darwin's tumor

May 18, 2006 7:22 PM

Ive read that cancer can be pretty clever - securing nutrition by releasing hormones which cause the creation of blood vessels, employing various methods of fooling our immune system etc. All these seem (at least to my layman's mind) to be pretty specialized eh... 'behaviors' i guess is the word. My question is, how would they evolve?

people with genes for really clever tumors would be much less likely to pass on those genes, and its not as if the cancer can reproduce (aside from metastasizing within the same body).

are all these behaviors that human tissue regularly carry out which would naturally be employed by a tumor?

(for the record, i don't buy into the ID crap. Id love to hear some fundy nut job suggest the god purposefully designed cancer just to give humanity the middle finger)

posted by Tryptophan-5ht to science & nature (22 comments total)

"Cancer" is just regular cells that have gone bad. They have lost the ability to stop from multiplying, so they grow unchecked, which is not what they're supposed to do. Imagine if the cells which make up your finger ceased to know when to stop growing; your finger would simply change into an ever-growing bulbous mass.

Wikipedia, as usual, has a good explanation of the process.

Nothing to do with evolution. Or intelligent design.
posted by jellicle at 7:34 PM on May 18, 2006

cancer does not evolve, it devolves. Some of what has been turned off is instructions to stop growing.

Cancer does not trick blood vessels into growing to it. Growing cancer is growing tissue and blood vessels naturally grow to growing tissue.
posted by MonkeySaltedNuts at 7:34 PM on May 18, 2006

people with genes for really clever tumors would be much less likely to pass on those genes, and its not as if the cancer can reproduce (aside from metastasizing within the same body).


Consider this, though: many people only develop cancer late in life, so it is perhaps incorrect to suppose that all forms and genetic dispositions would be selected against.
posted by clockzero at 7:35 PM on May 18, 2006

i understand what cancer is and the basics of how it happens (uncontrolled replication) but this business of fooling the immune system doesn't seem to jive with 'random fuck up during mitosis'.
posted by Tryptophan-5ht at 7:39 PM on May 18, 2006

Fooling the immune system is basically the same as "being regular cells", which they more or less are for these purposes. It's said a normal person may have multiple cancers through their life that never kill them, and are never detected, and sometimes go away since the immune system manages to recognize the normal cells from the cancerous ones.
posted by kcm at 7:45 PM on May 18, 2006

.. here, read about Interleukin-2, it may be of interest. It is used in some cancer treatments. Also, reverse transcriptase along those lines.
posted by kcm at 7:49 PM on May 18, 2006

OK, a little more - above, RT-PCR is used to develop experimental vaccines in some cases. This is done by taking the patient's own defects, as it were, and using it to create a "fix" for the defect. Also, you are on the right track, although it's more that some people have certain defects that allow the mutations to occur more often/easily/less detectedly. An example is BRCA (1 and 2).

I'm not even close to a doctor so I probably screwed some terminology and facts up, but you can read too. :)
posted by kcm at 7:54 PM on May 18, 2006

Carlo Maley of the Wistar Institute is using evolutionary biology to study how initial treatment evolves cancer cells resilient to future treatments.

While this doesn't seem to answer your initial question at first, keep in mind that cancer cells develop into tumors by surviving apoptosis and immunological responses, both mechanisms the body has evolved to control cells with broken division cycles. In a sense, oncogenes might have evolved to overcome natural defenses (just as those defenses have evolved to control the effect of oncogenes).

My own thought is that cancer could have co-evolved with sexual reproduction, which, along with senescence, offers another mechanism to kill off organisms after they've past the age where they will have likely reproduced. (The flip side of making sex work as a way to introduce better mutations into a population is that the parents have to die, eventually, and take themselves out of the gene pool.)
posted by Mr. Six at 7:59 PM on May 18, 2006

yeah, I messed up, RT-PCR is just for detection. The vaccines (peptides?) are based off making antigens more recognizable.. I think. :)
posted by kcm at 8:00 PM on May 18, 2006

Cancer does not trick blood vessels into growing to it. Growing cancer is growing tissue and blood vessels naturally grow to growing tissue.

Cancerous tissue does indeed release abnormally high levels of vascular endothelial growth factor (VEGF), a growth factor that promotes blood vessel development.

Microarrays and quantitative PCR show that VEGF genes are upregulated (overexpressed) in cells with abnormal oncogenes, compared with healthy tissue. VEGF inhibitors are being used to cut off blood vessel growth, starving the cancer tissue.
posted by Mr. Six at 8:09 PM on May 18, 2006

Cancer cells are basically deranged normal cells. To oversimplify, they don't really do new things that normal cells never could, they just lose the inhibitions that prevent them from doing things they shouldn't. Like growing uncontrollably, which is their most obvious craziness.

What makes a cancer cell a cancer cell is when it loses its differentiation (i.e., forgets it's a mature, well-regulated grownup) and starts acting like an undifferentiated progenitor or stem cell (i.e. behaves like a wild uncontrollable 2-year old).

For example, a differentiated (mature) epithelial cell has these characteristics: it stops growing when it touches other cells (called "contact inhibition"), it has a characteristic morphology due to a well-defined cytoskeleton, it makes certain proteins and only those proteins, etc. When it freaks out and turns into a cancer cell it loses contact inhibition, its morphology is borked because it's not making and assembling the cytoskeletal proteins correctly, and it starts making other proteins that don't belong in a self-respecting epithelial cell. (Interestingly, some of those proteins that shouldn't be there are proteases on the cell membrane, which allow metastatic cells to chew their way into other tissues and colonize other parts of the body). But the loss of contact inhibition is the most dramatically obvious - this is why tumors keep growing and growing, while normal cells form a lung or whatever, and then stop.

Cancer cells do the same things that happen in normal cells during various stages of embryonic development, childhood growth, wound healing and normal adult life - they just do them in the wrong time, place and order.

As far as fooling the immune system, it's not too surprising since the cancer cell looks a lot like "self", at least compared to a bacterium or other really foreign cell. The immune system does recognize and kill a certain number of tumor cells (using antibodies and NK cells) but it's tricky since cancer cells have the wrong combination of "self" proteins on their surface, not some bizarro totally alien antigens. (I'm not an immunologist, so apologies if I've mis-represented something in this paragraph.)

On preview: or, what everybody else said.
posted by Quietgal at 8:12 PM on May 18, 2006 [1 favorite has favorites]

Hmm, should have said "loss of contact inhibition is one reason tumors keep growing and growing ..."
posted by Quietgal at 8:16 PM on May 18, 2006

kcm - the vaccines are for stimulating an immune response against antigens expressed by the cancer cells and not healthy cells.

Quietgal - cancer cells often express truncated or fused versions of normal proteins, constitutively active Ras being an example. This is what the antibodies go for.

Tryptophan - you're right that some of the behaviors are fairly sophisticated. There are two things going on. First, as clockzero mentioned, people who have cancer may have already reproduced, making those genes fitness-neutral. Second, cancer cells do things that normal cells of the tissue from which the cancer came don't, and those things can be selected for as the cancer proliferates.

If a cell of the immune system, for example, gets its an antibody-making promoter in front of a survival-promoting gene, that gene will start to be pumped out in high doses, and that creates the potential for cancer. Once the cells have started to divide uncontrollably, natural selection suggests that the progeny cancer cells that happen to express angiogenic genes, for example, will divide at a faster rate and the resulting tumor will be composed largely of those cells.

The sophisticated behaviors may happen pre-cancer, and are required for the cancer to develop, but the behaviors may also arise by natural selection as the cancer divides. Both processes can and do occur together.
posted by Mr. Gunn at 8:25 PM on May 18, 2006

Some cancers are caused by viruses. Perhaps there's something about the cancers that is beneficial to the virus?
posted by frogan at 8:50 PM on May 18, 2006

This is my field; way too much info out there for a MeFi post.

On evolution: it is well known that, while tumor is a clonal proliferation of cells derived from a single cell, there is not a homogenous population of genetically identical cells. There is evolution, in a sense, within a tumor over time. Genetic mutations accumulate (remember, we are already dealing with a system with a derangement in the processes which control random mutation in normal cells) over time, conferring selective growth advantages to a certain group or groups of tumor cells. Those with selective growth advantages will have a survival advantage over those who don't, therefore, their progeny will likely be more predominant than other tumor cell groups.

Not to mention, that between different types of cancer, the mechanisms for mutation and proliferation can be wildly different. It can be quite difficult to generalize these concepts for all tumors.
posted by i_am_a_Jedi at 8:53 PM on May 18, 2006

An interesting snippet in Nature demonstrated that increasing the expression of the p53 tumour suppressor genes can lead to accelerated ageing in lab mice. This is an interesting result. The conventional answer as to why cancers tends to increase in frequency as we age relies on the fact that much of the increase happens after reproductive maturity, and the notion thus arises that late-stage cancers tend to arise because there is no strong selection pressure to remove them. They are thus seen as merely incidental.

This snippet instead suggests that at least some of the processes that lead to cancer are part of the body's defences against premature ageing. Some cancers may be simply the price we pay for living so damn long compared to many animals. This also problematizes some of the putative treatments for organ replacment or repair that involve adding lots of new stem cells, many of which have a higher probability to become cancerous than other body cell types.

I think a good parallel is with gout, where an unusually high serum level of urate crystals (approaching the solubility limit) leads to inevitable deposition of condensed crystals in joints and kidneys, leading to gout and kidney stones. Why have levels so high? Primates that diverged dozens of millions of years ago from the human lineage have much lower urate serum levels, but they don't live as long. It turns out that urate is a highly potent scavenger of reactive oxygen. Thus gout is an evolved price we pay for extending our life spans and reducing the probability of genetic mutations that would lead to tumour growth.

p53 mutant mice that display early ageing-associated phenotypes
For both p53+/m and pL53 transgenic models, we propose that the early ageing phenotypes are in part a result of enhanced activity of wild-type p53 in some tissues. The reduced cellularity and mass in organs of the older p53+/m mice suggests that organ cell numbers are not maintained. Moreover, some of the ageing phenotypes suggest a reduction in proliferation of stem cells. With the ageing process, this proliferative reserve may decline more rapidly in the p53+/m mice as their stem cells undergo replicative senescence sooner than their p53+/+ counterparts. The accumulation of genetic insults in the stem cells of p53+/m mice may provoke enhanced arrest responses that gradually result in fewer division-competent cells. The p53+/m mice eventually reach a point in which the proliferative capacity of stem cells is so reduced that sufficient numbers of mature cells cannot be provided to maintain organ homeostasis. The resulting phenotypes may include reductions in organ mass, function and tolerance for stress.
posted by meehawl at 9:02 PM on May 18, 2006

Great stuff here.
I also highly suggesting reading The Hallmarks of Cancer. It's a great primer those interested in knowing more about cancer and why cancer is what it is.
posted by jmd82 at 9:14 PM on May 18, 2006

re: evolution. This depends on what you consider the unit of selection. Cancer in and of itself is not an organism. Also, many cancers aren't inherited--selection acts on heritable traits. For example if I expose myself to carcinogens (e.g. tan & get melanomas, smoke & get lung cancer), that's not heritable (only possible if the mutation that causes the cancer is in a germ cell). Or if I contract a virus that causes cancer, (e.g. if i have HPV & get cervical cancer), that can't be passed on to future offspring genetically.

If I incur a mutation that causes cancer in a somatic cell, I can't pass that cancer on to my children. Like if my arm gets cut off, I can't pass that event on genetically. Though it's unfavorable, evolution isn't acting on an "one arm only" trait--all the gametes I make will code for two-armed babies. However, if that mutation is present in a germ cell, it can get passed on to my offspring, predisposing them to cancer. Normally it's only one of the two copies of the gene with the mutation, and it's usually recessive (a tumor suppressor gene). If it were dominant (an oncogene) or there were two copies inherited, the offspring often wouldn't stand a chance to make it to a functional embryo. (exception) So yes, when mutated genes that lead to cancer are inherited, you do see the effects of selection against the detrimental trait--like when the "offspring" doesn't even make it to a fetus, let alone survive to reproduce. And people who inherit a genetic predisposition to cancer often have earlier age of onset that the rest of the population, cutting into the timespan of reproductive fitness.

Of course, this response is predicated on thinking of the individual as the unit of selection. Looking at the survival of cancer cells over other cells gives a different perspective and a different answer. Though I remember reading that the original source of protooncogenes (functional genes that when mutated several times can lead to cancer) were from a portion of viral genome inserted into the genome of an ancestor (way way back in the evolutionary line). Which would mean that the action of tumors do originate from another organism, in a way.
posted by neda at 10:10 PM on May 18, 2006

wow. lots of good answers here.

this in particular was helpful:
"it is well known that, while tumor is a clonal proliferation of cells derived from a single cell, there is not a homogenous population of genetically identical cells. There is evolution, in a sense, within a tumor over time. Genetic mutations accumulate (remember, we are already dealing with a system with a derangement in the processes which control random mutation in normal cells) over time, conferring selective growth advantages to a certain group or groups of tumor cells. Those with selective growth advantages will have a survival advantage over those who don't, therefore, their progeny will likely be more predominant than other tumor cell groups."
posted by Tryptophan-5ht at 1:09 AM on May 19, 2006

According to a recent story, at least a couple of cancers have made it to the big time: they have become independent organisms in the wild in their own right. Of course, this has become almost routine in the laboratory; monoclonal antibodies are produced by hyridomas which are capable of living and reproducing indefinitely in that peculiar habitat, and there are many cancers which are transferred from individual to individual within inbred mouse lines.

I suppose the probability this could happen in human beings must be very low, if only because we are so much more genetically diverse than lab mice or Tasmanian Devils, but if I were a pathologist with time on my hands, some grant money, and a taste for extreme long shots, I'd be tempted to take a look at uterine and cervical cancers. After all, the uterus is designed to foster the growth of a mass of cells with alien genes and alien antigens.
posted by jamjam at 8:23 PM on May 20, 2006

Jamjam, that link goes to /null.
posted by metaculpa at 5:51 PM on May 28, 2006

story
posted by jamjam at 5:16 PM on May 29, 2006