tsRNAsearch: a pipeline for the identification of tRNA and ncRNA fragments from small RNA-sequencing data.

MOTIVATION: tRNAs were originally considered uni-functional RNA molecules involved in the delivery of amino acids to growing peptide chains on the ribosome. More recently, the liberation of tRNA fragments from tRNAs via specific enzyme cleavage has been characterized. Detection of tRNA fragments in sequencing data is difficult due to tRNA sequence redundancy and the short length of both tRNAs and their fragments. RESULTS: Here, we introduce tsRNAsearch, a Nextflow pipeline for the identification of differentially abundant tRNA fragments and other non-coding RNAs from small RNA-sequencing data. tsRNAsearch is intended for use when comparing two groups of datasets, such as control and treatment groups. tsRNAsearch comparatively searches for tRNAs and ncRNAs with irregular read distribution profiles (a proxy for RNA cleavage) using a combined score made up of four novel methods and a differential expression analysis, and reports the top ranked results in simple PDF and TEXT files. In this study, we used publicly available small RNA-seq data to replicate the identification of tsRNAs from chronic hepatitis-infected liver tissue data. In addition, we applied tsRNAsearch to pancreatic ductal adenocarcinoma (PDAC) and matched healthy pancreatic tissue small RNA-sequencing data. Our results support the identification of miR135b from the original study as a potential biomarker of PDAC and identify other potentially stronger miRNA biomarkers of PDAC. AVAILABILITY AND IMPLEMENTATION: https://github.com/GiantSpaceRobot/tsRNAsearch. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The spleen as a sanctuary site for residual leukemic cells following ABT-199 monotherapy in ETP-ALL.

B-cell lymphoma 2 (BCL-2) has recently emerged as a therapeutic target for early T-cell progenitor acute lymphoblastic leukemia (ETP-ALL), a high-risk subtype of human T-cell ALL. The major clinical challenge with targeted therapeutics, such as the BCL-2 inhibitor ABT-199, is the development of acquired resistance. We assessed the in vivo response of luciferase-positive LOUCY cells to ABT-199 monotherapy and observed specific residual disease in the splenic microenvironment. Of note, these results were confirmed by using a primary ETP-ALL patient-derived xenograft. Splenomegaly has previously been associated with poor prognosis in diverse types of leukemia. However, the exact mechanism by which the splenic microenvironment alters responses to specific targeted therapies remains largely unexplored. We show that residual LOUCY cells isolated from the spleen microenvironment displayed reduced BCL-2 dependence, which was accompanied by decreased BCL-2 expression levels. Notably, this phenotype of reduced BCL-2 dependence could be recapitulated by using human splenic fibroblast coculture experiments and was confirmed in an in vitro chronic ABT-199 resistance model of LOUCY. Finally, single-cell RNA-sequencing was used to show that ABT-199 triggers transcriptional changes in T-cell differentiation genes in leukemic cells obtained from the spleen microenvironment. Of note, increased expression of CD1a and sCD3 was also observed in ABT199-resistant LOUCY clones, further reinforcing the idea that a more differentiated leukemic population might display decreased sensitivity toward BCL-2 inhibition. Overall, our data reveal the spleen as a site of residual disease for ABT-199 treatment in ETP-ALL and provide evidence for plasticity in T-cell differentiation as a mechanism of therapy resistance.

Polymorphic centromere locations in the pathogenic yeast Candida parapsilosis.

Centromeres pose an evolutionary paradox: strongly conserved in function but rapidly changing in sequence and structure. However, in the absence of damage, centromere locations are usually conserved within a species. We report here that isolates of the pathogenic yeast species Candida parapsilosis show within-species polymorphism for the location of centromeres on two of its eight chromosomes. Its old centromeres have an inverted-repeat (IR) structure, whereas its new centromeres have no obvious structural features but are located within 30 kb of the old site. Centromeres can therefore move naturally from one chromosomal site to another, apparently spontaneously and in the absence of any significant changes in DNA sequence. Our observations are consistent with a model in which all centromeres are genetically determined, such as by the presence of short or long IRs or by the ability to form cruciforms. We also find that centromeres have been hotspots for genomic rearrangements in the C. parapsilosis clade.

Identification of an Exceptionally Long Intron in the HAC1 Gene of Candida parapsilosis.

The unfolded protein response (UPR) in the endoplasmic reticulum (ER) is well conserved in eukaryotes from metazoa to yeast. The transcription factor HAC1 is a major regulator of the UPR in many eukaryotes. Deleting HAC1 in the yeast Candida parapsilosis rendered cells more sensitive to DTT, a known inducer of the UPR. The deletion strain was also sensitive to Congo red, calcofluor white, and the antifungal drug ketoconazole, indicating that HAC1 has a role in cell wall maintenance. Transcriptomic analysis revealed that treatment of the wild type with DTT resulted in the increased expression of 368 genes. Comparison with mutant cells treated with DTT reveals that expression of 137 of these genes requires HAC1 Enriched GO term analysis includes response to ER stress, cell wall biogenesis and glycosylation. Orthologs of many of these are associated with UPR in Saccharomyces cerevisiae and Candida albicans Unconventional splicing of an intron from HAC1 mRNA is required to produce a functional transcription factor. The spliced intron varies in length from 19 bases in C. albicans to 379 bases in Candida glabrata, but has not been previously identified in Candida parapsilosis and related species. We used RNA-seq data and in silico analysis to identify the HAC1 intron in 12 species in the CTG-Ser1 clade. We show that the intron has undergone major contractions and expansions in this clade, reaching up to 848 bases. Exposure to DTT induced splicing of the long intron in C. parapsilosisHAC1, inducing the UPR.IMPORTANCE The unfolded protein response (UPR) responds to the build-up of misfolded proteins in the endoplasmic reticulum. The UPR has wide-ranging functions from fungal pathogenesis to applications in biotechnology. The UPR is regulated through the splicing of an unconventional intron in the HAC1 gene. This intron has been described in many fungal species and is of variable length. Until now it was believed that some members of the CTG-Ser1 clade such as C. parapsilosis did not contain an intron in HAC1, suggesting that the UPR was regulated in a different manner. Here we demonstrate that HAC1 plays an important role in regulating the UPR in C. parapsilosis We also identified an unusually long intron (626 bp) in C. parapsilosisHAC1 Further analysis showed that HAC1 orthologs in several species in the CTG-Ser1 clade contain long introns.

TPP riboswitch-dependent regulation of an ancient thiamin transporter in Candida.

Riboswitches are non-coding RNA molecules that regulate gene expression by binding to specific ligands. They are primarily found in bacteria. However, one riboswitch type, the thiamin pyrophosphate (TPP) riboswitch, has also been described in some plants, marine protists and fungi. We find that riboswitches are widespread in the budding yeasts (Saccharomycotina), and they are most common in homologs of DUR31, originally described as a spermidine transporter. We show that DUR31 (an ortholog of N. crassa gene NCU01977) encodes a thiamin transporter in Candida species. Using an RFP/riboswitch expression system, we show that the functional elements of the riboswitch are contained within the native intron of DUR31 from Candida parapsilosis, and that the riboswitch regulates splicing in a thiamin-dependent manner when RFP is constitutively expressed. The DUR31 gene has been lost from Saccharomyces, and may have been displaced by an alternative thiamin transporter. TPP riboswitches are also present in other putative transporters in yeasts and filamentous fungi. However, they are rare in thiamin biosynthesis genes THI4 and THI5 in the Saccharomycotina, and have been lost from all genes in the sequenced species in the family Saccharomycetaceae, including S. cerevisiae.

Identification of fungi in shotgun metagenomics datasets.

Metagenomics uses nucleic acid sequencing to characterize species diversity in different niches such as environmental biomes or the human microbiome. Most studies have used 16S rRNA amplicon sequencing to identify bacteria. However, the decreasing cost of sequencing has resulted in a gradual shift away from amplicon analyses and towards shotgun metagenomic sequencing. Shotgun metagenomic data can be used to identify a wide range of species, but have rarely been applied to fungal identification. Here, we develop a sequence classification pipeline, FindFungi, and use it to identify fungal sequences in public metagenome datasets. We focus primarily on animal metagenomes, especially those from pig and mouse microbiomes. We identified fungi in 39 of 70 datasets comprising 71 fungal species. At least 11 pathogenic species with zoonotic potential were identified, including Candida tropicalis. We identified Pseudogymnoascus species from 13 Antarctic soil samples initially analyzed for the presence of bacteria capable of degrading diesel oil. We also show that Candida tropicalis and Candida loboi are likely the same species. In addition, we identify several examples where contaminating DNA was erroneously included in fungal genome assemblies.

Identification of Non-Coding RNAs in the Candida parapsilosis Species Group.

The Candida CTG clade is a monophyletic group of fungal species that translates CTG as serine, and includes the pathogens Candida albicans and Candida parapsilosis. Research has typically focused on identifying protein-coding genes in these species. Here, we use bioinformatic and experimental approaches to annotate known classes of non-coding RNAs in three CTG-clade species, Candida parapsilosis, Candida orthopsilosis and Lodderomyces elongisporus. We also update the annotation of ncRNAs in the C. albicans genome. The majority of ncRNAs identified were snoRNAs. Approximately 50% of snoRNAs (including most of the C/D box class) are encoded in introns. Most are within mono- and polycistronic transcripts with no protein coding potential. Five polycistronic clusters of snoRNAs are highly conserved in fungi. In polycistronic regions, splicing occurs via the classical pathway, as well as by nested and recursive splicing. We identified spliceosomal small nuclear RNAs, the telomerase RNA component, signal recognition particle, RNase P RNA component and the related RNase MRP RNA component in all three genomes. Stem loop IV of the U2 spliceosomal RNA and the associated binding proteins were lost from the ancestor of C. parapsilosis and C. orthopsilosis, following the divergence from L. elongisporus. The RNA component of the MRP is longer in C. parapsilosis, C. orthopsilosis and L. elongisporus than in S. cerevisiae, but is substantially shorter than in C. albicans.