T-REX17 is a transiently expressed non-coding RNA essential for human endoderm formation

Curation statements for this article:
  • Curated by eLife

    eLife logo

    eLife assessment

    Based on a large set of complementary experiments, the authors propose that the lncRNA LNCSOX17 regulates human definitive endoderm differentiation, although its function is not related to the adjacent SOX17 gene in the same topological domain (TAD). The findings are important and supported by convincing data, although the molecular mechanism by which LNCSOX17 regulates endoderm differentiation stays unresolved.

This article has been Reviewed by the following groups

Read the full article

Abstract

Long non-coding RNAs (lncRNAs) have emerged as fundamental regulators in various biological processes, including embryonic development and cellular differentiation. Despite much progress over the past decade, the genome-wide annotation of lncRNAs remains incomplete and many known non-coding loci are still poorly characterized. Here, we report the discovery of a previously unannotated lncRNA that is transcribed 230 kb upstream of the SOX17 gene and located within the same topologically associating domain. We termed it T-REX17 ( T ranscript R egulating E ndoderm and activated by so X17 ) and show that it is induced following SOX17 activation but its expression is more tightly restricted to early definitive endoderm. Loss of T-REX17 affects crucial functions independent of SOX17 and leads to an aberrant endodermal transcriptome, signaling pathway deregulation and epithelial to mesenchymal transition defects. Consequently, cells lacking the lncRNA cannot further differentiate into more mature endodermal cell types. Taken together, our study identified and characterized T-REX17 as a transiently expressed and essential non-coding regulator in early human endoderm differentiation.

Article activity feed

  1. Author Response

    Reviewer #2 (Public Review):

    The paper has two key messages: the discovery and the function of LncSox17. Claims of gene discovery are today untrivial, given the large number of genome-wide datasets. Of course, I understand the authors cannot check everything but I feel some more clear and deep analysis of current databases is lacking.

    The reviewer is right when stating that there is an extremely high number of publicly available datasets and resources. In the current manuscript, we used Ensembl genes, Genecode V36 and Genecode V36 lncRNAs (commonly used datasets for gene and transcript annotation) and could not find reports of long non-coding RNAs with similar location, length and strand of T-REX17 (see Fig. 1). To further ensure that we did not overlook it, during the revision we inspected these datasets again, coming to the same conclusion that T-REX17 has not been previously reported at this locus.

    As we show, T-REX17 is only very transiently expressed in definitive endoderm and given that there are few available RNA-seq datasets covering this developmental transition from hiPSCs it is not entirely surprising that it has been missed in the past.

    Also, the exact coordinates of the lncRNA are not easy to find in the manuscript.

    This is certainly an important annotation we missed in the manuscript. We now updated the legend of Figure 1A to include the exact genomic location of T-REX17.

    Many statistical analyses are rather lacking. In particular I did not find details of how the DEGs were identified during differentiation (FDR? How many replicates?).

    We thank the reviewer for pointing this out. We now specify in the Methods section (page 42, lines 1037-1039) and in the figure legends (page 54, lines 1269-1271) how the DEGs have been identified, which thresholds have been used, and number of replicates performed.

    The results of the smFISH are surprising, since the level of expression seems rather low in comparison to the qPCR (only 4 times less expressed than Sox17) or the RNA-seq.

    Direct quantitative comparisons between smFISH and qPCR (or RNA-seq) assays are in general quite hard since the two technologies rely on different biochemical principles. qPCR and RNA-seq include an amplification step, and therefore their interpretation should be considered as relative rather than absolute. On the other hand, smFISH offers a more absolute quantitative information and provides clues about the subcellular localization of the investigated target. At the same time, in smFISH experiments, individual foci could represent the accumulation of more than one molecule, making it hard to accurately infer gene expression levels from images. Throughout the manuscript we combine the two assays in an attempt to provide more robust information about T-REX17 expression dynamics.

    We would also like to note the high specificity of our smFISH signal, given that we do not observe any detectable foci for T-REX17 in undifferentiated cells (Fig. 2C) or T-REX17 depleted endoderm cells (Fig. 3C).

  2. eLife assessment

    Based on a large set of complementary experiments, the authors propose that the lncRNA LNCSOX17 regulates human definitive endoderm differentiation, although its function is not related to the adjacent SOX17 gene in the same topological domain (TAD). The findings are important and supported by convincing data, although the molecular mechanism by which LNCSOX17 regulates endoderm differentiation stays unresolved.

  3. Reviewer #1 (Public Review):

    Landshammer et al. characterized the role of LNCSOX17, a previously not annotated lncRNA, in the regulation of human endoderm differentiation, further reinforcing the importance of lncRNAs in the regulation of human stem cells differentiation and embryonic development. LNCSOX17 is a unique lncRNA as it does not regulate neighboring SOX17 gene within the TAD.

    Employing different loss-of-function methods (i.e. CRISPR-Cas9 MECP2, CRISPR-pAS), the authors manage to untangle the complexity of the LNCSOX17 locus, showing that it contains a distal enhancer of SOX17, a transcription factor crucial for the determination of endodermal cell fate, and, on the other hand, it operates as RNA transcript to guarantee endodermal cell differentiation.
    Although lncRNA LNCSOX17 does not regulate SOX17 levels and chromatin occupancy, the authors show that its loss leads to the impairment of definitive endodermal differentiation, in line with the downregulation of endoderm-related genes and markers (eg CXCR4). These data fit well with the LNCSOX17 expression profile, which indeed appears to be restricted to early human definitive endoderm. The combination of multiple genomic techniques to manipulate the LNCSOX17 locus, together with the evidence of a clear phenotype upon loss of this lncRNA, constitutes the strength of the paper.

    The mechanism of how LNCSOX17 regulates endoderm differentiation is not clear and should be strengthened. The reader had a feeling that the identification of the LNCSOX17 molecular mechanism in definitive endoderm differentiation was not the focus of the work, but at the same time, it was also clear that the authors put a lot of effort to address this biological question by employing several-omics approaches (i.e. RNA pulldown, RNA-seq, CRISPR, HI-C).

    Overall the conclusions are supported by the data but some methods used in this manuscript (eg RIP, Pulldown) should be strengthen with alternative tools. The manuscript is easy to read and the figures are nicely represented.

  4. Reviewer #2 (Public Review):

    The paper has two key messages: the discovery and the function of LncSox17. Claims of gene discovery are today untrivial, given the large number of genome-wide datasets. Of course, I understand the authors cannot check everything but I feel some more clear and deep analysis of current databases is lacking. Also, the exact coordinates of the lncRNA are not easy to find in the manuscript.

    Many statistical analyses are rather lacking. In particular I did not find details of how the DEGs were identified during differentiation (FDR? How many replicates?).

    The results of the smFISH are surprising, since the level of expression seems rather low in comparison to the qPCR (only 4 times less expressed than Sox17) or the RNA-seq.

  5. Reviewer #3 (Public Review):

    Definitive endoderm is an important transient, progenitor tissue formed in the embryo that gives rise to most of the internal organ systems. Studying how definitive endoderm arises in development is important for understanding several common diseases and also for improving methods to specialise pluripotent stem cells in culture towards functional cell types with applications in regenerative medicine. The aim of the current study was to identify and characterise new genetic factors that contribute to these processes. The authors identified a previously-overlooked gene that they named LNCSOX17 and showed that this gene is needed for cells in culture to maintain their definitive endoderm identity. Similar genes have been shown previously to function by controlling other nearby genes, but the authors showed that this is not what is happening for LNCSOX17. Instead, it is likely that LNCSOX17 affects other processes in the cell, beyond the nearby gene. This research provides a nice example of how a noncoding gene that is expressed in a very restricted developmental stage can have strong effects on cell lineage control. Because there are thousands of other long, noncoding transcripts, most of which are largely uncharacterised, this study emphasises the urgent need to examine this type of transcript in further detail.

    Overall, the main conclusions of the manuscript are well supported by the evidence.

    A key strength of the work is that the authors use state-of-the-art genetic methods in human pluripotent stem cells to address the function and regulation of LNCSOX17 and nearby regulatory elements. It is clear that disabling LNCSOX17 does not affect SOX17, establishing that the long noncoding transcript does not function in cis.

    Robust cellular assays also provide strong evidence that the LNCSOX17 transcript is required for the continued development of endoderm cells (but not for the initial specification).

    Whether LNCSOX17 operates in trans is not fully established, but the authors present evidence that supports this viewpoint and they put forward a plausible model for how this might be mediated (albeit very preliminary, as they acknowledge).