ADVERTISEMENT
  About the SA Blog Network













Oscillator

Oscillator


Notes, thoughts, and news on synthetic biology.
Oscillator Home

I Heard You Like Feedback Loops

The views expressed are those of the author and are not necessarily those of Scientific American.


Email   PrintPrint



In some labs, biology and computer science are converging. On the one hand, computer scientists are working towards creating computer chips inspired by the circuitry of the brain, and on the other, some synthetic biologists are aiming to create biological computers inside living cells. Scientists and engineers have seen biology inside computers and computers inside biology for a long time, and the weird and wonderful blending and mixing and overlapping of the two is only becoming more pronounced.

A really cool recent paper announced the construction of cyborg yeast: cells that contain engineered genetic pathways that can be controlled by external signals that come from a computer. The researchers took advantage of two proteins from plants that control how the plant behaves in response to different wavelengths of light—in red light the two proteins bind to each other and turn on a gene, in near-infrared light the proteins unbind and the gene turns off. The yeast can be engineered with these proteins so that the expression of a yellow fluorescent protein can be controlled by light.

Feedback
The way that both engineered systems and living cells are controlled depends on feedback loops. Positive feedback amplifies a signal, like when a microphone picks up the sound coming out of its speaker, which then goes back to the speaker and on and on in a screechy vicious cycle.

While positive feedback amplifies signals, negative feedback stabilizes them by outputting the opposite of the input signal. A thermostat is a negative feedback device; when the temperature gets higher than the set point, the thermostat turns down the heat, and when the temperature gets too low the thermostat turns the heat up. Negative feedback devices have been used in engineering since antiquity, but weren’t theoretically formalized until the James Clerk Maxwell‘s paper “On Governors” in 1868. Governors were part of steam engines and made sure that the speed of the engine was kept constant. When the engine speeds up, the spindle of the governor moves faster and the centrifugal force pushes the metal balls up, pulling a lever that turns down the engine throttle. When the engine slows down the balls go back down and the throttle opens.

In the 1940s, researchers working in electrical engineering, control theory, communication theory, and neuroscience began to see feedback loops everywhere in both engineering and biology. A new field was born that brought together biology and engineering, with biology interpreted through engineering paradigms and engineering design influenced by how biological systems are controlled. In Cybernetics: or Control and Communication in the Animal and the Machine, Norbert Wiener discusses naming this new field:

[A]s happens so often to scientists, we have been forced to coin at least one artificial neo-Greek expression to fill the gap. We have decided to call the entire field of control and commuication theory, whether in the machine or in the animal, by the name Cybernetics, which we form from the Greek κυβερνήτης or steersman. In choosing this term, we wish to recognize…Clerk Maxwell…and that governor is derived from the Latin corruption of κυβερνήτης.

Feedback is everywhere in biology, from neural circuits to hormonal pathways to gene expression loops, maintaining homeostasis or amplifying signals like those that control embryo development. Norbert Wiener began to see the connections between control theory and biology when trying to build systems for shooting down airplanes during World War II. Feedback was necessary to constantly figure out how far off the system was from the plane and adjust accordingly. As happens so often in science, the crucial realization came from understanding how feedback loops fail; when there is a delay between the signal input and the negative feedback loop output, the output can overshoot the desired target and set the system into an uncontrolled oscillation as the feedback loop constantly tries to stabilize the system and overshoots the target from above and below. Wiener wondered whether the feedback loops guiding the anti-aircraft missiles had any similarity to the feedback loops that guide how the brain senses the position of the body and makes adjustments. Talking to a neurologist he learned that a similar condition occurs after some kinds of injuries to the cerebellum, where a patient will overshoot when trying to reach something and go into an oscillation known as a purpose tremor. The feedback loop that tells the brain how far away the hand is from the object isn’t getting the signal fast enough to stabilize the hand’s motion, setting off the oscillation. Much more recently, these kinds of oscillations have shown up in negative feedback genetic circuits, where a gene turns itself off. When there is a delay between when the gene is activated and when the protein is produced, the activity of the gene will oscillate with a period that depends on the length of the delay.

One fascinating part of Wiener’s Cybernetics discusses how “black boxes,” systems where we know the input and the output but none of the internal processing, can be synthesized through the connection of multiple “white boxes,” simple systems whose internal workings are pre-defined. While I don’t understand all of the electrical engineering equations involved with Wiener’s boxes, I think that this is a really interesting metaphor for synthetic biology. Synthetic biologists try to design and build biological functions, the “black boxes” of the cell, by combining the “white boxes” of genetic parts with well-characterized functions. Negative feedback loops can make switches, and oscillators, while positive feedback loops can amplify signals to help cells remember.

Cyborg Yeast
Designing a biological negative feedback loop that can keep the production of a protein constant is possible but very challenging, because even the best studied biological “white boxes” can interact with the rest of the cell in mysterious ways. A collaborative project between the labs of Mustafa Khammash and John Lygeros at ETH Zurich and Hana El-Samad at UCSF attempted to outsource this feedback loop to a computer, to keep the levels of yellow fluorescent protein constant in their engineered yeast cells. These “cyborg yeast” cells were kept in the dark with a computer monitoring the level of yellow fluorescence. When the computer sensed that the level of fluorescence was decreasing, it flashed a red light. When it sensed that the fluorescence was too high, it flashed near-infrared light that turned the gene off. With the help of the computer the same cells could maintain the protein level constant at any arbitrarily chosen level, a task that would be nearly impossible with “traditional” synthetic biology.

These cyborg yeast are a fascinating mixture of cybernetic loops, computer feedback loops controlling biological homeostasis, with applications not only in experimental biology, with the ability to hold a gene constant and study how downstream pathways are affected, but also in biotechnology, where the production of molecules has to be strictly controlled by the conditions of the culture, and maybe someday in medical devices that can control the release of medicines at precise moments. The metaphors and mixtures of computers and biology have always been complex, and the feedback between synthetic biology and computer science will only continue to amplify itself and spin off new and weird cyborgs.

Christina Agapakis About the Author: Christina Agapakis is a biological designer who blogs about biology, engineering, engineering biology, and biologically inspired engineering. Follow on Twitter @thisischristina.

The views expressed are those of the author and are not necessarily those of Scientific American.





Rights & Permissions

Comments 1 Comment

Add Comment
  1. 1. Leukippos_Institute 10:45 am 12/16/2011

    A nice news and views. Thanks Christina Agapakis.
    Very interesting that you mentioned Systems Theory.
    This provides in my opinion the new paradigm in science. We have moved from a reductionistic hypothesis driven science to a data driven scientific concept. Two computer developments are responsible for the paradigm change. The first is the enormous storage capacity in the cloud. We talk about a magnitude in the petabyte scale. One petabyte is equal to one quadrillion bytes or 1000 tetrabytes. This is 10exp15 bytes. Google processes about 24 petabytes per day. The second is that a huge number of computers have been connected and organized in social networks. a huge number of computers are now connected via the Internet. We now have access to the huge data- amount from many locations. In addition: We have twitter, facebook and many other social networks, which give a context for the enormous data-amount.

    These changes have resulted in huge quantities of data and complex systems. We have a data flooding. This a problem normal science cannot solve. The structures of the science world were designed to fit a pre-computer age. The hypothesis method can deal with simple correlations between A and B. But the method fails if the problem becomes more complex with many factors, eg A to I or even more. This is too complex for deducing an empirical consequence.

    A novel theoretical base for science has evolved. Groundbreaking theories have been published:
    As you have mentioned: 1948 – Norbert Wiener: Cybernetics or Control and communication in the animal and machine.
    Moreover interesting is also:
    1955 – William Ross Ashby: Introduction to cybernetics.
    1968 – Ludwig Bertalanffy: General System theory: Foundations, Development, Applications.
    Notable further developments of systems theory are: Heinz Foerster’s second order-cybernetics, Ilya Prigogine’s work on self-organization and his systems theory concepts with thermodynamics, and Mitchell Feigenbaum’s work on chaos theory.
    Systems theory is a fundament for software and thus influences the scientific method. Contemporary applications of systems theory are systems biology and synthetic biology. This philosophical movement has two components: Idealism and systems theory.
    Idealism:
    Idealism is based on Plato’s theory of forms (ideas). In line with this theory, cybernetics assumes that the human nervous system calculates reality. This means:
    Our brain calculates a model of an object.
    The human reception is the basis of the scientific method.
    The key tools are mathematics and logic.
    The result of this calculation is a model, which is not identical to the object.
    Thus, it is impossible to achieve knowledge about the world such as it exists independent of us
    Systems theory:
    In systems theory, contemporary science moves from reductionism to a more holistic position. A system is a set of interacting or independent components forming an integrated whole. Systems behavior involves input, processing and output of data, and can be self-organizing and self-regulating by feedback. Ref http://bit.ly/kjOVs2 and http://bit.ly/lMbKc8

    Link to this

Add a Comment
You must sign in or register as a ScientificAmerican.com member to submit a comment.

More from Scientific American

Scientific American Holiday Sale

Give a Gift &
Get a Gift - Free!

Give a 1 year subscription as low as $14.99

Subscribe Now! >

X

Email this Article

X