Two-faced proteins?

Proteins that both hinder and spur cancer progression may not be as uncommon as previously thought

By | January 11, 2011

Cancer researchers are identifying an increasing number of proteins that have a dual nature when it comes to cancer -- they may initially promote the development of tumors, but in the long run make them less aggressive, or vice versa.
Image: Wikimedia commons
"There's just so many different things going on [in cancer] that it's not at all surprising to me that you come up with examples of proteins that hurt and help," cell biologist linkurl:James Bear; of the Lineberger Comprehensive Cancer Center at the University of North Carolina School of Medicine told The Scientist. NEDD9, for example, is a scaffolding protein that supports tumor cell invasion and metastasis in many types of cancer. Accordingly, mice lacking the NEDD9 protein show delayed development of mammary tumors in mouse models of the disease. Late last year, however, linkurl:Erica Golemis; of the Fox Chase Cancer Center and her colleagues showed that once tumors do develop in these mutant mice, they are hyper-aggressive compared to the tumors in mice with working copies of NEDD9. This is not hugely surprising, Golemis said, given the range of critical signaling pathways known to depend on NEDD9, such as those that control proliferation, migration, and the cell cycle. "This protein has got multiple functions [and] multiple partners," she said. "When you take it out, you're not going to be affecting any one of these pathways, you're going to be affecting all of them." Similarly, some researchers have demonstrated that the WAVE protein, which helps initiate the reorganization of the actin cytoskeleton, facilitates cancer, while other groups have looked at the same protein and found the exact opposite, Bear said -- that WAVE tends to inhibit tumor progression. In addition to having different effects at different time points in cancer progression, proteins may function differently depending on their level of activity. "I think it probably is a common thing," said mouse geneticist linkurl:Norman Sharpless,; also of The Lineberger Comprehensive Cancer Center at the UNC School of Medicine. "It stands to reason there are a lot of biological processes where there's sort of a sweet spot in terms of being a cancer -- too much activity not so good, too little activity not so good either." DNA repair proteins, for example, are necessary for tumor growth, as the cancer cells must still be able to replicate their genetic code, but a slight dysfunction in certain repair proteins could promote tumor development, as a cell must accrue a series of mutations to become tumorigenic. So some mutations to repair proteins will encourage cancer, but mutations that shut down those same proteins will also shut down cancer. "A modest decrease in DNA repair function leads to increased genomic instability and cancer, whereas more complete defect in DNA repair is toxic to proliferating tumors," Sharpless said. A complete loss of ATR, for example, which plays a role in sensing DNA damage and activating repair pathways, is often toxic to cells, but "reduced levels may be oncogenic," he said. The same story is true for DNA ligase IV, a critical component of the non-homologous end joining pathway that repairs double-strand breaks. Despite these examples, there are relatively few documented cases of individual proteins that both help and hurt cancer progression. One problem in identifying such two-faced proteins may stem from the fact that these opposing effects are rarely demonstrated in the same research paper. As a result, different methodologies often apply to the different findings, making direct comparisons next to impossible. Many studies investigate different tumor types, for example, said Bear, or employ different end points, such as tumor size, metastatic burden, or cellular proliferation. "They are all honestly a reflection of cancer progression, but it's difficult to compare sometimes between papers," he said. "You really have to have a very detailed understanding of exactly what the researchers are reporting." Furthermore, scientists say, there is little incentive to embark on the difficult task of investigating the effects of a particular protein under all possible conditions in search of such opposing effects, because such a complicated picture is often unattractive to journals and funding agencies.

Download Flash player to listen to a conversation on publishing complex biological research

Oncologist Erica Golemis, PLoS Biology editor Theodora Bloom, and former BMJ editor Richard Smith discuss the best way to present cutting edge science.
Golemis, for example, who is a fan of complex, highly nuanced stories, found that reviewers often are not. "We tried to put our data about the tumor-promoting effect of NEDD9 loss together with the tumor-slowing effect, and the reviewers hated it," she said. "They kept saying, well, then the gene can't be that important. The editors essentially told us we had to split the results." Thus, there may be many more two-faced proteins out there, but only one side of them has been revealed so far by research. "I'm guessing this kind of preference for a simple story may be hindering investigation of the issue," Golemis said. "It's probably going to be a more complicated picture."
**__Related stories:__***linkurl:A new epigenetic cancer;
[11th May 2009]*linkurl:Master protein for tumor growth;
[12th March 2008]*linkurl:A new way to suppress apoptosis?;
[9th February 2007]


Avatar of: Mike Waldrep

Mike Waldrep

Posts: 155

January 11, 2011

Avatar of: Kay-Uwe Wagner

Kay-Uwe Wagner

Posts: 1

January 11, 2011

"Proteins that may initially promote the development of tumors, but in the long run make them less aggressive" - this is not at all a new concept. For example, there as been a previous report on the role of Akt1 in ErbB2-induced mammary cancer. Its overexpression accelerates tumor formation, but it also inhibits metastasis at the same time. Another example is TGF-beta, which restricts growth of normal cells prior to neoplastic transformation, but it is suggested to promote cancer growth, invasion, and metastasis, i.e. the exact opposite. Why is it so difficult for some people to believe that signaling networks change during cancer progression? We have also shown in two papers in 2009 and 2010 that signaling proteins such as the Janus kinase 2 (Janus, the god with the two faces, how appropriate for this article) are necessary for mammary tumor initiation, but they are no longer required for cancer growth. Sorry, but with all that evidence out there, I don't see anything new in this report or the disputed concept.\n\nBy the way, I now had a quick look at the NEDD9 paper on mammary tumorigenesis published in 2009 from the Fox Chase team. I am not involved in this research, nor have I reviewed this work. But looking at the M&M section, I can see that the knockout came from a C57 background, and the tumor incidence were compared +/+ versus -/- mice (Fig. 1). The PyMT transgene comes from FvB. So, if the authors compared the original PyMT mice that have wildtype (+/+) NEDD9 with a backcross of the PyMT into the knockout (-/-) that came from a C57 background, this is a BIG NO-NO because the C57 alleles will likely affect tumorigenesis!! So, at least for me, what was shown in that report is no clean evidence that NEDD9 has anything to do with cancer initiation. I suggest the authors or someone else with interest in that area should repeat that experiment in a pure FvB genetic background that is known to facilitate PyMT-induced tumorigenesis.
Avatar of: Nirmal Mishra

Nirmal Mishra

Posts: 22

January 11, 2011

There are several examples of proteins playing double and contrasting role. DNA polymerase can be cited as one.\nProf. Nirmal Kumar Mishra\nRetd. University Professor of Zoology, Patna University, Patna
Avatar of: Gil Lawton

Gil Lawton

Posts: 42

January 12, 2011

Isn't it beyond high time for the old "what does a gene code for" view, and "are genes selfish" view to be trashed? \n\nA biological organism is a complex system, even as a television set, an aircraft, a computer is a system. Would we be surprised to find that a particular key on a computer keyboard -- just because it is a "switch" doesn't predict what letter is printed on the screen. It is the sub-circuit that switch functions as a part of, that remains electrically open, or closed, that prints the letter. A capacitor does not, because it is a capacitor, determine what channel a television set projects on its screen. Almost every sub-circuit in a television utilizes capacitors.\n\nDo motor vehicle design engineers ask one another, "What one function is served by a fuse perform in an automobile?" No, because any circuit can contain, as a part of its overall system dynamic, a fuse.\n\nReason must guide rhetoric. Rhetoric must not be allowed to guide reason.\n\nDespite all the enormously useful and enlightening things that have come out of molecular research so far, it is still in its infancy. As Einstein put it, we have only "...lifted a corner of the veil and peeked under."\n\nIn a very practical sense, the role of molecular researchers today can be compared to that of a group of linguists who have learned of a language spoken on another planet. A verb form used in one context can indicate a particular meaning, tense, gender, number... but in another a different connotation and denotation entirely. In one syntax it can be flattering to a person on that planet, but in another it can be so vile an insult as to invite a fight to the death. \n\nThe only means of learning that language by a group of linguists communicating with the population of that planet would be by the process of immersion... that is, by just diving in somewhere and inter-communicating in it. Translating to our own language can be more difficult even than just learning (and accepting) that it doesn't work like our first-out-of-the-gate presuppositions about it.\n\nThe only access is by way of observing, analyzing, comparing, analyzing, describing... And,of course it's okay to try to extrapolate on basis of partial information aggregated to date. But when engaged in what is tantamount to LEARNING ENABLED ONLY BY REVERSE ENGINEERING, one can get bogged down in yesterday's model, terms, definitions, models that don't fit today's new data.\n\nWe have just enough information about what cells are, and do, what proteins code for, and how the result can serve (like those verb forms cited above) different purposes in different sub-circuits,that we have to be flexible enough to accept, and move on, beyond what were our prior translations of newly observed phenomena into our prior definitions, predictions, and models.\n\nWith increasing frequency, now results are requiring re-thinking whole areas of interpretation that seemed on basis of yesterday's lesser level of information, to be what is going on intercellularly, extrcellularly, in systemic signal transmission and response, between a cell and its neighbors.\n\nALL areas of this veritable deciphering of how nature works in this sub-system or that one is subject to individual differences, even within a species. One human's body, we know all too well, is not an identical system to the body of another. Commonalities can be found, yes; but the variation from individual to individual in, say, the effort to predict, prevent, diagnose, treat (and even in some cases to cure) cancer is complex.\n\nEvery area of molecular biology may be a gray area. Or, to tie that to the "language learning" or "reverse engineering" analogies above... every increment of new learning fits into a what remains a gray picture of complex systems having things in common, and individual systems "wired" as it were, a little differently in their organization.\n\nPerhaps we should not be surprised at much of anything, but should reflect that in the sciences generally, as Lincoln Barnett wrote in his The Universe and Dr. Einstein, "As new data are obtained, as new concepts emerge to rationalize them, the picture grows ever more remote from\neveryday human experience." (My sincerest apologies to Mr. Barnett for paraphrasing; but this is, I think, his gist.)\n\nMolecular science must slog on, and readapt and readapt and readapt, and let nature speak for itself. \n\n

Popular Now

  1. Secret Eugenics Conference Uncovered at University College London
  2. Like Humans, Walruses and Bats Cuddle Infants on Their Left Sides
  3. How Do Infant Immune Systems Learn to Tolerate Gut Bacteria?
  4. Scientists Continue to Use Outdated Methods