After restriction of a DNA sample with restriction enzyme(s), the fragments often require separation and analysis.

DNA fragments can be separated on the basis of their size using gel electrophoresis (or ultra centrifugation).

Units of size = (mega)Daltons [molecular weight] or (kilo)base pairs - (k)bp. The latter units are better. Why??

Fragments >500 bp (i.e. >0.5 kb) can be separated using agarose gel (e.g. at 0.8%). Polyacrylamide gel (e.g. at 20%) is used for smaller fragments - can discriminate 1 bp difference.

DNA is negatively charged and moves to anode (+) and, therefore, DNA samples must be loaded into wells at the cathode (-) end.

Use a standard of fragments of known size, e.g. restricted wild type lambda DNA.

Visualize bands using ethidium bromide and UV light.

Standards are included in every gel because no two runs will be the same. Standards enable the size of unknown fragments to be determined. A commonly used standard is wild type lambda DNA cut with HindIII which gives 8 fragments with a good size range, although at most only 7 fragments can usually be seen since the eighth fragment is too small to be visualized. 


CATHODE (-) END...........................................ANODE (+) END


1 | ...|... .....| ... .....|... | .........|. .| ....... .| ...... ......

2 | ......................|

3 |.................. | ....|

4 | .............|

5 | ...|... .....| ... .....|... | .........|. .| ....... .| ...... ......

6 | ...............................| ..........| ..|

7 | .......| .| .......|

8 | ..........................................................| ......|

...............    migration>>>>>

(Standards are in wells 1 and 5 with samples in the other wells.)

Migration rate (mobility) is inversely proportional to log10 size of DNA fragments:

1/mobility = k x log10 size

Or for larger fragments, the reciprocal of mobility is proportional to size:

1/mobility = k x size

The constant, k, will vary depending on factors affecting each gel run such as gel porosity, gel thickness, pH, voltage and time - hence the need to use standards for each gel.

Note: Large DNA fragments deviate from this linear relationship.

A rough estimation of the size of an unknown DNA sample can be made from examination of the gel but for more accurate sizing the distances of the bands from the wells must be measured and a graph drawn. The diagram below shows a standard consisting of wild-type lambda DNA cut with HindIII restriction enzyme. This enzyme cuts the linear lambda DNA 7 times yielding 8 fragments showing a good size range for a standard. Note that the units used in the diagram below are base pairs rather than the more commonly used kilobase pairs.

From a graph of migration distance v. log size of standard fragments the exact size of unknown fragments can be extrapolated.

Once a graph has been plotted, the straight part of it can be used to calculate the size of unknown DNA fragments. (Ignore larger DNA fragments when drawing the 'line of best fit' of the graph. The mobility of large fragments does not fit a linear relationship.) Calculate the size of an unknown using direct extrapolation from the graph by drawing a vertical line up from the x axis at the distance migrated by the unknown. Where this line intersects the graph line, draw a horizontal line across to the y axis. This value is the log of the size of the unknown DNA sample. The value must be anti-logged to obtain actual size.  

Alternatively, the formula for a straight line can be used. The formula for a straight line is:

y = mx + c

y = vertical axis (in this case log10 size of known DNA fragments, i.e. the standard)
x = distance migrated by the standard fragments
m = slope of graph (in this case it is a negative value)
c = a constant

Using the formula to calculate the size of unknown DNA samples is more accurate than using the graph extrapolation method but the formula must obviously be known first. Excel spreadsheets can be used to input the standard data and the program can then draw a graph as an X-Y scatter graph (Excel calls graphs 'charts'). However, remember not to include data for the larger standard fragments such as the first HindIII-digested WT lambda fragment which is approximately 23 kilobases in size - this is too large to fit a linear relationship and must be ignored when plotting the line of best fit. It is also possible to request the program to provide the complete formula for the straight line including the m and c values. It is then possible to calculate the size of unknown samples by using the formula and inserting the relevant x (distance) values.

But mobility is not only determined by size but also by the configuration of the DNA:






Open, circular DNA has a "nick" in one of its strands. If supercoiled DNA is "nicked", it unravels into open, circular form.



Under most conditions supercoiled (SC) DNA runs faster than linear (L) followed by open circular (OC) and covalently closed circular (CCC), which are the slowest of all. Open circular and covalently closed circular run at identical speeds and cannot be distinguished or separated on a gel.

As the standard is generally linear, correct sizing can only be performed on unknown linear fragments. Using linear standards to size supercoiled samples will usually give an artificially small size whilst CCC and OC samples will give an artificially large size when compared to linear standards. You can only compare "like with like".

Bacterial plasmids are mainly present in the cell in supercoiled form. However, there will be smaller quantities of CCC and OC form also present due to the fact that supercoils have to removed from DNA prior to replication and transcription inside the cell and also because extraction of plasmids from the bacterial cell may cause some damage (nicking).

Gels can be used:

Other types of gel electrophoresis include:

Details of practical techniques and laboratory protocols can be found in reference books, e.g

Sambrook, J. (2001).
Molecular Cloning: a Laboratory Manual. 3 Volumes. 
Cold Spring Harbor Laboratory Press.

DNA can be recovered from gels by:

1. Electroelution

Remove gel slice containing band of interest.

Place in dialysis bag and apply current.

DNA is attracted to the anode (+).


DNA is released from gel and can be recovered from inside of bag.

2. Low melting point agarose

Special agarose used - melts at 37oC (before DNA).

Cut out slice/band.

Melt it.

DNA is released.

But expensive!!

3. Special

Commercial systems are now available where the DNA contained in a gel slice is adsorbed to material in a tube that has a high affinity for DNA.


Developed by E.M. Southern.

Ref: J. Mol. Biol. (1975) 98, 503-517.

Used to detect fragments of DNA in a gel separated by electrophoresis that are complementary to known DNA or RNA sequences (probes).


Sample DNA may be first cut with restriction enzymes, as in RFLP's. But not necessarily.

1. Pre-treat gel with acid shorter fragments.

2. Treat with alkali denaturation (ssDNA).

3. Neutralize.

4. Set up blot. (See diagram below.)

DNA is drawn up towards filter membrane (eg. nitrocellulose) and sticks to it.

5. Bake filter at 80oC. DNA binds to filter.

6. Expose filter to labelled probe (RNA, ssDNA or oligonucleotide).

Probe hybridizes to complementary sequence(s) in sample.

7. Wash well. Why??

8. Detect presence of probe, eg. for radio-labelled probes using X-ray film and autoradiography .

Known sequences in probe can be matched with unknown fragments/bands.

Click on thumbnail below for diagram of Southern blotting technique

Southern blots are very sensitive and can even detect single-copy genes in a mixture.

Southern blots have various uses. Can you suggest some??

But they do have one major disadvantage. Do you know what this is??


Polymorphism = a variation (natural or induced) in the base sequence of DNA, e.g. insertion, deletion, substitution.

RFLP = variation in the size of DNA fragments generated by restriction enzymes.

DNA sample is treated with restriction endonuclease(s) fragments of different sizes.

The size of fragments and their distribution will depend on the DNA sequence of the sample. The presence of a mutation (i.e. a change in the DNA sequence of a gene) may produce different sized fragments from normal because:

After electrophoresis and hybridization with labelled probe(s), i.e. Southern blotting, a distinctive pattern may be recognized. Extra, or fewer, bands may be produced when the mutation is present compared to when it is absent.

A particular pattern may be diagnostic ("informative") for a certain disease, e.g. due to gene mutation, gene alteration, abnormal gene, or a disease process.

Diseases diagnosed using this (or similar) technique include:



A 'point mutation' is a change in only one nucleotide (or base) in a gene and is better called a single nucleotide polymorphism (SNP). In the case of sickle cell anaemia one change in a nucleotide (from A to T) in a particular section of the HBB gene coding for part of the haemoglobin molecule causes the amino acid glutamic acid to be replaced by valine. This has a very dramatic effect on the haemoglobin and the red blood cells and will cause the disease if present in the homozygous state in an individual.

However, not all point mutations affect the activities or functions of genes or their products and cause disease. Some point mutations have no observable effect whilst some can affect a patient's response to some drugs - sometimes causing them to be ineffective or producing adverse drug reactions. For example, a specific SNP is associated with abnormal metabolism of beta blocker drugs. This explains why beta blockers are unable to lower the blood pressure of some patients with hypertension. Some SNPs are associated with predisposition to certain diseases or conditions, e.g. a SNP in the human G protein beta-3 sub-unit is associated with hypertension, obesity and low birth weight.

SNPs can be detected by a variety of techniques including PCR, probes, blots, and ARMS.

Proving the link between a SNP and drug response or predisposition to disease will require the testing of a large number of subjects or patients and statistical analysis to show how significant the link is.


Northern blots

RNA does not bind to nitrocellulose!

RNA transferred from a gel to, e.g. chemically treated paper, before detection by labelled probe.

DBM (DiazoBenzyloxyMethyl) can be used for the paper treatment.

This will also bind denatured (ss) DNA.


Western blots

Proteins transferred from a gel after iso-electric focusing to nitrocellulose or nylon membrane.

eg. used for antibodies.

Second labelled antibody may be used for visualization.

What about Eastern blots??

Dot blots

Can be used with unfractionated DNA or RNA to measure amount of target sequence in a sample.

Simple, quick.


Colony and plaque blots

Used for the detection of recombinant clones produced by a host-vector system.

plasmid vectors colonies

lambda phage vectors plaques

cosmid vectors colonies

e.g. Grunstein-Hogness colony hybridization method for detection of recombinant clones uses replica plating, hybridization with a radio-labelled probe and autoradiography.

Grunstein.gif (22548 bytes)

Other methods of DNA analysis include:

1. Nucleotide/base sequencing (using Maxam-Gilbert or Sanger-Coulson method).

2. Determination of G=C (guanine-cytosine) ratio relative to A=T.

3. DNA and RNA hybridizations for assessment of homology.

4. Detection or analysis of sequences/genes using probes and/or PCR.

5.  Microarrays ("gene chips") are microscopic ordered arrays of specific probes
    on a planar surface. They are capable of diagnosing a range of diseases in one

Standard curve exercise 01
































Using the number of base pairs, rather than molecular weight, means that the number of codons in a sample of DNA can easily be calculated (by dividing the number of base pairs by 3) and also the approximate number of amino acids in a polypeptide coded for by that sample.











To remove unattached/unbound probe that has not hybridized with the target sequence.














Southern blots can be used to detect any DNA sequence or change in a sequence. A sequence unique to a microorganism can be detected and thus the infectious disease diagnosed. Particular genes or mutations can be detected which may be diagnostic for an inherited disease.














The main disadvantage of Southern blots is that they are rather time consuming and, therefore, rarely used in routine medical laboratories. Some tasks previously carried out by Southern blot are now done using the polymerase chain reaction (PCR) which is faster.














They don't exist. Complete boxing of the compass is still awaited!!