Transcription Factor Occupancy Research Articles

SBR-Blood: systems biology repository for hematopoietic cells

Extensive research into hematopoiesis (the development of blood cells) over several decades has generated large sets of expression and epigenetic profiles in multiple human and mouse blood cell types. However, there is no single location to analyze how gene regulatory processes lead to different mature blood cells. We have developed a new database framework called hematopoietic Systems Biology Repository (SBR-Blood), available online at http://sbrblood.nhgri.nih.gov, which allows user-initiated analyses for cell type correlations or gene-specific behavior during differentiation using publicly available datasets for array- and sequencing-based platforms from mouse hematopoietic cells. SBR-Blood organizes information by both cell identity and by hematopoietic lineage. The validity and usability of SBR-Blood has been established through the reproduction of workflows relevant to expression data, DNA methylation, histone modifications and transcription factor occupancy profiles.

Transcription factor trapping by RNA in gene regulatory elements.

Transcription factors (TFs) bind specific sequences in promoter-proximal and -distal DNA elements to regulate gene transcription. RNA is transcribed from both of these DNA elements, and some DNA binding TFs bind RNA. Hence, RNA transcribed from regulatory elements may contribute to stable TF occupancy at these sites. We show that the ubiquitously expressed TF Yin-Yang 1 (YY1) binds to both gene regulatory elements and their associated RNA species across the entire genome. Reduced transcription of regulatory elements diminishes YY1 occupancy, whereas artificial tethering of RNA enhances YY1 occupancy at these elements. We propose that RNA makes a modest but important contribution to the maintenance of certain TFs at gene regulatory elements and suggest that transcription of regulatory elements produces a positive-feedback loop that contributes to the stability of gene expression programs.

Large-scale identification of sequence variants influencing human transcription factor occupancy in vivo.

The function of human regulatory regions depends exquisitely on their local genomic environment and cellular context, complicating experimental analysis of the expanding pool of common disease- and trait-associated variants that localize within regulatory DNA. We leverage allelically resolved genomic DNaseI footprinting data encompassing 166 individuals and 114 cell types to identify >60,000 common variants that directly impact transcription factor occupancy and regulatory DNA accessibility in vivo. The unprecedented scale of these data enable systematic analysis of the impact of sequence variation on transcription factor occupancy in vivo. We leverage this analysis to develop accurate models of variation affecting the recognition sites for diverse transcription factors, and apply these models to discriminate nearly 500,000 common regulatory variants likely to affect transcription factor occupancy across the human genome. The approach and results provide a novel foundation for analysis and interpretation of noncoding variation in complete human genomes, and for systems-level investigation of disease-associated variants.

Epigenetic program and transcription factor circuitry of dendritic cell development.

Dendritic cells (DC) are professional antigen presenting cells that develop from hematopoietic stem cells through successive steps of lineage commitment and differentiation. Multipotent progenitors (MPP) are committed to DC restricted common DC progenitors (CDP), which differentiate into specific DC subsets, classical DC (cDC) and plasmacytoid DC (pDC). To determine epigenetic states and regulatory circuitries during DC differentiation, we measured consecutive changes of genome-wide gene expression, histone modification and transcription factor occupancy during the sequel MPP-CDP-cDC/pDC. Specific histone marks in CDP reveal a DC-primed epigenetic signature, which is maintained and reinforced during DC differentiation. Epigenetic marks and transcription factor PU.1 occupancy increasingly coincide upon DC differentiation. By integrating PU.1 occupancy and gene expression we devised a transcription factor regulatory circuitry for DC commitment and subset specification. The circuitry provides the transcription factor hierarchy that drives the sequel MPP-CDP-cDC/pDC, including Irf4, Irf8, Tcf4, Spib and Stat factors. The circuitry also includes feedback loops inferred for individual or multiple factors, which stabilize distinct stages of DC development and DC subsets. In summary, here we describe the basic regulatory circuitry of transcription factors that drives DC development.

An autoregulatory enhancer controls mammary-specific STAT5 functions.

Signal Transducers and Activators of Transcription (STATs) are principal transcription factors downstream of cytokine receptors. Although STAT5A is expressed in most tissues it remains to be understood why its premier, non-redundant functions are restricted to prolactin-induced mammary gland development and function. We report that the ubiquitously expressed Stat5a/b locus is subject to additional lineage-specific transcriptional control in mammary epithelium. Genome-wide surveys of epigenetic status and transcription factor occupancy uncovered a putative mammary-specific enhancer within the intergenic sequences separating the two Stat5 genes. This region exhibited several hallmarks of genomic enhancers, including DNaseI hypersensitivity, H3K27 acetylation and binding by GR, NFIB, ELF5 and MED1. Mammary-specific STAT5 binding was obtained at two canonical STAT5 binding motifs. CRISPR/Cas9-mediated genome editing was used to delete these sites in mice and determine their biological function. Mutant animals exhibited an 80% reduction of Stat5 levels in mammary epithelium and a concomitant reduction of STAT5-dependent gene expression. Transcriptome analysis identified a class of mammary-restricted genes that was particularly dependent on high STAT5 levels as a result of the intergenic enhancer. Taken together, the mammary-specific enhancer enables a positive feedback circuit that contributes to the remarkable abundance of STAT5 and, in turn, to the efficacy of STAT5-dependent mammary physiology.

New roles for DNA cytosine modification, eRNA, anchors, and superanchors in developing B cell progenitors

B-cell fate is orchestrated by a series of well-characterized developmental regulators. Here, we found that the onset of B-cell development was accompanied by large-scale changes in DNA cytosine modifications associated with promoters, enhancers, and anchors. These changes were tightly linked to alterations in transcription factor occupancy and nascent RNA (eRNA) transcription. We found that the prepro-B to the pro-B-cell transition was associated with a global exchange of DNA cytosine modifications for polycomb-mediated repression at CpG islands. Hypomethylated regions were found exclusively in the active/permissive compartment of the nucleus and were predominantly associated with regulatory elements or anchors that orchestrate the folding patterns of the genome. We identified superanchors, characterized by clusters of hypomethylated CCCTC-binding factor (CTCF)-bound elements, which were predominantly located at boundaries that define topological associated domains. A particularly prominent hypomethylated superanchor was positioned down-stream of the Ig heavy chain (Igh) locus. Analysis of global formaldehyde-cross-linking studies indicated that the Igh locus superanchor interacts with the VH region repertoire across vast genomic distances. We propose that the Igh locus superanchor sequesters the VH and DHJH regions into a spatial confined geometric environment to promote rapid first-passage times. Collectively, these studies demonstrate how, in developing B cells, DNA cytosine modifications associated with regulatory and architectural elements affect patterns of gene expression, folding patterns of the genome, and antigen receptor assembly.

Aging-Dependent Demethylation of Regulatory Elements Correlates with Chromatin State and Improved β Cell Function

Aging is driven by changes of the epigenetic state that are only partially understood. We performed a comprehensive epigenomic analysis of the pancreatic β cell, key player in glucose homeostasis, in adolescent and very old mice. We observe a global methylation drift resulting in an overall more leveled methylome in old β cells. Importantly, we discover targeted changes in the methylation status of β cell proliferation and function genes that go against the global methylation drift, are specific to β cells, and correlate with repression of the proliferation program and activation of metabolic regulators. These targeted alterations are associated with specific chromatin marks and transcription factor occupancy in young β cells. Strikingly, we find β cell function improved in aged mice, as predicted by the changes in methylome and transcriptome. Thus, aging of terminally differentiated cells in mammals is not always coupled to functional decline.

Abstract IA24: Regulation of acute leukemia initiation and progression by long noncoding RNAs.

Abstract Recent efforts to functionally annotate the human genome have revealed that up to 70% of our DNA is transcriptionally active. Since a very small portion of our genome encodes proteins many have hypothesized that a portion of this pervasive transcription results in production of non-coding RNA. By generating high depth RNA-Seq datasets integrated with chromatin features, several labs have revealed the presence of many thousands of previously un-annotated long non-coding RNAs (lncRNA), which are dynamically expressed in response to various stimuli in diverse cellular contexts. Despite these compelling advances, the vast majority of putative lncRNAs have not been proven to be functionally important, although a small portion have clearly been shown to play key regulatory roles. Most importantly, very little is known on the biological role of lncRNAs for human disease and specifically cancer. We provide here, using a combination of deep transcriptome sequencing, high-resolution transcription factor occupancy mapping and chromosome conformation capture data, the first comprehensive identification and characterization of lncRNA expression and function in T-cell acute lymphoblastic leukemia, an aggressive hematologic malignancy affecting both children and adults, driven by oncogenic NOTCH1 mutations and transcriptional activation. Moreover, we provide the first map of oncogene (NOTCH1)-targeted lncRNAs in this disease and identify an individual lncRNA (called LUNAR1) that appears to be essential for tumor growth as it controls cytokine (IGF1) signaling. Indeed, we were able to show that the LUNAR1 locus loops on a novel IGF1R enhancer controlling the expression of this important cytokine receptor. These studies suggest that: a) lncRNAs could be used as both biomarkers and therapy targets in human cancer, b) more efficient methods for large-scale inference and validation of lncRNA function are needed to fully understand their biological significance and, c) LUNAR1 is one of the first lncRNAs that can control growth of acute leukemia and a potential therapeutic target. These points will be addressed in this presentation. Citation Format: Iannis Aifantis. Regulation of acute leukemia initiation and progression by long noncoding RNAs. [abstract]. In: Proceedings of the AACR Special Conference on Hematologic Malignancies: Translating Discoveries to Novel Therapies; Sep 20-23, 2014; Philadelphia, PA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(17 Suppl):Abstract nr IA24.

Nonconsensus Protein Binding to Repetitive DNA Sequence Elements Significantly Affects Eukaryotic Genomes.

Recent genome-wide experiments in different eukaryotic genomes provide an unprecedented view of transcription factor (TF) binding locations and of nucleosome occupancy. These experiments revealed that a large fraction of TF binding events occur in regions where only a small number of specific TF binding sites (TFBSs) have been detected. Furthermore, in vitro protein-DNA binding measurements performed for hundreds of TFs indicate that TFs are bound with wide range of affinities to different DNA sequences that lack known consensus motifs. These observations have thus challenged the classical picture of specific protein-DNA binding and strongly suggest the existence of additional recognition mechanisms that affect protein-DNA binding preferences. We have previously demonstrated that repetitive DNA sequence elements characterized by certain symmetries statistically affect protein-DNA binding preferences. We call this binding mechanism nonconsensus protein-DNA binding in order to emphasize the point that specific consensus TFBSs do not contribute to this effect. In this paper, using the simple statistical mechanics model developed previously, we calculate the nonconsensus protein-DNA binding free energy for the entire C. elegans and D. melanogaster genomes. Using the available chromatin immunoprecipitation followed by sequencing (ChIP-seq) results on TF-DNA binding preferences for ~100 TFs, we show that DNA sequences characterized by low predicted free energy of nonconsensus binding have statistically higher experimental TF occupancy and lower nucleosome occupancy than sequences characterized by high free energy of nonconsensus binding. This is in agreement with our previous analysis performed for the yeast genome. We suggest therefore that nonconsensus protein-DNA binding assists the formation of nucleosome-free regions, as TFs outcompete nucleosomes at genomic locations with enhanced nonconsensus binding. In addition, here we perform a new, large-scale analysis using in vitro TF-DNA preferences obtained from the universal protein binding microarrays (PBM) for ~90 eukaryotic TFs belonging to 22 different DNA-binding domain types. As a result of this new analysis, we conclude that nonconsensus protein-DNA binding is a widespread phenomenon that significantly affects protein-DNA binding preferences and need not require the presence of consensus (specific) TFBSs in order to achieve genome-wide TF-DNA binding specificity.

Role of DNA Methylation in Modulating Transcription Factor Occupancy.

Although DNA methylation is commonly invoked as a mechanism for transcriptional repression, the extent to which it actively silences transcription factor (TF) occupancy sites in vivo is unknown. To study the role of DNA methylation in the active modulation of TF binding, we quantified the effect of DNA methylation depletion on the genomic occupancy patterns of CTCF, an abundant TF with known methylation sensitivity that is capable of autonomous binding to its target sites in chromatin. Here, we show that the vast majority (>98.5%) of the tens of thousands of unoccupied, methylated CTCF recognition sequences remain unbound upon abrogation of DNA methylation. The small fraction of sites that show methylation-dependent binding in vivo are in turn characterized by highly variable CTCF occupancy across cell types. Our results suggest that DNA methylation is not a primary groundskeeper of genomic TF landscapes, but rather a specialized mechanism for stabilizing intrinsically labile sites.

Differential epigenetic and transcriptional response of the skeletal muscle carnitine palmitoyltransferase 1B (CPT1B) gene to lipid exposure with obesity.

The ability to increase fatty acid oxidation (FAO) in response to dietary lipid is impaired in the skeletal muscle of obese individuals, which is associated with a failure to coordinately upregulate genes involved with FAO. While the molecular mechanisms contributing to this metabolic inflexibility are not evident, a possible candidate is carnitine palmitoyltransferase-1B (CPT1B), which is a rate-limiting step in FAO. The present study was undertaken to determine if the differential response of skeletal muscle CPT1B gene transcription to lipid between lean and severely obese subjects is linked to epigenetic modifications (DNA methylation and histone acetylation) that impact transcriptional activation. In primary human skeletal muscle cultures the expression of CPT1B was blunted in severely obese women compared with their lean counterparts in response to lipid, which was accompanied by changes in CpG methylation, H3/H4 histone acetylation, and peroxisome proliferator-activated receptor-δ and hepatocyte nuclear factor 4α transcription factor occupancy at the CPT1B promoter. Methylation of specific CpG sites in the CPT1B promoter that correlated with CPT1B transcript level blocked the binding of the transcription factor upstream stimulatory factor, suggesting a potential causal mechanism. These findings indicate that epigenetic modifications may play important roles in the regulation of CPT1B in response to a physiologically relevant lipid mixture in human skeletal muscle, a major site of fatty acid catabolism, and that differential DNA methylation may underlie the depressed expression of CPT1B in response to lipid, contributing to the metabolic inflexibility associated with severe obesity.

SeqGL Identifies Context-Dependent Binding Signals in Genome-Wide Regulatory Element Maps.

Genome-wide maps of transcription factor (TF) occupancy and regions of open chromatin implicitly contain DNA sequence signals for multiple factors. We present SeqGL, a novel de novo motif discovery algorithm to identify multiple TF sequence signals from ChIP-, DNase-, and ATAC-seq profiles. SeqGL trains a discriminative model using a k-mer feature representation together with group lasso regularization to extract a collection of sequence signals that distinguish peak sequences from flanking regions. Benchmarked on over 100 ChIP-seq experiments, SeqGL outperformed traditional motif discovery tools in discriminative accuracy. Furthermore, SeqGL can be naturally used with multitask learning to identify genomic and cell-type context determinants of TF binding. SeqGL successfully scales to the large multiplicity of sequence signals in DNase- or ATAC-seq maps. In particular, SeqGL was able to identify a number of ChIP-seq validated sequence signals that were not found by traditional motif discovery algorithms. Thus compared to widely used motif discovery algorithms, SeqGL demonstrates both greater discriminative accuracy and higher sensitivity for detecting the DNA sequence signals underlying regulatory element maps. SeqGL is available at http://cbio.mskcc.org/public/Leslie/SeqGL/.

Abstract 62: Molecular Basis of Regulatory Variation at Coronary Heart Disease-Associated Loci

Genome-wide association studies have identified 46 replicated genetic loci for coronary heart disease (CHD), and 104 loci associated at a 5% false discovery rate. However, the regulatory mechanisms of these associations largely remain elusive. Given that the majority of these CHD-associated loci reside in non-coding regions, they are predicted to function via context-specific gene regulation. Recent high-throughput assays of regulatory function include the assay for transposase-accessible chromatin using sequencing (ATAC-seq) and chromatin immunoprecipitation-sequencing (ChIP-seq). ATAC-seq utilizes a Tn5 transposase to fragment and tag accessible DNA sequences, which are often coupled to transcription factor occupancy identified by ChIP-seq. Importantly, this assay may reveal the spatio-temporal regulatory profiles in limited numbers of primary cells. Using ATAC-seq in human coronary artery smooth muscle cells (HCASMC) we identified 147,173 accessible chromatin peaks in control versus 198,976 peaks in TGF-beta-stimulated cells (136,446 shared peaks). Using de novo motif enrichment analysis we identified significant enrichment of specific AP-1 family members (29.2% vs. 5.1% background), chromatin remodeling, and SMC differentiation transcription factors. Using functional enrichment analysis of ChIP-seq and CHD-overlapping regions we observed enrichment of the hypoxia inducible factor 1 (HIF-1) and TGF-beta signaling pathways (1.5x10 -22 and 5.6x10 -18 , respectively) and relevant phenotypes, including cell migration and blood vessel morphology. Finally, we utilized these regulatory maps to explore the causal mechanisms underlying CHD-associated variants at four loci using haplotype-specific chromatin immunoprecipitation (haploChIP) and luciferase reporter assays. Taken together, these results suggest that genome-wide approaches such as ATAC-seq can be leveraged to map context-specific regulatory mechanisms of non-coding variants associated with complex diseases such as CHD, and reveal new biological and molecular insights into targeting heritable disease risk.

Occupancy by key transcription factors is a more accurate predictor of enhancer activity than histone modifications or chromatin accessibility.

BackgroundRegulated gene expression controls organismal development, and variation in regulatory patterns has been implicated in complex traits. Thus accurate prediction of enhancers is important for further understanding of these processes. Genome-wide measurement of epigenetic features, such as histone modifications and occupancy by transcription factors, is improving enhancer predictions, but the contribution of these features to prediction accuracy is not known. Given the importance of the hematopoietic transcription factor TAL1 for erythroid gene activation, we predicted candidate enhancers based on genomic occupancy by TAL1 and measured their activity. Contributions of multiple features to enhancer prediction were evaluated based on the results of these and other studies.ResultsTAL1-bound DNA segments were active enhancers at a high rate both in transient transfections of cultured cells (39 of 79, or 56%) and transgenic mice (43 of 66, or 65%). The level of binding signal for TAL1 or GATA1 did not help distinguish TAL1-bound DNA segments as active versus inactive enhancers, nor did the density of regulation-related histone modifications. A meta-analysis of results from this and other studies (273 tested predicted enhancers) showed that the presence of TAL1, GATA1, EP300, SMAD1, H3K4 methylation, H3K27ac, and CAGE tags at DNase hypersensitive sites gave the most accurate predictors of enhancer activity, with a success rate over 80% and a median threefold increase in activity. Chromatin accessibility assays and the histone modifications H3K4me1 and H3K27ac were sensitive for finding enhancers, but they have high false positive rates unless transcription factor occupancy is also included.ConclusionsOccupancy by key transcription factors such as TAL1, GATA1, SMAD1, and EP300, along with evidence of transcription, improves the accuracy of enhancer predictions based on epigenetic features.Electronic supplementary materialThe online version of this article (doi:10.1186/s13072-015-0009-5) contains supplementary material, which is available to authorized users.

Native Elongating Transcript Sequencing Reveals Human Transcriptional Activity at Nucleotide Resolution

Major features of transcription by human RNA polymerase II (Pol II) remain poorly defined due to a lack of quantitative approaches for visualizing Pol II progress at nucleotide resolution. We developed a simple and powerful approach for performing native elongating transcript sequencing (NET-seq) in human cells that globally maps strand-specific Pol II density at nucleotide resolution. NET-seq exposes a mode of antisense transcription that originates downstream and converges on transcription from the canonical promoter. Convergent transcription is associated with a distinctive chromatin configuration and is characteristic of lower-expressed genes. Integration of NET-seq with genomic footprinting data reveals stereotypic Pol II pausing coincident with transcription factor occupancy. Finally, exons retained in mature transcripts display Pol II pausing signatures that differ markedly from skipped exons, indicating an intrinsic capacity for Pol II to recognize exons with different processing fates. Together, human NET-seq exposes the topography and regulatory complexity of human gene expression.

High Resolution Architecture of Human Transcriptional Activity Revealed by Native Elongating Transcript Sequencing

Major features of transcription by human RNA polymerase II (Pol II) including pausing, antisense transcription, and alternative splicing remain poorly defined due to the lack of quantitative approaches for visualizing RNA synthesis at nucleotide resolution. We developed a powerful approach for performing native elongating transcript sequencing (NET‐seq) in human cells that exploits a quantitative biochemical purification of engaged Pol II along with the associated nascent RNA. This approach allows the genome‐wide mapping of Pol II positions in a strand‐specific manner and at nucleotide resolution. Application of NET‐seq exposes and classifies the high‐resolution transcriptional architecture of human promoter‐proximal regions. We uncover a novel mode of antisense transcription that originates downstream and converges on transcription from the canonical promoter. Convergent transcription is associated with a distinctive promoter chromatin configuration, is characteristic of lower‐expressed genes and is coupled to promoter‐proximal pausing. Integration of NET‐seq with high‐resolution genomic footprinting data reveals stereotypic Pol II pausing patterns mediated by transcription factor occupancy. Finally, exons retained in mature transcripts display Pol II pausing signatures that differ markedly from skipped exons, indicating an intrinsic capacity for Pol II recognition of exon structures with different processing fates. Together, human NET‐seq exposes novel features of the topography and regulatory complexity of human gene expression.

The pluripotent regulatory circuitry connecting promoters to their long-range interacting elements

The mammalian genome harbors up to one million regulatory elements often located at great distances from their target genes. Long-range elements control genes through physical contact with promoters and can be recognized by the presence of specific histone modifications and transcription factor binding. Linking regulatory elements to specific promoters genome-wide is currently impeded by the limited resolution of high-throughput chromatin interaction assays. Here we apply a sequence capture approach to enrich Hi-C libraries for >22,000 annotated mouse promoters to identify statistically significant, long-range interactions at restriction fragment resolution, assigning long-range interacting elements to their target genes genome-wide in embryonic stem cells and fetal liver cells. The distal sites contacting active genes are enriched in active histone modifications and transcription factor occupancy, whereas inactive genes contact distal sites with repressive histone marks, demonstrating the regulatory potential of the distal elements identified. Furthermore, we find that coregulated genes cluster nonrandomly in spatial interaction networks correlated with their biological function and expression level. Interestingly, we find the strongest gene clustering in ES cells between transcription factor genes that control key developmental processes in embryogenesis. The results provide the first genome-wide catalog linking gene promoters to their long-range interacting elements and highlight the complex spatial regulatory circuitry controlling mammalian gene expression.

An integrative analysis of TFBS-clustered regions reveals new transcriptional regulation models on the accessible chromatin landscape

DNase I hypersensitive sites (DHSs) define the accessible chromatin landscape and have revolutionised the discovery of distinct cis-regulatory elements in diverse organisms. Here, we report the first comprehensive map of human transcription factor binding site (TFBS)-clustered regions using Gaussian kernel density estimation based on genome-wide mapping of the TFBSs in 133 human cell and tissue types. Approximately 1.6 million distinct TFBS-clustered regions, collectively spanning 27.7% of the human genome, were discovered. The TFBS complexity assigned to each TFBS-clustered region was highly correlated with genomic location, cell selectivity, evolutionary conservation, sequence features, and functional roles. An integrative analysis of these regions using ENCODE data revealed transcription factor occupancy, transcriptional activity, histone modification, DNA methylation, and chromatin structures that varied based on TFBS complexity. Furthermore, we found that we could recreate lineage-branching relationships by simple clustering of the TFBS-clustered regions from terminally differentiated cells. Based on these findings, a model of transcriptional regulation determined by TFBS complexity is proposed.

Amplification of pico-scale DNA mediated by bacterial carrier DNA for small-cell-number transcription factor ChIP-seq

BackgroundChromatin-Immunoprecipitation coupled with deep sequencing (ChIP-seq) is used to map transcription factor occupancy and generate epigenetic profiles genome-wide. The requirement of nano-scale ChIP DNA for generation of sequencing libraries has impeded ChIP-seq on in vivo tissues of low cell numbers.ResultsWe describe a robust, simple and scalable methodology for ChIP-seq of low-abundant cell populations, verified down to 10,000 cells. By employing non-mammalian genome mapping bacterial carrier DNA during amplification, we reliably amplify down to 50 pg of ChIP DNA from transcription factor (CEBPA) and histone mark (H3K4me3) ChIP. We further demonstrate that genomic profiles are highly resilient to changes in carrier DNA to ChIP DNA ratios.ConclusionsThis represents a significant advance compared to existing technologies, which involve either complex steps of pre-selection for nucleosome-containing chromatin or pre-amplification of precipitated DNA, making them prone to introduce experimental biases.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-014-1195-4) contains supplementary material, which is available to authorized users.

Reassessing the "duon" hypothesis of protein evolution.

There are two distinct types of DNA sequences, namely coding sequences and regulatory sequences, in a genome. A recent study of the occupancy of transcription factors (TFs) in human cells suggested that protein-coding sequences also serve as the codes of TF occupancy, and proposed a "duon" hypothesis in which up to 15% of codons of human protein genes are constrained by the additional coding requirements that regulate gene expression. This hypothesis challenges our basic understanding on the human genome. We reanalyzed the data and found that the previous study was confounded by ascertainment bias related to base composition. Using an unbiased comparison in which G/C and A/T sites are considered separately, we reveal a similar level of conservation between TF-bound codons and TF-depleted codons, suggesting largely no extra purifying selection provided by the TF occupancy on the codons of human genes. Given the generally short binding motifs of TFs and the open chromatin structure during transcription, we argue that the occupancy of TFs on protein-coding sequences is mostly passive and evolutionarily neutral, with to-be-determined functions in the regulation of gene expression.