DELVE: feature selection for preserving biological trajectories in single-cell data.

Single-cell technologies can measure the expression of thousands of molecular features in individual cells undergoing dynamic biological processes. While examining cells along a computationally-ordered pseudotime trajectory can reveal how changes in gene or protein expression impact cell fate, identifying such dynamic features is challenging due to the inherent noise in single-cell data. Here, we present DELVE, an unsupervised feature selection method for identifying a representative subset of molecular features which robustly recapitulate cellular trajectories. In contrast to previous work, DELVE uses a bottom-up approach to mitigate the effects of confounding sources of variation, and instead models cell states from dynamic gene or protein modules based on core regulatory complexes. Using simulations, single-cell RNA sequencing, and iterative immunofluorescence imaging data in the context of cell cycle and cellular differentiation, we demonstrate how DELVE selects features that better define cell-types and cell-type transitions. DELVE is available as an open-source python package: https://github.com/jranek/delve .

Cell cycle plasticity underlies fractional resistance to palbociclib in ER+/HER2- breast tumor cells.

The CDK4/6 inhibitor palbociclib blocks cell cycle progression in Estrogen receptor-positive, human epidermal growth factor 2 receptor-negative (ER+/HER2-) breast tumor cells. Despite the drug's success in improving patient outcomes, a small percentage of tumor cells continues to divide in the presence of palbociclib-a phenomenon we refer to as fractional resistance. It is critical to understand the cellular mechanisms underlying fractional resistance because the precise percentage of resistant cells in patient tissue is a strong predictor of clinical outcomes. Here, we hypothesize that fractional resistance arises from cell-to-cell differences in core cell cycle regulators that allow a subset of cells to escape CDK4/6 inhibitor therapy. We used multiplex, single-cell imaging to identify fractionally resistant cells in both cultured and primary breast tumor samples resected from patients. Resistant cells showed premature accumulation of multiple G1 regulators including E2F1, retinoblastoma protein, and CDK2, as well as enhanced sensitivity to pharmacological inhibition of CDK2 activity. Using trajectory inference approaches, we show how plasticity among cell cycle regulators gives rise to alternate cell cycle "paths" that allow individual tumor cells to escape palbociclib treatment. Understanding drivers of cell cycle plasticity, and how to eliminate resistant cell cycle paths, could lead to improved cancer therapies targeting fractionally resistant cells to improve patient outcomes.

Inheritance of OCT4 predetermines fate choice in human embryonic stem cells.

It is well known that clonal cells can make different fate decisions, but it is unclear whether these decisions are determined during, or before, a cell's own lifetime. Here, we engineered an endogenous fluorescent reporter for the pluripotency factor OCT4 to study the timing of differentiation decisions in human embryonic stem cells. By tracking single-cell OCT4 levels over multiple cell cycle generations, we found that the decision to differentiate is largely determined before the differentiation stimulus is presented and can be predicted by a cell's preexisting OCT4 signaling patterns. We further quantified how maternal OCT4 levels were transmitted to, and distributed between, daughter cells. As mother cells underwent division, newly established OCT4 levels in daughter cells rapidly became more predictive of final OCT4 expression status. These results imply that the choice between developmental cell fates can be largely predetermined at the time of cell birth through inheritance of a pluripotency factor.

The Cell Cycle Browser: An Interactive Tool for Visualizing, Simulating, and Perturbing Cell-Cycle Progression.

The cell cycle is driven by precise temporal coordination among many molecular activities. To understand and explore this process, we developed the Cell Cycle Browser (CCB), an interactive web interface based on real-time reporter data collected in proliferating human cells. This tool facilitates visualizing, organizing, simulating, and predicting the outcomes of perturbing cell-cycle parameters. Time-series traces from individual cells can be combined to build a multi-layered timeline of molecular activities. Users can simulate the cell cycle using computational models that capture the dynamics of molecular activities and phase transitions. By adjusting individual expression levels and strengths of molecular relationships, users can predict effects on the cell cycle. Virtual assays, such as growth curves and flow cytometry, provide familiar outputs to compare cell-cycle behaviors for data and simulations. The CCB serves to unify our understanding of cell-cycle dynamics and provides a platform for generating hypotheses through virtual experiments.

Apical targeting of the P2Y(4) receptor is directed by hydrophobic and basic residues in the cytoplasmic tail.

The P2Y(4) receptor is selectively targeted to the apical membrane in polarized epithelial cell lines and has been shown to play a key role in intestinal chloride secretion. In this study, we delimit a 23 amino acid sequence within the P2Y(4) receptor C-tail that directs its apical targeting. Using a mutagenesis approach, we found that four hydrophobic residues near the COOH-terminal end of the signal are necessary for apical sorting, whereas two basic residues near the NH(2)-terminal end of the signal are involved to a lesser extent. Interestingly, mutation of the key hydrophobic residues results in a basolateral enrichment of the receptor construct, suggesting that the apical targeting sequence may prevent insertion or disrupt stability of the receptor at the basolateral membrane. The signal is not sequence specific, as an inversion of the 23 amino acid sequence does not disrupt apical targeting. We also show that the apical targeting sequence is an autonomous signal and is capable of redistributing the normally basolateral P2Y(12) receptor, suggesting that the apical signal is dominant over the basolateral signal in the main body of the P2Y(12) receptor. The targeting sequence is unique to the P2Y(4) receptor, and sequence alignments of the COOH-terminal tail of mammalian orthologs reveal that the hydrophobic residues in the targeting signal are highly conserved. These data define the novel apical sorting signal of the P2Y(4) receptor, which may represent a common mechanism for trafficking of epithelial transmembrane proteins.

Charged residues in the C-terminus of the P2Y1 receptor constitute a basolateral-sorting signal.

The P2Y(1) receptor is localized to the basolateral membrane of polarized Madin-Darby canine kidney (MDCK) cells. In the present study, we identified a 25-residue region within the C-terminal tail (C-tail) of the P2Y(1) receptor that directs basolateral sorting. Deletion of this sorting signal caused redirection of the receptor to the apical membrane, indicating that the region from the N-terminus to transmembrane domain 7 (TM7) contains an apical-sorting signal that is overridden by a dominant basolateral signal in the C-tail. Location of the signal relative to TM7 is crucial, because increasing its distance from the end of TM7 resulted in loss of basolateral sorting. The basolateral-sorting signal does not use any previously established basolateral-sorting motifs, i.e. tyrosine-containing or di-hydrophobic motifs, for function, and it is functional even when inverted or when its amino acids are scrambled, indicating that the signal is sequence independent. Mutagenesis of different classes of amino acids within the signal identified charged residues (five basic and four acidic amino acids in 25 residues) as crucial determinants for sorting function, with amidated amino acids having a lesser role. Mutational analyses revealed that whereas charge balance (+1 overall) of the signal is unimportant, the total number of charged residues (nine), either positive or negative, is crucial for basolateral targeting. These data define a new class of targeting signal that relies on total charge and might provide a common mechanism for polarized trafficking of epithelial proteins.

Key determinants of nucleotide-activated G protein-coupled P2Y(2) receptor function revealed by chemical and pharmacological experiments, mutagenesis and homology modeling.

The P2Y(2) receptor, which is activated by UTP, ATP, and dinucleotides, was studied as a prototypical nucleotide-activated GPCR. A combination of receptor mutagenesis, determination of its effects on potency and efficacy of agonists and antagonists, homology modeling, and chemical experiments was applied. R272 (extracellular loop EL3) was found to play a gatekeeper role, presumably responsible for recognition and orientation of the nucleotides. R272 is also directly involved in binding of dinucleotides, which behaved as partial agonists. Y118A (3.37) mutation led to dramatically reduced efficacy of agonists; it is part of the entry channel as well as the triphosphate binding site. While the Y114A (3.33) mutation did not have any effect on agonist activities, the antagonist Reactive Blue 2 (6) was completely inactive at that mutant. The disulfide bridge Cys25-Cys278 was found to be important for agonist potency but neither for agonist efficacy nor for antagonist potency.

Asymmetrical distributions of muscarinic receptor binding in the hippocampus of female rats.

Asymmetries in muscarinic receptor binding were investigated in the hippocampus of female rats by in vitro autoradiography. Coronal sections from 18 brains were incubated with the muscarinic receptor antagonist [3H]quinuclidinyl benzilate, the muscarinic M1 receptor antagonist [3H]pirenzepine, or the muscarinic M2 receptor antagonist [3H]AF-DX 384. Binding of these radioligands was higher on the right than the left side of CA1, CA3, and dentate gyrus in almost every brain confirming hemispheric asymmetry at the neurochemical level. The ovarian hormone, estradiol, did not alter the asymmetry in muscarinic binding. Neurochemical asymmetries within hippocampal subfields may have implications for physiological and behavioral functions.

The apical targeting signal of the P2Y2 receptor is located in its first extracellular loop.

P2Y2 and P2Y4 receptors, which have 52% sequence identity, are both expressed at the apical membrane of Madin-Darby canine kidney cells, but the locations of their apical targeting signals are distinctly different. The targeting signal of the P2Y2 receptor is located between the N terminus and 7TM, whereas that of the P2Y4 receptor is present in its C-terminal tail. To identify the apical targeting signal in the P2Y2 receptor, regions of the P2Y2 receptor were progressively substituted with the corresponding regions of the P2Y4 receptor lacking its targeting signal. Characterization of these chimeras and subsequent mutational analysis revealed that four amino acids (Arg95, Gly96, Asp97, and Leu108) in the first extracellular loop play a major role in apical targeting of the P2Y2 receptor. Mutation of RGD to RGE had no effect on P2Y2 receptor targeting, indicating that receptor-integrin interactions are not involved in apical targeting. P2Y2 receptor mutants were localized in a similar manner in Caco-2 colon epithelial cells. This is the first identification of an extracellular protein-based targeting signal in a seven-transmembrane receptor.

Polarized expression of human P2Y receptors in epithelial cells from kidney, lung, and colon.

Eight human G protein-coupled P2Y receptors (P2Y(1), P2Y(2), P2Y(4), P2Y(6), P2Y(11), P2Y(12), P2Y(13), and P2Y(14)) that respond to extracellular nucleotides have been molecularly identified and characterized. P2Y receptors are widely expressed in epithelial cells and play an important role in regulating epithelial cell function. Functional studies assessing the capacity of various nucleotides to promote increases in short-circuit current (I(sc)) or Ca(2+) mobilization have suggested that some subtypes of P2Y receptors are polarized with respect to their functional activity, although these results often have been contradictory. To investigate the polarized expression of the family of P2Y receptors, we determined the localization of the entire P2Y family after expression in Madin-Darby canine kidney (MDCK) type II cells. Confocal microscopy of polarized monolayers revealed that P2Y(1), P2Y(11), P2Y(12), and P2Y(14) receptors reside at the basolateral membrane, P2Y(2), P2Y(4), and P2Y(6) receptors are expressed at the apical membrane, and the P2Y(13) receptor is unsorted. Biotinylation studies and I(sc) measurements in response to the appropriate agonists were consistent with the polarized expression observed in confocal microscopy. Expression of the G(q)-coupled P2Y receptors (P2Y(1), P2Y(2), P2Y(4), P2Y(6), and P2Y(11)) in lung and colonic epithelial cells (16HBE14o- and Caco-2 cells, respectively) revealed a targeting profile nearly identical to that observed in MDCK cells, suggesting that polarized targeting of these P2Y receptor subtypes is not a function of the type of epithelial cell in which they are expressed. These experiments highlight the highly polarized expression of P2Y receptors in epithelial cells.

P2Y receptors modulate ion channel function through interactions involving the C-terminal domain.

Nucleotide stimulation of G(q)-coupled P2Y receptors expressed in Xenopus laevis oocytes produces the activation of an endogenous voltage-gated ion channel, previously identified as the transient inward (T(in)) channel. Expression of human P2Y(1), human P2Y(2), rat P2Y(6), human P2Y(11), or skate P2Y receptors in oocytes resulted in modulation of the voltage dependence and inactivation gating of the channel. Expression of the human P2Y(4) receptor, rat M(1)-muscarinic receptor, and human B(1)-bradykinin receptor did not alter the properties of the T(in) channel. Replacement of the C-terminal domain of the human B(1)-bradykinin receptor with the C-terminal domains of either the human P2Y(1) or human P2Y(2) receptor resulted in voltage dependence and inactivation-gating properties, respectively, of the T(in) channel that were similar to those elicited by the respective native P2Y receptor. Systematic truncation of the C-terminal region of the human P2Y(1) receptor identified a short region responsible for modulation of the T(in) channel. This region contains a conserved sequence motif found in all P2Y receptors that modulates the voltage dependence of the T(in) channel. Synthetic 20-mer peptides from the C-terminal domains of human P2Y(1) and P2Y(2) receptors produced a shift in the voltage dependence and slowed inactivation gating, respectively, after injection into oocytes expressing human B(1)-bradykinin or truncated human P2Y(1) receptors. These results indicate that certain P2Y receptors are capable of modulating the voltage sensitivity and inactivation gating of an endogenous oocyte ion channel through interactions involving the C-terminal region of the receptor. Such modulation of ion channel function could also exist in native mammalian cells that express P2Y receptors.