Sci has recently did a post on p53. She finds it to be a fascinating little guy, and it might just become her new obsession. It appears to be everywhere, the little protein behind the scenes making things happen.
(Awww, look at it, all hugging the DNA like that. What a cutie!)
And one of the things that it’s involved in is cancer. And Sci will explain part of why that is in a moment.
Feng et al. “Declining p53 function in the aging process: a possible mechanism for the increased tumor incidence in older populations” PNAS, 2007.
Before we start, Sci would like to note that PNAS papers are often GREAT for blogging. Good work, interesting ideas, and in nice, digestable chunks of data. No major monoliths. Sci’s not saying one should always use the least publishable unit, but when it comes to blogging, a simple elegant story is often the one Sci will pick.
So we know that p53 is involved in cancer. Why is this, you ask?
Because of what p53 has to do with the cell cycle.
Here you can see the cell cycle. p53 plays an important role here, in the “R” area between the G1 phase and the S phase, when the cell has to decide whether it’s going to go ahead with the replication process or not.
How does that happen, you ask? Well, it’s complicated, but Sci will try her best. When p53 is activated by things like stress, it can activate DNA repair proteins to repair possibly damaged DNA. It can ALSO activate a protein called p21, which can stop the cell cycle entirely (the relationship between p53 and p21 is a very tightly controlled one and has lots of interesting things associated with it). In fact, p53 is so important in stopping the unregulated growth of cells that it’s called a “tumor suppressor” gene. It’s also such a big deal that more than 50{9f43b4361d9a125bc126dd2a2d1949be02545ec69880430bc4fed2272fd72da3} of tumors in humans have problems with p53 function.
What this basically means (as far as Sci can determine, and she’s very new to this area) is that high levels of p53 are the sign of a stressed cell, but LOW levels of p53 (low than normal, good healthy cells have low levels of p53, but we’re talking lower than that) are a sign of a cell that cannot necessarily respond correctly to stress. By stress, Sci doesn’t mean cell exams, rather she means things like exposure to radiation, oxidative stress, things that cause a cell to stay up and worry at night.
So now we know a bit about p53, let’s think about p53, aging, and cancer. Cancer is sometimes described as a “disease of aging”. Obviously people can get cancer at any age, but most people start getting tumors in the last quarter or so of their lives. For humans, that’s after about age 50-55 (given a 75 year median lifespan), for mice, that about 1.5-2 years old (at best mice live about 3 years. This is news to Sci who has seen mice at about 18 months and they look OLD. Usually also fat and cuddly, but OLD).
Now obviously we know that p53 is a tumor-suppressor gene, and we know that cancer most often occurs in the elderly, and we know that p53 is involved in more than 50{9f43b4361d9a125bc126dd2a2d1949be02545ec69880430bc4fed2272fd72da3} of cancers, often in mutations that render p53 ineffective. This leads us to the rather obvious question (often, the most important questions are the most obvious) of whether p53, or its ability to function, decreases with age, making older animals more susceptible to problems with p53 and thus more susceptible to cancers. Decreased p53 with age would make any mutations that make p53 less effective that much more problematic and that much more likely to result in cancers.
So what these authors did was to take younger mice (about 6 months old, definitely a full adult, but not an older one), and older mice (20-28 months old), and compare their p53. They were interested not in the levels of p53 (which in healthy cells are normally pretty low), but in how p53 responded to stressors in these mice. So they took the mice, exposed them to irradiation (which is a pretty stressful event to a cell), and then looked at levels of p53, and ALSO also the things that p53 controls, including another transcription factor, p21, as well as other targets like Fas, MDM2 (which sounds like a band!), and Cyclin G1. And here’s what they got (Sci is presenting the figures that she wants to show entirely out of order from the way the paper did it. That’s the presentation that worked best for them, but Sci’s using the one that she thinks will make the most sense to you).
So here you can see the levels of p53 that were induced by irradiation. Under normal conditions, p53 doesn’t stay around a long time in the cell, it’s rapidly degraded, and so levels are low. Following a stress like irradiation, though, p53 becomes more stable, and so more of it accumulates in the cell, meaning there’s more around to function. The gray bars are the control, the black bars are the irradiated mice, and along the bottom you can see the different ages of mice they tested. You can see that the 6 month old mice (left set of bars) had a massive induction of p53, showing a good response to irradiation (a high p53 response would help arrest the cell cycle and check for DNA damage). But the OLDER mice (Two right sets of bars) had much lower, rather pathetic looking p53 responses. Well, that looks icky.
And what about how effective p53 was at getting its job done? Well, obviously if there’s less p53 it’s going to be less effective. The authors here looked at two measures, the downstream effects of p53 (like increases in p21), and apoptosis.
So above you can see the values for p21 expression, induced by p53 when cells are exposed to radiation. Sci had to graph this one up herself (TABLES! WHAT IS WITH THE TABLES! Lovely data, dudes, please give me a cool pic to go with it. But that’s ok. Sci graphs because she loves, and at least this gives her the opportunity to graph in pretty colors instead of black and white).
And you can see that the levels of p21 induced by p53 are a lot lower in older mice (the dark purple) in all the tissues measured (oh also, authors, I LOVE this paper. It’s an elegant story and beautifully told. But please include your stats. It’s not that hard. As it was, I ran your stats for you, and the results were beautifully significant). So the low levels of p53 are having major effects on what p53 does in the body, as measured by the proteins that p53 induces.
The authors also looked at another measure of p53 activity. I’m sure you all know that one of the features (eh, it’s a bug, it’s a feature) of tumor cells is that they can divide entirely without restraint. One of the cool things about p53 that makes it a tumor-suppressor protein is that it can induce the apoptosis of screwed-up cells. So the scientists measured the amount of apoptosis in the mouse’s cells after irradiation, which should cause some major screwed up cells. Therefore, if p53 is working correctly, a lot of those cells should be programmed to self destruct.
So here you can see control groups of cells (on the left), and irradiated cells (on the right). What you are looking for is cells stained red, which indicate cells going through apoptosis.
And check out the colors! You can see that the top two panels, which is a knockout for p53, obviously had no apoptosis, because they have no p53 to induce it. The second set down in the younger mice at 6 months old, and they have a nice hefty pattern of staining. The third and fourth sets down are older groups of mice, and you can see that the apoptosis has decreased a LOT, suggesting that p53 is not doing its job there. And since they then showed lower levels of p53 and lower levels of proteins that are induced by p53, it looks like a pretty tight story.
Finally, the authors looked further upstream from p53, at a protein called ATM kinase. This is a protein that is directly activated by DNA damage, and which then goes on to activate p53, making it more stable in the cell and causing it to accumulate.
You can see here levels of mRNA for ATM kinase. The grey bars are the youngest mice, the black are 20 months old, and the white are 28 months old. You can see that levels of ATM kinase mRNA declined a LOT as the mice aged. This implies that ATM kinase might be the thing decreases, which then decreases the ability of p53 to stay around, which THEN stops p53 from being effective because there is less of it.
The authors also checked (Sci bets this was a reviewer request), to see if these p53 changes could be induced in response to other types of stress, like Taxol or actinomycin D, and it does indeed appear to be a general effect of cellular stress.
So the conclusions? p53 function declines with age, and this might be due to decreases in ATM kinase (though Sci bets it might even be further upstream than that). These decreases in p53 function result in an inability to respond to cellular stress like irradiation (as shown in the paper), and so might play a big role in the increase in tumors that occur with age. Obviously, this isn’t the only answer, but it’s another piece in the puzzle of how cancer may develop, and another target for possible prevention.
Feng, Z., Hu, W., Teresky, A., Hernando, E., Cordon-Cardo, C., & Levine, A. (2007). Declining p53 function in the aging process: A possible mechanism for the increased tumor incidence in older populations Proceedings of the National Academy of Sciences, 104 (42), 16633-16638 DOI: 10.1073/pnas.0708043104
