CHALLENGE SUMMARY
No cell is an island, to paraphrase John Donne. We have somewhere between 10 and 70 trillion cells in our bodies all with the same genetic material we inherit from our parents. How then do we become who we are with all our organs and tissues integrated into the whole organism and how do these organs remember what they are throughout life? These questions are fundamental problems in biology—still. What causes organismality to emerge from individual cells, achieving control of conflict at lower levels so the organism becomes the unit of adaptation?
We have gained a tremendous understanding of the sequence, alphabet, and language of the genome, our knowledge of molecules and many of signaling pathways in normal and cancer cells has filled thousands and thousands of pages of journals, yet our fundamental understanding of how the organism in its collectivity becomes the unit of physiological function is at best quite rudimentary. Whereas “the role of individual variation in collective behaviors has been woefully understudied” in population studies, the reverse is true in biological sciences: We are losing the organism for the cells or to use a more common expression: We are losing the forest for the trees!
But the above have inescapable relevance to why we age and why we get cancer. So why do we still know so little about the fundamentals behind the how and the why? This may be because the majority of cell and molecular biologists have been intensely interested in single cells and molecules, developmental biologists have been involved intensely in genetic manipulations to discover single players in pathways in developing organs,
and the physiologists have dealt essentially with the already formed complex organism or organs. The bioinformaticists—who these days are dealing with huge data sets to make sense of the whole—often have not studied biology, so the models become speculations at best. In each of these disciplines, we are poorly equipped to think about the other areas; what is worse, we are forced by the system of rewards and survival to go solo, look only under the light and inside the box, and stick with our own tool box.
Under these circumstances, simple logic often is not played: if we have all these cells that have the same genes shouldn’t there be something else that brings specificity? Some of us have argued that this is the microenvironment and the context the cells and tissues find themselves.
A movement is afoot to get the physicists and engineers together with the biologists. This is good. But the problem is that these disciplines are far from each other and will take decades to truly train both sides to understand each other and there is much arrogance to go around! But why shouldn’t physiologists and molecular biologists who are much closer in discipline collaborate to answer our fundamental problems, for example?
The challenge for the working group is to answer the following question:
What causes organismality to emerge from individual cells, achieving control of conflict at lower levels so the organism becomes the unit of adaptation? How do we define “microenvironment” vs the “genes,” and whether it makes sense to ask: what is the hen and what is the egg?
Could we explore if there are similarities or differences between the ways we study single cells and groups of cells within a tissue in biology and those tools used and similar questions posed in social sciences for individuals versus groups? Can we learn from the behavior of other species in biology such as honeybees and microbes? Also, why have biologists been hampered in their understanding of how an organism comes to pass from its individual cells? This fundamental question needs to be better understood before we can also understand why and how we age and why cancer cells prefer not to cooperate with each other.
Key Questions
Do you believe that “no cell is an island”? Does this analogy work? If yes or no, why? Do you think the statement that we have 10-70 trillion cells in our bodies and that these essentially have the same genes, correct?
If yes, how would you go about explaining why your nose and elbow are different? How would you explain this point to your students if you were to write a textbook? Was there anything about this point in your last biology textbook you looked at? What is the definition of the microenvironment here and is it important?
Does it make sense to have such a broad umbrella for this NAKFI meeting? Can social sciences teach anything useful to biologists and vice versa?
Does IDR Team Challenge 1 “Using our understanding of cooperation in cognitive organisms to understand cooperation in organisms or entities without brains and vice versa” have relevance to this challenge?
Do single cells by themselves and individual cells within a tissue or organ differ functionally? If so, what is the evidence? What are the advantages if these were similar? What are the disadvantages? Are the movements from single cell to tissue reversible? When tissues are digested, do single cells retain their memory of the group or do they revert to the original, “uneducated” form?
How much “plasticity” is there? How much can be tolerated? Can cells perform tasks outside the body that do not usually exist in the body? How would they do it? What would it mean? Can we debate the following: “Don’t ask what a cell can do, ask what it does do”?
Is epigenetics relevant to the above questions? Have you studied this area? What do you think it means? How broadly should we define this term? Is it important? How does the epigenome play a role in the first question above?
Do you, or should you, think about Descartes? Did he say anything that is useful? Are we too uncritical of Darwin, i.e., are we too Darwinian? Are there analogies between Darwin and Freud? If so, why and what are the lessons—good and bad?
Is there anything in biology that is linear? Should we continue to think of signaling pathways as linear? How did you learn signaling from a textbook? Did any of your professors argue that biology in general is or is not linear? Is anything in biology “flat”? If so, how would it work?
Is there anything in biology that is self-organizing? What does this mean to you?
The title of a volume of a journal is “From Single Cell to Biology.” Does this make sense? Do you think there is something missing? Should it say from single cells to organs (or some such)? Why is this question being posed?
Reading
Bhat R and Bissell MJ. Of plasticity and specificity: Dialectics of the microenvironment and macroenvironment and the organ phenotype. Wire Developmental Biology 2013;3(2):147-163.
Bissell MJ and Hines WC. Why don’t we get more cancer? Nature Medicine 2011;17:320-329.
Bissell MJ. Website: http://www.lbl.gov/lifesciences/BissellLab/main.html (This includes TED talk.)
Hayden EC. Human genome at ten: Life is complicated. Nature 2010;464:664-667.
Zhu L, Aonoa M, Kim S-J, and Haraa M. Amoeba-based computing for traveling salesman problem: Long-term correlations between spatially separated individual cells of Physarum polycephalum. BioSystems 2013;112(1):1-10.
Moiseffi A and Copeland J. Firefly synchrony: A behavioral strategy to minimize visual clutter. Science 2010;329(5988):181.
Queller DC and Strassmann JE. Beyond society: The evolution of organismality. Philosophical Transactions of the Royal Society B 2009;364(1533):3143-3155.
Queller D and Strassmann JE. The veil of ignorance can favor biological cooperation. Biology Letters 2013;9(6):20130365.
Furuta S, Jeng Y-M, Zhou L, Huang L, Kuhn I, Bissell MJ, and Lee W-H. IL-25 causes apoptosis of IL-25R–expressing breast cancer cells without toxicity to nonmalignant cells. Science Translational Medicine 2011;3(78):78ra31.
Sathie S, Khetan N, and Nanjundiah V. Interspecies and intraspecies interactions in social amoebae. Journal of Evolutionary Biology 2014;27(2):349-362.
These very famous words quoted from Meditation XVII, a prose work published in 1624, remind us of the relevance to the above:
“No man is an island, entire of itself; every man is a piece of the continent, a part of the main; if a clod be washed away by the sea, Europe is the less, as well as if a promontory were, as well as if a Manor of thy friends or of thine own were; any man’s death diminishes me, because I am involved in Mankind; and therefore never send to know for whom the bell tolls; it tolls for thee.”
-John Donne
Because of the popularity of this topic, three groups explored this subject. Please be sure to review the other write-ups, which immediately follow this one.
IDR TEAM MEMBERS—GROUP A
Treena L. Arinzeh, New Jersey Institute of Technology
Robert H. Austin, Princeton University
Oleg A. Igoshin, Rice University
Sarah Lewin, New York University
Wolfgang Losert, University of Maryland
David K. Lubensky, University of Michigan
Ichiro Matsumura, Emory University School of Medicine
Mustafa A. H. Mir, University of California at Berkeley
Mercedes V. Talley, W.M. Keck Foundation
Sarah Lewin, NAKFI Science Writing Scholar New York University
IDR Team 8A was assigned the task of investigating what causes organismality to emerge from individual cells, achieving control of conflict at lower levels so the organism becomes the unit of adaptation. Each of the three groups with this task approached it in a different way, and IDR Team 8A focused primarily on identifying the qualities most important to making multicellular organisms work—and fail.
Experimenters have seen normal cells changing type entirely or growing out of control when put in a petri dish. They have seen tumor cells reverting to normal activity when put back into the right environment. Scientists who study wound healing, breast development, and neuron growth see isolated cells reacting very differently from those embedded in a system. Researchers in regenerative medicine test different complex three-dimensional scaffolds to grow cells into functional tissues, and roboticists, synthetic biologists, and computational modelers try to build complicated systems with specialized units working synergistically. And yet, there’s no way to carry practical knowledge of what works across different scales, levels of complexity, or types of organism.
Multicellular organisms make up so much of the world, and so much of ourselves, but we don’t have generalizable knowledge of how they work or a framework for how to combine the knowledge and techniques of different fields to get there. IDR Team 8A sought to understand the properties, intercellular interactions, and cell-environment interactions that allow multicellular organisms to function as a single unit and exhibit collective behavior.
Fulfilling the Social Contract . . . or Going Rogue
If we had a bag of cells, could we pick out the ones that came from a multicellular organism? Team 8A noted that such cells might be less hardy than single-celled organisms and they might be more specialized—they have generally had to compromise acting on their own behalf for a particular, globally important function.
A key characteristic of cells in many multicellular systems, including our own bodies, is the tendency toward differentiation: the cells start out uniform, but will diverge into dramatically different types based on their environment and surrounding cells as they reproduce. The more they differentiate, the less flexible they are and the more focused role they play within the body. And cells that have differentiated to fill a specific function make the ultimate sacrifice: they have a limited number of divisions before they die, and they will undergo planned cell death to help overall development of the organism.
The cells that make up multicellular organisms are not robust on their own but gain adaptability as part of the group by sharing energetic tasks and differentiating or specializing to different levels to suit the organism’s overall needs. But in some cases, such as when a cancer cell detaches from its original tissue and migrates to a different part of the body in the process of metastasis, a cell will entirely abandon its responsibilities and race elsewhere to grow. Scientists do not yet fully understand the characteristics of a cell that allow it to work within a multicellular unit and the stressors that can disrupt that or prompt a transition.
Hierarchy of Properties and Interactions
In order to run sensible experiments on the behavior of multicellular organisms, Team 8A discussed how researchers have to choose variables to restrict using a two-dimensional substrate or three-dimensional scaffold, limiting the interaction to only one or a few different types of cells, or narrowing down which aspects of the cells’ microenvironment are incorporated into the system. The team also considered whether any particular stimulus is needed to form a multicellular organism and whether multiple or diverging cell types are necessary. They noted that changing the physical shape of a tissue changes its function, and observed that it is not yet possible to predict how that will happen.
The process right now cannot be standardized: nobody has quantified
which elements are needed for cells to form a multicellular unit, or carry out particular functions, and which are just part of the environment. Knowing that would let researchers design targeted, efficient tests and build working multicellular systems without unnecessary complexity.
Toward Understanding
It can seem like a paradox: how do you start investigating before you know which elements of the cells and environment are important? But by simultaneously experimenting and modeling, researchers could reverse-engineer how multicellular systems work while narrowing down to the important variables.
Essentially, there would be two levels of complexity. Researchers would measure as many physical properties of component cells and their microenvironment as possible, looking at how they interact with one another on a small scale. They would then let the system grow and measure the characteristics and collective behavior of the tissues that emerge. At the same time, they would form a model of the cellular system with the measured parameters and then let that model develop, investigating the emergent patterns that form. By comparing the real and virtual tissues and changing the breadth of characteristics included in the virtual cells, the researchers will be able to tell which elements of the real cells are contributing to successful multicellular function. That way, when tissue engineers or cancer researchers set out to influence cellular systems and induce collective behaviors, they will know which aspects are most important to control.
Most of the technology needed for this kind of in-depth experimentation is already available: single-cell sequencing, identification and characterization of proteins, chemical analysis of the microenvironment, and microscopy; we also have tools to look at the physical changes the system goes through over time and how easily cells can move around. As technology moves forward, researchers will be able to measure properties less invasively and even change cell properties without interrupting the whole system. Tools that allowed scientists to characterize the three-dimensional microenvironment without interrupting an experiment and take time-dependent measurements with single-cell resolution would make this kind of analysis even easier and more productive.
Opening New Research Doors
Ultimately, filling in the framework would let researchers figure out what functional tissues and other multicellular structures have in common. They would find the smallest necessary set of properties, interactions, and environmental cues to generate tissue, and they would understand how the properties apply to multicellular organisms on different scales. They might even identify alternate, simpler ways to build those same structures. Researchers would also be able to clarify how multicellular systems respond to stress and understand what specifically can incite cells to break off independently, as in cancer metastasis.
As the team’s final-day presenter put it: “We can identify those key knobs, and once we do that we can begin to play with them.” And, through that play, researchers would get a handle on how organisms develop and build a bridge between basic biological principles and broader, macrocosmic interactions. Their discoveries would apply everywhere from biological threats like cancer to our own origins and the development of life.
IDR TEAM MEMBERS—GROUP B
Edward Archer, University of Alabama, Birmingham
Gabor Balazsi, Stony Brook University
Matteo Cavaliere, University of Edinburgh
Richard S. Conroy, National Institute of Biomedical Imaging and Bioengineering
Joseph D. Dougherty, Washington University
Michael D. Goodisman, Georgia Tech
Maria Pellegrini, W.M. Keck Foundation
Luis M. Rocha, Indiana University
Daniel Segre, Boston University
Blanka Sharma, University of Florida
Elizabeth Van Volkenburgh, University of Washington
Douglas White, Georgia Institute of Technology
Douglas White, NAKFI Science Writing Scholar Georgia Tech and Emory University
Introduction: A Multidisciplinary Approach to Answering a Timeless Question—How Did Complex Life Evolve?
IDR Team 8B was asked to address the emergence of multicellular organisms from individual cellular units. This question had two distinct parts which were each important to consider. First, what are the properties that define multicellular organisms? Second, what factors in the environment cause evolution to act on groups of cells instead of just single cells? The synergism of these two ideas captures the massive scope of the question posed to IDR Team 8. Understanding the evolution of multicellularity would provide insights understanding tissue regeneration, cancer, and metabolic diseases (e.g., type 2 diabetes mellitus) which are open and important challenges in human health.
IDR Team 8B consisted of a mix of computer scientists, neurologists, plant biologists, physiologists, geneticists, and experts on social insects. Using their multidisciplinary approach, they drew upon concepts from metabolism, computer science, and evolutionary biology to try and explain how multicellular organisms evolved from single cells.
What Defines a Multicellular Organism?
The team started by tackling the first half of this question: what makes a multicellular organism different from just a group of cells? The idea of specialization—that each cell can take on a specific role—came up early in the conversation. The freedom of cells to specialize is a hallmark of a multicellular organism as specialization allows the organism to perform more complex tasks with greater ease. Each cell can now focus on becoming a master at a specific task instead of having to become a jack of all trades. Interestingly, the team also observed that while specialization may be necessary in some cases, there is also a cost associated with specializing, and the system as a whole becomes more vulnerable with specialization.
The team also discussed what other properties are necessary to define an organism. The idea of cells as Turing machines was used to explain cell behavior. A Turing machine is a theoretical concept which consists of two
parts: a piece of memory which stores information, and a machine capable of storing information on that memory. This analogy provided many useful parallels when considering cells and how they evolve with respect to their environment and themselves. In the first iteration of this, the DNA was the information, while the cellular machinery responsible for reading and writing that DNA was considered the machine. However, several other types of information storage were considered including proteins, carbohydrates, voltage potentials, and RNA. One team member brought up an interesting concept that all of these types of information have a trade-off between the time it takes to read the information versus the stability of the information over time. While DNA stores information in an incredibly stable manner, it takes hours to respond to a signal. In contrast electrical potentials are transient phenomena but can open and close protein channels on the cell membrane on the order of seconds.
After hashing out the different mediums a cell can use to store information, an inevitable discussion of what has access to this cellular memory ensued. Can the environment influence cellular memory, or does only the cell have access to it? This was simply a surreptitiously packaged form of the age-old nature versus nurture debate which the group ultimately answered using the abstraction of the Turing machine. The cell as the machine can write to the memory, but that does not stop external forces, such as the environment, from also writing to the same memory. Thus, the group came to a consensus that, in the context of the evolution of multicellular organisms, both the environment and cellular genetic makeup are important.
A New Way to Understand Evolutionary Theory: A Computer Science Approach
The theory of evolution specifies that evolution is driven by some gain in fitness. The IDR team decided to subsequently explore what this means specifically with respect to the evolution of multicellular organisms. The requirements put in place via the environment certainly constrain how organisms evolve, and the team chose to think of these constraints as a fitness landscape. The team often chose to use an organism’s access to energy as a proxy for fitness. Given the question of how an organism evolves to maximize its access to energy, the team drew on another analogy from computer science: the space, time, communication complexity axes. Briefly, the idea behind using these axes in computer science is that they provide a quantitative way to compute the cost associated with completing a given task.
For example, if a computer is asked to solve a very large problem there are several ways it can accomplish this task. If time is not an issue, one computer can run until it solves the problem. This method would have a high time cost, but does not require complex communication or more than one computer. In another solution multiple computers could run in parallel to solve the problem. This would have a high space cost because more than one computer is required to solve the problem, but the two in parallel would solve the problem faster than a single computer which corresponds to a lower time cost. In a final example two computers divide the task up, and each computer specializes on solving a different part. This would also minimize the time cost but the two computers are required to share information to complete the task; thus there is an increase in communication complexity. Thus, though multiple different solutions exist for a single problem, each of them has a different cost associated with the solution.
In the context of biological systems, the team decided to rename these axes as the communication cost, space cost, and time cost required to solve any given biological problem. The team again related this to energy, and in a thought experiment, asked how different single cells could solve a task. In this task a cell had to drill through a barrier to get to a food supply. The group proposed three main solutions, which were analogous to those described in the computer examples. In the first example, a single cell could attempt to drill through the wall, sense how close it was to the food, and then move to the food and consume it. This would have a long time cost, but would not pay a cost for space or communication. In the second example, multiple cells could try to chew through the wall all in tandem. They would breach the wall faster, reducing the time cost, but the space cost would increase as more cells are needed to accomplish the task. In the third example a group of cells divide the task into movement, chewing and sensing to ultimately reach the food. They would reach the food much faster than the previous two examples, thus a low time cost, but pay a cost both in terms of the communication between the various specialized units, and the increase in the number of cells necessary.
The group used these analogies to analyze how various different organisms would fall on the space, time, and complexity axes. Bacteria and other single-celled organisms can grow into large colonies if the food source is plentiful. However, the rate at which they can consume food is directly proportional to the number of bacteria, and thus this strategy only works if a lot of bacteria are present (high space cost), or they have a long time to consume the food (high time cost). If there was a way to consume the
energy faster, this would provide a fitness advantage over this single-celled approach. In multicellular organisms cells take on specific roles related to sensing, moving, and harnessing energy resources. Due to this cooperative approach multicellular organisms can consume food much faster than their single-cell counter parts, but all of this specialization requires an increase in communication between these cell types (high communication cost). Interestingly, plants have addressed this problem by taking advantage of sunlight, which is a nearly infinite and plentiful resource but is chemically extremely difficult to harness. Thus considering biological evolution in the context of this three-axis energy landscape was extremely useful in helping to describe some of the underlying principals governing the evolution of multicellular systems.
Conclusions
IDR Team 8B took a unique and refreshing approach on tackling an age-old problem in biological theory. By combining ideas from evolutionary theory with modern abstractions from computer science, the group was able to come up with a framework for understanding the evolution of multicellular organisms. Though the concepts and equations governing this approach are still rough and undeveloped, it provides a starting point and a common language to enhance discussions about future study. Using this approach the team hopes to probe the fitness landscape to understand the conditions necessary for multicellular life to evolve. Understanding these principals could allow the creation of artificial multicellular organisms designed to complete specific tasks. Similarly, this approach could be harnessed to understand the evolution of cancer and metabolic diseases which would provide an avenue to developing more effective therapeutics and designs.
IDR TEAM MEMBERS—GROUP C
Josh N. Adkins, Pacific Northwest National Laboratory
Bahareh Behkam, Virginia Tech
John C. Doyle, Caltech
Jeffrey J. Fredberg, Harvard School of Public Health
Warren L. Grayson, Johns Hopkins University
Claire Maldarelli, New York University
William C. Ratcliff, Georgia Institute of Technology
Vanessa Sperandio, UT Southwestern Medical Center
Daniel K. Wells, Northwestern University
Mingming Wu, Cornell University
Claire Maldarelli, NAKFI Science Writing Scholar New York University
Team 8C was asked to answer a question seen universally in nature: What causes organismality to emerge from individual cells, achieving control of conflict at lower levels so the organism becomes the unit of adaptation?
Team 8C initially approached the problem by looking at specific examples in nature where individual cells interact with each other to become a greater functional unit, such as how a biofilm forms, how cancer cells grow into a tumor, or how an ant colony functions. Cells are not autonomous. Single cells within a multicellular organism do not work by themselves. The functionality of a multicellular organism does not come from every singular cell working on its own; it comes from the cells working together. This multicellular system and the systems around it create a microenvironment, which all multicellular organisms must rely on to function. By understanding the microenvironment of an organism, for example a type of cancer’s microenvironment, a better understanding of that cancer’s functionality can be inferred. We can then begin to understand what it is in the cellular environment that makes a cell become a cancer cell.
The team quickly realized that in order to succeed in life, most organisms follow a randomization strategy. In a human, most cells have almost the same genetic information, but slight differences allow cells to differentiate, to become different tissues and organs. By following a randomized strategy, slight differences develop and life thrives, allowing for adaptation to a changing environment. This system of randomization is seen universally throughout nature.
The team also noticed that within all functional units, a universal mechanism is at play. To establish a common model of functionality on an organismal level, the group likened its organizational hierarchy to Apple’s iPad. This architecture, they believe, governs how any group of cells functions as its own organ or organism. By studying biological systems and other systems in nature in adherence with this architecture, generalizations can be
drawn and predictions can be made. At the top of this layered architecture is the hardware. As in an iPad or any electronic device, hardware is fast, yet costly and not very diverse—for any given device, most hardware systems operate in the same way. The next level in this design is the operating system. This system is unique in that it is not very diverse—very few operating systems exist—yet from those few, an enormous amount of diversity arises. A very large number of apps can work on an iPad’s operating system. And when writing an app, only the type of operating system that the app is going to use needs to be considered. So, in general, it is very easy and inexpensive to write an app.
The group was also able to see this layered architecture in every biological system they discussed. For example, in biofilms, the hardware is the cell. The operating system is the intercellular communication, the chemical gradients, and the mechanical forces that are at play. On top of that there are the functions, or apps, that are implemented on the layer of the biofilm. The biofilm’s “apps” would be the cellular differentiation that develops as well as biofilm-level reproduction.
Using this design, they found, allows for life to grow, change, and adapt—and quickly. The reason bacteria can swap genes is that they are running exactly the same operating system. A gene comes in and it is simply a “plug and play”—they plug into the existing system that they are already familiar with and start working. However, the team noted that there is a major problem with this architecture. The universally shared operating system is hijackable. In bacteria, just as genes can be transferred, so can viruses. The way cancer develops is very similar to a biofilm. The hardware is the cells and the operating system is the intercellular communication and chemical and mechanical forces at play. The apps for a cancer cell are initiation, invasion, and eventually metastasis. Specifically, the operating system in a cancer cell is there to do tissue regeneration of wound healing. The cancer cell has essentially hijacked a normal cell in this process. So while it is normally beneficial to have this beautifully layered architecture, it allows for unwanted invaders, such as viruses or cancer, to take over.
The team concluded that with this architecture comes a balance of trade-offs. While the hardware is fast, it is inflexible, and while the operating system gives rise to a huge diversity of apps, it is hijackable. These patterns of trade-off with speed and flexibility are essential in understanding what drives organismality to emerge from individual cells and how biological noise allows organisms to grow, change, and find niche areas in which to thrive—whether we want them to or not.
