Search bioRxiv⌕ Search
Home/Archive/bioRxiv

bioRxiv · 10.1101/629097

Performance comparison of reverse transcriptases for single-cell studies

Zucha, D.·Androvic, P.·Kubista, M.·Valihrach, L.
View original source ↗Export metadata ↓

Abstract

BackgroundRecent technical advances allowing quantification of RNA from single cells are revolutionizing biology and medicine. Currently, almost all single-cell transcriptomic protocols rely on conversion of RNA to cDNA by reverse transcription (RT). However, RT is recognized as highly limiting step due to its inherent variability and suboptimal sensitivity, especially at minute amounts of RNA. Primary factor influencing RT outcome is reverse transcriptase (RTase). Recently, several new RTases with potential to decrease the loss of information during RT have been developed, but the thorough assessment of their performance is missing.\n\nMethodsWe have compared the performance of 11 RTases in RT-qPCR on single-cell and 100-cell bulk templates using two priming strategies: conventional mixture of random hexamers with oligo(dT)s and reduced concentration of oligo(dT)s mimicking common single-cell RNA-Seq library preparation protocols. Based on the performance, two RTases were further tested in high-throughput single-cell experiment.\n\nResultsAll RTases tested reverse transcribed low-concentration templates with high accuracy (R2 > 0.9445) but variable reproducibility (median CVRT = 40.1 %). The most pronounced differences were found in the ability to capture rare transcripts (0 - 90% reaction positivity rate) as well as in the rate of RNA conversion to cDNA (7.3 - 124.5 % absolute yield). Finally, RTase performance and reproducibility across all tested parameters were compared using Z-scores and validity of obtained results was confirmed in a single-cell model experiment. The better performing RTase provided higher positive reaction rate and expression levels and improved resolution in clustering analysis.\n\nConclusionsWe performed a comprehensive comparison of 11 RTases in low RNA concentration range and identified two best-performing enzymes (Maxima H-; SuperScript IV). We found that using better-performing enzyme (Maxima H-) over commonly-used below-average performer (SuperScript II) increases the sensitivity of single-cell experiment. Our results provide a reference for the improvement of current single-cell quantification protocols.

Source connections

Explore related subjects

molecular biologyPreprint

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Zucha, D., Androvic, P., Kubista, M., Valihrach, L.. 2019-05-07. Performance comparison of reverse transcriptases for single-cell studies. https://doi.org/10.1101/629097

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

Genomic features and instability of linezolid resistance in vancomycin-resistant Enterococcus faecalis ST525

This study investigated the mechanisms of oxazolidinone resistance and genomic background in clinical ST525 Enterococcus faecalis isolates, a poorly characterized lineage reported in Brazilian hospitals as either vancomycin-resistant (VRE) or linezolid- and vancomycin-resistant (LVRE). Five LVREfs from intensive care unit patients (linezolid MICs, 8-32 mg/L; vancomycin MICs, >256 mg/L) carried only G2576T and G2505A (in three of four rrl copies) as known linezolid resistance determinants, despite variable linezolid MICs. During serial in vitro passages without antibiotic pressure, isolates with initial MICs of 8-16 mg/L exhibited progressive MIC reductions accompanied by reversion of mutant to wild-type rrl copies, whereas the isolate with an initial MIC of 32 mg/L maintained high-level resistance without reversion. Additional domain V substitutions emerged during MIC decline, while growth kinetics remained broadly similar. Whole-genome sequencing of a representative isolate (LVREfs L7) revealed a ~2.8-Mb chromosome enriched in mobile genetic elements (~8.6%), lacking a functional CRISPR-Cas system. A plasmid-borne prototype Tn1546 carrying vanA, devoid of internal insertion sequences, was found inserted into a Tn3-family transposase gene, a platform not typically associated with vanA mobilization. Together, these findings indicate that ST525 LVREfs, whose genomic features are consistent with long-term persistence in hospitals, may pose a challenge to routine susceptibility testing. A single mutational mechanism can yield variable linezolid MICs, and the ratio of mutated vs wild-type rrl copies can become heterogeneous in the absence of antibiotic pressure, without correlating directly with these values, contrary to the well-documented copy number-MIC correlation.

molecular biology↗
bioRxiv2026

Global repression of mesenchymal genes in epithelial cells by the histone demethylase LSD1

Epithelial-mesenchymal transition (EMT) is an essential process in development, tissue repair, and tumour metastasis, and is triggered by SNAI, ZEB and TWIST transcription factors, that recruit lysine-specific demethylase 1 (LSD1) to repress epithelial genes such as E-cadherin. In mesenchymal cells, LSD1 inhibition restores E-cadherin expression and reduces migration and invasion, consistent with LSD1 inhibition reverting EMT, and offering a therapeutic modality for EMT-associated disease. Recent data, however, suggest context-dependent roles for LSD1, as its inhibition in epithelial cells increases mesenchymal marker expression, potentially promoting EMT. To investigate this discrepancy, we mapped genome-wide binding of LSD1 in HCT116 epithelial cells and used rapid degron-mediated ablation of endogenous LSD1 to measure changes in nascent and steady-state transcription and chromatin accessibility. In epithelial cells, LSD1 binds constitutively-accessible, primed mesenchymal promoters. These are rapidly derepressed after LSD1 ablation, with activation of SNAI2 and ZEB1, which repress epithelial hallmarks such as E-cadherin. Re-introduction of LSD1 downregulates SNAI2 and ZEB1 but, surprisingly, further represses E-cadherin, indicating sustained loss of epithelial characteristics. Collectively, our results show that LSD1 maintains epithelial identity by repressing mesenchymal genes, questioning the therapeutic utility of LSD1 inhibition for EMT control.

molecular biology↗
bioRxiv2026

The long noncoding RNA FOSIL shapes the dynamics of FOS induction

Immediate-early genes require rapid, transient, and precisely tuned transcriptional activation in response to extracellular signals, yet how this kinetic precision is achieved remains poorly understood. We characterize FOSIL (LINC01220), an inducible lncRNA locus downstream of FOS. Depleting FOSIL RNA delays heregulin (HRG)-induced FOS mRNA and protein accumulation and prolongs the response, indicating that FOSIL sharpens both the onset and the shutoff of the induction wave. Capture Micro-C maps show that FOSIL depletion weakens the chromatin boundary between FOS and FOSIL and increases local compaction across the pre-assembled FOS-FOSIL neighborhood during serum starvation, demonstrating that basal FOSIL production maintains a locally insulated, poised conformation prior to signal arrival. Upon HRG stimulation, FOSIL depletion blunts the rapid movement of the locus toward nuclear speckles, diminishes early speckle-proximal FOS transcription, and extends late firing at speckle-distant alleles. At single transcription active sites, nascent FOSIL activity correlates with closer speckle proximity and higher FOS output, and activating the endogenous FOSIL promoter by CRISPRa, but not supplying FOSIL RNA in trans, enhances both speckle association and FOS induction. Together, these findings support a mechanism where endogenous FOSIL activity coordinates pre-stimulus chromatin folding with dynamic nuclear speckle positioning to fine-tune the onset, amplitude, and resolution of immediate-early gene expression.

molecular biology↗