Making good decisions in early drug discovery.

The earliest phases of drug discovery require diverse scientific disciplines to work hand in hand to address many unknowns. Good decision making is crucial for success in this context and, yet, the topic of sound planning has rarely been addressed for the earliest stages of drug discovery. We propose a tailored, qualitative 'decision quality' process that can serve as a guide toward generating project plans optimized to address a given project situation. Furthermore, we propose a visual flow-chart format for the selected plan that includes key decisions and activities, together forming a decision roadmap of the plan. We illustrate each step of the process by means of a real-life example and provide recommendations for its implementation.

Details Matter in Structure-based Drug Design.

Successful structure-based drug design (SBDD) requires the optimization of interactions with the target protein and the minimization of ligand strain. Both factors are often modulated by small changes in the chemical structure which can lead to profound changes in the preferred conformation and interaction preferences of the ligand. We draw from examples of a Roche project targeting phosphodiesterase 10 to highlight that details matter in SBDD. Data mining in crystal structure databases can help to identify these sometimes subtle effects, but it is also a great resource to learn about molecular recognition in general and can be used as part of molecular design tools. We illustrate the use of the Cambridge Structural Database for identifying preferred structural motifs for intramolecular hydrogen bonding and of the Protein Data Bank for deriving propensities for protein-ligand interactions.

Gut-derived bacterial flagellin induces beta-cell inflammation and dysfunction.

Hyperglycemia and type 2 diabetes (T2D) are caused by failure of pancreatic beta cells. The role of the gut microbiota in T2D has been studied, but causal links remain enigmatic. Obese individuals with or without T2D were included from two independent Dutch cohorts. Human data were translated in vitro and in vivo by using pancreatic islets from C57BL6/J mice and by injecting flagellin into obese mice. Flagellin is part of the bacterial locomotor appendage flagellum, present in gut bacteria including Enterobacteriaceae, which we show to be more abundant in the gut of individuals with T2D. Subsequently, flagellin induces a pro-inflammatory response in pancreatic islets mediated by the Toll-like receptor (TLR)-5 expressed on resident islet macrophages. This inflammatory response is associated with beta-cell dysfunction, characterized by reduced insulin gene expression, impaired proinsulin processing and stress-induced insulin hypersecretion in vitro and in vivo in mice. We postulate that increased systemically disseminated flagellin in T2D is a contributing factor to beta-cell failure in time and represents a novel therapeutic target.

The Torsion Library: Semiautomated Improvement of Torsion Rules with SMARTScompare.

The Torsion Library is a collection of torsion motifs associated with angle distributions, derived from crystallographic databases. It is used in strain assessment, conformer generation, and geometry optimization. A hierarchical structure of expert curated SMARTS defines the chemical environments of rotatable bonds and associates these with preferred angles. SMARTS can be very complex and full of implications, which make them difficult to maintain manually. Recent developments in automatically comparing SMARTS patterns can be applied to the Torsion Library to ensure its correctness. We specifically discuss the implementation and the limits of such a procedure in the context of torsion motifs and show several examples of how the Torsion Library benefits from this. All automated changes are validated manually and then shown to have an effect on the angle distributions by correcting matching behavior. The corrected Torsion Library itself is available including both PDB as well as CSD histograms in the Supporting Information and can be used to evaluate rotatable bonds at https://torsions.zbh.uni-hamburg.de.

What We Learned in 25 Years of Interactive Molecular Design Sessions.

We retrace Prof. François Diederich's consultancy work for Roche and its impact over the years he worked with us. François Diederich uniquely shaped our approach to molecular design, and interactions with him and his research group at ETH Zurich have created deep insights into molecular recognition. Herein we share how his style and approach continue to inspire us.

Fasting increases microbiome-based colonization resistance and reduces host inflammatory responses during an enteric bacterial infection.

Reducing food intake is a common host response to infection, yet it remains unclear whether fasting is detrimental or beneficial to an infected host. Despite the gastrointestinal tract being the primary site of nutrient uptake and a common route for infection, studies have yet to examine how fasting alters the host's response to an enteric infection. To test this, mice were fasted before and during oral infection with the invasive bacterium Salmonella enterica serovar Typhimurium. Fasting dramatically interrupted infection and subsequent gastroenteritis by suppressing Salmonella's SPI-1 virulence program, preventing invasion of the gut epithelium. Virulence suppression depended on the gut microbiota, as Salmonella's invasion of the epithelium proceeded in fasting gnotobiotic mice. Despite Salmonella's restored virulence within the intestines of gnotobiotic mice, fasting downregulated pro-inflammatory signaling, greatly reducing intestinal pathology. Our study highlights how food intake controls the complex relationship between host, pathogen and gut microbiota during an enteric infection.

Interleukin-37 regulates innate immune signaling in human and mouse colonic organoids.

Intestinal epithelial cells (IEC) reside in close proximity to the gut microbiota and are hypo-responsive to bacterial products, likely to prevent maladaptive inflammatory responses. This is in part due to their strong expression of Single Ig IL-1 related receptor (SIGIRR), a negative regulator of interleukin (IL)-1 and toll-like receptor signaling. IL-37 is an anti-inflammatory cytokine that inhibits innate signaling in diverse cells by signaling through SIGIRR. Despite the strong expression of SIGIRR by IEC, few studies have examined whether IL-37 can suppress their innate immune signaling. We characterized innate immune responses of human and murine colonoids to bacteria (FliC, LPS) and host (IL-1ß) products and the role of IL-37/SIGIRR in regulating these responses. We demonstrated that human colonoids responded only to FliC, but not to LPS or IL-1ß. While colonoids derived from different donors displayed significant inter-individual variability in the magnitude of their innate responses to FliC stimulation, all colonoids released a variety of chemokines. Interestingly, IL-37 attenuated these responses through inhibition of p38 and NFκB signaling pathways. We determined that this suppression by IL-37 was SIGIRR dependent, in murine organoids. Along with species-specific differences in IEC innate responses, we show that IL-37 can promote IEC hypo-responsiveness by suppressing inflammatory signaling.

Culture Methods to Study Apical-Specific Interactions using Intestinal Organoid Models.

The lining of the gut epithelium is made up of a simple layer of specialized epithelial cells that expose their apical side to the lumen and respond to external cues. Recent optimization of in vitro culture conditions allows for the re-creation of the intestinal stem cell niche and the development of advanced 3-dimensional (3D) culture systems that recapitulate the cell composition and the organization of the epithelium. Intestinal organoids embedded in an extracellular matrix (ECM) can be maintained for long-term and self-organize to generate a well-defined, polarized epithelium that encompasses an internal lumen and an external exposed basal side. This restrictive nature of the intestinal organoids presents challenges in accessing the apical surface of the epithelium in vitro and limits the investigation of biological mechanisms such as nutrient uptake and host-microbiota/host-pathogen interactions. Here, we describe two methods that facilitate access to the apical side of the organoid epithelium and support the differentiation of specific intestinal cell types. First, we show how ECM removal induces an inversion of the epithelial cell polarity and allows for the generation of apical-out 3D organoids. Second, we describe how to generate 2-dimensional (2D) monolayers from single cell suspensions derived from intestinal organoids, comprised of mature and differentiated cell types. These techniques provide novel tools to study apical-specific interactions of the epithelium with external cues in vitro and promote the use of organoids as a platform to facilitate precision medicine.

Intestinal restriction of Salmonella Typhimurium requires caspase-1 and caspase-11 epithelial intrinsic inflammasomes.

We investigated the role of the inflammasome effector caspases-1 and -11 during Salmonella enterica serovar Typhimurium infection of murine intestinal epithelial cells (IECs). Salmonella burdens were significantly greater in the intestines of caspase-1/11 deficient (Casp1/11-/-), Casp1-/- and Casp11-/- mice, as compared to wildtype mice. To determine if this reflected IEC-intrinsic inflammasomes, enteroid monolayers were derived and infected with Salmonella. Casp11-/- and wildtype monolayers responded similarly, whereas Casp1-/- and Casp1/11-/- monolayers carried significantly increased intracellular burdens, concomitant with marked decreases in IEC shedding and death. Pretreatment with IFN-γ to mimic inflammation increased caspase-11 levels and IEC death, and reduced Salmonella burdens in Casp1-/- monolayers, while high intracellular burdens and limited cell shedding persisted in Casp1/11-/- monolayers. Thus caspase-1 regulates inflammasome responses in IECs at baseline, while proinflammatory activation of IECs reveals a compensatory role for caspase-11. These results demonstrate the importance of IEC-intrinsic canonical and non-canonical inflammasomes in host defense against Salmonella.

Inhibition of SARS-CoV-2 Infections in Engineered Human Tissues Using Clinical-Grade Soluble Human ACE2.

We have previously provided the first genetic evidence that angiotensin converting enzyme 2 (ACE2) is the critical receptor for severe acute respiratory syndrome coronavirus (SARS-CoV), and ACE2 protects the lung from injury, providing a molecular explanation for the severe lung failure and death due to SARS-CoV infections. ACE2 has now also been identified as a key receptor for SARS-CoV-2 infections, and it has been proposed that inhibiting this interaction might be used in treating patients with COVID-19. However, it is not known whether human recombinant soluble ACE2 (hrsACE2) blocks growth of SARS-CoV-2. Here, we show that clinical grade hrsACE2 reduced SARS-CoV-2 recovery from Vero cells by a factor of 1,000-5,000. An equivalent mouse rsACE2 had no effect. We also show that SARS-CoV-2 can directly infect engineered human blood vessel organoids and human kidney organoids, which can be inhibited by hrsACE2. These data demonstrate that hrsACE2 can significantly block early stages of SARS-CoV-2 infections.

The gastrointestinal pathogen Campylobacter jejuni metabolizes sugars with potential help from commensal Bacteroides vulgatus.

Although the gastrointestinal pathogen Campylobacter jejuni was considered asaccharolytic, >50% of sequenced isolates possess an operon for L-fucose utilization. In C. jejuni NCTC11168, this pathway confers L-fucose chemotaxis and competitive colonization advantages in the piglet diarrhea model, but the catabolic steps remain unknown. Here we solved the putative dehydrogenase structure, resembling FabG of Burkholderia multivorans. The C. jejuni enzyme, FucX, reduces L-fucose and D-arabinose in vitro and both sugars are catabolized by fuc-operon encoded enzymes. This enzyme alone confers chemotaxis to both sugars in a non-carbohydrate-utilizing C. jejuni strain. Although C. jejuni lacks fucosidases, the organism exhibits enhanced growth in vitro when co-cultured with Bacteroides vulgatus, suggesting scavenging may occur. Yet, when excess amino acids are available, C. jejuni prefers them to carbohydrates, indicating a metabolic hierarchy exists. Overall this study increases understanding of nutrient metabolism by this pathogen, and identifies interactions with other gut microbes.

Enteroids Derived From Inflammatory Bowel Disease Patients Display Dysregulated Endoplasmic Reticulum Stress Pathways, Leading to Differential Inflammatory Responses and Dendritic Cell Maturation.

BACKGROUND AND AIMS: Endoplasmic reticulum [ER] stress in intestinal epithelial cells [IECs] contributes to the pathogenesis of inflammatory bowel disease [IBD]. We hypothesized that ER stress changes innate signalling in human IECs, augmenting toll-like receptor [TLR] responses and inducing pro-inflammatory changes in underlying dendritic cells [DCs]. METHODS: Caco-2 cells and primary human colon-derived enteroid monolayers were exposed to ATP [control stressor] or thapsigargin [Tg] [ER stress inducer], and were stimulated with the TLR5 agonist flagellin. Cytokine release was measured by an enzyme immunoassay. ER stress markers CHOP, GRP78 and XBP1s/u were measured via quantitative PCR and Western blot. Monocyte-derived DCs [moDCs] were cultured with the IEC supernatants and their activation state was measured. Responses from enteroids derived from IBD patients and healthy control participants were compared. RESULTS: ER stress enhanced flagellin-induced IL-8 release from Caco-2 cells and enteroids. Moreover, conditioned media activated DCs to become pro-inflammatory, with increased expression of CD80, CD86, MHCII, IL-6, IL-15 and IL-12p70 and decreased expression of CD103 and IL-10. Flagellin-induced IL-8 production correlated with DC activation, suggesting a common stress pathway. Moreover, there were distinct differences in cytokine expression and basal ER stress between IBD and healthy subject-derived enteroid monolayers, suggesting a dysregulated ER stress pathway in IBD-derived enteroids. CONCLUSIONS: Cellular stress enhances TLR5 responses in IECs, leading to increased DC activation, indicating a previously unknown mechanistic link between epithelial ER stress and immune activation in IBD. Furthermore, dysregulated ER stress may be propagated from the intestinal epithelial stem cell niche in IBD patients.

Suppressive and Gut-Reparative Functions of Human Type 1 T Regulatory Cells.

BACKGROUND & AIMS: T-regulatory (Treg) cells suppress the immune response to maintain homeostasis. There are 2 main subsets of Treg cells: FOXP3 (forkhead box protein 3)-positive Treg cells, which do not produce high levels of effector cytokines, and type 1 Treg (Tr1) cells, which are FOXP3-negative and secrete interleukin (IL) 10. IL10 is an anti-inflammatory cytokine, so Tr1 cells might be used in the treatment of inflammatory bowel diseases. We aimed to develop methods to isolate and expand human Tr1 cells and define their functions. METHODS: We obtained blood and colon biopsy samples from patients with Crohn's disease or ulcerative colitis or healthy individuals (controls). CD4+ T cells were isolated from blood samples and stimulated with anti-CD3 and anti-CD28 beads, and Tr1 cells were purified by using an IL10 cytokine-capture assay and cell sorting. FOXP3-positive Treg cells were sorted as CD4+CD25highCD127low cells from unstimulated cells. Tr1 and FOXP3-positive Treg cells were expanded, and phenotypes and gene expression profiles were compared. T cells in peripheral blood mononuclear cells from healthy donors were stimulated with anti-CD3 and anti-CD28 beads, and the suppressive abilities of Tr1 and FOXP3-positive Treg cells were measured. Human colon organoid cultures were established, cultured with supernatants from Tr1 or FOXP3-positive cells, and analyzed by immunofluorescence and flow cytometry. T84 cells (human colon adenocarcinoma epithelial cells) were incubated with supernatants from Tr1 or FOXP3-positive cells, and transepithelial electrical resistance was measured to determine epithelial cell barrier function. RESULTS: Phenotypes of Tr1 cells isolated from control individuals vs patients with Crohn's disease or ulcerative colitis did not differ significantly after expansion. Tr1 cells and FOXP3-positive Treg cells suppressed proliferation of effector T cells, but only Tr1 cells suppressed secretion of IL1B and tumor necrosis factor from myeloid cells. Tr1 cells, but not FOXP3-positive Treg cells, isolated from healthy individuals and patients with Crohn's disease or ulcerative colitis secreted IL22, which promoted barrier function of human intestinal epithelial cells. Tr1 cell culture supernatants promoted differentiation of mucin-producing goblet cells in intestinal organoid cultures. CONCLUSIONS: Human Tr1 cells suppress proliferation of effector T cells (adaptive immune response) and production of IL1B and TNF by myeloid cells (inmate immune response). They also secrete IL22 to promote barrier function. They might be developed as a cell-based therapy for intestinal inflammatory disorders.

Bacterial AB5 toxins inhibit the growth of gut bacteria by targeting ganglioside-like glycoconjugates.

The AB5 toxins cholera toxin (CT) from Vibrio cholerae and heat-labile enterotoxin (LT) from enterotoxigenic Escherichia coli are notorious for their roles in diarrheal disease, but their effect on other intestinal bacteria remains unexplored. Another foodborne pathogen, Campylobacter jejuni, can mimic the GM1 ganglioside receptor of CT and LT. Here we demonstrate that the toxin B-subunits (CTB and LTB) inhibit C. jejuni growth by binding to GM1-mimicking lipooligosaccharides and increasing permeability of the cell membrane. Furthermore, incubation of CTB or LTB with a C. jejuni isolate capable of altering its lipooligosaccharide structure selects for variants lacking the GM1 mimic. Examining the chicken GI tract with immunofluorescence microscopy demonstrates that GM1 reactive structures are abundant on epithelial cells and commensal bacteria, further emphasizing the relevance of this mimicry. Exposure of chickens to CTB or LTB causes shifts in the gut microbial composition, providing evidence for new toxin functions in bacterial gut competition.

The Muc2 mucin coats murine Paneth cell granules and facilitates their content release and dispersion.

Paneth cells are a key subset of secretory epithelial cells found at the base of small intestinal crypts. Unlike intestinal goblet cells, which secrete the mucin Muc2, Paneth cells are best known for producing an array of antimicrobial factors. We unexpectedly identified Muc2 staining localized around Paneth cell granules. Electron microscopy (EM) confirmed an electron lucent halo around these granules, which was lost in Paneth cells from Muc2-deficient (-/-) mice. EM and immunostaining for lysozyme revealed that Muc2-/- Paneth cells contained larger, more densely packed granules within their cytoplasm, and we detected defects in the transcription of key antimicrobial genes in the ileal tissues of Muc2-/- mice. Enteroids derived from the small intestine of wild-type and Muc2-/- mice revealed phenotypic differences in Paneth cells similar to those seen in vivo. Moreover, lysozyme-containing granule release from Muc2-/- enteroid Paneth cells was shown to be impaired. Surprisingly, Paneth cells within human ileal and duodenal tissues were found to be Muc2 negative. Thus Muc2 plays an important role in murine Paneth cells, suggesting links in function with goblet cells; however human Paneth cells lack Muc2, highlighting that caution should be applied when linking murine to human Paneth cell functions. NEW & NOTEWORTHY We demonstrate for the first time that murine Paneth cell granules possess a halo comprised of the mucin Muc2. The presence of Muc2 exerts an impact on Paneth cell granule size and number and facilitates the release and dispersal of antimicrobials into the mucus layer. Interestingly, despite the importance of Muc2 in murine Paneth cell function, our analysis of Muc2 in human intestinal tissues revealed no trace of Muc2 expression by human Paneth cells.

Learning Medicinal Chemistry Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADMET) Rules from Cross-Company Matched Molecular Pairs Analysis (MMPA).

The first large scale analysis of in vitro absorption, distribution, metabolism, excretion, and toxicity (ADMET) data shared across multiple major pharma has been performed. Using advanced matched molecular pair analysis (MMPA), we combined data from three pharmaceutical companies and generated ADMET rules, avoiding the need to disclose the full chemical structures. On top of the very large exchange of knowledge, all companies involved synergistically gained approximately 20% more rules from the shared transformations. There is good quantitative agreement between the rules based on shared data compared to both individual companies' rules and rules published in the literature. Known correlations between log  D, solubility, in vitro clearance, and plasma protein binding also hold in transformation space, but there are also interesting exceptions. Data pools such as this allow focusing on particular functional groups and characterizing their ADMET profile. Finally the role of a corpus of robustly tested medicinal chemistry knowledge in the training of medicinal chemistry is discussed.

Frontline defenders: goblet cell mediators dictate host-microbe interactions in the intestinal tract during health and disease.

Goblet cells (GCs) are the predominant secretory epithelial cells lining the luminal surface of the mammalian gastrointestinal (GI) tract. Best known for their apical release of mucin 2 (Muc2), which is critical for the formation of the intestinal mucus barrier, GCs have often been overlooked for their active contributions to intestinal protection and host defense. In part, this oversight reflects the limited tools available to study their function but also because GCs have long been viewed as relatively passive players in promoting intestinal homeostasis and host defense. In light of recent studies, this perspective has shifted, as current evidence suggests that Muc2 as well as other GC mediators are actively released into the lumen to defend the host when the GI tract is challenged by noxious stimuli. The ability of GCs to sense and respond to danger signals, such as bacterial pathogens, has recently been linked to inflammasome signaling, potentially intrinsic to the GCs themselves. Moreover, further work suggests that GCs release Muc2, as well as other mediators, to modulate the composition of the gut microbiome, leading to both the expansion as well as the depletion of specific gut microbes. This review will focus on the mechanisms by which GCs actively defend the host from noxious stimuli, as well as describe advanced technologies and new approaches by which their responses can be addressed. Taken together, we will highlight current insights into this understudied, yet critical, aspect of intestinal mucosal protection and its role in promoting gut defense and homeostasis.

A simple, cost-effective method for generating murine colonic 3D enteroids and 2D monolayers for studies of primary epithelial cell function.

Cancer cell lines have been the mainstay of intestinal epithelial experimentation for decades, due primarily to their immortality and ease of culture. However, because of the inherent biological abnormalities of cancer cell lines, many cellular biologists are currently transitioning away from these models and toward more representative primary cells. This has been particularly challenging, but recent advances in the generation of intestinal organoids have brought the routine use of primary cells within reach of most epithelial biologists. Nevertheless, even with the proliferation of publications that use primary intestinal epithelial cells, there is still a considerable amount of trial and error required for laboratories to establish a consistent and reliable method to culture three-dimensional (3D) intestinal organoids and primary epithelial monolayers. We aim to minimize the time other laboratories spend troubleshooting the technique and present a standard method for culturing primary epithelial cells. Therefore, we have described our optimized, high-yield, cost-effective protocol to grow 3D murine colonoids for more than 20 passages and our detailed methods to culture these cells as confluent monolayers for at least 14 days, enabling a wide variety of potential future experiments. By supporting and expanding on the current literature of primary epithelial culture optimization and detailed use in experiments, we hope to help enable the widespread adoption of these innovative methods and allow consistency of results obtained across laboratories and institutions.NEW & NOTEWORTHY Primary intestinal epithelial monolayers are notoriously difficult to maintain culture, even with the recent advances in the field. We describe, in detail, the protocols required to maintain three-dimensional cultures of murine colonoids and passage these primary epithelial cells to confluent monolayers in a standardized, high-yield and cost-effective manner.

Identification of New ATG4B Inhibitors Based on a Novel High-Throughput Screening Platform.

Autophagy is an evolutionarily conserved homeostasis process through which aggregated proteins or damaged organelles are enveloped in a double-membrane structure called an autophagosome and then digested in a lysosome-dependent manner. Growing evidence suggests that malfunction of autophagy contributes to the pathogenesis of a variety of diseases, including cancer, viral infection, and neurodegeneration. However, autophagy is a complicated process, and understanding of the relevance of autophagy to disease is limited by lack of specific and potent autophagy modulators. ATG4B, a Cys-protease that cleaves ATG8 family proteins, such as LC3B, is a key protein in autophagosome formation and maturation process. A novel time-resolved fluorescence resonance energy transfer (TR-FRET) assay measuring protease activity of ATG4B was developed, validated, and adapted into a high-throughput screening (HTS) format. HTS was then conducted with a Roche focus library of 57,000 compounds. After hit confirmation and a counterscreen to filter out fluorescence interference compounds, 267 hits were confirmed, constituting a hit rate of 0.49%. Furthermore, among 65 hits with an IC50 < 50 µM, one compound mimics the LC3 peptide substrate (-TFG-). Chemistry modification based on this particular hit gave preliminary structure activity relationship (SAR) resulting in a compound with a 10-fold increase in potency. This compound forms a stable covalent bond with Cys74 of ATG4B in a 1:1 ratio as demonstrated by liquid chromatography/tandem mass spectrometry (LC/MS/MS). Furthermore, this compound displayed cellular ATG4B inhibition activity. Overall, the novel TR-FRET ATG4B protease assay plus counterscreen assay provides a robust platform to identify ATG4B inhibitors, which would help to elucidate the mechanism of the autophagy pathway and offer opportunities for drug discovery.

Tricellular Tight Junction Protein Tricellulin Is Targeted by the Enteropathogenic Escherichia coli Effector EspG1, Leading to Epithelial Barrier Disruption.

Enteropathogenic Escherichia coli (EPEC)-induced diarrhea is often associated with disruption of intestinal epithelial tight junctions. Although studies have shown alterations in the expression and localization of bicellular tight junction proteins during EPEC infections, little is known about whether tricellular tight junction proteins (tTJs) are affected. Using Caco-2 cell monolayers, we investigated if EPEC is capable of targeting the tTJ protein tricellulin. Our results demonstrated that at 4 h postinfection, EPEC induced a significant reduction in tricellulin levels, accompanied by a significant loss of transepithelial resistance (TEER) and a corresponding increase in paracellular permeability. Conversely, cells overexpressing tricellulin were highly resistant to EPEC-induced barrier disruption. Confocal microscopy revealed the distribution of tricellulin into the plasma membrane of infected epithelial cells and confirmed the localization of EPEC aggregates in close proximity to tTJs. Moreover, infections with EPEC strains lacking genes encoding specific type III secreted effector proteins demonstrated a crucial role for the effector EspG1 in modulating tricellulin expression. Complementation studies suggest that the EspG-induced depletion of tricellulin is microtubule dependent. Overall, our results show that EPEC-induced epithelial barrier dysfunction is mediated in part by EspG1-induced microtubule-dependent depletion of tricellulin.