DNA Restriction Mapping

    Techniques based on the use of restriction endonucleases and chromatographic separation of DNA fragments by size allow very precise gene mapping.  The same basic techniques - e.g. restriction endonuclease digestion, separation of nucleic acid fragments by size, using radioactivity or sometimes specially constructed analogs to label molecules of interest and to  distinguish them from other similar molecules, comparison of DNA sequences by hybridization, etc. - in a variety of combinations and with different substrates, enable us to accomplish a wide range of goals.  Whole chromosomes can be mapped, DNA sequence analysis allows intragenic mapping of actual base sequences within or surrounding a transcribed region,  allelic variation can be determined at a molecular level, even in the absence of any measurable phenotypic difference.  This lab exercise will serve as an introduction to some of these techniques and their possible uses, focusing primarily on molecular mapping methods.

    Restriction endonucleases are enzymes which recognize, bind, and hydrolyze DNA at specific nucleotide sequences. Each restriction endonuclease recognizes one specific nucleotide sequence, usually 4-6 base pairs (bp) long.  Hydrolysis may occur at one or the other end of this sequence, or internally, depending on the enzyme.  (see Fig 1.) Thus, complete digestion of DNA by a single restriction endonuclease results in a distinct set of fragments representing the intervals between restriction endonuclease recognition sites.  When separated by size, this pattern is characteristic of that DNA - restriction endonuclease combination.

    The most common method for separating DNA fragments by size is gel electrophoresis (similar methods are used to separate RNA or protein).  A mixture of DNA fragments is placed on a strip of gel (polyacrylamide and agarose are the most common gel substances).  A voltage potential is established across the gel which draws charged particles through the gel. The movement of larger molecules is retarded more than that of smaller molecules, so that eventually distinct size classes may be distinguished in the gel.  A similar method, called liquid chromatography, using a particulate matrix through which substances percolate at rates based on their size and shape is sometimes used instead - mostly for automated DNA sequencing.

    Before DNA can be sequenced it is necessary to generate a quantity of identical, fairly short pieces because the techniques can only accommodate a few hundred bp at a time.  DNA fragments can be cloned by inserting them individually into a vector - usually a plasmid, episome, or viral DNA - and then the altered vector is placed into an appropriate host cell - usually bacteria but sometimes a eukaryotic cell line.  The vector can replicate independently of the host cell cycle, so large numbers may be generated in each cell. Thus as the culture grows, the number of copies of the DNA of interest grows even faster.  Each colony derived from a single altered cell consists entirely of cells genetically identical to each other and to the original cell.  Thus, all contain the vector containing the inserted sequence.  Large quantities of this DNA can be generated under controlled conditions.  The vector can be constructed so that the DNA of interest can be recovered readily, in large quantities, with little contamination.

    A piece of DNA prepared in this way can be sequenced by one of two methods developed by Sanger, and Maxam and Gilbert, respectively.   Briefly, in the Sanger method the DNA to be sequenced is replicated using (a) a 5' end label and (b) 2',3'-dideoxy NTPs to terminate replication.  Resulting fragments are analyzed by chromatographic separation.  This is the method which has proven to be most easily automated through the use of fluorescent, rather than radioactive, labels.  The Maxam and Gilbert method involves (a) radioactive label on the 5' end, and (b) partial digestion with a different sort of restriction endonuclease which cleaves after one or another nucleotide residue in a base-specific manner.

    Analysis of base sequences is very precise, but other methods are also very useful.  Analysis of long pieces of DNA can be done using restriction endonucleases without the need to clone or sequence the DNA.  Using one restriction endonuclease at a time, target DNA is hydrolysed.  This generates a mixture of fragments which can be separated by gel electrophoresis and visualized as a characteristic pattern depending on the DNA and enzyme used.  One restriction endonuclease cuts at one specific sequence which occurs at intervals along the DNA.  These intervals are unique to a DNA sequence.  Thus the fragments' sizes, representing these intervals between  restriction sites, will characterize the particular DNA.  Using several restriction endonucleases in separate experiments generates a restriction map of the DNA which is unique to that DNA sequence.  These maps can be used to identify a particular species of DNA.  One increasingly common use of such a map is called restriction endonuclease fragment length polymorphism (RFLP) analysis.

    RFLP analysis has been used successfully to identify different alleles in a population, and ultimately to locate the genes.  Specifically, several human genes responsible for genetic abnormalities or diseases are being investigated using this methodology.  The approach requires that an abnormality is identified as being inherited in a family.  Samples of DNA from as many members of the family are prepared for analysis by cleavage with a series of endonucleases and separation of the resulting fragments by gel electrophoresis.  From the pedigree it is possible to infer those who are homozygous normal, and at least those who express the trait, if not some heterozygotes.  Thus, DNA from individuals known to carry at least one mutant allele is compared to DNA from individuals known to be homozygous normal.  The restriction digest patterns are compared.  Any variant DNA which either lacks one of the restriction sites, or has an extra one, compared to the normal DNA, will generate a different pattern of fragments when subjected to the same hydrolysis and electrophoresis.  If there has been a sizable insertion or deletion between two restriction sites, that altered pattern can also be identified.  Thus, for the family under study, a recognizable RFLP pattern can be identified which represents the normal allele, and a distinct pattern would represent the mutant allele.  It is important to note that these patterns are valid within a family only, because it is necessary to reduce the likelihood of other genetic variation being superimposed on the relevant restriction pattern.

    One other step is necessary for analysis of RFLPs generated from total eukaryotic genomic digests.  Eukaryotic genomes are so large that a typical restriction digest appears as a smear on a normal gel because there are so many restriction sites, and so many fragments.  In order to generate an interpretable pattern it is necessary to limit the number of fragments visualized at any one time.  The most convenient way to do this is to use a DNA probe which will hybridize to a specific DNA sequence.  The probe is labeled, for example with radioactivity.  After the restriction digest is electrophoresed, the DNA is transferred to a nitrocellulose membrane, and then exposed to the labeled probe.  The labeled probe DNA will hybridize - form base pairs with - its complementary sequences, which are found, on only a subset of the restriction fragments.  An X-ray film placed over the membrane will be exposed by the radioactive emission, so bands will be seen only where there was radioactivity.  Thus a few bands in the smear will be visualized, while the rest remains unhybridized, unlabeled, and invisible.

    Whenever possible the probe is chosen because it contains sequences believed to be related to the gene of interest or close to the gene on the chromosome (from other sorts of analyses).  So what is actually being mapped in these experiments is an RFLP which appears to be LINKED to the gene of interest.  The presence or absence of the RFLP correlates with the presence or absence of the mutant trait.  The RFLP, then, is an marker in its own right.  For this analysis to be useful the RFLP need not be part of the gene under study.  However, the closer the mutated restriction site to the gene of interest, the more closely linked will be the two traits.  If RFLP analysis is being used to predict the presence of one or two mutant alleles, the closer the linkage the more accurate the prediction.  If the goal is to map the gene, it is clear that the closer the two alleles, the more accurate the map.

    Restriction mapping is also used extensively in genetic engineering.  There are several ways in which restriction endonucleases are used.  One is to identify a DNA which one has or hopes to manipulate.  Restriction endonucleases are used to accomplish the manipulation as well.  Since they cleave DNA at specific sequences, they permit the combination of two pieces of DNA at a predetermined specific site.  Today's lab exercise will illustrate some of these methods.

    OH - MY - GOODNESS !!! DISASTER !!!    We have four plasmids in the stockroom freezer and somehow the labels have all fallen off!! How can we tell which is which?  Luckily we have, on file, partial restriction maps for each one.  [  these must be obtained from the instructor ]  So we quickly get out the three restriction endonucleases used for the maps, and digest samples of each unknown plasmid with each of the three enzymes (all in separate tubes).  The resulting fragments from each digest are separated by electrophoresis and the results presented below.  Match the digest data with the plasmid map by inferring the approximate relative sizes of fragments generated by each enzyme.  Write the name of the plasmid on the line below each set of data (three lanes for each plasmid).

    Now that we have relabeled our plasmid stocks with indelible marker, we are ready to do experiments.  Our research demands that we genetically modify one or another of these plasmids in different ways in order to test the function of certain DNA sequences.  We must analyze the products to identify which samples contain the desired alterations.  In order for the technician to identify the correct pattern, predict the patterns that would be generated by the following mutations.  Sketch your predicted electrophoretic patterns in the empty gel lanes.

A. The region of pA1 indicated by the box and * is removed from pA1.  What would digests of pA1 with the same 3 restriction endonucleases look like now?

B, C. The fragment removed from pA1 is inserted into a Pst I site in pR43.  Draw both possible outcomes.  Can you distinguish between them?

D. The Bgl II site indicated by an * in pX3 suffers a point mutation which renders it unrecognizable to Bgl II.  How would we know?


1.  Why are sites sensitive to some restriction endonucleases found more frequently than sites sensitive to others?  Given an enzyme with known specificity, how would you predict whether it is likely to find many target sites or few?

2. Genetic manipulation can be accomplished using restriction endonucleases to cut DNA at specific points.  How are these enzymes used to facilitate linking together DNA from different sources?

3. What is meant by "blunt end" cut? -- "sticky end"?

4. Why would an RFLP pattern shown to correlate with a genetic abnormality in one family be a poor indicator in another family?  Under what circumstances would you expect an RFLP pattern to be a good indicator in most or all people?

5. What are some of the arguments for and against genetic testing?
•    Which sorts of diseases or predispositions would you want to be tested for? Which would you rather not know?
•    Should prospective employers be allowed to take a DNA sample for analysis?  Why or why not?
    •    insurance companies?
    •    the government?
    •    an information clearinghouse that promises confidentiality?
    •    a sperm or ovum bank?
    •     an adoption agency - from the parents of the child so that the information is available in case of future need?
    •    etc.?