sign in sign up
<<
Home , Transcription (biology)

Insights & Perspectives

What do you mean by rate?

Think again The conceptual difference between nascent transcription rate and mRNA synthesis rate is essential for the proper understanding of transcriptomic analyses

JoseE. Perez-Ort ın1)*, Daniel A. Medina1), Sebastian Chavez 2) and Joaquın Moreno1)

mRNA synthesis in all is performed by RNA polymerases, which work knowledge of expression by re- as nanomachines on DNA templates. The rate at which their product is made is vealing the relative weight of synthesis and decay on mRNA changes during an important parameter in . Transcription rate encompasses global transcriptional responses [1, 2]. two related, yet different, concepts: the nascent transcription rate, which RNA polymerase (RNA pol) density can measures the in situ mRNA production by RNA polymerase, and the rate of easily be determined -wide [2, 3] synthesis of mature mRNA, which measures the contribution of transcription to (Fig. 1) and converted into a transcrip- the mRNA concentration. Both parameters are useful for molecular biologists, tion rate by assuming uniform RNA pol but they are not interchangeable and they are expressed in different units. It is speed during transcription. In this way, the number of mRNA molecules being important to distinguish when and where each one should be used. We propose synthesized per unit time and gene copy that for functional the use of nascent transcription rates should be can be calculated. We will call this restricted to the evaluation of the transcriptional process itself, whereas mature measure a “nascent” transcription rate mRNA synthesis rates should be employed to address the transcriptional input (nTR) to distinguish it from other uses of to mRNA concentration balance leading to variation of gene expression. the term (Box 1). Accordingly, nTR is an estimation of transcriptional activity at Keywords: the gene level. However, a different aspect of un- ; mRNA synthesis; nascent transcription; transcription rate; yeast . derstanding the role of transcription in gene expression is achieved by evaluat- ing the change in the mature mRNA to become a mature form that can be concentration ([mRNA]). The [mRNA] is Introduction: What is translated by the . Besides, a determined by both the synthesis (SR) transcription rate? mature mRNA molecule is translated and the degradation (DR) rates. mRNA only a number of times as regulated by synthesis is considered to be indepen- The rate at which the are mRNA degradation processes. dent of its concentration, whereas decay transcribed is the first and probably The appearance of genomic techni- is assumed to follow first-order kinetics one of the most important regulated ques to measure the concentration, as with rate constant kd [1, 4]. Thus, the steps along the flux of genetic informa- well as the transcription and degrada- rate of [mRNA] change can be written in tion. In eukaryotes, synthesized mRNA tion rates, of the mRNAs of all genes in its simplest form as: undergoes processing and transport an has vastly improved our d½mRNA ¼ SR k ½ðmRNA 1Þ dt d DOI 10.1002/bies.201300057 Although SR and kd may vary with time, [mRNA] is bound to reach a steady 1) Departamento de Bioquı´mica y Biologı´a Abbreviations: state whenever both SR and k remain Molecular, Universitat de Vale` ncia, Burjassot, DR, degradation (decay) rate; dsDNA, double d Spain stranded DNA; HL, mRNA half-; kd, degradation unchanged for a time interval which is 2) Departamento de Gene´ tica, Universidad de constant; [mRNA], cytoplasmatic mRNA concen- longer than (1/kd) [1]. This condition Sevilla, Sevilla, Spain tration; nTR, nascent transcription rate; RNA pol, may never be achieved by genes of RNA polymerase; SR, mRNA synthesis rate. *Corresponding author: Jose´ E. Pe´ rez-Ortı´n cyclic expression, but it is usually E-mail: [email protected] fulfilled for most genes under fixed

1056 www.bioessays-journal.com Bioessays 35: 1056–1062, ß 2013 WILEY Periodicals, Inc...... Insights & Perspectives J. E. Pe´ rez-Ortı´n et al.

understanding of mRNA and expression strategies. hn again Think

Nascent transcription rate can be measured from in vivo RNA pol density

RNA polymerases are highly processive , which work on dsDNA. Processive nanomachines have been described as “cars on a single-lane highway” [11, 12]. Thus, RNA pol travels at a variable speed and gets into traffic jams, crashes with other RNA pol, stops and backtracks. In kinetic terms, the two most important features of tran- scribing RNA pol are its average speed Figure 1. Scheme of the gene expression process in eukaryotes showing the different and the density of the molecules on techniques that measure the nTR and mature mRNA SR. The nascent TR is determined the track. The movement of RNA pol from the RNA pol (colored ovals) density on templates. Some of the RNA pol is characterized by a given speed or molecules are elongating (green), others are initiating (yellow) and some are backtracked elongation rate (in kb/min) that might (red). They can all be detected by chromatin immunoprecipitation, but only those that are be variable [13, 14]. The product of actually elongating are measured by Genomic run-on. If cells are incubated for a pulse with thiolated UTP precursors, the newborn mature mRNA can be purified and quantified together elongation speed and RNA pol density is with total non-labeled mRNA. The current techniques cannot distinguish between the mature a measure of “transcriptional produc- mRNA transiently in nucleus (supposed to be a minor part) and the cytoplasmic one. By tivity” (in nascent RNA molecules/min). assuming steady-state conditions, this protocol, called comparative Dynamic We propose the acronym nTR to design Analysis (cDTA [6]), allows the determination of the SR. See [27] for further details on the the number of full nascent mRNA techniques depicted herein. molecules per unit time produced by the RNA pol molecules acting on a single gene copy. nTR is determined by environmental conditions leading to a (kd) are known [2]. SR is also a measure the number of transcription initiation steady state in which the synthesis and of transcriptional activity and is fre- events at the , by RNA pol degradation rates are equal: quently referred to as “transcription speed changes, and by elongation rate” [1, 3, 7–10]. We use here “synthesis failure (premature drop-off) [15], all SR ¼ DR ¼ kd ½mRNAð2Þ rate” instead (as used in references [5, three being regulated processes [16]. 6]) to avoid confusion. Indeed, SR At a given average speed, the In that instance, [mRNA] can be represents another aspect of the mRNA transcriptional activity on a gene is expressed as a ratio between the biosynthetic process different from directly proportional to the total num- synthesis rate and kd, which is inversely nTR (Box 1). As mentioned above, SR ber of elongating RNA pol molecules on related to the mRNA half-life (HL): relates to the rate of appearance of the track. Nevertheless, as nTR com- mature (in contrast to nascent) mRNA. putes the production of full-length ln2 Furthermore, because [mRNA] balance mRNAs and genes have different HL ¼ ð3Þ kd in the results from mass action lengths, the relevant parameter for law, SR in Equations (1) and (2) should estimating nTR is not the number of The appearance of mature mRNA express the change in concentration, RNA pol per gene, but its density (in can also be determined genome-wide in and not the change in absolute amount molecules/kb) [17]. Genome-wide nTR a direct or indirect manner (Fig. 1) [5, 6]. (as does nTR). Therefore, SR is a can be quantified by measuring the Moreover, because this appearance is measure of the transcriptional input density of the polymerases acting on related to [mRNA] and its DR through to mature mRNA concentration at the each gene by either genomic run-on [3] Equation (1) (or particular embodiments cellular level. or chromatin immunoprecipitation of of it), techniques In this paper, we compare nTR and RNA pol II [7, 18] (Fig. 1). A caveat are currently used to infer one of the SR (both of which are frequently alluded of these techniques is that not all parameters from the experimental de- in the bibliography as “transcription detected RNA pol molecules would termination of the other two. For rate” in an undistinguishable way) and finish transcription because of a sto- example, the steady-state condition discuss their proper use when studying chastic elongation failure that may described above (Equation 2) can be gene expression in eukaryotes. This cause polymerase drop-off [6, 15]. Thus, used to determine the SR if the concen- leads to a re-appraisal of some pub- because longer genes have higher tration of the mature mRNA and the HL lished results, and also to a better number of drop-off events, the

Bioessays 35: 1056–1062, ß 2013 WILEY Periodicals, Inc. 1057 J. E. Pe´ rez-Ortı´n et al. Insights & Perspectives.....

molecules is to be divided by the Box 1 numberofgenecopiestogettheactual “Concept and units of nTR and SR” nTR (Box 1). This corrects for the case of multiple copies of the same gene or The nascent transcription rate (nTR) measures the work done by the biological when comparing cells with different machine RNA polymerase (RNA pol) and can be evaluated in terms of the levels of ploidy (see later). product (RNA) made per unit time. If and RNA pol molecules are not limiting, the number of nascent RNA molecules for a given gene should be proportional to its copy number. To keep the original significance of nTR mRNA synthesis rate is as a standard measure of transcriptional activity, we propose dividing the the relevant measure in Think again rate of mRNA synthesis by the average number of gene copies to obtain a normalized “per gene copy” nTR. Thus, the nTR units might be: kinetic equilibria

mRNA molecules=min=gene copy: The scenario is different when studying changes in the level of translatable The synthesis rate (SR) represents the transcriptional input to mature mRNA. In this case, the transcriptional mRNA concentration. It reflects the maximum hypothetical rate of change in input should rather be measured as the the mature mRNA concentration due exclusively to its synthesis, i.e. the rate of appearance of mature mRNA concentration change that would occur if there were no decay. Therefore, to in the . Eukaryotic mature calculate SR, the number of mature mRNA molecules released in the mRNA (usually determined as polyA- cytoplasm per unit time has to be divided by the cytoplasmic volume. Thus, RNA) results from nascent mRNA after a the units of SR might be: number of steps, which it might fail to pass. For instance, problems during mRNA molecules=L=min splicing and further processing may preclude maturation of nascent mRNA. From the above definitions, it can be concluded that nTR and SR are As a result, only a fraction of nascent different in two significant ways. First, nTR refers to nascent mRNAs (i.e. full mRNA reaches the cytoplasm. At this copies of the gene message as released by the RNA pol) while SR counts only point, the impact on further gene mature mRNAs (i.e. the fraction of nascent mRNAs that undergo correct expression depends on cytoplasmic processing and transport as to become translatable). Second, nTR is a rate mRNA concentration change. Although amount of production of of mRNA while SR is a rate of change of mRNA the [mRNA] balance assumed in Equa- concentration . tions (1) and (2) is undoubtedly a gross To illustrate the different biological meaning of nTR and SR one may oversimplification of a complex process consider a cell growing in size. Even if nTR is maintained constant for a that takes place in a spatially compart- particular gene, SR would monotonically decrease because of the fixed mentalized environment, it is still a synthetic input being divided by an ever increasing volume. In this case, a valid assumption that the DR depends constant nTR will reflect a steady work done by the RNA polymerase, while the on [mRNA] rather than on mRNA reduction in SR will reveal the decreasing impact on mRNA concentration amount. Therefore, to fit dimensionally (hence, on mRNA ) of a fixed transcriptional input on a growing in Equation (1), SR should be expressed volume. This example highlights the functional difference between these two as the mature mRNA molecules released parameters addressing distinct aspects of transcription. per unit time and unit volume. This magnitude has been called “transcrip- tion rate” in many publications [1, 9, 10] but we propose to call this magnitude calculated nTR is overestimated in a previously determined by other authors mRNA synthesis rate (SR) to distinguish gene-length-dependent way. This bias in the same experimental conditions from nTR (Box 1). Because mRNA has to be corrected from an estimate of [23] was used to convert all yeast genes translation events are also kinetically the drop-off rate. Besides, in contrast to nTRs into absolute values [3]. controlled, production is single-cell techniques that can measure Once correctly expressed as the expected to depend on [mRNA], too. mRNA production in absolute units (i.e. number of mRNA molecules synthe- Therefore, by expressing SR in terms of mRNA molecules/min) [19–22], com- sized per unit time and gene copy, concentration and time, it also reflects mon genome-wide techniques (averag- nTR is a valuable indicator of tran- how efficiently transcriptional changes ing millions of cells) deliver only scriptional activity, and it is directly (as measured by nTR) are transmitted relative values in arbitrary units. How- impacted by all transcriptional regula- into the downstream steps of gene ever, because these values are internally tory mechanisms acting on the poly- expression. As such, SR is a biologically proportional, different genes can be merase or the template. As nTR relevant parameter in its own right, and compared [2, 3] and arbitrary units measures productivity per gene copy, is related (but is not equivalent) to nTR. can be converted into absolute ones cell volume does not enter its calcula- Values for SR can be obtained through an external reference. For tion. But the experimentally deter- indirectly by applying Equation (2). In instance, the SR of the yeast HIS3 gene mined number of nascent mRNA fact, the first genome-wide calculation

1058 Bioessays 35: 1056–1062, ß 2013 WILEY Periodicals, Inc...... Insights & Perspectives J. E. Pe´ rez-Ortı´n et al.

of SR was carried out in yeast using experimentally determined [mRNAs] Box 2 and HLs, and assuming steady-state “mRNA concentration, not quantity, is pertinent for again Think conditions [24]. These indirect SR values obtained at the steady state can be transcriptomics” plotted against nTR values (in arbitrary units), which have been experimentally The relevant factor in chemical reaction kinetics is the concentration, not the determined under the same environ- amount, of molecules present. Thus, comparisons made between different mental conditions, to find a conversion samples in transcriptomics studies should be raised at the [mRNA] level. factor [2, 3, 25]. This is a more robust For instance, a new protocol for performing global expression analyses alternative to get real units for nTR than from samples that may differ as far as the amount of RNA/cell is concerned using a single gene value (see above). was recently proposed [41]. The authors concluded that the common This factor might be later applied to assumption that samples of the same number of cells under different transform nTR data under non-steady treatments have equivalent total amounts of mRNAs is not always true. They state conditions into SR. This procedure used data from two other reports [42, 43] in which an elevated induced assumes that the efficiency of nuclear expression of human c-Myc brought about a general mRNA degradation and export remains increase in the levels of thousands of mRNAs, which in turn, increased the total constant for every gene. That is obvi- amount of mRNA per cell. The authors warn about the usual normalization ously a simplification, but it is a valid practice of processing the same amount of mRNA from each sample, which conversion from nTR into SR because has been used in most experiments for many years. However, we wish to the use of [mRNA] in calculating stress that the comparison made between samples should not be based on standard SRs (those used for normali- mRNA/cell content, but on [mRNA]. This does not imply a major difference in zation) introduces the correct volume most cases in which cells of different samples are similar in size. However factor. in these reports [42, 43], the cell types used in the experiments were of Recently, genome-wide techniques different sizes. This means that the actual mRNA concentrations in the cells for directly measuring the rate of overexpressing c-Myc come closer to those in the control cells than suspected apparition of mature mRNA have been since the control cells were smaller. The actual sizes of cells were not directly implemented in yeast [5, 6] and mam- provided in the papers, but from the plots in Fig. S6A of [42], it is stated that cell malian cells [9, 10] (Fig. 1). They are volume increases about 1.3 times between 1.5 and 6 h after c-Myc induction. based on the capture of newborn mature This means that the observed rise in mRNA per cell (1.5 times between 0 mRNA, which is labeled during a short and 8 h) is probably compensated. Therefore, what the cited papers actually pulse with thiolated or uridine reveal is that c-Myc overexpression causes a general increase in nTR but, analogs. The subsequent isolation and because of the parallel increase in cell volume, it does not produce a parallel quantification of labeled mRNA can be rise in [mRNA] or in SR. used to evaluate SR. Because no subcel- lular fractionation is done, the mature mRNA measured includes both nuclear and cytoplasmic fractions, although the expected to maintain a better propor- What is the biological latter is considered to be much more tionality with cell (or cytoplasmic) relationship between nTR abundant (Fig. 1). Quantification of non- volume than the number of cells. labeled or total mRNA facilitates the From all the above, it seems clear and SR? determination of [mRNA] in the same that nTR and SR are conceptually, experiment and the subsequent estima- dimensionally, and numerically differ- nTR and SR are related biologically tion of HLs by assuming a steady-state ent magnitudes which should be used in relevant parameters, but involve dis- situation [26–28]. By using these proto- different contexts. They might be pro- tinct constraints. nTR measures the cols, SR can be obtained directly in real portional if cell volume, DNA content productivity of the transcriptional ma- absolute units after dividing by the cell and mRNA maturation efficiency remain chinery on a particular gene. This volume. Furthermore, it is a common constant (a common, but not universal, machinery is ruled by physicochemical practice to divide directly the amount of situation that is usually assumed), laws. Its speed and performance depend synthesized mRNA by the number of but will diverge otherwise. Despite all on temperature, DNA sequence, nucle- cells, leading to SR data expressed as a this, and because most available tech- osome positioning, and so on. With a rate of [mRNA] change per cell. This niques determine primarily nTR, it is gene transcribed at a given speed, any procedure is acceptable provided that a regular practice to use the latter increase in nTR requires a proportional the cell volume does not vary between in calculations in which, according to increase in RNA pol density. This the samples being compared. However, Equation (2), SR should be used instead. involves a higher of initiation if it is not possible to verify that cell The incorrect use of nTR values dis- events by free polymerases, which are volume remains constant, it might be regarding the volume factor operating probably in excess [2]. RNA pol density preferable to normalize the data to total on SR may lead to misleading conclu- is also limited by physical constrains, cell mass or total protein amount (see sions drawn from valuable experimen- such as the size of the transcriptional Box 3) because these parameters are tal data (Boxes 2 and 3). apparatus. The “Christmas trees”

Bioessays 35: 1056–1062, ß 2013 WILEY Periodicals, Inc. 1059 J. E. Pe´ rez-Ortı´n et al. Insights & Perspectives.....

expression, as in cancer and some Box 3 neurodevelopmental disorders [31]. In “SR and the transcriptional strategy to cope with those cases in which a small number of genes shift to an altered copy number, cell size” the observed RNA pol II density per gene copy (and the nTR) is expected to SR redefinition offers new views to some published data. The first case is the remain unchanged, as far as each copy Saccharomyces cerevisiae comparative analysis of HLs, SRs, and [mRNA] for behaves independently from the others Schizosaccharomyces pombe (Sc) and (Sp) [6]. The authors discovered that and RNA pol II is in excess, as is usually the [mRNA] in both yeasts are quite similar because the mRNA molecules in assumed (discussed in [2]). However, SR Sp are 3.1 times more abundant than in Sc, while its volume is also 2.7 times

Think again will increase proportionally to copy k larger. The median HLs in Sp are 5 times longer (0.2 factor difference in d; see number amplification. This example Equation 3), rendering a ratio of 0.23 between the DRs in Sp and Sc. At further illustrates the physiological the steady state, this ratio should match the equivalent ratio for SRs (see difference between nTR and SR. Equation 2). In spite of using a method (cDTA in Fig. 1) that experimentally In the special case of whole genome measures production of mature mRNAs, values for SR are given in nTR units duplications (e.g. from diploid to tetra- (44 mRNAs/cell- in Sp vs. 53 in Sc), which suggests a much higher ratio ploid), a parallel increase in all the (0.83), thus casting doubts on the steady-state assumption. However, the 3.6- components of the transcriptional ma- fold difference (0.83 vs. 0.23) results from using the incorrect units. Since Sp chinery (and subsequently in the com- is almost three times larger, the published data fit quite well after having ponents of other metabolic and gene corrected the calculated nTR by the volume (i.e. converted back to SR). Thus, expression pathways) also occurs. This the data indeed support a steady state around a similar [mRNA]. may lead to an increased nuclear Another example is the analysis of Sp mutants of variable size [7, 8, 44, 45]. volume, which in turn, brings about A main conclusion drawn was that “global transcription rates scale with increased cell volume, resulting in near size” [45]. The transcription rates were calculated after RNA pol ChIP and pulse constant concentrations for almost all labeling of nascent mRNA [7, 44]. Thus, they determined nTR. However, the SR molecules [8]. In fact, the ratio between (the relevant parameter for mRNA homeostasis) remained constant because nuclear and cell volumes remains of the parallel increase of nTR and cell size. The authors noted that “the constant with ploidy (reviewed transcription rate per protein” remained constant. Since the protein amount in [35]). Thus, in the case of genome increases linearly with cell size, that parameter is the equivalent of SR. Because duplication, nTR remains constant be- the authors also reported similar mRNA decay rates and [mRNA] values, it can cause the effect of twice as many be concluded that the kd were maintained across mutants with different size. transcribing RNA pol is corrected by This supports the idea that the same steady-state conditions apply in the wild the presence of two gene copies (accord- type and the mutants with no change in SR. Interestingly, it was also found that ing to our definition; see Box 1). Simi- nTR increases with growth only until cells reached twice the wild-type volume. larly, SR does not vary because the Thereafter, nTR did not increase, probably because of physicochemical doubled input of mature mRNA is constraints. Thus, SR decreased steadily above this volume threshold and compensated in concentration terms cells could no longer support the required increases in protein synthesis [7]. by the doubling of the cell volume. Both examples suggest that cells establish a similar steady-state value for Conversely, for those cells with the k mature [mRNA] by adapting nTR and/or d to cell size up to the performance same genome, but different cell vol- limits of the transcriptional machinery. umes, SR ought to be adapted. As observed in yeast, slight changes in volume can be buffered by proportional changes in nTR in order to keep SR and formed by RNA pol I when transcribing when a higher nTR is required, as in the [mRNA] constant (see Box 3). However rDNA (7.3 RNA pol/kb on average and common case for rDNA (150 repeats of for human cells, which can differ up to up to 24 molecules/kb) [29] probably the 35S gene in Saccharomyces cerevi- 1,000 times in volume, it is difficult to represents the absolute maximum siae, and 350 copies per haploid ge- imagine how nTR can adapt to such [30, 31]. The highest density measured nome in human cells) [33]. There are huge changes. Since the relevant vol- for RNA pol II in yeast is, at least, other examples of highly transcribed ume for the equilibrium in Equation (2) threefold lower [2]. In higher eukar- genes that obtain high global TR by is that of the cytoplasm, the “effective” yotes, the average RNA pol II density is adding new copies to a single genome. cytoplasmic volume can be reduced by probably lower than in yeast [27], For instance, in the haploid genome of increasing the volume of some organ- although high RNA pol II densities, S. cerevisiae most ribosomal protein elles (such as the vacuole in yeast). up to 10 RNA pol II/kb, have been genes are duplicated, and the genes Alternatively, mRNAs can be actively observed in HIV or rat-kidney [22, 32]. encoding methalotionein CUP1 get am- recruited to specialized sites (where The scarcity of RNA pol II elongating plified in response to the presence of they rise to higher concentration) as molecules and their physically limited heavy metals in the medium [34]. Con- in the axons of neurons or the mRNA elongation speed may have forced cells versely, pathological gradients in the Drosophila syncytium to increase the number of gene copies may be detrimental due to imbalanced blastoderm [36]. Another possibility to

1060 Bioessays 35: 1056–1062, ß 2013 WILEY Periodicals, Inc...... Insights & Perspectives J. E. Pe´ rez-Ortı´n et al.

buffer the effect of cell volume increase size-dependent control of [mRNA] in to V. Pelechano for critically reading the is to lower the DR by decreasing kd.In yeast (see [7], and discussion in [8]). manuscript. J.E.P-O. is supported by fact, mRNA stability is higher in cells Thus, it seems that nTR is the key grants from the Spanish MCINN and hn again Think with a larger volume, such as Schizo- regulatory point. nTR can be controlled the European Union (FEDER) (BFU2010- saccharomyces pombe relative to S. at either the initiation or the elongation 21975-C03-01), and from the Generalitat cerevisiae [6]. However, this strategy level. It has been shown that some Valenciana (PROMETEO 2011/088 and has a cost in terms of the reduced , such as Rpb4/7 or Ccr4-Not, ACOMP/2012/001). J.M. is supported by speed at which the cell can adapt to acting during transcription initiation or a grant from the Spanish MCINN new environments by changing gene elongation are charged onto nascent (BFU2009-11965). S.C. is supported by expression [1]. Although cells of multi- mRNAs and determine their cyto- grants from the Spanish MCINN cellular organism may tolerate substan- plasmic fate [27, 37–39]. Contrariwise, (BFU2010-21975-C03-02) the Regional tially longer adaptation times, there some cytoplasmic decay factors affect Andalusian Government (P07-CVI- must be a limiting size beyond which transcription initiation and elongation 02623 and P08-CVI-03508), and the this strategy cannot be further exploited after being imported to the nucleus European Union (FEDER). D.A.M. was and the cell has to settle for a reduced [28, 40]. For instance, impairing mRNA a recipient of a Santiago Grisolı´a fellow- [mRNA] level. Accordingly, [mRNA] degradation by deleting deadenylase ship from the Generalitat Valenciana. appears to be at least 2.5 times lower subunits of the Ccr4-Not complex in human cells than in yeast cells, even causes decreased decay rates as if the HLs of mammalian mRNAs are on expected, but also decreased synthesis References average 15 to 20 times longer [28]. rates [6]. Thus, gene expression in It should be finally remarked that eukaryotes seems to have evolved to 1. Perez-Ortin JE, Alepuz PM, Moreno J. 2007. Genomics and gene transcription an independent determination of nTR keep total [mRNA] within certain limits kinetics in yeast. Trends Genet 23: 250–7. and SR for a single organism should by a permanent cross-talk between 2. Pelechano V, Chavez S, Perez-Ortin JE. be a suitable approach for detecting mRNA synthesis and degradation acting 2010. A complete set of nascent transcription rates for yeast genes. PloS One 5: e15442. differential patterns in mRNA process- on nTR as the main target. 3. Garcia-Martinez J, Aranda A, Perez-Ortin ing efficiency (i.e. in RNA splicing or JE. 2004. Genomic run-on evaluates tran- in mRNA transport) between genes. scription rates for all yeast genes and identifies gene regulatory mechanisms. Mol To date, this particular study has not Conclusions Cell 15: 303–13. yet been carried out on a genomic 4. Steiger MA, Parker R. 2002. Analyzing scale. The wealth of transcription data provid- mRNA decay in Saccharomyces cerevisiae. Methods Enzymol ed by the current genomic techniques is 351: 648–60. 5. Miller C, Schwalb B, Maier K, Schulz D, contributing to illuminate the underly- et al. 2011. Dynamic transcriptome analysis Gene expression is a ing complexity of the phenomenon. In measures rates of mRNA synthesis and decay order to understand the complex and in yeast. Mol Syst Biol 7: 458. circular process: 6. Sun M, Schwalb B, Schulz D, Pirkl N,etal. variable transcriptional machinery and 2012. Comparative dynamic transcriptome Cross-talk between nTR to make sense of its regulatory value, it analysis (cDTA) reveals mutual feedback and mRNA stability is first necessary to define the pertinent between mRNA synthesis and degradation. Genome Res parameters and to clarify the nomencla- 22: 1350–9. 7. Zhurinsky J, Leonhard K, Watt S, Mar- [mRNA] homeostasis requires cross-talk ture. In particular, it must be realized guerat S,etal. 2010. A coordinated global between the synthesis and degradation that the common term “transcription control over cellular transcription. Curr Biol rates [28, 37]. According to the discus- rate” has been used by different authors 20: 2010–5. 8. Marguerat S, Bahler J. 2012. Coordinating sion above, the parameter that should to refer to different parameters, such as genome expression with cell size. Trends balance the DR in order to reach a given rate of nascent [3, 7, 19, 20], the Genet 28: 560–5. steady [mRNA] is SR, which can be production of mature mRNAs [9, 10] or 9. Rabani M, Levin JZ, Fan L, Adiconis X,etal. 2011. Metabolic labeling of RNA uncovers changed by affecting nTR (i.e. the even the RNA pol speed [11, 21, 22] in principles of RNA production and degradation recruitment and activity of the RNA different experimental contexts. Our dynamics in mammalian cells. Nat Biotechnol pol on the chromatin template) or by proposal here is a reasoned plea to 29: 436–42. altering the cell volume. However, as clarify nomenclature and also to distin- 10. Schwanhausser B, Busse D, Li N, Dittmar G,etal. 2011. Global quantification of volume changes affect the steady guish between nascent transcription mammalian gene expression control. Nature [mRNA], and subsequently the DR, in rates and mature synthesis rates, two 473: 337–42. a counteracting manner, this is not a physiologically relevant but different 11. Dennis PP, Ehrenberg M, Fange D, Bremer H. 2009. Varying rate of RNA chain elongation feasible alternative. The control over the aspects of RNA synthesis, thereby during rrn transcription in Escherichia coli. DR cannot operate on [mRNA] for the avoiding confusion in the interpretation J Bacteriol 191: 3740–6. 12. Tuller T, Carmi A, Vestsigian K, Navon S, same reason, but should act on kd. and discussion of the data. Therefore for the coordinated regulation et al. 2010. An evolutionarily conserved mechanism for controlling the efficiency of of mRNA synthesis and degradation, the protein translation. Cell 141: 344–54. critical parameters are those acting at Acknowledgments 13. Danko CG, Hah N, Luo X, Martins AL,etal. the molecular level: nTR and the mRNA The authors are grateful to the members 2013. Signaling pathways differentially affect RNA polymerase II initiation, pausing, and HL. However, mRNA degradation does of the laboratories in Valencia and elongation rate in cells. Mol Cell 50: 212– not seem to play an important role in the Seville for discussion and support and 22.

Bioessays 35: 1056–1062, ß 2013 WILEY Periodicals, Inc. 1061 J. E. Pe´ rez-Ortı´n et al. Insights & Perspectives.....

14. Marcello A. 2012. RNA polymerase II 25. Pelechano V, Perez-Ortin JE. 2010. There is rangements fuel rapid adaptation in yeast transcription on the fast lane. Transcription a steady-state transcriptome in exponentially populations. PLoS Genet 9: e1003232. 3: 29–34. growing yeast cells. Yeast 27: 413–22. 35. Huber MD, Gerace L. 2007. The size-wise 15. Mason PB, Struhl K. 2005. Distinction and 26. Friedel CC, Dolken L. 2009. Metabolic nucleus: nuclear volume control in eukar- relationship between elongation rate and tagging and purification of nascent RNA: yotes. J Cell Biol 179: 583–4. processivity of RNA polymerase II in vivo. implications for transcriptomics. Mol Biosyst 36. Holt CE, Bullock SL. 2009. Subcellular Mol Cell 17: 831–40. 5: 1271–8. mRNA localization in cells and why it 16. Cheung AC, Cramer P. 2012. A movie of 27. Perez-Ortin JE, de Miguel-Jimenez L, matters. Science 326: 1212–6. RNA polymerase II transcription. Cell 149: Chavez S. 2012. Genome-wide studies of 37. Goler-Baron V, Selitrennik M, Barkai O, 1431–7. mRNA synthesis and degradation in eukar- Haimovich G,etal. 2008. Transcription in 17. Hirayoshi K, Lis JT. 1999. Nuclear run-on yotes. Biochim Biophys Acta 1819: 604–15. the nucleus and mRNA decay in the cytoplasm assays: assessing transcription by measuring 28. Perez-Ortin JE, Alepuz P, Chavez S, are coupled processes. Genes Dev 22: 2022–7. density of engaged RNA polymerases. Meth- Choder M. 2013. Eukaryotic mRNA decay: 38. Harel-Sharvit L, Eldad N, Haimovich G,

Think again ods Enzymol 304: 351–62. methodologies, pathways, and links to other Barkai O,etal. 2010. RNA polymerase II 18. Mayer A, Lidschreiber M, Siebert M, Leike stages of gene expression. J Mol Biol,in subunits link transcription and mRNA decay K,etal. 2010. Uniform transitions of the press, DOI: 10.1016/j.jmb.2013.02.029 to translation. Cell 143: 552–63. general RNA polymerase II transcription 29. French SL, Osheim YN, Cioci F, Nomura M, 39. Reese JC. 2013. The control of elongation by complex. Nat Struct Mol Biol 17: 1272–8. et al. 2003. In exponentially growing Saccha- the yeast Ccr4-not complex. Biochim Biophys 19. Zenklusen D, Larson DR, Singer RH. 2008. romyces cerevisiae cells, rRNA synthesis is Acta 1829: 127–33. Single-RNA counting reveals alternative determined by the summed RNA polymerase 40. Haimovich G, Medina DA, Causse SZ, modes of gene expression in yeast. Nat I loading rate rather than by the number of Garber M,etal. 2013. Gene expression is Struct Mol Biol 15: 1263–71. active genes. Mol Cell Biol 23: 1558–68. circular: factors for mRNA degradation also 20. Larson DR, Zenklusen D, Wu B, Chao JA, 30. Lucchini R, Sogo JM. 1998. The dynamic foster mRNA synthesis. Cell 153: 1000–11. et al. 2011. Real-time observation of tran- structure of ribosomal RNA gene chromatin. 41. Loven J, Orlando DA, Sigova AA, Lin CY, scription initiation and elongation on an In Paule MR, ed; Transcription of Ribosomal et al. 2012. Revisiting global gene expression endogenous yeast gene. Science 332: 475–8. RNA Genes by Eukaryotic RNA Polymerase I. analysis. Cell 151: 476–82. 21. Maiuri P, Knezevich A, De Marco A, Mazza Austin TX: Landes Bioscience. p. 255–76. 42. Lin CY, Loven J, Rahl PB, Paranal RM,etal. D,etal. 2011. Fast transcription rates of RNA 31. Tang YC, Amon A. 2013. Gene copy-number 2012. Transcriptional amplification in tumor polymerase II in human cells. EMBO Rep 12: alterations: a cost-benefit analysis. Cell 152: cells with elevated c-Myc. Cell 151: 56–67. 1280–5. 394–405. 43. Nie Z, Hu G, Wei G, Cui K,etal. 2012. c-Myc 22. Femino AM, Fay FS, Fogarty K, Singer RH. 32. Maiuri P, Knezevich A, Bertrand E, Mar- is a universal amplifier of expressed genes in 1998. Visualization of single RNA transcripts cello A. 2011. Real-time imaging of the HIV-1 lymphocytes and embryonic stem cells. Cell in situ. Science 280: 585–90. transcription cycle in single living cells. 151: 68–79. 23. Iyer V, Struhl K. 1996. Absolute mRNA levels Methods 53: 62–7. 44. Marguerat S, Schmidt A, Codlin S, Chen W, and transcriptional initiation rates in Saccha- 33. Sakai K, Ohta T, Minoshima S, Kudoh J, et al. 2012. Quantitative analysis of fission romyces cerevisiae. Proc Natl Acad Sci USA et al. 1995. Human ribosomal RNA gene yeast and proteomes in pro- 93: 5208–12. cluster: identification of the proximal end liferating and quiescent cells. Cell 151: 671– 24. Holstege FC, Jennings EG, Wyrick JJ, containing a novel sequence. 83. Lee TI,etal. 1998. Dissecting the regulatory Genomics 26: 521–6. 45. Dungrawala H, Manukyan A, Schneider BL. circuitry of a eukaryotic genome. Cell 95: 34. Chang SL, Lai HY, Tung SY, Leu JY. 2013. 2010. Gene regulation: global transcription 717–28. Dynamic large-scale chromosomal rear- rates scale with size. Curr Biol 20: R979–81.

1062 Bioessays 35: 1056–1062, ß 2013 WILEY Periodicals, Inc.