Biochemistry - The Chemical Reactions of Living Cells, Volume 3 - D. Metzler 1980

Biochemical Genetics and the Synthesis of Nucleic Acids and Proteins
Translation of Genetic Information. Protein Synthesis
Protein Synthesis

The formation of polypeptide bonds on Ribosomes is generally divided into three processes: initiation, elongation, and termination [98]. Protein Synthesis begins with an initiation codon, most commonly the Methionine codon AUG. The GUG codon, properly positioned within the mRNA chain, can also serve as an initiation codon; in this case, it specifies methionine rather than valine. The base sequence preceding the initiation codon can also play a crucial role in recognizing the "start" signal, as evidenced by the fact that AUG and GUG codons occur not only at initiation sites.

a. Initiation [99]

In Bacteria, peptide chains always begin with The amino acid N-formylmethionine. Thus, the first step in protein synthesis consists of aligning the initiation codon at the required site on the ribosome and binding a tRNA molecule "loaded" with N-formylmethionine to it^{1) 2}. This process is relatively complex, partly because ribosomes must select the true initiation codon from many AUG and GUG codons contained within the interior of the mRNA chain. This is apparently achieved through base pairing between the ACCUCCU sequence located at the 3' end of the 16S ribosomal RNA and the complementary initiation sequence in the mRNA (Fig. 15-14) [100, 101]. Such binding also requires the ribosomal protein S1 (also known as i-factor). Its function is believed to be the Maintenance of the 3' end of the 16S rRNA in an "open" conformation rather than as a hairpin loop [102].

In addition to ribosomal Proteins, three other proteins required for initiation are known as initiation factors: IF-1, IF-2, and IF-3. The latter exists in at least two forms: IF-3a and IF-3ß. The Role of initiation factor IF-3 appears to be the Formation of the mRNA–16S rRNA initiation complex (Fig. 15-15, stage a) and its stabilization. At this stage, the "loaded" formylmethionyl-tRNA (fMet-tRNA) molecule binds to a second initiation factor, IF-2, which has previously bound a GTP molecule (stage b)²⁾. Subsequently, this complex binds to the 30S subunit, presumably in such a way that the tRNA anticodon forms complementary Base Pairs with the mRNA start codon (Fig. 15-5, stage c). The binding of the IF-2 complex is aided in some way by a third initiation factor, IF-1, whose exact role remains fully undetermined. IF-3 then leaves the complex (stage d), and the 50S ribosomal subunit joins to form an intact ribosome, after which IF-1 is released (stage e).

¹⁾ This binding is accompanied by a side effect: a shift in the equilibrium between the 30S and 50S subunits and intact 70S ribosomes toward increased dissociation [91].

²⁾ Some researchers believe that the fMet-tRNA–IF-2 complex binds to the ribosomal 30S subunit prior to the mRNA [103].

Although much remains unclear regarding the exact events occurring during these reactions, it is now well established that the "loaded" methionyl-tRNA molecule does indeed bind, as shown in Fig. 15-5, to the so-called P site (peptidyl site) of the 50S ribosomal subunit. According to an earlier hypothesis—which cannot yet be considered entirely disproven—initial binding occurs at the "A site" (aminoacyl site), followed by translocation to the P site. This hypothesis was prompted by the fact that GTP Hydrolysis occurs during The final stage of initiation (Fig. 15-15, stage f), where IF-2 is released as an IF-2–GDP complex. Data discussed in the next section indicate that GTP hydrolysis is required for translocation during the growth (elongation) of the peptide chain.

Class="center">

FIG. 15-14. Proposed hydrogen-bonding scheme between a 16S rRNA fragment and the A-protein initiation site in phage R17 RNA. The Introduction/11.html">Secondary Structure of the 16S rRNA fragment shown in this diagram is assumed to be stable under physiological conditions. An alternative hydrogen-bonding scheme, corresponding to an equally stable base complex whose protruding loop increases to 9 bases while a quartet of bases CCUU at the 3' end pairs with AAGG at the 5' end of the lower stem, is not shown [101].

That GTP hydrolysis is indeed necessary for initiation is evidenced by the fact that 5'-guanylylmethylenediphosphonate (a GTP analogue in which the central and terminal phosphorus atoms are linked by a methylene bridge) can be used in place of GTP at the initial binding stage. This analogue can replace GTP in all Cytology/cytology/16.html">Early stages of initiation up to binding with the 50S ribosome, but fails to function at the final step because it is not hydrolyzed. The precise significance of GTP hydrolysis has not yet been established. It is possible that it provides the energy required for the rearrangement of ribosomal components (discussed above), or simply serves to release the IF-2–GDP complex (for instance, if the IF-2–GTP complex is tightly bound while the IF-2–GDP complex is weakly bound).

FIG. 15-15. Initiation of Protein synthesis in bacterial ribosomes.

Obviously, these two hypotheses regarding the potential significance of GTP hydrolysis are not mutually exclusive.

Some insight into the spatial arrangement of ribosomal proteins involved in initiation can be gained from the observation that Antibodies against proteins S19 and S21 block the formation of the initiation complex with fMet-tRNA. It is also known that antibodies against proteins S2, S18, and S20 block the binding of factor IF-3. Cross-linking experiments have shown that IF-2 and S19 are located close to each other, whereas IF-3 lies adjacent to S12.

b. Elongation of peptide chains [104]

Once the fMet-tRNA initiation complex has occupied the P site, the growth of the peptide chain can begin, during which amino acid residues are sequentially added to the C-terminus of the growing polypeptide chain. Elongation proceeds in three successive steps that are repeated continuously until the entire polypeptide is completed:

1. Codon-specific binding of a tRNA molecule, loaded with the next amino acid, to the A site.

2. Peptide bond formation. At this step, the tRNA residing in the P site is discharged and, along with the growing peptide chain, is transferred to the A site.

3. "Translocation" of the peptidyl-tRNA from the A site to the P site. This step also involves the release of the used tRNA from the P site and the movement of the mRNA, bringing the next codon into the A site. The energy required for translocation is derived from the hydrolysis of a single GTP molecule.

FIG. 15-16. Elongation of the growing peptide chain.

The first of these three steps—namely, the binding of aminoacyl-tRNA to the A site—depends on the protein elongation factor T, or EF-T. This factor is a mixed Ts·Tu dimer, where one component, Ts, is a stable protein with a Molecular Weight of ~42,000. The other component, Tu, is a membrane-associated protein [104a] with a molecular weight of approximately 44,000, and its concentration in EF-T several times exceeds that of Ts. Protein Tu forms a complex with GTP and aminoacyl-tRNA molecules (though it binds fMet-tRNA poorly). As shown in Fig. 15-16, the interaction of GTP with the Ts·Tu complex releases the Ts protein and forms a GTP-Tu complex (stage a), which then binds to aminoacyl-tRNA and the ribosome (stages b and c). These reactions take place at the peptidyltransferase center on the 50S subunit, in a region containing proteins L7 and L12.

In the subsequent step (stage d), the peptide chain is transferred to the a-amino group of the aminoacyl-tRNA occupying the A site via a simple displacement reaction. In reality, however, this reaction is more complex than depicted in the figure. It is accompanied by the Cleavage of bound GTP and the release of Pi and the Tu–GDP complex. The latter, as shown in the figure, interacts with Ts, regenerating the Tu·Ts dimer and releasing GDP. Thus, the overall reaction consists of GTP cleavage coupled with peptide bond synthesis. The Chemistry of the reaction does not strictly require GTP hydrolysis. Nevertheless, it remains unclear how close the ends of two neighboring tRNA molecules lie to one another; the distance between them could be considerable. Proteins L7 and L12 are exceptionally rich in Alanine and feature a high proportion of a-helical regions, resembling the Muscle protein Myosin in this regard. Consequently, it has been suggested that these proteins act as part of a "mini-muscle" which, using The energy released during GTP hydrolysis, shifts specific Regions of the ribosomal complex to bring the amino group and the peptidyl group into close proximity during the peptidyltransferase reaction.

A chemically interesting and challenging question concerned which hydroxyl group (the 2'- or 3'-) of the terminal adenosine residue of tRNA carries the activated aminoacyl or peptidyl group. It is now believed that rapid equilibrium can be established between these groups via an orthoester (tetrahedral adduct) [equation (15-7)].

A modified Yeast phenylalanine tRNA complex ending in 3'-deoxyadenosine attaches an amino acid via the amino acid-activating enzyme (aminoacyl-tRNA synthetase) [Equation (11-2)], whereas phenylalanine tRNA ending in 2^,-deoxyadenosine fails to react. This suggests that the 2'-hydroxyl group serves as the initial aminoacylation site. However, studies on the ability of synthetic 2'-phenylalanine tRNA ending in 3'-deoxyadenosine to act as a peptide acceptor in the peptidyltransferase reaction on ribosomes revealed that it is inactive in this reaction. The 3'-phenylalaninedeoxyisomer proved to be active [105]. Thus, aminoacylation most likely occurs initially at the 2' position, followed by isomerization prior to the transpeptidation reaction [Equation (15-7)] to yield 3'-aminoacyl-tRNA. Nevertheless, later studies showed that in both baker's yeast and E. coli, some tRNAs are initially aminoacylated at the 2'-hydroxyl groups and others at the 3'-hydroxyl groups [106, 107].

The Third Stage of Polypeptide chain elongation on ribosomes depends on another "elongation factor," namely EF-G, which possesses "GTPase" activity. Evidence indicates that factor G also binds to ribosomal proteins L7 and L12 or in their immediate vicinity and competes with EF-Tu for positioning. The role of factor EF-G, much like that of factor EF-T, remains unclear, yet it is known that GTP hydrolysis is required for the translocation reaction to proceed, and the used tRNA cannot dissociate from the P-site until it binds to EF-G.

Mutants in the genes encoding factors EF-Tu, EF-Ts, and EF-G have been discovered, which may facilitate the in vivo study of these proteins' roles (analogous mutants for initiation factors IF-1, IF-2, and IF-3 have not yet been obtained).

c. Termination of Polypeptide Synthesis [108]

Ribosomes faithfully translate Genetic information, adding Amino Acids to the peptide chain until a stop codon is reached. At this point, a termination factor comes into play, apparently binding directly to both the stop codon and the mRNA. According to available data, in E. coli the termination factor RF-1, a protein with a molecular weight of ~ 44,000, recognizes the UAA and UAG codons, whereas factor RF-2 (molecular weight ~47,000) can recognize UAA or UGA. There are several hundred molecules of these termination factors in The Cell. Somehow, the termination factor must not only recognize the correct codon but also catalyze the hydrolytic Cleavage of the peptide chain from the tRNA. After a significant portion of The nucleotide sequence of viral mRNA molecules was deciphered, it was unexpectedly discovered that the genes encoding the coat proteins of RNA-containing phages (Fig. 15-19) sometimes end with two consecutive stop codons. This evidently ensures the reliability of Translation termination even if one of the stop codons is lost. Note that at the end of the i Gene of the E. coli lac Operon, alongside the TGA codon shown in Fig. 15-4, There is a second stop codon located five codons further downstream. It is currently unclear how general the duplication of stop codons at gene ends might be.

d. Polyribosomes

Under certain conditions, ribosomes isolated from Cells precipitate as clusters (polyribosomes) consisting of six or more ribosomes. It can be demonstrated that in these polyribosomes (or polysomes), the ribosomes are linked together by an mRNA strand. It is now generally accepted that polyribosome formation is due to a single mRNA molecule being translated simultaneously by multiple ribosomes. As soon as the 5' end of the mRNA leaves a ribosome, it can immediately bind to another ribosome, initiating the Translation of the next peptide chain, and so on. The number of ribosomes assembled into a polyribosome is determined by the length of the mRNA.

e. Protein Synthesis in Eukaryotic cells

There are many similarities between protein synthesis in bacterial and eukaryotic cells. In eukaryotes, the role of the initiator tRNA is played by a special methionyl-tRNA (Met-tRNA_f). (The subscript indicates that this tRNA can be formylated by the bacterial enzyme system. However, in The Eukaryotic Cell, this methionyl-tRNA remains unformylated.) Just as in bacteria, the incorporation of methionine into internal positions of a protein is carried out using a different transport RNA (Met-tRNA_m). Protein synthesis in eukaryotic cells requires at least three initiation factors: eIF-1 through eIF-3 [109]. A major difference from prokaryotic systems is that in eukaryotic cells, the initiator aminoacyl-tRNA binds to the ribosome first, followed by the mRNA. For eIF-3 recognition at the 5' end of mRNA in eukaryotes, the presence of a "cap" containing 7-methylguanosine may be required [109a]. Eukaryotic cells also contain elongation factors [110] EF-1 and EF-2, and only a single termination factor rather than two as in bacteria.

f. Codon-Anticodon Pairing

We now return to the fundamental problem: how the correct amino acid is delivered to the ribosome at the right time during Polypeptide chain synthesis. This process is based on the accurate "recognition" of the complementary codon in mRNA by the tRNA anticodon. An unexpected finding was the presence of inosine (I) in the anticodons of yeast tRNAs (though not in most E. coli tRNAs). It was also unexpected that a single cell contains fewer than 61 types of tRNA molecules (61 = 64 - 3 stop codons). Based on these facts, Crick formulated the so-called Wobble Hypothesis in 1966 [111]. According to this hypothesis, the first two bases at the 5' end of the codon (and the 3' end of the anticodon) must pair According to the same rules as the bases in DNA. As for the third base pair (the 3' end of the codon and the 5' end of the anticodon), steric constraints are less stringent, meaning that some deviations from exact complementarity are possible in this case. Crick proposed the following "rule" for third-base pairing:

Based on this hypothesis, all observed deviations from classical base pair formation (A-U, C-G) in the Watson-Crick helix can be explained. Thus, an anticodon with G at the 5' end can pair with codons ending in C or U. Anticodons ending in C or A will pair precisely. Anticodons ending in U can form pairs with codons having A or G at the 3' position. Anticodons with I at the 5' position can recognize codons

with any of these three bases at the third position. Comparing all these considerations with the data presented in Table 15-2 makes it immediately clear why fewer than 61 anticodons may suffice for a single cell. The same amino acid corresponds to multiple codons, and often The Nature of the base at the 3' position does not affect the meaning of the codon. This results in a certain economy in nature, consisting in the utilization of only a subset of anticodons rather than all of them. Crick demonstrated from a chemical standpoint that his hypothesis is plausible even if one assumes that the spatial relationships between bases during "wobble" pairing differ from classical Watson-Crick pairing. This is illustrated in Fig. 15-17, which shows the binding of inosine to C (a standard Watson-Crick base pair), as well as to A and U. The term wobble does not capture The Essence of the hypothesis particularly well, yet the hypothesis itself has allowed for A number of correct predictions. For instance, in accordance with this hypothesis, a mere three tRNAs are sufficient to recognize six Serine codons. Indeed, three tRNAs for serine have been discovered in E. coli.

FIG. 15-17. Pairing of inosine with cytosine (Watson-Crick pairing) and inosine with uracil (wobble pairs) (Watson J., Molecular Biology of the Gene, New York: W. A. Benjamin, 1976).

g. Aminoacyl-tRNA synthetases [112]

Another process associated with recognition is also of crucial importance for protein synthesis: the Selection of the correct amino acid by aminoacyl-tRNA synthetase and its transfer in an activated form to the corresponding tRNA [Equation (11-2)]. In bacteria, there is one aminoacyl-tRNA synthetase for each of the 20 amino acids. Each such synthetase must select a specific Amino Acid and its corresponding tRNA; the exact same enzyme can transfer the amino acid to all isoaccepting tRNAs specific for that amino acid.

Many attempts have been undertaken to determine which specific part (or parts) of the tRNA molecules participates in their recognition by the synthetase. To this end, nucleotide sequences in isoaccepting tRNAs have been compared, and studies of chemically modified and fragmented tRNA molecules have been conducted. A number of other approaches have also been utilized. The results of all these studies have suggested that no universal recognition method exists. In several cases where the anticodon structure was disrupted, the synthetase failed to attach amino acids to the chemically modified tRNA. This is a remarkable fact because, as experiments with crystalline tRNAs have shown, the codon is located approximately 7.5 nm away from the CCA end of the tRNA (Fig. 2-24). Despite the fact that synthetases are large Enzymes (the molecular weight of most is approximately 100,000), it remains difficult to understand how precise anticodon recognition can occur without the tRNA molecules undergoing significant conformational changes that bring the anticodon close to the CCA end of the tRNA. In some tRNAs, the anticodon does not participate in recognition. For example, in the case of yeast Phe-tRNA, it has been proven that residues located in the dihydrouridine stem and the upper part of the acceptor stem play a decisive role in recognition [112].

Rich drew attention to the fact that one side of the tRNA molecule in its normal configuration has a relatively constant structure, and suggested that this part is intended for interaction with some component of the ribosomal system. The structure of the other side of the tRNA molecule (Fig. 2-24), including the dihydrouridine loop and the acceptor stem, is considerably more variable. It is possible that this very region serves as the binding site for aminoacyl-tRNA synthetases. It has also been suggested that aminoacylation and binding to ribosomes may be accompanied by conformational changes at one or more sites on the tRNA molecule [113]. It is possible that the TψCG loop, inaccessible in native tRNA, opens up and interacts with 5S RNA in the ribosome. The short arm of this structure may swing relative to the end of the longer arm bearing the anticodon. The most effective method for detecting conformational changes is proton magnetic Resonance (PMR) (Ch. 2, Sec. 3.7), which makes it possible to obtain signals directly from each proton of the tRNA molecule involved in intramolecular Hydrogen bond formation [114].

The selection of the appropriate amino acid by aminoacyl-tRNA synthetase is of paramount importance. However, it is difficult to imagine an Active Site capable of clearly distinguishing the structure of two such similar compounds as isoleucine and valine. One pathway for precise amino acid selection could be a kinetic "proofreading" mechanism, similar to the one described for DNA polymerase I (Sec. D.4). Indeed, it has been demonstrated that valine erroneously attached to isoleucine tRNA undergoes rapid hydrolysis by the synthetase [114a], significantly reducing the probability of valine being incorporated into a protein in the wrong position.

h. Effect of Antibiotics on Ribosomes

The action of many currently known and most effective antibiotics is based on the blockade of Protein Synthesis on Ribosomes. The high efficacy of these remarkable medicinal agents is explained by the fact that they inhibit protein synthesis by bacterial 70-S ribosomes without affecting eukaryotic cell ribosomes. In other cases, the selective toxicity of antibiotics is due to the significantly higher permeability of bacterial membranes compared to those of animal cells.

The list of antibiotics acting at the ribosomal level is quite extensive [115, 116]. It includes, in particular, compounds that have played an important role in elucidating The Mechanism of protein synthesis. Although the aminoglycoside antibiotic streptomycin (Supplement 12-A), neomycins, and kanamycin share a common structural group, they all bind to ribosomes differently. As a result of streptomycin's unique action, ribosomes begin to misread the code. In doing so, it is primarily the first Base of the codon that is misread. Thus, for example, if poly(U) is used as informational RNA, instead of ordinary polyphenylalanine, a product containing 40% isoleucine is formed.

By exposing a bacterial population to antibiotics, mutants capable of growing in the presence of the corresponding antibiotic can be selected. In this manner, E. coli mutants resistant to streptomycin have been obtained, among others (though it should be noted that their frequency of appearance was very low: approximately 10^-12). It was established that the altered gene (rpsL or strA) is located on the genetic map in the region corresponding to 72 min¹⁾. It was subsequently shown that streptomycin binds to ribosomal protein S12, and rpsL is the gene for this protein. Among streptomycin-resistant bacteria, mutants can be selected that have become dependent on this antibiotic and are incapable of growing in its absence. It has been demonstrated that such streptomycin dependence arises from alterations in ribosomal protein S4. These experiments clearly show that a single point mutation altering just a single amino acid is sufficient to substantially change a living Organism's sensitivity to a specific toxin or even render the organism dependent on that toxin.

As follows from the analysis of resistant mutants, the antibiotic spectinomycin binds to protein S5. THE POSITION OF the spcA gene has been mapped to the 64th minute, suggesting that the ribosomal protein operon is located in this region of the E. coli chromosome. Kasugamycin inhibits the binding of fMet-tRNA (initiation). In this case, resistant mutants emerge in which the 16S rRNA is modified rather than the protein subunit; in mutant strains, the 16S rRNA is methylated to a lesser extent than in wild-type strains.

Briefly, several Other Antibiotics can be mentioned. Tetracyclines (Fig. 12-10) inhibit the binding of aminoacyl-tRNA to the A-site of the 30S ribosomal subunit. Lincomycin, sparsomycin, and chloramphenicol (Fig. 14-25) inhibit peptide bond formation and act on the 50S subunit. Chloramphenicol also induces the accumulation of ppGpp (Section B,2,k). The action of Polypeptide antibiotics—thiostrepton, bryamycin, and siomycin—is associated with their effects on factors G and Tu. Erythromycin (Fig. 12-10), along with other macrolide antibiotics, cycloheximide (Fig. 12-10), and fusidic acid (Chapter 12, Section I, 4), blocks translocation. Furthermore, fusidic acid inhibits the accumulation of ppGpp. Puromycin (Fig. 15-18) binds to the 50S subunit and causes premature termination of polypeptide chain synthesis. A glance at its chemical structure reveals how this effect is achieved: puromycin structurally mimics the aminoacylated 3'-end of a tRNA molecule down to the finest details, with the sole exception that it lacks an aminoacyl group. Once the growing polypeptide chain is transferred to it, further elongation becomes impossible.

¹⁾ The genes responsible for the Synthesis of the ribosomal protein S7 and the elongation factors EF-G and EF-Tu are located in this same region. It is believed that all of them are PARTS OF THE same transcriptional unit [116a].

FIG. 15-18. Structures of some inhibitory nucleoside analogs found in Nucleic Acids (Suhadolnik R. J., Nucleoside Antibiotics, Wiley-Interscience, New York, 1970, and R. Meyers, V. A. Malathi, R. P. Cox, S. Siler, JBC 248, 5909–5913, 1973).

i. Analysis of Nucleotide Sequences in mRNA

Only recently has it become possible to determine The nucleotide sequences of mRNA fragments and directly study the genetic information encoded within them. A convenient source of mRNA molecules

carrying unique genetic information is provided by RNA-containing Bacteriophages [117]. The genetic information of these Viruses is carried by RNA molecules consisting of only 3,500–4,500 NUCLEOTIDES and containing merely three genes (Supplement 4-G). The RNAs of phages f2, R17, MS2, and the more distant phage Qβ have been studied in considerable detail. The Amino acid sequences of several proteins encoded by these RNA molecules have been completely established, along with the full nucleotide sequence of one viral RNA molecule [118].

A convenient method for identifying gene start sites was proposed by Steitz [119]. THE PRINCIPLE OF this method is as follows. Isolated E. coli ribosomes are incubated under conditions that allow them to bind to mRNA, forming an initiation complex (Fig. 15-15). In the absence of additional Amino Acids and tRNA molecules, the initiation complex remains stable. When this complex is treated with pancreatic or T1 RNase, most of the RNA is hydrolyzed, except for a specific segment protected by the ribosome to which it is bound. It was found that the ribosome-protected RNA fragment of phage Qβ has the following sequence:

Note that the protected region includes the initiation codon AUG, and that the sequences of the subsequent codons precisely match the known N-terminal Amino Acid Sequence of the viral coat protein. Another interesting feature of this sequence is that the two regions indicated by braces with asterisks can base-pair with each other. As a result, the initiation codon can form a loop (hairpin structure). Although such hairpins are not always present in mRNA initiation regions, they occur quite frequently.

Moving slightly further upstream (to the left) in the nucleotide sequence of phage Qβ, one encounters a tetranucleotide group that can bind to 16S rRNA in the same manner as shown in Fig. 15-14 for the initiation region of the A protein in phage R17. Similar ribosome-protected initiation sequences have been discovered in the molecules of many viral RNAs, as well as in certain specific mRNAs [101, 102].

The complete sequence of all 3,569 nucleotides has been determined for phage MS2 RNA [118]. Parts of this sequence are shown in Fig. 15-19. The 5'-end (the central part of the structure depicted in the upper left corner) still carries the triphosphate group of the initiating GTP. Following a series of hairpins is the ribosome-protected region [119a], which begins with the initiation codon GUG. This provides direct evidence that GUG, like AUG, acts as a biologically important initiation codon. The nucleotide sequence immediately following the initiation codon accurately encodes almost the entire established amino acid sequence of the viral protein. The termination codon UAG is boxed in the figure. It is followed by a short intergenic region containing one arm of a hairpin, at the tip of which lies the initiation codon AUG for the next gene. Further along is a nucleotide sequence that precisely corresponds to the experimentally determined N-terminal amino acid sequence of the coat protein [120]. Another interesting feature of this sequence is the presence of a UGA termination codon (boxed) located immediately after THE START OF the gene encoding the coat protein (position 1390). This termination signal is out of frame with the AUG initiation codon and therefore cannot serve as the termination point for the coat protein gene. At the same time, it is in frame with the UAG codon, which is the termination codon of the A protein. In the presence of various host cell amber suppressor genes (Sections D,5 and D,6), the polypeptide chain of the A protein is extended until its synthesis terminates at this UGA signal.

FIG. 15-19. Selected regions of the nucleotide sequence of bacteriophage MS2 RNA and a schematic representation of their secondary structure. The initiation and termination codons of each of the three genes—encoding the A protein, coat protein, and replicase—as well as the second stop signal (which is in frame with the A protein gene but not the coat protein gene) are boxed. The entire coat protein gene is shown, whereas the nucleotide sequence is presented for less than one-third of its length [118, 119a, 120].

The coat protein gene, containing only 390 nucleotides, is shown in its entirety. Its proposed secondary structure resembles a flower [120]. The gene terminates with a double stop signal, UAAUAG. This is followed by a 36-nucleotide intergenic sequence, after which a very long replicase gene begins with an AUG codon. This gene ends at nucleotide 3,395, followed by a 374-nucleotide untranslated region at the 3'-end.

Another interesting example of ribosome-mediated protection of nucleic acids against Enzymatic hydrolysis comes from experiments with the single-stranded DNA of bacteriophage $\phi$X174 [121]. In this case, ribosomes protected a nucleotide sequence that included the ATG initiation codon. This codon and the seven subsequent codons corresponded to the known N-terminal amino acid sequence of the phage G gene-encoded spike protein.

Supplement 15-G

Replication of RNA-Containing Bacteriophages

Small icosahedral RNA-containing bacteriophages are of considerable interest because they contain a minimal number of genes, an attribute that allows a detailed investigation of their replication mechanisms^a,b. The complete sequence of the 3,569 nucleotides in phage MS2 RNA has now been fully deciphered (Fig. 15-19). The three genes of this phage encode protein A (phage maturation protein), the coat protein, and a replicase subunit. A mature viral particle contains a single molecule of protein A, which is essential for proper RNA encapsulation and for binding the phage to host cell pili. The RNA molecule is surrounded by a capsid composed of approximately 180 protein molecules. The replicase is required for the duplication of RNA molecules. Phage Qβ has a somewhat more complex structure; its RNA is 4.5 kb in length and, in addition to the maturation protein (designated protein A₂), contains several molecules of a fourth protein, A₁. Protein A₁ is unusual in that its N-terminal sequence of 130 amino acids is identical to that of the coat protein. Synthesis of the coat protein terminates at a UGA stop codon. However, even in wild-type E. coli cells, a small amount of UGA-specific suppressor tRNA allows translation to continue for approximately another 270 residues until a double stop signal, UGAUAA, is reached, which triggers chain termination. As a result, a large amount of coat protein and a small amount of protein A₂ are produced.

The replicase of phage Qβ has been studied in quite some detail. The formation of a fully active replicase complex requires three bacterial proteins in addition to the phage genome-encoded subunit: the ribosomal protein S1 and the elongation factors EF-Tu and EF-Ts. All three of these proteins normally participate in mRNA Translation, but the phage co-opts their RNA-binding abilities for an entirely different purpose.

Replication of a single-stranded phage must proceed in two stages. First, a complementary minus strand is synthesized on the plus strand contained within the phage particle. The initiation of this stage requires an additional bacterial protein, specifically host factor HF^b, and GTP. The resulting minus strands do not remain attached to the plus strands; they appear to be released from the replicase in single-stranded form and fold into highly ordered molecules containing numerous hairpins (similar to the plus strands of phage MS2 RNA shown in Fig. 15-19). Subsequently, the minus strands are copied (without The Need for factor HF) to produce A large number of new plus strands, which are then packaged into nascent phage particles.

Phage Qβ replicase is capable of synthesizing chains *in vitro* that are fully complementary to both plus and minus viral RNA molecules. However, the system is highly specific for viral RNA and cannot copy any other polynucleotides. Initiation of replication may require specific sequences at the 3'-end. In vitro replication is prone to errors, such as premature chain termination and base-mispairing. This leads to the generation of mutant RNA forms, making it possible to obtain RNA molecules significantly smaller than the viral RNA that are still readily replicated by the phage Qβ replicase system. The nucleotide sequence of one such fragment, comprising merely 114 nucleotides, has been determined^c.

^a Weissmann C., FEBS Letters, 40, S10–S78 (1974).

^b Senear A. W., Steitz J. A., JBC, 251, 1902–1912 (1976).

^b Mills D. R., Kramer F. R., Dobkin C., Nishihara T., Spiegelman S., PNAS, 72, 4252—4256 (1975).

k. Other Functions of ribosomes

Ribosomes not only synthesize proteins but also actively participate in regulatory mechanisms affecting the entire cell. One such enigmatic phenomenon is known as the stringent response [122, 123]. When deprived of a required amino acid in the growth medium, many amino acid-requiring (auxotrophic) mutants of E. coli and other bacteria promptly respond by downregulating the synthesis of ribosomal RNA, ribosomal proteins, purine nucleoside triphosphates, Lipids, and other essential compounds. However, Mutations in the rel gene (from relaxed) cause rRNA synthesis to continue even in the absence of the required amino acid (effectively relaxing the stringent response). In addition to this already complex behavior, it was discovered that stringent strains (rel⁺) accumulate the guanosine polyphosphates ppGpp and pppGpp (initially referred to as MS or "magic spot" compounds), whereas rel^- strains do not. The concentration of ppGpp under these conditions reaches 1 mM. It is now well established that guanosine polyphosphates are synthesized on ribosomes via The transfer of a pyrophosphoryl group from ATP:

ATP + GDP (GTP) → ppGpp (pppGpp) + AMP. (15-8)

This reaction requires a specific stringent factor—a ribosomal protein consisting of a single polypeptide chain with a molecular weight of approximately 75,000 [124]. In this process, the ribosomes must be bound to mRNA and contain uncharged tRNAs selected by codons in their A-sites.

Mutations in the rel gene abolish these regulatory effects by blocking the synthesis of guanosine polyphosphates. Conversely, the binding of fusidic acid and tetracycline mimics these mutations, allowing ribosomal RNA Synthesis to continue in stringent strains under conditions of amino acid starvation. The exact mechanism by which ppGpp and pppGpp affect the synthesis of tRNA or rRNA remains to be fully elucidated; it is quite possible that this process is mediated by Transcriptional Regulation involving a complex enzyme system [125—127].

It was unexpectedly discovered that E. coli phosphatidylserine synthetase is tightly bound to ribosomes [128]. This enzyme, which catalyzes the incorporation of serine into Phospholipids (step g in Fig. 12-8), is responsible for the synthesis of the major membrane Lipid Components of E. coli. The localization of this crucial enzyme on ribosomes may be somehow related to the Coordinated regulation of protein and lipid synthesis.