“Let’s stay together.” – Al Green

Some of the biggest questions in evolutionary biology deal with the origin of life. For example, if I go back one generation, I find my parents. Two generations, my grandparents. Ten generations are human beings who may or may not have looked like me. Five hundred thousand are, oh, I don’t know. Maybe a bipedal hominid? Anyway, if we continue going backward like this, we inevitably get to time zero and encounter some big-time questions that can really cause a brain to cramp up.

One of these major questions that can cause someone to drool on their shirt in amazement of evolution is the transition of life from unicellular, sovereign entities to cooperative multicellular organisms. A recent paper by Ratcliff et al. (2012) from the University of Minnesota posits that the first step towards multicellular organisms is cellular clustering; they then proceed to evolve clustering in unicellular yeast and ask questions about the clusters.


Premise: Bigger things settle in solution faster than smaller things.

(Oversimplified) Materials: Unicellular yeast (Saccharomyces cerevisiae), test tubes, solution that the yeast can eat, time

Step 1: Suspend unicellular yeast in solution in a test tube.

Step 2: Wait 45 minutes.

Step 3: Transfer the cells at the bottom of the tube to a new tube with fresh solution.

Step 4: Return to Step 2 60 times.

Step 5: Look in microscope.

How to go from one cell to multicellular clusters. Photos from Figure 1 of Ratcliff et al. (2012).

That on the right is the snowflake phenotype (their name for the multicellular clusters); every experimental replicate that Ratcliff et al. did produced it. The selection imposed by (1) gravity and (2) time was strong enough for the yeast to repeatedly evolve the snowflake (they experimentally determine the snowflake had a 34% fitness advantage over the unicellular phenotype).

The first major question the authors wanted to answer was how the clusters were forming. Were they unrelated cells sticking together (aggregation) or was a single cell producing baby cells without fully dividing (thus forming clusters of genetically related cells or “postdivision adhesion”)? Turns out, it’s the latter – the cells within a snowflake are related. The reason this makes pretty good evolutionary sense is discussed below.

Some other major findings first: the snowflakes produced baby snowflakes (instead of unicellular offspring, which some organisms apparently do).  Also, the snowflake phenotype was stable even when the selective pressures were removed – they didn’t revert to populations of single celled yeast when transferred 35 times without gravitational selection. Furthermore, the authors found snowflakes had a juvenile phase and an adult phase. This means that reproduction was delayed until a minimum size was reached. Additionally, the snowflakes had determinate growth – once they got to a certain size, they grew no more.

When the gravitational selection was strengthened, the snowflakes settled faster. The strong-selection snowflakes grew larger, contained more cells and delayed the adult phase. Because stronger selection caused these phenotypic changes in the snowflakes, the authors conclude that selection is acting on the snowflake as a single unit instead of acting on the cells that comprise the snowflake.

Another aspect of multicellularity the authors investigate is that of “division of labor”. In humans, different cells in our bodies do different things. How did THAT evolve? (Would you call it cooperation?) The authors show that the snowflakes do a version of cellular differentiation. Some of the cells undergo apoptosis (cell death) for the benefit of the whole. This is a pretty interesting concept, if you ask me, and it’s the reason that the cells being related makes sense. In order for division of labor to be favored in this way (some cells forgoing reproduction and instead undergoing cell death), two things must occur. Cells that die must be genetically related to cells that reproduce and the benefit of the death must exceed the cost of one more cell not reproducing. The authors use propagule size and number to determine if these conditions are met; the reason some cells undergo apoptosis is to create breaks in the snowflake that allow the baby snowflakes to be released into the world. This is adaptive because it allows more babies to be made than if the snowflakes used traditional cell divisions.

In summary: Selection for larger size (as determined by gravity) caused multicellular clustering to evolve in unicellular yeast. The resulting “snowflakes” were clusters of genetically related cells that produced little baby snowflakes, sometimes at the expense of individual cells dying. And the kicker: seems like it was pretty easy, right? I’m like two supplies and a billion dollar microscope short of doing this in my kitchen tonight!

 Ratcliff WC, Denison RF, Borrello M & Travisano M. 2012. Experimental evolution of multicellularity. PNAS (Early Edition). DOI: 10.1073/pnas.1115323109

3 comments on ““Let’s stay together.” – Al Green

  1. Crazy cool stuff!
    Do yeast have iron in them? If you had a sharp magnetic field gradient once you have snowflake phenotypes, could you get a phenotype that tends to have higher iron concentration?

  2. Interesting idea, Luke—I don’t know anything about yeast’s baseline abilities in this regard, but it’d make sense that if you provided iron in the growth medium, then selected “survivor” colonies using a magnet as you suggest, eventually you’d get a strain that was really good at accumulating iron.

  3. […] Bjørn Østman on Michael Behe. Also represented: recent work from this very blog, by Noah Reid and Sarah Hird. Go check it out! Share and enjoy:Like this:LikeBe the first to like this post. This entry was […]

Comments are closed.