biology is technology

Rob Carlson :: Ch 4

"I must tell you that I can prepare urea without requiring a kidney of an animal, either man or dog.”*  With these words, in 1828 Friedrich Wohler announced he had irreversibly changed the world.  In a letter to his former teacher Joens Jacob Berzelius, Wohler wrote that he had witnessed, “The great tragedy of science, the slaying of a beautiful hypothesis by an ugly fact.”  The beautiful idea to which he referred was vitalism, the notion that organic matter, exemplified in this case by urea, was animated and created by a vital force and that it could not be synthesized from inorganic components.  The ugly fact was a dish of urea crystals on his laboratory bench, produced by heating inorganic salts.  Thus was born the field of synthetic organic chemistry.

Around the dawn of the 19th century, chemistry was in revolution right along with the rest of the western world.  The study of chemical transformation, then still known as alchemy, was undergoing systematic quantification.  Rather than rely on vague and mysterious incantations, scientists such as Antoine Lavoisier wanted to create what historian of science and technology Bruce Hevly calls an “objective vocabulary” for chemistry.  Through careful measurement, a set of clear rules governing the synthesis of inorganic, non-living materials gradually emerged.

In contrast, in the early 1800s the study of organic molecules was primarily concerned with understanding how molecules already in existence were put together.  It was a study of chemical compositions and reactions.  Unlike the broader field of chemistry taking shape from alchemy, making new organic things was of lesser concern because it was thought by many that organic molecules were beyond synthesis.  Then, in 1828, Wohler synthesized urea.  Suddenly, with one experiment, the way scientists did organic chemistry changed.  The ability to assemble organic molecules from inorganic components altered the way people viewed a large fraction of the natural world because they could conceive of building much of it from simpler pieces.  Building something from scratch, or modifying an existing system, requires understanding more details about the system than simply looking at it, poking it, and describing how it behaves.  This new approach to chemistry helped open the door to the world we live in today.  Products of synthetic organic chemistry dominate our environment, and the design of those products is possible only because understanding the process of novel assembly revealed new principles.

It was this step of moving to Synthetic Chemistry, and then to an engineering of chemistry, which radically changed the way people understood chemistry.  Chemists had to learn rules that weren’t apparent before.  In the same way that Chemical Engineering changed our understanding of nature, as we begin engineering biological systems we will learn considerably more about the way biological pieces work together.  Challenges will arise that aren’t obvious just from watching things happen.  With time, we will understand and address those challenges, and our use of biology will change dramatically in the process.  The analogy at this point should be clear; we are well on our way to developing Synthetic Biology.

Before going further, it is worth noting that this is not the original incantation of the phrase “synthetic biology”.  Whatever the reception this time around, the first time it was a flop.  In her history of the modern science of biology, Making Sense of Life, Evelyn Fox Keller recounts efforts at the turn of the 20th Century to discover the secret of life through construction of artificial, and synthetic, living systems; “To many authors writing in the early part of the [20th] century, the [path] seemed obvious: the question of what life is was to be answered not by induction but by production, not be analysis but by synthesis.”(Keller, p.18)  This offshoot of experimental biology reached its pinnacle, or nadir, depending on your point of view, in attempts by Stephané Leduc to assemble purely physical and chemical systems that demonstrated behaviors reminiscent of biology.  As part of his program to demonstrate “the essential character of the living being”(ibid, p.28) at both the sub-cellular and cellular level, Leduc constructed chemical systems that he claimed displayed mitotic division, growth, development, and even cellular motility.  He described these patterns and forms in terms of the well-understood physical phenomena of diffusion and osmotic pressure.  It is important to note that these efforts to synthesize life-like forms relied as much on experiment as upon theory developed to describe the relevant physics and chemistry.  That is, this was a specific program to use physical principles to explain biological phenomena.  These efforts were described in a review paper at the time as “La Biologie synthetique”(ibid, p.31-32).

While the initial reception to this work was somewhat favorable, Leduc’s grandiose claims about the implications of his work, and a growing general appreciation for complicated biological mechanisms determined through experiments with living systems, led to something of a backlash against the approach of understanding biology through construction.  By 1913, one reviewer wrote, “The interpretations of M. Leduc are so fantastic…that it is impossible to take them seriously”(ibid, p.31).  Keller chronicles this episode within the broader historical debate over the role of construction and theory in biology.   History regards the folks in the synthetic camp, and related efforts to build mathematical descriptions of biology, particularly in the area of growth and development, as poorly regarded by their peers.  Perhaps inspired by the contemporaneous advances in physics, it seems that the mathematical biologists and the synthetic biologists of the day pushed the interpretation of their work further than was warrented by available data.

In response to what he viewed as theory run rampant, Charles Davenport suggested in 1934 that, “What we require at the present time is more measurement and less theory…There is an unfortunate confusion at the present time bewteen quantitative biology and bio-mathematics…Until quantitative measurement has provided us with more facts of biology, I prefer the former science to the latter”(ibid, p.86).  I think these remarks are still valid today.  Leduc, and the approach he espoused, failed because real biological parts are more complex, and obey different rules, than his simple chemical systems, however beautiful they were.  And it is quite clear that vast forests have been felled to publish theory papers that have little to do with the biology we see out the window.  But theory, drawn from physics, chemistry, and engineering, does have a role to play in describing biological systems.  Resistance to the tools of theory has been, in part, cultural.  There has always been a certain tension in biology over the utility of mathematical and physical approaches to the subject;

To put it simply, one could say that biologists do not accept the Kantian view of mathematics (or, rather, mathematization) as the measure of a true science; indeed, they have often actively and vociferously repudiated any such criterion.  Nor have practicing biologists shown much enthusiasm for the use of mathematics as a heuristic guide in their studies of biological problems.(Keller, p. 81)

Fortunately, this appears to be changing. Mathematical approaches are flourishing in biology, particularly in the interpretation of large data sets produced by genomic and proteomic studies.  Physicists and engineers are making fundamental contributions to the quantitative understanding of how individual proteins work in their biological context.  But I think it is important to acknowledge that not all biologists think a synthetic, bottom up, approach will yield truths applicable to complex systems that have evolved over billions of years.  Such concerns are not without merit, because as the quotation from Charles Davenport suggests, biology has traditionally had more success when driven by good data rather than theory.  The challenge today is to build quantitatively predictive design tools based on the measured device physics of real biological parts, and to implement designs within organisms in ways that work in the real world.

The modern reincarnation of the effort to build living systems has many faces in many locations around the world.  From Stanley Miller’s 1953 synthesis of amino acids under conditions thought to be common throughout the universe, to modern efforts to expand upon terrestrial biochemistry and thereby change not just the sequence but also the content of the genetic code by including new amino acids, to engineered proteins and genetic circuits, Synthetic Biology is in full ferment, and the phrase itself has slowly reemerged in scientific and popular literature.  The most explicit, and nuts and bolts, efforts are lead by a group at MIT.  From the SyntheticBiology.org website:

Synthetic biology refers to both (a) the design and fabrication of biological components and systems that do not already exist in the natural world and (b) the re-design and fabrication of existing biological systems.

Biologists began the process of creating a modern Synthetic Biology three decades ago, with the advent of recombinant DNA technology.  Splicing a gene from one organism into an unrelated organism is a step loosely analogous to Wohler’s synthesis of urea: it is the fundamental capability required to build synthetic biological systems and thereby begin determining design rules.

After exploring the development of aviation technology in the last chapter, it is natural to ask what it will take to build biological systems “to spec” in the same way we now build airplanes.  We have seen that the differences between the two kinds of systems are enormous.  There are vastly many more moving parts in a cell than in an airplane, and we simply don’t yet know how to fly a cell.  As with early aviation, biological technology is in the process of moving from qualitative stories to quantitative models. 

The design of airplanes is clearly a highly mature field compared to where we are with designing biological systems.  There are also fundamental differences in the starting components we have to work with.  Unlike early aviation, where all the components were fabricated from scratch, building new biological systems currently requires using molecular components that have been evolving for several billion years.  The main difficulty caused by this requirement is that we just don’t yet know how all those components work.  One reason for such ignorance is that the physics governing the behavior of molecules is different than the physics we are used to at much larger length scales.  The mechanical intuition we have for gears, motors, beams, and cantilevers cannot be simply transferred to molecules.  The molecular scale of the components also means they are very hard to interact with directly, which complicates testing models through experiment.  Fundamental aspects of basic components used to turn genes on and off, and make proteins, are still mysterious.  As explored in Chapter 2, designing biological systems just isn’t like imagining a new Lego structure.  But even given our inadequate knowledge of biological details, the design philosophy still has utility.  This is where cartoons, based on generalizations and abstractions of molecular details, can be of some assistance.

Cartoons are used in chemistry to describe the states of atoms and molecules, and if you know how to read them can help predict the outcomes of chemical reactions.  To push this further, in high energy physics diagrammatical systems have been developed that allow interactions between particles to be predictively, and conveniently, visualized, and that can also be used to calculate.  Feynman diagrams, initially developed by Richard Feynman, are cartoons that have strict mathematical interpretations.  There is an algebra – a grammar – governing manipulation of the diagrams and their translation into mathematics. 

Cartoons are already playing a role in understanding biology, whether they illustrate molecular interactions or the information content of genes.  With a reasonable knowledge of genetics and biochemistry, gene and protein sequences can be read as if they were English.  Chemical structures may help figure out how two biological molecules will interact.  Where phenomena are not directly observable via microscopy, cartoons are often used to suggest models of cellular organization and function. 

But at this cartoon level it is much more complicated to predict the outcomes of assembling new biological structures, be they molecules or organisms, than it is even to envision what a new Lego construction will look like.  In addition to there being many more kinds of biological Legos than there are plastic bricks, the biological building bricks tend to change their shape and color every time two bits are snapped together.

With this in mind, I will proceed with a “Lego” or cartoon level description of molecular biology that will certainly miss important details, but that will suffice for my immediate purposes**.

The ubiquitous cartoon representation of DNA is a ladder twisted around its long axis into a double helix that makes one revolution about every 10 rungs.  The rungs of that ladder are molecules that constitute genetic information.  These molecules are often abbreviated by their first letters of their names, A, T, G, and C.  There are two complimentary base pairs to a rung; A’s paired with T’s, G’s with C’s.  A human genome consists of approximately three billion of these bases, separated into twenty-three pairs of chromosomes packed into the nucleus.  Each chromosome is a contiguous molecule of DNA, many millions of bases long, wherein the bases are organized into genes.  Genes are about a thousand bases long, though many genes do not consist of contiguous bases, and many genomes contain long stretches of DNA that do not appear to contain any genes whatsoever.

Information output from the genome begins with gene expression.  A polymerase molecule reads a gene, proceeding in a particular direction along one of the strands, transcribing DNA into RNA.  The cartoon function of RNA is to transfer information from genes in the nucleus to a cell’s protein making machinery.  A gene is delineated from the rest of the letters in a chromosome by short sequences of bases at either end that represent “Start” and “Stop” to the polymerase, telling the polymerase where to start transcribing and where to stop.  This is, effectively, the molecular definition of a gene – it is the letters between Start and Stop that are read by the polymerase, transcribed into RNA, and eventually translated into proteins.  At this stage of the Genome Project, this is in fact the only criterion for identification of many genes.  Little more is known about most human genes because it is much easier to read a gene than to determine its function.  Indeed, we have not yet associated any function with most of the human DNA sequences identified via the molecular definition of “gene”.

The word “gene” has another meaning in studies of Mendelian heredity, in which the notion of a gene is used to keep track of traits traced through generations.  Classical genetic studies of hair and eye color, or diseases such as breast cancer and sickle cell anemia, demonstrate that for many traits there is a basic unit of heredity passed from one generation to the next.  A major triumph of modern biology is the identification of some heritable traits with the information content of a sequence of bases, that is, with a single gene.  Diseases such as sickle cell anemia, cystic fibrosis, and some breast cancers are each attributable to a single gene.  These cases are the easiest to identify, the low hanging fruit of molecular medicine.  Many traits, however, seem to be determined by several genes, or the interaction of several genes.  We are some distance away from understanding these complexities.  Until then, most researchers have their hands full sorting out what individual genes do when expressed.

Returning to the cartoon description of gene expression, a polymerase transcribes DNA into RNA, which is in turn translated into protein by a ribosome.  Regulatory sequences of DNA adjacent to genes can influence how often, and under what conditions, they are expressed.  Part of the “Central Dogma” of molecular biology, as Francis Crick called it in 1957[1], is that information flows from DNA, to RNA, to protein.  A working knowledge of the Central Dogma is enough to understand the elementary details of most current developments in Synthetic Biology.

This rest of this chapter utilizes the above cartoon description of basic molecular biology, and is built around three examples of designed biological systems that demonstrate various challenges in 1) defining the device physics of molecular components, 2) building and simulating models, 3) using those components, and, most importantly, 4) building new biological entities based on quantitatively predictive designs.

Some synthetic biological circuits are more sensitive to this uncertainty than others.  In the first example below, the circuit operates as predicted by simulations in large part because the design was insensitive to unknown device physics.  Constructed with a small number of unengineered moving parts, it was designed to maintain a stable response to a signal after the signal ended.  It is perhaps the simplest example of a biological system built to respond to its environment with a stable change.

The second example is a circuit that was less predictable precisely because it assumed a more complete understanding of certain molecules than existed at the time.  It was designed to exhibit predictable, periodic oscillation.  This behavior required a more complex simulation than the first example, and the simulation revealed that the design required modified proteins.

The third and final example involves building a design tool for a large number of parts in two complex organisms.  It is an attempt to model viral infection at the level of single cells, an audacious first step towards building something akin to a “flight model” for a biological systems.

...

1.         Crick, F., Symposium of the Society of Experimental Biology, 1957. 12: p. 138-163.

*http://classes.yale.edu/02-03/chem220a/studyaids/history/lighter%20side/wohler%20and%20urea.htm 

the letter: http://classes.yale.edu/chem125a/125/history99/4RadicalsTypes/UreaLetter1828.html

Nature 407, 677 (12 October 2000); A victim of truth, SUNETRA GUPTA, http://www.nature.com/cgi-taf/DynaPage.taf?file=/nature/journal/v407/n6805/full/407677a0_r.html&filetype=&dynoptions=

**The best cartoon description of biology is quite literally a cartoon.  Larry Gonick’s Cartoon Guide to Genetics is a masterful description of molecular biology written at a level that is easily accessible to non-scientists.  Dog-eared copies can also be found on the bookshelves of many professionals.