The global process of gene expression
The main steps of the gene expression process
Gene expression consists in cellular processes from gene to protein (Fig. 1). Which proteins are present, and their precise amounts are crucial for cell performance in terms of growth, adaptation and survival.
The first step of gene expression is transcription, the process of RNA synthesis from DNA (Fig. 1). Catalyzed by RNA polymerase, each gene in double-stranded DNA is transcribed to numerous copies of single-stranded RNA, in which the deoxyribose sugar is replaced by ribose and the thymine base by uracil. The three steps of transcription are initiation, elongation and termination. Briefly, during transcription initiation, RNA polymerase binds to the promoter located at -35 and -10 nucleotides upstream of the transcriptional start site and separates the DNA strands, providing the single-stranded template needed for transcription. During elongation, the RNA polymerase « reads » the template strand one base at a time and builds from 5′ to 3’ an RNA molecule out of the complementary nucleotides. Once the terminator signals are transcribed, they cause the transcript to be released from the RNA polymerase.
Second, translation occurs to synthetize a polypeptide chain (a protein) from the mRNA template. Translation is carried out by ribosomes. In prokaryotes, the ribosome is composed of two subunits, the small 30S subunit and the large 50S subunit, leading to a sedimentation coefficient of 70S for the assembled ribosome. Translation is composed of three steps, initiation, elongation and termination. The process of translation will be described in more details in section III.
In time, the mRNA concentration varies because of mRNA degradation (Fig. 1). In prokaryotes, and particularly in E. coli, the average mRNA half-life is relatively short, about few minutes (Bernstein et al., 2002). The degradosome is a large protein complex which catalyzes most of mRNA degradation in E. coli (Carpousis, 2007). The main components of E. coli degradosome are the RNase E, the polynucleotide phosphorylase (PNPase), the RNA helicase B (RhlB) and the enolase. RNase E is an endoribonuclease that preferentially cleaves single-stranded A/U rich regions, PNPase is a 3’- 5’ exoribonuclease and RhlB unwinds RNA. The role of enolase is not yet well defined. In E. coli, initiation of mRNA degradation starts with the endonucleolytic cleavage by the RNase E (Bouvier and Carpousis, 2011).
As the mRNA molecule, protein can also be degraded over time (Fig. 1). However, proteins have relatively longer stability than mRNAs with half-lives in the range of hours. Protein degradation is mediated by protein complexes usually composed of a protein with an ATPase activity (from the AAA family such as ClpA, ClpX andt ClpY) and a protease for the proteolysis (for example ClpP and ClpQ) (Chandu and Nandi, 2004). The role of the ATPase is to unfold the protein via conformational changes driven by ATP hydrolysis and to address the unstructured protein to the protease. Then the protease degrades the protein into smaller peptides and, finally, into amino acids, which are recycled into the cellular pool. Only the ClpAP, ClpXP et ClpYQ associations seem to be present in E. coli (Picard et al., 2009). In some cases (like for Lon and FtsH) the enzymes harbor in the same polypeptide the two domains conferring the ATPase and proteolysis activities.
At last, both the mRNAs and proteins are diluted over time as cells divide (Fig.1). The rate of dilution of mRNAs and proteins due to cellular growth depends on the generation time of the microbial cells.
Regulations of gene expression
Gene expression is constantly regulated in response to environmental changes, (e.g. nutrient availability, physico-chemical conditions, etc). In prokaryotes, transcriptional regulation has been thought for many years as the main regulation of protein synthesis. Transcription regulation involves both global and specific regulations (Berthoumieux et al., 2014; Gerosa et al., 2014). An example of global regulation that affects the transcription of all the genes, is the increase of the RNA polymerase quantity and activity with the growth rate (Ehrenberg et al., 2013; Klumpp and Hwa, 2008; Klumpp et al., 2009). Specific regulation affects gene expression of a rather limited number of genes via transcriptional factors binding to specific promoters (Martínez-Antonio, 2011). The transcriptional factors act either as enhancer, repressor or inducer of transcription initiation in response to environmental changes.
However, it was shown that post-transcriptional regulations are also involved in mRNA level regulation. mRNA stability differs between E. coli transcripts depending on sequence determinants such as 5’end secondary structure (Arnold et al., 1998) or codon bias (Esquerré et al., 2015). The mRNA half-life is also dependent on growth conditions, it increases in E. coli when the growth rate decreases (Chen et al., 2015; Esquerré et al., 2014). Concerning the dilution of the mRNA molecules due to cell division, this post-transcriptional process is often neglected in E. coli because the generation time (in the range of hours) is generally much longer than most of the mRNA half-lives (in the range of minutes).
Variation in mRNA level is however not sufficient to explain protein level variability. Protein level is not totally correlated with the mRNA abundance in bacteria. Correlation coefficients between 0.2 and 0.6 were reported between the proteome and transcriptome data in in E. coli (Lu et al., 2007), Lactococcus lactis (Dressaire et al., 2009), and Desulfovibrio vulgaris (Nie et al., 2006). Taniguchi and colleagues confirmed in E. coli at the single cell level this low correlation (Taniguchi et al., 2010). Therefore, the role of translational regulations in gene expression regulation was largely suspected.
Stochasticity of gene expression
Stochastic gene expression is one reason of phenotype heterogeneity between genetically identical cells. Variability of gene expression between cells is manifested at the levels of transcription and translation. Taniguchi and colleagues performed a fluorescent reporter library to measure transcriptome and proteome in E. coli and showed different mRNA and protein copy numbers between isogenic cells (Taniguchi et al., 2010). The heterogeneity of proteome, which was not correlated to transcriptome, suggests the existence of translation variation between cells. Measurement of translatome at the single-cell level could be performed to quantify this phenomenon. Transcription and translation, like other biological processes, are inherently stochastic based on the random encounter between molecules (for example between RNA polymerase and gene promoter for transcription or ribosome and mRNA for translation). This probability of encounter depends on the numbers of molecules (RNA polymerase, ribosomes, mRNAs) which could be different between cells.
Stochasticity of gene expression may occur at another level: within the same cell. Indeed, the translational levels of mRNA copies of the same transcript could be different. Experiments using fluorescence method to monitor translation of single molecule in a single human cell (Morisaki et al., 2014) showed that there was a variability of translation activity between mRNA copies of one mRNA. Within the same cell, heterogeneity may come from the stochastic diffusion of molecules. The local concentrations of ribosomes and mRNAs are crucial factors for the random encounter between mRNAs and ribosomes during initiation. Ribosomes and mRNAs are not homogeneously localized within the cell: polysomes are enriched at the polar zones, spatially excluded from the nucleoid (Bakshi et al., 2012, 2015) whereas mRNAs are synthetized at the nucleoid. In bacteria, mRNAs could be present at a very low copy number and sparsely distributed in the cell, the main source of variability according to some models (Kaern et al., 2005). Moreover, cells are crowed with macromolecules that could limit the diffusion and random encounter between mRNAs and ribosomes (Klumpp et al., 2013).
In the next sections, we will describe the process of translation (III), its regulation (IV) and the interactions of translation with other cellular processes (V).
How does translation work?
The principal actor of the translational machinery is the ribosome, a ribonucleoprotein complex composed of around 60% of ribosomal RNA (rRNA) and 40% of proteins. Ribosome is composed of two subunits: the large one is constituted from two rRNAs, 5S and 23S (116 nts and 2900 nts respectively) and 33 ribosomal proteins (L1 to L33). The small subunit includes the 16s rRNA and the 21 ribosomal proteins (S1 to S21) (Fig. 2).
Figure 2: Structure of E. coli ribosome. The small subunit is shown in the left with the 16S rRNA in blue and the small (S) ribosomal proteins in orange. The large subunit is in the right with the 5S and 23S rRNAs in red and the large (L)-proteins in green (Shajani et al., 2011).
The small 30S subunit assures the ribosome binding to mRNA and controls translation fidelity via precise binding between mRNA and tRNA. The large 50S subunit catalyzes the peptide bond formation. The tRNA makes the connection between the mRNA sequence read by nucleotide triplet (corresponding to a unit named codon) and the neosynthesized peptide. One charged tRNA comprises at one end the anticodon, a sequence complementary to the codon read on the mRNA and at the other end the corresponding amino acid. Note that there is a tRNA specific to the start codon: noted fMet-tRNA, this initiator tRNA is loaded with a formylated methionine amino acid.
In E. coli, genes coding for rRNAs are organized in operons and present in 7 copies. Each rrn operon contains 3 genes coding for the 5S, 16S and 23S rRNAs. These operons are under the control of two promoters P1 and P2. The two promoters are positively regulated by the transcription factor Fis and negatively regulated by the H-NS factor. The concentrations of Fis and H-NS depend on the growth rate with a low Fis concentration at the low growth rate or on the onset of the stationary phase whereas H-NS level increases in stationary phase (Schneider et al., 2003).
Table of contents :
Multiplexing of polysome profiling experiments to study
translation in Escherichia coli
Analysis of RNA-Seq data to study translatome
Co-directional regulations of transcription and translation in Escherichia coli: more concentrated mRNAs are more efficiently translated
CONCLUSIONS ET PERSPECTIVES