From Nucleotides to Genetic Information: DNA and RNA

How a molecule stores information that can be copied

A plant repeatedly produces new leaves with related forms, and a growing yeast population produces more yeast cells. Cells do more than divide a collection of finished parts: they retain information that guides subsequent manufacture and copy that information during reproduction. DNA combines a comparatively persistent chemical backbone with a sequence that can vary. RNA has related chemistry but performs diverse roles in information transfer, recognition, catalysis and regulation. Nucleic acids become easier to understand when these tasks are separated into three questions: how the chain is connected, how its sequence is copied, and how cells turn a sequence into something that works.

A book uses a limited alphabet to express many different messages, and DNA similarly uses a small set of building blocks in many sequences. Molecules, however, do not consciously interpret text. Information here means that an arrangement influences the arrangement and behaviour of subsequent products through identifiable chemical processes. A DNA segment can specify an RNA sequence, which can influence an amino acid sequence through the translation machinery. Many RNAs are themselves functional products and are never translated into protein. Genetic information is therefore a traceable structural relationship, rather than an abstract instruction floating apart from chemistry or a claim that every DNA segment directly specifies a protein.

How was DNA identified as a carrier of heritable traits? Avery, MacLeod and McCarty's 1944 work connected a substance capable of changing bacterial characteristics with its chemical composition. The central strategy was to separate candidate material, determine whether the resulting change could be inherited, and test which molecular component accompanied the activity. This was more discriminating than asking which substance was most abundant in a cell. With only a few familiar bases, DNA had initially seemed less diverse than protein. The experiments pointed toward another source of complexity: the enormous number of possible arrangements along a long chain.

Hershey and Chase's 1952 bacteriophage experiments tracked genetic material from another direction. Separate phosphorus and sulfur isotope labels distinguished nucleic acid from protein, allowing distributions inside and outside infected cells to be compared. Isotopes are atoms of the same element with different masses; they can provide a tracking signal while largely retaining the element's chemistry. The interpretation concerned the main distributions and their relationship to subsequent reproduction, rather than assuming perfect separation of every labelled atom. Connecting molecular identity with destination became a recurring experimental strategy in studies of replication, transcription and protein synthesis.

Cellular organisms generally store their genomes in DNA, while some viruses use RNA. DNA is not confined to a nucleus: bacteria have no membrane-enclosed nucleus, and mitochondria and plant chloroplasts also possess DNA. Some mature cells lose their nuclei or much of their nucleic acid during differentiation. Molecular identity, cellular location and life-cycle stage should therefore be distinguished. The same underlying chemistry can operate within different organisational arrangements. This distinction also keeps a familiar drawing of a nucleus from becoming an unintended definition of where all genetic information must reside.

Sugars, phosphates and bases make a directional chain

The basic unit of a nucleic acid is a nucleotide, consisting of a sugar, a nitrogen-containing base and phosphate. Sugar plus base is a nucleoside; addition of phosphate makes a nucleotide. The term base names the nitrogen-containing ring compound and does not imply that the whole nucleic acid forms an alkaline solution. Its phosphate backbone carries substantial negative charge under common aqueous conditions. DNA's usual four bases are adenine, A; guanine, G; cytosine, C; and thymine, T. RNA commonly uses uracil, U, in place of T. A and G have two fused rings and are purines; C, T and U are pyrimidines with a single six-membered principal ring.

DNA contains deoxyribose, whereas RNA contains ribose. Their central difference is at the sugar's 2′ position: RNA has a hydroxyl group where DNA normally has hydrogen. The prime mark distinguishes sugar numbering from numbering within the base. The 1′ position connects to the base, while 3′ and 5′ positions participate in backbone connections. The sugar ring is not a flat pentagon; it adopts puckered conformations, and the relative orientation of sugar and base can change. These geometric variations influence chain shape. A flat structural drawing records connectivity, rather than fixing the molecule into one planar pose in water.

Common DNA and RNA backbones use 3′,5′ phosphodiester linkages. The 3′ oxygen of one sugar and the 5′ oxygen of another are connected through phosphate. The two ester connections link sugars along the backbone, with bases projecting from it. One end is called the 5′ end and the other the 3′ end. These names identify backbone direction and do not require every real molecule to have identical terminal chemical groups. Sequences are conventionally written from 5′ to 3′. Thus, the direction in which a string of letters is recorded is itself part of the structural information.

Imagine a row of directional connectors: joining them end to end does not make the two ends interchangeable. Cellular nucleic acid polymerases normally add a new nucleotide to the growing chain's 3′ hydroxyl, extending that chain in the 5′-to-3′ direction. The template is read in the opposite direction. This refers to chemical orientation, not whether a molecule points left or right in a picture. Rotating the drawing does not exchange the molecule's 5′ and 3′ identities. Establishing this idea early makes complementary sequences, primers and translation direction much easier to follow.

RNA's extra 2′ hydroxyl is chemically consequential. It participates in hydrogen bonding and folding and, under suitable acid–base conditions, can attack a neighbouring phosphate and promote backbone cleavage. DNA lacks this group and consequently has different stability toward that reaction. Special 2′,5′ linkages, circular RNAs and branched intermediates also occur in biology. The familiar linear 3′,5′ backbone describes the principal arrangement without exhausting all possible structures. For a biomolecular material, these small differences affect persistence, enzyme recognition and assembly, connecting atomic-scale chemistry to how the material behaves over time.

Why a double helix pairs and still bends

Two DNA chains do not simply wind together like arbitrary pieces of rope. In a common duplex, the backbones run in opposite directions, bases face inward, and sugars and phosphates are more exposed to water. A usually forms a Watson–Crick pair with T, while G pairs with C. Combining one larger and one smaller ring system supports a fairly consistent width, and suitable hydrogen-bond donors and acceptors align across each pair. One sequence therefore constrains the other. An upper strand written 5′-AGTC-3′ pairs with 3′-TCAG-5′ below it. Writing that lower strand in the usual 5′-to-3′ convention gives 5′-GACT-3′.

This arrangement explains a classical composition relationship. Across an entire duplex built from standard pairs, each A is accompanied by a T and each G by a C. The counts need not match within one isolated strand, and single-stranded DNA or unusual pairing requires separate consideration. Matching composition also does not establish matching sequence. Identical numbers of A, T, G and C can be arranged into many different chains, just as the same letters can form different words. Composition supplied useful constraints on structure; the order of the bases carries more detailed information.

Franklin and Gosling's 1953 DNA-fibre work supplied important diffraction evidence about helicity and hydration. In the same year, Watson and Crick proposed the double-helix model, and Wilkins and colleagues published related structural evidence. X-ray diffraction does not photograph one DNA molecule as a greatly enlarged object. Interference between scattered waves records repeated structure, which is interpreted together with chemical constraints. Features of fibre patterns helped constrain helical dimensions, while composition and plausible atomic geometry constrained pairing. The discovery is particularly instructive as an example of measurements and models limiting one another's possibilities.

Common hydrated B-DNA is about 2 nanometres across, with an average axial rise of roughly 0.34 nanometres per base pair and typically about 10.5 base pairs per turn in solution. A nanometre is one billionth of a metre, much smaller than the separation an ordinary optical microscope resolves directly. Local twist and bending vary with sequence, environment and bound proteins; A- and Z-form structures further demonstrate that nucleic acids do not have one immutable helical geometry. Average rise estimates a long chain's contour length, the distance measured along its axis. That quantity is different from the separation between its two ends in solution.

The two backbones do not divide the duplex circumference into equal halves. Their attachment positions to the bases leave two surface channels of different widths, the major and minor grooves. Base edges expose hydrogen-bond donors, acceptors and other groups within these grooves, allowing some proteins to recognise local sequence features while the duplex remains paired. Groove shape also changes with sequence and environment. A- and B-form DNA are normally right-handed helices, while Z-DNA formed under suitable conditions is left-handed. Handedness describes rotation while advancing along the helix axis; it does not exchange the chemical identities of the 5′ and 3′ ends. Sequence, backbone direction and overall helical shape are therefore three distinct structural descriptions.

Duplex stability combines several effects. Pairing provides directional selectivity; stacking between neighbouring bases, interactions with water and ionic screening of the negatively charged backbone also contribute. GC-rich regions often melt at higher temperatures under otherwise comparable conditions, but counting three hydrogen bonds instead of two does not predict every sequence. Neighbouring base combinations, length, salt and mismatch position all matter. DNA can also bend while retaining many paired bases because backbone angles and rotations alter local direction. Treating it as a perfectly straight glass rod would miss the coiling, packaging and mechanical response of a long polymer in a cell.

What replication preserves: reasoning with heavy nitrogen

The double helix suggests a plausible copying mechanism: separate the old strands and build a complementary new strand beside each. A plausible model still requires a discriminating test. The question is where old material appears in the products. A conservative model retains the complete old duplex and makes a wholly new one. Semiconservative replication places one old strand in each daughter duplex. A dispersive model distributes pieces of old strands through newly made chains. All can increase total DNA, yet they predict different distributions of parental atoms. Measuring total abundance alone cannot distinguish these possibilities.

Meselson and Stahl's 1958 experiment labelled existing DNA with a heavier nitrogen isotope and then allowed growth with a lighter nitrogen source. Newly synthesised strands were consequently lighter. After sufficient centrifugation in a caesium chloride density gradient, DNA accumulated where the surrounding solution matched its buoyant density. The separation was about density, not a simple rule that a longer molecule falls farther. Heavy–heavy, heavy–light and light–light duplexes occupied distinguishable positions. An otherwise invisible question about strand inheritance thus became a question about observable bands.

After one round, the semiconservative model predicts that every duplex contains an old heavy strand and a new light strand, producing an intermediate-density band. The conservative alternative predicts separate heavy and light bands. Some dispersive possibilities still need to be distinguished after the first observation. The second round supplies another test: semiconservative replication gives half heavy–light and half light–light molecules. In later rounds, the original heavy strands remain in a fixed number of duplexes while the total population continues to grow. The fraction at intermediate density therefore halves with each additional round.

Semiconservative describes the distribution of old strands; it does not imply that separated DNA performs the whole copying task unaided. Helicases help open duplexes, polymerases select and connect nucleotides, primases provide starting segments, and ligases seal appropriate backbone discontinuities. Because new chains normally extend only 5′ to 3′, synthesis on the two sides of a replication fork is organised differently. One side can follow fork movement relatively continuously, while the other is made in segments that are subsequently joined. The asymmetry follows from a shared chemical direction rule, rather than from different base-pairing rules on the two templates.

Copying also requires selection and correction. Suitable pairing favours a geometry compatible with reaction, proofreading by some polymerases can remove mismatched termini, and cells possess subsequent repair processes. Rare base tautomers, chemical damage and other events can nevertheless generate errors. Genetic stability is therefore a product of several molecular safeguards, rather than absolute correctness at every hydrogen-bond encounter. A retained sequence change can have different consequences depending on its location, cellular processing and later selection. A mutation is not automatically equivalent to complete loss of a protein's activity.

RNA carries messages and forms working structures

RNA often participates in cells as a single strand, but a single strand need not be an unstructured cord. Complementary regions within one chain can come together to form stems, loops, bulges and more complex three-dimensional folds. RNA also uses common noncanonical pairs, including G–U wobble pairing. Local stability depends on sequence, ions, temperature and the surrounding fold, rather than a universal minimum number of paired bases. Double-stranded RNA occurs in biological processes and some viral genomes. The statement that DNA is always double-stranded and RNA always single-stranded is therefore only a very rough initial picture.

Messenger RNA, or mRNA, delivers translatable information to ribosomes. Researchers once needed to distinguish two possibilities: whether each protein required a new specialised ribosome, or whether existing machinery could read different temporary messages. Brenner, Jacob and Meselson's 1961 experiments traced relationships between old ribosomes and newly synthesised RNA, supporting the delivery of new information to existing ribosomes. Changing the production programme therefore need not require building an entirely new machine. This picture helps explain how regulation of transcription and RNA lifetime can alter protein production on relatively short timescales.

Many eukaryotic mRNAs acquire a 5′ cap, undergo splicing and receive a polyadenosine tail at the 3′ end during maturation. The cap has a distinctive 5′–5′ triphosphate connection, unlike the ordinary backbone linkage. These features participate in protection, transport and translation control. Poly(A) length changes, and some mRNAs do not use this kind of tail. Introns are commonly removed during splicing; regions retained in the mature RNA are exons. Exons can include untranslated regions. Equating exon with amino-acid-coding sequence would therefore confuse two different levels: RNA processing and protein translation.

Transfer RNA, or tRNA, acts as an adaptor between nucleotide and amino acid sequences. One region recognises an mRNA codon through an anticodon, while the other end normally carries an amino acid through its mature CCA terminus. Its secondary-structure diagram is often drawn as a cloverleaf, whereas the actual three-dimensional fold is commonly L-shaped. The cloverleaf records which regions pair; the L shape records how those regions assemble in space. These are complementary descriptions, not rival pictures. Modified nucleotides in tRNA also influence recognition and stability, giving this short polymer a chemical vocabulary beyond the four ordinary letters.

A typical cloverleaf can be followed along one continuous tRNA chain. Regions near the 5′ and 3′ ends pair to form the acceptor stem, from which the 3′ CCA terminus projects. Three other stem–loop regions are commonly called the D arm, anticodon arm and T arm, with a variable region between arms. D refers to the common modified residue dihydrouridine; the T-arm name is associated with ribothymidine, pseudouridine and other characteristic residues. The anticodon loop exposes the three bases used for decoding, while other regions support folding and recognition by synthetases and ribosomes. Contacts between D and T regions bring distant parts of the secondary-structure drawing together to organise the L-shaped elbow. Some tRNAs, including organellar variants, lack particular canonical features, so these parts describe a useful common architecture.

Ribosomal RNA, or rRNA, contributes the principal structural and catalytic core of the ribosome. The 1982 self-splicing RNA work of Kruger, Cech and colleagues helped establish that catalytic capability was not confined to proteins. Subsequent ribosome structures showed that RNA organises the centre responsible for peptide-bond formation. RNA therefore combines two capabilities: its sequence supplies pairing information, and its fold supplies functional geometry. For biomolecular materials, this is a particularly useful relationship: a sequence biases assembly, and the assembled structure performs a task.

Measuring amount, length and pairing

A nucleic acid solution often appears clear and colourless, so molecular changes must be converted into an instrumental signal. Electronic systems in the bases absorb ultraviolet light, giving substantial absorption near 260 nanometres. Comparing incident and transmitted light defines absorbance. In a dilute solution, more absorbing molecules or a longer light path produces greater attenuation, allowing concentration to be estimated with an appropriate absorption coefficient. Different sequences and single- versus double-stranded states do not have identical coefficients. Matching the coefficient to the sample is part of the conversion from an optical measurement to a molecular amount.

Absorbance answers only part of the analytical question. Proteins, free nucleotides and other substances can influence a spectrum, and a long DNA molecule broken into many fragments can still absorb strongly. Gel electrophoresis is commonly used to inspect length distributions. Negatively charged nucleic acids move in an electric field, while gel pores impede molecules according to size and shape. For linear double-stranded DNA within a suitable resolving range, smaller fragments generally move faster, and unknown bands can be compared with a length standard in the same gel. Circular, supercoiled or partly single-stranded molecules also respond through their shape, so every band position cannot be interpreted as length alone.

A gel resembles a mesh with an adjustable pore size. A mesh that is too coarse poorly separates very small fragments, while one that is too fine hinders larger molecules. Field strength, temperature and buffer conditions affect movement and heating. RNA readily forms structure, so measuring its length often calls for bringing molecules into comparable unfolded states. Preventing nuclease-mediated degradation is a separate requirement. Distinguishing unfolding from chain cleavage explains why a treatment can help reveal length without replacing the need for suitable sample handling and preservation.

Centrifugation also distinguishes nucleic acids and ribosomes through their motion. Rate-zonal separation compares how far particles travel within a specified time, depending on mass, shape, hydration and solvent density. The isopycnic separation in the replication experiment instead approaches equilibrium and compares final buoyant-density positions. Dividing sedimentation speed by the applied centrifugal acceleration gives a sedimentation coefficient. Its unit, the svedberg or S, equals ten to the power minus thirteen seconds. It is not a mass unit, and subunit S values are not simply additive. A typical bacterial 30S small ribosomal subunit contains 16S rRNA, while its 50S large subunit contains 23S and 5S rRNA; together they form a 70S ribosome. Typical eukaryotic cytoplasmic 40S and 60S subunits form an 80S ribosome, with 18S rRNA in the small subunit and 28S, 5.8S and 5S in the large one. These numerical names describe sedimentation behaviour, with molecular composition matched to the biological system.

Heating a duplex disrupts pairing and stacking, allowing paired regions to separate: melting or denaturation. If the backbone remains intact, separation and reassociation can occur under suitable conditions. Cleaving backbone bonds is degradation, a different transformation. Unfolded nucleic acids often absorb more strongly at 260 nanometres, giving the hyperchromic effect. Plotting absorbance against temperature follows changes in pairing and permits a transition midpoint to be defined between suitable baselines. Curves depend on sequence, length, salt and concentration; RNA can also contain highly stable local structures. A single temperature range is not a universal property of all nucleic acids.

Complementary strands finding one another and pairing again is renaturation or annealing. Pairing between complementary regions from different molecules is often called hybridisation. Its rate depends on concentration, existing structure and the search for a correct alignment as well as temperature. Rapid cooling may trap some conformations, while short suitable partners can still pair rapidly; sudden cooling does not universally prevent reassociation. Hybridisation supplies a sequence-selective connection for materials. Components carrying complementary segments can assemble through pairing, while temperature and solution conditions tune the stability of that connection. Understanding the measurement therefore leads directly to a useful design principle.

From genes to codons: changing molecular language

Transcription transfers DNA sequence information into RNA; translation converts a nucleotide sequence into an amino acid sequence. The first uses complementary copying between related nucleic acid alphabets, while the second requires a correspondence between triplets and amino acids. For a particular transcription unit, polymerase uses one DNA strand as its template. Different genes along a chromosome can occupy opposite orientations, so an entire chromosome does not have one universally transcribed strand and one permanently inactive counterpart. Transcription also produces functional rRNAs, tRNAs and regulatory RNAs, rather than only messages destined to become proteins.

A protein-coding gene also needs appropriate initiation, processing and termination. A promoter is a DNA region that helps the transcription machinery choose an initiation region and direction. The transcription start is where RNA synthesis begins; the start codon is where the ribosome begins reading amino-acid information. A transcribed but untranslated 5′ region commonly separates these positions. A 3′ untranslated region can follow the stop codon, and transcription termination uses different signals and machinery from translation termination. Distant regulatory sequences can influence transcription without all being called promoters. Common bacterial operons place several coding regions in one mRNA, each with its own translation-start and stop arrangements. Many eukaryotic mRNAs predominantly encode one polypeptide, while alternative splicing produces different RNAs from the same gene. These arrangements explain why a gene cannot be defined by one fixed length or one universal DNA-to-product count.

The central dogma concerns detailed residue-by-residue sequence information. DNA can be copied into DNA or transcribed into RNA; reverse transcription uses RNA as a template for DNA, and some processes copy RNA into RNA. Crick's 1970 explanation emphasised whether detailed sequence information can transfer from protein back to nucleic acid or into another protein. It was not an attempt to compress every cellular influence into one unidirectional arrow. A protein can regulate transcription or alter cell state without using its amino acid sequence as a template to write a nucleotide sequence.

A codon consists of three consecutive nucleotides. Why consider at least this length? Four letters offer four choices at each position. One position gives four combinations, two give sixteen, and three give sixty-four, enough to assign the twenty common amino acids and control signals. Multiplication here counts combinations of independently chosen positions. It does not prove that biology uses triplets; genetic and biochemical experiments established the reading mechanism. In the standard genetic code, several codons can specify the same amino acid, a property called degeneracy. The three ordinary stop codons indicate termination rather than three additional amino acids.

Nirenberg and Matthaei's 1961 use of artificial poly(U) illustrated how a template could be connected to its product. Providing a simple RNA to a cell-free protein-synthesis system allowed the investigators to ask which amino acid was incorporated into newly made material. Poly(U) promoted a product rich in phenylalanine, establishing a measurable link between nucleotide information and amino acid selection. The complete code emerged from subsequent experiments with defined templates and related approaches. Counting possible triplets and experimentally deciphering their meanings are distinct achievements.

Reading also requires a starting position. Grouping the same letters into threes from position one or position two produces different codons: different reading frames. Inserting or deleting one nucleotide commonly changes downstream grouping. Replacing one nucleotide may change an amino acid, preserve it or create a stop signal. Because the code is degenerate, an amino acid sequence generally does not specify one unique DNA sequence in reverse. Genetic-code variants and special recoding events also occur. Rules should be matched to the organism and cellular compartment being discussed, which explains why equal DNA lengths do not imply equal protein products.

How the ribosome joins amino acids carried by adaptors

ATP and GTP are adenosine triphosphate and guanosine triphosphate. Ribose and A or G form the nucleoside, with three phosphates attached on its 5′ side. TP names those three phosphates; one fewer gives a diphosphate and two fewer a monophosphate. During amino acid loading, ATP participates in formation of an aminoacyl-adenylate intermediate and release of pyrophosphate, two phosphate groups still linked together. Transfer of the aminoacyl group to tRNA then releases adenosine monophosphate, AMP. DNA polymerases instead use four deoxynucleoside triphosphates, collectively called dNTPs: d identifies the sugar lacking a 2′ hydroxyl and N represents different bases. These molecules supply new-chain building blocks and participate in the phosphate-transfer chemistry that joins them.

Translation first requires loading the appropriate amino acid onto a tRNA. Aminoacyl-tRNA synthetases recognise features of both partners and use ATP-dependent chemistry to produce charged tRNA. The amino acid's carboxyl group becomes ester-linked to a sugar hydroxyl on the terminal adenosine. This makes the adaptor analogy concrete: an anticodon identifies an mRNA codon while the other end carries the amino acid to be incorporated. If loading is wrong, reading the anticodon correctly does not automatically repair the cargo. Selection and editing during aminoacylation are therefore important parts of translation accuracy in their own right.

ATP and GTP provide chemistry that couples different steps. Energy should be discussed through the overall free-energy change from reactants to products, rather than as energy automatically released by cutting one high-energy bond. Breaking an isolated bond requires energy. Hydrolysis simultaneously forms other bonds and changes hydration, charge and molecular distributions; the overall reaction can release usable free energy under the relevant conditions. Amino acid loading stores part of a reaction's chemical potential in the activated connection, while GTP hydrolysis couples factor selection, conformational changes and directionality. This is more informative than imagining one ATP directly driving every peptide bond at the instant it forms.

The ribosome has large and small subunits and binds mRNA together with several tRNAs. A, P and E name three major tRNA positions. The A site accepts incoming aminoacyl-tRNA, the P site carries the growing peptide, and the E site participates in departure of unloaded tRNA. After a new amino acid is positioned, the existing chain is transferred to it, forming a new peptide bond. Relative movement of ribosome and mRNA by one codon then prepares the growing chain for another cycle. A, P and E are working positions, not three permanently distinct tRNA species. The tRNAs move through states as the ribosome undergoes coordinated conformational changes.

Nissen and colleagues' 2000 structural work clarified the molecular environment of peptide-bond formation. The catalytic centre is organised principally by rRNA, rather than by a separate protein peptidyl-transferase enzyme hidden inside the large subunit. Ribosomal proteins remain essential contributors to assembly, stability and coordinated function. This structural finding connects RNA folding directly to protein manufacture. It also shows why treating RNA as nothing more than a temporary written message overlooks one of its central molecular roles.

Initiation establishes the reading frame before the repeating elongation cycle begins. Many bacterial mRNAs contain an upstream Shine–Dalgarno sequence that can pair with a region of small-subunit 16S rRNA, helping position the start codon; other bacterial initiation arrangements also exist. A specialised initiator tRNA places the first amino acid in the P site. Common bacterial translation uses formylmethionine, whereas eukaryotic cytoplasmic translation normally uses methionine. Occupying P first leaves A available for the next aminoacyl-tRNA. During elongation, a GTP-bound factor helps deliver and check new tRNA, while another factor participates in translocation. These stages govern recognition and movement separately, so successful pairing does not by itself complete the entire cycle.

Which end of a chain grows was itself an experimental question. In his 1961 investigation of haemoglobin-chain assembly, Dintzis supplied traceable amino acid to a protein-producing system and compared labels in different fragments of chains completed during short intervals. If growth proceeds from the amino toward the carboxyl end, many partly completed chains preferentially incorporate the newly available label into their later-made carboxyl-terminal regions. Longer labelling permits broader distribution through complete chains. Relating time to position makes N-to-C growth an experimentally testable inference. A related decoding example arises when a changed tRNA anticodon recognises a stop codon and sometimes competes with release to continue translation. Such suppressor-tRNA activity depends on the inserted amino acid and protein position for any functional recovery; it does not remove the original DNA mutation.

mRNA is normally read 5′ to 3′, while a polypeptide grows from its amino terminus toward its carboxyl terminus. At a suitable stop codon, release factors help recognise the signal and promote hydrolysis of the peptide–tRNA connection. The product is released and the machinery recycled. Bacteria and eukaryotes differ in initiation signals, factors and compartmental organisation. Eukaryotic transcription and cytoplasmic translation are generally separated by the nuclear envelope, whereas bacterial transcription and translation can overlap in time. Several ribosomes can work simultaneously on one mRNA, each making a chain, to form a polysome. Folding, localisation and possible chemical modification then turn new chains into mature functional products.

Chemically making a specified nucleic acid sequence

Specified nucleic acid sequences provide probes for complementary targets, short connections between material components and model fragments for structural studies. Chemical synthesis faces a different control problem from cellular polymerisation: a nucleoside has several potentially reactive groups, so how can one selected unit be joined at only the intended position each time? Protecting groups temporarily mask positions, while a solid support retains the growing chain as excess reagents are washed away. The tRNA-gene synthesis reported by Khorana's team in 1970 demonstrated how chemically prepared short segments could be aligned by complementarity and joined enzymatically into a longer DNA structure.

Early phosphodiester and phosphotriester methods both directed a nucleotide phosphate to react with another nucleoside hydroxyl. A phosphodiester connection retains an ionisable phosphate position. A phosphotriester strategy protects that position with a removable group, changing the intermediate charge and reaction behaviour. Diester and triester count the phosphate connections through oxygen to organic groups. Phosphoramidite chemistry instead performs coupling with a different, readily activated phosphorus reagent and then oxidises the resulting phosphite to a phosphate. Solid phase and solution phase describe whether the growing chain is attached to a support; they describe process organisation. Diester, triester and phosphoramidite describe connection chemistry, so the two classifications answer different questions. In each route, temporary protection is ultimately removed as required for the desired product.

Common solid-phase phosphoramidite synthesis attaches the first nucleoside through its 3′ end and extends toward its 5′ side. This conventional arrangement differs from cellular polymerase growth. Each cycle exposes a reactive hydroxyl and couples an activated incoming monomer. Ends that failed to extend can be capped, reducing their participation in later cycles and the formation of deletion sequences. The new phosphite linkage is oxidised to a more stable phosphate form. Finally, protecting groups are removed, the chain is released from its support and products are separated. Understanding each stage's job explains sequence control more clearly than memorising a string of reagent abbreviations.

Beaucage and Caruthers' 1981 phosphoramidite work helped develop this controlled cycle. The aim was not to make every group maximally reactive at once, but to activate the required monomer at the required time while maintaining useful handling properties at other stages. Even highly efficient individual steps accumulate incomplete or erroneous products when repeated many times. If the probability of a correct extension is taken as constant and other losses are considered separately, repeated multiplication estimates the full-length target fraction. This explains why a favourable single-step efficiency does not by itself describe the final long-chain product.

Improving step efficiency, separating different lengths and verifying sequence address different problems. Separation can remove some short chains, while incorrect sequences of the intended length may require more specific analysis. Assembly through overlapping complementary regions extends the accessible structure but also introduces questions about mismatches, omissions and joining order. No single maximum synthetic length applies permanently to every platform. Length, modification chemistry, sequence-dependent structure, route and purity requirements jointly determine the attainable product. Comparing methods through these actual conditions is more informative than carrying an old instrument limit into a universal rule.

Complementarity also provides a construction rule for materials. Components carrying selected complementary segments can assemble into particular combinations. Some regions become duplexes, while others remain flexible connectors or exposed recognition sites. Sequence establishes local pairing preferences; concentration, salt and temperature influence whether the desired assembled state is reached. Obtaining a fragment of the expected length does not establish every feature of its three-dimensional organisation, so structural measurements retain an independent role. Nucleic acids are therefore both objects of genetic study and polymers whose interaction patterns can be arranged through sequence.

Amplification increases signal; verification identifies it

Sometimes the goal is to make more copies of an existing DNA region rather than build a sequence afresh from monomers. The polymerase chain reaction, PCR, uses two short primers to define a target region through repeated strand separation, primer pairing and polymerase extension. Each primer supplies both sequence recognition and an extendable 3′ end. On opposite templates, their growing ends point into the intended region. During the first few cycles, products bounded by both primer positions become established and subsequently dominate. This geometry explains why choosing any two short chains that can bind somewhere on DNA does not guarantee the intended product.

Thermostable polymerases made repeated temperature cycling more convenient. Saiki and colleagues' 1988 study demonstrated their usefulness in PCR and helped advance its operation and applications. Thermal stability concerns continued activity after heating, whereas fidelity concerns nucleotide selection and proofreading; these are different properties. Taq polymerase naturally lacks the common 3′-to-5′ exonucleolytic proofreading activity. Other polymerases have different capabilities. Length, sequence, required accuracy and subsequent use therefore guide system choice, rather than an early performance range of one enzyme becoming a permanent limit of PCR itself.

If every cycle doubles the amplifiable population, repeated cycles produce exponential growth. An exponent is shorthand for repeated multiplication: the same growth factor is applied once per cycle. In practice, the added population per cycle can be smaller, and later reagent depletion, product accumulation and changes in enzyme activity produce a plateau. The following model holds early-cycle efficiency constant so that cumulative effects can be understood. It describes a defined growth regime rather than predicting unlimited physical accumulation. The accompanying graph shows how seemingly modest differences per cycle become substantial over repeated cycles.

Large amplification and correct amplification must be assessed separately. Background molecules with sufficiently matching primer sites can also increase. Template damage, copying errors and small amounts of material carried into the sample influence later results. A no-template control probes reagent background, a known positive control tests whether the system operates, and length, specific recognition or sequence measurements establish product identity. Primer specificity is not guaranteed by counting how rare a short string would be in an ideal random four-letter sequence. Real genomes contain repeats, composition biases and related sequences, and pairing can tolerate some mismatches.

Different primer strategies address different information constraints. If only a protein segment is known, genetic-code degeneracy allows several corresponding DNA sequences. Degenerate primers mix candidate oligonucleotides to cover those possibilities, while reducing the concentration of each exact sequence within the mixture. Nested PCR first uses outer primers and then another pair inside the first product. A target must satisfy two successive positional selections, which commonly improves selectivity, while transfer between rounds requires control of carried-over contamination. Hot-start systems delay polymerase activity until an appropriate heating stage, reducing unwanted extension during cooler preparation. These approaches alter candidate-sequence coverage, a second positional selection and reaction-start timing respectively. Product identity is still established through evidence such as length and sequence.

From duplex structure to PCR, the same relationships recur. Complementary pairing supplies selectivity, backbone chemistry supplies direction and stability, enzymes and environment govern the process, and measurements turn molecular change into interpretable evidence. RNA folding and ribosome function further show that an information-bearing chain can itself become a working structure. Understanding nucleic acids therefore means being able to explain a concrete observation: which molecule is present, how it formed, what determines its abundance, and whether the sequence and function required for its purpose have been retained.