Day 2 Afternoon Lecture Notes

Steve Williams, Smith College

June 7, 2004

The RevT gene is found in a repeating sequence called LIN1 (for long-interspersed sequence 1) which is 7 kb long. LIN1 is a reverse transposon and is scattered around all the chromosomes. Transposons splice themselves in and out of genes and change position. Retrotransposons are so-called because they move via an RNA intermediate. Retroviruses may have evolved from retrotransposons.

LIN1 accounts for >10% of the mouse genome. The 7 kbp sequence has two EcoR1 sites. The actual RevT coding sequence of 1.4 kbp lies between these EcoR1 sites. The huge number of RevT repeats in the genome gives a distinct band in an EcoR1 digest at 1.4 kbp. Stained gels can be misleading because higher MW fragments bind more ethidium bromate dye than smaller fragments. Not all retrotransposons contain a sequence that will code for the reverse transcriptase protein; some are non-functional "pseudogenes." Reverse transcriptase makes DNA from RNA.

SINEs are "short interspersed repeats" or "short tandem repeats." The famous ALU repeat is 300 base pairs long. Repeated sequences make up about 50% of mammalian genomes.

Wild-type lambda phage has 5 EcoR1 sites. The genetically engineered cloning vector Lambda Zap Express has one precisely positioned EcoR1 site separating the longer "left arm" (28.3 kbp) and shorter "right arm" (10.5 kbp). In experiment 1 we will insert the mouse RevT gene between the two arms of Lambda Zap Express. DNA ligase will will form new phosphodiester bonds between the 3' and 5' ends of the backbone for a total of 4 ligations. The mouse DNA is then integral with the lambda DNA.

During such a process, misligations (mouse-mouse, right-left, left-left, circular mouse, etc.) will occur. Only left-mouse-right is viable along with left-right. Other possibilities are missing some of the genes lambda needs to reproduce. How can we prevent the religation of the two lambda arms without the mouse insert?

We use calf intestinal alkaline phosphatase (CIAP). This enzyme removes phosphates from the 5' ends of the vector. Now the left and right arms, like phosphates, can't religate to each other. The mouse insert isn't treated with CIAP so it retains the 5' phosphates and the ability to ligate to the left and right arms. Only insert-5' to vector-3' will work as the vector-5' has no phosphate. With 2 out of the 4 backbones bonded along with the bases, the molecule will be reasonably stable. When the ligated molecule is put into E. coli, a cellular enzyme will repair it.

Nothing mentioned so far will prevent the insertion of more than one mouse RevT gene. This misligation is prevented by using few insert molecules in the reaction. To get 1:1 insert:vectors, set (ng vector/length vector) = (ng insert/length insert). Following p. 36 of the lab protocols, (200 ng insert/4000 bp)/(1000 ng vector/39x10³ bp) = 2, the maximum amount advisable. Here the high 2:1 ratio accounts for loss in dialysis and other procedures.

How is Lambda Zap Express modified from wild-type phage? It has only one EcoR1 site and its length is reduced from 48.5 kbp to 39 kbp by removing untranscribed DNA. Due to limited space in the head of the phage, the maximum length of the lambda genome is 51 kbp. Therefore LZE can accommodate inserts up to 12 kbp long as opposed to only 2.5 kbp for wild-type phage. The molecular motor which packages DNA into the head doesn't recognize a chromosome shorter than 39 kbp, so that's the effective lower bound.

The LZE product is DNA which is already cut and dephosphorylated (i.e. terminal phosphate groups are removed). Nonetheless there is still a small amount of viable left-right ligation product without an insert.

Why do we have to heat the lambda MW markers before running them on the gel? In wild-type lambda, the 5' ends have a 12-nucleotide (nt) overhang. These "COS" ends can ligate and form a ring. Heating the lambda marker unbinds these ends and linearizes the DNA. 65C is hot enough to unbind the 12 hydrogen bonds that hold the ends together. Various restriction enzyme digests of lambda (Lambda BSTEII, Lambda Hind for example) are useful as they give a number of predictable bands. The fragments would bind together to form higher MW products if the markers were not heated. The enzyme that packs phage DNA into the head also linearizes the circular form.

Viable lambda bacteriophage can easily be assembled in a test tube: mix proteins, stir and wait 2 hours. Stratagene (San Diego) and Promega both sell a "packaging extract" that include head proteins, tail proteins and assembly proteins. Just add DNA and you have all the ingredients to make phage. The DNA must have 39 kbp <= length <= 51 kbp; it must have both the left and right arms; and it must have the COS ends.

Other methods besides phage and plasmids can insert DNA into E. coli, including electroporation and "gene guns," but phage infection is by far the most efficient in term of cells transformed per unit starting DNA. Phage form clear "plaques" on the "bacterial lawn" of a Petri dish. Each plaque may have a different insert from the mouse genome. Special E. coli strains with no restriction sites in their genome are used for transformation. A normal Petri dish can accommodate about 10⁴ inserts.

Since the RevT gene is about 10% of the mouse genome, a plate with a hundred plaques will almost always have a RevT plaque. A single-copy gene will require a much larger plate in order to find a plaque. The reason why plaques self-limit in size is not known. The phage must be low enough in concentration that the plaque boundaries can be visualized.