Recent postings have provoked numerous questions about my application of the term "cognitive" to cell regulatory processes. I base this usage on the notion that cognitive actions are knowledge-based and involve decisions appropriate to acquired information. It is common today for molecular, cell and developmental biologists to speak of cells "knowing" and "choosing" what to do under various conditions. While most scientists using these terms would insist they are just handy metaphors, I argue here that we should take these instinctive words more literally. Cell cognition may well prove itself a fruitful scientific concept.
Choosing the right sugar to eat: Jacques Monod and bacterial control regimes
We can trace the origins of molecular understanding of how cells control reading DNA sequence information to Monod's pioneering studies (in 1942 Nazi-occupied Paris!) on how bacteria respond to a choice of sugars. They completely consume the sugar that provides the most rapid growth (typically glucose) before switching to the less efficiently digested one.
Unrelated bacteria use the same transport system to sense the presence or absence of glucose in their environment, but they employ different molecular components to transmit that information to the genome. So this important sensory mechanism evolved at least twice in bacterial history. In E. coli, the signal that glucose is absent is an intracellular "second messenger" molecule that is purely symbolic; it has no structural connection to the glucose transport process.
Recognition and metabolism of the less-preferred sugar is a sophisticated process. It involves transport components, metabolic enzymes and specialized regulatory proteins. Together, they function as microprocessors controlling expression from the corresponding DNA regions in response to each sugar and the second messenger. This integration ensures that proteins needed for digestion only appear when appropriate.
While many assert that the bacterial control system is purely mechanical, not enough experiments have been done to show whether cells respond in a deterministic way. What we know for sure is that even these smallest cells use sophisticated sensory and intracellular communication processes to discriminate between alternative nutrients.
Passing through the cell cycle successfully: the checkpoint concept
Complex cells with a nucleus (eukaryotes) display a tightly controlled multistage cell division cycle. Each stage involves intricate processes such as cell growth, DNA replication, and accurate transmission of genome copies to daughter cells. Elaborate biochemical reactions regulate passage from one stage to the next.
On top of the stage-to-stage control circuitry, a self-monitoring system makes sure everything comes out right. If the different biochemical and biomechanical processes fall out of synch, or if there is either a mistake or damage, sensory molecules detect the problem. They activate a "checkpoint" to hold up the entire cycle until everything has been set right for renewed progress.
Cells set distinct checkpoint systems for growth and division. The easiest to appreciate is the "spindle checkpoint." This makes sure that each daughter cell gets one and only one copy of each duplicated chromosome. The reliability of cell division depends on this sensory process. Left to random chromosome distribution, less than one in a billion divisions of our own cells would be successful. If any pair of chromosomes is not correctly aligned on the spindle apparatus to ensure equal transmission of the copies, the checkpoint apparatus senses the misalignment and emits a signal to halt cell division. Once all chromosomes are properly aligned, the checkpoint is released and division follows quickly.
Although each checkpoint could be deemed just another intricate mechanism, it is hard to consider the entire integrated cell cycle-checkpoint system purely mechanical. This is because the network is capable of responding to completely unpredictable events, such as external damage or experimental interventions. It displays reliability enviable in any complex human manufacturing process. Note that a dividing cell has far more components than any man-made device.
"To be, or not to be. That is the question."
There is widespread agreement in the molecular cell biology community that programmed cell death following trauma is the result of a cell decision-making process. When eukaryotic cells suffer injury (particularly well studied in the case of DNA damage), there are at least two different outcomes:
(1) a checkpoint, repair of the damage, and then resumption of the cell cycle,
(2) or cell suicide following an organized cascade of events, labelled "apoptosis." The dying cell disintegrates in an orderly way.
The cell chooses between repair and survival, on the one hand, and apoptosis, on the other, based on its environment. The key environmental features for human cells are nutrition and the presence of intercellular signaling molecules, called cytokines.
Some signals favor cell death, like Tumor Necrosis Factors (TNFs). A damaged cell detects them by means of so-called "death receptors" on the outer surface of its membrane. Other signals favor survival and proliferation, and are generally called "growth factors." Each type of growth factor has one or more specific receptors used to sense its presence.
A cell's response to a particular source of damage, such as X-rays, depends dramatically on the cytokines present in the environment. The response to trauma, where a cell decides either to adjust its activities and survive or undergo an orderly cell death, is an excellent candidate for empirical investigation as a cognitive process.
Interestingly, even so-called "simple" cells like bacteria have programmed cell death routines subject to control by signals from other cells.
In all cases, the cell suicide routine may be interpreted as a benefit to the multicellular community. This is also true of bacteria, which spend the majority of their active existence as multicellular organisms, not isolated single cells.
Although controversial as a general feature of bacterial life when first suggested, we now know that most bacteria display various multicellular behaviors. They involve emitting, receiving and processing a large vocabulary of chemical symbols.
Based on all the above, I think it can reasonably be argued that cell cognition and intercellular communication are central to all levels of life. They deserve detailed empirical and theoretical analysis.
Li, Z. et al. (2010). "Decision making of the p53 network: Death by integration." J Theor Biol. http://www.ncbi.nlm.nih.gov/pubmed/21130774.
Li, Z. et al. (2011). "Bacteria-based AND logic gate: a decision-making and self-powered biosensor." Chem Commun (Camb) 47(11): 3060-3062. http://www.ncbi.nlm.nih.gov/pubmed/21203618.
Lindqvist, A. et al. (2009). "The decision to enter mitosis: feedback and redundancy in the mitotic entry network." J Cell Biol 185(2): 193-202. http://www.ncbi.nlm.nih.gov/pubmed/19364923.
Nakagaki, T. (2001). "Smart behavior of true slime mold in a labyrinth." Res Microbiol 152(9): 767-770. http://www.ncbi.nlm.nih.gov/pubmed/11763236.
Nakagaki, T. et al. (2004). "Obtaining multiple separate food sources: behavioural intelligence in the Physarum plasmodium." Proc Biol Sci 271(1554): 2305-2310. http://www.ncbi.nlm.nih.gov/pubmed/15539357.
Nakagaki, T. et al. (2004). "Smart network solutions in an amoeboid organism." Biophys Chem 107(1): 1-5. http://www.ncbi.nlm.nih.gov/pubmed/14871595.
Sanges, D. and M. P. Cosma (2010). "Reprogramming cell fate to pluripotency: the decision-making signalling pathways." Int J Dev Biol 54(11-12): 1575-1587. http://www.ncbi.nlm.nih.gov/pubmed/21305473.
Zhang, X. P. et al. (2009). "Cell fate decision mediated by p53 pulses." Proc Natl Acad Sci U S A 106(30): 12245-12250. http://www.ncbi.nlm.nih.gov/pubmed/19617533.
Some additional experimental and theoretical references include:
Avraham, R. and Y. Yarden (2011). "Feedback regulation of EGFR signalling: decision making by early and delayed loops." Nat Rev Mol Cell Biol 12(2): 104-117. http://www.ncbi.nlm.nih.gov/pubmed/21252999.
Bitbol, M. and P. L. Luisi (2004). "Autopoiesis with or without cognition: defining life at its edge." J R Soc Interface 1(1): 99-107. http://www.ncbi.nlm.nih.gov/pubmed/16849156.
Koseska, A., A. Zaikin, et al. (2009). "Timing cellular decision making under noise via cell-cell communication." PLoS One 4(3): e4872. http://www.ncbi.nlm.nih.gov/pubmed/19283068.
Latty, T. and M. Beekman (2011). "Irrational decision-making in an amoeboid organism: transitivity and context-dependent preferences." Proc Biol Sci 278(1703): 307-312. http://www.ncbi.nlm.nih.gov/pubmed/20702460.
"What is the scientific evidence that makes it necessary to assign "cognitive" functions (as defined by the usual use of the word, which is reserved for very high level functions in humans and animals) to (albeit complex) biochemical processes in cells?"
IMHO such an assignment requires an enormous amount of scientific evidence that needs to show laser sharp focus on the question as to why thermodynamic models of such processes are unable to predict the behaviour of such cells and why psychological terms for information processing are necessary to make scientific progress.
I think this is a sufficiently precise scientific question to have a very good answer in available publications. So far I have not been shown any peer reviewed publication that would contain any evidence (let alone a sufficient amount) for the assertions made by Dr. Shapiro. I am more then willing to discuss details of such publications, but I would ask to be shown particular passages that are supposed to contain the required evidence.
Adrian, J., S. Torti, et al. (2009). "From decision to commitment: the molecular memory of flowering." Mol Plant 2(4): 628-642. http://www.ncbi.nlm.nih.gov/pubmed/19825644.
Ambravaneswaran, V., I. Y. Wong, et al. (2010). "Directional decisions during neutrophil chemotaxis inside bifurcating channels." Integr Biol (Camb) 2(11-12): 639-647. http://www.ncbi.nlm.nih.gov/pubmed/20676444.
Balazsi, G., A. van Oudenaarden, et al. (2011). "Cellular decision making and biological noise: from microbes to mammals." Cell 144(6): 910-925. http://www.ncbi.nlm.nih.gov/pubmed/21414483.
Carpenter, A. C. and R. Bosselut (2010). "Decision checkpoints in the thymus." Nature Immunology 11: 666 - 673. .
Falschlehner, C., C. H. Emmerich, et al. (2007). "TRAIL signalling: decisions between life and death." Int J Biochem Cell Biol 39(7-8): 1462-1475. http://www.ncbi.nlm.nih.gov/pubmed/17403612.
Festjens, N., T. Vanden Berghe, et al. (2007). "RIP1, a kinase on the crossroads of a cell's decision to live or die." Cell Death Differ 14(3): 400-410. http://www.ncbi.nlm.nih.gov/pubmed/17301840.
Goodnow, C. C., C. G. Vinuesa, et al. (2010). "Control systems and decision making for antibody production." Nature Immunology 11: 681 - 688. http://www.ncbi.nlm.nih.gov/pubmed/20644574.
Halley, J. D., F. R. Burden, et al. (2009). "Stem cell decision making and critical-like exploratory networks." Stem Cell Res 2(3): 165-177. http://www.ncbi.nlm.nih.gov/pubmed/19393588.
Helikar, T., J. Konvalina, et al. (2008). "Emergent decision-making in biological signal transduction networks." Proc Natl Acad Sci U S A 105(6): 1913-1918. http://www.ncbi.nlm.nih.gov/pubmed/18250321.
Kobayashi, T. J. (2010). "Implementation of dynamic Bayesian decision making by intracellular kinetics." Phys Rev Lett 104(22): 228104. http://www.ncbi.nlm.nih.gov/pubmed/20867209.
Kurosaki, T., H. Shinohara, et al. (2010). "B cell signaling and fate decision." Annu Rev Immunol 28: 21-55. http://www.ncbi.nlm.nih.gov/pubmed/19827951.
Loewer, A. and G. Lahav (2006). "Cellular conference call: external feedback affects cell-fate decisions." Cell 124(6): 1128-1130. http://www.ncbi.nlm.nih.gov/pubmed/16564006.
Mojzisch, A. and K. Krug (2008). "Cells, circuits, and choices: social influences on perceptual decision making." Cogn Affect Behav Neurosci 8(4): 498-508. http://www.ncbi.nlm.nih.gov/pubmed/19033244.
Ngalim, S. H., A. Magenau, et al. (2010). "How do cells make decisions: engineering micro- and nanoenvironments for cell migration." J Oncol 2010: 363106. http://www.ncbi.nlm.nih.gov/pubmed/20652046.
Perkins, T. J. and P. S. Swain (2009). "Strategies for cellular decision-making." Mol Syst Biol 5: 326. http://www.ncbi.nlm.nih.gov/pubmed/19920811.
• “YOU HAVE TO PRODUCE EVIDENCE” is one of Swift’s repeated complaints. In fact, this posting contains 17 links to published articles (each with further references), to summary discussions or reference lists from my web page (with multiple citations), and to Wikipedia entries (with links to articles and databases). The evidence for sensory, information transfer and decision-making processes in the cell systems I discussed IS readily available to the interested reader.
• Swift demonstrates a deep confusion about mathematical models, empirical observations, and scientific explanations. In referring readers to “…simulation tools for cellular regulatory networks,” he makes the astonishing statement that “These efforts are only a google search away for those who care about the cold reality of it.” Anyone who understands scientific rigor would not confuse attempts at computer simulations with “cold reality.” Scientific explanations are tentative and continually modified abstract representations of what we observe about the natural world. The reason to investigate concepts like cognition is to see if they lead us to better explanations for empirical observations, including the sensory-based adjustments in cell behavior that I outlined. Swift argues for close-mindedness, but his alternatives are hardly realistic or satisfying.
My references to simulation serve as a reminder that all the processes you are referring to are regarded by the scientific community as being computable by systems of differential equations and similar models of reality.
You will be hard pressed to find any successful references in the scientific literature to the simulation of proper cognitive processes by equally trivial mathematical approaches.
This should serve as a hands on reminder that cellular networks can be understood in similar terms as thermodynamics and chemical process controls and analog electronic circuits.
No such understanding of "cognition" exists (neural network theory has not advanced sufficiently to map high level cognitive processes to the network layer). In my opinion it is safe to conclude that these differences are more than just knee deep. They are, indeed, overwhelming.
Again, I am more than open to listening to a serious suggestion by you or other researchers, of how we should research cell cognition. Unfortunately, I can not find any reference to such a program in any of your submissions to this day.
I am very surprised by your need to call me "closed-minded". I am, indeed, very open minded, but I do require at least a minimal amount of scientific evidence to change my mind. This requirement happens to the foundation of science. We do not make changes or additions to existing and working frameworks until we are being forced to it by nature, which can supply us at any time with new data that contradicts the explanations that were developed to understand the older data sets.
Even then older theories are usually not supplanted by completely new ones but they are extended and, in some cases, integrated into the newer frameworks.
In this context I have yet to find a contradiction between molecular biology and Darwinian theory. If anything, modern molecular biology supplies a wonderful microscopic arsenal of tools to support Darwinian theory. The latter explain why certain birds lose flight and certain animals develop long necks. Molecular biology can determine which cellular and molecular changes had to happen on the level of genetics for these genotype changes.
To me this is no different than the interplay between Newtonian mechanics, which is just as valid today as it was at the time of Newton, and quantum mechanics. Quantum mechanics extends Newtonian mechanics, it does not make it obsolete.
The truth is that DNA sequence analysis contradicts Darwinian theory in many ways. Let me just take one example, Darwin's argument that evolutionary change is the gradual result of numerous "successive, slight" changes in hereditary traits. We now know that many key transitions in the evolutionary history of yeast, Paramecium, flowering plants (Darwin's "abominable mystery"), and vertebrates involved whole genome duplication (WGD) events, which can hardly be considered "slight" changes. The evidence for WGD in evolutionary history can be found in the references posted at http://shapiro.bsd.uchicago.edu/ExtraRefs.WholeGenomeDoublingCriticalStagesEvolution.shtml. My book explains this discovery and many other ways that the findings of molecular biology have led us to take a new perspective on evolutionary theory.
"Quantum mechanics extends Newtonian mechanics, it does not make it obsolete. "
As Thomas Kuhn pointed out, scientific ideas do not change in a gradual fashion but by the replacement of one conceptual framework with another. He called this conceptual transition a "paradigm shift." This was the case when Newtonian assumptions about the structure of space and time were replaced by Einstein's relativistic view. General relativity, not quantum mechanics, is the modern alternative to the foundations of Newtonian physics.
Misstatements like these two are the reason that I wrote you are "closed-minded."
Shapiro's statement (below) that it is general relativity, not quantum mechanics that is the modern alternative to the foundations of Newtonian physics is of course correct. Indeed, we can go one major step forward. Newtonian mechanics remains an acceptable approximation in many instances (but by no means all) if used to calculate planetary orbits, rocket trajectories etc. However, the underlying assumptions (paradigm if you wish) of Newtonian mechanics are flat out wrong.
Newton assumed an absolute 3 dimensional space and an independent unvarying one dimensional time parameter. Einstein demonstrated that both on these concepts are wrong. In their place is 'space-time', which is not absolute but is influenced by matter-energy, and which in turn governs the motion of matter-energy. The passage of time is not absolute but depends on the frame of reference of the observer.
The important difference between the Newtonian and Einsteinian views of the universe is the difference between the governing paradigms. Shapiro is presenting a new paradigm which has major implications for the interpretation and understanding of data concerning the behavior of living organisms. This paradigm is not some sort of 'magic' introduced to explain the behavior of organisms but is simply a new way of looking at the current data and formulating experiments to find new data.
A fitting word for these, sometimes very complex, regulatory functions is "homeostasis". It is well established and carries the proper meaning to describe what is happening in cellular regulatory networks.
Berthajane Vandegrift
A Few Autistic Questions about Freud, Marx and Darwin
Biology used to be a much more qualitative than quantitative science. But that has changed. Modern biologists are modelling cellular systems with differential equations describing the chemical equilibria in cells just as physicists are modelling planetary motion and electrical engineers simulate analog electronic circuits.
If you want to see how modern biology really deals with the problem of complex chemical dynamics in cells, look at simulation tools for cellular regulatory networks:
http://people.cc.ku.edu/~grobe/cell-sim/projects.html
http://www.brc.dcs.gla.ac.uk/projects/bps/links.html
etc.. These efforts are only a google search away for those who care about the cold reality of it.
There is no animism at work in these mathematical approaches, at all. It's all fairly straight forward math describing trivial thermodynamics.
We have (roughly) a continuum. Humans are the most complex, individual cells the least. If the behavior of cells can be fully described without the concept of 'cognition' but the behavior of human beings cannot, where do we draw the line? At invertebrates? At worms?
In particular, since 'science' can include the concept of 'cognition' when describing human behavior, why is it a violation of 'science' to include the same concept when describing cellular behavior?
Yes, you can ask that. And then, if you want to be taken seriously in science, YOU HAVE TO PRODUCE EVIDENCE that they do not (and please note that the differential equation models are just one of several layers of increasingly more detailed and precise mechanistic explanation, so they are expected to break down on some level, only to be replaced with a model of higher precision). Dr. Shapiro has not produced such evidence.
"Lets first consider human beings."
Dr. Shapiro is not talking about human beings. Neither am I. Cognition is a perfectly fitting term for humans. But we are talking about single cells here. So now you have to answer the scientific question of a complex function like "cognition" by doing experiments on single cells, not on complete human beings and you have to show conclusively that the function of a single cell reaches the level of what we COMMONLY call "cognition". Not a watered down, redefined meaning of "cognition", but cognition as it is defined in philosophy and psychology and used commonly in science.
Of course not. And none of that matters because we are still not talking about humans but about single cells.
"We have (roughly) a continuum."
That's an assumption of yours for which you would have to present EVIDENCE. And even then you are running into the problem that the term "cognition" does not describe a continuum but requires a rather high threshold of information processing, which, for instance, your computer can NOT, yet, reach. And your computer already happens to be much more complex than many single cells.
"In particular, since 'science' can include the concept of 'cognition' when describing human behavior, why is it a violation of 'science' to include the same concept when describing cellular behavior? "
It is not a violation... as soon as you can find an example of a cell that has similar cognitive capabilities compared to humans. So far, I believe, no cell has been found that can write a sonnet or teach a unit on general relativity.
Cells don't do that, based on everything I have seen, read, and understood.
If a cell chooses cell death over continued function, that has certainly got to be a matter of the physical contents of its surroundings and physical reactions taking place within it. In fact, that would be how we make decisions as well....wouldn't it? It is all physical responses on a cellular level, based on a massive number of physical variables.
I am not convinced that "cognition" in the way you wrote about it is happening, or has ever happened in anything!
If you believe our cognitive processes are nothing more than "physical reactions," then clearly you will not believe the same of cells. However, if think those physical/cognitive processes in ourselves are worthy of study, then why should we deny that at the cellular level?
There may be a typo in my just posted reply to your comment. I meant to say that if you believe our own cognition is nothing more than "physical processes," then you will clearly believe the same of cells. However, if you think human cognition is worth studying, why deny the same to the physical/cognitive processes that go on in cells?
NONE of that can be said for the cellular regulatory pathways that you are referring to. Your language is not based on evidence that these networks can do any of this but is merely your personal choice of words. Moreover, I am pretty certain that you are gravely misrepresenting the general scientific community with your assertions. Why you feel a need for this is something only you can answer.
Really? The molecular biologists I know do not talk that way. And under no circumstances do they use animistic terminology to promote bad science.
"Really?". Go to Google Scholar. Search using the following two phrases: 'cells know' and 'cells choose'. Yes, really.
I saw some recent article on research that proved the organic chemistry basis for information coding in microtubules - something that could be the memory mechanism in neurons and etc.
I have long suspected that the collective forraging and division of labor (division of resources) dynamics in larger multi-cellular animals would have to exist on a cellular scale.
Have you heard about that new theory of swarms? It covers everything from chemistry to ecosystems. Its thought provoking.
Thanks for your comment. I use the term "cognitive" for cell information-processing because it includes sensing, is limited and can be defined in a reasonable way. That will make it possible to investigate what cognition comprises.
In my opinion, it is a mistake to leap from "cognitive" to "mental." That brings in a whole host of issues, many of which are far from science (human amour-propre, for example) and obscures the attempt to figure out how cells actually operate.
If we agree that cells operate in a more-than-mechanical fashion, then we have to figure out what that means and how they do it.
For years people have talked about "Junk DNA" and "vestigial organs" and we've seen those terms become less appropriate with each passing year.
I like what Wendell Reed said about dogs, flies etc. We naturally assume bacteria are less smart than we are, but we don't really know that. What if in some respects they are smarter? Let's lay our assumptions on the table.
Maybe words like "mental" are too loaded as Dr. Shapiro says. But if we are willing to accept terms like cognition, perhaps we'll watch more closely and more easily acknowledge what cells are truly capable of. Only then can we learn from them.