15. Translation

Ribosome Structure and Function

The ribosome is composed of two asymmetric subunits, each containing both rRNA and proteins. The landmark crystal structures by Venki Ramakrishnan, Thomas Steitz, and Ada Yonath (Nobel Prize 2009) revealed that the ribosome is fundamentally a ribozyme: the peptidyl transferase center is composed entirely of rRNA (23S in prokaryotes, 28S in eukaryotes), with no protein atoms within 18 angstroms of the active site.

Three tRNA Binding Sites

The ribosome has three tRNA binding sites that span both subunits:

• A (aminoacyl) site: accepts incoming aminoacyl-tRNA; decoding occurs here
• P (peptidyl) site: holds tRNA carrying the growing peptide chain
• E (exit) site: deacylated tRNA exits here after translocation

Derivation 1: The Genetic Code — Degeneracy and Wobble

The genetic code was deciphered by Marshall Nirenberg, Har Gobind Khorana, and Robert Holley (Nobel Prize 1968). It is a triplet, non-overlapping, degenerate, nearly universal code mapping 64 codons to 20 amino acids and 3 stop signals.

Why Triplets?

The information-theoretic argument: with 4 nucleotide bases, the minimum codon length to encode 20 amino acids is:

Triplets are the minimum, but 64 codons for 20 amino acids (+ stops) means the code is degenerate: most amino acids are specified by more than one codon. This degeneracy is not random — it follows a pattern that minimizes the impact of mutations.

The Wobble Hypothesis

Francis Crick proposed the wobble hypothesis (1966) to explain how fewer tRNAs (~45 in mammals) can decode all 61 sense codons. Non-standard base pairing is allowed at the 3rd codon position (wobble position):

The wobble rules explain why degeneracy is concentrated at the 3rd codon position. The first two positions follow strict Watson-Crick rules, so mutations there are more likely to change the encoded amino acid.

Error Minimization

The genetic code is organized to minimize the phenotypic impact of point mutations. Conservative amino acid substitutions (similar physicochemical properties) are overrepresented:

This ~2-fold enrichment suggests the code was optimized by natural selection, though the degree of optimization is debated. Most single-nucleotide mutations either change the amino acid to a similar one or are silent (synonymous) — a built-in buffer against translational and replication errors.

Derivation 2: The Translation Mechanism

Initiation

Translation initiation differs fundamentally between prokaryotes and eukaryotes:

Elongation: The Three-Step Cycle

Each amino acid addition involves three steps, consuming 2 GTP:

Codon Usage Bias and Translation Efficiency

Although the genetic code is degenerate, synonymous codons are not used equally. Highly expressed genes preferentially use codons corresponding to the most abundant tRNAs, a phenomenon called codon usage bias.

The Codon Adaptation Index (CAI)

The CAI quantifies how well a gene's codon usage matches the optimal codons of the organism:

where $w_i$ is the relative adaptiveness of the $i$-th codon and $L$ is the gene length in codons. CAI ranges from 0 (poor adaptation) to 1 (optimal). Highly expressed genes (ribosomal proteins, glycolytic enzymes) have CAI > 0.8 in E. coli.

Practical Applications

Codon optimization is critical for recombinant protein expression: when expressing a human protein in E. coli (or vice versa), the gene sequence is redesigned to use the host's preferred codons. This can increase protein yields 10-100 fold. The mRNA vaccines for COVID-19 were codon-optimized for human codon usage and used modified nucleosides (N1-methylpseudouridine) to enhance translation efficiency and reduce innate immune recognition.

Rare codons also have functional roles: they can slow translation at specific positions, allowing co-translational protein folding of individual domains before the next domain begins to emerge from the ribosome. Deliberately placed rare codons act as "translational speed bumps" that improve folding efficiency for complex multidomain proteins.

Derivation 3: Energetics of Translation

Translation is the most energetically expensive biosynthetic process in the cell — consuming up to ~75% of a cell's ATP in rapidly growing bacteria.

Cost per Amino Acid

For each amino acid incorporated, the cell expends:

For a typical 300-residue protein, the translation cost is ~1200 ATP. Adding initiation factors, termination, and quality control increases this to ~1300-1400 ATP per protein. For perspective, a single glucose molecule yields only ~32 ATP — so synthesizing one protein requires the oxidation of ~40 glucose molecules.

Translation Speed and Fidelity

In E. coli, the ribosome incorporates ~20 amino acids/sec. The error rate (~$10^{-4}$ per codon) is achieved through a kinetic proofreading mechanism (Hopfield, 1974):

The first selection step is initial binding ($\sim 10^{-2}$ discrimination). GTP hydrolysis then provides an irreversible step that allows a second selection ($\sim 10^{-2}$), amplifying the fidelity at the cost of energy (the sacrificed GTP).

Aminoacyl-tRNA Synthetases: The Second Genetic Code

The aminoacyl-tRNA synthetases (aaRS) are responsible for correctly matching each amino acid to its cognate tRNA — a process Paul Schimmel called the "second genetic code" because errors here are propagated into the protein with no opportunity for correction.

The Aminoacylation Reaction

Each of the 20 aaRS catalyzes a two-step reaction:

The PP$_i$ is immediately hydrolyzed by pyrophosphatase ($\text{PP}_i \rightarrow 2\;\text{P}_i$), making the overall reaction irreversible (equivalent to 2 ATP).

Two Classes of aaRS

Editing (Proofreading) by aaRS

Some amino acids are structurally similar (e.g., isoleucine vs valine, threonine vs serine), and the initial discrimination by the active site is insufficient. Several aaRS have a separate editing domain that hydrolyzes mischarged AA-tRNA. Isoleucyl-tRNA synthetase (IleRS) is the classic example:

The editing site uses a "double sieve" mechanism: the synthetic site excludes amino acids larger than isoleucine (steric exclusion), while the editing site excludes amino acids as large as isoleucine but hydrolyzes the smaller valine-AMP. This achieves $\sim 1/40{,}000$ error rate, well below the ~1/3,000 rate from the synthetic site alone.

Co-Translational Protein Folding and Chaperones

Newly synthesized polypeptides do not fold in a vacuum. The cellular environment is extremely crowded (~300-400 mg/mL protein), increasing the risk of misfolding and aggregation. A sophisticated chaperone system assists folding:

• Trigger factor (bacteria) / NAC (eukaryotes): bind the ribosome exit tunnel and shield the nascent chain during co-translational folding
• Hsp70/DnaK system: ATP-dependent chaperone that binds hydrophobic stretches on unfolded proteins, preventing aggregation and allowing iterative folding attempts
• Hsp60/GroEL-GroES (chaperonin): a barrel-shaped cavity where proteins up to ~60 kDa fold in isolation from the crowded cytoplasm; uses ~130 ATP per folding cycle
• Hsp90: specialized chaperone for signaling proteins (kinases, steroid receptors, transcription factors); target of the anticancer drug geldanamycin

The Unfolded Protein Response (UPR)

When misfolded proteins accumulate in the endoplasmic reticulum, the unfolded protein response (UPR) is activated through three ER transmembrane sensors:

If the UPR fails to resolve ER stress, the cell activates apoptosis via CHOP (C/EBP homologous protein). Chronic ER stress and UPR activation are implicated in diabetes (beta-cell failure), neurodegeneration (Parkinson's, Alzheimer's), and liver disease. Protein misfolding diseases (amyloidoses) — including Alzheimer's ($\beta$-amyloid), Parkinson's ($\alpha$-synuclein), and prion diseases (PrP$^{\text{Sc}}$) — result from the failure of quality control to prevent toxic aggregation.

Derivation 4: Termination and Quality Control

Termination

When a stop codon (UAA, UAG, or UGA) enters the A site, no aminoacyl-tRNA matches. Instead, release factors recognize the stop codon and catalyze peptide release:

• Prokaryotes: RF1 (recognizes UAA/UAG), RF2 (recognizes UAA/UGA), RF3 (GTPase for RF recycling)
• Eukaryotes: eRF1 (recognizes all three stop codons), eRF3 (GTPase)

Ribosome-Associated Quality Control

Three quality control pathways handle aberrant mRNAs:

Polyribosomes and Translational Control

Multiple ribosomes can simultaneously translate a single mRNA, forming polyribosomes (polysomes). The spacing between ribosomes is ~80 nucleotides (~27 codons), meaning a typical 1 kb mRNA can accommodate ~12 ribosomes simultaneously, dramatically increasing the protein production rate.

mRNA Circularization

Eukaryotic mRNAs adopt a closed-loop conformation during active translation, mediated by the interaction between the 5' cap-binding complex (eIF4E-eIF4G) and the 3' poly(A)-binding protein (PABP):

This circularization enhances translation efficiency by allowing ribosomes that have just terminated to be "recycled" directly to the 5' cap for re-initiation, without dissociating into the cytoplasm. It also ensures that only intact mRNAs (with both a cap and poly(A) tail) are efficiently translated — providing quality control against truncated or decapped mRNAs.

Global Translational Control

Under stress conditions, cells globally reduce translation to conserve energy. The key mechanism involves phosphorylation of eIF2$\alpha$ at Ser-51 by four stress-responsive kinases:

• GCN2: activated by uncharged tRNAs (amino acid starvation)
• PERK: activated by ER stress (unfolded protein response)
• HRI: activated by heme deficiency (erythroid cells)
• PKR: activated by double-stranded RNA (viral infection)

Paradoxically, eIF2$\alpha$ phosphorylation increases translation of specific mRNAs with upstream open reading frames (uORFs) in their 5' UTR, including ATF4 (stress-responsive transcription factor) and CHOP (pro-apoptotic). This integrated stress response (ISR) allows the cell to reprogram gene expression while reducing overall protein synthesis — a sophisticated survival strategy.

mTORC1 and Cap-Dependent Translation

The mTORC1 kinase promotes cap-dependent translation by phosphorylating two key targets:

In the unphosphorylated state, 4E-BP1 sequesters eIF4E, blocking cap recognition. mTORC1-mediated phosphorylation releases eIF4E to join the eIF4F complex, enabling cap-dependent translation. Rapamycin (sirolimus) inhibits mTORC1, reducing translation of mRNAs with complex 5' UTRs (often encoding growth factors and oncoproteins) — hence its use as an immunosuppressant and anticancer agent.

Derivation 5: Post-Translational Modifications

The proteome's diversity vastly exceeds the genome's coding capacity because proteins are extensively modified after translation. Over 400 types of post-translational modifications (PTMs) have been identified, expanding the functional repertoire from 20 amino acids to effectively thousands of distinct chemical states.

Phosphorylation

The most common regulatory PTM, catalyzed by >500 protein kinases in the human genome (the "kinome"):

Phosphorylation adds 2 negative charges at physiological pH, causing conformational changes and creating docking sites for phospho-binding domains (SH2, PTB, 14-3-3). ~30% of human proteins are phosphorylated at any given time. Phosphatases reverse the modification, creating a dynamic on/off switch.

Glycosylation

Two major types: N-linked (to Asn in the sequon Asn-X-Ser/Thr; begins in ER with dolichol-linked oligosaccharide) and O-linked (to Ser/Thr; occurs in Golgi). Glycosylation affects protein folding, stability, and cell-cell recognition. About 50% of human proteins are glycosylated. Congenital disorders of glycosylation (CDGs) cause multisystem disease.

Ubiquitination

The attachment of ubiquitin (76 amino acids) to lysine residues marks proteins for proteasomal degradation. The ubiquitin-proteasome system (Aaron Ciechanover, Avram Hershko, Irwin Rose — Nobel Prize 2004):

Polyubiquitin chains linked via Lys-48 of ubiquitin signal proteasomal degradation. Other chain types (Lys-63, linear) serve non-degradative functions (NF-$\kappa$B signaling, DNA repair). The human genome encodes >600 E3 ubiquitin ligases, providing remarkable substrate specificity. The N-end rule pathway links protein half-life to the identity of the N-terminal amino acid.

Protein Sorting: Directing Proteins to Their Destinations

Newly synthesized proteins must be directed to the correct subcellular compartment. In eukaryotic cells, proteins can be targeted to the ER/secretory pathway, mitochondria, nucleus, peroxisomes, or remain in the cytoplasm. Targeting depends on signal sequences — short peptide motifs recognized by specific receptors:

ER-Associated Degradation (ERAD)

Misfolded proteins in the ER are retrotranslocated to the cytoplasm and degraded by the proteasome — a process called ERAD. The ER quality control system uses a lectin-based mechanism: the N-linked glycan on newly synthesized proteins is sequentially trimmed by glucosidases and mannosidases. If the protein fails to fold properly within the allotted time, mannose trimming marks it for ERAD:

ERAD is clinically relevant: the most common cystic fibrosis mutation ($\Delta$F508-CFTR) produces a protein that folds slowly but is functional. ERAD prematurely degrades it before it reaches the plasma membrane. Corrector drugs (lumacaftor, elexacaftor in Trikafta) stabilize the mutant CFTR, allowing it to escape ERAD and reach the cell surface — a triumph of molecular understanding informing therapy.

Applications in Medicine and Research

Historical Context

The central dogma (DNA $\rightarrow$ RNA $\rightarrow$ protein) was articulated by Francis Crick in 1958. Nirenberg's landmark experiment (1961) used poly-U mRNA in a cell-free system to show that UUU encodes phenylalanine — the first codon assignment. By 1966, all 64 codons were deciphered. The ribosome crystal structures (2000) revealed that peptide bond formation is catalyzed by RNA, supporting the "RNA world" hypothesis — that RNA preceded proteins as catalysts in early life.

The Signal Recognition Particle and Protein Targeting

Proteins destined for the secretory pathway, plasma membrane, or organelles contain signal sequences that direct them to the endoplasmic reticulum during translation. Gunter Blobel proposed the signal hypothesis (Nobel Prize 1999):

The signal recognition particle (SRP) is a ribonucleoprotein (6 proteins + 7SL RNA) that binds the hydrophobic signal peptide as it emerges from the ribosome exit tunnel, pausing translation until the ribosome docks at the ER membrane. The signal peptide is cleaved by signal peptidase in the ER lumen, and the protein is co-translationally folded by ER chaperones (BiP/GRP78, calnexin, calreticulin).

Protein Degradation: The N-End Rule

Alexander Varshavsky discovered the N-end rule (1986), which relates a protein's half-life to the identity of its N-terminal amino acid. In eukaryotes, destabilizing N-terminal residues (Arg, Lys, His, Phe, Trp, Tyr, Leu, Ile) are recognized by specific E3 ubiquitin ligases (N-recognins, UBR1/2), targeting the protein for proteasomal degradation:

The N-end rule has been expanded to include the Ac/N-end rule (N-terminally acetylated residues) and the Pro/N-end rule (proline-specific). These pathways regulate the half-lives of ~30% of cellular proteins and play critical roles in chromosome segregation, cardiovascular development, and neurodegeneration.

Selenocysteine: The 21st Amino Acid

Selenocysteine (Sec) is co-translationally incorporated at UGA codons (normally a stop codon) when a specific SECIS element (selenocysteine insertion sequence) is present in the 3' UTR of the mRNA. The process requires a specialized tRNA (tRNA$^{\text{Sec}}$), a dedicated elongation factor (eEFSec), and SECIS-binding protein 2 (SBP2). The human genome encodes 25 selenoproteins, including glutathione peroxidases, thioredoxin reductases, and iodothyronine deiodinases — all involved in redox homeostasis and thyroid hormone metabolism.

Python Simulations