Sanger Sequencing

Sanger sequencing (formally the dideoxynucleotide chain-termination method) is a first-generation enzymatic DNA sequencing technique that determines nucleotide order through the stochastic incorporation of chain-terminating 2′,3′-dideoxynucleoside triphosphates (ddNTPs) by DNA polymerase during in vitro template replication, followed by high-resolution electrophoretic separation of the resulting nested DNA fragments. Developed in 1977 by British biochemist Frederick Sanger, Steven Nicklen, and Alan R. Coulson at the Medical Research Council (MRC) Laboratory of Molecular Biology in Cambridge—earning Sanger his second Nobel Prize in Chemistry in 1980—the method provided the primary sequencing platform of molecular genetics and genomics for three decades.

Using dideoxy chain termination, Sanger and colleagues decoded the first viral DNA genome, the 5,386-nucleotide single-stranded bacteriophage ΦX174 (1977), followed by the 16,569-bp human mitochondrial genome (1981) and the 48,502-bp bacteriophage λ genome (1982). Subsequent automation via four-color fluorescence dye-terminator chemistry and 96/384-capillary array electrophoresis enabled the international Human Genome Project (1990–2003) and Celera Genomics' whole-genome shotgun sequencing effort to assemble the 3-gigabase human reference genome. Today, alongside high-throughput second-generation platforms, automated Sanger sequencing remains the clinical and research reference for single-amplicon validation (700–1,000 bp reads at 99.99% [Q40] accuracy).

Nucleophilic Attack and the Missing 3′-Hydroxyl Group

During standard template-directed DNA synthesis, DNA polymerase anneals a synthetic oligonucleotide primer (18–24 nt) to a denatured single-stranded DNA template and catalyzes stepwise 5′ → 3′ elongation. In each catalytic cycle, the 3′-hydroxyl (3′-OH) group of the terminal deoxyribose sugar at the 3′ primer terminus functions as a nucleophile (activated by a catalytic Mg²⁺ ion in the polymerase two-metal-ion active site), attacking the α-phosphorus atom of an incoming canonical 2′-deoxynucleoside triphosphate (dNTP: dATP, dTTP, dCTP, or dGTP) to form a phosphodiester bond and release inorganic pyrophosphate (PP_i).

Sanger sequencing exploits 2′,3′-dideoxynucleoside triphosphates (ddNTPs: ddATP, ddTTP, ddCTP, and ddGTP)—synthetic nucleoside analogs that lack hydroxyl groups at both the 2′ and 3′ carbons of the ribose ring, carrying a non-reactive 3′-hydrogen (3′-H) atom in place of the 3′-OH group. When DNA polymerase incorporates a complementary ddNTP across from its template base, the nascent phosphodiester bond forms normally at the 5′-phosphate of the ddNTP, but the absence of a free 3′-OH group on the incorporated dideoxy sugar prevents further nucleophilic attack on subsequent incoming dNTPs. Chain elongation halts at that nucleotide position.

Stoichiometric dNTP : ddNTP Ratio and Nested Fragment Generation

By spiking a sequencing reaction with a molar excess of all four standard dNTPs alongside a small, calibrated fraction (~0.5% to 1%) of ddNTPs (typically ~100–200 μM dNTPs to ~1–2 μM ddNTPs), chain termination occurs stochastically at only a small percentage of template encounters for each base position. Across billions of template copies amplified over 25–30 linear thermal cycles (cycle sequencing using thermostable Taq polymerase), the reaction generates a population of single-stranded DNA fragments that all share an identical 5′ primer terminus and terminate with a specific ddNTP at every single A, T, C, and G position across a 700- to 1,000-nucleotide window.

Four-Color Dye-Terminator Chemistry and FRET Donors

In Sanger's original 1977 protocol, the template had to be split into four separate reaction tubes (one for ddATP, ddTTP, ddCTP, and ddGTP) spiked with radioactive [α-³²P]-dATP or [α-³⁵S]-dATP, run side by side in four lanes of a denaturing 7 M urea polyacrylamide gel, and visualized via autoradiography on X-ray film. Between 1985 and 1987, Leroy Hood, Lloyd M. Smith, Michael Hunkapiller, and Tim Hunkapiller at the California Institute of Technology—together with researchers at DuPont (Prober et al., 1987)—automated the workflow by covalently attaching four spectrally separated fluorescent dyes to either the 5′ sequencing primer (dye-primer sequencing) or directly to the four ddNTP bases (dye-terminator sequencing, linked at the C5 position of pyrimidines ddTTP/ddCTP and the C7 position of 7-deazapurines ddATP/ddGTP).

Attaching four different fluorophores directly to the four ddNTPs allowed all four chain-termination reactions to be executed in a single reaction tube and separated down a single electrophoretic lane or capillary. In modern BigDye Terminator chemistry, each ddNTP carries a Fluorescence Resonance Energy Transfer (FRET) cassette consisting of a shared fluorescein donor dye excited by a single 488-nm or 505-nm solid-state laser, linked via a rigid propargylethynylbenzamide spacer to one of four dichlororhodamine (dRhodamine) acceptor dyes emitting at separated wavelengths (~525 nm [green, A], ~555 nm [black/yellow, G], ~580 nm [red, T], and ~605 nm [blue, C]).

F667Y Active-Site Engineering of Taq DNA Polymerase

Wild-type Family A DNA polymerases—including Escherichia coli Pol I (Klenow fragment) and native Thermus aquaticus (Taq) DNA polymerase—discriminate against 2′,3′-dideoxynucleotides, incorporating standard dNTPs 100- to 1,000-fold faster than ddNTPs and producing uneven chromatogram peak heights depending on local sequence context. In 1995, Stanley Tabor and Charles C. Richardson at Harvard Medical School solved this kinetic bottleneck by identifying a conserved aromatic amino acid in the O-helix of the polymerase fingers domain—Phe667 in Taq polymerase (corresponding to Phe762 in E. coli Pol I and Tyr526 in bacteriophage T7 polymerase)—that sterically senses the 3′-OH group of the incoming nucleotide.

Engineering a single site-directed point mutation substituting phenylalanine with tyrosine (F667Y Taq polymerase, commercialized as Thermo Sequenase and AmpliTaq DNA Polymerase, FS) introduced a phenolic hydroxyl group on the enzyme side chain that compensates for the missing 3′-OH on the incoming ddNTP. The F667Y substitution increases ddNTP incorporation efficiency by >250-fold, equalizing peak heights across the sequencing trace and enabling Phil Green's Phred algorithm to assign accurate Phred quality scores (Q = −10 log₁₀ P_error) directly from peak spacing and resolution.

96-Capillary Array Electrophoresis

During the final phase of the Human Genome Project (1998–2003), manual pouring of polyacrylamide slab gels was replaced by capillary array electrophoresis on instruments such as the Applied Biosystems Prism 3700 and 3730xl analyzers. Denatured extension products in formamide are electrokinetically injected into 96 or 384 flexible fused-silica capillaries (50 μm inner diameter, 36–80 cm length) filled with replaceable linear polydimethylacrylamide polymer (POP-7). Under an electric field of 150–300 V/cm, Joule heat dissipates across the high surface-to-volume ratio of the micro-capillaries, allowing fragments from 20 nt to 1,000 nt to migrate past a laser detection window in under 120 minutes, generating a four-color fluorescence electropherogram (chromatogram) stored in .ab1 format.

Although high-throughput short-read platforms generate millions of reads per run at lower per-base cost, Sanger sequencing retains a central role in molecular biology workflows:

• Plasmid, Cloning, and CRISPR Edit Verification: Confirming the sequence fidelity of site-directed mutagenesis constructs, synthetic gene inserts, viral vectors, and targeted CRISPR-Cas9 knock-in/knockout edits across 500–1,000 bp amplicons.
• Clinical Variant Confirmation and Familial Segregation: Orthogonal validation of pathogenic single-nucleotide variants (SNVs) and small indels detected in clinical exome panels, including within high-GC exons (when supplemented with betaine/DMSO) or regions masked by homologous pseudogenes (such as PMS2 or CYP2D6).
• Heterozygote and Mosaicism Detection: In diploid genomic DNA, a heterozygous point mutation appears in the Sanger chromatogram as two overlapping fluorescence peaks of roughly equal (~50:50) height at a single nucleotide position, with a lower limit of somatic mosaicism detection of ~15% to 20% variant allele frequency (VAF).

Why does a Sanger sequencing reaction require both standard dNTPs and 2′,3′-ddNTPs in a specific ~100:1 molar ratio?

If a reaction contained only 2′,3′-ddNTPs, DNA polymerase would incorporate a chain-terminating dideoxynucleotide at the first template base adjacent to the 3′ end of the primer, halting every strand at length n + 1. Conversely, if the reaction contained only standard dNTPs, every strand would extend all the way to the 5′ end of the template without terminating. Using a ~100:1 molar ratio of dNTPs to ddNTPs ensures that at each template position, ~99% of elongating strands incorporate a normal dNTP and continue growing while ~1% incorporate a ddNTP and terminate, producing a balanced distribution of fragment lengths spanning 1 to 1,000 nucleotides.

Why does Sanger chromatogram quality degrade in the first 15–40 bases ("dye blobs") and beyond 850–1,000 bases?

Within the first 15–40 bases downstream of the sequencing primer, unincorporated fluorescently labeled ddNTPs and residual primer dimers co-migrate through the capillary polymer alongside the shortest extension products unless removed by paramagnetic bead or Sephadex spin-column cleanup, producing background fluorescence peaks known as dye blobs. At the opposite end of the read (>850–1,000 bases), two physical factors degrade single-base resolution: (1) the fractional mass and length difference between a 900-nt fragment and a 901-nt fragment is only 0.11%, causing electrophoretic diffusion bands to overlap; and (2) the number of surviving un-terminated strands declines geometrically with read length, reducing the signal-to-noise ratio.

What happens to a Sanger sequencing chromatogram when a patient carries a heterozygous insertion or deletion (indel)?

In a diploid sample carrying a heterozygous indel (for example, a 1-bp deletion on one allele and the wild-type sequence on the homologous chromosome), both alleles amplify during PCR and are sequenced simultaneously in the same capillary. From the 5′ primer up to the site of the indel, the chromatogram displays clean, single-color peaks. Downstream of the indel position, the two alleles are out of phase by the length of the insertion or deletion, causing every subsequent nucleotide position to display superimposed double peaks across the remainder of the trace—which computational tools such as TIDE or PolyPeakParser deconvolve into the two constituent alleles.

1. Sanger, F., Nicklen, S., & Coulson, A. R. (1977). "DNA Sequencing with Chain-Terminating Inhibitors." Proceedings of the National Academy of Sciences, 74(12), 5463–5467. https://doi.org/10.1073/pnas.74.12.5463
2. Smith, L. M., Sanders, J. Z., Kaiser, R. J., Hughes, P., Dodd, C., Connell, C. R., Heiner, C., Kent, S. B., & Hood, L. E. (1986). "Fluorescence Detection in Automated DNA Sequence Analysis." Nature, 321(6071), 674–679. https://doi.org/10.1038/321674a0
3. Prober, J. M., Trainor, G. L., Dam, R. J., Hobbs, F. W., Robertson, C. W., Zagursky, R. J., Cocuzza, A. J., Jensen, M. A., & Baumeister, K. (1987). "A System for Rapid DNA Sequencing with Fluorescent Chain-Terminating Dideoxynucleotides." Science, 238(4825), 336–341. https://doi.org/10.1126/science.2443975
4. Tabor, S., & Richardson, C. C. (1995). "A Single Residue in DNA Polymerases of the Escherichia coli DNA Polymerase I Family Is Critical for Distinguishing Between Deoxy- and Dideoxyribonucleotides." Proceedings of the National Academy of Sciences, 92(14), 6339–6343. https://doi.org/10.1073/pnas.92.14.6339