four distinct human and mouse cell lines, eight different conditions, and three technical replicates each. I demonstrated the robustness of RiboWich by successfully profiling 96 samples in parallel, obtaining high-quality and consistent data across all replicates. Crucially, RiboWich not only satisfies all standard quality metrics for ribosome profiling but also accurately captures condition-specific translational signatures. Overall, due to its inherent characteristics, RiboWich represents the first true high-throughput ribosome profiling method, making it the perfect technology to process large numbers of samples with minimal starting material. Building upon this solid background, and with the goal of moving toward single-cell applications, I started scaling down RiboWich. Strategically, I evaluated three different approaches in parallel, exploiting three distinct compartmentalization methods and two barcoding strategies, with the ultimate goal of finding the most efficient balance between resolution, throughput, costs, and automation dependency. The first proposed method is in-cell RiboWich, which exploits the cell itself as the compartment in which reactions occur. While RPFs are physically separated, cells are actually processed in bulk within a single tube. In detail, in-cell RiboWich combines cell permeabilization, in-cell nuclease digestion to generate RPFs, and in-cell RiboWich ligation coupled with barcoding via a split-and-pool strategy. In this context, I obtained preliminary indications that the different steps work, although robust optimization is still needed before moving towards a true single-cell experiment. Next, in-capsule RiboWich uses semi-permeable capsules (SPCs), developed by Atrandi Biotechnology, for isolating single cells. As in the previous case, while RPFs are separated because they are compartmentalized within capsules, cells are processed as a bulk mixture. In detail, in-capsule RiboWich integrates cell encapsulation, lysis, in-capsule nuclease digestion for RPF generation, in-capsule RiboWich ligation, and in-capsule library preparation. Split-and-pool barcoding occurs post-PCR amplification at the very end of the library preparation. With the aim of testing the compatibility between RiboWich and the SPC technology, I obtained promising results and encouraging indications regarding the digestion and ligation steps, although further optimizations are required before advancing to single-cell applications. Finally, in-plate RiboWich utilizes wells within multi-well plates to separate individual cells. Unlike the other two methods, this approach is based on the physical isolation of cells, where reactions occur sequentially within the same well. In detail, in-plate RiboWich involves cell sorting into plates and automated robotic handling for library preparation. Indeed, lysis, nuclease digestion for RPF generation, RiboWich ligation, and subsequent steps for library construction occur via automation. What sets this method apart is that, to increase throughput, it exploits the previously described dual-step multiplexing strategy for barcoding. In this context, I demonstrated that in-plate RiboWich successfully handles the automated, parallel profiling of 168 samples with minimal starting material and captures bona fide RPFs. Among the three proposed compartmentalization and barcoding strategies, the in-plate format currently represents the most technically mature and reliable approach. The method can readily scale to cover an entire 384-well plate, establishing a feasible and highly reproducible foundation for future single-cell applications. After establishing RiboWich as a robust, mature technology capable of transitioning ribosome profiling from a low-throughput tool to an unprecedented, scalable high-throughput platform, my next objective was to validate its utility in a biologically relevant context. Specifically, I sought to showcase its potential for drug screening and validation in Spinal Muscular Atrophy (SMA). SMA is an autosomal recessive neurodegenerative disorder primarily caused by deletions or mutations in the SMN1 gene, resulting in insufficient levels of the Survival Motor Neuron (SMN) protein. While most current approved therapies increase SMN protein levels, they remain non-curative and are limited by variable patient responses. Emerging evidence highlights that SMN functions as a ribosome-associated protein, regulating the translation of a specific subset of transcripts characterized by translational enhancer sequences in their 5’ UTR and rare codons at the beginning of the coding sequence (CDS). Consequently, SMA is characterized by decreased ribosome occupancy on these target mRNAs and global translational defects that can be rescued by restoring SMN levels. Based on this background, before the beginning of my PhD, I hypothesized that identifying small molecules capable of modulating translation represents a promising, unseen therapeutic frontier, either as standalone treatments or as synergistic adjuncts to existing SMN-dependent therapies. To test this hypothesis, a small-molecule screening was performed, and four candidate hits capable of increasing SMN-specific translation were identified. Among these, the kinase inhibitor PD98059 proved most promising, exerting the strongest effect on my SMN sensor. Capitalizing on the simplicity and dual-step multiplexing capacity of RiboWich, I decided to expand the scope of this investigation. RiboWich was deployed with a dual purpose: to profile translatomes in a disease-relevant context, thereby staging its potential as a powerful drug-screening methodology, and to comprehensively analyze the genome-wide effects of PD98059. RiboWich demonstrated that the compound rescues defects in ribosome occupancy on mRNAs sharing features with SMN-specific transcripts. Crucially, I showed that PD98059’s effect is independent of both SMN and the canonical ERK/MEK pathway, for which PD98059 is a well-known classical inhibitor. This unexpected finding suggested that the small molecule operates through an alternative molecular target in our SMA cellular model at the concentration employed. To identify this alternative mechanism, I performed limited proteolysis coupled with mass spectrometry (LiP-MS), which led to the unambiguous identification of a single protein target: the threonyl-tRNA synthetase 1 (TARS1) enzyme. Molecular dynamics simulations identified a putative binding pocket at the dimerization interface of the TARS1 homodimer, offering new perspectives on the mechanism of action of PD98059. Moreover, preliminary experiments support a model of enzymatic inhibition, wherein PD98059-mediated inhibition of TARS1 induces a specific increase in the translation of SMN-specific transcripts. Overall, this part of the work demonstrates that RiboWich serves as an ideal, high-throughput tool for small-molecule screening and validation and identifies the first translation-based, SMN-independent molecule that exerts beneficial effects not only in vitro, but also in vivo in SMA animal models.