back to Molecular Biology Homepage

back to Homework Page

Here are some full-credit student answers to Homework #3, and some notes from me...

The mechanism of ampicillin resistance is that the "amp/r" gene encodes an enzyme capable of degrading ampicillin. In the absence of this enzyme, ampicillin interferes will bacterial cell wall synthesis, effectively lysing the bacteria.

Given this information, why do you think the 30-60 min incubation is necessary? What is happening during this time that is important for ampicillin resistance?

Transforming the resisance gene into a bacterium does not confer immediate resistance. Enough time must be given after transformation to allow for the transcription and translation of the gene into the "amp/r" enzyme.

(1 pt for answers that said that the extra time was required to complete DNA uptake. If DNA uptake required an extra 30 - 60 min, then you would have to wait ANOTHER 30-60 minutes to allow gene expression before plating on ampicillin.)

2. (2 pts) To obtain oligonucleotide primers for use in PCR, as probes, or as sequencing primers, chemical synthesis is used. (This process is outlined in figure 7-9 in your textbook.) There are companies that specialize in doing this efficiently and cheaply, and most researchers (including us) simply order primers from these companies as we need them.

A limitation on the chemical synthesis of oliogonucleotides is "coupling efficiency," the efficiency with which new nucleotides are added to the growing chains. The company we're using claims a coupling efficiency of about 99% for every nucleotide after the first one (for which "coupling efficiency" is 100% by definition). In other words, 99% of the single nucleotides attached to the column will be successfully joined to the second nucleotide in the chain. 99% of these will be successfully joined to the third, and so on.

Given this efficiency, what is the percentage of oligonucleotides that will be full-length if the primer ordered is 20 nt long? 83% (0.99 to the 19th power)

What is the percentage if the primer ordered is 100 nt long? 37% (0.99 to the 99th power)

Comment on the usefulness of this procedure for accurately synthesizing long oligonucleotides.

This procedure is good for making shorter length oligonucleotides, but as the length of the olionucleotide increases the efficiency decreases drastically. With an efficiency of 37% for a 100 nt oligonucleotide it seems that this technique is not the best choice, unless you are able to purify the full-length oligonucleotides away from the others.

3. (2 pts) The size of restriction enzyme recognition sites varies by enzyme, usually in the range of 4 - 8 basepairs. EcoRI, for example, has a 6-bp recognition site: GAATTC. AluI's recognition site is only four bp: AGCT.

You have a 5000 bp plasmid about which you have no sequence information. To begin to characterize this plasmid, you decide to do some restriction enzyme digests and agarose gels. From previous experience, you've learned that this is most successful if you have less than 8 or 10 bands per digest, and most are in the range of 800 - 7000 bp.

Based simply on probability, how often will an AluI site appear in DNA, on average? How many fragments would you get from an average 5000 bp fragment?

once every 256 basepairs, or about 20 times in a 5000 bp plasmid. 20 cuts of a circular plasmid would give you 20 fragments.

What about EcoRI? How often will it cut, on average, and how many fragments would you expect?

once every 4096 basepairs, or about once in a 5000 bp plasmid. You'd get one linear fragment.

Which enzyme would be a better choice for your analysis?

[Neither would be ideal. I accepted either answer if it was accompanied by a reasonable explanation. Here are two full-credit answers...]

1) Ideally you'd like to have 8 - 10 bands for analysis. I believe that for this reason the AluI enzyme would be the better of the two to use for this analysis, based on the variety provided by the multiple fragments of AluI.

2) I would choose EcoRI because it makes fragments of 800 - 7000 bp long, although it cuts at only one spot. It seems like 20 fragments are too hard to deal with and they would likely be very small, making detection by agarose gel inaccurate.

NOTE 1: Confronted with this situation, I'd actually use a combination of different 6-base cutters like EcoRI.

NOTE 2: Several of you suggested using an AluI partial digest. This would actually make the situation worse, not better. This is because in a partial digest, no one particular site is recognized any better than another--the "decision" of where to cut is random. Thus, for example, if you let a digest go long enough that each plasmid was cut approximately 3 times, you'd end up not with three fragments, but with a set of fragments representing all of the possible combinations of 3 sites.

4. (4 pts)

[disclaimer. The following is a fictional research situation, invented for the purposes of this question. Any resemblance to actual research questions is purely accidental, and probably minimal.]

You are an investigator studying changes in brain function. Specifically, you've long been interested in the biochemical changes that correlate with symptoms of Alzheimer's disease. This question, however, has been very difficult to pursue experimentally.

Recently there has been a breakthrough of sorts. A brain syndrome has been described in a simpler mammal, the furster, which seems to parallel Alzheimer's in humans. Some concerned furster owners have donated the brains of their recently deceased pets (both with and without the Alzheimer's-like syndrome) to your laboratory, and you have discovered to your surprise that you are able to grow cells from these brains in petri dishes. You have cultures of both normal and diseased cells.

Through biochemical analysis, you have identified a short protein which is consistently present in diseased cells and absent from normal ones. You've developed an assay that demonstrates that this new protein, which you've named plackogen, accumulates in plaque-like structures that resemble the plaques found in Alzheimer's tissue. However, preliminary sequencing of the protein has not revealed any homology to known proteins from any organism.

You are interested to know whether expression of plackogen is sufficient to trigger the formation of plaques in otherwise normal cells. To test this, you want to add plackogen protein to normal cells. However, you have been unable to get the normal cells to take up significant levels of plackogen protein when it is simply added to the medium. As an alternative, you propose to clone the gene for plackogen and transfer it to the normal cells. It isn't directly clonable by PCR, because you have been unable to obtain protein sequence from the ends of the protein.

a) What kind of library will you construct to clone the plackogen gene? Why?

A cDNA library would likely be the best. The genome of the furster is likely to be too large to easily make a genomic library. In addition, the plackogen gene is known to be expressed in diseased cells, thus a source of mRNA is readily available.

b) In broad terms (at the level of detail that was used in class), how will you construct this library?

mRNA can be isolated from diseased furster cells by the use of an oligo-dT string that will base pair with the poly-A tails of mRNA after the cells have been lysed. Once the mRNA is isolated, reverse transcriptase can be used to create a cDNA of each expressed gene. After adding restriction sites, the cDNA can be ligated into plasmids and then transformed into bacterial colonies.

NOTE: to be useful, a library must be carried in either bacterial or viral hosts, each of which carries a separate DNA insert and replicates to form a clone of identical individuals. It is not sufficient to make the cDNA fragments--they must be ligated into a vector (either a plasmid or a lambda vector) and transformed into bacteria or packaged into virions.

c) How will you screen the library to find the clone you want? Specifically, what will you use as a probe, and how will you design it? (You needn't describe the steps involved in screening, just the design of the probe.)

In designing the oligonucleotide probe (most likely degenerate), I would select an area of amino acid sequence that is as non-degenerate as possible out of the section of known protein sequence. I would make a probe about 20 nt in length, in order to ensure high specificity, hopefully without introducing too many variables. Onto this sequence, I would attach a label. (The label would be a radioactive phosphate at the 5' end if autoradiography were to be used.)

d) Once you have the gene, what changes will you need to make to your construct before you introduce it into cultured furster brain cells? (Assume that an efficient method of DNA delivery into cultured furster brain cells exists.)

After identifying which plaques contain the desired gene, we need to prepare a plasmid that will work in brain cells. We will need a plasmid with the following characteristics: a marker which allows us to recognize which brain cells have taken up the plasmid, an origin of replication which works in these brain cells, and splice sites which allow us to insert the gene after we chop it out of the lambda vector.

e) Go back and think about your answer to d again. There are some fairly obvious changes to make, and some more subtle ones.

Because we want the gene to be expressed, we will also need a promoter for the gene which works in these brain cells.

top of page

back to Homework Page

back to Molecular Biology Homepage