Interview with Professor Charles Perou (University of North Carolina School of Medicine, USA)
Next generation gene sequencing
The session I was involved with was focussed on next generation sequencing and whole genome sequencing and its application to breast cancer, of course. And my talk in particular was focussed on using sequencing technology for studying gene expression – how that compares and contrasts with micro array studies on human breast cancer specimens and really focussed on the additional information that we get from the sequencing versus DNA micro arrays.
What applications does this have?
Well I think it’s useful in that by using the sequencing we’re able to get a huge amount of information above and beyond what we can get with DNA micro arrays. So we get the same type of information that we get from micro arrays, which is a quantitative measure of gene expression now for all human genes and all non-coding RNAs and other RNA species that we can’t see by DNA micro arrays we can now get with this technology.
And so the hope is that by having this broader net and this more complete net we can identify better prognostic and predictive expression signatures. So as I alluded to, in addition to the gene expression quantitative levels we can also get sequence information. So from the gene expression data we can now also get splicing information and I gave examples of important alternative splices between different sub-types of breast cancer. So in one instance it looked, for example, that in breast cancer cell lines we could see that the CD44 gene, which is an important gene for some of the cancer stem cell experiments, that in one group of breast cancers they express the whole gene and in another group they were missing three exons, which seems like it might have important functional consequences. So I think by the sequencing technologies, these are the types of new discoveries that will be made and we’ll see if they are of clinical importance. I think some of them will be; some of them won’t be of clinical importance, they’re all of biologic importance, but now we can begin to see things we couldn’t even try and look at before.
Another important feature of the sequencing approach for gene expression as well is the ability to see mutations. So not only can we see the gene is expressed but now we can see if it has a sequence variant. And so in my talk I gave examples of being able to determine P53 mutation status in these tumours; we could also see there was a rare AKT3 mutation and we could also see the presence of these, what we call, fusion proteins. So in this particular case a region of the genome was deleted and the two genes at the boundary of the deletion got put together and actually you could see the transcript that was this now hybrid protein. Now, whether that’s of biological or clinical importance or it was just a coincidence will take some functional studies to figure out but now the excitement is we can see all these very interesting events and begin to sort out whether they’re causative of cancer or not.
What sort of timescale are we looking at?
We’re certainly using them in the research laboratory today. Now when they might make it to the clinical setting, it will probably take a year or two or three, in part because of the computational challenges that we now must address. So, as I pointed out in our session, it’s now easier to generate the data than it is to analyse it and the data is so rich in information content that the analysis can take many months. So we’re now in this phase where we can, with relative ease, generate the data and with not relative ease analyse the data. So I would imagine in a year or two it will be more automated and straightforward to analyse the data and then it will probably be more clinically applicable.
Do you need better computers or more people?
All of the above. So we, and many others, as we set up the sequencing facilities and machines we’re simultaneously beefing up the computing infrastructure. We need more computers; we need a higher bandwidth to get the data from one place to another and then once it gets there we need additional people to develop the analysis programmes and then yet more people to actually mine the data for the biologic insights. And so we’re all ramping up and, again, I think in a year or so we will have ramped up and certain techniques will have become more established and then we’ll be able to do higher numbers of samples; look for the correlations with clinical response to important drugs or prognosis and that’s when this new technology may make the clinical impact that we think that it will have.
Why should a patient be interested in this research?
As with all these new technologies, in the short term the patient might not have an immediate benefit for her or him, but certainly they’re contributing to the knowledge base that will lay the foundation for the next generation of clinical discoveries to be made.
And depending on that patient, how much information that patient would like to receive, this new technology gives us more information than we’ve ever had before and so it’s going to tell us more about the biology of that tumour and certainly that would likely be of interest to the patient and to the physician. Now, whether you can act upon that information is a different story and the answer is today we probably can’t act upon these new discoveries because we don’t, in fact, know what they mean and how to act upon them. But that will come with time.