Deciphering cell-cell interactions and communication from gene expression.

Cell-cell interactions orchestrate organismal development, homeostasis and single-cell functions. When cells do not properly interact or improperly decode molecular messages, disease ensues. Thus, the identification and quantification of intercellular signalling pathways has become a common analysis performed across diverse disciplines. The expansion of protein-protein interaction databases and recent advances in RNA sequencing technologies have enabled routine analyses of intercellular signalling from gene expression measurements of bulk and single-cell data sets. In particular, ligand-receptor pairs can be used to infer intercellular communication from the coordinated expression of their cognate genes. In this Review, we highlight discoveries enabled by analyses of cell-cell interactions from transcriptomic data and review the methods and tools used in this context.

Gαs is dispensable for ß-arrestin coupling but dictates GRK selectivity and is predominant for gene expression regulation by ß2-adrenergic receptor.

ß-arrestins play a key role in G protein-coupled receptor (GPCR) internalization, trafficking, and signaling. Whether ß-arrestins act independently of G protein-mediated signaling has not been fully elucidated. Studies using genome-editing approaches revealed that whereas G proteins are essential for mitogen-activated protein kinase activation by GPCRs., ß-arrestins play a more prominent role in signal compartmentalization. However, in the absence of G proteins, GPCRs may not activate ß-arrestins, thereby limiting the ability to distinguish G protein from ß-arrestin-mediated signaling events. We used ß2-adrenergic receptor (ß2AR) and its ß2AR-C tail mutant expressed in human embryonic kidney 293 cells wildtype or CRISPR-Cas9 gene edited for Gαs, ß-arrestin1/2, or GPCR kinases 2/3/5/6 in combination with arrestin conformational sensors to elucidate the interplay between Gαs and ß-arrestins in controlling gene expression. We found that Gαs is not required for ß2AR and ß-arrestin conformational changes, ß-arrestin recruitment, and receptor internalization, but that Gαs dictates the GPCR kinase isoforms involved in ß-arrestin recruitment. By RNA-Seq analysis, we found that protein kinase A and mitogen-activated protein kinase gene signatures were activated by stimulation of ß2AR in wildtype and ß-arrestin1/2-KO cells but absent in Gαs-KO cells. These results were validated by re-expressing Gαs in the corresponding KO cells and silencing ß-arrestins in wildtype cells. These findings were extended to cellular systems expressing endogenous levels of ß2AR. Overall, our results support that Gs is essential for ß2AR-promoted protein kinase A and mitogen-activated protein kinase gene expression signatures, whereas ß-arrestins initiate signaling events modulating Gαs-driven nuclear transcriptional activity.

Avant-garde: an automated data-driven DIA data curation tool.

Several challenges remain in data-independent acquisition (DIA) data analysis, such as to confidently identify peptides, define integration boundaries, remove interferences, and control false discovery rates. In practice, a visual inspection of the signals is still required, which is impractical with large datasets. We present Avant-garde as a tool to refine DIA (and parallel reaction monitoring) data. Avant-garde uses a novel data-driven scoring strategy: signals are refined by learning from the dataset itself, using all measurements in all samples to achieve the best optimization. We evaluate the performance of Avant-garde using benchmark DIA datasets and show that it can determine the quantitative suitability of a peptide peak, and reach the same levels of selectivity, accuracy, and reproducibility as manual validation. Avant-garde is complementary to existing DIA analysis engines and aims to establish a strong foundation for subsequent analysis of quantitative mass spectrometry data.

Activity of the PI3K-δ,γ inhibitor duvelisib in a phase 1 trial and preclinical models of T-cell lymphoma.

Duvelisib (IPI-145) is an oral inhibitor of phosphatidylinositol 3-kinase (PI3K)-δ/γ isoforms currently in clinical development. PI3K-δ/γ inhibition may directly inhibit malignant T-cell growth, making duvelisib a promising candidate for patients with peripheral (PTCL) or cutaneous (CTCL) T-cell lymphoma. Inhibition of either isoform may also contribute to clinical responses by modulating nonmalignant immune cells. We investigated these dual effects in a TCL cohort from a phase 1, open-label study of duvelisib in patients with relapsed or refractory PTCL (n = 16) and CTCL (n = 19), along with in vitro and in vivo models of TCL. The overall response rates in patients with PTCL and CTCL were 50.0% and 31.6%, respectively (P = .32). There were 3 complete responses, all among patients with PTCL. Activity was seen across a wide spectrum of subtypes. The most frequently observed grade 3 and 4 adverse events were transaminase increases (40% alanine aminotransferase, 17% aspartate aminotransferase), maculopapular rash (17%), and neutropenia (17%). Responders and nonresponders had markedly different changes in serum cytokine profiles induced by duvelisib. In vitro, duvelisib potently killed 3 of 4 TCL lines with constitutive phospho-AKT (pAKT) vs 0 of 7 lines lacking pAKT (P = .024) and exceeded cell killing by the PI3K-δ-specific inhibitor idelalisib. Administration of duvelisib to mice engrafted with a PTCL patient-derived xenograft resulted in a shift among tumor-associated macrophages from the immunosuppressive M2-like phenotype to the inflammatory M1-like phenotype. In summary, duvelisib demonstrated promising clinical activity and an acceptable safety profile in relapsed/refractory TCL, as well as preclinical evidence of both tumor cell-autonomous and immune-mediated effects. This trial was registered at www.clinicaltrials.gov as #NCT01476657.

Reduced-representation Phosphosignatures Measured by Quantitative Targeted MS Capture Cellular States and Enable Large-scale Comparison of Drug-induced Phenotypes.

Profiling post-translational modifications represents an alternative dimension to gene expression data in characterizing cellular processes. Many cellular responses to drugs are mediated by changes in cellular phosphosignaling. We sought to develop a common platform on which phosphosignaling responses could be profiled across thousands of samples, and created a targeted MS assay that profiles a reduced-representation set of phosphopeptides that we show to be strong indicators of responses to chemical perturbagens.To develop the assay, we investigated the coordinate regulation of phosphosites in samples derived from three cell lines treated with 26 different bioactive small molecules. Phosphopeptide analytes were selected from these discovery studies by clustering and picking 1 to 2 proxy members from each cluster. A quantitative, targeted parallel reaction monitoring assay was developed to directly measure 96 reduced-representation probes. Sample processing for proteolytic digestion, protein quantification, peptide desalting, and phosphopeptide enrichment have been fully automated, making possible the simultaneous processing of 96 samples in only 3 days, with a plate phosphopeptide enrichment variance of 12%. This highly reproducible process allowed ∼95% of the reduced-representation phosphopeptide probes to be detected in ∼200 samples.The performance of the assay was evaluated by measuring the probes in new samples generated under treatment conditions from discovery experiments, recapitulating the observations of deeper experiments using a fraction of the analytical effort. We measured these probes in new experiments varying the treatments, cell types, and timepoints to demonstrate generalizability. We demonstrated that the assay is sensitive to disruptions in common signaling pathways (e.g. MAPK, PI3K/mTOR, and CDK). The high-throughput, reduced-representation phosphoproteomics assay provides a platform for the comparison of perturbations across a range of biological conditions, suitable for profiling thousands of samples. We believe the assay will prove highly useful for classification of known and novel drug and genetic mechanisms through comparison of phosphoproteomic signatures.

Tumor microenvironmental determinants of high-risk DCIS progression.

Ductal carcinoma in situ (DCIS) constitutes an array of morphologically recognized intraductal neoplasms in the mammary ductal tree defined by an increased risk for subsequent invasive carcinomas at or near the site of biopsy detection. However, only 15-45% of untreated DCIS cases progress to invasive cancer, so understanding mechanisms that prevent progression is key to avoid overtreatment and provides a basis for alternative therapies and prevention. This study was designed to characterize the tumor microenvironment and molecular profile of high-risk DCIS that grew to a large size but remained as DCIS. All patients had DCIS lesions >5cm in size with at least one additional high-risk feature: young age (<45 years), high nuclear grade, hormone receptor negativity, HER2 positivity, the presence of comedonecrosis, or a palpable mass. The tumor immune microenvironment was characterized using multiplex immunofluorescence to identify immune cells and their spatial relationships within the ducts and stroma. Gene copy number analysis and whole exome DNA sequencing identified the mutational burden and driver mutations, and quantitative whole-transcriptome/gene expression analyses were performed. There was no association between the percent of the DCIS genome characterized by copy number variants (CNAs) and recurrence events (DCIS or invasive). Mutations, especially missense mutations, in the breast cancer driver genes PIK3CA and TP53 were common in this high-risk DCIS cohort (47% of evaluated lesions). Tumor infiltrating lymphocyte (TIL) density was higher in DCIS lesions with TP53 mutations (p=0.0079) compared to wildtype lesions, but not in lesions with PIK3CA mutations (p=0.44). Immune infiltrates were negatively associated with hormone receptor status and positively associated with HER2 expression. High levels of CD3+CD8- T cells were associated with good outcomes with respect to any subsequent recurrence (DCIS or invasive cancer), whereas high levels of CD3+Foxp3+ Treg cells were associated with poor outcomes. Spatial proximity analyses of immune cells and tumor cells demonstrated that close proximity of T cells with tumor cells was associated with good outcomes with respect to any recurrence as well as invasive recurrences. Interestingly, we found that myoepithelial continuity (distance between myoepithelial cells surrounding the involved ducts) was significantly lower in DCIS lesions compared to normal tissue (p=0.0002) or to atypical ductal hyperplasia (p=0.011). Gene set enrichment analysis identified several immune pathways associated with low myoepithelial continuity and a low myoepithelial continuity score was associated with better outcomes, suggesting that gaps in the myoepithelial layer may allow access/interactions between immune infiltrates and tumor cells. Our study demonstrates the immune microenvironment of DCIS, in particular the spatial proximity of tumor cells and T cells, and myoepithelial continuity are important determinants for progression of disease.

YAP-Driven Malignant Reprogramming of Epithelial Stem Cells at Single Cell Resolution.

Tumor initiation represents the first step in tumorigenesis during which normal progenitor cells undergo cell fate transition to cancer. Capturing this process as it occurs in vivo, however, remains elusive. Here we employ cell tracing approaches with spatiotemporally controlled oncogene activation and tumor suppressor inhibition to unveil the processes underlying oral epithelial progenitor cell reprogramming into cancer stem cells (CSCs) at single cell resolution. This revealed the rapid emergence of a distinct stem-like cell state, defined by aberrant proliferative, hypoxic, squamous differentiation, and partial epithelial to mesenchymal (pEMT) invasive gene programs. Interestingly, CSCs harbor limited cell autonomous invasive capacity, but instead recruit myeloid cells to remodel the basement membrane and ultimately initiate tumor invasion. CSC transcriptional programs are conserved in human carcinomas and associated with poor patient survival. These findings illuminate the process of cancer initiation at single cell resolution, thus identifying candidate targets for early cancer detection and prevention.

Direct reprogramming of oral epithelial progenitor cells to cancer stem cells at single cell resolution in vivo.

Tumor initiation represents the initial step in tumorigenesis during which normal progenitor cells undergo cell fate transition to cancer. Most studies investigating cancer-driving mechanisms in solid tumors rely on analyses of established malignant lesions, and thus cannot directly capture processes underlying the reprogramming of normal progenitor cells into cancer cells. Here, using spatiotemporally controlled oncogene expression in a genetically engineered system we demonstrate that concomitant YAP activation and HPV E6-E7 -mediated inhibition of tumor suppressive pathways is sufficient to rapidly reprogram oral epithelial progenitor cells (OEPCs) into cancer stem cells (CSCs). Single cell analyses of these nascent CSCs revealed hallmark transcriptional programs driving tumor initiation. Importantly, these CSC-enriched expression signatures distinguish normal tissue from malignant head and neck tumors and are associated with poor patient survival. Elucidating mechanisms underlying OEPC to CSC reprogramming may offer new insights to halt the conversion of premalignant cells into invasive carcinoma. HIGHLIGHTS: YAP and HPV E6-E7 reprogram oral epithelial progenitor cells into cancer stem cells. Single cell analyses reveal the transcriptional architecture of tumor initiation.CSC transcriptional programs distinguish normal tissue from carcinoma.CSC signatures are associated with poor head and neck cancer survival.

Integrated Molecular Characterization of Intraductal Papillary Mucinous Neoplasms: An NCI Cancer Moonshot Precancer Atlas Pilot Project.

Intraductal papillary mucinous neoplasms (IPMN) are cystic precursor lesions to pancreatic ductal adenocarcinoma (PDAC). IPMNs undergo multistep progression from low-grade (LG) to high-grade (HG) dysplasia, culminating in invasive neoplasia. While patterns of IPMN progression have been analyzed using multiregion sequencing for somatic mutations, there is no integrated assessment of molecular events, including copy-number alterations (CNA) and transcriptional changes that accompany IPMN progression. We performed laser capture microdissection on surgically resected IPMNs of varying grades of histologic dysplasia obtained from 23 patients, followed by whole-exome and whole-transcriptome sequencing. Overall, HG IPMNs displayed a significantly greater aneuploidy score than LG lesions, with chromosome 1q amplification being associated with HG progression and with cases that harbored co-occurring PDAC. Furthermore, the combined assessment of single-nucleotide variants (SNV) and CNAs identified both linear and branched evolutionary trajectories, underscoring the heterogeneity in the progression of LG lesions to HG and PDAC. At the transcriptome level, upregulation of MYC-regulated targets and downregulation of transcripts associated with the MHC class I antigen presentation machinery as well as pathways related to glycosylation were a common feature of progression to HG. In addition, the established PDAC transcriptional subtypes (basal-like and classical) were readily apparent within IPMNs. Taken together, this work emphasizes the role of 1q copy-number amplification as a putative biomarker of high-risk IPMNs, underscores the importance of immune evasion even in noninvasive precursor lesions, and reinforces that evolutionary pathways in IPMNs are heterogenous, comprised of both SNV and CNA-driven events. SIGNIFICANCE: Integrated molecular analysis of genomic and transcriptomic alterations in the multistep progression of IPMNs, which are bona fide precursors of pancreatic cancer, identifies features associated with progression of low-risk lesions to high-risk lesions and cancer, which might enable patient stratification and cancer interception strategies.

Stem-like breast cancer cells in the activated state resist genetic stress via TGFBI-ZEB1.

Breast cancer cells with stem-like properties are critical for tumor progression, yet much about these cells remains unknown. Here, we characterize a population of stem-like breast cancer cells expressing the integrin αvß3 as transcriptionally related to activated stem/basal cells in the normal human mammary gland. An unbiased functional screen of genes unique to these cells identified the matrix protein TGFBI (BIG-H3) and the transcription factor ZEB1 as necessary for tumorsphere formation. Surprisingly, these genes were not required for cell proliferation or survival, but instead maintained chromosomal stability. Consistent with this finding, CRISPR deletion of either gene synergized with PARP inhibition to deplete αvß3+ stem-like cells, which are normally resistant to this therapy. Our findings highlight a critical role for TGFBI-ZEB1 protection against genetic stress as a key attribute of activated stem-like cells and suggest that disrupting this ability may enhance their "BRCAness" by increasing sensitivity to PARP inhibitors.

The breast pre-cancer atlas illustrates the molecular and micro-environmental diversity of ductal carcinoma in situ.

Microenvironmental and molecular factors mediating the progression of Breast Ductal Carcinoma In Situ (DCIS) are not well understood, impeding the development of prevention strategies and the safe testing of treatment de-escalation. We addressed methodological barriers and characterized the mutational, transcriptional, histological, and microenvironmental landscape across 85 multiple microdissected regions from 39 cases. Most somatic alterations, including whole-genome duplications, were clonal, but genetic divergence increased with physical distance. Phenotypic and subtype heterogeneity was frequently associated with underlying genetic heterogeneity and regions with low-risk features preceded those with high-risk features according to the inferred phylogeny. B- and T-lymphocytes spatial analysis identified three immune states, including an epithelial excluded state located preferentially at DCIS regions, and characterized by histological and molecular features of immune escape, independently from molecular subtypes. Such breast pre-cancer atlas with uniquely integrated observations will help scope future expansion studies and build finer models of outcomes and progression risk.

Stromal remodeling regulates dendritic cell abundance and activity in the tumor microenvironment.

Stimulatory type 1 conventional dendritic cells (cDC1s) engage in productive interactions with CD8+ effectors along tumor-stroma boundaries. The paradoxical accumulation of "poised" cDC1s within stromal sheets is unlikely to simply reflect passive exclusion from tumor cores. Drawing parallels with embryonic morphogenesis, we hypothesized that invasive margin stromal remodeling generates developmentally conserved cell fate cues that regulate cDC1 behavior. We find that, in human T cell-inflamed tumors, CD8+ T cells penetrate tumor nests, whereas cDC1s are confined within adjacent stroma that recurrently displays site-specific proteolysis of the matrix proteoglycan versican (VCAN), an essential organ-sculpting modification in development. VCAN is necessary, and its proteolytic fragment (matrikine) versikine is sufficient for cDC1 accumulation. Versikine does not influence tumor-seeding pre-DC differentiation; rather, it orchestrates a distinctive cDC1 activation program conferring exquisite sensitivity to DNA sensing, supported by atypical innate lymphoid cells. Thus, peritumoral stroma mimicking embryonic provisional matrix remodeling regulates cDC1 abundance and activity to elicit T cell-inflamed tumor microenvironments.

Proteomic profiling dataset of chemical perturbations in multiple biological backgrounds.

While gene expression profiling has traditionally been the method of choice for large-scale perturbational profiling studies, proteomics has emerged as an effective tool in this context for directly monitoring cellular responses to perturbations. We previously reported a pilot library containing 3400 profiles of multiple perturbations across diverse cellular backgrounds in the reduced-representation phosphoproteome (P100) and chromatin space (Global Chromatin Profiling, GCP). Here, we expand our original dataset to include profiles from a new set of cardiotoxic compounds and from astrocytes, an additional neural cell model, totaling 5300 proteomic signatures. We describe filtering criteria and quality control metrics used to assess and validate the technical quality and reproducibility of our data. To demonstrate the power of the library, we present two case studies where data is queried using the concept of "connectivity" to obtain biological insight. All data presented in this study have been deposited to the ProteomeXchange Consortium with identifiers PXD017458 (P100) and PXD017459 (GCP) and can be queried at https://clue.io/proteomics .

Mutational profiling of micro-dissected pre-malignant lesions from archived specimens.

BACKGROUND: Systematic cancer screening has led to the increased detection of pre-malignant lesions (PMLs). The absence of reliable prognostic markers has led mostly to over treatment resulting in potentially unnecessary stress, or insufficient treatment and avoidable progression. Importantly, most mutational profiling studies have relied on PML synchronous to invasive cancer, or performed in patients without outcome information, hence limiting their utility for biomarker discovery. The limitations in comprehensive mutational profiling of PMLs are in large part due to the significant technical and methodological challenges: most PML specimens are small, fixed in formalin and paraffin embedded (FFPE) and lack matching normal DNA. METHODS: Using test DNA from a highly degraded FFPE specimen, multiple targeted sequencing approaches were evaluated, varying DNA input amount (3-200 ng), library preparation strategy (BE: Blunt-End, SS: Single-Strand, AT: A-Tailing) and target size (whole exome vs. cancer gene panel). Variants in high-input DNA from FFPE and mirrored frozen specimens were used for PML-specific variant calling training and testing, respectively. The resulting approach was applied to profile and compare multiple regions micro-dissected (mean area 5 mm2) from 3 breast ductal carcinoma in situ (DCIS). RESULTS: Using low-input FFPE DNA, BE and SS libraries resulted in 4.9 and 3.7 increase over AT libraries in the fraction of whole exome covered at 20x (BE:87%, SS:63%, AT:17%). Compared to high-confidence somatic mutations from frozen specimens, PML-specific variant filtering increased recall (BE:85%, SS:80%, AT:75%) and precision (BE:93%, SS:91%, AT:84%) to levels expected from sampling variation. Copy number alterations were consistent across all tested approaches and only impacted by the design of the capture probe-set. Applied to DNA extracted from 9 micro-dissected regions (8 PML, 1 normal epithelium), the approach achieved comparable performance, illustrated the data adequacy to identify candidate driver events (GATA3 mutations, ERBB2 or FGFR1 gains, TP53 loss) and measure intra-lesion genetic heterogeneity. CONCLUSION: Alternate experimental and analytical strategies increased the accuracy of DNA sequencing from archived micro-dissected PML regions, supporting the deeper molecular characterization of early cancer lesions and achieving a critical milestone in the development of biology-informed prognostic markers and precision chemo-prevention strategies.

A Library of Phosphoproteomic and Chromatin Signatures for Characterizing Cellular Responses to Drug Perturbations.

Although the value of proteomics has been demonstrated, cost and scale are typically prohibitive, and gene expression profiling remains dominant for characterizing cellular responses to perturbations. However, high-throughput sentinel assays provide an opportunity for proteomics to contribute at a meaningful scale. We present a systematic library resource (90 drugs × 6 cell lines) of proteomic signatures that measure changes in the reduced-representation phosphoproteome (P100) and changes in epigenetic marks on histones (GCP). A majority of these drugs elicited reproducible signatures, but notable cell line- and assay-specific differences were observed. Using the "connectivity" framework, we compared signatures across cell types and integrated data across assays, including a transcriptional assay (L1000). Consistent connectivity among cell types revealed cellular responses that transcended lineage, and consistent connectivity among assays revealed unexpected associations between drugs. We further leveraged the resource against public data to formulate hypotheses for treatment of multiple myeloma and acute lymphocytic leukemia. This resource is publicly available at https://clue.io/proteomics.