A complete set of nascent transcription rates for yeast genes

Abstract

The amount of mRNA in a cell is the result of two opposite reactions: transcription and mRNA degradation. These reactions are governed by kinetics laws, and the most regulated step for many genes is the transcription rate. The transcription rate, which is assumed to be exercised mainly at the RNA polymerase recruitment level, can be calculated using the RNA polymerase densities determined either by run-on or immunoprecipitation using specific antibodies. The yeast Saccharomyces cerevisiae is the ideal model organism to generate a complete set of nascent transcription rates that will prove useful for many gene regulation studies. By combining genomic data from both the GRO (Genomic Run-on) and the RNA pol ChIP-on-chip methods we generated a new, more accurate nascent transcription rate dataset. By comparing this dataset with the indirect ones obtained from the mRNA stabilities and mRNA amount datasets, we are able to obtain biological information about posttranscriptional regulation processes and a genomic snapshot of the location of the active transcriptional machinery. We have obtained nascent transcription rates for 4,670 yeast genes. The median RNA polymerase II density in the genes is 0.078 molecules/kb, which corresponds to an average of 0.096 molecules/gene. Most genes have transcription rates of between 2 and 30 mRNAs/hour and less than 1% of yeast genes have >1 RNA polymerase molecule/gene. Histone and ribosomal protein genes are the highest transcribed groups of genes and other than these exceptions the transcription of genes is an infrequent phenomenon in a yeast cell.

Figure 1. Comparison of the direct and indirect TR datasets with modifications related to elongation activity.

Comparison of the new nascent TR dataset (this study) with a newly computed indirect TR datasets using the mRNA amount and the mRNA stability that takes into account the effect of the dilution (see discussion), the indirect TR dataset from , and mRNA amount with different chromatin related parameters associated with active elongation as H3 trimethylation (H3K36me3, A) or presence of elongation related transcription factors the histone acetyltranferases Esa1p (B) or Gcn5p (C) . To avoid any bias depending on the wideness of the datasets Spearman rank correlation is used. Note that the correlations are always higher for the nascent TR dataset than for the indirect datasets, and that the indirect TR datasets show a better correlation than the mRNA amount data set.

Figure 2. Comparison of the direct and indirect TR datasets with presence of RNA pol II.

Comparison of the new direct TR dataset (this study) and indirect TR datasets computed using the mRNA amount and the mRNA stability , as in Fig. 1 with RNA pol II data (average IP along the coding region of Rpb3-TAP) . To avoid any bias depending on the wideness of the datasets Spearman rank correlation is used. Note that the correlations are always higher for the nascent TR dataset (A) than for the indirect datasets (B, C), and that the indirect TR datasets show a better correlation than the mRNA amount data set (D).

Figure 3. Nascent TR data set.

A) Histogram of the yeast nascent productive TR data set in real units (molecules per hour, mol/hour) and RNA pol II densities (RNA pol II molecules per kb, mol/kb) for all the genes. The RNA pol I density is shown as a reference with a dashed red line. The natural values of RNA pol II density and TR corresponding to −2, −1 and 0 units in log scale are shown as dashed blue lines. B) Comparison of nascent TR, RNA pol II molecules per gene, RNA amount (RA, using RNA-Seq data corrected with GATC-PCR, see Methods) and mRNA stability from ref. for the different selected groups of genes (GO categories). The box plots represent the median and quartiles of the data. The whiskers show the maximum and minimum of the data set (excluding the outliers which lie beyond 1.5 times the inter quartile range).

Figure 4. A snapshot of the yeast transcriptome.

A) Selected data from the yeast nascent TR dataset. It is shown the median values for the TR for some selected groups of genes in exponential growth in YPD (glucose, this study) as well as in different conditions using previously published GRO datasets normalized with the newly computed nascent data: ^a cells growing exponentially in YPGal, (galactose: time 5 from [10]); ^b 16 min after the oxidative stress due to the addition of 0.1 mM t-BOOH ^c 10 min in 0.4 M NaCl . It is also shown the percentage of RNA pol II engaged in productive transcription (total transcription activity) relative to the exponential growth in glucose. ^e In the case of yeast cells growing in galactose this percentage is not significant because it is not possible to assume that the total amount of RNA pol II has not changed after several cell cycles using galactose as carbon source. On the other hand, we assume that the total amount of RNA pol II has not change during the short stresses. B) Schema of the yeast nucleus showing the computed distribution of transcriptional machinery during the exponential growing in rich medium. The size of the respective circles reflects the amount of RNA pol II molecules in each category.