Archive for the ‘transcription’ Category

The transcription or the expression of a gene(the process by which the DNA sequence is converted into a functional product like protein or RNA) is controlled by the region of the DNA generally present upstream of the gene. This region consists of several short segments(also known as motifs) which act as binding sites to proteins called transcription factors. It is generally believed that genes that share the same multiple regulators must show similar expression profiles or vice versa the genes that are show close expression patterns could be regulated by the same set of transcription factors.

If we look closer at the regulatory regions of a known set of co-expressed genes in a particular tissue, will it give a rule for how the architecture(the min or maximum number of binding sites, spacing of these binding sites, orientation of these binding sites etc.) of such regulatory regions look like and explain something about their evolution ?

This is exactly what the authors have done in this paper in science(subscription required). They have used the 19 genes that are co-expressed in muscle cells of developing urochordate Ciona embryo. Of these 19 genes, 17 function in the same macromolecular complex, underscoring the requirement for tight coexpression. These 19 genes include six single-copy loci (sequences in a genome that do not share homology with any other sequences in the same genome). Seven genes are composed of two or three members(paralogs) of multicopy gene families. We also know that genes that are expressed in Ciona are predominantly regulated by three different binding elements in their regulatory regions. These elements are 1) cAMP response element called CRE 2)MyoD motif 3) Tbx6 motif. These elements can be described in terms of DNA sequence bases they are composed of.

The authors study the distribution, composition and strength of these motifs in the upstream regulatory regions of the 19 genes. They found that there was high degree of heterogenity in these regulatory genes. There was no common feature they could discern from all of these 19 loci. So how to account for the co-exp of these genes, the authors show that it is done by conserving locus specific distribution of these features. This can be see more clearly in the following picture (B).

(B) Distribution of cis-regulatory function at the 19 loci of this study. Cs, Ciona. savignyi; Ci, Ciona. intestinalis. Labels below axes indicate distance to transcription start site. Area of circle is proportional to estimated motif activity. Motifs are depicted as circles, and color indicates motif type: CRE (red), MyoD (green), and Tbx6 (blue).

We do not see any commonality in the locus, but we see that the architecture of the regulatory regions are conserved in the specific locus, for example the Ck.ci(creatine kinase gene from Ciona intesinalis) and Ck.cs (from Ciona savignyi) have the same distribution of the motifs. This locus specific conservation the authors saw only in the six single copy genes. There was a higher degree of heterogenity in the paralogous cluster of genes in terms of both sequence and functional turn over.

Thus the authors conclude “Thus, the syntactical rules governing this regulatory function are flexible but become highly constrained evolutionarily once they are established in a particular element.”
Brown, C.D., Johnson, D.S., Sidow, A. (2007). Functional Architecture and Evolution of Transcriptional Elements That Drive Gene Coexpression. Science, 317(5844), 1557-1560. DOI: 10.1126/science.1145893


Read Full Post »

Last week I blogged on the talk by Peter Fraser where he showed evidence for the existence of transcription factories. He showed pictures as follows to demonstrate the theory of transcription factory. transcription factory

Fig. 1. Transcription factories are concentrated foci of active RNA polymerase. Immuno-detection of the hyper-phosphorylated form of RNAPII reveals their focal existence in a limiting number of transcription factories. Shown is a deconvoluted, single optical section of a mouse E12.5 fetal liver nucleus. Scale bar, 5 μm. Image courtesy of L. Chakalova. (taken from Seminars in Cell & Developmental Biology).

A new paper in molecular cell has appeared where the authors show evidence for the classical view that Pol II can be recruited to gene loci for activated transcription in living cells. They show this in polytene nuclei in salivary glands of Drosophila. Chromosomal DNA at loci containing the HS(heat shock) protein genes locally decondenses upon HS to form “puffs,” as can be readily observed in fixed, spread polytene chromosomes. Accompanying this decondensation is the strong recruitment of Pol II molecules that become densely packed along the activated transcription units. This is clearly demonstrated by this movie in live cells. Here is a picture which shows the transcription in the polytene chromosomes. The green color shows the presence of active transcription units where RNA pol II(which is ligated to enhaced Green flourscent protein) has been recruited.

polytene chromosome

Figure 1. Recruitment of Pol II to Major HS Puffs at 87A and 87C Observed in Living Cells

(A and B) Two-photon optical sections of a polytene nucleus expressing Rpb3-EGFP under NHS and HS conditions. (A) NHS: yellow arrows indicate Pol II-enriched sites that are transcriptionally active during normal development. (B) HS: red arrows show newly formed Pol II-concentrated sites upon HS.

(C) Recruitment of Pol II to 87A and 87C after HS. Times after HS are shown in minutes.

(D–F) Three sections of the same polytene nucleus under HS show distinct Pol II enriched sites (green). Chromosomes are stained with Hoechst33342 (red). 87A and 87C sites are indicated by the white arrows in (F) (Yao et al., 2006).

(G) A maximum intensity projection image (shown in pseudocolor) of all optical section series of this nucleus expressing EGFP-Rpb3 under HS (D–F). Scale bars, 10 μm.

Next the authors also have performed the dual-color FISH analyses on HS genes(figure 3). Labeled bacterial artificial chromosomes (BACs) containing HS genes were used to probe the interphase, diploid nuclei in larval imaginal disc tissues (Figure 3A). It is well known that HS genes are robustly activated in all larval tissues after HS, and therefore DNA-FISH signal during HS can represent active gene loci. The FISH signal indicates that HS genes occupy spatially distinct domains in most cases. Two Hsp70 loci 87A and 87C that are cytogenetically very close (separated by one subdivision not, vert, similar400 kb) also show distinct FISH signal. From these results, the authors conclude that, in diploid cells, HS genes colocalize at a very low frequency, and therefore, these coregulated genes are not generally cotranscribed within shared transcription factories during activation.no colocalization

Figure 3. Intranuclear Positions of HS Gene Loci in Drosophila Imaginal Disc Diploid Nuclei

(A) A FISH image of imaginal disc nuclei (red, small hsp locus 67B; green, Hsp70 locus 87A). Scale bar, 5 μm.

(B) FISH images on HS genes before and after HS (87C, Hsp70 locus; 63B, hsp83 locus).

(C) Summary of FISH analyses on HS gene pairs before and after HS.

(D) Positions of 87A, 67B, and 63B loci relative to the nuclear periphery or interior regions.

It is real exciting and waiting game to see which of the two theories are going to fail: Is transcription factory a distinct sub-nuclear compartment or the classical view of RNA pol II recruited to the gene loci !!


Read Full Post »

Today we had a very fascinating story by Peter Fraser (from the Babraham Institute, Cambridge), suggesting that genes migrate to specialized sites called transcription factories in the nucleus for transcription. These transcription factories are nothing but discrete foci in the nucleus containing high concentrations of RNA polII(the eukaryotic polymerase responsible for transcription of protein-coding genes). This is very much opposed to the well known text book theory that the RNA polymerase is recruited to promoters and RNA are transcribed from de novo transcription start sites.

The number of discrete factories visible in the nucleus is less than the number of expressed genes, which means that multiple genes are transcribed from one factory. This is also evident from the fact that the hbb(hemoglobin beta chain- a component of oxygen carrying hemoglobin protein in blood) gene is  transcribed in the same factory as that other similarly expressed genes in the erythroid cell. These genes are seperated by more than 50 megabases in cis(present of the same chromosome). Also unlinked genes in trans(present on different chromosomes) co-localized in the same the transcription factory but the frequency of such co-localization was low. (ex hbb & hba genes). The most interesting fact was that the quiet allele of active genes are located away from the transcription factories.

He also discussed more about the nuclear co-localization and regulatory functional implication of such localizations, described in detail in ” Nuclear organization of the genome and the potential for gene regulation Nature. 2007 May 24;447(7143):413-7

The interacting cis and trans genes completely dissociated when RNA initiation was inhibited by heat-shocking cells at 45 degrees, while the remained associated when only the RNA elongation was inhibited using a chemical DRB. Futhermore the transcription factories persisted even in the absence of transcription Which he presented as a compelling evidence that these transcription factories are distinct nuclear sub-compartments and not just local concentrated foci of RNA pol II. ref:” Transcription factories are nuclear subcompartments that remain in the absence of transcription Mitchell JA, Fraser P 2008 Jan 1;22(1):20-5

Read Full Post »

Today I read a paper from Genome Research 17:746-759, 2007. The authors have tried to annotate transcript products(RNA) from 399 annotated protein-coding loci which represent 1% of the human genome. Just as a reminder, from a single gene locus a large number of diverse transcripts can be produced, this is majorly through alternative splicing. This process generates on an average more than 5.4 transcript variants per locus. In addition we also have mammalian genes regulated by alternative promoters also producing transcript variants.

The authors in this paper show that for 82% of the gene locus tested had transcript variants with new perviously unknown internal exons or 5′ distal exons. More than half of the new transcript variants also span large segments of genome sequences and overlap with the upstream unannotated genes. Also they have shown that 20% of the transcripts have open reading frames fusing with that of adjacent genes. this they describe as chimeric transcript, where the fused ORFs produce a new protein.

But most of novel exons of the new transcripts did not show significant evolutionary conservation, which makes doubt the biological significance of the novel exons, as by definition conservation is a measure of some kind of biological function. Which leaves open that these novel exons could represent transcriptional noise as explained in Kevin struhl’s commentary on the infidelity of RNA Pol II.

Finally the authors have shown that the novel 5′ distal exons to be associated with hallmarks of transcription start site. But both functional RNA and transcriptional noise are produced by the same machinery and will obviously share many characteristics. Hence I believe such regulatory properties are inadequate criteria to distinguish biological significance from transcriptional noise.

I am not sure if these transcripts have been discovered due to increased sensitivity of experimental procedures or if they really are functional biological products. As Kevin struhl has pointed it would be “useful to have an experimental measurement of transcriptional noise that is not cofounded by the possibilty that the observed RNAs are biologically significant”.

Read Full Post »