Search

Motivation The advancement of high-throughput technology characterizes a wide variety of epigenetic modifications and noncoding RNAs across the genome involved in disease pathogenesis via regulating gene expression. The high dimensionality of both epigenetic/noncoding RNA and gene expression data make it challenging to identify the important regulators of genes. Conducting univariate test for each possible regulator–gene pair is subject to serious multiple comparison burden, and direct application of regularization methods to select regulator–gene pairs is computationally infeasible. Applying fast screening to reduce dimension first before regularization is more efficient and stable than applying regularization methods alone. Results We propose a novel screening method based on robust partial correlation to detect epigenetic and noncoding RNA regulators of gene expression over the whole genome, a problem that includes both high-dimensional predictors and high-dimensional responses. Compared to existing screening methods, our method is conceptually innovative that it reduces the dimension of both predictor and response, and screens at both node (regulators or genes) and edge (regulator–gene pairs) levels. We develop data-driven procedures to determine the conditional sets and the optimal screening threshold, and implement a fast iterative algorithm. Simulations and applications to long noncoding RNA and microRNA regulation in Kidney cancer and DNA methylation regulation in Glioblastoma Multiforme illustrate the validity and advantage of our method. Availability and implementation The R package, related source codes and real datasets used in this article are provided at https://github.com/kehongjie/rPCor. Supplementary information Supplementary data are available at Bioinformatics online.

BACKGROUND: Substance use disorders (SUDs) are associated with disruptions in circadian rhythms. Both human and animal work has shown the integral role for circadian clocks in the modulation of reward behaviors. Interestingly, astrocytes have emerged as key regulators of circadian rhythmicity. However, no studies to date have identified the role of circadian astrocyte function in the nucleus accumbens (NAc), a hub for reward regulation, or determined the importance of these rhythms for reward-related behavior. METHODS: Using astrocyte-specific RNA-sequencing across time-of-day, we first characterized diurnal variation of the NAc astrocyte transcriptome. We then investigated the functional significance of this circadian regulation through viral-mediated disruption of molecular clock function in NAc astrocytes, followed by assessment of reward-related behaviors, metabolic-related molecular assays, and whole-cell electrophysiology in the NAc. RESULTS: Strikingly, ~43% of the astrocyte transcriptome has a diurnal rhythm and key metabolic pathways were enriched among the top rhythmic genes. Moreover, mice with a viral-mediated loss of molecular clock function in NAc astrocytes show a significant increase in locomotor response to novelty, exploratory drive, operant food self-administration and motivation. At the molecular level, these animals also show disrupted metabolic gene expression, along with significant downregulation of both lactate and glutathione levels in the NAc. Importantly, loss of NAc astrocyte clock function also significantly altered glutamatergic signaling onto neighboring medium spiny neurons, alongside upregulated glutamate-related gene expression. CONCLUSIONS: Taken together, these findings demonstrate a novel role for astrocyte circadian molecular clock function in the regulation of the NAc and reward-related behaviors.

The striatum in the brain is involved in various behavioral functions, including reward, and disease processes, such as opioid use disorder (OUD). Further understanding of the role of striatal subregions in reward behaviors and their potential associations with OUD requires molecular identification of specific striatal cell types in human brain. The human striatum contains subregions based on different anatomical, functional, and physiological properties, with the dorsal striatum further divided into caudate and putamen. Both caudate and putamen are associated with alterations in reward processing, formation of habits, and development of negative affect states in OUD. Using single nuclei RNA- sequencing of human postmortem caudate and putamen, we identified canonical neuronal cell types in striatum (e.g.,dopamine receptor 1 or 2 expressing neurons, D1 or D2) and less abundant subpopulations, including D1/D2 hybrid neurons and multiple classes of interneurons. By comparing unaffected subjects to subjects with OUD, we found neuronal-specific differences in pathways related to neurodegeneration, interferon response, and DNA damage. DNA damage markers were also elevated in striatal neurons of rhesus macaques following chronic opioid administration. We identified sex-dependent differences in the expression of stress-induced transcripts (e.g., FKBP5) among astrocytes and oligodendrocytes from female subjects with OUD. Thus, we describe striatal cell types and leverage these data to gain insights into molecular alterations in human striatum associated with opioid addiction.

Opioid craving and the vulnerability to relapse is associated with severe and persistent disruptions to sleep and circadian rhythms. Investigations into the cellular and molecular pathways in the human brain underlying the relationship between circadian rhythms and OUD remain limited. In human subjects with OUD, previous transcriptomics work implicated a role for circadian regulation of synaptic processes in key cognitive- and reward-related brain regions, dorsolateral prefrontal cortex (DLPFC) and nucleus accumbens (NAc). To provide further insights into the synaptic alterations associated with OUD, we used mass-spectrometry based proteomics to deeply profile protein alterations in tissue homogenates and synaptosomes from both NAc and DLPFC of unaffected and OUD subjects. Between unaffected and OUD subjects, we identified 43 differentially expressed (DE) proteins in NAc homogenates and 55 DE proteins in DLPFC homogenates. In synaptosomes, we found 56 DE proteins in NAc of OUD subjects and 161 DE proteins in DLPFC. Examining synaptosome enrichment of specific proteins enabled us to identify brain region- and synapse-specific pathway alterations in NAc and DLPFC associated with OUD. Across both regions, we found OUD-associated protein alterations primarily in pathways involved in GABAergic and glutamatergic synaptic functions, as well as circadian rhythms. Using time-of-death (TOD) analyses, where the TOD of each subject is used as a time-point across a 24-hour cycle, we were able to map circadian-related changes in the synaptic proteomes in NAc and DLPFC associated with OUD. In OUD, TOD analysis revealed significant circadian changes in endoplasmic reticulum to Golgi vesicle-mediated transport and protein membrane trafficking in NAc synapses, accompanied by changes in platelet derived growth factor receptor beta signaling in DLPFC synapses. Together, our results lend further support for molecular disruption of circadian regulation of synaptic signaling in the human brain as a key factor in opioid addiction.

Supplemental Fig S5. Imprinted genes display approximately 50% DNA methylation. Gene regions known to be imprinted were chosen to confirm the validity of the Methyl-Seq assay. For each gene a β value was calculated for each window. The frequency of each particular β value is graphically represented for each gene.

Supplemental Figure S2. All samples from both time points were used in a hierarchal clustering analysis. 10 samples at the early time point and 11 samples at the late time point were included.(E=early, L=late, P=parous, NP=nulliparous)

Supplemental Fig S8. Hypermethylation of the Igf1r aligns with sites of histone methylation. (A) The UCSC Genome Browser was used to align the Igf1r with other epigenetic modifications. The red bar indicates the region of differential DNA methylation identified in the SureSelectXT Methyl-Seq Assay. Black bars represent sites of histone methylation, the intensity corresponds to the degree of methylation. (B) Igf1r bisulfite sequencing data are depicted. Each animal is numbered and each horizontal line represents DNA from a separate clone from that animal. The percent methylation of each animal is listed under each CpG site. Open circles = unmethylated, and closed circles = methylated. The lower panel represents a combined result from all clones in all animals.

Supplemental Fig S7. Igf1r is central to a top signaling network in Ingenuity pathway analysis. Ingenuity pathway analysis (IPA) identified this signaling network to be significantly altered by DNA methylation. The hyper and hypomethylated genes identified to be persistently differentially methylated by the meta-analysis were input into IPA (Supplemental Data 1). Red symbols represent hypermethylation, green symbols represent hypomethylation and intensity depicts methylation difference between nulliparous and parous.

Primers used for quantitative PCR and bisulfite sequencing.

Supplemental Fig S9. Differentially methylated regions of IGF pathway genes align with histone methylation sites. The UCSC Genome Browser was used to align the Irs1, Igf1, and Igfbp4 with other epigenetic modifications. The red bars indicate the regions of differential DNA methylation identified in the SureSelectXT Methyl-Seq Assay. Black bars represent sites of histone methylation, the intensity corresponds to the degree of methylation.

The most effective natural prevention against breast cancer is an early first full-term pregnancy. Understanding how the protective effect is elicited will inform the development of new prevention strategies. To better understand the role of epigenetics in long-term protection, we investigated parity-induced DNA methylation in the mammary gland. FVB mice were bred or remained nulliparous and mammary glands harvested immediately after involution (early) or 6.5 months following involution (late), allowing identification of both transient and persistent changes. Targeted DNA methylation (109 Mb of Ensemble regulatory features) analysis was performed using the SureSelectXT Mouse Methyl-seq assay and massively parallel sequencing. Two hundred sixty-nine genes were hypermethylated and 128 hypomethylated persistently at both the early and late time points. Pathway analysis of the persistently differentially methylated genes revealed Igf1r to be central to one of the top identified signaling networks, and Igf1r itself was one of the most significantly hypermethylated genes. Hypermethylation of Igf1r in the parous mammary gland was associated with a reduction of Igf1r mRNA expression. These data suggest that the IGF pathway is regulated at multiple levels during pregnancy and that its modification might be critical in the protective role of pregnancy. This supports the approach of lowering IGF action for prevention of breast cancer, a concept that is currently being tested clinically. Cancer Prev Res; 8(10); 1000–9. ©2015 AACR.

Supplemental Figure S3. Mapping efficiency was above 50 for all samples in both time points. One sample was an outlier in the early time point (P12) and one at the late time point (NP4) and were removed from further analysis. Both Pearson Correlation plots and Principal Component Analysis are displayed.

All Supplementary Figure legends are included in this document.

Supplemental Figure S6. Supervised hierarchical clustering and cell type content analysis. (A) Heatmaps displaying supervised hierarchical clustering of hypermethylated and hypomethylated windows including dendrograms (Red=Methylated, Green=Unmethylated).

Supplemental Data 1. This spreadsheet contains the results of the differential methylation analysis, the meta-analysis and the pathway analysis. In the meta-analysis sheets, the first 3 columns represent the chromosomal locations, followed by the significance level (stat, p-value, and FDR). Weight.1 and weight.2 represent the weight of early and late data respectively. For the meta-analysis we weighted both as 1. The effect size represents the difference in DNA methylation between NP and P at the early and late time points (effect size.1=early, effect size.2=late), and the p-value for each is next depicted. We then filtered the data by trimmed means as in the Materials and Methods section and the new values are listed in columns N and O. We then annotated the gene name, which genomic feature the loci resides in, and the distance to the nearest promoter (distance to feature). We then annotated the mean coverage and the number of CpG sites in the differentially methylated region. The differential methylation sheets include similar data without the columns pertaining to the meta-analysis. The 3Ingenuity Pathway Analysis (IPA) spreadsheets contain the pathway analysis result. The pathway name, followed by the statistical significance, and molecules represented in our data for each pathway are detailed on each sheet. The analysis was first conducted on the hyper and hypomethylated genes separately, then the combined hyper and hypermethylated genes were analyzed together.

Supplementary Figure S1. Experimental design and quality control analysis for Methyl-Seq data. (A) FVB mice were bred to become pregnant (parous, P) and underwent a complete pregnancy (Pr), lactation (L), and Involution (I). Animals were then euthanized immediately after involution (Early, E), or 6.5 months post-involution (Late, Lt). Nulliparous (NP) age matched controls were included at each time point. Mammary glands were harvested and genomic DNA was isolated and processed for input into the SureSelectXT Mouse Methy-Seq Enrichment Platform for DNA methylation analysis. (B) Quantitative RT-PCR of cell type specific transcripts. Each data point represents an animal, and bars represent mean ? SEM.

Antibody and shRNA sequence information.

Supplementary Data from Hotspot ESR1 Mutations Are Multimodal and Contextual Modulators of Breast Cancer Metastasis

Constitutively active estrogen receptor α (ER/ESR1) mutations have been identified in approximately one-third of ER+ metastatic breast cancers. Although these mutations are known as mediators of endocrine resistance, their potential role in promoting metastatic disease has not yet been mechanistically addressed. In this study, we show the presence of ESR1 mutations exclusively in distant but not local recurrences in five independent breast cancer cohorts. In concordance with transcriptomic profiling of ESR1-mutant tumors, genome-edited ESR1 Y537S and D538G-mutant cell models exhibited a reprogrammed cell adhesive gene network via alterations in desmosome/gap junction genes and the TIMP3/MMP axis, which functionally conferred enhanced cell–cell contacts while decreasing cell-extracellular matrix adhesion. In vivo studies showed ESR1-mutant cells were associated with larger multicellular circulating tumor cell (CTC) clusters with increased compactness compared with ESR1 wild-type CTCs. These preclinical findings translated to clinical observations, where CTC clusters were enriched in patients with ESR1-mutated metastatic breast cancer. Conversely, context-dependent migratory phenotypes revealed cotargeting of Wnt and ER as a vulnerability in a D538G cell model. Mechanistically, mutant ESR1 exhibited noncanonical regulation of several metastatic pathways, including secondary transcriptional regulation and de novo FOXA1-driven chromatin remodeling. Collectively, these data provide evidence for ESR1 mutation–modulated metastasis and suggest future therapeutic strategies for targeting ESR1-mutant breast cancer.Significance:Context- and allele-dependent transcriptome and cistrome reprogramming in mutant ESR1 cell models elicit diverse metastatic phenotypes related to cell adhesion and migration, which can be pharmacologically targeted in metastatic breast cancer.

Supplementary Text

Invasive lobular carcinoma (ILC) is the second most common subtype of breast cancer following invasive ductal carcinoma (IDC) and characterized by the loss of E-cadherin–mediated adherens junctions. Despite displaying unique histologic and clinical features, ILC still remains a chronically understudied disease, with limited knowledge gleaned from available laboratory research models. Here we report a comprehensive 2D and 3D phenotypic characterization of four estrogen receptor–positive human ILC cell lines: MDA-MB-134, SUM44, MDA-MB-330, and BCK4. Compared with the IDC cell lines MCF7, T47D, and MDA-MB-231, ultra-low attachment culture conditions revealed remarkable anchorage independence unique to ILC cells, a feature not evident in soft-agar gels. Three-dimensional Collagen I and Matrigel culture indicated a generally loose morphology for ILC cell lines, which exhibited differing preferences for adhesion to extracellular matrix proteins in 2D. Furthermore, ILC cells were limited in their ability to migrate and invade in wound-scratch and transwell assays, with the exception of haptotaxis to Collagen I. Transcriptional comparison of these cell lines confirmed the decreased cell proliferation and E-cadherin–mediated intercellular junctions in ILC while uncovering the induction of novel pathways related to cyclic nucleotide phosphodiesterase activity, ion channels, drug metabolism, and alternative cell adhesion molecules such as N-cadherin, some of which were differentially regulated in ILC versus IDC tumors. Altogether, these studies provide an invaluable resource for the breast cancer research community and facilitate further functional discoveries toward understanding ILC, identifying novel drug targets, and ultimately improving the outcome of patients with ILC.Significance: These findings provide the breast cancer research community with a comprehensive assessment of human invasive lobular carcinoma (ILC) cell line signaling and behavior in various culture conditions, aiding future endeavors to develop therapies and to ultimately improve survival in patients with ILC. Cancer Res; 78(21); 6209–22. ©2018 AACR.

Phenotype–gene association studies can uncover disease mechanisms for translational research. Association with multiple phenotypes or clinical variables in complex diseases has the advantage of increasing statistical power and offering a holistic view. Existing multi-variate association methods mostly focus on SNP-based genetic associations. In this paper, we extend and evaluate two adaptive Fisher’s methods, namely AFp and AFz, from the p-value combination perspective for phenotype–mRNA association analysis. The proposed method effectively aggregates heterogeneous phenotype–gene effects, allows association with different data types of phenotypes, and performs the selection of the associated phenotypes. Variability indices of the phenotype–gene effect selection are calculated by bootstrap analysis, and the resulting co-membership matrix identifies gene modules clustered by phenotype–gene effect. Extensive simulations demonstrate the superior performance of AFp compared to existing methods in terms of type I error control, statistical power and biological interpretation. Finally, the method is separately applied to three sets of transcriptomic and clinical datasets from lung disease, breast cancer, and brain aging and generates intriguing biological findings.

Substance use disorders (SUDs) are associated with disruptions in sleep and circadian rhythms that persist during abstinence and may contribute to relapse risk. Repeated use of substances such as psychostimulants and opioids may lead to significant alterations in molecular rhythms in the nucleus accumbens (NAc), a brain region central to reward and motivation. Previous studies have identified rhythm alterations in the transcriptome of the NAc and other brain regions following the administration of psychostimulants or opioids. However, little is known about the impact of substance use on the diurnal rhythms of the proteome in the NAc. We used liquid chromatography coupled to tandem mass spectrometry-based (LC-MS/MS) quantitative proteomics, along with a data-independent acquisition (DIA) analysis pipeline, to investigate the effects of cocaine or morphine administration on diurnal rhythms of proteome in the mouse NAc. Overall, our data reveals cocaine and morphine differentially alters diurnal rhythms of the proteome in the NAc, with largely independent differentially expressed proteins dependent on time-of-day. Pathways enriched from cocaine altered protein rhythms were primarily associated with glucocorticoid signaling and metabolism, whereas morphine was associated with neuroinflammation. Collectively, these findings are the first to characterize the diurnal regulation of the NAc proteome and demonstrate a novel relationship between phase-dependent regulation of protein expression and the differential effects of cocaine and morphine on the NAc proteome.

Estrogen receptor alpha (ER/ESR1) is mutated in 30-40% of endocrine resistant ER-positive (ER+) breast cancer. ESR1 mutations cause ligand-independent growth and increased metastasis in vivo and in vitro. Despite the distinct clinical features and changes in therapeutic response associated with ESR1 mutations, there are no data about their potential role in intrinsic subtype switching.Applying five different luminal and basal gene set pairs derived from cell lines and tumors, ESR1 mutant cell models and clinical samples showed a significant enrichment of basal subtype markers. Among them, the six basal cytokeratins (BCKs) were the most enriched genes (KRT5/6A/6B/14/16/17) uniquely in ESR1 mutant cells but not other endocrine resistant cell models. BCKs were observed to heterogeneously express in a minor cell subpopulation in ESR1 mutant cell models and clinical specimens. ER ChIP-seq showed the mutant-specific induction of BCKs was independent of ER binding and instead selectively expressed in clones with low ER expression. In contrast, BCKs are associated with chromatin reprogramming centered around a progesterone receptor-orchestrated 154 kb insulated neighborhood at the KRT14/16/17 genomic region. Stronger CTCF binding was detected at the bases of chromatin loop in ESR1 mutant cells. Knockdown of progesterone receptor but not glucocorticoid receptor drastically blocked the induction of KRT14/16/17 in ESR1 mutant cells. Unexpectedly, high BCK expression in ER+ primary breast cancer is associated with good prognosis, and these tumors show enriched activation of a number of immune pathways, a distinctive feature shared with ESR1 mutant tumors. While the BCK-associated immune activation is not related to tumor mutation burdens, S100A8 and S100A9 were identified as the most highly induced immune mediators shared between high-BCKs ER+ and ESR1 mutant tumors, which was further validated in the plasma samples of a cohort of 18 patients with ER+ metastases (11 WT vs 7 mutant). Finally, single-cell RNA-seq analysis in an ER+ bone metastasis case inferred the involvement of S100A8 and S100A9 in paracrine crosstalk between epithelial and stromal cells, particularly macrophages and fibroblasts through TLR4 signaling. Collectively, these observations demonstrate that ESR1 mutant tumors gain basal features with induction of basal cytokeratins via epigenetic mechanisms in rare subpopulation of cells. This is associated with increased immune activation, encouraging additional studies of immune therapeutic vulnerabilities in ESR1 mutant tumors. Citation Format: Zheqi Li, Yang Wu, Olivia Mcginn, Amir Bahreini, Nolan M. Priedigkeit, Kai Ding, Sayali Onkar, Carol A. Sartorius, Lori Miller, Margaret Rosenzweig, Nikhil Wagle, Jennifer K. Richer, William J. Muller, Laki Buluwela, Simak Ali, Dario A.A. Vignali, Yusi Fang, Li Zhu, George C. Tseng, Jason Gertz, Jennifer M. Atkinson, Adrian V. Lee, Steffi Oesterreich. Esr1 mutant breast cancers show elevated basal cytokeratins and immune activation [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr PD1-08.

Constitutively active estrogen receptor α (ER/ESR1) mutations have been identified in approximately one-third of ER+ metastatic breast cancers. Although these mutations are known as mediators of endocrine resistance, their potential role in promoting metastatic disease has not yet been mechanistically addressed. In this study, we show the presence of ESR1 mutations exclusively in distant but not local recurrences in five independent breast cancer cohorts. In concordance with transcriptomic profiling of ESR1-mutant tumors, genome-edited ESR1 Y537S and D538G-mutant cell models exhibited a reprogrammed cell adhesive gene network via alterations in desmosome/gap junction genes and the TIMP3/MMP axis, which functionally conferred enhanced cell–cell contacts while decreasing cell-extracellular matrix adhesion. In vivo studies showed ESR1-mutant cells were associated with larger multicellular circulating tumor cell (CTC) clusters with increased compactness compared with ESR1 wild-type CTCs. These preclinical findings translated to clinical observations, where CTC clusters were enriched in patients with ESR1-mutated metastatic breast cancer. Conversely, context-dependent migratory phenotypes revealed cotargeting of Wnt and ER as a vulnerability in a D538G cell model. Mechanistically, mutant ESR1 exhibited noncanonical regulation of several metastatic pathways, including secondary transcriptional regulation and de novo FOXA1-driven chromatin remodeling. Collectively, these data provide evidence for ESR1 mutation–modulated metastasis and suggest future therapeutic strategies for targeting ESR1-mutant breast cancer. Significance: Context- and allele-dependent transcriptome and cistrome reprogramming in mutant ESR1 cell models elicit diverse metastatic phenotypes related to cell adhesion and migration, which can be pharmacologically targeted in metastatic breast cancer.

Background Diurnal rhythms in gene expression have been detected in the human brain. Previous studies found that males and females exhibit 24-hour rhythms in known circadian genes, with earlier peak expression in females. Whether there are sex differences in large-scale transcriptional rhythms in the cortex that align with observed sex differences in physiological and behavioral rhythms is currently unknown. Methods Diurnal rhythmicity of gene expression was determined for males and females using RNA sequencing data from human postmortem dorsolateral prefrontal cortex (DLPFC) and anterior cingulate cortex (ACC). Sex differences among rhythmic genes were determined using significance cutoffs, threshold-free analyses, and R2 difference. Phase concordance was assessed across the DLPFC and ACC for males and females. Pathway and transcription factor analyses were also conducted on significantly rhythmic genes. Results Canonical circadian genes had diurnal rhythms in both sexes with similar amplitude and phase. When analyses were expanded to the entire transcriptome, significant sex differences in transcriptional rhythms emerged. There were nearly twice as many rhythmic transcripts in the DLPFC in males and nearly 4 times as many rhythmic transcripts in the ACC in females. Results suggest a diurnal rhythm in synaptic transmission specific to the ACC in females (e.g., GABAergic [gamma-aminobutyric acidergic] and cholinergic neurotransmission). For males, there was phase concordance between the DLPFC and ACC, while phase asynchrony was found in females. Conclusions There are robust sex differences in molecular rhythms of genes in the DLPFC and ACC, providing potential mechanistic insights into how neurotransmission and synaptic function are modulated in a circadian-dependent and sex-specific manner.

As one of the most successful cancer therapeutic targets, estrogen receptor-α (ER/ESR1) has been extensively studied in decade-long. Sequencing technological advances have enabled genome-wide analysis of ER action. However, reproducibility is limited by different experimental design. Here, we established the EstroGene database through centralizing 246 experiments from 136 transcriptomic, cistromic and epigenetic datasets focusing on estradiol-treated ER activation across 19 breast cancer cell lines. We generated a user-friendly browser (https://estrogene.org/) for data visualization and gene inquiry under user-defined experimental conditions and statistical thresholds. Notably, documentation-based meta-analysis revealed a considerable lack of experimental details. Comparison of independent RNA-seq or ER ChIP-seq data with the same design showed large variability and only strong effects could be consistently detected. We defined temporal estrogen response metasignatures and showed the association with specific transcriptional factors, chromatin accessibility and ER heterogeneity. Unexpectedly, harmonizing 146 transcriptomic analyses uncovered a subset of E2-bidirectionally regulated genes, which linked to immune surveillance in the clinical setting. Furthermore, we defined context dependent E2 response programs in MCF7 and T47D cell lines, the two most frequently used models in the field. Collectively, the EstroGene database provides an informative resource to the cancer research community and reveals a diverse mode of ER signaling.

Model organisms are instrumental substitutes for human studies to expedite basic, translational, and clinical research. Despite their indispensable role in mechanistic investigation and drug development, molecular congruence of animal models to humans has long been questioned and debated. Little effort has been made for an objective quantification and mechanistic exploration of a model organism’s resemblance to humans in terms of molecular response under disease or drug treatment. We hereby propose a framework, namely Congruence Analysis for Model Organisms (CAMO), for transcriptomic response analysis by developing threshold-free differential expression analysis, quantitative concordance/discordance scores incorporating data variabilities, pathway-centric downstream investigation, knowledge retrieval by text mining, and topological gene module detection for hypothesis generation. Instead of a genome-wide vague and dichotomous answer of “poorly” or “greatly” mimicking humans, CAMO assists researchers to numerically quantify congruence, to dissect true cross-species differences from unwanted biological or cohort variabilities, and to visually identify molecular mechanisms and pathway subnetworks that are best or least mimicked by model organisms, which altogether provides foundations for hypothesis generation and subsequent translational decisions.

The immune tumor microenvironment (TME) of epithelial ovarian cancer (EOC) carries both effector and suppressive functions. To define immune correlates of chemotherapy-induced tumor involution, we performed longitudinal evaluation of biomarker expression on serial biological specimens collected during intraperitoneal (IP) platinum-based chemotherapy. Serial biological samples were collected at several time points during IP chemotherapy. RNA from IP fluid cells and tumor tissue was analyzed via NanoString. Meso Scale Discovery (MSD) multiplex assay and ELISA for MUC1 antibodies were performed on plasma and IP fluid. Differentially expressed genes in IP fluid demonstrate an upregulation of B cell function and activation of Th2 immune response along with dampening of Th1 immunity during chemotherapy. MSD analysis of IP fluid and gene expression analysis of tumor tissue revealed activation of Th2 immunity and the complement system. Anti-MUC1 antibodies were detected in IP fluid samples. IP fluid analysis in a secondary cohort also identified chemotherapy-induced B cell function genes. This study shows that serial IP fluid sampling is an effective method to capture changes in the immune TME during chemotherapy and reveals treatment induced changes in B cell function and Th2 immunity.

Summary Clustering with variable selection is a challenging yet critical task for modern small-n-large-p data. Existing methods based on sparse Gaussian mixture models or sparse $K$-means provide solutions to continuous data. With the prevalence of RNA-seq technology and lack of count data modeling for clustering, the current practice is to normalize count expression data into continuous measures and apply existing models with a Gaussian assumption. In this article, we develop a negative binomial mixture model with lasso or fused lasso gene regularization to cluster samples (small $n$) with high-dimensional gene features (large $p$). A modified EM algorithm and Bayesian information criterion are used for inference and determining tuning parameters. The method is compared with existing methods using extensive simulations and two real transcriptomic applications in rat brain and breast cancer studies. The result shows the superior performance of the proposed count data model in clustering accuracy, feature selection, and biological interpretation in pathways.

Clinicopathologic differences between histological subtypes of invasive breast cancer are increasingly being appreciated. Mixed invasive ductal lobular carcinomas (mDLC) are thought to be composed of both ductal and lobular components, and we sought to determine whether mDLC clinically align more closely with invasive ductal (IDC) or invasive lobular (ILC) carcinoma subtypes or if they display intermediate or unique features dissimilar to either type. Key clinical and histologic parameters were compared between cohorts of patients with mDLC (N = 410), IDC (N = 12,979), and ILC (N = 1,569) identified from cancer registry data of a single large healthcare system.Patients with mDLC were older (59 years (49 - 68)) than those with IDC (57 years (48 - 67), p = 0.014)) and younger than those with ILC (61 years (51 - 70), p = 0.006). Tumor size in mDLC was larger (19mm (12 - 27)) than IDC (16mm (10 - 25), p < 0.001) and smaller than ILC (20mm (12 - 35), p = 0.036). Similar to ILC, mDLC were more likely than IDC to be ER+ (92% vs 78% in IDC, p < 0.001), and less likely to be HER2+ (8% vs 15% in IDC, p = 0.04). mDLC were also similar to ILC with regards to higher likelihood of diagnosis at higher stage (p < 0.001), yet with lower grade (p < 0.001), at diagnosis as compared to IDC. Heatmap visualization, as well as dimension reduction by multidimentional scaling (MDS), demonstrates significant overlap of the mDLC and ILC cohorts. Furthermore, an elastic net regression model based on clinicopathologic parameters predicts mDLC to align more closely with ILC than IDC. For patients for whom oncotype Dx scores were available, there was a trend for enrichment of low risk RS scores with rare high-risk RS tumors in mDLC, similar to ILC. With regards to response to neoadjuvant chemotherapy, a subset of the aforementioned cohorts who had received neoadjuvant chemotherapy, mDLC (N = 17), IDC (N = 180), and ILC (N = 57), were compared. Among patients in whom breast conserving surgery (BCS) was attempted, patients with IDC were more likely to have a successful BCS than those with ILC, with less margin positivity thereby avoiding re-excision and/or completion mastectomy (70% vs 32%, respectively; p = 0.003). Successful BCS was achieved with mDLC 56% of the time, although compared to IDC and ILC statistical significance was not reached. In a limited cohort receiving neoadjuvant endocrine therapy (mDLC (N = 7), IDC (N = 37), and ILC (N = 21)) no differences with regard to rates of successful BCS were identified. Pathologic complete response rates (pCR) were additionally evaluated, although small study numbers precluded our ability to perform statistical analysis.Collectively, the aforementioned findings support a higher concordance between mDLC and ILC as compared to IDC. It is feasible that the lobular component of mDLC tumors is predominant, leading to the observed histopathologic similarities noted between mDC and ILC cohorts. We are planning meta-analyses including data from other institutions, and molecular studies to further understand complexities of mDLC.The authors acknowledge grant support from ASCO Conquer Cancer (to NW and AN). Citation Format: Azadeh Nasrazadani, Yujia Li, Yusi Fang, Osama Shah, Jennifer M Atkinson, Joanna S Lee, Priscilla F McAuliffe, Adrian V Lee, George Tseng, Peter Lucas, Steffi Oesterreich, Norman Wolmark. Mixed invasive ductal lobular carcinomas (mDLC) are clinically more similar to invasive lobular carcinoma (ILC) than to invasive ductal carcinoma (IDC) [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS7-15.