The DNA molecule has a long backbone of phosphate molecules on each strand; the backbone can be compared to the uprights on a ladder.  A number of different important laboratory techniques have been developed based on the physicochemical properties of the phosphate backbone; even more techniques are designed around the base pairs, which we discuss in other boxes.  In this box, we describe one of the most important techniques based on the phosphate backbone—gel electrophoresis.

Consider a solution in which DNA molecules of different lengths or sizes are present.  This is a common laboratory situation, either because there has been a reaction to produce more DNA and the in vitro synthesis has resulted in molecules of different sizes or because a much larger DNA molecule has been sheared or broken down into smaller pieces, for example.  What method can be used to separate the molecules of different sizes from each other?

Each DNA molecule is negatively charged because of the phosphate backbone.  In addition, the negative charges are evenly spaced, with one negative charge in the backbone between each base pair.

The negative charge on the phosphate backbone is the property underlying the technique of electrophoresis to separate DNA molecules.  If we put DNA into an ionic solution such as a salt solution, and apply an electric current to the solution, DNA will migrate to different positions based on its size and electric charge.  In fact, DNA is an anion—it has a net negative charge—so it will always run towards the positive pole.

We could put DNA into a salt solution and apply a current but this will not provide a readily visible method to separate DNA by size.  To achieve that objective, we first make a colloidal substance or a gel for the procedure of gel electrophoresis.  We then put the DNA in the gel, immerse that in a salt solution, and apply an electric current for electrophoresis.  The gel anchors the DNA in place before the current is applied.

However, the gel introduces another property that affects how the DNA will migrate in an electric field, and that property is actually more significant that the charge.  If we put DNA molecules of different sizes into a gel and apply an electric current to them, you may think, based on what we have said, that the longer molecules will migrate further (or faster) since they have more negative charges.  That is not true because the gel offers resistance to macromolecular movement and acts like a sieve.  Shorter molecules migrate faster than long ones because shorter ones can move through the pores of the gel more easily.  Think of the gel as being a forest, and the DNA molecules being animals running through the forest.  Small animals, like mice and chipmunks, can run through a forest more quickly than large animals like bears and elk because they get move through small openings between the trees more quickly. In addition, the more densely packed the trees in the forest, the greater the difference between the pace of migration between small and large animals. In the same way, small molecules of DNA run faster than large molecules in the same gel, and both the length and the charge on the backbone determine its position on the gel. In fact, there is a logarithmic relationship between the size of a linear DNA molecule and the distance it migrates on a gel.

Since DNA is negatively charged and migrates towards the positive pole, we make wells in the gel next to the negative pole for loading the DNA before we apply the current.  The gel is usually made of agarose, a synthetic version of the natural polymer agar, although molecules of DNA less than 150 base pairs long can be separated using a different polymer called polyacrylamide.  Polyacrylamide is routinely used to separate proteins by a similar procedure of electrophoresis.  Agarose gels are made by dissolving agarose in a buffered ionic solution; the most commonly used solution has the buffer Tris to regulate the pH, acetate ions to provide the ionic environment, and a chemical called EDTA, so this buffer is called TAE.  The role of EDTA is to reduce the degradation or breakdown of the DNA during the procedure.  Enzymes that degrade DNA, known as nucleases, require magnesium ions for their activity.  EDTA chelates or attaches to magnesium ions in an inactive form, so that nucleases are also inactive.  A typical gel is about 1.5% agarose, although lower percentages (such as 0.8%) can be used for long DNA molecules and higher percentages (such as 2%) can be used for shorter DNA molecules.

While the gel is still a liquid, a plastic comb is inserted at one end; when the gel is solid, the comb is removed.  The teeth of the comb form the wells into which DNA can be added.  The gel is immersed in TAE, the comb is removed, and the DNA is pipetted or loaded into the wells.  The wells are placed at the negative terminus, and an electric current is applied.  DNA molecules migrate towards the positive pole, which separates them by size.  In order to have a standard basis for comparisons, one lane of the gel includes commercially available size markers—that is, a solution that has DNA molecules of defined sizes.  A commercially available loading dye is also typically included in each lane so that the progress of the DNA through the gel can be monitored visually; the loading dye is suspended in a dense liquid such as glycerol so that the DNA solution will sink to the bottom of the well when it is being loaded.

Once DNA molecules have been separated, it is necessary to see where they have migrated.  The most common way to detect DNA molecules in the gel is to use a molecule such as ethidium bromide.  Ethidium bromide is typically added to the agarose solution before it has solidified, and the gel is examined under UV light.

As described here, agarose gel electrophoresis separates linear, double-stranded DNA molecules based on the charge on the phosphate backbone and the length of the molecule.  Separation by gel electrophoresis depends on the size of the DNA molecule rather than on its base sequence, so two molecules of about the same length but of different sequences will migrate similarly.  Related methods can be used to separate RNA and single-stranded DNA molecules, but the process is a bit more complicated since a single-stranded molecule can make intra-strand base pairs.  It can also be used for circular DNA molecules, such as plasmids, but the size estimates are dependent upon the conformation of the circular molecules, which can be hard to predict.