20.109(S07): Yeast transformation

Introduction

Two toddlers, a can of white paint and a new flat screen TV..."what could possibly go wrong?" Microsoft Windows operating systems for web-enabled ATMs..."what could possibly go wrong?" There is no question that optimism is a fundamentally important element of every experiment we do. We work to design solutions; we plan for success not failure. Even difficult projects should be tackled with the idea that they can, should and will work. Attitude alone, however, does not overcome pitfalls and there is much to be gained in applying some of that optimistic energy to answering: "what if...?". And oddly, spending some time being negative before you jump in is likely to improve your chances of a successful outcome since some of the pitfalls we encounter are avoidable if they can be anticipated.

In the laboratory, "controls" help answer the "what if" questions. They are equally or perhaps more important than your experimental samples and often more numerous. If the controls in an experiment haven't worked there is very little point in considering the data you have collected. Even experienced researchers often wish they had included a better or a different control for an experiment since data often leads to more questions, some of which might have been anticipated.

Your experiment today is straightforward: transform your yeast cells with your PCR product and look for cells that can grow on media lacking uracil. What if you return to lab next time and saw no yeast colonies growing on any of your petri dishes? What if you return to find yeast covering the plates completely? How can you be sure these are even yeast and not bacterial colonies? You've transformed and transfected cells before...what controls were used for these experiments and can they be applied here?

The negative control helps confirm that any positive experimental data is arising from the experimental sample and not some random event. For example, leaving the DNA out of a transformation reaction is a negative control. No colonies should grow from that sample. If colonies do grow with the negative control then perhaps the petri plates are contaminated, or perhaps the plates are the wrong kind, or perhaps the yeast are already able to grow on those plates without your experimental DNA. The negative control can't distinguish these possible explanations but it can tell you that something needs to be re-worked.

The positive control helps eliminate trivial explanations for failed experiments (these are experiments after all, so you don't really know what the outcome should be....). Transforming plasmid DNA into competent cells is a positive control. If no colonies grow, then something is wrong with the cells or the plates. If cells only grow on your positive control and not on your experimental sample, then the experiment should be refined and repeated.

Even "perfect" results can be misleading and need to be carefully considered. Beyond the negative and positive controls, a good researcher thinks through the experimental sample itself to see if "perfect" data could be explained in more than one way. Today, for example, you will remove an aliquot of your PCR product and transform that DNA into competent yeast. What if you returned to lab next time to find these results:

Sample	number of colonies growing on SC-ura
No DNA	0
Plasmid DNA with URA3 marker	1000
PCR product	1000

It seems like this experiment "worked," really really well, but in fact the number of colonies on the experimental sample is too "good." The PCR product isn't expected to give rise to ura+ cells as often as plasmid DNA can. After all we're relying on recombination of the 40 bases flanking the product to integrate the URA3 gene. And if the SAGA-subunit deletion makes the cells even a little bit sicker than wild type, the correct product will be even harder to get. So how can this "too perfect" data be explained?

Remember what is in that PCR sample:

• product (hopefully....in a "real" lab setting you'd check to see that the reactions actually worked before transforming!)
• leftover primers....those should be degraded by the cell
• leftover dNTPs....also transparent to the cell once inside
• leftover Taq...not a problem
• template DNA...hmmm...

What is to stop the URA3 template DNA, just another plasmid after all, from giving rise to cells that grow on SC-ura? To solve this problem the PCR product (expected to be around 900 base pairs) could be purified from the template (around 4 or 5 Kb) using an agarose gel and a Qiagen kit as you did before. A more elegant solution though, is what you will use since the template you were given last time has no yeast origin of replication. Thus the yeast will not be able to copy the plasmid and it will be diluted then lost from your transformed samples. This "trick" has saved some time in the lab but did require some anticipation. Hopefully you can see the value.

Protocols

Part 3: Transformation

DONE!