The importance of experimental failure

It is only natural for us to want to succeed in our endeavors, but sometimes failure is more valuable. As Sir Winston Churchill once commented: "Success consists of going from failure to failure without loss of enthusiasm". Replace "Success" with "Science" and you have your future as a Scientist in a nutshell! I am sure that Andy Murray would agree that this philosophy applies just as much to sport, or indeed any creative activity. Failure is simply an outcome that is unexpected and often unwanted. But it may give you an insight into your lack of appreciation of the complete set of variables associated with your experiment.

1         2         3        m

When we set out last week to investigate the composition of the mealworm larvae, the outcome of the manual grinding and extraction was a little disappointing. I expected that grinding would release a significant amount of protein, and indeed the first centrifugation looked very promising. After centrifugation, the Falcon tubes had 4 layers (from the bottom up): exoskeleton (brown), disrupted cells (pale grey), around 30ml yellow supernatant and finally a fatty layer floating on the supernatant. Typically, a viscous yellow supernatant would be protein rich. However, looking at the samples run on SDS PAGE (right) suggests that the protein concentration is low, with one high molecular weight protein predominating (see lanes 1 and 2). Those of you who took spectral readings noted an absorbance maximum around 450nm. So whey was the "proteome" so simple and the total protein concentration so low? 

Some of you attempted to fractionate the sample using a Q Sepharose column. A brown colouration at the top of the column at low ionic strength (PBS buffer) was displaced by 1M NaCl. I thought this again looked promising, but SDS PAGE revealed a protein concentration too low to easily detect by Coomassie Blue staining. So what do you do under these circumstances?. Well this apparent failure, led us to see if we could concentrate the protein using ammonium sulphate and acid precipitation. As you can see left. this did do the trick (to some extent, and ideally, we should have dialysed the samples before loading, to neutralise the pH and remove the salt), and indeed confirmed a lower than expected protein concentration in the extract the supernatant, as before. We still see a major band at high molecular weight and little in the body of the gel.

M      1       2       3

I was bothered that the larvae contained a major protein and little else and I also wondered why the supernatant was so viscous. The lipid was largely removed, but I wondered about nucleic acids: DNA and RNA. So Michael ran an agarose gel and stained for nucleic acid with ethidium bromide. The first impression is that there is high molecular weight DNA (top of the gel) and a smear of RNA (which at this stage I expect might be rRNA, but this needs further work). Since our starting material was freeze dried,  the high molecular weight of the DNA is a bonus (it is indicative of minimal degradation), but the RNA looks to be degraded. The salt in the samples is probably the reason for the "half smile" of the fuzzy RNA bands. Lanes 1, 2 and 3 contain different amounts of sample, nonetheless, it is clear that the supernatant contains significant amounts of nucleic acids as well as protein. This now presents us with an opportunity to use the mealworms for a wide range of experiments in proteomics and genomics.

We have systematically begun to investigate the potential of the mealworm, Tenebrio molitor as our model organism. I believe the results are very encouraging, but I will now look at similar experiments in Drosophila and mosquito for comparisons and methods. The most important aspect of these experiments, is the approach to research in which an experiment is carried out, the results evaluated and experience and knowledge brought to bear on the outcome. The follow up experiments we carried out begin to bring insight into the investigation, but the following questions remain. Why is the proteome so limited? Do we have levels of degradation as a result of using lyophilised material? What is the origin of the brown colouration? The spectroscopy measurements may give us a clue. These are the questions we shall investigate in the coming weeks. 

The message from this is that we failed to obtain high yields of a diversity of proteins from the larvae, but it may be that our methodology is inefficient or our source material is degraded (or prone to degradation through the day). It is also possible that the protein content of the larvae is limited and the high molecular weight band is a storage protein with some developmental role. The unexpected outcome is that the larvae are a good source of genomic DNA, which opens the door to Molecular Biology based studies. This is an example of serendipity, which I shall discuss in due course.