A DNA molecule is built from three chemical parts repeated over and over: a sugar (deoxyribose), a phosphate group, and a nitrogenous base. Two long chains of alternating sugars and phosphates twist around each other in the famous double helix, while pairs of bases bridge the gap between them like rungs on a spiraling ladder. That basic architecture stores all of the genetic information a cell needs to function, and the specific arrangement of its parts explains why DNA can be copied, read, and repaired with remarkable precision.
The Sugar-Phosphate Backbone
The outer rails of the DNA ladder are made of two alternating components: a five-carbon sugar called deoxyribose and a phosphate group. Each phosphate links the third carbon of one sugar to the fifth carbon of the next, forming a long, repeating chain. This sugar-phosphate backbone gives DNA its structural rigidity and determines the direction of each strand. Biochemists refer to the two ends of a strand as the 5ʹ end (where a free phosphate sits) and the 3ʹ end (where a free hydroxyl group sits). In a double helix, the two strands run in opposite directions, which matters when cellular machinery reads or copies the molecule.
The “deoxy” in deoxyribose means the sugar is missing one oxygen atom compared with ribose, the sugar found in RNA. That seemingly minor difference has real consequences for stability. Studies that separated the effects of the missing oxygen from other chemical differences between DNA and RNA showed that the presence or absence of the 2ʹ-hydroxyl group on the sugar can either stabilize or destabilize a nucleic acid helix depending on the structure involved, while the methyl group unique to DNA’s thymine base is consistently stabilizing.1PubMed. Origins of the large differences in stability of DNA and RNA helices: C-5 methyl and 2′-hydroxyl effects The net result is that double-stranded DNA sits in a sweet spot: chemically stable enough to store information for decades, yet flexible enough that enzymes can pry the strands apart when needed.
Phosphate groups also give DNA its negative electrical charge. Every phosphate carries a negative charge at the pH found inside cells, which means the entire backbone is studded with negative charges. Positively charged ions and proteins in the cell neutralize some of that charge, helping the molecule fold and pack tightly. Without those interactions, the two negatively charged strands would repel each other and the helix would fall apart.
The Four Nitrogenous Bases
Attached to each sugar is one of four nitrogenous bases, and these bases are where the genetic information actually lives. Two of them, adenine (A) and guanine (G), are larger, double-ringed structures called purines. The other two, cytosine (C) and thymine (T), are smaller, single-ringed structures called pyrimidines. The specific sequence of these four letters along a strand spells out genes, regulatory signals, and all the other instructions encoded in a genome.
A pairs with T, and G pairs with C. This rule is not arbitrary. Early spectroscopic work showed that guanine derivatives and cytidine derivatives form stronger complexes with each other than with themselves, and the same selective pairing was observed between adenine and uracil (RNA’s stand-in for thymine). When researchers tested every possible combination of the purines and pyrimidines, only the G-C and A-U (or A-T) pairs formed stable mixed dimers.2Science. Hydrogen Bonding Specificity of Nucleic Acid Purines and Pyrimidines in Solution The geometry is part of the explanation: a purine paired with a pyrimidine produces a rung of roughly uniform width, keeping the helix regular. Two purines together would be too wide; two pyrimidines, too narrow.
The bonds holding each base pair together are hydrogen bonds, which are individually weak compared with the covalent bonds in the backbone. Computational studies that evaluated the strength of hydrogen bonds across 15 different nucleic acid base pairs found that nearly all of them qualify as medium-strength hydrogen bonds.3PubMed. Detection and evaluation of hydrogen bond strength in nucleic acid base pairs G-C pairs have three hydrogen bonds, while A-T pairs have only two, which is why DNA regions rich in G-C pairs are harder to pull apart than A-T-rich regions. But hydrogen bonding alone does not account for the stability of the whole helix.
What Actually Holds the Double Helix Together
It is natural to assume that the hydrogen bonds between base pairs are the glue that keeps the double helix intact, but the reality is more nuanced. A detailed thermodynamic analysis of the forces stabilizing double-stranded DNA found that base stacking, the interaction between bases stacked on top of one another along the helix, is the main stabilizing factor across a wide range of temperatures and salt concentrations. In that same analysis, A-T pairing was actually found to be mildly destabilizing, and G-C pairing contributed almost no net stabilization on its own.4PubMed Central. Base-stacking and base-pairing contributions into thermal stability of the DNA double helix That counterintuitive result does not mean base pairing is unimportant; it is essential for the specificity of the code, ensuring that A always faces T and G always faces C. But the physical force that keeps two strands wound around each other comes primarily from the way flat base surfaces stack like coins in a roll.
Stacking interactions are driven largely by hydrophobic effects: the flat, relatively water-repelling surfaces of the bases prefer to nestle against each other rather than face the surrounding water.5PubMed Central. Hydrophobic catalysis and a potential biological role of DNA unstacking induced by environment effects Van der Waals forces between neighboring bases also contribute. Taken together, stacking and hydrogen bonding, supplemented by van der Waals contacts, account for the bulk of double helix stability.6PubMed. On the conformational stability of oligonucleotide duplexes and tRNA molecules
Water plays a part too. When a DNA helix melts (the two strands separate), water molecules are released from the structure. Measurements using osmotic stressors have shown that roughly four water molecules per base pair are released during duplex melting.7PubMed Central. Effects of hydration, ion release, and excluded volume on the melting of triplex and duplex DNA Those bound waters form an ordered shell around the helix, and their release is part of the energetic equation that governs when and how easily DNA strands come apart.
The Classic Double Helix and Its Variants
When most people picture DNA, they picture B-DNA: a right-handed helix with about ten base pairs per full turn and a diameter of roughly two nanometers. B-DNA is the predominant form under normal cellular conditions, but it is not the only shape the molecule can adopt. Crystallographic studies have identified at least three families of DNA helix: A, B, and Z, each with distinct geometry and chain-folding rules.8PubMed. The anatomy of A-, B-, and Z-DNA
A-DNA is a wider, more compact right-handed helix that forms when DNA is dehydrated or in certain protein-DNA complexes. Z-DNA is particularly unusual: it is left-handed, meaning it spirals in the opposite direction from B-DNA, and tends to form in stretches of alternating purine-pyrimidine sequence under specific ionic conditions. While Z-DNA was once considered a laboratory curiosity, research has since linked it to roles in gene regulation and immune signaling. The conditions that trigger transitions between B-form and these alternative conformations, including changes in salt concentration, hydration, and mechanical stress from processes like transcription, have been studied extensively using plasmid DNA models.9PubMed Central. Topological Behavior of Plasmid DNA
Non-Canonical Structures
Beyond the three helical families, DNA can fold into shapes that look nothing like a simple double helix. Two of the most studied are G-quadruplexes and i-motifs. G-quadruplexes form in guanine-rich sequences, where four guanines arrange themselves into a flat quartet stabilized by unconventional hydrogen bonds, and multiple quartets stack to produce a four-stranded structure. I-motifs arise from cytosine-rich sequences and are also four-stranded, held together by pairs of cytosines sharing a proton.
Both structures have been observed inside living human cells, not just in test tubes. G-quadruplexes have been visualized using specific antibodies, and an estimated half of human genes contain potential G-quadruplex-forming sequences near their regulatory regions, suggesting widespread roles in controlling gene expression. I-motifs were similarly confirmed in human cell nuclei, and proteins that specifically bind i-motifs and regulate gene activity have been discovered.10Nucleic Acids Research. A DNA G-quadruplex/i-motif hybrid I-motifs are now thought to participate in genome stability, transcription, replication, and telomere and centromere function.11Trends in Genetics. Unveiling the biology of DNA i-motifs Researchers are also exploring non-canonical DNA structures as potential drug targets, because small molecules that stabilize or disrupt a quadruplex near an oncogene might dial gene expression up or down.12PubMed. Non-canonical DNA structures: Comparative quantum mechanical study
NMR spectroscopy has become a key tool for studying the dynamics of these alternative folds, capturing how G-quadruplexes, i-motifs, triplexes, and Z-DNA structures shift and flex, and how they interact with the proteins that recognize them.13PubMed Central. Dynamics Studies of DNA with Non-canonical Structure Using NMR Spectroscopy
Chemical Tags That Change What DNA Does Without Changing Its Sequence
The parts of DNA described so far are the ones encoded by the genome itself, but living cells routinely tack on small chemical groups that alter how genes behave. The best known is the addition of a methyl group to cytosine, creating 5-methylcytosine. This modification typically suppresses gene activity and plays essential roles in development, cell identity, and silencing of transposable elements. More recently, additional modifications have been discovered, including 5-hydroxymethylcytosine, which is generated when a family of enzymes oxidizes the methyl group on 5-methylcytosine.14PubMed Central. Epigenetics of Modified DNA Bases: 5-Methylcytosine and Beyond
These modifications do not change the sequence of A, T, G, and C letters, but they do change the physical properties of the molecule. Molecular dynamics simulations have shown that methylation and hydroxymethylation alter the local twist and bending angles of the helix at the modified sites, shifting how DNA interacts with proteins that read its surface.15PubMed Central. Understanding the structural and dynamic consequences of DNA epigenetic modifications: computational insights into cytosine methylation and hydroxymethylation You can think of these chemical tags as annotations written in the margins of a textbook: they do not change the printed words, but they tell the reader which passages to pay attention to and which to skip.
DNA Outside the Nucleus
Most discussions of DNA parts focus on the long, linear chromosomes inside the nucleus, but cells carry DNA in other compartments too. Human mitochondria, the organelles that generate most of a cell’s energy, each contain a small circular DNA molecule of 16,569 base pairs encoding 37 genes: two ribosomal RNAs, 22 transfer RNAs, and 13 proteins involved in energy production.16PubMed. The mitochondrial genome: structure, transcription, translation and replication The chemical parts of mitochondrial DNA are identical to nuclear DNA (the same sugars, phosphates, and bases), but the packaging is different. Mitochondrial DNA lacks the elaborate histone-wrapping system that nuclear chromosomes use, and it is inherited almost exclusively from the mother.
Plants add another layer: their chloroplasts also carry circular DNA molecules, encoding proteins essential for photosynthesis. Bacteria, too, typically store their genome as a single circular DNA molecule rather than the multiple linear chromosomes found in animal and plant nuclei. The chemical vocabulary is the same everywhere, but the organizational grammar differs across life’s domains.
When Parts Break
Understanding the parts of DNA also means understanding what happens when those parts sustain damage. One of the most common injuries comes from ultraviolet light, which can cause two adjacent pyrimidine bases (usually thymines) on the same strand to fuse together into a dimer. These pyrimidine dimers distort the helix and, if left unrepaired, can lead to mutations and eventually cancer.17PubMed Central. DNA excision repair: where do all the dimers go? The two main types of UV-induced lesions are cyclobutane pyrimidine dimers and pyrimidine (6-4) pyrimidone photoproducts, and accumulation of either type can be lethal to cells.18PubMed. Insight in DNA Repair of UV-induced Pyrimidine Dimers by Chromatographic Methods
Cells cope through a set of repair pathways. Nucleotide excision repair, for example, recognizes the distortion in the helix, clips out a short stretch of the damaged strand, and uses the intact opposite strand as a template to fill the gap. The complementary base-pairing rules make this possible: because A always pairs with T and G with C, the undamaged strand carries a perfect copy of the lost information. This is one of the most elegant practical consequences of DNA’s double-stranded structure.
UV damage is just one type. Oxidative damage, spontaneous loss of bases, and errors during replication all threaten the molecule’s integrity. Cells maintain an arsenal of repair enzymes tailored to different kinds of lesions, all exploiting the same structural feature: the redundancy built into having two complementary strands.
Synthetic Alternatives and XNA
Researchers have spent decades asking which parts of DNA are truly essential for storing and transmitting genetic information, and one way to test that is to swap parts out. Synthetic genetic polymers, collectively known as XNAs, replace the natural deoxyribose sugar with alternative sugars while keeping the same or similar bases and phosphate linkages. Over 50 years of chemical design have produced a vast array of these sugar-modified nucleic acids.19Nucleic Acids Research. The XNA alphabet
What makes XNAs interesting is that some of them can be replicated and even evolved in the laboratory, much like natural DNA. Advances in polymerase engineering have enabled the synthesis and replication of several XNA types, producing molecules with improved chemical and biological stability compared with natural nucleic acids.20PubMed Central. Towards XNA nanotechnology: new materials from synthetic genetic polymers That improved stability is precisely why XNAs are attractive for therapeutic applications: natural DNA and RNA are quickly chewed up by enzymes in the body, but XNAs can resist degradation, making them candidates for next-generation drug molecules.21PubMed Central. Modified nucleic acids: replication, evolution, and next-generation therapeutics
XNA research also has implications for understanding life’s origins. A central puzzle in origin-of-life chemistry is why nature settled on deoxyribose and ribose rather than some other sugar. Laboratory synthesis has shown that the building blocks of DNA purines and RNA pyrimidines can form together under plausible prebiotic conditions, suggesting that a mixed system of deoxyribonucleosides and ribonucleosides may have coexisted before life as we know it emerged.22PubMed Central. Selective prebiotic formation of RNA pyrimidine and DNA purine nucleosides The fact that artificial sugar backbones can also support heredity and evolution hints that natural DNA’s particular chemical makeup may be one workable solution among many, rather than the only possible one.