BEACON Researchers at Work: The genetic basis of biofilm formation

This week’s BEACON Researchers at Work blog post is by University of Washington graduate student Elyse Hope.

Elyse Hope“Remember to finish your full course of antibiotics” is a phrase we have probably all heard from a doctor at least once. Intuition tells us that a long course of antibiotics is designed to completely eliminate any pathogens from our bodies, making sure we don’t have any left over that might become drug-resistant. The information that is left out, however, is how a pathogen might evolve drug resistance. One way in which microbes (bacteria, yeast, and others) might survive an antimicrobial or antibiotic treatment is to form a biofilm, which is the trait I study in yeast. The premise of a biofilm is this: microbes can create proteins on their cell surfaces and even outside the cell, allowing the microbes to stick to surfaces and to each other in a mass (Verstrepen et al, 2004). When exposed to a stressor like an antimicrobial or other chemical treatment, cells on the outside of the biofilm may die, but cells on the inside of the biofilm might survive (Smukalla et al, 2008) and – once the stress has passed – repopulate. Drug-resistant microbes pose an increasing healthcare problem, and part of combating this problem is understanding the genetics behind antimicrobial resistance and how those genetics contribute to the types of biofilms we see. The primary goal of my research is to better understand the genetic basis of yeast biofilm formation and how the ability to form a biofilm evolves in yeast.

In Maitreya Dunham’s lab at the University of Washington, I am using many different strains of budding yeast Saccharomyces cerevisiae (the same yeast that makes bread, wine, and beer) to broadly investigate the genetics underlying biofilm formation. Most of what we know about yeast biofilms comes from working with a few well-studied laboratory strains, but we haven’t known until now whether these lab strains are representative of what we would see in yeast in the wild, including strains of yeast involved in infection. Recent work from a group in France (Liti et al, 2009) generated a collection of wild yeast from different sources all over the world, from Ethiopia to Malaysia to Pennsylvania, and from wine to cactus to palm tree nectar. I wanted to know: if we look at biofilms formed by these wild strains, do we see the same characteristics predicted by laboratory strains?

There are many different ways to look at yeast biofilms established in the literature, from how well the yeast stick to each other and to surfaces, to how complex they look when they grow together in a colony (Granek and Magwene, 2010; Stovicek et al, 2010). We studied five different visible traits related to biofilms in yeast and showed that most of these traits are uncorrelated. This means that whether a yeast strain can stick to other cells (trait 3) has little to do with how well it can form a “lacy” colony (trait 1) or stick to a surface (trait 5).


We also found that each of these wild strains has an entirely unique set of traits, both in strength and complexity, and that the traits are different depending on whether a cell is haploid or diploid (“ploidy”, how many copies of its chromosomes it has), which is a more complex picture of the relationship between ploidy and biofilm traits than was previously known from laboratory strains (Galitski et al, 1999; Reynolds and Fink, 2001).

Our next step is to look at changes in single genes and how they specifically contribute to each of these traits. My final goal is to make it possible for us to engineer strains of yeast with very specific biofilm characteristics, and to know what gene variants to look for as potential targets for small molecules. If we could find targeted ways to disrupt biofilms, then the microbes inside could be much more susceptible to drug treatments.

For more, see our paper here.


I have come to feel passionate about the work I am doing, and believing in the larger goal helps me meet the smaller goals of day-to-day effort. When I started graduate school, however, I had never even seen a yeast cell before. I started college with a computational background and planned to be a physics major, but genetics drew me in with its blend of logical problem-solving and real-world applications, and I declared a biology major instead. My senior year I joined a lab that had just started a collaboration with a genomics group. My PI sent me a paper about genome sequencing to see if I would be interested in going that route with my research. I had never read anything more interesting in my life; I was astounded by what had been accomplished with sequencing so far, as well as the implications for what could be accomplished in the future. I later entered a Genome Sciences graduate program so I could work on realizing those possibilities. I had every intent of focusing on sequencing and staying computational, but I had an amazing conversation with a PI who worked on yeast and envisioned a very real-world project on biofilms that would integrate genetics, sequencing, and bench work. She took a chance on me that I would love yeast as much as she did, without any experience whatsoever, and she was right. 



Galitski, T., A. J. Saldanha, C. A. Styles, E. S. Lander and G. R. Fink, 1999 Ploidy Regulation of Gene Expression. Science 285: 251-254.

Granek, J. A., and P. M. Magwene, 2010 Environmental and genetic determinants of colony morphology in yeast. PLoS Genet 6: 1-12.

Liti, G., D. M. Carter, A. M. Moses, J. Warringer, L. Parts et al., 2009 Population genomics of domestic and wild yeasts. Nature 458: 337-341.

Reynolds, T. B., and G. R. Fink, 2001 Bakers’ yeast, a model for fungal biofilm formation. Science 291: 878-881.

Smukalla, S., M. Caldara, N. Pochet, A. Beauvais, S. Guadagnini et al., 2008 FLO1 is a variable green beard gene that drives biofilm-like cooperation in budding yeast. Cell 135: 726-737.

Stovicek, V., L. Vachova, M. Kuthan and Z. Palkova, 2010 General factors important for the formation of structured biofilm-like yeast colonies. Fungal Genet Biol 47: 1012-1022.

Verstrepen, K. J., T. B. Reynolds and G. R. Fink, 2004 Origins of variation in the fungal cell surface. Nat Rev Microbiol 2: 533-540.

For more information about Elyse’s work, you can contact her at ehope at u dot washington dot edu.

Posted in BEACON Researchers at Work | Tagged , , , , , , | Comments Off

BEACON Researchers at Work: Evolution and the nano-scale

Today’s BEACON Researchers at Work blog post is by NC A&T faculty Dr. Joseph L. Graves Jr.

Dr. Joseph L. Graves Jr. Associate Dean for Research & Professor of Biological Sciences Fellow, American Association for the Advancement of Science Section G: Biological Sciences Joint School of Nanoscience & Nanoenginneering North Carolina A&T State University & UNC Greensboro

Dr. Joseph L. Graves Jr.
Associate Dean for Research & Professor of Biological Sciences;
Fellow, American Association for the Advancement of Science
Section G: Biological Sciences;
Joint School of Nanoscience & Nanoenginneering,
North Carolina A&T State University & UNC Greensboro

One nanometer is defined as 1 x 10-9 meter. This is about the size of one glucose molecule. The nucleus of a human cell is about 600 nm’s across and the average bacterium is about 1000 nm’s in size. Clearly many important biological phenomena occur at the nanoscale, including molecular self-assembly. The rate of the development of nanotechnology over the last twenty years has been astounding. Its applications are all around us. From sunscreens and dental adhesives, to the components of high technology devices. Nanotechnology is being proposed to build bioactive and biodegradable scaffolds for tissue engineering and for controlled drug delivery to treat chronic diseases.

The global production of nanomaterials (NMs) is expected to grow exponentially. It is estimated that global production of NMs will reach 104–105 tons annually for structural applications by 2020. Other projections include: 103 tons in skin care products, information communications technology industries >103 tons, biotechnology 10 tons, and environmental industry 103–104 (source Royal Society and Royal Academy of Engineering Report, 2004). Yet little has been done to study the ecotoxicity of NMs, particularly issues such as bioaccumulation in food chains; or impacts of NMs on bacteria and other microorganisms. Indeed, metallic and metallic oxide nanoparticles are already being touted as the miracle “cure” for multi-drug resistant bacteria (Rai et al. 2012). This means that bacteria will be increasingly exposed to metallic/metallic oxide NPs and other NMs both intentionally (as antimicrobial applications) and unintentionally (run-off from industrial processes.)

Reading these glowingly optimistic reports from those working with engineered nanoparticles (eNPs) was disconcerting. There were two major flaws in this thinking. First was the idea that bacteria would be widely susceptible to noble metals (copper, silver, gold, etc.). Second was even if this were currently true, that bacteria would not rapidly evolve resistance to them. It turns out that neither assumption is true. Bacteria have an array of resistance mechanisms to heavy metals in general and silver in particular (Silver and Phung 2006; Mijnendonckx et al. 2013). It turns out that essentially all bacteria have genes for toxic metal ion resistances. Amongst those best studied include those for Ag+, AsO2-, AsO4(3-), Cd2+ Co2+, CrO4(2-), Cu2+, Hg2+, Ni2+,Pb2+, TeO3(2-), Tl+ and Zn2+. The largest group of resistance systems functions by energy-dependent efflux of toxic ions. Fewer involve enzymatic transformations (oxidation, reduction, methylation, and demethylation) or metal-binding proteins (for example, metallothionein SmtA, chaperone CopZ and periplasmic silver binding protein SilE). Some of the efflux resistance systems are ATPases and others are chemiosmotic ion/proton exchangers (Silver and Phung 2006).

Evolution of Nanoparticle Resistance

To test the second assumption, I decided to utilize a relatively “naïve” bacterium Escherichia coli K12MG1655 to determine how rapidly this strain could evolve increased silver nanoparticle resistance. This strain of E. coli did not have any of the sil (silver resistance) genetic elements in its genome. However, like most bacteria its genome normally contains heavy metal sensing genes such as the cus system. The Cus system aids in protection of cells from high concentrations of silver and copper. The histidine kinase CusS of the CusRS two-component system functions as a silver/copper responsive sensor kinase and is essential for the induction of the genes encoding the CusCFBA efflux pump. The efflux pump works by removing the toxic concentrations of the metal from the interior of the bacterial cell.

Scanning electron microscope picture of E. coli bacterium with AgNPs associated with cell wall (picture by M. Tajkarimi, JSNN).

Scanning electron microscope picture of E. coli bacterium with AgNPs associated with cell wall (picture by M. Tajkarimi, JSNN).

Our experiment was simple. We conducted an experimental evolution protocol using “off the shelf” E. coli K12 MG1655 and exposed it to increasing concentrations of 10nm citrate-coated spherical nanoparticles. We cultured the cells using Davis Minimal Broth with dextrose 10% as a sole carbon source, enriched with thiamine hydrochloride 0.1% in 10 ml of total culture volume maintained in 50 ml Erlenmeyer flasks. The flasks were placed in a shaking incubator at 37o C for 24 hours. This is generally considered a non-stressful growing media for E. coli. The cultures were propagated daily by transfers of 0.1 ml into 9.9 ml of DMB. The control populations were maintained in this medium without the addition of silver nanoparticles (AgNPs). The treatment populations were exposed to increasing concentrations of spherical 10nm citrate-coated AgNPs. Both the control and treatment groups were replicated five-fold. In this way we could determine if any of the mutations that arose in the control or treatment groups were shared by more than one of its populations.

After determining the minimum inhibitory concentration (MIC) for this strain of E. coli, we exposed the treatment group to a concentration less than MIC so that some bacteria could survive. We wanted to allow enough survivorship such that a sufficient number of bacteria were left in the culture with potential silver-resistant mutations. In the non-exposed bacteria we would normally observe a 2-log increase in the culture over 24 hours (this is about 6.5 generations resulting in 106 founders growing to about 108 per ml of culture). At the beginning of the day the cultures would look “clear” but by the end of the day they would be “turbid” to the eye. Our goal was to observe the same sort of growth in the silver nanoparticle treated populations. This happened rapidly.

The table below shows the number of generations in this experiment that were kept at a given AgNP concentration:

Generations Exposure Concentration
1 – 50 50 mg/l
51 – 140 100 mg/l
141 – 265 125 mg/l

Thus by generation 50 (~ 9 days later!) we were observing turbid cultures in the AgNP treated group exposed to 50 mg/L of 10nm AgNPs. After an additional 90 generations we were observing turbidity in 24 hours for the treated group at 100 mg/L of 10nm AgNPs and so on.

Undergraduates Quincy Cunningham and Herve Nonga explain MIC experiment results to Dr. Chandra Jack at Annual BEACON Congress, 2014.

Undergraduates Quincy Cunningham and Herve Nonga explain MIC experiment results to Dr. Chandra Jack at Annual BEACON Congress, 2014.

We tested the control and treatment groups at generation 250 for population growth over 24 hours in a range of concentrations of 10nm citrate-coated AgNPs, 10nm PVP-coated AgNPs, 40 nm citrate-coated AgNPs, 40 nm PVP-coated AgNPs, and bulk silver nitrate (AgNO3). Both controls and treatments were able to grow at 50 and 100 mg/L, but not surprisingly, the treatment populations showed superior growth compared to the controls at 250 mg/L, 500 mg/L, and 750 mg/L. In other words, the treatment group was now AgNP resistant, relative to the control bacteria.

Genomics of Resistance

The experimental evolution protocol used in this study indirectly demonstrated that genomic changes must have occurred between the control and treatment bacteria. Next generation sequencing was used to investigate the genomic changes more directly. This process is facilitated by the fact that E. coli K12MG1655 has already been fully sequenced, allowing us to compare the genomic features of both our control and treatment populations against a reference genome for this bacterium. Thus we sequenced the “off-the-shelf” MG1655, our controls, and our treatments and compared their genomes to the reference using the breseq bioinformatic pipeline (developed by the Dr. Jeffrey Barrick, U. Texas – he is another BEACON scientist.) We sequenced the ancestral bacteria (off-the-shelf), generation 100 controls and treatment; generation 150 controls and treatment; and generation 200 controls and treatment via Illumina sequencing technology. Their sequences are then trimmed of Illumina adaptors, aligned to the reference genome, and genetic variants called against the reference genome. The E. coli genome contains ~4.7 million base pairs – so this is not an unsubstantial computational task.

Going into the sequencing, I had expected to see a large number of genetic differences between the controls and the treatment populations. In fact, the results showed that not only did AgNP resistance evolve quickly, but that it didn’t take a great deal of genomic changes to achieve the result! The breseq pipeline allows one to investigate point mutations (SNPs), and deletions, insertions, insertion elements (indels). The genomic story was told mainly by three point mutations! As we are still finishing this study, I will not hang my hat on these results as of yet. However, stay tuned…the nature of science requires that we check our results. What we can say at this point is that it seems relatively easy for bacteria to evolve resistance to metallic nanoparticles (as they did to traditional antibiotics.) Care should be utilized before we intentionally and accidentally introduce these nanomaterials into our ecosystems. That is because evolution is always in action, even where you don’t suspect it.


Gudipaty, S.A. and McEvoy, M.M., The histidine kinase CusS senses silver ions through direct binding by its sensor domain, Biochim Biophys Acta 1844(9): 1656-61, 2014 doi: 10.1016/j.bbapap.2014.06.001.

Mijnendonckx, K., Leys, N., Mahillon, J., Silver, S., Van Houdt, R., Antimicrobial silver: uses, toxicity and potential for resistance, Biometals 26(4): 609-21, 2013. doi: 10.1007/s10534-013-9645-z.

Rai, M.K., Deshmukh, S.D., Ingle, A.P., and Gade, A. K., Silver nanoparticles: the powerful nanoweapon against multidrug-resistant bacteria, J. Appl. Microbiol. 112(5): 841-52, 2012. doi: 10.1111/j.1365-2672.2012.05253.x.

Silver, S. and Phung, le T. A bacterial view of the periodic table: genes and proteins for toxic inorganic ions, J. Ind. Microbiol. Biotechnol. 32(11-12): 587-605, 2005.


This research project would not have been possible without the work of the following individuals; Mehrdad Tajkarimi, graduate student, Nanoscience Department, UNC Greensboro; Quincy Cunningham, undergraduate student, NCATSU; Herve Nonga, undergraduate student, Michigan State University; Adero Campbell, undergraduate student, Bennett College; and Dr. Scott Harrison, Biology Department, NCATSU.

Posted in BEACON Researchers at Work | Tagged , , , , , | Comments Off

BEACON Researchers at Work: Climate effects on algae… and undergrads

This week’s BEACON Researchers at Work blog post is by MSU graduate students Jakob Nalley and Danny O’Donnell, with University of Texas undergraduate Farhana Haque.

From left to right: Jakob Nalley, Farhana Haque, Dr. Elena Litchman, Danny O'Donnell

From left to right: Jakob Nalley, Farhana Haque, Dr. Elena Litchman, Danny O’Donnell

Go ahead, take a deep breath, and let it out. Almost half of the oxygen you breathed in came from the phytoplankton, or algae, that live in nearly every body of water on the planet with access to light (even your fish tank). Although they are extremely small, phytoplankton are a fundamental component of the biosphere, forming the base of aquatic food webs, fixing large amounts of atmospheric carbon, and photosynthetically producing nearly half of the oxygen we breathe. Phytoplankton live in an environment that is ever-changing through space and time, so to persist they must be able to respond to rapid environmental change (e.g. temperature, nutrient availability, grazer density). Global climate change has potentially large direct and indirect effects on the diversity and abundance of phytoplankton around the world, with far-reaching consequences for organisms at higher trophic levels. However, it is unclear how phytoplankton may respond plastically (in the short term) or evolutionarily (on longer timescales) to increased temperatures and nutrient inputs, and how these changes might then reverberate through the ecosystem.

Dr. Elena Litchman’s lab focuses, in part, on investigating how global climate change may influence phytoplankton physiology, competitive ability, biomass production, and community structure. This summer, through the BEACON funded Research Experience for Undergraduates (REU) program at the Kellogg Biological Station (KBS), we were extremely fortunate to have the opportunity to continue this temperature work with an outstanding undergraduate, Farhana Haque, from the University of Texas – Austin. Farhana worked on two separate research projects, both focusing on how temperature influences the ecology and evolution of phytoplankton species and communities.

Photo of Litchman stocks to illustrate algal diversity

The REU program aims to expose young scientists to unique research experiences to which they may not otherwise have access at their home university. It is apparent from the success of her summer research (and its contribution to furthering the research goals of our lab) that Farhana’s experiences at KBS were unique in a number of ways. The most striking being she had to travel over one thousand miles to get her first taste of ecological and evolutionary research. After gaining a better understanding of the applications eco-evolutionary research can have, through the work we do in the Litchman lab, the simply “classical science” (as Farhana once described it) developed into a rich, very much modern, and extremely attractive science.

KBS has a unique way of cultivating enthusiasm for ecological research. It is a small field station where professors’ doors are always open, and graduate students are eager to interact with anyone that is willing to listen. Through some interesting eco-evolutionary research and tremendous opportunities to interact with a number of eminent scientists, this REU experience was clearly transformative for Farhana, and ecological and evolutionary research are now firmly on her radar.  Here are some of Farhana’s reflections on her research experience at KBS.

Apparently, undergraduates aren’t much different than algae. When I first arrived at KBS in Hickory Corners, Michigan at the beginning of summer I was forced to bundle up in the balmy 70-degree weather. As a native Texan, I handled three-digit temperatures without a sweat, but froze in the Gull Lake breeze. While I put on a jacket, algae also employ different strategies to deal with their changing environment. And with rising temperatures, the algae in the ocean will have to adapt to their environment.

Farhana at the KBS Undergraduate Research Symposium (8/6/14)

Farhana at the KBS Undergraduate Research Symposium (8/6/14)

This summer, as a BEACON REU student in the Litchman lab at KBS, I measured the change in cell-size in algae as a response to temperature. I worked under the guidance of two graduate students, Jake Nalley and Danny O’Donnell. At first, it was a bit confusing. Two different people working on two different doctoral theses. How could I, an undergraduate from another school, fit in? With the help of my mentors and our PI Dr. Elena Litchman, we carved out a project by using a major theme in ecological systems: evolution. By tracking the change of cell-size with temperature change, I was able to see significant trends. With short-term growth algae tends to shrink as it heats up and we attributed our change to plasticity, trait changes without genetic modification. Over long-term growth at a warmer temperature, algae cell-size is expected to shrink (though my tenure here was too short to see genetic change in our algae, you’ll have to go to Danny for updates!).

Evolution turned out to be my cohesive force, which I feel made my experience so much more satisfying. A common complaint I hear among my peers entering into the world of research (me included) is that our minimal projects are dead-ends. Collaboration is key to escaping the rut. As a biochemistry major, I felt like I brought more to the table when my mentors and I would modify experiments. UT has amazing computer scientists and statisticians, while MSU has a unique field station that brings the natural world to the lab bench. Danny and Jake saw nuances that ecologists at Michigan State would appreciate, while I saw my own brand of chemistry in everything. With a healthy dose of enthusiasm and hard work, we were able to work together and discover some exciting trends in algae. 

**For more about Jake’s work visit his website or email him at nalleyja at msu dot edu

**For more about Danny’s work email him at odonn146 at msu dot edu

Posted in BEACON Researchers at Work | Tagged , , , , , | Comments Off

BEACON Researchers at Work: Evolving Complex Traits

This week’s BEACON Researchers at Work blog post is by University of Idaho postdoc Matthieu Delcourt. 

putterThe evolution of complex traits is one of the major enigmas in evolutionary biology. While we know a lot about phenotypic variation within populations, much remains to be explored about the genetic causes and consequences of trait complexity. We know that the genome is composed of many genes that are located on same or different chromosomes and that their proximity influences their respective evolutionary histories across many generations. We also know that these same genes can interact together by repressing, activating or regulating protein expression. Taken together, these types of interactions between genes can affect phenotypes dramatically and this may also vary between individuals. Quantitative traits such as body size, life history traits or stress tolerance are good examples of traits that often show a great amount of variation within a population, and are also good examples of how genetic interactions can therefore influence the process of adaptation. Quantitative traits depend on the action of many interacting genes with small effects that often generate genetic correlations between multiple phenotypes. Those interactions can happen either at the chromosomal level or the cellular level and can either constrain or increase the nature of genetic variation that is relevant to selection.

Figure 1In Dr. Paul Hohenlohe’s lab at the University of Idaho, we explore the effects of gene network structure on the adaptation to stress tolerance in the budding yeast (Saccharomyces cerevisiae). We are interested in linking phenotypic to genomic patterns of variation within the context of quantitative genetics. These are very different approaches that gather information at various levels and that are often difficult to combine in a series of experiments. We have started a project looking in particular at the effect of size and degree of independence of gene networks involved in stress tolerance in salt, glycerol, and copper sulfate. We take advantage of the yeast as a microorganism to perform experimental evolution, phenotypic and genomic analyses in a polymorphic population that we created in our laboratory. This population has shown a large amount of phenotypic and genetic variation in stress response, meaning that the underlying genetic architecture differs between individual cell lines within that population. Interestingly, the level of connectivity between networks seemed to directly affect the genetic correlation between traits. Take the example of two osmotic stressors (salt and glycerol): the cellular response will involve very similar sets of genes (that are tightly connected) to equilibrate the osmotic pressure inside the cells in either case. On the other hand, when the gene networks involved are very distinct, for example when one network is involved in the cellular response to an osmotic stress (salt) and the other to an oxidative stress (copper sulfate), then very different sets of genes are regulated. We found that in the first case (see figure above), the stress responses were genetically correlated, meaning that an individual cell that had the machinery to deal with salt could also deal with glycerol very well. In the second case, there was no genetic correlation between the stress responses, meaning that these network were to some extent free to evolve independently genetically. One can wonder how this fits into the broader picture of trait complexity. This is actually a very exciting result that extends previous finding on multivariate phenotypic evolution to the level of gene-gene interaction. Now that our field gathers lots and lots of genetic data, we need to make sense of the effects of all of these genes and their interactions altogether on the phenotypes. Can we predict traits in individuals given we know something about their genotypes? Many human diseases, genetic syndromes or psychological disorders indeed depend on one, two or more gene networks that are inter-connected and these are often complex traits that vary tremendously within and among populations.

We are currently exploring where in the genome this variation in stress response originates and how many mutations are responsible for this variation. Do mutations tend to hit at the center or the periphery of a gene network? How does the degree of independence between gene networks constrain the evolution of stress response in yeast? These are some of the questions that we ask in our lab and that we hope will encourage research at the interface between phenotypic, quantitative genetic and genomic analyses.

For more information about Matthieu’s work, you can contact him at mdelcourt at uidaho dot edu.

Posted in BEACON Researchers at Work | Tagged , , , , , , , | Comments Off

BEACON Researchers at Work: The Invisible Hand of Evolution

This week’s BEACON Researchers at Work blog post is by MSU postdoc Jeff Morris.

IMG_0270As the 18th century dawned over Europe, pretty much everybody believed the world was as it was because of a mysterious divine plan. But during the period known as the Enlightenment, thinkers began working out the mechanics that structured both nature and human society. By century’s end many of the fundamental processes of physics and chemistry had been elucidated. But the greatest leap ahead in the history of biology – the discovery by Charles Darwin of evolution by natural selection – wouldn’t happen until the middle of the 19th century, and it would owe as much to the study of economics as to anything in the preceeding century’s life sciences.

Adam Smith
, an 18th century Scottish philosopher, was struck by how economics obeyed natural-like laws despite the often capricious and irrational behavior of economic actors. Smith envisioned an intrinsic interrelation between producers and consumers that pulled economic behavior toward certain norms as if “led by an invisible hand.” The net result of many financial actors behaving selfishly is a well-regulated, self-organized system: the parts don’t have any intention of working together, but screw up and do it anyway.

Young Charles Darwin was influenced by many, many Enlightenment thinkers, but it’s hard not to notice how similar the self-organization of nature by natural selection is to Smith’s ideas about economics. This Spectator article from 2009 covers Darwin’s economic influences in depth. It also uses one of my favorite weird words, “defenestrate”:

“Ideas evolve by descent with modification, just as bodies do, and Darwin at least partly got this idea from economists, who got it from empirical philosophers. Locke and Newton begat Hume and Voltaire who begat Hutcheson and Smith who begat Malthus and Ricardo who begat Darwin and Wallace. Before Darwin, the supreme example of an undesigned system was Adam Smith’s economy, spontaneously self-ordered through the actions of individuals, rather than ordained by a monarch or a parliament. Where Darwin defenestrated God, Smith had defenestrated government.” – Matt Ridley

Marketplace analogies are common in biology: here are two dealing with microbial evolution. And when applied to human social organization, evolution often draws the same criticisms as free market policies: can “nature red in tooth and claw” produce cooperation and charity, or must we rely on a benevolent dictator to give us those happy institutions?

The chief problem with evolving cooperation is the tragedy of the commons. Briefly, if cooperation has a cost, then a non-cooperator that can still get the benefits of cooperation will always have a fitness advantage over cooperators. Theoretically, this advantage will always exist even if the breakdown of cooperation totally trashes the environment. We know the tragedy doesn’t always happen because we see organisms in nature working together – but how does evolution escape it?

Bill Hamilton proved that self-sacrifice could evolve by natural selection if the recipients of the sacrifice were related to the sacrificer – something we’ve come to call kin selection. In the microbial world, kin selection can happen when microbes live in close physical association with each other. Since microbes reproduce by simple division, as long as they don’t move around they’re likely to be surrounded by close relatives. Therefore, if a cell spends resources to make a product that leaks into the environment, the cells most likely to benefit are also close relatives.

Bacteria living in a spatially structured environment like a seabed (left) are more likely to be related to their neighbors than the same organisms living in open, well-mixed water (right). Classic kin selection can happen in the scenario on the left, but not in the one on the right.

Bacteria living in a spatially structured environment like a seabed (left) are more likely to be related to their neighbors than the same organisms living in open, well-mixed water (right). Classic kin selection can happen in the scenario on the left, but not in the one on the right.

But kin selection can’t explain all the cooperation we observe in the microbial world. In fact, we see tons of it in the open ocean, whose turbulent waters are about as close as you’re likely to get to no spatial structure at all. The oceans are full of cells that are dependent on unrelated cells to make crucial metabolites, or in some cases to clean up environmental toxins. In order to understand how this arrangement evolved, we need to consider two theories:

  1. Streamlining theory arose from work with the omnipresent marine bacterium SAR11 done mostly by Steven Giovannoni’s group at Oregon State. It maintains that there is intense selection pressure on microbes to reduce the sizes of their genomes when nutrients are limiting and populations are large. During this process of gene loss, many cells lose the ability to perform vital functions and become dependent on neighboring cells that have retained those functions.
  2. The Black Queen Hypothesis, or BQH, which I originally proposed in my PhD thesis and then worked out more rigorously in a 2012 mBio paper with Erik Zinser and Richard Lenski, proposes that streamlined cells can get away with losing important functions as long as those functions leak their products into the environment. Like players of the card game Hearts want to get rid of the black Queen of Spades because of her high point cost, streamlining cells want to get rid of leaky functions. However, some cells have to end up holding the Black Queen, because once the leaked products are rare enough, additional streamlining moochers won’t have an advantage. BQH evolution thus produces communities of function-performers, or helpers, and moochers, or beneficiaries. These communities have the appearance of cooperation/altruism, but they arise by normal selfish Darwinian natural selection.

We originally proposed the BQH to explain how the marine photosynthetic bacterium Prochlorococcus had become dependent on its neighbors for protection from hydrogen peroxide, a toxin that is constantly produced in natural waters exposed to sunlight. Because peroxides move freely across cellular membranes, any cell protecting itself by breaking down the peroxides also lowers the environmental concentration of peroxide, unavoidably helping any mooching neighbors. This leakiness makes peroxide detoxification a Black Queen function and explains how Prochlorococcus can get away with not protecting itself.

In order to test the predictions of the BQH, we used a wimpy Prochlorococcus-like E. coli mutant that had all of its anti-peroxide defenses knocked out. We then gave this mutant a plasmid – an accessory piece of DNA – that allows the cells to make the peroxide-destroying enzyme KatG and evolved the resulting strain under strong, peroxide-generating light for 1,200 generations, or about 150 daily transfers of the cultures into fresh batches of growth medium.

Percentage of the evolving E. coli populations retaining the KatG plasmid. Values are means of 3 replicately evolved populations; error bars are standard deviations.

Percentage of the evolving E. coli populations retaining the KatG plasmid. Values are means of 3 replicately evolved populations; error bars are standard deviations.

Even though plasmid-free cells could barely survive on their own under the lights, they nevertheless evolved and stably co-existed with their plasmid-containing ancestors throughout the 1,200 generations (see above). Moreover, while the helpers underwent a number of evolutionary changes, there was no evidence that they were trying to be “stingy” with their production of KatG in order to outcompete the beneficiaries. Just the opposite – the evolved helpers made more KatG than their ancestors.

Family tree of genes that mutated in more than one of the three evolved populations.

Family tree of genes that mutated in more than one of the three evolved populations.

Not only did the two forms coexist, there’s also evidence that they diverged into different “species.” When we sequenced the genomes of helper and beneficiary clones taken from three replicate evolved lines, we found two mutated genes common to all helper lines, and a completely different set of two mutated genes in all beneficiary lines (see above). This indicates that the early choice to become a beneficiary fundamentally changed the adaptive landscape of these cells, meaning different mutations are adaptive for helpers than for beneficiaries. This is a barrier to gene flow, and based on the ecological species concept, these two types represent different species (or more properly for bacteria, ecotypes) – evolved in 150 days under a simple BQH regime!

Starting with a single, clonal population, Black Queen evolution produced an ecosystem with 2 distinct types, one of which was apparently altruistically helping the other. Of course, we know that they aren’t helping because they care about the beneficiaries; they just can’t avoid doing so. Each player is maximizing its own selfish advantage at the other’s expense, but the nature of the leaky function prevents either from winning the game and taking sole control of the evolutionary medium.

The BQH thus acts as an invisible hand stablizing the ecosystem and forcing the two types to play nice with each other. Are there any important differences between this kind of inadvertant cooperation and what we might think of as “intentional” cooperation? Maybe, maybe not. Either way, though, it’s clear that the BQH equilibrium can keep varieties of closely related organisms co-existing for a long time, and, unable to get rid of their pesky partner, it’s possible that other forms of cooperation might evolve in time.

Human attempts to manipulate complex systems, be they economies or ecosystems, are often studies in disaster. The invisible hand of the Black Queen shows us one of many ways that blind nature is often a better engineer than any intelligent designer.


Morris, JJ, SE Papoulis, and RE Lenski. 2014. Coexistence of evolving bacteria stabilized by a shared Black Queen function. Evolution.

Morris, JJ, RE Lenski, and ER Zinser. 2012. The Black Queen Hypothesis: evolution of dependencies through adaptive gene loss. mBio 3:e00036-12.

Giovannoni, SJ, JC Thrash, and B Temperton. 2014. Implications of streamlining theory for microbial ecology. ISME J 8:1553-1565.

Jeff Morris is a NASA Astrobiology Institute postdoctoral fellow working in Richard Lenski’s lab at Michigan State. He will be starting as an assistant professor in the Biology Department at the University of Alabama in Birmingham in January. Jeff (@ASDarwinist) blogs about science, politics, and heavy metal at

Posted in BEACON Researchers at Work | Tagged , , , , | Comments Off