Transcriptional interference: an unexpected layer of complexity in gene regulation

Systematic decoding of cis gene regulation defines context-dependent control of the multi-gene costimulatory receptor locus in human T cells

Cis-regulatory elements (CREs) interact with trans regulators to orchestrate gene expression, but how transcriptional regulation is coordinated in multi-gene loci has not been experimentally defined. We sought to characterize the CREs controlling dynamic expression of the adjacent costimulatory genes CD28, CTLA4 and ICOS, encoding regulators of T cell-mediated immunity. Tiling CRISPR interference (CRISPRi) screens in primary human T cells, both conventional and regulatory subsets, uncovered gene-, cell subset- and stimulation-specific CREs. Integration with CRISPR knockout screens and assay for transposase-accessible chromatin with sequencing (ATAC-seq) profiling identified trans regulators influencing chromatin states at specific CRISPRi-responsive elements to control costimulatory gene expression. We then discovered a critical CCCTC-binding factor (CTCF) boundary that reinforces CRE interaction with CTLA4 while also preventing promiscuous activation of CD28. By systematically mapping CREs and associated trans regulators directly in primary human T cell subsets, this work overcomes longstanding experimental limitations to decode context-dependent gene regulatory programs in a complex, multi-gene locus critical to immune homeostasis.

HIV-Tocky system to visualize proviral expression dynamics

Determinants of HIV-1 latency establishment are yet to be elucidated. HIV reservoir comprises a rare fraction of infected cells that can survive host and virus-mediated killing. In vitro reporter models so far offered a feasible means to inspect this population, but with limited capabilities to dissect provirus silencing dynamics. Here, we describe a new HIV reporter model, HIV-Timer of cell kinetics and activity (HIV-Tocky) with dual fluorescence spontaneous shifting to reveal provirus silencing and reactivation dynamics. This unique feature allows, for the first time, identifying two latent populations: a directly latent, and a recently silenced subset, with the latter having integration features suggestive of stable latency. Our proposed model can help address the heterogeneous nature of HIV reservoirs and offers new possibilities for evaluating eradication strategies.

Identification of critical lncRNAs for milk fat metabolism in dairy cows using WGCNA and the construction of a ceRNAs network

As key regulators, long non-coding RNAs (lncRNAs) play a crucial role in the ruminant mammary gland. However, the function of lncRNAs in milk fat synthesis from dairy cows is largely unknown. In this study, we used the weighted gene co-expression network analysis (WGCNA) to comprehensive analyze the expression profile data of lncRNAs from the group's previous Illumina PE150 sequencing results based on bovine mammary epithelial cells from high- and low-milk-fat-percentage (MFP) cows, and identify core_lncRNAs significantly associated with MFP by module membership (MM) and gene significance (GS). Functional enrichment analysis (Gene Ontology, Kyoto Encyclopedia of Genes and Genomes) of core_lncRNA target genes (co-localization and co-expression) was performed to screen potential lncRNAs regulating milk fat metabolism and further construct an interactive regulatory network of lipid metabolism-related competing endogenous RNAs (ceRNAs). A total of 4876 lncRNAs were used to construct the WGCNA. The MEdarkturquoise module among the 19 modules obtained was significantly associated with MFP (r = 0.78, p-value <0.05) and contained 64 core_lncRNAs (MM > 0.8, GS > 0.4). Twenty-four lipid metabolism-related lncRNAs were identified by core_lncRNA target gene enrichment analysis. TCONS_00054233, TCONS_00152292, TCONS_00048619, TCONS_00033839, TCONS_00153791 and TCONS_00074642 were key candidate lncRNAs for regulating milk fat synthesis. The 22 ceRNAs most likely to be involved in milk fat metabolism were constructed by interaction network analysis, and TCONS_00133813 and bta-miR-2454-5p were located at the network's core. TCONS_00133813_bta-miR-2454-5p_TNFAIP3, TCONS_00133813_bta-miR-2454-5p_ARRB1 and TCONS_00133813_bta-miR-2454-5p_PIK3R1 are key candidate ceRNAs associated with milk fat metabolism. This study provides a framework for the co-expression module of MFP-related lncRNAs in ruminants, identifies several major lncRNAs and ceRNAs that influence milk fat synthesis, and provides a new understanding of the complex biology of milk fat synthesis in dairy cows.

Potential multi-modal effects of provirus integration on HIV-1 persistence: lessons from other viruses

Despite antiretroviral therapy (ART), HIV-1 persists as proviruses integrated into the genomic DNA of CD4⁺ T cells. The mechanisms underlying the persistence and clonal expansion of these cells remain incompletely understood. Cases have been described in which proviral integration can alter host gene expression to drive cellular proliferation. Here, we review observations from other genome-integrating human viruses to propose additional putative modalities by which HIV-1 integration may alter cellular function to favor persistence, such as by altering susceptibility to cytotoxicity in virus-expressing cells. We propose that signals implicating such mechanisms may have been masked thus far by the preponderance of defective and/or nonreactivatable HIV-1 proviruses, but could be revealed by focusing on the integration sites of intact proviruses with expression potential.

Emerging Functions of lncRNA Loci beyond the Transcript Itself

Thousands of long noncoding RNAs (lncRNAs) are actively transcribed in mammalian genomes. This class of RNAs has important regulatory functions in a broad range of cellular processes and diseases. Numerous lncRNAs have been demonstrated to mediate gene regulation through RNA-based mechanisms. Simultaneously, non-functional lncRNA transcripts derived from the activity of lncRNA loci have been identified, which underpin the notion that a considerable fraction of lncRNA loci exert regulatory functions through mechanisms associated with the production or the activity of lncRNA loci beyond the synthesized transcripts. We particularly distinguish two main RNA-independent components associated with regulatory effects; the act of transcription and the activity of DNA regulatory elements. We describe the experimental approaches to distinguish and understand the functional mechanisms derived from lncRNA loci. These scenarios reveal emerging mechanisms important to understanding the lncRNA implications in genome biology.

Who let the DoGs out? - biogenesis of stress-induced readthrough transcripts

Readthrough transcription caused by inefficient 3'-end cleavage of nascent mRNAs has emerged as a hallmark of the mammalian cellular stress response and results in the production of long noncoding RNAs known as downstream-of-gene (DoG)-containing transcripts. DoGs arise from around 10% of human protein-coding genes and are retained in the nucleus. They are produced minutes after cell exposure to stress and can be detected hours after stress removal. However, their biogenesis and the role(s) that DoGs or their production play in the cellular stress response are incompletely understood. We discuss findings that implicate host and viral proteins in the mechanisms underlying DoG production, as well as the transcriptional landscapes that accompany DoG induction under different stress conditions.

SINEUPs: a novel toolbox for RNA therapeutics

RNA molecules have emerged as a new class of promising therapeutics to expand the range of druggable targets in the genome. In addition to ‘canonical’ protein-coding mRNAs, the emerging richness of sense and antisense long non-coding RNAs (lncRNAs) provides a new reservoir of molecular tools for RNA-based drugs. LncRNAs are composed of modular structural domains with specific activities involving the recruitment of protein cofactors or directly interacting with nucleic acids. A single therapeutic RNA transcript can then be assembled combining domains with defined secondary structures and functions, and antisense sequences specific for the RNA/DNA target of interest.

As the first representative molecules of this new pharmacology, we have identified SINEUPs, a new functional class of natural antisense lncRNAs that increase the translation of partially overlapping mRNAs. Their activity is based on the combination of two domains: an embedded mouse inverted SINEB2 element that enhances mRNA translation (effector domain) and an overlapping antisense region that provides specificity for the target sense transcript (binding domain). By genetic engineering, synthetic SINEUPs can potentially target any mRNA of interest increasing translation and therefore the endogenous level of the encoded protein.

In this review, we describe the state-of-the-art knowledge of SINEUPs and discuss recent publications showing their potential application in diseases where a physiological increase of endogenous protein expression can be therapeutic.

Hyperosmotic stress alters the RNA polymerase II interactome and induces readthrough transcription despite widespread transcriptional repression

Stress-induced readthrough transcription results in the synthesis of downstream-of-gene (DoG)-containing transcripts. The mechanisms underlying DoG formation during cellular stress remain unknown. Nascent transcription profiles during DoG induction in human cell lines using TT-TimeLapse sequencing revealed widespread transcriptional repression upon hyperosmotic stress. Yet, DoGs are produced regardless of the transcriptional level of their upstream genes. ChIP sequencing confirmed that stress-induced redistribution of RNA polymerase (Pol) II correlates with the transcriptional output of genes. Stress-induced alterations in the Pol II interactome are observed by mass spectrometry. While certain cleavage and polyadenylation factors remain Pol II associated, Integrator complex subunits dissociate from Pol II under stress leading to a genome-wide loss of Integrator on DNA. Depleting the catalytic subunit of Integrator using siRNAs induces hundreds of readthrough transcripts, whose parental genes partially overlap those of stress-induced DoGs. Our results provide insights into the mechanisms underlying DoG production and how Integrator activity influences DoG transcription.

The Temperature-Dependent Retention of Introns in GPI8 Transcripts Contributes to a Drooping and Fragile Shoot Phenotype in Rice

Attachment of glycosylphosphatidylinositols (GPIs) to the C-termini of proteins is one of the most common posttranslational modifications in eukaryotic cells. GPI8/PIG-K is the catalytic subunit of the GPI transamidase complex catalyzing the transfer en bloc GPI to proteins. In this study, a T-DNA insertional mutant of rice with temperature-dependent drooping and fragile (df) shoots phenotype was isolated. The insertion site of the T-DNA fragment was 879 bp downstream of the stop codon of the OsGPI8 gene, which caused introns retention in the gene transcripts, especially at higher temperatures. A complementation test confirmed that this change in the OsGPI8 transcripts was responsible for the mutant phenotype. Compared to control plants, internodes of the df mutant showed a thinner shell with a reduced cell number in the transverse direction, and an inhomogeneous secondary wall layer in bundle sheath cells, while many sclerenchyma cells at the tops of the main veins of df leaves were shrunken and their walls were thinner. The df plants also displayed a major reduction in cellulose and lignin content in both culms and leaves. Our data indicate that GPI anchor proteins play important roles in biosynthesis and accumulation of cell wall material, cell shape, and cell division in rice.

iRAPs curb antisense transcription in E. coli

RNA polymerase-binding RNA aptamers (RAPs) are natural RNA elements that control transcription in cis by directly contacting RNA polymerase. Many RAPs inhibit transcription by inducing Rho-dependent termination in Escherichia coli. Here, we studied the role of inhibitory RAPs (iRAPs) in modulation of antisense transcription (AT) using in silico and in vivo approaches. We revisited the antisense transcriptome in cells with impaired AT regulators (Rho, H-NS and RNaseIII) and searched for the presence of RAPs within antisense RNAs. Many of these RAPs were found at key genomic positions where they terminate AT. By exploring the activity of several RAPs both in a reporter system and in their natural genomic context, we confirmed their significant role in AT regulation. RAPs coordinate Rho activity at the antisense strand and terminate antisense transcripts. In some cases, they stimulated sense expression by alleviating ongoing transcriptional interference. Essentially, our data postulate RAPs as key determinants of Rho-mediated AT regulation in E. coli.

The Eleanor ncRNAs activate the topological domain of the ESR1 locus to balance against apoptosis

MCF7 cells acquire estrogen-independent proliferation after long-term estrogen deprivation (LTED), which recapitulates endocrine therapy resistance. LTED cells can become primed for apoptosis, but the underlying mechanism is largely unknown. We previously reported that Eleanor non-coding RNAs (ncRNAs) upregulate the ESR1 gene in LTED cells. Here, we show that Eleanors delineate the topologically associating domain (TAD) of the ESR1 locus in the active nuclear compartment of LTED cells. The TAD interacts with another transcriptionally active TAD, which is 42.9 Mb away from ESR1 and contains a gene encoding the apoptotic transcription factor FOXO3. Inhibition of a promoter-associated Eleanor suppresses all genes inside the Eleanor TAD and the long-range interaction between the two TADs, but keeps FOXO3 active to facilitate apoptosis in LTED cells. These data indicate a role of ncRNAs in chromatin domain regulation, which may underlie the apoptosis-prone nature of therapy-resistant breast cancer cells and could be good therapeutic targets.

Long term estrogen deprivation can result in apoptosis in breast cancer cells. Here, the authors show that this apoptosis is induced by the long-range chromatin interaction of loci containing the ESR1 and FOXO3 genes, resulting in FOXO3-mediated apoptosis.

Non-coding RNAs: Functional roles in the regulation of stress response in Brassica crops

Brassica crops face a combination of different abiotic and biotic stresses in the field that can reduce plant growth and development by affecting biochemical and morpho-physiological processes. Emerging evidence suggests that non-coding RNAs (ncRNAs), especially microRNAs (miRNAs) and long ncRNAs (lncRNAs), play a significant role in the modulation of gene expression in response to plant stresses. Recent advances in computational and experimental approaches are of great interest for identifying and functionally characterizing ncRNAs. While progress in this field is limited, numerous ncRNAs involved in the regulation of gene expression in response to stress have been reported in Brassica. In this review, we summarize the modes of action and functions of stress-related miRNAs and lncRNAs in Brassica as well as the approaches used to identify ncRNAs.

Transcription termination sequences support the expression of transgene product secreted with milk

Expression of the reporter gene in transgenic animals depends on the surrounding chromatin environment. Recent genome-wide studies have shown that, in mammals, the entire genome is transcribed. Transcription through a transgene often has a negative effect on the expression of a reporter gene. Here, we compared the ability of well-studied chicken chromatin insulator HS4 and bidirectional transcription terminators from the human genome to support high-level expression of the firefly luciferase gene (Fluc) under control of the previously characterized goat β-casein gene promoter. The insertion of HS4 or either of the two transcription terminators upstream of the promoter resulted in tenfold enhancement of Fluc expression in the mammary glands of transgenic mice. These results suggest that transcriptional terminators, similar to the HS4 insulator, can be used to improve the reporter gene expression in transgenic animals.

Discovery and expression analysis of novel transcripts of the human SR-related CTD-associated factor 1 (SCAF1) gene in human cancer cells using Next-Generation Sequencing

The human SR-related CTD associated factor 1 (SCAF1) gene is a new member of the human SR (Ser/Arg-rich) superfamily of pre-mRNA splicing factors, which has been discovered and cloned by members of our lab. SCAF1 interacts with the CTD domain of the RNA polymerase II polypeptide A and is firmly involved in pre-mRNA splicing. Although it was found to be expressed widely in multiple human tissues, its mRNA levels vary a lot. The significant relation of SCAF1 with cancer has been confirmed by many studies, since SCAF1 mRNA transcript was found to be overexpressed in breast and ovarian tumors, confirming its significant prognostic value as a cancer biomarker in both these human malignancies. In this study, we describe the discovery and cloning of fifteen novel transcripts of the human SCAF1 gene (SCAF1 v.2 - v.16), using nested PCR and NGS technology. In detail, extensive bioinformatic analysis revealed that these novel SCAF1 splice variants comprise a total of nine novel alternative splicing events between the annotated exons of the gene, thus producing seven novel SCAF1 transcripts with open-reading frames, which are predicted to encode novel SCAF1 isoforms and eight novel SCAF1 transcripts with premature termination codons that are likely long non-coding RNAs. Additionally, a novel 3' UTR was discovered and cloned using nested 3' RACE and was validated with Sanger sequencing. In order to validate the NGS findings as well as to investigate the expression profile of each novel transcript, RT-PCR experiments were carried out with the use of variant-specific primers. Since SCAF1 is implicated in many human malignancies, qualifying as a potential biomarker, the quantification of the presented novel transcripts in human samples may have clinical applications in different types of cancer.

Genome-wide identification of miRNAs and lncRNAs in Cajanus cajan

Background

Non-coding RNAs (ncRNAs) are important players in the post transcriptional regulation of gene expression (PTGR). On one hand, microRNAs (miRNAs) are an abundant class of small ncRNAs (~22nt long) that negatively regulate gene expression at the levels of messenger RNAs stability and translation inhibition, on the other hand, long ncRNAs (lncRNAs) are a large and diverse class of transcribed non-protein coding RNA molecules (> 200nt) that play both up-regulatory as well as down-regulatory roles at the transcriptional level. Cajanus cajan, a leguminosae pulse crop grown in tropical and subtropical areas of the world, is a source of high value protein to vegetarians or very poor populations globally. Hence, genome-wide identification of miRNAs and lncRNAs in C. cajan is extremely important to understand their role in PTGR with a possible implication to generate improve variety of crops.

Results

We have identified 616 mature miRNAs in C. cajan belonging to 118 families, of which 578 are novel and not reported in MirBase21. A total of 1373 target sequences were identified for 180 miRNAs. Of these, 298 targets were characterized at the protein level. Besides, we have also predicted 3919 lncRNAs. Additionally, we have identified 87 of the predicted lncRNAs to be targeted by 66 miRNAs.

Conclusions

miRNA and lncRNAs in plants are known to control a variety of traits including yield, quality and stress tolerance. Owing to its agricultural importance and medicinal value, the identified miRNA, lncRNA and their targets in C. cajan may be useful for genome editing to improve better quality crop. A thorough understanding of ncRNA-based cellular regulatory networks will aid in the improvement of C. cajan agricultural traits.

Electronic supplementary material

The online version of this article (10.1186/s12864-017-4232-2) contains supplementary material, which is available to authorized users.

Transposable elements and circular DNAs

Circular DNAs are extra-chromosomal fragments that become circularized by genomic recombination events. We have recently shown that yeast LTR elements generate circular DNAs through recombination events between their flanking long terminal repeats (LTRs). Similarly, circular DNAs can be generated by recombination between LTRs residing at different genomic loci, in which case the circular DNA will contain the intervening sequence. In yeast, this can result in gene copy number variations when circles contain genes and origins of replication. Here, I speculate on the potential and implications of circular DNAs generated through recombination between human transposable elements.

Association of Long Non-Coding RNA HOTAIR Polymorphisms with Cervical Cancer Risk in a Chinese Population

Long non-coding RNAs (lncRNAs), HOTAIR has been reported to be upregulated in cervical cancer development and progression. However, SNPs (single nucleotide polymorphisms) in the lncRNAs and their associations with cervical cancer susceptibility have not been reported. In the current study, we hypothesized that SNPs within the lncRNA HOTAIR may influence the risk of cervical cancer. We performed a case-control study including 510 cervical cancer patients (cases) and 713 cancer-free individuals (controls) to investigate the association between three haplotype-tagging SNPs (rs920778, rs1899663 and rs4759314) in the lncRNA HOTAIR and the risk of cervical cancer. We found a strong association between the SNP rs920778 in the intronic enhancer of the HOTAIR and cervical cancer (P<10⁻⁴). Moreover, the cervical cancer patients with homozygous TT genotype were significantly associated with tumor-node-metastasis (TNM) stage. In vitro assays with allele-specific reporter constructs indicated that the reporter constructs bearing rs920778T allele conferred elevated reporter gene transcriptional activity when compared to the reporter constructs containing rs920778C allele. Furthermore, HOTAIR expression was higher in cervical cancer tissues than that in corresponding normal tissues, and the high expression was associated with the risk-associated allele T. In summary, our studies provide strong functional evidence that functional SNP rs920778 regulates HOTAIR expression, and may ultimately influence the predisposition for cervical cancer.

The role of repressor kinetics in relief of transcriptional interference between convergent promoters

Transcriptional interference (TI), where transcription from a promoter is inhibited by the activity of other promoters in its vicinity on the same DNA, enables transcription factors to regulate a target promoter indirectly, inducing or relieving TI by controlling the interfering promoter. For convergent promoters, stochastic simulations indicate that relief of TI can be inhibited if the repressor at the interfering promoter has slow binding kinetics, making it either sensitive to frequent dislodgement by elongating RNA polymerases (RNAPs) from the target promoter, or able to be a strong roadblock to these RNAPs. In vivo measurements of relief of TI by CI or Cro repressors in the bacteriophage λ PR-PRE system show strong relief of TI and a lack of dislodgement and roadblocking effects, indicative of rapid CI and Cro binding kinetics. However, repression of the same λ promoter by a catalytically dead CRISPR Cas9 protein gave either compromised or no relief of TI depending on the orientation at which it binds DNA, consistent with dCas9 being a slow kinetics repressor. This analysis shows how the intrinsic properties of a repressor can be evolutionarily tuned to set the magnitude of relief of TI.

An antisense RNA-mediated mechanism eliminates a meiosis-specific copper-regulated transcript in mitotic cells

Sense and antisense transcripts produced from convergent gene pairs could interfere with the expression of either partner gene. In Schizosaccharomyces pombe, we found that the iss1(+) gene produces two transcript isoforms, including a long antisense mRNA that is complementary to the meiotic cum1(+) sense transcript, inhibiting cum1(+) expression in vegetative cells. Inhibition of cum1(+) transcription was not at the level of its initiation because fusion of the cum1(+) promoter to the lacZ gene showed that activation of the reporter gene occurs in response to low copper conditions. Further analysis showed that the transcription factor Cuf1 and conserved copper-signaling elements (CuSEs) are required for induction of cum1(+)-lacZ transcription under copper deficiency. Insertion of a multipartite polyadenylation signal immediately downstream of iss1(+) led to the exclusive production of a shorter iss1(+) mRNA isoform, thereby allowing accumulation of cum1(+) sense mRNA in copper-limited vegetative cells. This finding suggested that the long iss1(+) antisense mRNA could pair with cum1(+) sense mRNA, thereby producing double-stranded RNA molecules that could induce RNAi. We consistently found that mutant strains for RNAi (dcr1Δ, ago1Δ, rdp1Δ, and clr4Δ) are defective in selectively eliminating cum1(+) sense transcript in the G1 phase of the cell cycle. Taken together, these results describe the first example of a copper-regulated meiotic gene repressed by an antisense transcription mechanism in vegetative cells.

Non Coding RNA Molecules as Potential Biomarkers in Breast Cancer

The pursuit of minimally invasive biomarkers is a challenging but exciting area of research. Clearly, such markers would need to be sensitive and specific enough to aid in the detection of breast cancer at an early stage, would monitor progression of the disease, and could predict the individual patient's response to treatment. Unfortunately, to date, markers with such characteristics have not made it to the clinic for breast cancer. Past years, many studies indicated that the non-coding part of our genome (the so called 'junk' DNA), may be an ideal source for these biomarkers. In this chapter, the potential use of microRNAs and long non-coding RNAs as biomarkers will be discussed.

Deep sequencing reveals a novel class of bidirectional promoters associated with neuronal genes

Background

Comprehensive annotation of transcripts expressed in a given tissue is a critical step towards the understanding of regulatory and functional pathways that shape the transcriptome.

Results

Here, we reconstructed a cumulative transcriptome of the human prefrontal cortex (PFC) based on approximately 300 million strand-specific RNA sequence (RNA-seq) reads collected at different stages of postnatal development. We find that more than 50% of reconstructed transcripts represent novel transcriptome elements, including 8,343 novel exons and exon extensions of annotated coding genes, 11,217 novel antisense transcripts and 29,541 novel intergenic transcripts or their fragments showing canonical features of long non-coding RNAs (lncRNAs). Our analysis further led to a surprising discovery of a novel class of bidirectional promoters (NBiPs) driving divergent transcription of mRNA and novel lncRNA pairs and displaying a distinct set of sequence and epigenetic features. In contrast to known bidirectional and unidirectional promoters, NBiPs are strongly associated with genes involved in neuronal functions and regulated by neuron-associated transcription factors.

Conclusions

Taken together, our results demonstrate that large portions of the human transcriptome remain uncharacterized. The distinct sequence and epigenetic features of NBiPs, as well as their specific association with neuronal genes, further suggest existence of regulatory pathways specific to the human brain.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-457) contains supplementary material, which is available to authorized users.

Suppression of expression between adjacent genes within heterologous modules in yeast

Recent studies have shown that proximal arrangement of multiple genes can have complex effects on gene expression. For example, in the case of heterologous gene expression modules, certain arrangements of the selection marker and the gene expression cassette may have unintended consequences that limit the predictability and interpretability of module behaviors. The relationship between arrangement and expression has not been systematically characterized within heterologous modules to date. In this study, we quantitatively measured gene expression patterns of the selection marker (KlURA3 driven by the promoter, pKlURA) and the gene expression cassette (GFP driven by the galactose-inducible GAL1 promoter, pGAL1) in all their possible relative arrangements in Saccharomyces cerevisiae. First, we observed that pKlURA activity depends strongly on the relative arrangement and the activity of pGAL1. Most notably, we observed transcriptional suppression in the case of divergent arrangements: pKlURA activity was reduced when pGAL1 was inactive. Based on our nucleosome occupancy data, we attribute the observed transcriptional reduction to nucleosome repositioning. Second, we observed that pGAL1 activity also depends on the relative arrangement of pKlURA. In particular, strains with divergent promoters showed significantly different pGAL1 activation patterns from other strains, but only when their growth was compromised by lack of uracil. We reasoned that this difference in pGAL1 activation patterns arises from arrangement-dependent pKlURA activity that can affect the overall cell physiology (i.e., cell growth and survival in the uracil-depleted condition). Our results underscore the necessity to consider ramifications of promoter arrangement when using synthetic gene expression modules.

Computational analysis of conserved RNA secondary structure in transcriptomes and genomes

Transcriptomics experiments and computational predictions both enable systematic discovery of new functional RNAs. However, many putative noncoding transcripts arise instead from artifacts and biological noise, and current computational prediction methods have high false positive rates. I discuss prospects for improving computational methods for analyzing and identifying functional RNAs, with a focus on detecting signatures of conserved RNA secondary structure. An interesting new front is the application of chemical and enzymatic experiments that probe RNA structure on a transcriptome-wide scale. I review several proposed approaches for incorporating structure probing data into the computational prediction of RNA secondary structure. Using probabilistic inference formalisms, I show how all these approaches can be unified in a well-principled framework, which in turn allows RNA probing data to be easily integrated into a wide range of analyses that depend on RNA secondary structure inference. Such analyses include homology search and genome-wide detection of new structural RNAs.

Emerging epigenetic mechanisms of long non-coding RNAs

Long non-coding RNAs (lncRNAs) have been increasingly appreciated as an integral component of gene regulatory networks. Genome-wide features of their origin and expression patterns ascribed a prominent role for lncRNAs to the regulation of protein-coding genes, and also suggest a potential link to many human diseases. Recent studies have begun to unravel the intricate regulatory mechanism of lncRNAs occurring at multiple levels. The brain is one of the richest sources of lncRNAs, many of which have already shown a close relationship with genes or genetic loci implicated in a wide range of neurological disorders. This review describes recently emerging mechanistic principles of lncRNA functions to provide neuroscientists with molecular insights that will help future research on lncRNAs in the brain.

A unified model for yeast transcript definition

Identifying genes in the genomic context is central to a cell's ability to interpret the genome. Yet, in general, the signals used to define eukaryotic genes are poorly described. Here, we derived simple classifiers that identify where transcription will initiate and terminate using nucleic acid sequence features detectable by the yeast cell, which we integrate into a Unified Model (UM) that models transcription as a whole. The cis-elements that denote where transcription initiates function primarily through nucleosome depletion, and, using a synthetic promoter system, we show that most of these elements are sufficient to initiate transcription in vivo. Hrp1 binding sites are the major characteristic of terminators; these binding sites are often clustered in terminator regions and can terminate transcription bidirectionally. The UM predicts global transcript structure by modeling transcription of the genome using a hidden Markov model whose emissions are the outputs of the initiation and termination classifiers. We validated the novel predictions of the UM with available RNA-seq data and tested it further by directly comparing the transcript structure predicted by the model to the transcription generated by the cell for synthetic DNA segments of random design. We show that the UM identifies transcription start sites more accurately than the initiation classifier alone, indicating that the relative arrangement of promoter and terminator elements influences their function. Our model presents a concrete description of how the cell defines transcript units, explains the existence of nongenic transcripts, and provides insight into genome evolution.

Gene regulation by non-coding RNAs

The past two decades have seen an explosion in research on non-coding RNAs and their physiological and pathological functions. Several classes of small (20-30 nucleotides) and long (>200 nucleotides) non-coding RNAs have been firmly established as key regulators of gene expression in myriad processes ranging from embryonic development to innate immunity. In this review, we focus on our current understanding of the molecular mechanisms underlying the biogenesis and function of small interfering RNAs (siRNAs), microRNAs (miRNAs) and Piwi-interacting RNAs (piRNAs). In addition, we briefly review the relevance of small and long non-coding RNAs to human physiology and pathology and their potential to be exploited as therapeutic agents.

Transcription through enhancers suppresses their activity in Drosophila

BACKGROUND

Enhancer elements determine the level of target gene transcription in a tissue-specific manner, providing for individual patterns of gene expression in different cells. Knowledge of the mechanisms controlling enhancer action is crucial for understanding global regulation of transcription. In particular, enhancers are often localized within transcribed regions of the genome. A number of experiments suggest that transcription can have both positive and negative effects on regulatory elements. In this study, we performed direct tests for the effect of transcription on enhancer activity.

RESULTS

Using a transgenic reporter system, we investigated the relationship between the presence of pass-through transcription and the activity of Drosophila enhancers controlling the expression of the white and yellow genes. The results show that transcription from different promoters affects the activity of enhancers, counteracting their ability to activate the target genes. As expected, the presence of a transcriptional terminator between the inhibiting promoter and the affected enhancer strongly reduces the suppression. Moreover, transcription leads to dislodging of the Zeste protein that is responsible for the enhancer-dependent regulation of the white gene, suggesting a 'transcription interference' mechanism for this regulation.

CONCLUSIONS

Our findings suggest a role for pass-through transcription in negative regulation of enhancer activity.

Extensively duplicated and transcriptionally active recent lateral gene transfer from a bacterial Wolbachia endosymbiont to its host filarial nematode Brugia malayi

BACKGROUND

Lymphatic filariasis is a neglected tropical disease afflicting more than 120 million people, while another 1.3 billion people are at risk of infection. The nematode worm Brugia malayi is one of the causative agents of the disease and exists in a mutualistic symbiosis with Wolbachia bacteria. Since extensive lateral gene transfer occurs frequently between Wolbachia and its hosts, we sought to measure the extent of such LGT in B. malayi by whole genome sequencing of Wolbachia-depleted worms.

RESULTS

A considerable fraction (at least 115.4-kbp, or 10.6%) of the 1.08-Mbp Wolbachia wBm genome has been transferred to its nematode host and retains high levels of similarity, including 227 wBm genes and gene fragments. Complete open reading frames were transferred for 32 of these genes, meaning they have the potential to produce functional proteins. Moreover, four transfers have evidence of life stage-specific regulation of transcription at levels similar to other nematode transcripts, strengthening the possibility that they are functional.

CONCLUSIONS

There is extensive and ongoing transfer of Wolbachia DNA to the worm genome and some transfers are transcribed in a stage-specific manner at biologically relevant levels.

Long non-coding RNAs in haematological malignancies

Long non-coding RNAs (lncRNAs) are functional RNAs longer than 200 nucleotides in length. LncRNAs are as diverse as mRNAs and they normally share the same biosynthetic machinery based on RNA polymerase II, splicing and polyadenylation. However, lncRNAs have low coding potential. Compared to mRNAs, lncRNAs are preferentially nuclear, more tissue specific and expressed at lower levels. Most of the lncRNAs described to date modulate the expression of specific genes by guiding chromatin remodelling factors; inducing chromosomal loopings; affecting transcription, splicing, translation or mRNA stability; or serving as scaffolds for the organization of cellular structures. They can function in cis, cotranscriptionally, or in trans, acting as decoys, scaffolds or guides. These functions seem essential to allow cell differentiation and growth. In fact, many lncRNAs have been shown to exert oncogenic or tumor suppressor properties in several cancers including haematological malignancies. In this review, we summarize what is known about lncRNAs, the mechanisms for their regulation in cancer and their role in leukemogenesis, lymphomagenesis and hematopoiesis. Furthermore, we discuss the potential of lncRNAs in diagnosis, prognosis and therapy in cancer, with special attention to haematological malignancies.

Cellular and molecular mechanisms involved in the establishment of HIV-1 latency

Latently infected cells represent the major barrier to either a sterilizing or a functional HIV-1 cure. Multiple approaches to reactivation and depletion of the latent reservoir have been attempted clinically, but full depletion of this compartment remains a long-term goal. Compared to the mechanisms involved in the maintenance of HIV-1 latency and the pathways leading to viral reactivation, less is known about the establishment of latent infection. This review focuses on how HIV-1 latency is established at the cellular and molecular levels. We first discuss how latent infection can be established following infection of an activated CD4 T-cell that undergoes a transition to a resting memory state and also how direct infection of a resting CD4 T-cell can lead to latency. Various animal, primary cell, and cell line models also provide insights into this process and are discussed with respect to the routes of infection that result in latency. A number of molecular mechanisms that are active at both transcriptional and post-transcriptional levels have been associated with HIV-1 latency. Many, but not all of these, help to drive the establishment of latent infection, and we review the evidence in favor of or against each mechanism specifically with regard to the establishment of latency. We also discuss the role of immediate silent integration of viral DNA versus silencing of initially active infections. Finally, we discuss potential approaches aimed at limiting the establishment of latent infection.

A genome-wide investigation of expression characteristics of natural antisense transcripts in liver and muscle samples of pigs

Natural antisense transcripts are endogenous transcripts that are complementary to the sense-strand of DNA. These transcripts have been identified in various eukaryotic species and are involved in a broad range of regulatory events and biological processes. However, their general biological functions, expression characteristics and regulatory mechanisms are still unclear. In this study, 497 liver and 586 muscle samples were harvested from a White Duroc×Erhualian F₂ resource population. The expression profiles of sense and antisense transcripts were determined by tag-based RNA sequencing. We identified 33.7% and 20.4% of transcripts having both sense and antisense expression, and 12.5% and 6.1% of transcripts only expressing antisense transcripts in liver and muscle, respectively. More than 32.2% of imprinting or predicted imprinting genes in the geneimprint database were detected with both sense and antisense expression. The correlations between sense and antisense expression in sense-antisense pairs were diverse in both liver and muscle, showing positive, negative or absent correlation. Antisense expression increases gene expression variability. More interestingly, compared to eQTL mapping of sense transcripts in which more than one eQTL was mapped for a transcript, only one eQTL was identified for each antisense transcript, and the percentage of cis-eQTL in antisense eQTL was higher than that in sense eQTL. This suggests that the expressions of antisense transcripts tend to be cis-regulated by a single genomic locus. To our knowledge, this study is the first systematical investigation of antisense transcription in pigs. The findings improve our understanding of the complexity of porcine transcriptome.

Activity of the upstream component of tandem TERT/survivin promoters depends on features of the downstream component

We spliced the promoters of the human telomerase and human survivin genes (PhTERT and PhSurv, respectively) widely used for gene therapy and known to have the broadest cancer type spectrum of activity. Two head-to-tail constructs were obtained: the PhTERT-PhSurv and PhSurv-PhTERT tandems. The splicing caused quantitative and qualitative changes in the promoter features. In both constructs, only the promoter proximal to the transcribed gene retained its ability to initiate transcription, whereas the distal promoter was silent, the phenomenon never reported before. However, the distal promoter modulated the activity of the proximal one by increasing its strength and causing an appearance of additional transcription start sites. We suggested that this suppression might be due to the presence of Sp1 transcription factor binding sites in both promoters and Sp1-bridges between these sites. Such Sp1-bridges might convert the tandem promoter linear DNA into a stem-loop structure. If localized inside the formed loop, the distal promoter could lose its ability to initiate transcription. To test this hypothesis, we constructed two modified double promoters, where the proximal PhSurv promoter was replaced either by a shortened variant of the survivin promoter (PhSurv269) or by the mouse survivin promoter. Both PhSurv substitutes were considerably shorter than PhSurv and had different numbers and/or positions of Sp1 sites. In modified tandems, transcription was initiated from both promoters. We also prepared two mutant forms of the PhSurv-PhTERT tandem with two or four Sp1 sites removed from the distal "long" PhSurv promoter. In the first case, the distal PhSurv promoter remained silent, whereas the removal of four Sp1 binding sites restored its activity. In the majority of studied cancer cell lines the efficiency of transcription from the hTERT-(shortened hSurv269) promoter tandem was markedly higher than from each constituent promoter. In normal lung fibroblast cells, the tandem promoter activity was considerably lower.

The DMD locus harbours multiple long non-coding RNAs which orchestrate and control transcription of muscle dystrophin mRNA isoforms

The 2.2 Mb long dystrophin (DMD) gene, the largest gene in the human genome, corresponds to roughly 0.1% of the entire human DNA sequence. Mutations in this gene cause Duchenne muscular dystrophy and other milder X-linked, recessive dystrophinopathies. Using a custom-made tiling array, specifically designed for the DMD locus, we identified a variety of novel long non-coding RNAs (lncRNAs), both sense and antisense oriented, whose expression profiles mirror that of DMD gene. Importantly, these transcripts are intronic in origin and specifically localized to the nucleus and are transcribed contextually with dystrophin isoforms or primed by MyoD-induced myogenic differentiation. Furthermore, their forced ectopic expression in both human muscle and neuronal cells causes a specific and negative regulation of endogenous dystrophin full length isoforms and significantly down-regulate the activity of a luciferase reporter construct carrying the minimal promoter regions of the muscle dystrophin isoform. Consistent with this apparently repressive role, we found that, in muscle samples of dystrophinopathic female carriers, lncRNAs expression levels inversely correlate with those of muscle full length DMD isoforms. Overall these findings unveil an unprecedented complexity of the transcriptional pattern of the DMD locus and reveal that DMD lncRNAs may contribute to the orchestration and homeostasis of the muscle dystrophin expression pattern by either selective targeting and down-modulating the dystrophin promoter transcriptional activity.

The bromodomain-containing gene BRD2 is regulated at transcription, splicing, and translation levels

The human BRD2 gene has been linked and associated with a form of common epilepsy and electroencephalographic abnormalities. Disruption of Brd2 in the mouse revealed that it is essential for embryonic neural development and that viable Brd2(+/-) heterozygotes show both decreased GABAergic neuron counts and increased susceptibility to seizures. To understand the molecular mechanisms by which mis-expression of BRD2 might contribute to epilepsy, we examined its regulation at multiple levels. We discovered that BRD2 expresses distinct tissue-specific transcripts that originate from different promoters and have strikingly different lengths of 5' untranslated regions (5'UTR). We also experimentally confirmed the presence of a highly conserved, alternatively spliced exon, inclusion of which would result in a premature termination of translation. Downstream of this alternative exon is a polymorphic microsatellite (GT-repeats). Manipulation of the number of the GT-repeats revealed that the length of the GT-repeats affects the ratio of the two alternative splicing products. In vitro translation and expression in cultured cells revealed that among the four different mRNAs (long and short 5'UTR combined with regular and alternative splicing), only the regularly spliced mRNA with the short 5'UTR yields full-length protein. In situ hybridization and immunohistochemical studies showed that although Brd2 mRNA is expressed in both the hippocampus and cerebellum, Brd2 protein only can be detected in the cerebellar Purkinje cells and not in hippocampal cells. These multiple levels of regulation would likely affect the production of functional BRD2 protein during neural development and hence, its role in the etiology of seizure susceptibility.

Molecular Understanding of HIV-1 Latency

The introduction of highly active antiretroviral therapy (HAART) has been an important breakthrough in the treatment of HIV-1 infection and has also a powerful tool to upset the equilibrium of viral production and HIV-1 pathogenesis. Despite the advent of potent combinations of this therapy, the long-lived HIV-1 reservoirs like cells from monocyte-macrophage lineage and resting memory CD4+ T cells which are established early during primary infection constitute a major obstacle to virus eradication. Further HAART interruption leads to immediate rebound viremia from latent reservoirs. This paper focuses on the essentials of the molecular mechanisms for the establishment of HIV-1 latency with special concern to present and future possible treatment strategies to completely purge and target viral persistence in the reservoirs.

Noncoding RNAs involved in mammary gland development and tumorigenesis: there's a long way to go

The mammalian genome encodes thousands of noncoding RNAs. These noncoding transcripts are broadly categorized into short noncoding RNAs, such as microRNAs (miRNAs), and long noncoding RNAs (lncRNAs) of greater than 200 nt. While the role of miRNAs in development and cancer biology has been extensively studied, much less is known about the vast majority of noncoding transcripts represented by lncRNAs. LncRNAs are emerging as key regulators of developmental processes and as such, their frequent misregulation in tumorigenesis and disease in not unexpected. The role of lncRNAs in mammary gland development and breast cancer is just beginning to be elucidated. This review will discuss the role of lncRNAs in mammalian and mammary gland development. In addition, we will review the contributions of lncRNAs to the stepwise progression of tumorigenesis, highlighting the role of lncRNAs in breast cancer.

Tissue-specific alternative polyadenylation at the imprinted gene Mest regulates allelic usage at Copg2

The gene Mest (also known as Peg1) is regulated by genomic imprinting in the mouse and only the paternal allele is active for transcription. MEST is similarly imprinted in humans, where it is a candidate for the growth retardation Silver-Russell syndrome. The MEST protein belongs to an ancient family of hydrolases but its function is still unknown. It is highly conserved in vertebrates although imprinted expression is only observed in marsupials and eutherians, thus a recent evolutionary event. Here we describe the identification of new imprinted RNA products at the Mest locus, longer variants of the RNA, called MestXL, transcribed >10 kb into the downstream antisense gene Copg2. During development MestXL is produced exclusively in the developing central nervous system (CNS) by alternative polyadenylation. Copg2 is biallelically expressed in the embryo except in MestXL-expressing tissues, where we observed preferential expression from the maternal allele. To analyze the function of the MestXL transcripts in Copg2 regulation, we studied the effects of a targeted allele at Mest introducing a truncation in the mRNA. We show that both the formation of the MestXL isoforms and the allelic bias at Copg2 are lost in the CNS of mutants embryos. Our results propose a new mechanism to regulate allelic usage in the mammalian genome, via tissue-specific alternative polyadenylation and transcriptional interference in sense-antisense pairs at imprinted loci.

The viral protein Tat can inhibit the establishment of HIV-1 latency

The establishment of HIV-1 latency can result from limiting levels of transcription initiation or elongation factors, restrictive chromatin modifications, transcriptional interference, and insufficient Tat activity. Since the viral protein Tat can counteract many of these factors, we hypothesized that the presence of exogenous Tat during infection might inhibit the establishment of latency. This was explored using a Jurkat model of latency establishment and reactivation. PCR and reverse transcriptase PCR (RT-PCR) confirmed the latent state in this model and showed evidence of transcriptional interference. To address our hypothesis, cells undergoing infection were first exposed to either purified recombinant Tat or a transactivation-negative mutant. Only the former resulted in a modest inhibition of the establishment of latency. Next, Jurkat cells stably expressing intracellular Tat were used in our latency model to avoid limitations of Tat delivery. Experiments confirmed that intracellular Tat expression did not affect the susceptibility of these cells to viral infection. Eight weeks after infection, Jurkat cells expressing Tat harbored up to 1,700-fold fewer (P < 0.01) latent viruses than Jurkat cells that did not express Tat. Additionally, Tat delivered by a second virus was sufficient to reactivate most of the latent population. Our results suggest that inhibition of the establishment of latent infection is theoretically possible. In a hypothetical scenario of therapy that induces viral gene expression during acute infection, activation of viruses which would otherwise have entered latency could occur while concurrent highly active antiretroviral therapy (HAART) would prevent further viral spread, potentially decreasing the size of the established latent reservoir.

Regulation of the photorespiratory GLDPA gene in C(4) flaveria: an intricate interplay of transcriptional and posttranscriptional processes

The mitochondrial Gly decarboxylase complex (GDC) is a key component of the photorespiratory pathway that occurs in all photosynthetically active tissues of C(3) plants but is restricted to bundle sheath cells in C(4) species. GDC is also required for general cellular C(1) metabolism. In the Asteracean C(4) species Flaveria trinervia, a single functional GLDP gene, GLDPA, encodes the P-subunit of GDC, a decarboxylating Gly dehydrogenase. GLDPA promoter reporter gene fusion studies revealed that this promoter is active in bundle sheath cells and the vasculature of transgenic Flaveria bidentis (C(4)) and the Brassicacean C(3) species Arabidopsis thaliana, suggesting the existence of an evolutionarily conserved gene regulatory system in the bundle sheath. Here, we demonstrate that GLDPA gene regulation is achieved by an intricate interplay of transcriptional and posttranscriptional mechanisms. The GLDPA promoter is composed of two tandem promoters, P(R2) and P(R7), that together ensure a strong bundle sheath expression. While the proximal promoter (P(R7)) is active in the bundle sheath and vasculature, the distal promoter (P(R2)) drives uniform expression in all leaf chlorenchyma cells and the vasculature. An intron in the 5' untranslated leader of P(R2)-derived transcripts is inefficiently spliced and apparently suppresses the output of P(R2) by eliciting RNA decay.

Intronic L1 retrotransposons and nested genes cause transcriptional interference by inducing intron retention, exonization and cryptic polyadenylation

Background

Transcriptional interference has been recently recognized as an unexpectedly complex and mostly negative regulation of genes. Despite a relatively few studies that emerged in recent years, it has been demonstrated that a readthrough transcription derived from one gene can influence the transcription of another overlapping or nested gene. However, the molecular effects resulting from this interaction are largely unknown.

Methodology/Principal Findings

Using in silico chromosome walking, we searched for prematurely terminated transcripts bearing signatures of intron retention or exonization of intronic sequence at their 3′ ends upstream to human L1 retrotransposons, protein-coding and noncoding nested genes. We demonstrate that transcriptional interference induced by intronic L1s (or other repeated DNAs) and nested genes could be characterized by intron retention, forced exonization and cryptic polyadenylation. These molecular effects were revealed from the analysis of endogenous transcripts derived from different cell lines and tissues and confirmed by the expression of three minigenes in cell culture. While intron retention and exonization were comparably observed in introns upstream to L1s, forced exonization was preferentially detected in nested genes. Transcriptional interference induced by L1 or nested genes was dependent on the presence or absence of cryptic splice sites, affected the inclusion or exclusion of the upstream exon and the use of cryptic polyadenylation signals.

Conclusions/Significance

Our results suggest that transcriptional interference induced by intronic L1s and nested genes could influence the transcription of the large number of genes in normal as well as in tumor tissues. Therefore, this type of interference could have a major impact on the regulation of the host gene expression.

Functions of noncoding RNAs in neural development and neurological diseases

The development of the central nervous system (CNS) relies on precisely orchestrated gene expression regulation. Dysregulation of both genetic and environmental factors can affect proper CNS development and results in neurological diseases. Recent studies have shown that similar to protein coding genes, noncoding RNA molecules have a significant impact on normal CNS development and on causes and progression of human neurological disorders. In this review, we have highlighted discoveries of functions of noncoding RNAs, in particular microRNAs and long noncoding RNAs, in neural development and neurological diseases. Emerging evidence has shown that microRNAs play an essential role in many aspects of neural development, such as proliferation of neural stem cells and progenitors, neuronal differentiation, maturation, and synaptogenesis. Misregulation of microRNAs is associated with some mental disorders and neurodegeneration diseases. In addition, long noncoding RNAs are found to play a role in neural development by regulating the expression of protein coding genes. Therefore, examining noncoding RNA-mediated gene regulations has revealed novel mechanisms of neural development and provided new insights into the etiology of human neurological diseases.

A natural antisense transcript at the Huntington's disease repeat locus regulates HTT expression

Huntington's disease (HD) is a progressive neurodegenerative disorder caused by a CAG repeat expansion in exon 1 of huntingtin (HTT). Relatively little attention has been directed to the genomic features of the antisense strand at the HD locus, though the presence of a transcript from this strand has been suggested by a survey of the entire transcriptome and the existence of several EST tags. In this study, we identified huntingtin antisense (HTTAS), a natural antisense transcript at the HD repeat locus that contain the repeat tract. HTTAS is 5' capped, poly (A) tailed and contains three exons, alternatively spliced into HTTAS_v1 (exons 1 and 3) and HTTAS_v2 (exons 2 and 3). Exon 1 includes the repeat. HTTAS_v1 has a weak promoter, and is expressed at low levels in multiple tissue types and throughout the brain. Reporter assays indicate that while efficient promoter activity requires a short repeat, repeat expansion reduces promoter efficiency. Consistent with the reporter assays, levels of HTTAS_v1 are reduced in human HD frontal cortex. In cell systems, overexpression of HTTAS_v1 specifically reduces endogenous HTT transcript levels, while siRNA knockdown of HTTAS_v1 increases HTT transcript levels. Minigene constructs of the HD locus confirm the regulatory effect of HTTAS_v1 on HTT, and demonstrate that the effect is dependent on repeat length and is at least partially Dicer dependent. Together, these findings provide strong evidence for the existence of a gene antisense to HTT, with properties that include regulation of HTT expression.

Influence of host gene transcription level and orientation on HIV-1 latency in a primary-cell model

Human immunodeficiency virus type 1 (HIV-1) establishes a latent reservoir in resting memory CD4(+) T cells. This latent reservoir is a major barrier to the eradication of HIV-1 in infected individuals and is not affected by highly active antiretroviral therapy (HAART). Reactivation of latent HIV-1 is a possible strategy for elimination of this reservoir. The mechanisms with which latency is maintained are unclear. In the analysis of the regulation of HIV-1 gene expression, it is important to consider the nature of HIV-1 integration sites. In this study, we analyzed the integration and transcription of latent HIV-1 in a primary CD4(+) T cell model of latency. The majority of integration sites in latently infected cells were in introns of transcription units. Serial analysis of gene expression (SAGE) demonstrated that more than 90% of those host genes harboring a latent integrated provirus were transcriptionally active, mostly at high levels. For latently infected cells, we observed a modest preference for integration in the same transcriptional orientation as the host gene (63.8% versus 36.2%). In contrast, this orientation preference was not observed in acutely infected or persistently infected cells. These results suggest that transcriptional interference may be one of the important factors in the establishment and maintenance of HIV-1 latency. Our findings suggest that disrupting the negative control of HIV-1 transcription by upstream host promoters could facilitate the reactivation of latent HIV-1 in some resting CD4(+) T cells.