The covariance environment defines cellular niches for spatial inference.

A key challenge of analyzing data from high-resolution spatial profiling technologies is to suitably represent the features of cellular neighborhoods or niches. Here we introduce the covariance environment (COVET), a representation that leverages the gene-gene covariate structure across cells in the niche to capture the multivariate nature of cellular interactions within it. We define a principled optimal transport-based distance metric between COVET niches that scales to millions of cells. Using COVET to encode spatial context, we developed environmental variational inference (ENVI), a conditional variational autoencoder that jointly embeds spatial and single-cell RNA sequencing data into a latent space. ENVI includes two decoders: one to impute gene expression across the spatial modality and a second to project spatial information onto single-cell data. ENVI can confer spatial context to genomics data from single dissociated cells and outperforms alternatives for imputing gene expression on diverse spatial datasets.

The covariance environment defines cellular niches for spatial inference.

The tsunami of new multiplexed spatial profiling technologies has opened a range of computational challenges focused on leveraging these powerful data for biological discovery. A key challenge underlying computation is a suitable representation for features of cellular niches. Here, we develop the covariance environment (COVET), a representation that can capture the rich, continuous multivariate nature of cellular niches by capturing the gene-gene covariate structure across cells in the niche, which can reflect the cell-cell communication between them. We define a principled optimal transport-based distance metric between COVET niches and develop a computationally efficient approximation to this metric that can scale to millions of cells. Using COVET to encode spatial context, we develop environmental variational inference (ENVI), a conditional variational autoencoder that jointly embeds spatial and single-cell RNA-seq data into a latent space. Two distinct decoders either impute gene expression across spatial modality, or project spatial information onto dissociated single-cell data. We show that ENVI is not only superior in the imputation of gene expression but is also able to infer spatial context to disassociated single-cell genomics data.

O-GlcNAcylation and Regulation of Galectin-3 in Extraembryonic Endoderm Differentiation.

The regulation of proteins through the addition and removal of O-linked ß-N-acetylglucosamine (O-GlcNAc) plays a role in many signaling events, specifically in stem cell pluripotency and the regulation of differentiation. However, these post-translational modifications have not been explored in extraembryonic endoderm (XEN) differentiation. Of the plethora of proteins regulated through O-GlcNAc, we explored galectin-3 as a candidate protein known to have various intracellular and extracellular functions. Based on other studies, we predicted a reduction in global O-GlcNAcylation levels and a distinct galectin expression profile in XEN cells relative to embryonic stem (ES) cells. By conducting dot blot analysis, XEN cells had decreased levels of global O-GlcNAc than ES cells, which reflected a disbalance in the expression of genes encoding O-GlcNAc cycle enzymes. Immunoassays (Western blot and ELISA) revealed that although XEN cells (low O-GlcNAc) had lower concentrations of both intracellular and extracellular galectin-3 than ES cells (high O-GlcNAc), the relative secretion of galectin-3 was significantly increased by XEN cells. Inducing ES cells toward XEN in the presence of an O-GlcNAcase inhibitor was not sufficient to inhibit XEN differentiation. However, global O-GlcNAcylation was found to decrease in differentiated cells and the extracellular localization of galectin-3 accompanies these changes. Inhibiting global O-GlcNAcylation status does not, however, impact pluripotency and the ability of ES cells to differentiate to the XEN lineage.

Lactate Enhances Mouse ES Cell Differentiation Toward XEN Cells In Vitro.

Metabolism plays a crucial role for cell survival and function; however, recent evidence has implicated it in regulating embryonic development. In the embryo, the inner cell mass undergoes orchestrated cellular divisions resulting in the formation of pluripotent epiblast stem cells and primitive endoderm cells. However, both lineages can be captured in vitro as embryonic stem (ES) cells and extraembryonic endoderm (XEN) cells. Concomitantly, changes in the metabolic profile occurs during development, and are well documented in the embryonic lineages. However, a comprehensive multi-omic analysis of these features in XEN cells remains lacking. We observed that mouse XEN cells exhibited high sensitivity to glycolytic inhibition in addition to maintaining elevated intra- and extracellular lactate levels in vitro. Extraembryonic endoderm cells maintain high lactate levels by increased LDHA activity, and re-routing pyruvate away from the mitochondria resulting in reduced mitochondrial activity due to disruptions in electron transport chain stoichiometry. Importantly, exogenous lactate supplementation or promoting intracellular lactate accumulation enhances XEN differentiation in vitro. These results highlight how lactate contributes to XEN differentiation in vitro and may serve to enhance reprogramming efficiency of cells used for regenerative medicine.

Serum-dependent and -independent regulation of PARP2.

PARP2 belongs to a family of proteins involved in cell differentiation, DNA damage repair, cellular energy expenditure, and chromatin modeling. In addition to these overlapping functions with PARP1, PARP2 participates in spermatogenesis, T-cell maturation, extra-embryonic endoderm formation, adipogenesis, lipid metabolism, and cholesterol homeostasis. Knowledge of the functions of PARP2 is far from complete, and the mechanism(s) by which the gene and protein are regulated are unknown. In this study, we found that two different mechanisms are used in vitro to regulate PARP2 levels. In the presence of serum, PARP2 is degraded through the ubiquitin-proteasome pathway; however, when serum is removed or dialyzed with a 3.5 kDa molecular cut membrane, PARP2 rapidly becomes sodium dodecyl sulphate- and urea-insoluble. Despite the presence of a putative serum response element in the PARP2 gene, transcription is not affected by serum deprivation, and PARP2 levels are restored when serum is replaced. The loss of PARP2 affects cell differentiation and gene expression linked to cholesterol and lipid metabolism. These observations highlight the critical roles that PARP2 plays under different physiological conditions, and reveal that PARP2 is tightly regulated by distinct pathways.

Metabolic profile and differentiation potential of extraembryonic endoderm-like cells.

Glucose metabolism has a crucial role for providing substrates required to generate ATP and regulate the epigenetic landscape. We reported that F9 embryonal carcinoma stem-like cells require cytosolic reactive oxygen species to differentiate into extraembryonic endoderm; however, mitochondrial sources were not examined. To extend these studies, we examined the metabolic profile of early and late-passage F9 cells, and show that their ability to differentiate is similar, even though each population has dramatically different metabolic profiles. Differentiated early-passage cells relied on glycolysis, while differentiated late-passage cells transitioned towards oxidative phosphorylation (OXPHOS). Unexpectedly, electron transport chain protein stoichiometry was disrupted in differentiated late-passage cells, whereas genes encoding mitofusion 1 and 2, which promote mitochondrial fusion and favor OXPHOS, were upregulated in differentiated early-passage cells. Despite this, early-passage cells cultured under conditions to promote glycolysis showed enhanced differentiation, whereas promoting OXPHOS in late-passage cells showed a similar trend. Further analysis revealed that the distinct metabolic profiles seen between the two populations is largely associated with changes in genomic integrity, linking metabolism to passage number. Together, these results indicate that passaging has no effect on the potential for F9 cells to differentiate into extraembryonic endoderm; however, it does impact their metabolic profile. Thus, it is imperative to determine the molecular and metabolic status of a stem cell population before considering its utility as a therapeutic tool for regenerative medicine.

Wnt and Hedgehog Signaling Regulate the Differentiation of F9 Cells into Extraembryonic Endoderm.

Mouse F9 cells differentiate into primitive extraembryonic endoderm (PrE) when treated with retinoic acid (RA), and this is accompanied by an up-regulation of Gata6. The role of the GATA6 network in PrE differentiation is known, and we have shown it directly activates Wnt6. Canonical Wnt/ß-catenin signaling is required by F9 cells to differentiate to PrE, and this, like most developmental processes, requires input from one or more additional pathways. We found both RA and Gata6 overexpression, can induce the expression of Indian Hedgehog (Ihh) and a subset of its target genes through Gli activation during PrE induction. Chemical activation of the Hh pathway using a Smoothened agonist (SAG) also increased Gli reporter activity, and as expected, when Hh signaling was blocked with a Smoothened antagonist, cyclopamine, this RA-induced reporter activity was reduced. Interestingly, SAG alone failed to induce markers of PrE differentiation, and had no effect on Wnt/ß-catenin-dependent TCF-LEF reporter activity. The expected increase in Wnt/ß-catenin-dependent TCF-LEF reporter activity and PrE markers induced by RA was, however, blocked by cyclopamine. Finally, inhibiting GSK3 activity with BIO increased both TCF-LEF and Gli reporter activities. Together, we demonstrate the involvement of Hh signaling in the RA-induced differentiation of F9 cells into PrE, and while the activation of the Hh pathway itself is not sufficient, it as well as active Wnt/ß-catenin are necessary for F9 cell differentiation.

Mechanisms Regulating Stemness and Differentiation in Embryonal Carcinoma Cells.

Just over ten years have passed since the seminal Takahashi-Yamanaka paper, and while most attention nowadays is on induced, embryonic, and cancer stem cells, much of the pioneering work arose from studies with embryonal carcinoma cells (ECCs) derived from teratocarcinomas. This original work was broad in scope, but eventually led the way for us to focus on the components involved in the gene regulation of stemness and differentiation. As the name implies, ECCs are malignant in nature, yet maintain the ability to differentiate into the 3 germ layers and extraembryonic tissues, as well as behave normally when reintroduced into a healthy blastocyst. Retinoic acid signaling has been thoroughly interrogated in ECCs, especially in the F9 and P19 murine cell models, and while we have touched on this aspect, this review purposely highlights how some key transcription factors regulate pluripotency and cell stemness prior to this signaling. Another major focus is on the epigenetic regulation of ECCs and stem cells, and, towards that end, this review closes on what we see as a new frontier in combating aging and human disease, namely, how cellular metabolism shapes the epigenetic landscape and hence the pluripotency of all stem cells.

NOX1 and NOX4 are required for the differentiation of mouse F9 cells into extraembryonic endoderm.

Mouse F9 cells differentiate to primitive endoderm (PrE) when treated with retinoic acid (RA). Differentiation is accompanied by increased reactive oxygen species (ROS) levels, and while treating F9 cells with antioxidants attenuates differentiation, H2O2 treatment alone is sufficient to induce PrE. We identified the NADPH oxidase (NOX) complexes as candidates for the source of this endogenous ROS, and within this gene family, and over the course of differentiation, Nox1 and Nox 4 show the greatest upregulation induced by RA. Gata6, encoding a master regulator of extraembryonic endoderm is also up-regulated by RA and we provide evidence that NOX1 and NOX4 protein levels increase in F9 cells overexpressing Gata6. Pan-NOX and NOX1-specific inhibitors significantly reduced the ability of RA to induce PrE, and this was recapitulated using a genetic approach to knockdown Nox1 and/or Nox4 transcripts. Interestingly, overexpressing either gene in untreated F9 cells did not induce differentiation, even though each elevated ROS levels. Thus, the data suggests that ROS produced during PrE differentiation is dependent in part on increased NOX1 and NOX4 levels, which is under the control of GATA6. Furthermore, these results suggest that the combined activity of multiple NOX proteins is necessary for the differentiation of F9 cells to primitive endoderm.

Frizzled gene expression and negative regulation of canonical WNT-ß-catenin signaling in mouse F9 teratocarcinoma cells.

Mouse F9 cells differentiate into primitive endoderm (PrE) following the activation of the canonical WNT-ß-catenin pathway. The upregulation of Wnt6 and activation of ß-catenin-TCF-LEF-dependent transcription is known to accompany differentiation, but the Frizzled (FZD) receptor responsible for transducing the WNT6 signal is not known. Eight of the 10 Fzd genes were found to be expressed in F9 cells, with Fzd7 being the most highly expressed, and chosen for further analysis. To alter steady-state Fzd7 levels and test the effect this has on differentiation, siRNA and overexpression approaches were used to knock-down and ectopically express the Fzd7 message, respectively. siRNA knock-down of Fzd7 resulted in reduced DAB2 levels, and the overexpression activated a TCF-LEF reporter, but neither approach affected differentiation. Our focus turned to how canonical WNT6 signaling was attenuated to allow PrE cells to form parietal endoderm (PE). Dkk1, encoding a WNT antagonist, was examined and results showed that its expression increased in F9 cells treated with retinoic acid (RA) or overexpressing Wnt6. F9 cells overexpressing human DKK1 or treated with DKK1-conditioned medium and then treated with RA failed to differentiate, indicating that a negative feedback loop involving WNT6 and DKK1 attenuates canonical WNT-ß-catenin signaling, thereby allowing PE cells to differentiate.