Cloning DNA and Finding Genes

Isolating genes:

• "Easy" if the gene product (protein) is known:
• • Create a cDNA library using an expression vector.
  • Probe with antibodies that bind the gene product.
  • Isolate and sequence positive clones.
• More difficult if product is not known:
• • Identify a marker (microsatellite, RFLP) that is linked to the disease.
  • Use a technique called positional cloning to find the gene.
  • • e.g., cloning of the Huntington's Disease gene.

Huntington's Disease

• Dominant autosomal neurodegenerative disease
• fatal
• occurs in about 1/10,000 live births
• Late onset - age 35 to 45 (many have children before symptoms appear)

This was one of the early examples of successfully finding a gene for a major human disease.

Step 1: Find a linked marker

1. First they had to find a pedigree for a family with high frequency of the disease. They found a large family in Venezuela. They collected blood samples from family members with and without the disease, extracted the DNA, and examined lots of DNA markers.

2. They used "restriction fragment length polymorphisms" and looked for restriction sites that were correlated with the disease. What is a restriction enzyme? An enzyme that at a particular sequence of 4-6 bases. There are lots of different enzymes that recognize different sequences. See

See Fig. 7.1 and 7.2

We also talked about how fragments are separated on an agarose gel, and how radioactive probes can be used to label particular fragments (both the fragment are probe are denatured to make them single -stranded; when they reanneal the probe will hybridize to any fragments that have the complementary base sequence).

3. They got lucky and found marker only 4 cM away.

• How often should you find a person with Huntington's disease who does not also have that marker, given that it is 4 cM away?

Eventually they localized the gene to chromosome 4.

They examined many more markers and many more individuals and finally found markers that narrowed the region to about 500 kb.

Step 2: use those linked markers to find nearby DNA sequences

• Make genomic "library"
• Probe that library for the marker
• "walk" to extend known sequence

See Fig. 7.9, 8.8, 7.13

How do you make a library?

• Use a restriction enzyme to make lots of fragments of genomic DNA
• Use the same enzyme to cut the cloning vector (e.g. plasmid).
• The sticky ends will anneal together and they can be sealed with DNA ligase.

See figs 7.3, 7.4, 7.5

We talked about the way pUC plasmids have been engineered to be convenient for cloning- they have a penicillin resistance gene (to select bacteria that contain the plasmid) and they have a lacZ gene (that can cleave xGal to make a blue dye). If foreign DNA is inserted into the lacZ gene, no dye will be produced and the colony will be white.

How do you sequence the final fragments?

• Sequencing reactions build on our understanding of DNA replication. Mix a single-stranded template, short complementary primer, DNA polymerase, and deoxynucleotides and you can synthesize a complementary strand in a test tube.
• In the sequencing reaction, a small amount of di-deoxy ribonucleotides are added which stops synthesis wherever they are incorporated. (Why?)
• ddA, ddC, ddG, ddT are each added to separate tubes
• When fragments are separated on a gel, you can read off the sequence by reading the bases from bottom of the gel (shortest fragments) to the top

See Fig. 7.19 Dideoxy DNA sequencing of a theoretical DNA fragment

A current example: Parkinson's disease

Valente et al, Science, May 2004 characterized the gene for a rare form of Parkinson's

The procedure was similar to that used for Huntington's disease, but they could make use of the complete human genome sequence.

• Find markers linked to the disease using pedigree analysis of families from Italy and Spain
• • Many more markers are available today
  • Narrowed the region to 3 cM (2,800,000 bp)
• Searched the Human Genome database-- 40 genes in that region
• Narrowed the list of candidates:
• • Expressed in brain cells?
  • Similar to other known genes?
  • Sequencing the final candidate gene showed 2 mutations

Yet another example: Cloning Spider Silk gene

How did they do it? (simplified version)

• Knew the protein structure of silk:
• • Repetitive amino acid sequence
  • GGPXGGPX . . . .
• Used the protein to predict the underlying DNA sequence
• • Made degenerate oligonucleotide
  • • Gly = GGA,GGC,GGG, GGU
    • Pro = CCA, CCC, CCG, CCU
    • Make lots of different DNA sequences to match all possible codons for a short fragment of the protein
• Use that to probe a cDNA library
• • What is a cDNA library? Why would they want to do that?
  • (cDNA is only made from expressed genes, so all of the junk and introns have been removed, as well as any genes that are not transcribed in the tissue of interest. Makes a smaller library of clones to search through.)
• What they did:
• • Dissected spider silk glands
  • extracted RNA
  • pulled out sequences with a polyA tail
  • Used reverse transcriptase to copy the RNA to DNA
  • Cloned that DNA into a vector (as above)