The sequence of bases on mrna that indicates the beginning of the protein-building instructions.

3.6 The Genetic Code

Now processed, the mature mRNA transcript leaves the nucleus and is delivered to the ribosome, which is located in the cytosol. The ribosome acts as a protein factory, and the mature mRNA functions as the instructions for manufacturing. Proteins are made of amino acids, and most human proteins are 50–1000 amino acids in size. There are 20 different amino acids, and the sequence of mRNA determines the order in which the ribosome will assemble the amino acids into a protein.

The ribosome initially moves down the transcript one base at a time, reading the sequence in three-base words known as codons (Fig. 3.17). The ribosome starts translation, the assembly of a protein out of amino acids, when it encounters the start codon in the mRNA, which is the sequence AUG. The AUG codon is usually within the context of a slightly larger sequence, called the Kozak consensus sequence, which generally has the sequence GCCACCAUGG (the underlined adenine can also be a guanine). AUG codes for the amino acid methionine, and so all protein translation begins with methionine.

The start codon sets the reading frame: instead of continuing to move down the mRNA transcript one base at a time, the ribosome now reads the mRNA codons consecutively, three bases at a time (Fig. 3.18). The sequence of the triplet codon determines which amino acid is added next to the growing protein. When the ribosome reaches a stop codon, it falls off the mRNA, and the protein is complete. There are three variations of the stop codon: UGA, UAA, and UAG. The segment of mRNA before this starting point is not translated and is known as the 5′ untranslated region (5′ UTR) (Fig. 3.18B). Any mRNA past the stop codon will not be translated; this region is known as the 3′ UTR. The sequence from the start codon to the stop codon is known as an open reading frame because it is translatable.

Genetic Code Analogy

After binding to the mRNA, the ribosome begins translation at the start codon, AUG, and then moves down the mRNA transcript one codon (three nucleotides) at a time until it reaches a stop codon. Try finding the translated codons in the following sentence. The start codon—THE—will set the reading frame. The three stop codon possibilities are OKK, OOK, and OKO.

Based on the four nucleotides in RNA—adenine, guanine, cytosine, and uracil—there are 64 possible different 3-letter permutations (Fig. 3.19). There are only 20 amino acids, however, and so some of the codons are redundant, meaning that two or more codons encode the same amino acid. There are three stop codons, which end translation and do not encode any amino acid.

Many scientists worked to decipher the genetic code. Robert W. Holley, Har Gobind Khorana, and Marshall W. Nirenberg shared a Nobel Prize in physiology or medicine in 1968 for their work in determining the “key” to deciphering the genetic code. The Table in Fig. 3.19 reveals which amino acids are encoded by each codon. The code is universal: all living things have the same 20 amino acids that are encoded by these codons, indicating that this system originated very early in the development of life and has been evolutionarily conserved over time. Being that viruses take advantage of the host translational machinery and ribosomes, viral mRNAs use these same codons to encode the same amino acids in their proteins as do living things.

Study Break

Translate the two mRNA sequences found in Fig. 3.18A and B.

5.2 The Initiation Complex Assembles

Before protein synthesis starts, the two subunits of the ribosome are floating around separately. Because the 16S rRNA, with the anti-Shine-Dalgarno sequence, is in the small subunit of the ribosome, the mRNA binds to a free small subunit. Next the initiator tRNA, carrying fMet, recognizes the AUG start codon. Assembly of this 30S initiation complex needs three proteins (IF1, IF2, and IF3), known as initiation factors, which help arrange all the components correctly. IF2 physically contacts the acceptor stem of the fMet-tRNA, and this interaction is essential to stabilize the initiation complex.

IF3 recognizes the start codon and the matching anticodon end of the initiator tRNA. IF3 prevents the 50S subunit from binding prematurely to the small subunit before the correct initiator tRNA is present. Once the 30S initiation complex has been assembled, IF3 departs and the 50S subunit binds. IF1 and IF2 are now released, resulting in the 70S initiation complex (Fig. 13.19). This process consumes energy in the form of GTP, which is split by IF2.

Proteins known as initiation factors help the ribosomal subunits, mRNA and tRNA assemble correctly.

The ORF C and Core and Pre-core Proteins

This ORF has two in-frame ATG start codons. The core protein (HBcAg) is a product of 185 amino acids, encoded from the second start codon (position 1901). The pre-core protein has 214 amino acids and is translated from the first start codon (position 1814). The post-translational cleavage of the pre-core protein results in a protein, the hepatitis B “e” antigen (HBeAg) that is not used to generate the mature virion, but is secreted from infected cells as a soluble protein acting as an immunomodulator.

3.2.1 Translation blocking

Translation-blocking morpholinos bind to the start codon or 5′ UTR of a target transcript (typically anywhere between position − 50 and + 25) and sterically block progression of the ribosomal initiation complex. Unlike siRNA antisense technologies, morpholinos generally do not cause degradation of the targeted mRNA transcript. This means that RT-PCR is not a suitable means for ascertaining the efficacy of translation-blocking morpholinos. Ideally, the level of knockdown should be determined using an antibody to the protein of interest, although this is not always feasible. As an alternative, the knockdown efficiency can be assessed by coinjecting the morpholino and in vitro transcribed mRNA encoding a version of the target gene containing a tag (e.g., HA, FLAG, GFP).

Endogenous Regulation

Translation initiation in eukaryotes

In eukaryotes, the smaller subunit of the ribosome finds the AUG start codon in either of the following two ways: (1) It recognizes the AUG start codon by a special secondary structure of RNA in the vicinity of this codon, or (2) it first binds to the 5′-end of mRNA called the “5′ cap” (or RNA cap or RNA 7-methylguanosine cap or RNA m7G cap) and then scans the mRNA in the 5′ → 3′ direction until the AUG start codon is found.

In 1984, Marilyn Kozak discovered an inhibitory effect of “redundant” AUG triplets between the 5′ cap and the “true” AUG start codon. This effect is commonly known as Kozak's rule, which Igor Rogozin and Luciano Milanesi successfully retested on a tremendously large array of new genomic data in 2001. In 2002, Alexey Kochetov and Vladimir Shumny described composite translation initiation elements within the mRNA 5′-UTR in higher plants. In 2004, Akinori Sarai discovered translation-associated SNPs in plants. In 2006, Oxana Volkova and Nikolay Kolchanov examined potential open reading frames (ORFs) in eukaryotic mRNA 5′-UTRs. Finally, in 2011, Georgii Bazykin and Alexey Kochetov found the evolutionary conservatism of alternative AUG start codons in eukaryotic mRNAs.

2.2 Stable Transformation of Tomato Trichomes

Although particle bombardment can indicate trichome specificity, the number of trichomes hit might not be very high and the procedure is destructive. Alternatively, for a more in-depth study, one can make stable transgenic plants expressing a promoter–reporter construct. In this section, a construct with the SlTPS9 promoter driving a GUS–sYFP1 fusion was introduced into S. lycopersicum cv. Moneymaker to verify the promoter activity of SlTPS9 in stably transformed plants.

2.2.3 Transgene Expression

Five independent transgenic primary (T0) lines were obtained after PCR verification of the presence of the YFP gene on genomic DNA isolated from leaves. Based on YFP expression in T0 stems and leaves, one transgenic line was selected for further analysis. Transgene expression was further confirmed by observing YFP fluorescence using an EVOSfl inverted microscope (http://www.thermofisher.com) in stems and leaves collected from the transgenic T0 plants. The selected transgenic line exhibited strong YFP fluorescence in glandular stem and leaf trichomes (Fig. 4B and D). Fig. 4 shows that fluorescence was specific to the four glandular head cells of the type VI trichomes. This result confirms the trichome-specific expression of SlTPS9 as found by Bleeker, Diergaarde, et al. (2011) and Bleeker, Spyropoulou, et al. (2011). It furthermore shows that the method presented here is suitable to drive transgene expression specifically in the glandular trichomes of stably transformed tomatoes.

Fig. 4. The SlTPS9 promoter drives sYFP expression specifically in type VI glandular trichomes of stable transgenic tomato plants. (A) Stem trichomes of untransformed plant. (B) Stem trichomes of tomato transformed with the SlTPS9:GUS–sYFP construct. (C) Leaf trichomes of untransformed plants. (D) Leaf trichomes of tomato transformed with the SlTPS9:GUS–sYFP construct. Images: (i) Normal light (brightfield), (ii) GFP filter, and (iii) merged image of (i) and (ii). Images were taken using the EVOSfl inverted microscope (http://www.thermofisher.com). Arrows indicate type VI glandular trichomes and type V nonglandular trichomes. Scale bars: 400 μm.

2.5 Predicting translation initiation rates across a genome

The RBS Calculator's reverse engineering mode can predict the translation initiation rate of every start codon in a bacterial genome. These predictions enable you to find the correct start codons of open reading frames, estimate the expression levels of proteins within an operon, and identify internal start codons that exhibit significant translation and produce variant proteins. The RBS Calculator software uses MPI for parallel programming and its calculations may be distributed across many processors on a supercomputer.

For example, the E. coli K12 genome contains 629,738 start codons. Over 600 internal start codons have significant amounts of translation, compared to the annotated open reading frame. In particular, the RBS Calculator correctly predicts the significant translation (6700 au) of a five-amino acid peptide encoded within its 23S rRNA (Tenson et al., 1996).

2.3 Synthetic ribosome binding sites

Ribosome binding sites is a nucleotide sequence of mRNA generally located upstream from the start codon and it is used to lead the ribosome to the right position on the mRNA during the protein translation initiation. Selecting the translation initiation site (TIS) is a critical and the rate-limiting step during the translation progress. In prokaryotes, the RBSs with Shine-Dalgarno (SD) sequence play an significant role in the translation initiation site’s identification within mRNA by ribosome [19]. Bacterial translation initiation needs that the RBSs of mRNAs uses single-stranded conformations, and RBS sequences are rich in A and G, and it is localized from 3 to 14 base pairs upstream from the beginning of a gene where the start codon is usually AUG [20]. It is reported that the translation efficiency is related to the secondary structure in the mRNA initiation region, and the translation initiation is fully based on the entire initiation region’s spontaneous unfolding, while ribosomes seemingly do not recognize nucleotides outside the SD region and initiation codon [21]. However, the working mechanism of RBS is different in eukaryotic cells. For eukaryotes, ribosome recruitment is normally mediated by the cap at the 5’-end of the mRNA. After the ribosome fund the cap, it will go downstream until meet the first start codon “AUG,” which is usually contained in a special sequence named Kozak motif which is 5’-A(G)CCAUGG-3’. Marylin Kozak performed important researches on vertebrate mRNAs to characterize optimal TIS consensus sequence. Within this motif, nucleotides conservation in crucial positions, namely a purine at −3 and a G at +4, is critical for TIS recognition [22].

The RBS engineering is a valuable approach for modulating gene translation efficiency and thus optimizing gene expression, and such strategy was wieldy used, ranging from regulating gene circuits to metabolic pathway. The known RBSs were stored in some libraries, such as the BioBrick Registry (http://parts.igem.org), thus RBSs with interest can be selected to test their strengths of gene expression [52,53]. Luo et al. used designed RBSs and took the GFP as a reporter to enhance production of target protein under translational level in Streptomyces coelicolor M145 [19]. Eriksen et al. demonstrated that the decreased protein production is caused by the reduced stability of mRNA, and they used different RBS sequence from the E.coli lacZ gene contributed to about 100-fold span in protein expression [54]. Oesterle et al. established smart RBS-libraries based on algorithms design to sample the entire space of translation space using CRISPR/Cas9-based genome editing method, and such strategy allows for efficient and stable fine tuning of chromosomal functions with smallest changes [55].

On the other hand, the RBSs can be designed in silica to achieve special targets. In bacteria, the RBSs Calculator is a designed method which is used to predict and control translation initiation and protein expression [56]. This method could predict the translation initiation rate for every start codon during one mRNA transcript, which might also optimize the sequence of a synthetic RBS to obtain one targeted translation initiation rate [56]. Nguyen et al. developed a reference interaction site model (RISM) to predict the RNA molecules folding event in the presence of monovalent and divalent cations [57]. Clauwaer et al. present a novel neural network DeepRibo for delineation and annotation of expressed genes in prokaryotes [58].

RNA molecules use various mechanisms to control gene expression and one of such regulation examples is riboswitches. For bacteria, riboswitches are RNA structures, which are normally used to detect a kind of metabolites and ions to regulate gene expression at translational and transcriptional level. Currently, approximately 40 different classes of riboswitches which have been discovered [59]. In 2002, the riboswitches were firstly experimentally validated and open a new era to regulate gene expression [60-63]. Among the first discovered riboswitches, the glycine riboswitch includes two tandem glycine-binding aptamers in many examples, which is followed by a single expression platform to tune the expression of gene in response to glycine, and Babina et al. proposed the models of how tandem glycine riboswitches regulated its native locus within the genome and determined how its disruptions to glycine riboswitch function influence organism [64]. One of the most known riboswitches is RBS-based riboswitch. In generally, there are two typical RBS-based riboswitch, RBS-based transcriptional on-riboswitches and RBS-based transcriptional off-riboswitches (Fig. 2.2A and B). The ligand binding by molecules could recreate secondary structure of mRNA to promote translation process via exposing RBS for ribosome access in bacteria [65]. Another riboswitch is named ribozyme-based riboswitch, which consists of two factors, cis-acting hammerhead ribozyme (regulation domain) and RNA aptamer (sensor domain) (Fig. 2.2C). Like the RBS-based riboswitch, there are two types of ribozyme-based riboswitches, but the ribozyme-based on-riboswitches are widely used. For ribozyme-based riboswitch, the entrapping of ligand to the sensor domain controls the activity of ribozyme thus leads to change the RNA molecule stability and functional translation [66]. It is believed that across all existing domains of life, riboswitches appear to show one of the most highly conserved mechanisms for regulating a wide range of biochemical pathways [67]. Recently, the reaction states of riboswitch RNA were captured by mix-and inject XFEL serial crystallography [68]. Nowadays, the riboswitches were wildly used to specifically detect target molecules and regulate the report gene expression. Zhou et al obtained a synthetic glycin-OFF riboswitch to dynamically control the 5-aminolevulinic’s acid biosynthesis [69]. Dwidar et al. isolated a histamine-binding RNA aptamer and used it to design robust riboswitch to regulate the expression of genes in the presence of histamine [70].

A From ORFs to Genes

Open reading frames (ORFs) are defined as spans of DNA sequence between start and stop codons. Automatic extraction of all possible ORFs from error-free genomic DNA with known genetic code would seem a straightforward task (which can be performed online at, e.g. www.expasy.ch/www/dna.html or www.ncbi.nlm.nih.gov/gorf/gorf.html). In real life this step is complicated by DNA sequencing errors that may lead to missed or falsely assigned start/stop codons and consequently to extended or shortened ORFs. Given a list of all possible ORFs in a given genome, deciding on which of them constitute genes may be difficult. First, partially or fully overlapping ORFs often occur on the same DNA strand. Second, competing ORFs are commonly present on different DNA strands. Finally, even in the absence of contradictions there is no certainty that an ORF, particularly a short one, actually codes for a protein (Fickett, 1996).

In many cases, genes are identified based on statistically significant sequence similarity of translated ORFs with known protein sequences (Gish and States, 1993). This method is used, for example, in the Analysis and Annotation Tool (http://genome.cs.mtu.edu/aat.html) developed by X. Huang et al., (1997). In the absence of significant database hits, gene identification methods based on coding potential assessment and recognition of regulatory DNA elements must be applied. The most widely used program for finding prokaryotic genes, GeneMark (http://genemark.biology.gatech.edu/GeneMark; Borodovsky et al., 1994), employs a non-homogeneous Markov model to classify DNA regions into protein-coding, non-coding, and non-coding but complementary to coding. GeneMark and similar programs (see Fickett, 1996) rely on organism-specific recognition parameters and thus require a sufficiently large training set of known genes from a given organism for successful gene prediction.

Inferring genes by signal and by similarity represent the so-called intrinsic and extrinsic approaches (Borodovsky et al., 1994), which should ideally be used in combination. The quality of gene prediction can be further improved by using additional available evidence, such as operon structure, location of ribosome-binding sites and predicted signal peptides.