Science shines in a new virtual SY-Stem

In 2020, the in-person meetings that used to be at the core of a scientist's life came to a screeching halt due to the COVID-19 pandemic. One such cancelled meeting was the third Symposium for the Next Generation of Stem Cell Researchers (SY-Stem), normally held at the Vienna BioCenter in Austria. Undeterred, the organizers Elly Tanaka (IMP, Vienna, Austria), Juergen Knoblich, Ulrich Elling, Bon-Kyoung Koo and Sasha Mendjan (all IMBA, Vienna, Austria) revived the meeting in a virtual setting in 2021 (Fig. 1). They successfully designed a format where scientific talks via Zoom were combined with poster sessions and coffee breaks in an online networking platform.

With the decision to hold the conference virtually, the cost of the meeting was reduced significantly, making the meeting accessible to a larger number of researchers from across the globe. Indeed, this year's SY-Stem meeting was the largest to date, bringing together more than 300 researchers from over 28 countries. Although virtual conferences can't quite replicate or even replace the organic networking experience of face-to-face meetings, the online networking platform (‘Remo’) supported well-attended ‘meet-the-speaker chats’, where students and PIs mingled and exchanged ideas, as well as poster sessions that allowed the presenters to give an overview of their work to a small audience by sharing their results as a PDF or a presentation. Chance encounters that happen during in-person meetings were replaced by the random assignment of individuals into a poster room when entering the networking platform, dropping them in on lively discussions and giving these sessions a familiar feel of community. Moreover, after being welcomed by Juergen Knoblich, the attendants were treated to the first highlight of the meeting (Fig. 2), when Elly Tanaka together with members of her own lab and the Stark group from the IMP performed a piece by Austrian composer Joseph Haydn. Sara Sajko (on piano) and Daniel F. Azar (on vocals), both former PhD students at the Max Perutz Labs Vienna, then concluded SY-Stem 2021 in beauty with pieces by Gabriel Fauré, a second musical highlight of the meeting.

Over 3 days, the symposium covered different aspects of stem cell research ranging from early embryonic development and organogenesis to disease modelling and clinical applications. The symposium was anchored by outstanding keynote presentations from Amy Wagers (Harvard University, Cambridge, MA, USA) and Lorenz Studer (Memorial Sloan Kettering Cancer Institute, New York, USA). However, unlike other meetings, SY-Stem highlights next-generation researchers, including Master's and PhD students making their first big discoveries, postdocs who will be forming their own research groups, and young group leaders starting to shape the future of developmental and stem cell biology. Here, we summarize the excellent work presented by these researchers.

One recurring theme of the meeting related to species-intrinsic differences in developmental timing, specifically the relatively slow pace of human development, which can limit the maturity of cell types differentiated from human pluripotent stem cells (hPSCs). This relatively slow maturation of human cells is a challenge for the field, as disease modelling and regenerative medicine approaches often require cells that have more mature functional properties. Keynote speaker Lorenz Studer presented a pharmacological screen carried out by his PhD student Emiliano Hergenreder, who discovered a combination of four small molecules (nicknamed GENtoniK) that promotes the maturation of multiple cell types in both 2D and 3D culture. If broadly applicable, such methods have the potential to both improve the quality of hPSC-derived cells, and to shed light on the biological mechanisms underlying cell- and species-specific maturation processes in vitro.

Cantas Alev, a junior group leader at the ASHBi Institute at Kyoto University (Japan), presented mostly unpublished data on the segmentation clock, which governs the development of somites from paraxial mesoderm in vivo. This process can be recapitulated in vitro in a 3D culture system of hPSC-derived paraxial mesodermal cells that spontaneously form travelling waves of HES7 reporter expression during the coordinated formation of epithelial somites. This work is an extension of a recent study in which Cantas and colleagues could recapitulate aspects of the segmentation clock in vitro in a 2D context, without the formation of segments or somite-like structures (Matsuda et al., 2020). Interestingly, the period of these in vitro observed reporter oscillations reflects species-specific differences in the segmentation clock. This is a topic that Miki Ebisuya (EMBL, Barcelona, Spain) is studying in her group, using an expanded ‘zoo’ of model organisms. The HES7-based molecular oscillator is created by a negative-feedback loop whereby the HES7 protein inhibits its own gene promoter. As such, differences in the timing of the oscillator could reflect differences in transcription, RNA maturation, translation, protein degradation or other cellular factors. Indeed, dynamic SILAC labelling studies presented by Miki suggest that most human proteins in in vitro-derived presomitic mesoderm have a longer half-life than mouse proteins in a manner that is independent of their sequence, suggesting species-specific differences in development are partially due to intrinsic differences in the speed of their biochemical reactions (Matsuda et al., 2020).

Ruth Hornbacher, a Master's student in the group of Paulina Latos (Medical University Vienna, Austria), presented work on the role of MSX2 in syncytiotrophoblast development in the preimplantation human embryo. Using human trophoblast stem cells as a model, Ruth and her colleagues used gene knockdown followed by transcriptional profiling and ChIP-seq to show how MSX2 interacts with the SWI/SNF complex to bind to and silence genes expressed in syncytiotrophoblast cells (Hornbachner et al., 2021 preprint). Although in vitro model systems are powerful, the developing embryo remains the gold standard for studying morphogenesis and organogenesis. However, observing the development of post-implantation mammalian embryos is complicated as their ex utero development can be supported for only limited periods of time. Alejandro Aguilera Castrejon together with his PhD mentor Jacob Hanna (Weizmann Institute of Science, Rehovot, Israel) improved existing roller in vitro embryo culture systems by controlling O₂ levels, CO₂ levels and atmospheric gas pressure, enabling explanted mouse embryos to survive for almost 4 days, or up to E11 (Aguilera-Castrejon et al., 2021). This culture regime has the potential to revolutionize how we are able to work with the post-implantation mouse embryo. The quality of talks from these excellent budding scientists was remarkable and was one feature that made the symposium uniquely enjoyable.

Three-dimensional organoid cultures derived from hPSCs or adult tissues have exploded in popularity in the last few years, as their ability to self-organize and generate cellular diversity allows aspects of human development to be studied in vitro. Multiple talks focused on intestinal organoids. Georg Busslinger, a newly minted PI at CeMM in Vienna, Austria, presented his postdoctoral work from the lab of Hans Clevers (Hubrecht Institute, Utrecht, The Netherlands), showing that organoids derived from the esophagus, different parts of the stomach and duodenum are clearly distinct from each other, likely reflecting different functional roles (Busslinger et al., 2021). Interestingly, single-cell RNA sequencing (scRNA-seq) revealed a small subpopulation of cells expressing high levels of BEST4 and CHTF (BCHE) that regulate water secretion into the intestinal lumen to wash away the mucus layer and enable nutrient absorption by nearby cells. Also focusing on intestinal organoids, Denise Serra presented work from her time as a PhD student in the group of Prisca Liberali (Friedrich Miescher Institute for Biomedical Research, Basel, Switzerland). She discussed how the generation of cellular diversity in organoids requires spontaneous symmetry breaking mediated by Notch/DLL1 lateral inhibition that appears to be initiated at the 16-cell stage in intestinal organoids (Serra et al., 2019). Kim Jensen (BRIC and DanStem/University of Copenhagen, Denmark) presented unpublished work comparing fetal and adult intestinal organoids. Culturing primary intestinal cells from the developing fetus generates spheroids rather than organoids, but these can be induced to form organoids upon exposure to WNT agonists at seeding. Using CAGE-seq, ATAC-seq and scRNA-seq, Kim's group has been analyzing the transcriptional regulation of cellular identity, while a phenotypic CRISPR screen in fetal spheroids allowed them to identify genes whose loss of function can accelerate or block the transition towards adult-like organoids. This elegant system is an excellent demonstration of how combinations of new technologies (organoids, CRISPR screens, single-cell analysis and multi-omics) can accelerate the discovery of molecular mechanisms regulating tissue maturation.

While organoids can be used to study human development, Grayson Camp (Institute of Molecular and Clinical Ophthalmology, Basel, Switzerland) together with Jason Spence (University of Michigan, Ann Arbor, USA) are using intestinal organoids to explore evolutionary conservation and divergence between humans and chimpanzees. First, the team analyzed human intestine development and assessed the fidelity of organoids by comparing them with a reference atlas. Organoid transplantation into the kidney capsule promotes maturation, and the team used both morphological and scRNA-seq analysis to reveal the developmental time course of intestinal crypt and villi structure formation at unprecedented resolution (Yu et al., 2020 preprint). Besides revealing insights into the temporal order of developmental transitions, the two labs are also exploring the signaling interactions between cells and identified NRG1 as a potential factor that can enhance the maturation of intestinal organoids. What the team has achieved with human cells, they are now transferring to chimpanzee stem cells, which promises to shed light on features that are unique to humans.

Jens Puschhof, a PhD student from the group of Hans Clevers, shared a rather unconventional use of intestinal organoids. He injected them with different types of bacteria to study how bacterial toxins might affect DNA integrity (Pleguezuelos-Manzano et al., 2020). Mutation signature analysis of cloned and sequenced organoids revealed an AAATT T>N motif associated with the presence of pks+ bacteria that produce colibactin, a compound thought to alkylate DNA on adenine residues and to induce double-strand breaks. This discovery could be clinically relevant as some colorectal cancers have a high concentration of pks+ E. coli, and pks+ bacteria are included in some probiotics. Jens also showed that cells from the venom glands of poisonous snakes can be cultured as organoids that generate heterogeneous populations of secretory cells that produce either three-finger or c-type lectin toxins (Post et al., 2020). The production of sufficient quantities of toxins to aid in antivenom production remains challenging, but these studies highlight the versatility and therapeutic potential of organoids. Finally, Sasha Mendjan (IMBA, Vienna, Austria) systematically optimized culture conditions from a homogeneous mesodermal population to develop a self-organizing ‘cardioid’ model system (Hofbauer et al., 2020 preprint). These fluid-filled structures are composed of cardiomyocytes and are lined by endocardial-like cells; they show great promise as a system for studying heart development and damage.

In the last SY-Stem meeting, scRNA-seq was still an emerging technique. However, commercialization of the technology in recent years has effectively added it to the standard toolbox of many labs. In addition to the work already highlighted, Karl Köhler, a junior PI from the Children's Hospital/Harvard Medical School (Boston, MA, USA), presented beautiful studies of sensory organoid development from hPSCs (Koehler et al., 2017; Lee et al., 2020). Single-cell analysis nicely complemented immunohistochemical analysis of otic placode organoids, revealing a full spectrum of constituent cell lineages. Jinwook Choi, a postdoc from the lab of Joo-Hyeon Lee (Wellcome-MRC Cambridge Stem Cell Institute, UK) used single-cell analysis of primary lung organoid cultures to reveal a population of damage-associated transient progenitors (DATPs) that differentiate into alveolar cells once inflammation is resolved, but persist and accumulate if IL-1b signaling remains high, as seen in the lungs of patients who died from COVID-19 (Choi et al., 2020).

Annotating cell clusters from single-cell data can be challenging, especially when clusters are indistinct, or unexpected. Organoids provide a great opportunity to recapitulate brain development but recent reports have claimed that in vitro stress, caused by low oxygen and nutrient supply at the center of organoids, has a global, detrimental effect on cell specification. Ábel Vértesy from the lab of Jürgen Knoblich analyzed datasets from hPSC-derived cortical organoids generated by many groups and found that stress impacted a distinct subpopulation of cells, giving rise to a distinct cell cluster with signatures of starvation, glycolysis and ER/UPR stress. Removing these cells from the analysis can dramatically improve cell clustering and recapitulate textbook developmental trajectories.

scRNA-seq has also been elegantly used to better understand the process by which pluripotent stem cells initiate differentiation. Christa Buecker, a junior PI from the Max Perutz Labs in Vienna, Austria used scRNA-seq to study the exit from naive pluripotency in differentiation-impaired mutants, aiming to identify how gene regulatory networks are rewired if differentiation is perturbed. Finally, Florian Merkle (University of Cambridge, UK) presented a collaborative study in which over 200 genetically distinct hiPSC lines were pooled and differentiated into dopaminergic neurons (Jerber et al., 2021). This effort identified transcriptional signatures in the pluripotent state that predict neuronal differentiation efficiency, allowing groups to avoid working with cell lines recalcitrant to neuronal differentiation. Furthermore, analysis of over 1 million single cells from these cell lines revealed over 1000 disease-associated quantitative trait loci, of which nearly half had not been seen in the GTEx catalogue.

New facets of stem cell biology were revealed by other methods in addition to scRNA-seq. Emiel van Genderen, a PhD student in Derk ten Berge's group at Erasmus MC Rotterdam (The Netherlands), described a pluripotent ‘rosette stem cell’ state believed to be distinct from and sitting between the naive and primed states, in which cells have bivalent chromatin marks but do not appear to be fate committed (Neagu et al., 2020). Graziano Martello from the University of Padua (Italy) shared a recently published study showing that mitochondrial localisation of the transcription factor STAT3 may mediate the effects of LIF on the DNA methylases DNMT3A/B via the glutamine derivative α-ketoglutarate (Betto et al., 2021). Silvia Santos from the Francis Crick Institute (London, UK) shared work suggesting that FGF2-treated murine pluripotent stem cells form gastruloids more quickly than those treated with FGF4 and resemble E4.5 mouse embryos, although the origin of FGF2 at this developmental stage in vivo remains unresolved (Gharibi et al., 2020 preprint).

Lineage-tracing studies of prostate progenitors presented by Elisavet Tika (Universite Libre de Bruxelles, Belguim), a PhD student in Cedric Blanpain's group, revealed that basal cells in the prostate are multipotent during development, whereas luminal cells only give rise to lumen cells, but that at later stages only rare basal cells enriched in ECM factors retain multipotency (Centonze et al., 2020; Tika et al., 2019). Alice Rossi (King's College London, UK), a PhD student co-supervised by Sousa Nunes and Francois Guillemot, carried out a screen in Drosophila to identify genes that can affect the formation of quiescent neural stem cells, and identified a nucleoporin whose downregulation led to the nuclear retention of mRNAs in both flies and mice (Rossi et al., 2021 preprint). These studies are a useful reminder that tried-and-tested methods applied to in vivo model systems remain the foundation of stem cell and developmental biology research.

The symposium also featured mechanistic studies of regeneration and adaptive responses to stress. Meritxell Huch (Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany) presented work from her group showing that the TET1 demethylase is crucially important during cell fate transitions in liver regeneration; many genes gain hydroxymethylation marks immediately before cell division and TET1 depletion in the ductal department abolishes duct-mediated liver regeneration (Aloia et al., 2019). She also shared exciting unpublished studies performed with a custom-made microfluidic device to systematically alter organoid composition in order to study cell:cell interactions during liver regeneration.

Kate Miroshnikova, a postdoc in Sarah Wikström's lab (Helsinki Institute of Life Science, Finland) who is about to start her own group at NIH, showed that epithelial cells rapidly reorganise their chromatin by demethylation at histone H3K9 to reduce nuclear stiffness in response to shearing stress in a manner that may be mediated by the force-sensing protein PIEZO1 (Nava et al., 2020). One of the keynote speakers, Amy Wagers, shared a recent study from her group suggesting that the gene FOS is involved in the activation of muscle satellite cells. She revealed that FOS is induced upon injury, that FACS-purified FOS-GFP satellite cells are able to regenerate damaged cells more efficiently, and that FOS conditional knockouts exhibit long-term deficits in muscle repair (Almada et al., 2021).

Perhaps one of the most exciting applications of stem cell biology relates to the study and treatment of human disease. Alba Tristan-Noguero, a postdoc from Garcia Cazorla's lab (Fundació Sant Joan de Déu, Barcelona, Spain), who is closely collaborating with Antonella Consiglio's lab (IDIBELL/UB, Spain), modelled tyrosine hydroxylase deficiency and showed that hPSC-derived neurons lacking tyrosine hydroxylase exhibit morphological defects, providing a novel system with which to better understand the mechanisms underlying this rare disease. Florian Merkle presented work showing that hPSC-derived hypothalamic neurons produce the potent appetite-suppressing neuropeptide β-MSH in a manner that is dynamically regulated by the hormone leptin, establishing this culture system as a powerful tool for studying obesity. Nicola Valeri (Institute of Cancer Research, London, UK) reported how his group has been using perioperative biopsies from liquid or solid human cancers to generate organoids and to help inform clinical approaches, e.g. by testing drug responsiveness before drugs are given to patients. Furthermore, Lorenz Studer revealed that melanomas are likely derived from melanocyte precursors that express ATAD2, a gene that interacts with the melanoma-associated gene SOX10 and is associated with poor prognosis in individuals with melanoma (Baggiolini et al., 2020 preprint). Importantly, the enzyme ATAD2 can be targeted by drugs, suggesting new therapeutic strategies for this deadly disease.

In vitro disease modelling is powerful, but it is becoming clear that some diseases, such as Duschene's Muscular Dystrophy can also be treated in vivo, e.g. by correcting the frameshift mutation in the dystrophin gene using CRISPR/Cas9-mediated homology directed-repair. Indeed, the group of Amy Wagers has used adeno-associated viruses to restore in-frame dystrophin in about 1% of muscle satellite cells, which might be able to functionally rescue a larger percentage of muscle cells (Goldstein et al., 2019). Gene editing strategies that introduce double-strand breaks are problematic as they can lead to the positive selection of cancer-associated mutations. Gerald Schwank (University of Zurich, Switzerland) is addressing this issue using Cas9 cytidine base editors. As proof of principle, he showed that he could significantly reduce expression of the targeted gene PCSK9 in vivo. Blocking PCSK9 increases the concentration of LDL receptors on the cell surface, and the manipulation therefore led to an expected decrease in circulating levels of LDL (Marquart et al., 2020 preprint).

One of the most exciting yet challenging promises of regenerative medicine is cell replacement therapy. This was pursued decades ago in the case of Parkinson's disease, whereby dopaminergic neurons lost in disease were replaced by explants from the midbrain of fetal donors. However, the results were mixed due to imperfect patient selection, endpoint definition and heterogeneous cell populations present in the donor material. The group of Lorenz Studer has now developed a robust two-stage protocol to generate relatively pure populations of dopaminergic neurons from hPSCs by giving a high concentration pulse of a WNT agonist at a specific developmental time point to mimic signals from the midbrain/hindbrain boundary (Kim et al., 2021; Piao et al., 2021). This approach enables the production of billions of dopaminergic neurons that can be cryopreserved and delivered to hospitals around the world for transplantation. These exciting advances concretely demonstrate the value of basic research in developmental and stem cell biology, and herald other therapies that could emerge from the combined efforts of these fields.

Although SY-Stem participants were unable to see the beautiful surroundings in Vienna or benefit from serendipitous interactions, the conference was still a great success. Three days of exciting science were punctuated by discussion and networking. During this time, it became clear that technological advances have enabled leaps in our understanding, but that clever approaches for unravelling the enduring mystery of how a single cell can develop into something as complex as a tissue or organism are why the science in this field is so compelling. Most striking was the high quality of science and presentation by some of the youngest participants. Clearly, the future is bright in the hands of these promising early career scientists.

We thank the IMP/IMBA Graphics Department for permission to use the SY-Stem meeting poster illustration in this publication and the IMBA for the photograph of the Stanaka Quartett. We specifically thank Lea Klement for organizing the non-scientific aspects of the meeting. We also thank Cantas Alev for his helpful comments on an earlier draft of this manuscript.