The Justy mutation identifies Gon4-like as a gene that is essential for B lymphopoiesis

Genetically engineered mouse models of head and neck cancers

The head and neck region is one of the anatomic sites commonly afflicted by cancer, with ~1.5 million new diagnoses reported worldwide in 2020 alone. Remarkable progress has been made in understanding the underlying disease mechanisms, personalizing care based on each tumor's individual molecular characteristics, and even therapeutically exploiting the inherent vulnerabilities of these neoplasms. In this regard, genetically engineered mouse models (GEMMs) have played an instrumental role. While progress in the development of GEMMs has been slower than in other major cancer types, several GEMMs are now available that recapitulate most of the heterogeneous characteristics of head and neck cancers such as the tumor microenvironment. Different approaches have been employed in GEMM development and implementation, though each can generally recapitulate only certain disease aspects. As a result, appropriate model selection is essential for addressing specific research questions. In this review, we present an overview of all currently available head and neck cancer GEMMs, encompassing models for head and neck squamous cell carcinoma, nasopharyngeal carcinoma, and salivary and thyroid gland carcinomas.

Genome-Wide Gene Expression Profiles Reveal Distinct Molecular Characteristics of the Goose Granulosa Cells

Granulosa cells (GCs) are decisive players in follicular development. In this study, the follicle tissues and GCs were isolated from the goose during the peak-laying period to perform hematoxylin-eosin staining and RNA-seq, respectively. Moreover, the dynamic mRNA and lncRNA expression profiles and mRNA-lncRNA network analysis were integrated to identify the important genes and lncRNAs. The morphological analysis showed that the size of the GCs did not significantly change, but the thickness of the granulosa layer cells differed significantly across the developmental stages. Subsequently, 14,286 mRNAs, 3,956 lncRNAs, and 1,329 TUCPs (transcripts with unknown coding potential) were detected in the GCs. We identified 37 common DEGs in the pre-hierarchical and hierarchical follicle stages, respectively, which might be critical for follicle development. Moreover, 3,089 significant time-course DEGs (Differentially expressed genes) and 13 core genes in 4 clusters were screened during goose GCs development. Finally, the network lncRNA G8399 with CADH5 and KLF2, and lncRNA G8399 with LARP6 and EOMES were found to be important for follicular development in GCs. Thus, the results would provide a rich resource for elucidating the reproductive biology of geese and accelerate the improvement of the egg-laying performance of geese.

The transcriptional regulator GON4L is required for viability and hematopoiesis in mice

The Gon4l gene encodes a putative transcriptional regulator implicated in the control of both cell differentiation and proliferation. Previously, we described a mutant mouse strain called Justy in which splicing of pre-mRNA generated from Gon4l is disrupted. This defect severely reduces, but does not abolish, GON4L protein expression and blocks the formation of early B-lineage progenitors, suggesting Gon4l is required for B-cell development in vertebrates. Yet, mutations that disable Gon4l in zebrafish impair several facets of embryogenesis that include the initiation of primitive hematopoiesis, arguing this gene is needed for multiple vertebrate developmental pathways. To better understand the importance of Gon4l in mammals, we created mice carrying an engineered version of Gon4l that can be completely inactivated by Cre-mediated recombination. Breeding mice heterozygous for the inactivated Gon4l allele failed to yield any homozygous-null offspring, indicating Gon4l is an essential gene in mammals. Consistent with this finding, as well previously published results, cell culture studies revealed that loss of Gon4l blocks cell proliferation and compromises viability, suggesting a fundamental role in the control of cell division and survival. Studies using mixed bone marrow chimeras confirmed Gon4l is required for B-cell development but also found it is needed to maintain definitive hematopoietic stem/progenitor cells that are the source of all hematopoietic cell lineages. Our findings reveal Gon4l is an essential gene in mammals that is required to form the entire hematopoietic system.

Newly identified Gon4l/Udu-interacting proteins implicate novel functions

Mutations of the Gon4l/udu gene in different organisms give rise to diverse phenotypes. Although the effects of Gon4l/Udu in transcriptional regulation have been demonstrated, they cannot solely explain the observed characteristics among species. To further understand the function of Gon4l/Udu, we used yeast two-hybrid (Y2H) screening to identify interacting proteins in zebrafish and mouse systems, confirmed the interactions by co-immunoprecipitation assay, and found four novel Gon4l-interacting proteins: BRCA1 associated protein-1 (Bap1), DNA methyltransferase 1 (Dnmt1), Tho complex 1 (Thoc1, also known as Tho1 or HPR1), and Cryptochrome circadian regulator 3a (Cry3a). Furthermore, all known Gon4l/Udu-interacting proteins—as found in this study, in previous reports, and in online resources—were investigated by Phenotype Enrichment Analysis. The most enriched phenotypes identified include increased embryonic tissue cell apoptosis, embryonic lethality, increased T cell derived lymphoma incidence, decreased cell proliferation, chromosome instability, and abnormal dopamine level, characteristics that largely resemble those observed in reported Gon4l/udu mutant animals. Similar to the expression pattern of udu, those of bap1, dnmt1, thoc1, and cry3a are also found in the brain region and other tissues. Thus, these findings indicate novel mechanisms of Gon4l/Udu in regulating CpG methylation, histone expression/modification, DNA repair/genomic stability, and RNA binding/processing/export.

Gon4l/Udu regulates cardiomyocyte proliferation and maintenance of ventricular chamber identity during zebrafish development

Vertebrate heart development requires spatiotemporal regulation of gene expression to specify cardiomyocytes, increase the cardiomyocyte population through proliferation, and to establish and maintain atrial and ventricular cardiac chamber identities. The evolutionarily conserved chromatin factor Gon4-like (Gon4l), encoded by the zebrafish ugly duckling (udu) locus, has previously been implicated in cell proliferation, cell survival, and specification of mesoderm-derived tissues including blood and somites, but its role in heart formation has not been studied. Here we report two distinct roles of Gon4l/Udu in heart development: regulation of cell proliferation and maintenance of ventricular identity. We show that zygotic loss of udu expression causes a significant reduction in cardiomyocyte number at one day post fertilization that becomes exacerbated during later development. We present evidence that the cardiomyocyte deficiency in udu mutants results from reduced cell proliferation, unlike hematopoietic deficiencies attributed to TP53-dependent apoptosis. We also demonstrate that expression of the G1/S-phase cell cycle regulator, cyclin E2 (ccne2), is reduced in udu mutant hearts, and that the Gon4l protein associates with regulatory regions of the ccne2 gene during early embryogenesis. Furthermore, udu mutant hearts exhibit a decrease in the proportion of ventricular cardiomyocytes compared to atrial cardiomyocytes, concomitant with progressive reduction of nkx2.5 expression. We further demonstrate that udu and nkx2.5 interact to maintain the proportion of ventricular cardiomyocytes during development. However, we find that ectopic expression of nkx2.5 is not sufficient to restore ventricular chamber identity suggesting that Gon4l regulates cardiac chamber patterning via multiple pathways. Together, our findings define a novel role for zygotically-expressed Gon4l in coordinating cardiomyocyte proliferation and chamber identity maintenance during cardiac development.

Primary and Metastatic Melanoma With NTRK Fusions

A number of oncogenic driver mutations have been identified in melanocytic nevi and melanoma, but translocations also play a role in tumorigenesis and provide potential therapeutic targets for malignant lesions. Various translocations, such as those involving the anaplastic lymphoma kinase (ALK), neurotrophic tropomyosin receptor kinase 1 (NTRK1), and NTRK3 have been reported in spitzoid melanocytic neoplasms leading to kinase-fusion proteins that result in immunohistochemically detectable ALK or NTRK expression. We have previously reported that ALK expression can be found in nonspitzoid primary and metastatic cutaneous melanomas. In this study we report that nonspitzoid metastasizing melanomas of adults may also harbor NTRK fusions and that NTRK expression can be immunohistochemically detected in these tumors. Of 751 melanomas analyzed by next-generation sequencing, 4 metastatic melanomas were identified with NTRK fusions, 3 involving NTRK1, 1 involving NTRK2. They occurred in 3 women and 1 man. Two of the corresponding primary tumors were from the trunk, 1 from an extremity and 1 tumor arose in anal skin. One primary tumor displayed features of superficial spreading melanoma and 3 were nodular melanomas. All tumors were cytologically characterized by the presence of large epithelioid melanocytes. All tumors were immunoreactive with anti-Trk antibody. Next-generation sequencing documented that the NTRK1 fusion partners included TRIM63, DDR2, and GON4L. One tumor harbored an NTRK2-TRAF2 fusion. Thus, our findings document that NTRK kinase fusions can occur in nonspitzoid metastasizing melanomas of adults. The presence of an NTRK family fusion in these tumors may provide a therapeutic opportunity in a small subset of patients with metastatic melanoma.

Gon4l regulates notochord boundary formation and cell polarity underlying axis extension by repressing adhesion genes

Anteroposterior (AP) axis extension during gastrulation requires embryonic patterning and morphogenesis to be spatiotemporally coordinated, but the underlying genetic mechanisms remain poorly understood. Here we define a role for the conserved chromatin factor Gon4l, encoded by ugly duckling (udu), in coordinating tissue patterning and axis extension during zebrafish gastrulation through direct positive and negative regulation of gene expression. Although identified as a recessive enhancer of impaired axis extension in planar cell polarity (PCP) mutants, udu functions in a genetically independent, partially overlapping fashion with PCP signaling to regulate mediolateral cell polarity underlying axis extension in part by promoting notochord boundary formation. Gon4l limits expression of the cell–cell and cell–matrix adhesion molecules EpCAM and Integrinα3b, excesses of which perturb the notochord boundary via tension-dependent and -independent mechanisms, respectively. By promoting formation of this AP-aligned boundary and associated cell polarity, Gon4l cooperates with PCP signaling to coordinate morphogenesis along the AP embryonic axis.

Anteroposterior axis extension during gastrulation is dynamically coordinated, but how this is regulated at a molecular level is unclear. Here, the authors show in zebrafish that the chromatin factor Gon4l, encoded by ugly duckling, coordinates axis extension by modulating EpCAM and Integrinα3b expression.

Early B Cell Progenitors Deficient for GON4L Fail To Differentiate Due to a Block in Mitotic Cell Division

B cell development in Justy mutant mice is blocked due to a precursor mRNA splicing defect that depletes the protein GON4-like (GON4L) in B cell progenitors. Genetic and biochemical studies have suggested that GON4L is a transcriptional regulator that coordinates cell division with differentiation, but its role in B cell development is unknown. To understand the function of GON4L, we characterized B cell differentiation, cell cycle control, and mitotic gene expression in GON4L-deficient B cell progenitors from Justy mice. We found that these cells established key aspects of the transcription factor network that guides B cell development and proliferation and rearranged the IgH gene locus. However, despite intact IL-7 signaling, GON4L-deficient pro-B cell stage precursors failed to undergo a characteristic IL-7-dependent proliferative burst. These cells also failed to upregulate genes required for mitotic division, including those encoding the G₁/S cyclin D3 and E2F transcription factors and their targets. Additionally, GON4L-deficient B cell progenitors displayed defects in DNA synthesis and passage through the G₁/S transition, contained fragmented DNA, and underwent apoptosis. These phenotypes were not suppressed by transgenic expression of prosurvival factors. However, transgenic expression of cyclin D3 or other regulators of the G₁/S transition restored pro-B cell development from Justy progenitor cells, suggesting that GON4L acts at the beginning of the cell cycle. Together, our findings indicate that GON4L is essential for cell cycle progression and division during the early stages of B cell development.

GON4L Drives Cancer Growth through a YY1-Androgen Receptor-CD24 Axis

In principle, the inhibition of candidate gain-of-function genes defined through genomic analyses of large patient cohorts offers an attractive therapeutic strategy. In this study, we focused on changes in expression of CD24, a well-validated clinical biomarker of poor prognosis and a driver of tumor growth and metastasis, as a benchmark to assess functional relevance. Through this approach, we identified GON4L as a regulator of CD24 from screening a pooled shRNA library of 176 candidate gain-of-function genes. GON4L depletion reduced CD24 expression in human bladder cancer cells and blocked cell proliferation in vitro and tumor xenograft growth in vivo Mechanistically, GON4L interacted with transcription factor YY1, promoting its association with the androgen receptor to drive CD24 expression and cell growth. In clinical bladder cancer specimens, expression of GON4L, YY1, and CD24 was elevated compared with normal bladder urothelium. This pathway is biologically relevant in other cancer types as well, where CD24 and the androgen receptor are clinically prognostic, given that silencing of GON4L and YY1 suppressed CD24 expression and growth of human lung, prostate, and breast cancer cells. Overall, our results define GON4L as a novel driver of cancer growth, offering new biomarker and therapeutic opportunities. Cancer Res; 76(17); 5175-85. ©2016 AACR.

A conserved interaction that is essential for the biogenesis of histone locus bodies

Nuclear protein, ataxia-telangiectasia locus (NPAT) and FLICE-associated huge protein (FLASH) are two major components of discrete nuclear structures called histone locus bodies (HLBs). NPAT is a key co-activator of histone gene transcription, whereas FLASH through its N-terminal region functions in 3' end processing of histone primary transcripts. The C-terminal region of FLASH contains a highly conserved domain that is also present at the end of Yin Yang 1-associated protein-related protein (YARP) and its Drosophila homologue, Mute, previously shown to localize to HLBs in Drosophila cells. Here, we show that the C-terminal domain of human FLASH and YARP interacts with the C-terminal region of NPAT and that this interaction is essential and sufficient to drive FLASH and YARP to HLBs in HeLa cells. Strikingly, only the last 16 amino acids of NPAT are sufficient for the interaction. We also show that the C-terminal domain of Mute interacts with a short region at the end of the Drosophila NPAT orthologue, multi sex combs (Mxc). Altogether, our data indicate that the conserved C-terminal domain shared by FLASH, YARP, and Mute recognizes the C-terminal sequence of NPAT orthologues, thus acting as a signal targeting proteins to HLBs. Finally, we demonstrate that the C-terminal domain of human FLASH can be directly joined with its N-terminal region through alternative splicing. The resulting 190-amino acid MiniFLASH, despite lacking 90% of full-length FLASH, contains all regions necessary for 3' end processing of histone pre-mRNA in vitro and accumulates in HLBs.

Oligogenic germline mutations identified in early non-smokers lung adenocarcinoma patients

OBJECTIVES

A polygenic model is commonly assumed for the predisposition to common cancers. With respect to lung cancer, Genome Wide Association Studies (GWAS) have identified three loci at 15q25, 5p15.33, and 6p21. However, the relative risks associated with alleles at these loci are low; in addition, the data are limited to smokers, and have not been quite reproducible.

MATERIALS AND METHODS

In order to investigate genetic susceptibility we have adopted an entirely novel patient selection strategy. First, we have selected for adenocarcinoma (ADCA) histology only; second, we have selected non-smokers; third we have selected patients who developed ADCA of lung before the age of 60 and who had an older unaffected sib: we have identified 31 such sib-pairs. Among them, we selected two patients with very early age at disease onset (37- and 49-years old), and having a healthy sibling available for genome comparison older than at least 7 years.

RESULTS

On germline DNA samples of four subjects of two such pairs we have carried out whole exome sequencing. Truncating mutations were detected in 8 'cancer genes' in one affected, and in 5 cancer genes in the other affected subject: but none in the two healthy sibs (p=0.0026). Some of these mutant genes (such as BAG6, SPEN and WISP3) are recognized as major cancer players in lung tumors; others have been previously identified in other human cancers (JAK2, TCEB3C, NELFE, TAF1B, EBLN2), in mouse models (GON4L, NOP58, and RBMX) or in genome-wide association studies (KIAA2018, ZNF311).

CONCLUSIONS

This study identifies for the first time in non-smokers with lung adenocarcinoma specific sets of germline mutations that, together, may predispose to this tumor.

Dissecting mammalian immunity through mutation

Although mutation and natural selection have given rise to our immune system, a well-placed mutation can also cripple it, and within an expanding population we are recognizing more and more cases of single-gene mutations that compromise immunity. These mutations are an ideal tool for understanding human immunology, and there are more ways than ever to measure their physiological effects. There are also more ways to create mutations in the laboratory, and to use these resources to systematically define the function of every gene in our genome. This review focuses on the discovery and creation of mutations in the context of mammalian immunity, with an emphasis on the use of genome-wide chemical and CRISPR/Cas9 mutagenesis to reveal gene function.

Separation of plasmacytoid dendritic cells from B-cell-biased lymphoid progenitor (BLP) and Pre-pro B cells using PDCA-1

B-cell-biased lymphoid progenitors (BLPs) and Pre-pro B cells lie at a critical juncture between B cell specification and commitment. However, both of these populations are heterogenous, which hampers investigation into the molecular changes that occur as lymphoid progenitors commit to the B cell lineage. Here, we demonstrate that there are PDCA-1⁺Siglec H⁺ plasmacytoid dendritic cells (pDCs) that co-purify with BLPs and Pre-pro B cells, which express little or no CD11c or Ly6C. Removal of PDCA-1⁺ pDCs separates B cell progenitors that express high levels of a Rag1-GFP reporter from Rag1-GFP^low/neg pDCs within the BLP and Pre-pro B populations. Analysis of Flt3-ligand knockout and IL-7Rα knockout mice revealed that there is a block in B cell development at the all-lymphoid progenitor (ALP) stage, as the majority of cells within the BLP or Pre-pro B gates were PDCA-1⁺ pDCs. Thus, removal of PDCA-1⁺ pDCs is critical for analysis of BLP and Pre-pro B cell populations. Analysis of B cell potential within the B220⁺CD19⁻ fraction demonstrated that AA4.1⁺Ly6D⁺PDCA-1⁻ Pre-pro B cells gave rise to CD19⁺ B cells at high frequency, while PDCA-1⁺ pDCs in this fraction did not. Interestingly, the presence of PDCA-1⁺ pDCs within CLPs may help to explain the conflicting results regarding the origin of these cells.

The Justy mutant mouse strain produces a spontaneous murine model of salivary gland cancer with myoepithelial and basal cell differentiation

We previously identified a novel mutant mouse strain on the C3HeB/FeJ background named Justy. This strain bears a recessive mutation in the Gon4l gene that greatly reduces expression of the encoded protein, a nuclear factor implicated in transcriptional regulation. Here, we report that Justy mutant mice aged 6 months or older spontaneously developed carcinomas with myoepithelial and basaloid differentiation in salivary glands with an incidence of ∼25%. Tumors developed proximate to submandibular glands and to a lesser extent in the sublingual and parotid glands. Histologically, tumors often had central cavitary lesions filled with necrotic debris that were lined by tumor cells, and had spindle and epithelioid cell differentiation with lesser basaloid to clear cell features. Tumor tissue often had variable evidence of a high mitotic rate, pleomorphism, and invasion into adjacent salivary glands. Neoplastic cells had diffuse immunoreactivity for pancytokeratin (AE1/AE3) and p63. Although CK5/6 immunostaining was seen in the much of the tumor cells, it was often lacking in pleomorphic areas. Tumor cells lacked immunoreactivity for alpha-smooth muscle actin, S100, c-Kit, and glial fibrillary acid protein. In addition, tumors had immunoreactivity for phosphorylated and total epidermal growth factor receptor, suggesting that EGFR signaling may participate in growth regulation of these tumors. These findings indicate that the salivary gland carcinomas occur spontaneously in Justy mice, and that these tumors may offer a valuable model for study of EGFR regulation. In combination, our data suggest that Justy mice warrant further investigation for use as a mouse model for human salivary gland neoplasia.

Sedimentation and immunoprecipitation assays for analyzing complexes that repress transcription

Co-repressor proteins function as platforms for the assembly of multi-subunit complexes that mediate transcriptional repression. Common components of such complexes are histone deacetylases, which catalyze the removal of acetyl groups from the tails of histones within nucleosomes, resulting in chromatin compaction and gene repression. In addition, co-repressor complexes generally interact with sequence-specific DNA-binding proteins that direct association with regulatory elements in the genome. Thus, identifying proteins that stably associate with co-repressors can provide insights regarding the biochemical function and target gene specificity of these molecules. Here, we describe a density gradient fractionation method for determining whether a co-repressor is incorporated into high-molecular complexes within cells and for identifying potential constituents of these complexes. We also describe a co-immunoprecipitation assay for confirming and further studying interactions between co-repressors and other proteins that may represent functional partners.

Tim-1 regulates Th2 responses in an airway hypersensitivity model

T-cell immunoglobulin mucin-1 (Tim-1) is a transmembrane protein postulated to be a key regulator of Th2-type immune responses. This hypothesis is based in part upon genetic studies associating Tim-1 polymorphisms in mice with a bias toward airway hyperrespon-siveness (AHR) and the development of Th2-type CD4(+) T cells. Tim-1 expressed by Th2 CD4(+) T cells has been proposed to function as a co-stimulatory molecule. Tim-1 is also expressed by B cells, macrophages, and dendritic cells, but its role in responses by these cell types has not been firmly established. Here, we generated Tim-1-deficient mice to determine the role of Tim-1 in a murine model of allergic airway disease that depends on the development and function of Th2 effector cells and results in the generation of AHR. We found antigen-driven recruitment of inflammatory cells into airways is increased in Tim-1-deficient mice relative to WT mice. In addition, we observed increased antigen-specific cytokine production by splenocytes from antigen-sensitized Tim-1-deficient mice relative to those from controls. These data support the conclusion that Tim-1 functions in pathways that suppress recruitment of inflammatory cells into the airways and the generation or activity of CD4(+) T cells.

B-lymphocyte commitment: identifying the point of no return

Even though B-lymphocyte development is one of the best understood models for cell differentiation in the hematopoetic system, recent advances in cell sorting and functional genomics has increased this understanding further. This has suggested that already early lymphoid primed multipotent progenitor cells (LMPPs) express low levels of lymphoid restricted transcripts. The expression of these genes becomes more pronounced when cells enter the FLT-3/IL-7 receptor positive common lymphoid progenitor (CLP) stage. However, the expression of B-lineage specific genes is limited to a B-cell restricted Ly6D surface positive subpopulation of the CLP compartment. The gene expression patterns also reflect differences in lineage potential and while Ly6D negative FLT-3/IL-7 receptor positive cells represents true CLPs with an ability to generate B/T and NK cells, the Ly6D positive cells lack NK cell potential and display a reduced T-cell potential in vivo. These recent findings suggest that the CLP compartment is highly heterogenous and that the point of no return in B-cell development may occur already in B220(-)CD19(-) cells. These findings have allowed for a better understanding of the interplay between transcription factors like EBF-1, PAX-5 and E47, all known as crucial for normal B-cell development. In this review, we aim to provide a comprehensive overview of B-cell fate specification and commitment based on the recent advances in the understanding of molecular networks as well as functional properties of early progenitor populations.

The developmental regulator protein Gon4l associates with protein YY1, co-repressor Sin3a, and histone deacetylase 1 and mediates transcriptional repression

Genetic studies involving zebrafish and mice have demonstrated that the protein Gon4l (Gon4-like) is essential for hematopoiesis. These studies also suggested that Gon4l regulates gene expression during hematopoietic development, yet the biochemical function of Gon4l has not been defined. Here, we describe the identification of factors that interact with Gon4l and may cooperate with this protein to regulate gene expression. As predicted by polypeptide sequence conservation, Gon4l interacted and co-localized with the DNA-binding protein YY1 (Yin Yang 1). Density gradient sedimentation analysis of protein lysates from mouse M12 B cells showed that Gon4l and YY1 co-sediment with the transcriptional co-repressor Sin3a and its functional partner histone deacetylase (HDAC) 1. Consistent with these results, immunoprecipitation studies showed that Gon4l associates with Sin3a, HDAC1, and YY1 as a part of complexes that form in M12 cells. Sequential immunoprecipitation studies demonstrated that Gon4l, YY1, Sin3a, and HDAC1 could all associate as components of a single complex and that a conserved domain spanning the central portion of Gon4l was required for formation of this complex. When targeted to DNA, Gon4l repressed the activity of a nearby promoter, which correlated with the ability to interact with Sin3a and HDAC1. Our data suggest that Sin3a, HDAC1, and YY1 are co-factors for Gon4l and that Gon4l may function as a platform for the assembly of complexes that regulate gene expression.

Causal modeling using network ensemble simulations of genetic and gene expression data predicts genes involved in rheumatoid arthritis

Tumor necrosis factor α (TNF-α) is a key regulator of inflammation and rheumatoid arthritis (RA). TNF-α blocker therapies can be very effective for a substantial number of patients, but fail to work in one third of patients who show no or minimal response. It is therefore necessary to discover new molecular intervention points involved in TNF-α blocker treatment of rheumatoid arthritis patients. We describe a data analysis strategy for predicting gene expression measures that are critical for rheumatoid arthritis using a combination of comprehensive genotyping, whole blood gene expression profiles and the component clinical measures of the arthritis Disease Activity Score 28 (DAS28) score. Two separate network ensembles, each comprised of 1024 networks, were built from molecular measures from subjects before and 14 weeks after treatment with TNF-α blocker. The network ensemble built from pre-treated data captures TNF-α dependent mechanistic information, while the ensemble built from data collected under TNF-α blocker treatment captures TNF-α independent mechanisms. In silico simulations of targeted, personalized perturbations of gene expression measures from both network ensembles identify transcripts in three broad categories. Firstly, 22 transcripts are identified to have new roles in modulating the DAS28 score; secondly, there are 6 transcripts that could be alternative targets to TNF-α blocker therapies, including CD86 - a component of the signaling axis targeted by Abatacept (CTLA4-Ig), and finally, 59 transcripts that are predicted to modulate the count of tender or swollen joints but not sufficiently enough to have a significant impact on DAS28.

The collection and analysis of clinical data has played a key role in providing insights into the diagnosis, prognosis and treatment of disease. However, it is imperative that molecular and genetic data also be collected and integrated into the creation of network models, which capture underlying mechanisms of disease and can be interrogated to elucidate previously unknown biology. Bringing data from the clinic to the bench completes the cycle of translational research, which we demonstrate with this work. We built disease models from genetics, whole blood gene expression profiles and the component clinical measures of rheumatoid arthritis using a data-driven approach that leverages supercomputing. Genetic factors can be utilized as a source of perturbation to the system such that causal connections between genetics, molecular entities and clinical outcomes can be inferred. The existing TNF-α blocker treatments for rheumatoid arthritis are only effective for approximately 2/3 of the affected population. We identified novel therapeutic intervention points that may lead to the development of alternatives to TNF-α blocker treatments. We believe this approach will provide improved drug discovery programs, new insights into disease progression, increased drug efficacy and novel biomarkers for chronic and complex diseases.