Wednesday, November 22, 2006


Today I tackle the sxy manuscript discussion. We still have a little more work to do on the RNase data/figure/results section, but otherwise it is only the discussion that needs to be massaged from notes into a coherent flow. The next draft of the manuscript should be done by the end of today or tomorrow - watch out lab mates, it's editing time.

In non-science news, I spent nearly an hour and half yesterday in my continued search for an AC/DC adaptor for one of our weigh scales. The dead adaptor has a highly unusual combination of voltage, amperage, and polarity but I did eventually find a replacement - unfortunately, it comes from China and we must purchase at least a 1000 units. Amazingly, the units are only $1.25 each. This a clear example of the rape of China where a sizeable piece of electronics, wiring, plastic, and resulting pollution is valued so low yet still includes a profit for the manufacturer. On the plus side, we can have it labelled with any company name we want, so it could be a custom "Redfield lab" power supply.

Thursday, November 09, 2006

A sexy new machine

Yesterday I ran an experiment to test whether my failed Friday the 13th real-time PCR was due to bad master mix or a bad machine. I ran a very simple set of reactions using a dilution series of chromosomal DNA as template. Reactions were run with old or new master mix on the new real-time PCR machine in the lab next door. All primers generated nice amplification plots regardless of which master mix was used, thus my Friday the 13th problems were caused by our old machine. Today I have repeated my experiment from last month, but data analysis is going slowly as I learn how to use the new analysis software. Even with the frustrations that often arise when learning one's way through fancy new software, I am generally impressed with the newer machine and software. In the 5 years since our previous analysis software was developed, data visualization and interactiveness has come a long way. Now that I've had my coffee, it's time for another session of real-time analysis.

Monday, November 06, 2006

Opposite to expectations

In past blogs I have alluded to some analyses where I compared sxy mRNA and protein abundance between strains carrying mutations in sxy. I found that hypercompetent sxy mutants produce on average about twice as much transcript but 10-50 times as much protein as wild type cells. In other words, a sxy transcript in a hypercompetent mutant is translated much more efficiently than a transcript in a normal cell. Thus, sxy appears to be post-transcriptionally regulated

I thought these results may be telling us about how sxy expression is regulated in normal cells. I predicted that transferring cells from rich to starvation medium would result in a slight increase of sxy promoter activity and a large increase in translation of sxy. I have crunched the data and, very surprisingly, found the exact opposite result. Upon transfer to MIV, sxy transcript levels immediately skyrocket by 25 fold, then quickly return to pre-induction levels. Protein levels show a much more gradual increase to about 10 fold pre-induction levels, followed by a slow decline.

This result is consistent with what we know about the relative stability of the mRNA and protein – the former is generally very short lived and the latter is long lived. However, it means that cells do not use post-transcriptional regulation to induce sxy upon transfer to inducing conditions – or, if they do, the effect is masked by a very strong inducing signal acting at the sxy promoter. The result may also indicate that in MIV sxy transcript half-life is insufficient to allow significant translation, hence the only moderate increase in protein level. In other words, it is the opposite of what I described above. In a hypercompetent mutant in rich medium, one transcript can make a lot of protein, whereas in wild type cells in MIV, some transcripts may be degraded before even making one complete protein.

Friday, November 03, 2006

Fluorescence is the problem

It turns out that the problem with my real time PCR run on October 13 is something to do with fluorescence. In the last post I described how I didn't see any amplification, but the fluorescence levels appeared abnormally high from start to finish. Today I ran the PCR products out on a gel and found that the PCR worked very well, as expected with my well-used primer sets. The products were clean and lacked any primer dimers. So, if the PCR worked, it is either the fluorophores in the reaction mix that are bad or the machine is broken and doesn't detect them properly.

There are two fluorophores in the reaction mix we use: SYBR and ROX (newer machines don't use ROX). SYBR is the dye that binds to DNA and is used to track amplicon abundance between PCR cycles. Rox, on the other hand, is a passive dye that doesn't interact with anything the reaction. ROX and SYBR have non-overlapping excitation and emission spectra. Thus, the machine can idenpendently read the SYBR and ROX signals: the SYBR signal is expected to vary dramtically between wells and change throughout the PCR run while the stable ROX signal is used to correct for variance that arises because of differences in transparency between wells.

My next step is to compare our old Master Mix with the new suspect batch. I may run this experiment on the new machine next door to make sure it is not our old machine that is the problem. However, this requires learning to use a new machine and analysis software and burrowing PCR tubes from the machine's owner.