Copying, Sorting and Reading DNA: PCR, Gel Electrophoresis and Sequencing

From one molecule to a billion

A forensic scientist has a single hair root. A doctor has a swab with a few virus particles. A researcher wants to read one gene out of 3 billion base pairs. All three face the same problem: far too little DNA to work with, mixed in with far too much other DNA. Three techniques solve it: PCR copies just the region you want, gel electrophoresis sorts DNA pieces by size, and sequencing reads the bases. Expectation D3.5 names PCR and DNA sequencing as key processes in genetic modification and analysis.

PCR: the polymerase chain reaction

Kary Mullis invented PCR in 1983 and shared the 1993 Nobel Prize in Chemistry with the Canadian biochemist Michael Smith of the University of British Columbia (honoured for site-directed mutagenesis, a method for changing a gene at a chosen base).

Ingredients in a small tube:

• template DNA containing the target region;
• two primers, one for each end of the target, each complementary to a different strand (their 3′ ends point toward each other);
• Taq polymerase, a DNA polymerase from Thermus aquaticus, a bacterium first found in hot springs in Yellowstone National Park. It is heat-stable, so it survives the repeated heating;
• the four dNTPs (building blocks) and a buffer with magnesium ions, which the polymerase needs.

A machine called a thermal cycler repeats three steps about 25 to 35 times:

1. Denature at about 95 °C: hydrogen bonds break and the strands separate. (No helicase needed; heat does the job.)
2. Anneal at about 50 to 65 °C: the primers base-pair with their complementary sequences. (Primers are in large excess, so they bind before the long template strands can re-pair.)
3. Extend at 72 °C: Taq polymerase adds nucleotides to the 3′ end of each primer, copying the target.

Each cycle doubles the number of target copies (ideally), so after n cycles one molecule gives 2ⁿ copies. After 30 cycles that is over a billion. In practice the reaction slows down as ingredients run out.

Uses of PCR: detecting viruses and bacteria (the standard COVID-19 tests used RT-PCR, in which reverse transcriptase first copies the virus's RNA into DNA); diagnosing genetic conditions; forensic DNA profiles from tiny samples; studying ancient DNA; and copying genes to insert into vectors for genetic engineering. Quantitative (real-time) PCR adds a fluorescent signal so the amount of starting DNA can be measured.

Gel electrophoresis: sorting DNA by size

DNA is negatively charged (from its phosphate groups), with about the same charge per unit length whatever its size. So in an electric field, all DNA moves toward the positive electrode, and the only thing that separates the pieces is how easily they move through the gel.

1. A gel of agarose (a carbohydrate from seaweed) is made with a row of wells at one end and placed in a buffer.
2. DNA samples, mixed with a dense loading dye, are put into the wells, which are at the negative end.
3. Current is applied. DNA moves toward the positive end ("run to red": the positive lead is usually red). Small fragments slip through the gel's pores faster and travel farther; large ones lag behind.
4. The DNA is stained with a dye that glows under a special light, showing bands. Each band is millions of DNA molecules of the same length.
5. A DNA ladder of known sizes in one lane is used to estimate sizes. Distance travelled is roughly proportional to the logarithm of the size.

DNA profiling (fingerprinting)

About 99.9% of the DNA sequence is the same in any two people. Forensic scientists look at the places where people differ most: short tandem repeats (STRs), where a short sequence such as GATA is repeated a variable number of times. At one STR locus, you might have 8 repeats on one chromosome and 11 on the other. Labs use PCR to copy about 15 to 20 STR loci and measure the lengths (today usually with capillary electrophoresis, a gel inside a thin tube). The chance that two unrelated people match at all of them is tiny.

• Forensics: comparing crime-scene DNA with suspects. Canada's National DNA Data Bank, run by the RCMP, opened in 2000.
• Paternity and family testing: a child inherits one allele at each locus from each biological parent, so each of the child's bands must match a band from the mother or the father.
• Identifying remains and wildlife forensics (for example, identifying meat from a protected species).

Sequencing: reading the bases

In 1977 Frederick Sanger invented chain-termination sequencing (he won his second Nobel Prize for it).

1. Copy the template with DNA polymerase, a primer and normal dNTPs, plus a small amount of dideoxynucleotides (ddNTPs). A ddNTP has no 3′ OH, so once it is added the chain cannot grow.
2. Each time polymerase happens to add a ddNTP instead of the normal dNTP, the chain stops. The result is a set of fragments of every possible length, each ending in a known base.
3. Separate the fragments by size. In the original method, four reactions (ddA, ddC, ddG, ddT) ran in four gel lanes. Modern machines label each ddNTP with a different fluorescent colour and run all fragments in one capillary; a detector reads the colours as the fragments pass in order of size.
4. Read from the shortest fragment to the longest: that gives the new strand 5′→3′, which is complementary to the template.

Faster, cheaper, longer

The Human Genome Project (1990 to 2003) used Sanger sequencing and cost about US$2.7 billion. An almost complete, gap-free human sequence was published in 2022 by the Telomere-to-Telomere consortium (the Y chromosome followed in 2023). Since the late 2000s:

• Next-generation sequencing reads millions of short DNA fragments in parallel, and computers assemble them. The newest machines can sequence a human genome for a few hundred dollars in reagents.
• Long-read methods read much longer stretches. In nanopore sequencing, a single DNA strand is pulled through a tiny protein pore, and changes in electric current identify the bases; some devices are small enough to hold in your hand.

Results feed into bioinformatics, the computer analysis of biological data, a fast-growing career field. In Canada, the Centre for Applied Genomics at SickKids in Toronto is one of the major sequencing and analysis centres.