A T-cell engaging bispecific antibody with a tumor-selective bivalent folate receptor alpha binding arm for the treatment of ovarian cancer

The use of T-cell engagers (TCEs) to treat solid tumors is challenging, and several have been limited by narrow therapeutic windows due to substantial on-target, off-tumor toxicities due to the expression of low levels of target antigens on healthy tissues. Here, we describe TNB-928B, a fully human TCE that has a bivalent binding arm for folate receptor alpha (FRα) to selectively target FRα overexpressing tumor cells while avoiding the lysis of cells with low levels of FRα expression. The bivalent design of the FRα binding arm confers tumor selectivity due to low-affinity but high-avidity binding to high FRα antigen density cells. TNB-928B induces preferential effector T-cell activation, proliferation, and selective cytotoxic activity on high FRα expressing cells while sparing low FRα expressing cells. In addition, TNB-928B induces minimal cytokine release compared to a positive control TCE containing OKT3. Moreover, TNB-928B exhibits substantial ex vivo tumor cell lysis using endogenous T-cells and robust tumor clearance in vivo, promoting T-cell infiltration and antitumor activity in mouse models of ovarian cancer. TNB-928B exhibits pharmacokinetics similar to conventional antibodies, which are projected to enable favorable administration in humans. TNB-928B is a novel TCE with enhanced safety and specificity for the treatment of ovarian cancer.

Abstract 6244: A T cell engaging bispecific antibody utilizing a bivalent tumor-selective folate receptor alpha binding arm for the treatment of ovarian cancer

The use of T cell engagers (TCEs) to treat solid tumors in humans is challenging, and several have been limited by narrow therapeutic windows due to substantial toxicity. Dose-limiting toxicities of TCEs include cytokine release syndrome (CRS) and on-target, off-tumor toxicities since many solid tumor target antigens are also expressed in healthy tissues, albeit at lower levels. While there have been improvements in the clinical management of immune-mediated toxicities such as CRS, on-target, off-tumor effects remain one of the primary challenges of using TCEs for treatment of solid tumors. Here, we describe TNB-928B, a novel fully human TCE that is bivalent for folate receptor alpha (FRα) that selectively targets FRα overexpressing tumor cells yet avoids targeting cells that express low levels of FRα found in normal human cells. Selectivity is conferred by an avidity effect of the FRα targeting arm in combination with a unique low-affinity CD3 arm and is anticipated to reduce the risk of adverse events in normal cells expressing low levels of FRα. TNB-928B induces preferential effector T cell activation, proliferation, and cytotoxic activity with minimal cytokine release, compared to a positive control containing OKT3, thereby improving the therapeutic index. Moreover, TNB-928B exhibits substantial ex vivo tumor cell lysis using endogenous T cells and is effective in vivo promoting T cell infiltration and tumor regression in mouse models of ovarian cancer. Together, these results demonstrate TNB-928B represents a novel next generation immunotherapy modality with enhanced safety and specificity for the treatment of ovarian cancer.

Citation Format: Brian C. Avanzino, Kirthana Prabhakar, Pranjali Dalvi, Sharon Hartstein, Hannes Kehm, Aarti Balasubramani, Andrew A. Boudreau, Ben Buelow, Karen Chang, Laura M. Davison, Suhasini Iyer, Vidyut Kalwit, Harbani Malik-Chaudhry, Will Pierson, Geovanni Pineda, Udaya S. Rangaswamy, Sowmya Saiganesh, Ute Schellenberger, Harshad S. Ugamraj, Roland Buelow, Jocelyn Chapman, Nathan D. Trinklein, Katherine E. Harris. A T cell engaging bispecific antibody utilizing a bivalent tumor-selective folate receptor alpha binding arm for the treatment of ovarian cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 6244.

Attenuating CD3 affinity in a PSMAxCD3 bispecific antibody enables killing of prostate tumor cells with reduced cytokine release

BACKGROUND

Therapeutic options currently available for metastatic castration-resistant prostate cancer (mCRPC) do not extend median overall survival >6 months. Therefore, the development of novel and effective therapies for mCRPC represents an urgent medical need. T cell engagers (TCEs) have emerged as a promising approach for the treatment of mCRPC due to their targeted mechanism of action. However, challenges remain in the clinic due to the limited efficacy of TCEs observed thus far in solid tumors as well as the toxicities associated with cytokine release syndrome (CRS) due to the usage of high-affinity anti-CD3 moieties such as OKT3.

METHODS

Using genetically engineered transgenic rats (UniRat and OmniFlic) that express fully human IgG antibodies together with an NGS-based antibody discovery pipeline, we developed TNB-585, an anti-CD3xPSMA TCE for the treatment of mCRPC. TNB-585 pairs a tumor-targeting anti-PSMA arm together with a unique, low-affinity anti-CD3 arm in bispecific format. We tested TNB-585 in T cell-redirected cytotoxicity assays against PSMA+ tumor cells in both two-dimensional (2D) cultures and three-dimensional (3D) spheroids as well as against patient-derived prostate tumor cells. Cytokines were measured in culture supernatants to assess the ability of TNB-585 to induce tumor killing with low cytokine release. TNB-585-mediated T cell activation, proliferation, and cytotoxic granule formation were measured to investigate the mechanism of action. Additionally, TNB-585 efficacy was evaluated in vivo against C4-2 tumor-bearing NCG mice.

RESULTS

In vitro, TNB-585 induced activation and proliferation of human T cells resulting in the killing of PSMA+ prostate tumor cells in both 2D cultures and 3D spheroids with minimal cytokine release and reduced regulatory T cell activation compared with a positive control antibody that contains the same anti-PSMA arm but a higher affinity anti-CD3 arm (comparable with OKT3). In addition, TNB-585 demonstrated potent efficacy against patient-derived prostate tumors ex vivo and induced immune cell infiltration and dose-dependent tumor regression in vivo.

CONCLUSIONS

Our data suggest that TNB-585, with its low-affinity anti-CD3, may be efficacious while inducing a lower incidence and severity of CRS in patients with prostate cancer compared with TCEs that incorporate high-affinity anti-CD3 domains.

A bispecific antibody agonist of the IL-2 heterodimeric receptor preferentially promotes in vivo expansion of CD8 and NK cells

The use of recombinant interleukin-2 (IL-2) as a therapeutic protein has been limited by significant toxicities despite its demonstrated ability to induce durable tumor-regression in cancer patients. The adverse events and limited efficacy of IL-2 treatment are due to the preferential binding of IL-2 to cells that express the high-affinity, trimeric receptor, IL-2Rαβγ such as endothelial cells and T-regulatory cells, respectively. Here, we describe a novel bispecific heavy-chain only antibody which binds to and activates signaling through the heterodimeric IL-2Rβγ receptor complex that is expressed on resting T-cells and NK cells. By avoiding binding to IL-2Rα, this molecule circumvents the preferential T-reg activation of native IL-2, while maintaining the robust stimulatory effects on T-cells and NK-cells in vitro. In vivo studies in both mice and cynomolgus monkeys confirm the molecule's in vivo biological activity, extended pharmacodynamics due to the Fc portion of the molecule, and enhanced safety profile. Together, these results demonstrate that the bispecific antibody is a safe and effective IL-2R agonist that harnesses the benefits of the IL-2 signaling pathway as a potential anti-cancer therapy.

TNB-486 induces potent tumor cell cytotoxicity coupled with low cytokine release in preclinical models of B-NHL

The therapeutic potential of targeting CD19 in B cell malignancies has garnered attention in the past decade, resulting in the introduction of novel immunotherapy agents. Encouraging clinical data have been reported for T cell-based targeting agents, such as anti-CD19/CD3 bispecific T-cell engager blinatumomab and chimeric antigen receptor (CAR)-T therapies, for acute lymphoblastic leukemia and B cell non-Hodgkin lymphoma (B-NHL). However, clinical use of both blinatumomab and CAR-T therapies has been limited due to unfavorable pharmacokinetics (PK), significant toxicity associated with cytokine release syndrome and neurotoxicity, and manufacturing challenges. We present here a fully human CD19xCD3 bispecific antibody (TNB-486) for the treatment of B-NHL that could address the limitations of the current approved treatments. In the presence of CD19+ target cells and T cells, TNB-486 induces tumor cell lysis with minimal cytokine release, when compared to a positive control. In vivo, TNB-486 clears CD19+ tumor cells in immunocompromised mice in the presence of human peripheral blood mononuclear cells in multiple models. Additionally, the PK of TNB-486 in mice or cynomolgus monkeys is similar to conventional antibodies. This new T cell engaging bispecific antibody targeting CD19 represents a novel therapeutic that induces potent T cell-mediated tumor-cell cytotoxicity uncoupled from high levels of cytokine release, making it an attractive candidate for B-NHL therapy.

Perspective: Designing T-Cell Engagers With Better Therapeutic Windows

This perspective highlights the history and challenges of developing CD3-based bispecific T-cell engagers (TCEs) as cancer therapeutics as well as considerations and potential strategies for designing the next generation TCE molecules. The goal of this article is to raise awareness of natural T-cell biology and how to best harness the tumor cell killing capacity of cytotoxic T-cells with TCEs. In light of 30 years of concerted efforts to advance TCEs in early clinical development, many of the first-generation bispecific antibodies have exhibited lackluster safety, efficacy, and manufacturability profiles. As of January 2020, blinatumomab remains the only approved TCE. Many of the current set-backs in early clinical trials implicate the high-affinity CD3 binding domains employed and the respective bispecific platforms as potential culprits. The underlying conviction of the authors is that by taking corrective measures, TCEs can transform cancer therapy. Through openness, transparency, and much needed feedback from ongoing clinical studies, the field can continuously improve the design and effectiveness of next generation T-cell redirecting therapeutics.

Abstract 1554: Uncoupling tumor-cell cytotoxicity from cytokine release with novel T-cell engaging bispecific antibodies

T-cell recruiting bispecific antibody (T-BsAb) treatment is an emerging cancer immunotherapy notable for its highly-specific tumor killing activity in humans. However, the success of T-BsAbs in the clinic has been limited due to cytokine-release related toxicities. In order for this technology to reach its potential, the next generation of T-BsAbs must avoid this unintended toxicity by uncoupling cytokine-release from tumor-cell killing. To this end, we used a sequence-based discovery method on CD3-immunized OmniFlic rats, which express human fixed light-chain antibodies, to identify several novel anti-CD3 antibodies. These antibodies bind to multiple CD3 epitopes, each with varying affinities and degrees of T-cell activation. Furthermore, when placed in the context of a T-BsAb, these novel anti-CD3 antibodies achieve similar levels of T-cell mediated tumor-specific lysis, but with a thousand-fold range in potency. Among these antibodies, we identified one lead with particularly favorable properties, in that it stimulates very low levels of cytokine production without sacrificing its potent antigen-specific tumor lysis in vitro, in mouse xenograft assays, as well as ex vivo patient sample studies. This unique behavior gives our novel anti-CD3 antibody the potential to widen the therapeutic window when treating patients and ultimately usher in the next generation of T-BsAbs.

Citation Format: Kyle Lorentsen, Preethi Sankaran, Nathan D. Trinklein, Duy Pham, Ute Schellenberger, Ben Buelow, Andrew Boudreau, Priya Choudhry, Starlynn C. Clarke, Kevin Dang, Katherine E. Harris, Suhasini Iyer, Brett Jorgensen, Payal Pratap, Udaya S. Rangaswamy, Harshad S. Ugamraj, Omid Vafa, Arun P. Wiita, Wim van Schooten, Roland Buelow, Shelley Force Aldred. Uncoupling tumor-cell cytotoxicity from cytokine release with novel T-cell engaging bispecific antibodies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1554.

Efficient tumor killing and minimal cytokine release with novel T-cell agonist bispecific antibodies

T-cell-recruiting bispecific antibodies (T-BsAbs) have shown potent tumor killing activity in humans, but cytokine release-related toxicities have affected their clinical utility. The use of novel anti-CD3 binding domains with more favorable properties could aid in the creation of T-BsAbs with improved therapeutic windows. Using a sequence-based discovery platform, we identified new anti-CD3 antibodies from humanized rats that bind to multiple epitopes and elicit varying levels of T-cell activation. In T-BsAb format, 12 different anti-CD3 arms induce equivalent levels of tumor cell lysis by primary T-cells, but potency varies by a thousand-fold. Our lead CD3-targeting arm stimulates very low levels of cytokine release, but drives robust tumor antigen-specific killing in vitro and in a mouse xenograft model. This new CD3-targeting antibody underpins a next-generation T-BsAb platform in which potent cytotoxicity is uncoupled from high levels of cytokine release, which may lead to a wider therapeutic window in the clinic.

Multispecific Antibody Development Platform Based on Human Heavy Chain Antibodies

Heavy chain-only antibodies (HCAbs) do not associate with light chains and their V_H regions are functional as single domains, forming the smallest active antibody fragment. These V_H regions are ideal building blocks for a variety of antibody-based biologics because they tolerate fusion to other molecules and may also be attached in series to construct multispecific antibodies without the need for protein engineering to ensure proper heavy and light chain pairing. Production of human HCAbs has been impeded by the fact that natural human V_H regions require light chain association and display poor biophysical characteristics when expressed in the absence of light chains. Here, we present an innovative platform for the rapid development of diverse sets of human HCAbs that have been selected in vivo. Our unique approach combines antibody repertoire analysis with immunization of transgenic rats, called UniRats, that produce chimeric HCAbs with fully human V_H domains in response to an antigen challenge. UniRats express HCAbs from large transgenic loci representing the entire productive human heavy chain V(D)J repertoire, mount robust immune responses to a wide array of antigens, exhibit diverse V gene usage and generate large panels of stable, high affinity, antigen-specific molecules.

Sequence-Based Discovery Demonstrates That Fixed Light Chain Human Transgenic Rats Produce a Diverse Repertoire of Antigen-Specific Antibodies

We created a novel transgenic rat that expresses human antibodies comprising a diverse repertoire of heavy chains with a single common rearranged kappa light chain (IgKV3-15-JK1). This fixed light chain animal, called OmniFlic, presents a unique system for human therapeutic antibody discovery and a model to study heavy chain repertoire diversity in the context of a constant light chain. The purpose of this study was to analyze heavy chain variable gene usage, clonotype diversity, and to describe the sequence characteristics of antigen-specific monoclonal antibodies (mAbs) isolated from immunized OmniFlic animals. Using next-generation sequencing antibody repertoire analysis, we measured heavy chain variable gene usage and the diversity of clonotypes present in the lymph node germinal centers of 75 OmniFlic rats immunized with 9 different protein antigens. Furthermore, we expressed 2,560 unique heavy chain sequences sampled from a diverse set of clonotypes as fixed light chain antibody proteins and measured their binding to antigen by ELISA. Finally, we measured patterns and overall levels of somatic hypermutation in the full B-cell repertoire and in the 2,560 mAbs tested for binding. The results demonstrate that OmniFlic animals produce an abundance of antigen-specific antibodies with heavy chain clonotype diversity that is similar to what has been described with unrestricted light chain use in mammals. In addition, we show that sequence-based discovery is a highly effective and efficient way to identify a large number of diverse monoclonal antibodies to a protein target of interest.

When monoclonal antibodies are not monospecific: Hybridomas frequently express additional functional variable regions

Monoclonal antibodies are commonly assumed to be monospecific, but anecdotal studies have reported genetic diversity in antibody heavy chain and light chain genes found within individual hybridomas. As the prevalence of such diversity has never been explored, we analyzed 185 random hybridomas, in a large multicenter dataset. The hybridomas analyzed were not biased towards those with cloning difficulties or known to have additional chains. Of the hybridomas we evaluated, 126 (68.1%) contained no additional productive chains, while the remaining 59 (31.9%) contained one or more additional productive heavy or light chains. The expression of additional chains degraded properties of the antibodies, including specificity, binding signal and/or signal-to-noise ratio, as determined by enzyme-linked immunosorbent assay and immunohistochemistry. The most abundant mRNA transcripts found in a hybridoma cell line did not necessarily encode the antibody chains providing the correct specificity. Consequently, when cloning antibody genes, functional validation of all possible VH and VL combinations is required to identify those with the highest affinity and lowest cross-reactivity. These findings, reflecting the current state of hybridomas used in research, reiterate the importance of using sequence-defined recombinant antibodies for research or diagnostic use.

A Single Human Papillomavirus Vaccine Dose Improves B Cell Memory in Previously Infected Subjects

Although licensed human papillomavirus (HPV) vaccines are most efficacious in persons never infected with HPV, they also reduce infection and disease in previously infected subjects, indicating natural immunity is not entirely protective against HPV re-infection. The aim of this exploratory study was to examine the B cell memory elicited by HPV infection and evaluate whether vaccination merely boosts antibody (Ab) levels in previously infected subjects or also improves the quality of B cell memory. Toward this end, the memory B cells (Bmem) of five unvaccinated, HPV-seropositive subjects were isolated and characterized, and subject recall responses to a single HPV vaccine dose were analyzed. Vaccination boosted Ab levels 24- to 930-fold (median 77-fold) and Bmem numbers 3- to 27-fold (median 6-fold). In addition, Abs cloned from naturally elicited Bmem were generally non-neutralizing, whereas all those isolated following vaccination were neutralizing. Moreover, Ab and plasmablast responses indicative of memory recall responses were only observed in two subjects. These results suggest HPV vaccination augments both the magnitude and quality of natural immunity and demonstrate that sexually active persons could also benefit from HPV vaccination. This study may have important public policy implications, especially for the older 'catch-up' group within the vaccine's target population.

Identification of a Functional In Vivo p53 Response Element in the Coding Sequence of the Xeroderma Pigmentosum Group C Gene

The protein product of the xeroderma pigmentosum group C (XPC) gene is a DNA damage recognition factor that functions early in the process of global genomic nucleotide excision repair. Regulation of XPC expression is governed in part by p53 at the transcriptional level. To identify the regulatory elements involved in the p53-dependent control of XPC expression, we performed a quantitative PCR tiling experiment using multiple regularly spaced primer pairs over an 11-kb region centered around the XPC transcriptional start site. p53 chromatin immunoprecipitation was performed following ultraviolet irradiation, and DNA was analyzed for enrichment at each of 48 amplicons covering this region. A segment just upstream of the XPC translational initiation site was significantly enriched, whereas no enrichment of any other region was noted. In vitro promoter reporter assays and gel retardation assays were used to confirm the p53 responsiveness of this region and to define the minimal region with stimulating activity. We identified a p53 response element that has significant similarity to a consensus sequence, with 3 mismatches. This response element is unique in that part of the p53 binding site included the coding sequence for the first 2 amino acids in the XPC protein.

Functional analysis of transcription factor binding sites in human promoters

Background

The binding of transcription factors to specific locations in the genome is integral to the orchestration of transcriptional regulation in cells. To characterize transcription factor binding site function on a large scale, we predicted and mutagenized 455 binding sites in human promoters. We carried out functional tests on these sites in four different immortalized human cell lines using transient transfections with a luciferase reporter assay, primarily for the transcription factors CTCF, GABP, GATA2, E2F, STAT, and YY1.

Results

In each cell line, between 36% and 49% of binding sites made a functional contribution to the promoter activity; the overall rate for observing function in any of the cell lines was 70%. Transcription factor binding resulted in transcriptional repression in more than a third of functional sites. When compared with predicted binding sites whose function was not experimentally verified, the functional binding sites had higher conservation and were located closer to transcriptional start sites (TSSs). Among functional sites, repressive sites tended to be located further from TSSs than were activating sites. Our data provide significant insight into the functional characteristics of YY1 binding sites, most notably the detection of distinct activating and repressing classes of YY1 binding sites. Repressing sites were located closer to, and often overlapped with, translational start sites and presented a distinctive variation on the canonical YY1 binding motif.

Conclusions

The genomic properties that we found to associate with functional TF binding sites on promoters -- conservation, TSS proximity, motifs and their variations -- point the way to improved accuracy in future TFBS predictions.

Identification, replication, and functional fine-mapping of expression quantitative trait loci in primary human liver tissue

The discovery of expression quantitative trait loci ("eQTLs") can help to unravel genetic contributions to complex traits. We identified genetic determinants of human liver gene expression variation using two independent collections of primary tissue profiled with Agilent (n = 206) and Illumina (n = 60) expression arrays and Illumina SNP genotyping (550K), and we also incorporated data from a published study (n = 266). We found that ∼30% of SNP-expression correlations in one study failed to replicate in either of the others, even at thresholds yielding high reproducibility in simulations, and we quantified numerous factors affecting reproducibility. Our data suggest that drug exposure, clinical descriptors, and unknown factors associated with tissue ascertainment and analysis have substantial effects on gene expression and that controlling for hidden confounding variables significantly increases replication rate. Furthermore, we found that reproducible eQTL SNPs were heavily enriched near gene starts and ends, and subsequently resequenced the promoters and 3'UTRs for 14 genes and tested the identified haplotypes using luciferase assays. For three genes, significant haplotype-specific in vitro functional differences correlated directly with expression levels, suggesting that many bona fide eQTLs result from functional variants that can be mechanistically isolated in a high-throughput fashion. Finally, given our study design, we were able to discover and validate hundreds of liver eQTLs. Many of these relate directly to complex traits for which liver-specific analyses are likely to be relevant, and we identified dozens of potential connections with disease-associated loci. These included previously characterized eQTL contributors to diabetes, drug response, and lipid levels, and they suggest novel candidates such as a role for NOD2 expression in leprosy risk and C2orf43 in prostate cancer. In general, the work presented here will be valuable for future efforts to precisely identify and functionally characterize genetic contributions to a variety of complex traits.

Two-tiered approach identifies a network of cancer and liver disease-related genes regulated by miR-122

MicroRNAs function as important regulators of gene expression and are commonly linked to development, differentiation, and diseases such as cancer. To better understand their roles in various biological processes, identification of genes targeted by microRNAs is necessary. Although prediction tools have significantly helped with this task, experimental approaches are ultimately required for extensive target search and validation. We employed two independent yet complementary high throughput approaches to map a large set of mRNAs regulated by miR-122, a liver-specific microRNA implicated in regulation of fatty acid and cholesterol metabolism, hepatitis C infection, and hepatocellular carcinoma. The combination of luciferase reporter-based screening and shotgun proteomics resulted in the identification of 260 proteins significantly down-regulated in response to miR-122 in at least one method, 113 of which contain predicted miR-122 target sites. These proteins are enriched for functions associated with the cell cycle, differentiation, proliferation, and apoptosis. Among these miR-122-sensitive proteins, we identified a large group with strong connections to liver metabolism, diseases, and hepatocellular carcinoma. Additional analyses, including examination of consensus binding motifs for both miR-122 and target sequences, provide further insight into miR-122 function.

Sequence features that drive human promoter function and tissue specificity

Promoters are important regulatory elements that contain the necessary sequence features for cells to initiate transcription. To functionally characterize a large set of human promoters, we measured the transcriptional activities of 4575 putative promoters across eight cell lines using transient transfection reporter assays. In parallel, we measured gene expression in the same cell lines and observed a significant correlation between promoter activity and endogenous gene expression (r = 0.43). As transient transfection assays directly measure the promoting effect of a defined fragment of DNA sequence, decoupled from epigenetic, chromatin, or long-range regulatory effects, we sought to predict whether a promoter was active using sequence features alone. CG dinucleotide content was highly predictive of ubiquitous promoter activity, necessitating the separation of promoters into two groups: high CG promoters, mostly ubiquitously active, and low CG promoters, mostly cell line-specific. Computational models trained on the binding potential of transcriptional factor (TF) binding motifs could predict promoter activities in both high and low CG groups: average area under the receiver operating characteristic curve (AUC) of the models was 91% and exceeded the AUC of CG content by an average of 23%. Known relationships, for example, between HNF4A and hepatocytes, were recapitulated in the corresponding cell lines, in this case the liver-derived cell line HepG2. Half of the associations between tissue-specific TFs and cell line-specific promoters were new. Our study underscores the importance of collecting functional information from complementary assays and conditions to understand biology in a systematic framework.

A functional analysis of the CREB signaling pathway using HaloCHIP-chip and high throughput reporter assays

Background

Regulation of gene expression is essential for normal development and cellular growth. Transcriptional events are tightly controlled both spatially and temporally by specific DNA-protein interactions. In this study we finely map the genome-wide targets of the CREB protein across all known and predicted human promoters, and characterize the functional consequences of a subset of these binding events using high-throughput reporter assays. To measure CREB binding, we used HaloCHIP, an antibody-free alternative to the ChIP method that utilizes the HaloTag fusion protein, and also high-throughput promoter-luciferase reporter assays, which provide rapid and quantitative screening of promoters for transcriptional activation or repression in living cells.

Results

In analysis of CREB genome-wide binding events using a comprehensive DNA microarray of human promoters, we observe for the first time that CREB has a strong preference for binding at bidirectional promoters and unlike unidirectional promoters, these binding events often occur downstream of transcription start sites. Comparison between HaloCHIP-chip and ChIP-chip data reveal this to be true for both methodologies, indicating it is not a bias of the technology chosen. Transcriptional data obtained from promoter-luciferase reporter arrays also show an unprecedented, high level of activation of CREB-bound promoters in the presence of the co-activator protein TORC1.

Conclusion

These data suggest for the first time that TORC1 provides directional information when CREB is bound at bidirectional promoters and possible pausing of the CREB protein after initial transcriptional activation. Also, this combined approach demonstrates the ability to more broadly characterize CREB protein-DNA interactions wherein not only DNA binding sites are discovered, but also the potential of the promoter sequence to respond to CREB is evaluated.

The ets-related transcription factor GABP directs bidirectional transcription

Approximately 10% of genes in the human genome are distributed such that their transcription start sites are located less than 1 kb apart on opposite strands. These divergent gene pairs have a single intergenic segment of DNA, which in some cases appears to share regulatory elements, but it is unclear whether these regions represent functional bidirectional promoters or two overlapping promoters. A recent study showed that divergent promoters are enriched for consensus binding sequences of a small group of transcription factors, including the ubiquitous ets-family transcription factor GA-binding protein (GABP). Here we show that GABP binds to more than 80% of divergent promoters in at least one cell type. Furthermore, we demonstrate that GABP binding is correlated and associated with bidirectional transcriptional activity in a luciferase transfection assay. In addition, we find that the addition of a strict consensus GABP site into a set of promoters that normally function in only one direction significantly increases activity in the opposite direction in 67% of cases. Our findings demonstrate that GABP regulates the majority of divergent promoters and suggest that bidirectional transcriptional activity is mediated through GABP binding and transactivation at both divergent and nondivergent promoters.

Surveys of the locations of genes in the human genome have revealed that a surprising number of genes, greater than 10%, have transcription start sites within 1 kb of one another on opposite strands. These divergent gene pairs, sometimes referred to as bidirectional genes, are common in organisms such as bacteria and yeast, but it is unknown why such an arrangement exists in large, mammalian genomes. Recently, it has become apparent that the promoters of these divergent genes are regulated by a subset of transcription factors, and we have focused on one of these, GA-binding protein (GABP). We find that it regulates a large number of human genes, including the majority of divergent genes, and that its binding is associated with, correlated with, and sufficient for bidirectional transcriptional activity. Although clearly GABP is a major regulator of divergent genes, which carry out a variety of roles critical for the function and survival of the cell, these data also propose novel roles for GABP as a transcription factor. For example, the ability of GABP to promote bidirectional transcription may prove to be biologically relevant in generating many of the transcripts that have been observed outside of protein coding genes.

Systematic evaluation of variability in ChIP-chip experiments using predefined DNA targets

The most widely used method for detecting genome-wide protein-DNA interactions is chromatin immunoprecipitation on tiling microarrays, commonly known as ChIP-chip. Here, we conducted the first objective analysis of tiling array platforms, amplification procedures, and signal detection algorithms in a simulated ChIP-chip experiment. Mixtures of human genomic DNA and "spike-ins" comprised of nearly 100 human sequences at various concentrations were hybridized to four tiling array platforms by eight independent groups. Blind to the number of spike-ins, their locations, and the range of concentrations, each group made predictions of the spike-in locations. We found that microarray platform choice is not the primary determinant of overall performance. In fact, variation in performance between labs, protocols, and algorithms within the same array platform was greater than the variation in performance between array platforms. However, each array platform had unique performance characteristics that varied with tiling resolution and the number of replicates, which have implications for cost versus detection power. Long oligonucleotide arrays were slightly more sensitive at detecting very low enrichment. On all platforms, simple sequence repeats and genome redundancy tended to result in false positives. LM-PCR and WGA, the most popular sample amplification techniques, reproduced relative enrichment levels with high fidelity. Performance among signal detection algorithms was heavily dependent on array platform. The spike-in DNA samples and the data presented here provide a stable benchmark against which future ChIP platforms, protocol improvements, and analysis methods can be evaluated.

Integrated analysis of experimental data sets reveals many novel promoters in 1% of the human genome

The regulation of transcriptional initiation in the human genome is a critical component of global gene regulation, but a complete catalog of human promoters currently does not exist. In order to identify regulatory regions, we developed four computational methods to integrate 129 sets of ENCODE-wide chromatin immunoprecipitation data. They collectively predicted 1393 regions. Roughly 47% of the regions were unique to one method, as each method makes different assumptions about the data. Overall, predicted regions tend to localize to highly conserved, DNase I hypersensitive, and actively transcribed regions in the genome. Interestingly, a significant portion of the regions overlaps with annotated 3'-UTRs, suggesting that some of them might regulate anti-sense transcription. The majority of the predicted regions are >2 kb away from the 5'-ends of previously annotated human cDNAs and hence are novel. These novel regions may regulate unannotated transcripts or may represent new alternative transcription start sites of known genes. We tested 163 such regions for promoter activity in four cell lines using transient transfection assays, and 25% of them showed transcriptional activity above background in at least one cell line. We also performed 5'-RACE experiments on 62 novel regions, and 76% of the regions were associated with the 5'-ends of at least two RACE products. Our results suggest that there are at least 35% more functional promoters in the human genome than currently annotated.

Transcription factor binding and modified histones in human bidirectional promoters

Bidirectional promoters have received considerable attention because of their ability to regulate two downstream genes (divergent genes). They are also highly abundant, directing the transcription of approximately 11% of genes in the human genome. We categorized the presence of DNA sequence motifs, binding of transcription factors, and modified histones as overrepresented, shared, or underrepresented in bidirectional promoters with respect to unidirectional promoters. We found that a small set of motifs, including GABPA, MYC, E2F1, E2F4, NRF-1, CCAAT, YY1, and ACTACAnnTCC are overrepresented in bidirectional promoters, while the majority (73%) of known vertebrate motifs are underrepresented. We performed chromatin-immunoprecipitation (ChIP), followed by quantitative PCR for GABPA, on 118 regions in the human genome and showed that it binds to bidirectional promoters more frequently than unidirectional promoters, and its position-specific scoring matrix is highly predictive of binding. Signatures of active transcription, such as occupancy of RNA polymerase II and the modified histones H3K4me2, H3K4me3, and H3ac, are overrepresented in regions around bidirectional promoters, suggesting that a higher fraction of divergent genes are transcribed in a given cell than the fraction of other genes. Accordingly, analysis of whole-genome microarray data indicates that 68% of divergent genes are transcribed compared with 44% of all human genes. By combining the analysis of publicly available ENCODE data and a detailed study of GABPA, we survey bidirectional promoters with breadth and depth, leading to biological insights concerning their motif composition and bidirectional regulatory mode.

Identification and analysis of functional elements in 1% of the human genome by the ENCODE pilot project

We report the generation and analysis of functional data from multiple, diverse experiments performed on a targeted 1% of the human genome as part of the pilot phase of the ENCODE Project. These data have been further integrated and augmented by a number of evolutionary and computational analyses. Together, our results advance the collective knowledge about human genome function in several major areas. First, our studies provide convincing evidence that the genome is pervasively transcribed, such that the majority of its bases can be found in primary transcripts, including non-protein-coding transcripts, and those that extensively overlap one another. Second, systematic examination of transcriptional regulation has yielded new understanding about transcription start sites, including their relationship to specific regulatory sequences and features of chromatin accessibility and histone modification. Third, a more sophisticated view of chromatin structure has emerged, including its inter-relationship with DNA replication and transcriptional regulation. Finally, integration of these new sources of information, in particular with respect to mammalian evolution based on inter- and intra-species sequence comparisons, has yielded new mechanistic and evolutionary insights concerning the functional landscape of the human genome. Together, these studies are defining a path for pursuit of a more comprehensive characterization of human genome function.

Serum response factor binding sites differ in three human cell types

The serum response factor (SRF) is essential for embryonic development and maintenance of muscle cells and neurons. The mechanism by which this factor controls these divergent pathways is unclear. Here we present a genome-wide view of occupancy of SRF at its binding sites with a focus on those that vary with cell type. We used chromatin immunoprecipitation (ChIP) in combination with human promoter microarrays to identify 216 putative SRF binding sites in the human genome. We performed independent quantitative PCR validation at over half of these sites that resulted in 146 sites we assert to be true binding sites at over 90% confidence. Nearly half of the sites are bound by SRF in only one of the three cell types we tested, providing strong evidence for the diverse roles for SRF in different cell types. We also explore possible mechanisms controlling differential binding of SRF in these cell types by assaying cofactor binding, DNA methylation, histone methylation, and histone acetylation at a subset of sites bound preferentially in smooth muscle cells. Although we did not see a strong correlation between SRF binding and epigenetics modifications, at these sites, we propose that SRF cofactors may play an important role in determining cell-dependent SRF binding sites. ELK4 (previously known as SAP-1 [SRF-associated protein-1]) is ubiquitously expressed. Therefore, we expected it to occupy sites where SRF binding is common in all cell types. Indeed, 90% of SRF sites also bound by ELK4 were common to all three cell types. Together, our data provide a more complete understanding of the regulatory network controlled by SRF.

Comprehensive analysis of transcriptional promoter structure and function in 1% of the human genome

Transcriptional promoters comprise one of many classes of eukaryotic transcriptional regulatory elements. Identification and characterization of these elements are vital to understanding the complex network of human gene regulation. Using full-length cDNA sequences to identify transcription start sites (TSS), we predicted more than 900 putative human transcriptional promoters in the ENCODE regions, representing a comprehensive sampling of promoters in 1% of the genome. We identified 387 fragments that function as promoters in at least one of 16 cell lines by measuring promoter activity in high-throughput transient transfection reporter assays. These positive functional results demonstrate widespread use of alternative promoters. We show a strong correlation between promoter activity and the corresponding endogenous RNA transcript levels, providing the first experimental quantitative estimate of promoter contribution to gene regulation. Finally, we identified functional regions within a randomly selected subset of 45 promoters using deletion analyses. These experiments showed that, on average, the sequence -300 to -50 bp of the TSS positively contributes to core promoter activity. Interestingly, putative negative elements were identified -1000 to -500 bp upstream of the TSS for 55% of genes tested. These data provide the largest and most comprehensive view of promoter function in the human genome.

Direct isolation and identification of promoters in the human genome

Transcriptional regulatory elements play essential roles in gene expression during animal development and cellular response to environmental signals, but our knowledge of these regions in the human genome is limited despite the availability of the complete genome sequence. Promoters mark the start of every transcript and are an important class of regulatory elements. A large, complex protein structure known as the pre-initiation complex (PIC) is assembled on all active promoters, and the presence of these proteins distinguishes promoters from other sequences in the genome. Using components of the PIC as tags, we isolated promoters directly from human cells as protein-DNA complexes and identified the resulting DNA sequences using genomic tiling microarrays. Our experiments in four human cell lines uncovered 252 PIC-binding sites in 44 semirandomly selected human genomic regions comprising 1% (30 megabase pairs) of the human genome. Nearly 72% of the identified fragments overlap or immediately flank 5' ends of known cDNA sequences, while the remainder is found in other genomic regions that likely harbor putative promoters of unannotated transcripts. Indeed, molecular analysis of the RNA isolated from one cell line uncovered transcripts initiated from over half of the putative promoter fragments, and transient transfection assays revealed promoter activity for a significant proportion of fragments when they were fused to a luciferase reporter gene. These results demonstrate the specificity of a genome-wide analysis method for mapping transcriptional regulatory elements and also indicate that a small, yet significant number of human genes remains to be discovered.

Transcriptional regulation and binding of heat shock factor 1 and heat shock factor 2 to 32 human heat shock genes during thermal stress and differentiation

Transcription of mammalian heat shock genes can be regulated by heat shock factors (HSF) 1 and 2. Although it has been shown previously that these factors respond to distinct stimuli, a broad analysis of the induction and function of these factors in living cells has not been performed. In our study, we assayed binding of human HSF1 and HSF2 at the promoters of 32 genes identified through LocusLink as heat shock genes in response to elevated temperature and hemin-induced differentiation in human K562 erythroleukemic cells using the chromatin immunoprecipitation technique. We also measured the induced expression of these genes under these 2 conditions. We found that 17 of the 32 genes were transcriptionally induced during heat shock, and HSF1 binding was detected at 15 of the 17 promoters. Nearly all the genes induced by heat shock were also induced to a lesser degree during hemin treatment. However, some genes were induced significantly more during hemin treatment than during heat shock. A new finding is that HSF1 and HSF2 bind to the same targets, but HSF1 binding is activated more by heat than by hemin treatment, and HSF2 binding is only activated by hemin treatment and not by heat. This technology also identified previously unknown HSF1 binding sites near genes that were previously shown to be heat inducible that may contribute to gene-specific regulation.

Diverse and specific gene expression responses to stresses in cultured human cells

We used cDNA microarrays in a systematic study of the gene expression responses of HeLa cells and primary human lung fibroblasts to heat shock, endoplasmic reticulum stress, oxidative stress, and crowding. Hierarchical clustering of the data revealed groups of genes with coherent biological themes, including genes that responded to specific stresses and others that responded to multiple types of stress. Fewer genes increased in expression after multiple stresses than in free-living yeasts, which have a large general stress response program. Most of the genes induced by multiple diverse stresses are involved in cell-cell communication and other processes specific to higher organisms. We found substantial differences between the stress responses of HeLa cells and primary fibroblasts. For example, many genes were induced by oxidative stress and dithiothreitol in fibroblasts but not HeLa cells; conversely, a group of transcription factors, including c-fos and c-jun, were induced by heat shock in HeLa cells but not in fibroblasts. The dataset is freely available for search and download at http://microarray-pubs.stanford.edu/human_stress/Home.shtml.

An abundance of bidirectional promoters in the human genome

The alignment of full-length human cDNA sequences to the finished sequence of the human genome provides a unique opportunity to study the distribution of genes throughout the genome. By analyzing the distances between 23,752 genes, we identified a class of divergently transcribed gene pairs, representing more than 10% of the genes in the genome, whose transcription start sites are separated by less than 1000 base pairs. Although this bidirectional arrangement has been previously described in humans and other species, the prevalence of bidirectional gene pairs in the human genome is striking, and the mechanisms of regulation of all but a few bidirectional genes are unknown. Our work shows that the transcripts of many bidirectional pairs are coexpressed, but some are antiregulated. Further, we show that many of the promoter segments between two bidirectional genes initiate transcription in both directions and contain shared elements that regulate both genes. We also show that the bidirectional arrangement is often conserved among mouse orthologs. These findings demonstrate that a bidirectional arrangement provides a unique mechanism of regulation for a significant number of mammalian genes.

The role of heat shock transcription factor 1 in the genome-wide regulation of the mammalian heat shock response

Previous work has implicated heat shock transcription factor 1 (HSF1) as the primary transcription factor responsible for the transcriptional response to heat stress in mammalian cells. We characterized the heat shock response of mammalian cells by measuring changes in transcript levels and assaying binding of HSF1 to promoter regions for candidate heat shock genes chosen by a combination of genome-wide computational and experimental methods. We found that many heat-inducible genes have HSF1 binding sites (heat shock elements, HSEs) in their promoters that are bound by HSF1. Surprisingly, for 24 heat-inducible genes, we detected no HSEs and no HSF1 binding. Furthermore, of 182 promoters with likely HSE sequences, we detected HSF1 binding at only 94 of these promoters. Also unexpectedly, we found 48 genes with HSEs in their promoters that are bound by HSF1 but that nevertheless did not show induction after heat shock in the cell types we examined. We also studied the transcriptional response to heat shock in fibroblasts from mice lacking the HSF1 gene. We found 36 genes in these cells that are induced by heat as well as they are in wild-type cells. These results provide evidence that HSF1 does not regulate the induction of every transcript that accumulates after heat shock, and our results suggest that an independent posttranscriptional mechanism regulates the accumulation of a significant number of transcripts.

Identification and functional analysis of human transcriptional promoters

Genomic and full-length cDNA sequences provide opportunities for understanding human gene structure and transcriptional regulatory elements. The simplest regulatory elements to identify are promoters, as their positions are dictated by the location of transcription start sites. We aligned full-length cDNA clones from the Mammalian Gene Collection to the human genome rough draft sequence to estimate the start sites of more than 10,000 human transcripts. We selected genomic sequence just upstream from the 5' end of these cDNA sequences and designated these as putative promoters. We assayed the functions of 152 of these DNA fragments, chosen at random from the entire set, in a luciferase-based transfection assay in four human cultured cell types. Ninety-one percent of these DNA fragments showed significant transcriptional activity in at least one of the cell lines, whereas 89% showed activity in at least two of the lines. We analyzed the distributions of strengths of these promoter fragments in the different cell types and identified likely alternative promoters in a large fraction of the genes. These data indicate that this approach is an effective method for predicting human promoters and provide the first set of functional data collected in parallel for a large set of human promoters.