# Base Pair in DNA: Rules, Types and Examples A base pair in DNA is a specific partnership between one purine and one pyrimidine held together by hydrogen bonds across the two strands of the double helix. In the canonical Watson-Crick scheme, adenine (A) pairs with thymine (T) through two hydrogen bonds, and guanine (G) pairs with cytosine (C) through three hydrogen bonds. That single rule is the reason a DNA strand can be copied, why the two strands of a duplex are called complementary, and why the total amount of A in a double-stranded genome equals the total amount of T. It also explains why a G-C rich duplex resists melting better than an A-T rich one, and why the geometry of the helix is uniform even though the four bases are chemically different. ## The Four Bases and Their Chemical Classes DNA contains four nitrogenous bases attached to a deoxyribose sugar and a phosphate group. Each base is a flat, aromatic ring system that stacks in the interior of the helix. The purines are adenine and guanine. They have a fused two-ring structure, a pyrimidine ring fused to an imidazole ring, giving nine atoms in the bicyclic framework. The pyrimidines are thymine and cytosine. They have a single six-membered ring, giving six atoms in the ring system. This size difference matters. A purine is roughly twice the width of a pyrimidine. If two purines paired, the pair would be too wide for the helix. If two pyrimidines paired, the pair would be too narrow. A purine plus a pyrimidine gives a pair of consistent width, which is why the sugar-phosphate backbones of the two strands stay at a constant distance apart. The bases are not covalently bonded to each other across the helix. They are held by hydrogen bonds, which are individually weak but collectively strong. The covalent bonds in DNA run along each strand, connecting sugar to phosphate to sugar. The cross-strand connection is noncovalent. ## The Base Pairing Rules The rules that govern [base pairing](/knowledge/molecular-biology/base-pairing) are often called Watson-Crick complementarity. They have three parts. **Rule 1: A pairs with T.** Adenine and thymine form two hydrogen bonds. The donor and acceptor groups line up on the Watson-Crick face of each base, the edge that faces the helix interior. **Rule 2: G pairs with C.** Guanine and cytosine form three hydrogen bonds. The extra bond makes G-C pairs more stable than A-T pairs. **Rule 3: A purine always pairs with a pyrimidine.** A pairs with T, and G pairs with C. A does not pair with G, and C does not pair with T, in the standard duplex. This keeps the helix diameter constant. The hydrogen bonding pattern is directional. On the Watson-Crick face, each base presents a specific arrangement of hydrogen bond donors (amino groups) and acceptors (carbonyl oxygens and ring nitrogens). Adenine presents a donor and an acceptor. Thymine presents the complementary acceptor and donor. Guanine presents two donors and one acceptor, and cytosine presents the matching two acceptors and one donor. The fit is geometric as well as chemical. ### Why Hydrogen Bonds Are Not the Whole Story Hydrogen bonds alone do not hold the duplex together. Base stacking contributes a large share of the stability. The aromatic rings of adjacent base pairs stack on top of one another in the helix interior, and these stacking interactions, driven largely by van der Waals contact and hydrophobic effects, add substantial free energy. The distinction matters in practice. A duplex with the correct hydrogen bonding pattern but poor stacking would still be unstable. Conversely, mismatched bases can sometimes be accommodated if the stacking context is favorable, which is one reason DNA polymerases need proofreading activity to achieve high fidelity. Computational work on base pairs confirms the role of both forces. In one study, theoretical calculations using the self-consistent charge density functional tight-binding method verified the role of hydrogen bonding and van der Waals type interactions in a host-guest interface involving adenine and thymine [1]. The two contributions operate together. ## Chargaff's Rule Chargaff's rule states that in double-stranded DNA, the amount of adenine equals the amount of thymine, and the amount of guanine equals the amount of cytosine. This is a direct consequence of A-T and G-C pairing. Every A on one strand faces a T on the other. Every G faces a C. The rule applies to double-stranded DNA. It does not apply to single-stranded DNA, to most RNA molecules, or to individual strands of a duplex. A single strand can have any base composition at all. Only when the complementary strand is present does the symmetry appear. A related observation, sometimes called Chargaff's second parity rule, states that within a single strand, %A is approximately equal to %T and %G is approximately equal to %C. This is an empirical tendency rather than a strict rule, and it breaks down in some sequence contexts. Studies of microsatellites in *Caenorhabditis elegans* have shown that certain tandem repeat segments violate the second parity rule and display asymmetric folding energies between the top and bottom strands [2]. ### Chargaff's Rule Table The table below shows a worked example of a double-stranded [DNA molecule](/blog/guides/dna-molecule) with a base composition that satisfies Chargaff's rule. The percentages sum to 100. | Base | Percentage | Complementary base | Percentage of complement | |---|------|----------|-------------| | A | 30% | T | 30% | | T | 30% | A | 30% | | G | 20% | C | 20% | | C | 20% | G | 20% | | **Total** | **100%** | | **100%** | Read the table this way. If A is 30%, T must also be 30%, because every A is paired with a T. The remaining 40% is split equally between G and C, so each is 20%. The sum of all four percentages is always 100. The sum of A and T plus the sum of G and C is also always 100. If a problem gives you %A = 22%, then %T = 22%, and %G + %C = 56%, so %G = 28% and %C = 28%. If a problem gives you %G = 35%, then %C = 35%, %A + %T = 30%, and %A = %T = 15%. ## Base Pairing Table: Partner and Hydrogen Bond Count | Base | Class | Partner | Hydrogen bonds | Pair geometry | |---|----|-----|--------|--------| | Adenine (A) | Purine | Thymine (T) | 2 | Watson-Crick | | Thymine (T) | Pyrimidine | Adenine (A) | 2 | Watson-Crick | | Guanine (G) | Purine | Cytosine (C) | 3 | Watson-Crick | | Cytosine (C) | Pyrimidine | Guanine (G) | 3 | Watson-Crick | The hydrogen bond count is the reason G-C content correlates with duplex stability. A sequence that is 70% G-C has more three-bond pairs per unit length than a sequence that is 70% A-T, and it will generally have a higher melting temperature. This principle is used when designing PCR primers, where a balanced G-C content helps ensure the primer anneals at a predictable temperature. ## Worked Examples: Writing the Complementary Strand Complementary strand writing is the most common base-pairing exercise in [molecular biology](/blog/careers/molecular-biology). The rules are simple. Read the template strand from 5' to 3'. Write the complement antiparallel, meaning the new strand runs 3' to 5' relative to the template. ### Example 1: A Short Tetranucleotide Template strand: 5'-ATGC-3' The complement is 3'-TACG-5'. Read position by position. A pairs with T. T pairs with A. G pairs with C. C pairs with G. The complement strand is written antiparallel, so its 3' end aligns with the 5' end of the template. ### Example 2: A Longer Sequence Template strand: 5'-GGATC-3' The complement is 3'-CCTAG-5'. Position by position: G pairs with C, G pairs with C, A pairs with T, T pairs with A, C pairs with G. ### Example 3: Writing the Complement in the Same Orientation If you want to write the complement in the 5' to 3' orientation, reverse the order after writing the antiparallel complement. For the template 5'-ATGC-3', the antiparallel complement is 3'-TACG-5'. Written 5' to 3', it is 5'-GCAT-3'. This convention matters when you design primers or search a database. Most software tools report sequences in the 5' to 3' orientation, so a reverse complement is often what you actually need. ### Example 4: Calculating Base Composition from a Sequence Take the sequence 5'-AAGCT-3'. Count the bases: A appears twice, G once, C once, T once. The complement is 3'-TTCGA-5', which also contains two A, one G, one C, and one T. The double-stranded molecule therefore has %A = 30%, %T = 30%, %G = 20%, and %C = 20%, which satisfies Chargaff's rule. ## DNA Versus RNA Base Pairing The base pairing rules are nearly identical in RNA, with one substitution. RNA uses uracil (U) in place of thymine. Uracil is the unmethylated form of the pyrimidine ring. Thymine is 5-methyluracil. In RNA, A pairs with U through two hydrogen bonds, and G pairs with C through three hydrogen bonds. The same purine-pyrimidine geometry applies. | Feature | DNA | RNA | |-----|---|---| | Pyrimidine partner for A | Thymine (T) | Uracil (U) | | Hydrogen bonds in A pair | 2 | 2 | | Sugar | Deoxyribose | Ribose | | Typical strandedness | Double-stranded | Often single-stranded | | Chargaff's rule applies | Yes, to duplex DNA | Only to double-stranded regions | The reason DNA uses thymine rather than uracil is a matter of repair. Cytosine can spontaneously deaminate to uracil. If DNA used uracil as a normal base, the repair machinery could not distinguish a legitimate uracil from a deaminated cytosine. By using thymine, DNA marks uracil as foreign, and uracil DNA glycosylase can remove it. ## Non-Canonical Base Pairs The Watson-Crick pairs are the dominant geometry in B-form DNA, but they are not the only possible arrangement. Alternative pairings appear in specific structural contexts and are biologically important. **Hoogsteen base pairs.** In a Hoogsteen pair, the purine flips to the *syn* conformation around the glycosidic bond, and hydrogen bonding occurs on the Hoogsteen face rather than the Watson-Crick face. Hoogsteen pairs form transiently in duplex DNA and are stabilized in complexes with proteins and small molecules. Solid-state NMR spectroscopy using dynamic nuclear polarization can distinguish Watson-Crick from Hoogsteen geometry based on characteristic chemical shifts and internuclear dipolar couplings, and it has been used to observe Hoogsteen pairs in a [nucleosome core particle](/knowledge/molecular-biology/nucleosome-core-particle) [3]. **Wobble pairs.** A wobble pair forms when a purine and pyrimidine pair through an offset hydrogen bonding pattern, such as G pairing with U. Wobble pairs are common in RNA, particularly at the wobble position of the genetic code, and they can also occur as mismatches in DNA. **Water-mediated pairs.** Some non-canonical pairs are bridged by a water molecule. Crystal structures of DNA duplexes containing non-CpG methylated cytosine have revealed a water-mediated *cis* Watson-Crick/Hoogsteen geometry that coexists with the standard Watson-Crick geometry in the same base pair, with partial occupancies of about 0.1 and 0.9 respectively [4]. The methylated mC:G pair can adopt both conformations, and the alternative geometry is specific to non-CpG methylated sites in B-form DNA. **Metal-mediated pairs.** Artificial base pairs can be held together by metal ions rather than hydrogen bonds. Silver(I) ions mediate base pairs between artificial nucleobases and canonical bases, and studies of 7-deazaadenine/thymine sequences show a strong preference for canonical X-Ag(I)-T pairing over homobase arrangements [5]. These systems are used in DNA nanotechnology. **Expanded genetic alphabets.** Artificially Expanded Genetic Information Systems add unnatural nucleotide pairs to the natural G-C and A-T/U pairs. Cryo-EM structures of *E. coli* [RNA polymerase](/blog/guides/rna-polymerase) elongation complexes containing these unnatural pairs show that the enzyme recognizes them in a Watson-Crick geometry, indicating that the same mechanistic principles apply to natural and artificial pairs [6]. ## How Base Pairing Is Observed in Practice Several standard methods detect and quantify base pairing. **Melting temperature analysis.** Heating a DNA duplex causes the hydrogen bonds to break and the strands to separate. The temperature at which half the duplex is melted is the melting temperature, or Tm. Because G-C pairs have three hydrogen bonds, higher G-C content raises the Tm. This is measured by tracking absorbance at 260 nm, which increases as the bases unstack. **X-ray crystallography.** Crystal structures of DNA duplexes reveal the exact geometry of each base pair, including hydrogen bond distances and angles. This method has been used to show that a single base pair can adopt multiple conformations simultaneously, as in the mC:G case described above [4]. **NMR spectroscopy.** Solution and solid-state NMR can detect base pair geometry in solution and in large complexes. Dynamic nuclear polarization enhances sensitivity enough to study base pairs in assemblies approaching 200 kDa [3]. **Nanopore sensing.** Nanopore measurements can identify hydrogen bonding base pairing at single-base-pair resolution. One study used nanopore unzipping of thymine-melamine-thymine triplets to detect non-canonical thymine-melamine hydrogen bonding in DNA abasic sites, achieving single-base-pair resolution and revealing enhanced intramolecular stability compared to canonical pairs [7]. **Computational methods.** Density functional theory and related methods calculate the energetics of base pairing and stacking. Computational analysis of 5-fluorouracil adsorption on Watson-Crick and Hoogsteen base pairs showed that the drug prefers Watson-Crick geometry and that the adsorption energies are higher than those of pristine nucleobase pairs, indicating that base pair cleavage is less likely during drug binding [8]. ## Why Base Pairing Matters Base pairing is the physical basis of heredity. During replication, each strand serves as a template, and the polymerase reads the template base and inserts the complementary base. During transcription, RNA polymerase reads the template strand and synthesizes an RNA copy with U in place of T. During translation, codon-anticodon pairing at the ribosome uses the same rules, with wobble at the third position. The specificity of base pairing also underlies molecular biology techniques. PCR primers are designed to be complementary to a target sequence. Hybridization assays rely on complementary probe binding. CRISPR guide RNAs base pair with target DNA. Antisense oligonucleotides base pair with mRNA to block translation. Errors in base pairing cause mutations. The Löwdin model proposes that tautomerization of a G-C pair to G*-C* can lead to mispairing during replication, producing G*-T and A-C* intermediates that resolve to A-T after two rounds of replication. Kinetic calculations based on this model give a probability of transition mutation from G-C to A-T of about 1.31 × 10⁻⁸, consistent with previously reported mutation rates [9]. Other lesions, such as oxidized adenine, can pair with G or C and produce A to C or A to G mutations [10]. ## Common Mistakes and Limitations **Confusing hydrogen bonds with covalent bonds.** Base pairs are not covalently linked. The two strands of DNA are held together by hydrogen bonds and base stacking, not by covalent cross-links. Covalent bonds run along the backbone of each strand. **Thinking purines pair with purines.** Purines pair with pyrimidines. A pairs with T, and G pairs with C. A does not pair with G in the standard duplex, and C does not pair with T. **Assuming Chargaff's rule applies to single strands.** The rule applies to double-stranded DNA. A single strand can have any composition. **Forgetting the antiparallel orientation.** When writing a complementary strand, the new strand runs in the opposite direction. The 5' end of the template aligns with the 3' end of the complement. **Overlooking base stacking.** Hydrogen bonds alone do not explain duplex stability. Stacking interactions contribute substantially. **Treating Watson-Crick geometry as the only possibility.** Hoogsteen, wobble, water-mediated, and metal-mediated pairs all exist and have biological roles. **Assuming G-C content fully predicts Tm.** G-C content is a major factor, but sequence context, salt concentration, and length also affect melting temperature. Individual sequences and experimental conditions require case-specific analysis, and any diagnostic or clinical application should be reviewed by a qualified professional. ## Quick Review 1. A pairs with T through two hydrogen bonds. G pairs with C through three hydrogen bonds. 2. A purine always pairs with a pyrimidine, keeping the helix diameter constant. 3. Chargaff's rule: %A = %T and %G = %C in double-stranded DNA, and all four percentages sum to 100. 4. The complement of 5'-ATGC-3' is 3'-TACG-5', written antiparallel. 5. RNA uses uracil instead of thymine, so A pairs with U through two hydrogen bonds. 6. Base stacking, not just hydrogen bonding, stabilizes the double helix. 7. Non-canonical pairs such as Hoogsteen and wobble pairs exist and have biological functions. ## Frequently Asked Questions ### How many hydrogen bonds hold an A-T base pair together? Two. Adenine and thymine form two hydrogen bonds across the Watson-Crick face. ### How many hydrogen bonds hold a G-C base pair together? Three. Guanine and cytosine form three hydrogen bonds, which makes G-C pairs more stable than A-T pairs. ### What is the complement of 5'-ATGC-3'? The complement is 3'-TACG-5'. Read antiparallel, A pairs with T, T pairs with A, G pairs with C, and C pairs with G. ### Does Chargaff's rule apply to RNA? No, not to single-stranded RNA. Chargaff's rule applies to double-stranded DNA, where %A equals %T and %G equals %C. ### Why does DNA use thymine instead of uracil? Thymine allows the repair machinery to recognize uracil as a deaminated cytosine rather than a normal base. If DNA used uracil normally, that distinction would be lost. ### Are base pairs covalent bonds? No. Base pairs are held by hydrogen bonds and base stacking. The covalent bonds in DNA run along the sugar-phosphate backbone of each strand. ## Related Articles - [Purine Always Pairs with Pyrimidine: Base Pairing Rules](/knowledge/molecular-biology/purine-always-pair-with-pyrimidine) - [Base Pair Substitution: Types, Mechanisms, and Effects](/knowledge/molecular-biology/base-pair-substitution) - [RNA Base Pairs: A-U and G-C Pairing Explained](/blog/guides/rna-base-pairs) - [Nucleotide Bases: Structure, Types, and Roles in DNA and RNA](/knowledge/molecular-biology/nucleotide-base) - [Dna Bases](/blog/guides/dna-bases) - [Base Pairing: Rules, Mechanisms, and Biological Significance](/knowledge/molecular-biology/base-pairing) - [Chargaff's Rule: Base Pairing Explained with Examples](/knowledge/molecular-biology/chargaff-s-rule-base-pairing-explained-with-examples) ## Sources 1. [DNA-Mimicking Metal-Organic Frameworks with Accessible Adenine Faces for Complementary Base Pairing.](https://pubmed.ncbi.nlm.nih.gov/35373199/) 2. [Microsatellites that violate Chargaff's second parity rule have base order-dependent asymmetries in the folding energies of complementary DNA strands and may not drive speciation.](https://pubmed.ncbi.nlm.nih.gov/18579157/) 3. [Probing Watson-Crick and Hoogsteen base pairing in duplex DNA using dynamic nuclear polarization solid-state NMR spectroscopy.](https://pubmed.ncbi.nlm.nih.gov/35857870/) 4. [Structural basis of water-mediated cis Watson-Crick/Hoogsteen base-pair formation in non-CpG methylation.](https://pubmed.ncbi.nlm.nih.gov/38989613/) 5. [Silver Binding Dichotomy for 7-Deazaadenine/Thymine: Preference for Watson-Crick Pairing over Homobase Interactions in DNA.](https://pubmed.ncbi.nlm.nih.gov/40638818/) 6. [A unified Watson-Crick geometry drives transcription of six-letter expanded DNA alphabets by E. coli RNA polymerase.](https://pubmed.ncbi.nlm.nih.gov/38086811/) 7. [Single-Molecule Nanopore Detection of Non-Canonical Thymine-Melamine Hydrogen Bonding Base Pair in DNA Abasic Site.](https://pubmed.ncbi.nlm.nih.gov/40891541/) 8. [Unveiling the Intermolecular Interactions between Drug 5-Fluorouracil and Watson-Crick/Hoogsteen Base Pairs: A Computational Analysis.](https://pubmed.ncbi.nlm.nih.gov/38882136/) 9. [Kinetic Study of Transition Mutations from G-C to A-T Base Pairs in Watson-Crick DNA Base Pairs: Double Proton Transfers.](https://pubmed.ncbi.nlm.nih.gov/34516113/) 10. [Rapid excision of oxidized adenine by human thymine DNA glycosylase.](https://pubmed.ncbi.nlm.nih.gov/36460098/)