Name: Kelly McGowan
Position: Postdoctoral fellow in genetics, Stanford University, April, 2008 to current
Background: PhD in genetics, Stanford University, 1999 to 2008; resident in dermatology, Stanford University, 1996 to 1999
The wild-type protein p53 is often referred as the “guardian of the genome” for its role in suppressing tumors. But in mutant form p53 can contribute to cancer, and too much of it in the body is known to accelerate the aging process.
Studying the role of the protein in pigmentation in mice, a team of researchers, including those at Stanford University, stumbled across findings suggesting that p53 responds to disruptions in a cell’s protein factories and may be implicated in disease pathways in ways yet to be understood, particularly as the protein relates to Diamond-Blackfan syndrome, in which a patient’s bone marrow is unable to produce sufficient amounts of red blood cells.
The work is described in an article appearing in the July 20 online edition of Nature Genetics
, available here
ProteoMonitor spoke with Kelly McGowan, the first author on the study, this week. Below is an edited version of the interview.
What led you and your colleagues to do this study? I’m guessing you suspected that p53 had a role in disease pathways even if you weren’t sure what that role is?
The project started as a part of a large screen looking basically for mice with defects in pigmentation. The [laboratory of Gregory Barsh, a professor of genetics and pediatrics at Stanford, and a co-author on the Nature Genetics study] has had a long interest in using pigmentation as a model to study different genes not only involved in pigmentation and color, but through the years Greg has found that a lot of the pathways that are involved in pigmentation are also involved in other processes throughout the organism.
In particular, he’s used mice for that purpose. Basically, our results were sort of a byproduct of an investigation into the genes involved causing these mice to have dark feet or dark skin. It’s a little bit fortuitous in a way.
So it’s an accident that you made this finding?
Yes. These [mice] are mutant animals and … mice have been treated with a chemical that [allow you to] look at animals that come from a large screen and screen them for a variety of different phenotypes. We were interested in hair and skin color. Other people were interested in, for example, kidney function or other things like behavioral defects.
Basically, we were trying to identify genes involved in a specific phenotype of interest.
What would you say is the main message of the study?
I think we, using pigmentation, have been able to identify molecules, in particular p53, that’s a critical player in a human disease, and that [Diamond-Blackfan syndrome] is basically a bone marrow failure. There are patients who have mutations in these ribosomal proteins. These are the genes that we identified that are mutated in these dark-skinned mutants.
And patients with ribosomal mutations also have, like the mice, anemia and problems with their bone marrow. And so one of the take-home messages is that you may see phenotypes in the skin, but this pathway is also active in other organ systems where it’s causing other phenotypes.
What other organ systems? What other disease pathways?
The skin and the bone marrow are the two primary ones that we’ve looked at, but we anticipate this pathway, because ribosomal subunits are ubiquitous — they’re everywhere in every tissue — we anticipate there are probably other phenotypes that we haven’t examined yet.
I don’t anticipate that this is going to be specific to the skin and the bone marrow. These mice, for example, also are smaller than their litter mates or the controls, and that’s actually seen in the human disease, as well.
So patients with Diamond-Blackfan anemia, who have this mutation in ribosomal proteins, also have growth retardation and are small. It kind of mimics disease in many ways.
In follow-up studies, are you looking at other organ systems and pathways?
Because of the link with the bone marrow failure, that’s probably going to be our main focus, to try to explore the role of p53 in this disease, and then potentially try to identify targets for therapy. That’s the ultimate goal, to see if we can modulate that pathway to potentially bring something to help the patients.
Is it purely the activation of p53 that creates this disease pathway, or does it work in concert with certain other molecules to trigger disease?
Basically what we did was take our mutants in which we have a mutation in a ribosomal protein and genetically removed p53. We got rid of the phenotype completely, and basically reversed the anemia phenotype.
We think it’s a necessary component. I think there are clearly downstream effectors that are causing the anemia, so p53 is known to do more kind of classic things like cause apoptosis or cell death or slow down the ability of the cell to proliferate. My guess would be that p53 is probably affecting one of those processes in bone marrow, but we haven’t specifically looked at what those downstream targets are yet.
There are clearly molecules downstream of p53 that are important.
Are you going to be looking at some of these molecules?
Yes, that’s the goal. The model that we have right now has a fairly mild anemia phenotype, so one of the ideas that we have is to try to strengthen or make that phenotype a little more robust so that we can actually look at downstream effectors with a greater sensitivity.
Have you identified any kind of candidate or suspect downstream molecules?
The only thing we know right now is that this pathway affects or causes increased cell death or apoptosis in bone marrow progenitor cells. That’s really the only data point that we have right now. There are some very good candidates, some of the very classic molecules that p53 triggers, things in the apoptosis pathways: things like Bax … or Perp, these are classic p53-regulated genes known to be involved in apoptosis.
So I think we have some very good candidates that we can start with.
Has there been prior research that has looked at p53 and the role it plays in Diamond-Blackfan syndrome?
There is one paper that just came out and that’s in a zebrafish model … where they used a method which is very common in zebrafish to knock down ribosomal proteins in zebrafish. Basically, some of the phenotypes they showed in their model were reversed when you also remove p53.
So it’s a similar pathway. They didn’t look at the blood or blood development in the zebrafish, but there have been other links to ribosomal biology and p53.
Is p53 a biomarker of any sort, or is it too early to say?
I think probably it’s too early to say. I think you’d … basically need to isolate human patient cells to look at that.
The study says this finding opens up new avenues for diagnostics and treatments. What do you have in mind?
I think finding the downstream targets of p53 and modulating them with drugs is probably the way that you would approach therapy at this point for human patients.
We’re not talking about silencing the entire molecule then?
No, no, no. I think that that would be a tricky proposition. … We know that in this setting that p53 is stabilized or more active than in a wild-type animal and you would have to have a very fine-tuned balance to regulate the levels of p53. If you brought it down too much, you could put patients at risk for cancer.
I think you need downstream targets to really effectively go after therapeutics.
Any data from your follow-up work that you can share?
We’re just starting, so we’re pretty excited to explore the role of p53 not only in this disease. This disease has a variety of phenotypes. Some patients have very severe anemia while other patients with the same mutation might have a very mild phenotype.
One of the things that we’re exploring right now is why that happens. One of the possibilities is that p53 is controlled by other pathways in the genome in different ways. Someone who might have a very severe phenotype might have a very robust activation p53, but someone who might have a more mild phenotype might have a less severe [activation].
We’re interested in developing a model of that to understand why you get such a varied expression of the phenotype.
Does Diamond-Blackfan have a profile that would explain these wide differences in phenotypes?
Well, no. It’s completely perplexing, and I think that that’s one of the intriguing aspects to this study. With this model, we can now try to explore what other things are feeding into that pathway.