Monday, July 22, 2013

Zac - How do we know what we are doing?

One of the biggest questions in scientific research is - how do we know that our results are reliable?  This is one of the most important questions that a young budding scientist can ask - and he/she should ask it frequently. So how do we know? 

For example, my work in the Maslakova Lab involves injecting cells with morpholinos and/or fluorescent markers.  If we see changes in the morphology of the cell and/or the development of the resulting embryo, how do we know that they are due to to the morpholinos?  Could the effects that we see possibly be due to the fact that we just stuck a great big needle into these cells and injected them with a foreign liquid?  How do we know that the morpholinos are not causing some non-specific cell toxicity? Could it be possible that the morpholinos we are injecting - which are supposed to knock down the production of a single protein with high specificity - are also halting the production of other proteins unrelated to our target of interest?

These are very important questions that we must continually ask ourselves, and if we are ever to draw any conclusions from the results we see, we must prove without a doubt that the changes we see are a result of the things that we say they are.  Herein lies the importance of including an experimental control: a series of experiments of a known outcome that demonstrates the reliability of one's materials and methods.

In the context of our experiments using morpholino injections it is extremely important to eliminate any possibility that the morpholinos are doing anything except what they should be doing: knocking down the production one (and only one) protein. As a positive control we injected morpholinos designed to knock out the production of Beta-catenin. This experiment has been done previously (by Henry et al 2008) and it produced a known phenotype in the larvae of a related nemertean: Cerebratulus lacteus. By performing the same experiment on Micrura alaskensis we demonstrate not only that morpholinos have an effect in M. alaskensis, but also that the effect is the same as previously demonstrated in C. lacteus. As a negative control we intend to inject a controlled morpholino sequence designed to do absolutely nothing. If we inject this control morpholino and subsequently see anything other than a perfectly normal looking embryo we will know that we ought to think twice about the specificity of our other morpholino injections.  

Another experiment that we intend to perform - the gold standard for demonstrating that the morpholino we inject is targeting the gene that we think it is targeting - is referred to as an mRNA rescue. Remember: morpholinos deplete a growing embryo of a specific protein by blocking the translation of that protein from the endogenous mRNA within the cell.  If one were to inject a morpholino against a particular gene and simultaneously inject a solution of mRNA coding for that exact protein - but modified ever so slightly as to render it immune to morpholino treatment - one would expect that the developing embryo would no longer display the phenotype expected from the injection of that morpholino.  This technique, in concert with various staining and imaging techniques, can confirm within a shadow of a doubt that our morpholinos are depleting the proteins that we say they are depleting.  

This coming week will largely be consumed with staining fixed larvae with antibodies and other fluorescent labels specific to certain tissue types.  These techniques will allow us to see some of the effects of these morpholinos otherwise invisible with light microscopy.  Stay tuned for some fluorescence microscopy pictures!

1 comment:

  1. Great question you are posing here, glad they are training you to think this way!!

    ReplyDelete