DNA mismatches reveal conformational penalties in protein-DNA recognition.

Transcription factors recognize specific genomic sequences to regulate complex gene-expression programs. Although it is well-established that transcription factors bind to specific DNA sequences using a combination of base readout and shape recognition, some fundamental aspects of protein-DNA binding remain poorly understood1,2. Many DNA-binding proteins induce changes in the structure of the DNA outside the intrinsic B-DNA envelope. However, how the energetic cost that is associated with distorting the DNA contributes to recognition has proven difficult to study, because the distorted DNA exists in low abundance in the unbound ensemble3-9. Here we use a high-throughput assay that we term SaMBA (saturation mismatch-binding assay) to investigate the role of DNA conformational penalties in transcription factor-DNA recognition. In SaMBA, mismatched base pairs are introduced to pre-induce structural distortions in the DNA that are much larger than those induced by changes in the Watson-Crick sequence. Notably, approximately 10% of mismatches increased transcription factor binding, and for each of the 22 transcription factors that were examined, at least one mismatch was found that increased the binding affinity. Mismatches also converted non-specific sites into high-affinity sites, and high-affinity sites into 'super sites' that exhibit stronger affinity than any known canonical binding site. Determination of high-resolution X-ray structures, combined with nuclear magnetic resonance measurements and structural analyses, showed that many of the DNA mismatches that increase binding induce distortions that are similar to those induced by protein binding-thus prepaying some of the energetic cost incurred from deforming the DNA. Our work indicates that conformational penalties are a major determinant of protein-DNA recognition, and reveals mechanisms by which mismatches can recruit transcription factors and thus modulate replication and repair activities in the cell10,11.

DFF-ChIP: a method to detect and quantify complex interactions between RNA polymerase II, transcription factors, and chromatin.

Recently, we introduced a chromatin immunoprecipitation (ChIP) technique utilizing the human DNA Fragmentation Factor (DFF) to digest the DNA prior to immunoprecipitation (DFF-ChIP) that provides the precise location of transcription complexes and their interactions with neighboring nucleosomes. Here we expand the technique to new targets and provide useful information concerning purification of DFF, digestion conditions, and the impact of crosslinking. DFF-ChIP analysis was performed individually for subunits of Mediator, DSIF, and NELF that that do not interact with DNA directly, but rather interact with RNA polymerase II (Pol II). We found that Mediator was associated almost exclusively with preinitiation complexes (PICs). DSIF and NELF were associated with engaged Pol II and, in addition, potential intermediates between PICs and early initiation complexes. DFF-ChIP was then used to analyze the occupancy of a tight binding transcription factor, CTCF, and a much weaker binding factor, glucocorticoid receptor (GR), with and without crosslinking. These results were compared to those from standard ChIP-Seq that employs sonication and to CUT&RUN which utilizes MNase to fragment the genomic DNA. Our findings indicate that DFF-ChIP reveals details of occupancy that are not available using other methods including information revealing pertinent protein:protein interactions.

Author Correction: Metabolic gatekeeper function of B-lymphoid transcription factors.

In Fig. 3c of this Letter, the the effects of CRISPR-Cas9-mediated deletion of NR3C1, TXNIP and CNR2 in patient-derived B-lineage leukaemia cells were shown. For curves depicting NR3C1 (left graph), data s for TXNIP (middle graph) were inadvertently plotted. This figure has been corrected online, and the original Fig. 3c is shown as Supplementary Information to this Amendment for transparency. The error does not affect the conclusions of the Letter. In addition, Source Data files have been added for the Figs. 1-4 and Extended Data Figs. 1-10 of the original Letter.

IL6Myc mouse is an immunocompetent model for the development of aggressive multiple myeloma.

Multiple Myeloma (MM) is a plasma cell neoplasm originating in the bone marrow and is the second most common blood cancer in the United States. One challenge in understanding the pathogenesis of MM and improving treatment is a lack of immunocompetent mouse models. We previously developed the IL6Myc mouse that generates plasmacytomas at 100% penetrance that phenotypically resemble aggressive MM. Using comprehensive genomic analysis, we found that the IL6Myc tumors resemble aggressive MM by RNA and protein expression. We also found that IL6Myc tumors accumulated fusions and missense mutations in genes that overlap significantly with human myeloma, indicating that the mouse is good model for studying disease etiology. Lastly, we derived cell lines from IL6Myc tumors that express cell surface markers typical of MM and readily engraft into mice, home to the bone marrow, and induce osteolytic disease. The cell lines may be useful in developing immunotherapies directed against BAFF-R and TACI, though not BCMA, and may also be a good model for studying dexamethasone resistance. These data indicate that the IL6Myc model is useful for studying development of aggressive MM and for developing new treatments against such forms of the disease.

Metabolic gatekeeper function of B-lymphoid transcription factors.

B-lymphoid transcription factors, such as PAX5 and IKZF1, are critical for early B-cell development, yet lesions of the genes encoding these transcription factors occur in over 80% of cases of pre-B-cell acute lymphoblastic leukaemia (ALL). The importance of these lesions in ALL has, until now, remained unclear. Here, by combining studies using chromatin immunoprecipitation with sequencing and RNA sequencing, we identify a novel B-lymphoid program for transcriptional repression of glucose and energy supply. Our metabolic analyses revealed that PAX5 and IKZF1 enforce a state of chronic energy deprivation, resulting in constitutive activation of the energy-stress sensor AMPK. Dominant-negative mutants of PAX5 and IKZF1, however, relieved this glucose and energy restriction. In a transgenic pre-B ALL mouse model, the heterozygous deletion of Pax5 increased glucose uptake and ATP levels by more than 25-fold. Reconstitution of PAX5 and IKZF1 in samples from patients with pre-B ALL restored a non-permissive state and induced energy crisis and cell death. A CRISPR/Cas9-based screen of PAX5 and IKZF1 transcriptional targets identified the products of NR3C1 (encoding the glucocorticoid receptor), TXNIP (encoding a glucose-feedback sensor) and CNR2 (encoding a cannabinoid receptor) as central effectors of B-lymphoid restriction of glucose and energy supply. Notably, transport-independent lipophilic methyl-conjugates of pyruvate and tricarboxylic acid cycle metabolites bypassed the gatekeeper function of PAX5 and IKZF1 and readily enabled leukaemic transformation. Conversely, pharmacological TXNIP and CNR2 agonists and a small-molecule AMPK inhibitor strongly synergized with glucocorticoids, identifying TXNIP, CNR2 and AMPK as potential therapeutic targets. Furthermore, our results provide a mechanistic explanation for the empirical finding that glucocorticoids are effective in the treatment of B-lymphoid but not myeloid malignancies. Thus, B-lymphoid transcription factors function as metabolic gatekeepers by limiting the amount of cellular ATP to levels that are insufficient for malignant transformation.

Nascent transcript analysis of glucocorticoid crosstalk with TNF defines primary and cooperative inflammatory repression.

The glucocorticoid receptor (NR3C1, also known as GR) binds to specific DNA sequences and directly induces transcription of anti-inflammatory genes that contribute to cytokine repression, frequently in cooperation with NF-kB. Whether inflammatory repression also occurs through local interactions between GR and inflammatory gene regulatory elements has been controversial. Here, using global run-on sequencing (GRO-seq) in human airway epithelial cells, we show that glucocorticoid signaling represses transcription within 10 min. Many repressed regulatory regions reside within "hyper-ChIPable" genomic regions that are subject to dynamic, yet nonspecific, interactions with some antibodies. When this artifact was accounted for, we determined that transcriptional repression does not require local GR occupancy. Instead, widespread transcriptional induction through canonical GR binding sites is associated with reciprocal repression of distal TNF-regulated enhancers through a chromatin-dependent process, as evidenced by chromatin accessibility and motif displacement analysis. Simultaneously, transcriptional induction of key anti-inflammatory effectors is decoupled from primary repression through cooperation between GR and NF-kB at a subset of regulatory regions. Thus, glucocorticoids exert bimodal restraints on inflammation characterized by rapid primary transcriptional repression without local GR occupancy and secondary anti-inflammatory effects resulting from transcriptional cooperation between GR and NF-kB.

Systematic in vitro profiling of off-target affinity, cleavage and efficiency for CRISPR enzymes.

CRISPR RNA-guided endonucleases (RGEs) cut or direct activities to specific genomic loci, yet each has off-target activities that are often unpredictable. We developed a pair of simple in vitro assays to systematically measure the DNA-binding specificity (Spec-seq), catalytic activity specificity (SEAM-seq) and cleavage efficiency of RGEs. By separately quantifying binding and cleavage specificity, Spec/SEAM-seq provides detailed mechanistic insight into off-target activity. Feature-based models generated from Spec/SEAM-seq data for SpCas9 were consistent with previous reports of its in vitro and in vivo specificity, validating the approach. Spec/SEAM-seq is also useful for profiling less-well characterized RGEs. Application to an engineered SpCas9, HiFi-SpCas9, indicated that its enhanced target discrimination can be attributed to cleavage rather than binding specificity. The ortholog ScCas9, on the other hand, derives specificity from binding to an extended PAM. The decreased off-target activity of AsCas12a (Cpf1) appears to be primarily driven by DNA-binding specificity. Finally, we performed the first characterization of CasX specificity, revealing an all-or-nothing mechanism where mismatches can be bound, but not cleaved. Together, these applications establish Spec/SEAM-seq as an accessible method to rapidly and reliably evaluate the specificity of RGEs, Cas::gRNA pairs, and gain insight into the mechanism and thermodynamics of target discrimination.

Relapse-associated AURKB blunts the glucocorticoid sensitivity of B cell acute lymphoblastic leukemia.

Glucocorticoids (GCs) are used in combination chemotherapies as front-line treatment for B cell acute lymphoblastic leukemia (B-ALL). Although effective, many patients relapse and become resistant to chemotherapy and GCs in particular. Why these patients relapse is not clear. We took a comprehensive, functional genomics approach to identify sources of GC resistance. A genome-wide shRNA screen identified the transcriptional coactivators EHMT2, EHMT1, and CBX3 as important contributors to GC-induced cell death. This complex selectively supports GC-induced expression of genes contributing to cell death. A metaanalysis of gene expression data from B-ALL patient specimens revealed that Aurora kinase B (AURKB), which restrains GC signaling by phosphorylating EHMT1-2, is overexpressed in relapsed B-ALL, suggesting it as a potential contributor to relapse. Inhibition of AURKB enhanced GC-induced expression of cell death genes, resulting in potentiation of GC cytotoxicity in cell lines and relapsed B-ALL patient samples. This function for AURKB is distinct from its canonical role in the cell cycle. These results show the utility of functional genomics in understanding mechanisms of resistance and rapidly identifying combination chemotherapeutics.

SelexGLM differentiates androgen and glucocorticoid receptor DNA-binding preference over an extended binding site.

The DNA-binding interfaces of the androgen (AR) and glucocorticoid (GR) receptors are virtually identical, yet these transcription factors share only about a third of their genomic binding sites and regulate similarly distinct sets of target genes. To address this paradox, we determined the intrinsic specificities of the AR and GR DNA-binding domains using a refined version of SELEX-seq. We developed an algorithm, SelexGLM, that quantifies binding specificity over a large (31-bp) binding site by iteratively fitting a feature-based generalized linear model to SELEX probe counts. This analysis revealed that the DNA-binding preferences of AR and GR homodimers differ significantly, both within and outside the 15-bp core binding site. The relative preference between the two factors can be tuned over a wide range by changing the DNA sequence, with AR more sensitive to sequence changes than GR. The specificity of AR extends to the regions flanking the core 15-bp site, where isothermal calorimetry measurements reveal that affinity is augmented by enthalpy-driven readout of poly(A) sequences associated with narrowed minor groove width. We conclude that the increased specificity of AR is correlated with more enthalpy-driven binding than GR. The binding models help explain differences in AR and GR genomic binding and provide a biophysical rationale for how promiscuous binding by GR allows functional substitution for AR in some castration-resistant prostate cancers.

Suppression of B-cell development genes is key to glucocorticoid efficacy in treatment of acute lymphoblastic leukemia.

Glucocorticoids (GCs), including dexamethasone (dex), are a central component of combination chemotherapy for childhood B-cell precursor acute lymphoblastic leukemia (B-ALL). GCs work by activating the GC receptor (GR), a ligand-induced transcription factor, which in turn regulates genes that induce leukemic cell death. Which GR-regulated genes are required for GC cytotoxicity, which pathways affect their regulation, and how resistance arises are not well understood. Here, we systematically integrate the transcriptional response of B-ALL to GCs with a next-generation short hairpin RNA screen to identify GC-regulated "effector" genes that contribute to cell death, as well as genes that affect the sensitivity of B-ALL cells to dex. This analysis reveals a pervasive role for GCs in suppression of B-cell development genes that is linked to therapeutic response. Inhibition of phosphatidylinositol 3-kinase δ (PI3Kδ), a linchpin in the pre-B-cell receptor and interleukin 7 receptor signaling pathways critical to B-cell development (with CAL-101 [idelalisib]), interrupts a double-negative feedback loop, enhancing GC-regulated transcription to synergistically kill even highly resistant B-ALL with diverse genetic backgrounds. This work not only identifies numerous opportunities for enhanced lymphoid-specific combination chemotherapies that have the potential to overcome treatment resistance, but is also a valuable resource for understanding GC biology and the mechanistic details of GR-regulated transcription.

JUNB is a key transcriptional modulator of macrophage activation.

Activated macrophages are crucial for restriction of microbial infection but may also promote inflammatory pathology in a wide range of both infectious and sterile conditions. The pathways that regulate macrophage activation are therefore of great interest. Recent studies in silico have putatively identified key transcription factors that may control macrophage activation, but experimental validation is lacking. In this study, we generated a macrophage regulatory network from publicly available microarray data, employing steps to enrich for physiologically relevant interactions. Our analysis predicted a novel relationship between the AP-1 family transcription factor Junb and the gene Il1b, encoding the pyrogen IL-1ß, which macrophages express upon activation by inflammatory stimuli. Previously, Junb has been characterized primarily as a negative regulator of the cell cycle, whereas AP-1 activity in myeloid inflammatory responses has largely been attributed to c-Jun. We confirmed experimentally that Junb is required for full expression of Il1b, and of additional genes involved in classical inflammation, in macrophages treated with LPS and other immunostimulatory molecules. Furthermore, Junb modulates expression of canonical markers of alternative activation in macrophages treated with IL-4. Our results demonstrate that JUNB is a significant modulator of both classical and alternative macrophage activation. Further, this finding provides experimental validation for our network modeling approach, which will facilitate the future use of gene expression data from open databases to reveal novel, physiologically relevant regulatory relationships.

Response Element Composition Governs Correlations between Binding Site Affinity and Transcription in Glucocorticoid Receptor Feed-forward Loops.

Combinatorial gene regulation through feed-forward loops (FFLs) can bestow specificity and temporal control to client gene expression; however, characteristics of binding sites that mediate these effects are not established. We previously showed that the glucocorticoid receptor (GR) and KLF15 form coherent FFLs that cooperatively induce targets such as the amino acid-metabolizing enzymes AASS and PRODH and incoherent FFLs exemplified by repression of MT2A by KLF15. Here, we demonstrate that GR and KLF15 physically interact and identify low affinity GR binding sites within glucocorticoid response elements (GREs) for PRODH and AASS that contribute to combinatorial regulation with KLF15. We used deep sequencing and electrophoretic mobility shift assays to derive in vitro GR binding affinities across sequence space. We applied these data to show that AASS GRE activity correlated (r(2) = 0.73) with predicted GR binding affinities across a 50-fold affinity range in transfection assays; however, the slope of the linear relationship more than doubled when KLF15 was expressed. Whereas activity of the MT2A GRE was even more strongly (r(2) = 0.89) correlated with GR binding site affinity, the slope of the linear relationship was sharply reduced by KLF15, consistent with incoherent FFL logic. Thus, GRE architecture and co-regulator expression together determine the functional parameters that relate GR binding site affinity to hormone-induced transcriptional responses. Utilization of specific affinity response functions and GR binding sites by FFLs may contribute to the diversity of gene expression patterns within GR-regulated transcriptomes.

A naturally occurring insertion of a single amino acid rewires transcriptional regulation by glucocorticoid receptor isoforms.

In addition to guiding proteins to defined genomic loci, DNA can act as an allosteric ligand that influences protein structure and activity. Here we compared genome-wide binding, transcriptional regulation, and, using NMR, the conformation of two glucocorticoid receptor (GR) isoforms that differ by a single amino acid insertion in the lever arm, a domain that adopts DNA sequence-specific conformations. We show that these isoforms differentially regulate gene expression levels through two mechanisms: differential DNA binding and altered communication between GR domains. Our studies suggest a versatile role for DNA in both modulating GR activity and also in directing the use of GR isoforms. We propose that the lever arm is a "fulcrum" for bidirectional allosteric signaling, conferring conformational changes in the DNA reading head that influence DNA sequence selectivity, as well as conferring changes in the dimerization domain that connect functionally with remote regulatory surfaces, thereby influencing which genes are regulated and the magnitude of their regulation.

Spermine oxidase maintains basal skeletal muscle gene expression and fiber size and is strongly repressed by conditions that cause skeletal muscle atrophy.

Skeletal muscle atrophy is a common and debilitating condition that remains poorly understood at the molecular level. To better understand the mechanisms of muscle atrophy, we used mouse models to search for a skeletal muscle protein that helps to maintain muscle mass and is specifically lost during muscle atrophy. We discovered that diverse causes of muscle atrophy (limb immobilization, fasting, muscle denervation, and aging) strongly reduced expression of the enzyme spermine oxidase. Importantly, a reduction in spermine oxidase was sufficient to induce muscle fiber atrophy. Conversely, forced expression of spermine oxidase increased muscle fiber size in multiple models of muscle atrophy (immobilization, fasting, and denervation). Interestingly, the reduction of spermine oxidase during muscle atrophy was mediated by p21, a protein that is highly induced during muscle atrophy and actively promotes muscle atrophy. In addition, we found that spermine oxidase decreased skeletal muscle mRNAs that promote muscle atrophy (e.g., myogenin) and increased mRNAs that help to maintain muscle mass (e.g., mitofusin-2). Thus, in healthy skeletal muscle, a relatively low level of p21 permits expression of spermine oxidase, which helps to maintain basal muscle gene expression and fiber size; conversely, during conditions that cause muscle atrophy, p21 expression rises, leading to reduced spermine oxidase expression, disruption of basal muscle gene expression, and muscle fiber atrophy. Collectively, these results identify spermine oxidase as an important positive regulator of muscle gene expression and fiber size, and elucidate p21-mediated repression of spermine oxidase as a key step in the pathogenesis of skeletal muscle atrophy.

Glucocorticoids and Cancer.

Unlike other steroid hormone receptors, the glucocorticoid receptor (GR) is not considered an oncogene. In breast cancer, the estrogen receptor (ER) drives cell growth, proliferation, and metastasis, and the androgen receptor (AR) plays a similar role in prostate cancer. Accordingly, treatment of these diseases has focused on blocking steroid hormone receptor function. In contrast, glucocorticoids (GCs) work through GR to arrest growth and induce apoptosis in lymphoid tissue. Glucocorticoids are amazingly effective in this role, and have been deployed as the cornerstone of lymphoid cancer treatment for decades. Unfortunately, not all patients respond to GCs and dosage is restricted by immediate and long term side effects. In this chapter we review the treatment protocols that employ glucocorticoids as a curative agent, elaborate on what is known about their mechanism of action in these cancers, and also summarize the palliative uses of glucocorticoids for other cancers.

Reversible targeting of noncatalytic cysteines with chemically tuned electrophiles.

Targeting noncatalytic cysteine residues with irreversible acrylamide-based inhibitors is a powerful approach for enhancing pharmacological potency and selectivity. Nevertheless, concerns about off-target modification motivate the development of reversible cysteine-targeting strategies. Here we show that electron-deficient olefins, including acrylamides, can be tuned to react with cysteine thiols in a rapidly reversible manner. Installation of a nitrile group increased the olefins' intrinsic reactivity, but, paradoxically, eliminated the formation of irreversible adducts. Incorporation of these electrophiles into a noncovalent kinase-recognition scaffold produced slowly dissociating, covalent inhibitors of the p90 ribosomal protein S6 kinase RSK2. A cocrystal structure revealed specific noncovalent interactions that stabilize the complex by positioning the electrophilic carbon near the targeted cysteine. Disruption of these interactions by protein unfolding or proteolysis promoted instantaneous cleavage of the covalent bond. Our results establish a chemistry-based framework for engineering sustained covalent inhibition without accumulating permanently modified proteins and peptides.

PI3Kδ inhibition potentiates glucocorticoids in B-lymphoblastic leukemia by decreasing receptor phosphorylation and enhancing gene regulation.

Glucocorticoids, including dexamethasone and prednisone, are the cornerstone of B-lymphoblastic leukemia (B-ALL) therapy. Because response to glucocorticoids alone predicts overall outcomes for B-ALL, enhancing glucocorticoid potency is a route to improving outcomes. However, systematic toxicities prevent the use of higher dose and more potent glucocorticoids. We therefore took a functional genomic approach to identify targets to enhance glucocorticoid activity specifically in B-ALL cells. Here we show that inhibition of the lymphoid-restricted PI3Kδ, signaling through the RAS/MAPK pathway, enhances both prednisone and dexamethasone activity in almost all ex vivo B-ALL specimens tested. This potentiation is most synergistic at sub-saturating doses of glucocorticoids, approaching the EC50. Potentiation correlates with global enhancement of glucocorticoid-induced gene regulation, including regulation of effector genes that drive B-ALL cell death. Idelalisib reduces phosphorylation of the glucocorticoid receptor (GR) at MAPK1/ERK2 targets S203 and S226, and ablation of these phospho-acceptor sites enhances glucocorticoid potency. We further show that phosphorylation of S226 reduces the affinity of GR for DNA in vitro, which impairs DNA binding. We therefore propose that PI3Kδ inhibition improves glucocorticoid efficacy in B-ALL in part by decreasing GR phosphorylation, increasing DNA binding affinity, and enhancing downstream gene regulation. The overall enhancement of GR function suggests that idelalisib will provide benefit to most patients with B-ALL by improving outcomes for patients whose disease is less responsive to glucocorticoid-based therapy, including high-risk disease, and allowing less toxic glucocorticoid-sparing strategies for patients with standard-risk disease.

PI3Kδ Inhibition Potentiates Glucocorticoids in B-lymphoblastic Leukemia by Decreasing Receptor Phosphorylation and Enhancing Gene Regulation.

Glucocorticoids are the cornerstone of B-lymphoblastic leukemia (B-ALL) therapy. Because response to glucocorticoids alone predicts overall outcomes for B-ALL, enhancing glucocorticoid potency should improve treatment. We previously showed that inhibition of the lymphoid-restricted PI3Kδ with idelalisib enhances glucocorticoid activity in B-ALL cells. Here, we show that idelalisib enhances glucocorticoid potency in 90% of primary B-ALL specimens and is most pronounced at sub-saturating doses of glucocorticoids near the EC50. Potentiation is associated with enhanced regulation of all glucocorticoid-regulated genes, including genes that drive B-ALL cell death. Idelalisib reduces phosphorylation of the glucocorticoid receptor (GR) at PI3Kδ/MAPK1 (ERK2) targets S203 and S226. Ablation of these phospho-acceptor sites enhances sensitivity to glucocorticoids with ablation of S226 in particular reducing synergy. We also show that phosphorylation of S226 reduces the affinity of GR for DNA in vitro. We propose that PI3Kδ inhibition improves glucocorticoid efficacy in B-ALL in part by decreasing GR phosphorylation, increasing DNA binding affinity, and enhancing downstream gene regulation. This mechanism and the response of patient specimens suggest that idelalisib will benefit most patients with B-ALL, but particularly patients with less responsive, including high-risk, disease. This combination is also promising for the development of less toxic glucocorticoid-sparing therapies.

De novo distillation of thermodynamic affinity from deep learning regulatory sequence models of in vivo protein-DNA binding.

Transcription factors (TF) are proteins that bind DNA in a sequence-specific manner to regulate gene transcription. Despite their unique intrinsic sequence preferences, in vivo genomic occupancy profiles of TFs differ across cellular contexts. Hence, deciphering the sequence determinants of TF binding, both intrinsic and context-specific, is essential to understand gene regulation and the impact of regulatory, non-coding genetic variation. Biophysical models trained on in vitro TF binding assays can estimate intrinsic affinity landscapes and predict occupancy based on TF concentration and affinity. However, these models cannot adequately explain context-specific, in vivo binding profiles. Conversely, deep learning models, trained on in vivo TF binding assays, effectively predict and explain genomic occupancy profiles as a function of complex regulatory sequence syntax, albeit without a clear biophysical interpretation. To reconcile these complementary models of in vitro and in vivo TF binding, we developed Affinity Distillation (AD), a method that extracts thermodynamic affinities de-novo from deep learning models of TF chromatin immunoprecipitation (ChIP) experiments by marginalizing away the influence of genomic sequence context. Applied to neural networks modeling diverse classes of yeast and mammalian TFs, AD predicts energetic impacts of sequence variation within and surrounding motifs on TF binding as measured by diverse in vitro assays with superior dynamic range and accuracy compared to motif-based methods. Furthermore, AD can accurately discern affinities of TF paralogs. Our results highlight thermodynamic affinity as a key determinant of in vivo binding, suggest that deep learning models of in vivo binding implicitly learn high-resolution affinity landscapes, and show that these affinities can be successfully distilled using AD. This new biophysical interpretation of deep learning models enables high-throughput in silico experiments to explore the influence of sequence context and variation on both intrinsic affinity and in vivo occupancy.

Transcriptional coactivation by EHMT2 restricts glucocorticoid-induced insulin resistance in a study with male mice.

The classical dogma of glucocorticoid-induced insulin resistance is that it is caused by the transcriptional activation of hepatic gluconeogenic and insulin resistance genes by the glucocorticoid receptor (GR). Here, we find that glucocorticoids also stimulate the expression of insulin-sensitizing genes, such as Irs2. The transcriptional coregulator EHMT2 can serve as a transcriptional coactivator or a corepressor. Using male mice that have a defective EHMT2 coactivation function specifically, we show that glucocorticoid-induced Irs2 transcription is dependent on liver EHMT2's coactivation function and that IRS2 play a key role in mediating the limitation of glucocorticoid-induced insulin resistance by EHMT2's coactivation. Overall, we propose a model in which glucocorticoid-regulated insulin sensitivity is determined by the balance between glucocorticoid-modulated insulin resistance and insulin sensitizing genes, in which EHMT2 coactivation is specifically involved in the latter process.