Tag Archives: LTEE

You gotta know when to hold ‘em

I was honored and humbled to speak at the Doctoral Hooding Ceremony last weekend at the University of North Carolina at Chapel Hill. I received my Ph.D. there in 1982. It was great to be back in Chapel Hill, seeing some old friends and making many new ones.

There was also one of those interesting small-world connections: UNC Chancellor Carol Folt is an ecologist. I first met Carol when she was an assistant professor at Dartmouth and I was commuting from Amherst, where I was a postdoc, to Dartmouth, to teach evolution as a sabbatical replacement for one semester. Carol is such a positive person, always smiling, and an energetic chancellor.

Anyhow, I had never given a talk like this before, so it was a challenge to prepare. Here’s what I had to say to new doctorates; maybe some of you will find it useful as well.

~~~ ~~~ ~~~

Let me begin by congratulating all of the new PhDs and recipients of other doctoral degrees. Each of you climbed a mountain that no one before you had ever climbed. That’s what made it a doctorate — your original research leading to new knowledge.

My remarks today are about constancy versus change, and about luck versus skill. They turn out to be core themes in the research I do, and they also have a lot to do with life, including the decisions we make in our professional careers.

Speaking of constancy, some things hardly seem to change. I got my degree here in 1982. And who won the NCAA men’s basketball title that year? Yep, it was the Tar Heels, just like this year.

Of course, there have also been a lot of changes since I was a student. Music, for example. When we went to the bar, we had these awesome communal listening devices, called jukeboxes. You didn’t even need headphones to hear the music.

Kool & the Gang’s “Celebration” was hot then — and it’s still a great song if you’ve got a party tonight! Cross-over country music was big, too.

Kenny Rogers had a hit called “The Gambler”, about advice from an old poker player. You’ve probably heard it. It goes like this:

“You got to know when to hold ‘em, Know when to fold ‘em, Know when to walk away, Know when to run.”

Of course, the song is about life, using poker as a metaphor. Just as in our careers and lives, poker requires making decisions in the face of uncertainty.

I had a lot of very good luck at Carolina. I went to a party where I happened to meet Madeleine, a graduate student in the School of Public Health, who is now my wife.

However, I also faced some difficulties, and while I managed to get through them, they led me to change the direction of my research.

I came to UNC to study ecology, which focuses on species and their interactions in nature. I got interested in biology when I took a non-majors course as an undergraduate at Oberlin College, and I saw the sweep of discoveries from molecular biology to vertebrate evolution.

As I contemplated graduate school, I focused on ecology because it was filled with interesting and unanswered questions that, to my naïve self back then, seemed like they wouldn’t be too hard to study.

Many ecologists are superb naturalists, including Nelson Hairston, my advisor here at Carolina, who loved the salamanders he studied, and who knew their biology inside and out.

Or Charles Darwin, who was fond of beetles. On a collecting trip, he already had two beetles he wanted, one in each hand, when he came upon a third that he also wanted to keep. He was so in love with his beetles that he popped one into his mouth to free up a hand. Well, it turns out that the one he put in his mouth was a bombardier beetle. To escape predators, they combine and squirt out two chemicals in an explosive exothermic reaction. Needless to say, Darwin lost all three of those beetles.*

As a kid, I loved being outdoors, hiking and playing sports. But I wasn’t a naturalist; I didn’t know very much about any particular group of animals or plants. At least partly because of that lack of familiarity with organisms in the wild, my first efforts at doing ecological research were failures.

Let me give one example, because it’s kind of funny — at least in hindsight. I tried to do a field experiment using praying mantises. I reared batches of them in the lab from egg cases, and then released them on small plots with two treatments. I had painstakingly cleared the vegetation around each plot by hand to keep the mantises where I put them. Well, the next time I went to see how they were doing, I couldn’t find a single one! Maybe some birds were watching me when I released the mantises, wondering: “What is this crazy guy doing?” before gobbling them up. I have no idea what happened, but that experiment was a total bust.

With hindsight, I was lucky that this project failed right away. The treatment effect I was looking for would probably not have given a significant outcome, even if the mantises had stayed put. So even failures can sometimes be valuable, by keeping us from wasting time—and by forcing us to change direction.

Maybe some of you had failed projects, too, before you found your bearings. It’s a normal part of science and scholarship, though it’s upsetting when it happens.

I had another project that also failed. But this second failure led me to the study system that became my dissertation, which was about the effects of forest cutting and competition on a certain group of insects, called ground beetles.

I loved being outdoors in the mountains of western North Carolina, although the frequent rainstorms often flooded the traps that I used to catch the beetles, drenching both the beetles and me. But this project, at last, was successful, leading to my dissertation and some papers.

But I also had doubts that this line of research was a good fit for my interests and skills. Maybe some of you are at similar points in your career.

I’m sure some of you have found work that you hope to continue for the rest of your life. If so, that’s terrific and more power to you.

Others of you might be pondering or even planning a change—using your degree and experience, but setting off in a new direction. Maybe not right away, but perhaps keeping an eye out for some opportunity that better fits your own skills and interests.

In my case, an exciting opportunity dawned in a graduate reading group, when we read a paper about the coevolution of bacteria and viruses that attack bacteria. Even though I had no experience in microbiology, I wrote the head of that lab with an idea for a project related to the paper, and—lucky for me—he hired me as a postdoc.

Before I started my new position, I was worried about working in an area where, once again, I had no experience. Well, I soon discovered that I enjoyed the work. I wasn’t good at it right away, but I liked the rhythm of a microbiology lab. Unlike praying mantises, the bacteria stayed put in their flasks. Unlike the beetles in the mountains, there weren’t any rainstorms in the lab. And sometimes you could see the results of an experiment the very next day.

Down the road, there were more hurdles. In my first year of looking for a faculty position, I applied for dozens of jobs. I got one interview and no offers. Meanwhile, the grant that funded my research wasn’t renewed, and I had a growing family to support. I even thought about leaving science — and I would have if Lady Luck hadn’t come through for me yet again.

The grant was renewed on the second try, and in my second year on the job market I got two offers. So I headed out to Irvine, California, where I started a project that continues to this day.

The project is an evolution experiment. In fact, the experiment was set up to address the same themes as my talk today—luck and skill, constancy and change—although in a scientific context, rather than a personal one.

In evolution, genetic mutations are random events, while the process that Darwin discovered—adaptation by natural selection, sometimes called “survival of the fittest”—multiplies the best competitors across the generations. I wanted to see how luck and skill—that is, mutation and selection—would play out if we could watch evolution over and over and over.

So I set up 12 populations of E. coli bacteria, all started from the same genetic stock, and I put them in identical flasks, with identical food, the same temperature, etc.

I wanted to know: Would they all change and adapt in the same way, showing the power of natural selection to shape life? Or would each population evolve along a different path, highlighting the importance of random mutation?

One thing that makes bacteria great for this experiment is that we can freeze samples and then later revive them as living cells. In essence, our freezers are time-travel machines for the bacteria, allowing us to directly compare and even compete bacteria that lived at different times.

You’ve all heard about our close relatives, the Neanderthals, who went extinct about 40,000 years ago. Some of you might know that their DNA has been recovered from fossils, allowing their genomes to be analyzed. It’s even been discovered that most of us have stretches of Neanderthal DNA in our own genomes.

But despite these amazing advances, we don’t really know what the Neanderthals were like and how similar they would be to us, if they were raised in our world. How well would they play chess, or music, or basketball? What topics would they choose for their dissertations? What would they talk about if they were at this podium?

Back to the experiment with bacteria: We’ve seen many parallel changes in the bacteria across the 12 replicate populations, showing that natural selection can sometimes make evolution predictable, despite the randomness of mutation. But we’ve also seen differences emerge, including in one lineage a surprising new ability to grow on a resource that other E. coli cannot use. And using new technologies that didn’t exist when the experiment was started, we’ve sequenced hundreds of genomes to find the mutations in samples from across the generations and populations, allowing us to test the repeatability of evolution at the level of the DNA itself.

I sometimes call it “the experiment that keeps on giving.” I originally intended the experiment to run for 2,000 generations, which would take about a year. Well, today it’s been running for almost 30 years, and the bacteria have been evolving for 67,000 generations.

This experiment keeps on giving because the bacteria keep evolving in interesting and sometimes unexpected ways, and because students bring new questions and ideas to the project. My hope is that it will continue long after I’m gone.

While the experiment gets a lot of nice press and compliments these days, there have been some obstacles along the way, as there always are in life and science.

When the first paper was submitted, one reviewer was very negative and even hostile. That reviewer wrote: “I feel like a professor giving a poor grade to a good student” — ouch! — without any suggestions for how to improve it. In fact, the reviewer even wrote: “This paper has merit and no errors, but I do not like it.” Well, I wasn’t going to fold — I liked the cards in this hand. So I wrote a rebuttal, and the paper was accepted. In fact, it went on to receive the journal’s award for best paper of the year.

A second obstacle was one of my own making. I came across another experimental system that I found fascinating, and still do — artificial life in the form of computer programs that can replicate themselves and evolve. At the time, I thought maybe the long-term experiment with bacteria had run its course. Well, unlike in poker, when you face important decisions in your research and career, you can ask other people for advice. It’s a good thing, because I was able to have my cake and eat it, too. Everyone told me: “Don’t end the experiment with bacteria. It’s too valuable.” So my lab has kept it going and it has continued to be a scientific gold mine.

Along the way, some creationists have criticized our work. Some don’t believe our results, while others believe us but say: “See, they’re still only bacteria” — as though any scientist would expect to see worms or monkeys or whatever emerge from this experiment.

There can be many reasons for misunderstandings between scientists and the public: problems of education, politics, and communication. The third problem — communication — is one that we can strive to overcome by explaining our work not only to our close colleagues, but also to the general public.

A couple of years ago I had a wonderful opportunity to communicate science to a broad public audience. I was asked by the producer of “Through the Wormhole with Morgan Freeman” to do a segment about our research on bacteria for that show.

One of the scenes had me playing poker with a few of my students. It shows how the effect of a random event—a particular card in a game of poker—depends on the context in which it occurs. The same is true in evolution. A particular mutation that might be advantageous in one species could be detrimental or even lethal in another.

Let’s have a look**:

“When there was a Queen and a King of Hearts on the table and you have the 10 and Ace of Hearts in your hand, you are set up to potentially make a Royal Flush, the most powerful hand in poker. All you need is for the final card to be the Jack of Hearts.”

I’ve been lucky in life. I was born to parents who nurtured me. I was born in a nation dedicated to life, liberty, and the pursuit of happiness. And like those of you receiving your degrees today, I was fortunate to get a superb education here at Carolina.

The French scientist Louis Pasteur — who in the 1800s disproved spontaneous generation, invented what we now call pasteurization, and developed the first rabies vaccine — said: “chance favors the prepared mind.”

Thanks to your Carolina education, and the hard work that brought you here today, you have a prepared mind. You will encounter many uncertainties, probably some obstacles, and hopefully some terrific opportunities as the cards of life are dealt to you.

Play them well: Know when to hold them, know when to fold them. And sometimes you won’t really know what to do, so you’ll just have to give it your best shot.

Thank you, and congratulations again to all of you receiving your doctoral degrees today.

~~~ ~~~ ~~~

*This story is told in the autobiographical chapter of The Life and Letters of Charles Darwin, edited by his son Francis Darwin. I should have checked the source instead of relying on my memory, as Darwin says he lost only two of the three beetles.  The details of the bombardier beetle’s chemical defense system were worked out in the 1960s by Thomas Eisner and others.

**Thanks to Tony Lund, who produced the television show, for also making the short clip that I showed in my talk. You can see a longer clip here.



Filed under Education, Humor, Science

Optimizing the product of the wow factor and the beneficial mutation supply rate

This post follows up on my post from yesterday, which was about choosing a dilution factor in a microbial evolution experiment that avoids the loss of too many beneficial mutations during the transfer bottleneck.

If we only want to maximize the cumulative supply of beneficial mutations that survive dilution, then following the reasoning in yesterday’s post, we would chose the dilution factor (D) to maximize g Ne = (g2) Nmin = (g2) Nmax / (2g), where Nmax is a constant (the final population size) and D = 1 / (2g). Thus, we want to maximize (g2) / (2g) for g > 0, which gives g = ~2.885 and D = ~0.1354, which is in agreement with the result of Wahl et al. (2002, Genetics), as noted in a tweet by Danna Gifford.

The populations would therefore be diluted and regrow by ~7.4-fold each transfer cycle. But as discussed in my previous post, this approach does not account for the effects of clonal interference, diminishing-returns epistasis, and perhaps other important factors. And if I had maximized this quantity, the LTEE would only now be approaching a measly 29,000 generations!

So let’s not be purists about maximizing the supply of beneficial mutations that survive bottlenecks. There’s clearly also a “wow” factor associated with having lots and lots of generations.  This wow factor should naturally and powerfully reflect the increasing pleasure associated with more and more generations.  So let’s define wow = ge, which is both natural and powerful.  Therefore, we should maximize wow (g2) / (2g), which provides the perfect balance between the pleasure of having lots of generations and the pain of losing beneficial mutations during the transfer bottlenecks.

It turns out that the 100-fold dilution regime for the LTEE is almost perfect!  It gives a value for wow (g2) / (2g) of 75.93.  You can do a tiny bit better, though, with the optimal ~112-fold dilution regime, which gives a value of 76.03.


Filed under Humor, Science


Every day, we propagate the E. coli populations in the long-term evolution experiment (LTEE) by transferring 0.1 ml of the previous day’s culture into 9.9 ml of fresh medium. This 100-fold dilution and regrowth back to stationary phase—when the bacteria have exhausted the resources—allow log2 100 = 6.64 generations (doublings) per day. We round that to six and two-thirds generations, so every 15 days equals 100 generations and every 75 days is 500 generations.

A few weeks ago, I did the 10,000th daily transfer, which corresponds to 66,667 generations. Not bad! But as I was walking home today, I thought about one of the decisions I had to make when I was designing the LTEE. What dilution factor should I use?

If … if I had chosen to use a 1,000-fold dilution instead of a 100-fold dilution, the LTEE would be past 100,000 generations. That’s because log2 1,000 = ~10 generations per day. In that case, we’d have reached a new power of 10, which would be pretty neat. As it is, it will take us (or rather the next team to take over the LTEE) another 14 years or so to get there.

I’ll discuss my thinking as to why I chose a 100-fold dilution factor in a bit. But first, here’s a question for you, which you can vote on in the poll below.

Let’s say that we had done a 1,000-fold daily dilution all along. And let’s say we measured fitness (relative to the ancestral strain, as we usually do) after 10,000 days.  Do you think that the mean fitness of the evolved populations subjected to 1,000-fold dilutions after 100,000 generations (on day 10,000) would be higher or lower than that of the evolved populations subjected to 100-fold dilutions after 66,667 generations (also day 10,000)?

I’ll begin by mentioning a couple of practical issues, but then set them aside, as they aren’t so interesting. First, a 100-fold dilution is extremely simple to perform given the volumes involved (i.e., 0.1 and 9.9 ml). And the LTEE was designed to be simple, in order to increase its reliability. A 1,000-fold dilution isn’t quite as easy, as it involves either an intermediate dilution or the transfer of a smaller volume (0.01 ml), which in my experience tends to be a bit less accurate. Second, the relative importance of the various phases of growth—lag, exponential, transition, and stationary—for fitness would change a bit (Vasi et al., 1994).

Setting those issues aside, here was my thinking about the dilution factor when I planned the LTEE. In asexual populations that start without any standing genetic variation, the extent of adaptive evolution depends on both the number of generations and the supply rate of beneficial mutations. The supply rate of beneficial mutations, in turn, depends on the mutation rate (m) times the fraction of mutations that are beneficial (f) times the effective population size (Ne).

There are many different uses and meanings of effective population size in population genetics, depending on the problem at hand: the question is “effective” with respect to what process? Without going into the details, we would like to express Ne such that it takes into account the expected loss of beneficial mutations during the daily dilutions. To a first approximation, theory shows that the relevant Ne is equal to the product of the “bottleneck” population size right after the dilution (Nmin) and the number of generations (g) between Nmin and the final population size during each transfer cycle (Lenski et al., 1991).

The final population size in the LTEE is ~5 x 108 cells (10 ml x 5 x 107 cells per ml), and it is the same regardless of the dilution factor, provided that the bacteria have enough time to reach that density between transfers. The 1,000-fold dilution regime would reduce Nmin by 10-fold relative to the 100-fold regime, although the 50% increase in the number of generations per cycle would offset that reduction with respect to the effective population size. Nonetheless, Ne would be ~6.7-fold higher in the 100-fold regime than in the 1,000-fold regime.

The greater number of generations in 10,000 days under the 1,000-fold regime would also increase the cumulative supply of beneficial mutations by 50%. Nonetheless, the extent of adaptive evolution, which is (under this simple model) proportional to the product of the elapsed generations and Ne, would be ~44% greater under the 100-fold dilution regime than the 1,000-fold dilution regime. So that’s why I chose the 100-fold dilution regime … I was more interested in making sure we would see substantial adaptation than in getting to a large number of generations.

Now you know why the LTEE has only reached 67,000 or so generations.

Of course, I could also have chosen a 10-fold regime, and by this logic the populations might have achieved even higher fitness levels. I could also have chosen a much higher dilution factor; even with a 1,000,000-fold dilution the ancestral strain could double 20 times in 24 h, allowing them to persist. Or at least they could persist for a while. With severe bottlenecks, natural selection becomes unable to prevent the accumulation of deleterious mutations by random drift, so that fitness declines. And if fitness declines to the degree that the populations can no longer double 20 times in 24 h, then the bacteria would go extinct as the result of a mutational meltdown.

Returning to the cases where the bottlenecks are not so severe, the theory that led me to choose the 100-fold dilution regime ignores a number of complicating factors, such as clonal interference (Gerrish and Lenski, 1998; Lang et al., 2013; Maddamsetti et al., 2015) and diminishing-returns epistasis (Khan et al., 2011; Wiser et al., 2013; Kryazhimskiy et al., 2014). It’s predicated, I think, on the assumption that the supply rate of beneficial mutations limits the speed of adaptation.

When the LTEE started, I had no idea what fraction of mutations would be beneficial. I think it was generally understood that beneficial mutations were very rare. But the LTEE and other microbial evolution experiments have shown that beneficial mutations, while rare, are not so rare as we once thought, especially once an experiment has run long enough (Wiser et al., 2013) or otherwise been designed (Perfeito et al., 2007; Levy et al., 2015) to allow beneficial mutations with small effects to be observed and counted.

So I think it remains an open question whether my choice of the 100-fold dilution regime was the right one, in terms of maximizing fitness gains.

And that makes me think about redoing the LTEE. OK, maybe not starting all over, as we do have a fair bit invested in the last 29 years of work. But maybe expanding the LTEE on the fly, as it were. We could, for example, expand from 12 populations to 24 populations without too much trouble. We’d keep the 12 original populations going, of course, but we’d spin off 12 new ones in a paired design (i.e., one from each of the 12 originals) where we changed the dilution regime. What do you think? Is this a good idea for a grant proposal? And if so, what dilution factor would you suggest we add?

Feel free to expand on your thoughts in the comments section below!

Note: See my next post for a bit more of the mathematics, along with a tongue-in-cheek suggestion for combining the effects of the beneficial mutation supply rate and a “wow” factor associated with having lots of generations.


Gerrish, P. J., and R. E. Lenski. 1998. The fate of competing beneficial mutations in an asexual population. Genetica 102/103:127-144.

Khan, A. I., D. M. Dinh, D. Schneider, R. E. Lenski, and T. F. Cooper. 2011. Negative epistasis between beneficial mutations in an evolving bacterial population. Science 332: 1193-1196.

Kryazhimskiy, S., D. P. Rice, E. R. Jerison, and M. M. Desai. 2014. Global epistasis makes adaptation predictable despite sequence-level stochasticity. Science 344: 1519-1522.

Lang, G. I., D. P. Rice, M. J. Hickman, E. Sodergren, G.M. Weinstock, D. Botstein, and M. M. Desai. 2013. Pervasive genetic hitchhiking and clonal interference in forty evolving yeast populations. Nature 500: 571-574.

Lenski, R. E., M. R. Rose, S. C. Simpson, and S. C. Tadler. 1991. Long-term experimental evolution in Escherichia coli. I. Adaptation and divergence during 2,000 generations. American Naturalist 138: 1315-1341.

Levy, S. F., J. R. Blundell, S. Venkataram, D. A. Petrov, D. S. Fisher, and G. Sherlock. 2015. Quantitative evolutionary dynamics using high-resolution lineage tracking. Nature 519: 181-186.

Maddamsetti, R., R.E. Lenski, and J. E. Barrick. 2015. Adaptation, clonal interference, and frequency-dependent interactions in a long-term evolution experiment with Escherichia coli. Genetics 200: 619-631.

Perfeito, L., L. Fernandes, C. Mota, and I. Gordo. 2007. Adaptive mutations in bacteria: high rate and small effects. Science 317: 813-815.

Vasi, F., M. Travisano, and R. E. Lenski. 1994. Long-term experimental evolution in Escherichia coli. II. Changes in life-history traits during adaptation to a seasonal environment. American Naturalist 144: 432-456.

Wiser, M. J., N. Ribeck, and R. E. Lenski. 2013. Long-term dynamics of adaptation in asexual populations. Science 342: 1364-1367.


Filed under Science

What Was I Thinking?

The LTEE has run for over 10,000 days and almost 67,000 generations. It’s time to shut it down, as of today.

It’s been a hell of a lot of work, and we have almost nothing to show for it. As some astute commentators have noted around the web, the creatures in the flasks are still just bacteria—creatures, just as they were created.

If you read the first LTEE paper*, you’ll see we predicted the bacteria should become yeast by about 5,000 generations, nematodes at 15,000 generations or so, and fruit flies by 30,000 generations, maybe 35,000 at the outside.

After that, we’d have to stop the experiment anyhow, because we wouldn’t be able to freeze and bring them back alive any longer.

Plus, we’d have to get IRB approval for human experimentation if we ran it much past 50,000 generations.

Well, we’ve given the LTEE all this time, and still … they ’re just bacteria. I guess we’ve proven that Charles Darwin was wrong after all.

As an astute reviewer pointed out when we submitted that first paper, “I feel like a professor giving a poor grade to a good student.”  I should’ve listened and quit way back then. It would’ve saved everyone a lot of time and effort.

Now it’s going to be a hell of a lot of work next week emptying the freezers and autoclaving all those samples.

* Lenski, R. E., M. R. Rose, S. C. Simpson, and S. C. Tadler. 1991. Long-term experimental evolution in Escherichia coli. I. Adaptation and divergence during 2,000 generations. American Naturalist 138: 1315-1341.


Filed under Humor, Uncategorized

Some Wrinkles in Time

Today is another milestone for the E. coli long-term evolution experiment—the LTEE, for short. I did the 10,000th daily transfer today at about noon.

REL doing LTEE transfer 10,000 with Neerja keeping a close eye on me

[Yours truly, doing the 10,000th LTEE transfers. Technician Neerja Hajela is keeping a close eye on me, and with good reason. Photo by Thomas LaBar.]

Some of you will remember we just celebrated the LTEE’s 29th birthday a few weeks ago, on February 24th. And if you’re quick with math, you might be thinking: “Wait a second: 29 years times 365 days per year is a lot more than 10,000 days. Have Lenski and his team screwed up?”

The answer is both yes and no. Let me explain.

The LTEE began on February 24, 1988 [1, 2].

From February 24, 1988, to March 13, 2017, equals 10,609 days on which we could have done transfers. But we’ve only had 10,000 transfers. What happened to those other days?

In short, the bacteria spent the 609 “lost” days in a freezer at –80°C or in a refrigerator at 4°C.

One chunk of days was lost when the LTEE was moved from my lab at UC-Irvine, where I started the experiment, to MSU, where it is today. Moving a lab is difficult: it requires moving people, moving equipment and materials, often renovating space, obtaining new supplies and equipment, hiring new people, and trouble-shooting and otherwise getting everything organized to resume work [3].

We lost 191 days from April 8, 1992, when the 10,000-generation samples went into the freezer at UCI, to October 16, 1992, when the LTEE restarted from the frozen samples at MSU.

Most of the other days have been lost as a result of various accidents. I’m often asked, when I give talks on the LTEE, how we’ve kept the experiment going so long without contamination, broken flasks, equipment failure, etc.

The short answer is that we haven’t. Many accidents have happened along the way.

There are 3 main types of accidents, each of which involves a different sort of interruption and recovery.

Little mistakes: Sometimes a flask has a hairline crack; when you take it out of the incubator the next day, there’s just a puddle of salt on the bottom. Or maybe someone knocked over a flask while doing the daily transfers. In cases like these where a mistake occurs that is immediately recognized, we go back in time (and lose) one day.

How do we do that? Each day, after the transfers have been made, we don’t immediately discard the previous day’s cultures. Instead, we put them in a refrigerator, where we can use them to restart the experiment after these little mistakes. The bacteria have finished growing long before each day’s transfer, so they are in stationary phase, and their metabolic activity is even lower sitting there at 4°C. Restarting the populations from the refrigerated cultures is a perturbation, of course, but a tiny one in the scheme of things.

When these little mistakes happen to one population, we go back a day for all the populations. We do that so that the rhythm of the experiment, which involves quality-control checks and freezing samples at regular intervals, is the same for all of the populations.

Bigger slipups: Another sort of problem can occur if the entire experiment is compromised in a way that is not immediately recognized. For example, the autoclave might not be working properly, and we realize that bottles of media that we’ve been using for a few days are contaminated. In that case, the cultures stored in the refrigerator won’t help us.

But we don’t have to start the LTEE all over at t = 0. (If we did, then the experiment wouldn’t be here today!) Instead, we go back to the last time that we froze samples, just like we did when we restarted the experiment after the move from UCI to MSU. Importantly, we restart the LTEE from whole-population samples, not individual clones, so that we do not lose the diversity that is present in an evolving population.

Of course, moving the bacteria into and out of the freezer is a perturbation, involving the addition of a cryoprotectant, freezing the cells, thawing them, and re-acclimating them to the conditions of the LTEE. Still, it happens only occasionally. Moreover, all of the samples used in competitions or other assays go into the freezer, come out, and are re-acclimated to the relevant conditions before measurements are made.

Dreaded cross-contamination: The third kind of accident is when bacteria from one LTEE population “migrate” into another population. That’s not supposed to happen, because it compromises the statistical independence of the populations, which are units of replication on which many analyses rest. I worried about this issue before I started the LTEE, because one of the central questions that motivated the experiment is the reproducibility of evolution. And I’m glad I worried about it. Fortunately, there was a pretty easy way of dealing with this concern from the outset.

Six of the 12 populations started from cells of an ancestral strain, REL606, that cannot grow on the sugar arabinose; they are phenotypically Ara. The others started from cells of a mutant, REL607, that can grow on arabinose; these populations are Ara+. There is no arabinose in the LTEE environment, and the mutation that allows growth on arabinose has no measurable affect on fitness in that environment. However, when Araand Ara+ cells grow on Tetrazolium Arabinose (TA) agar in a petri dish, they make red and white (or pink) colonies, respectively.


[Mix of Araand Ara+ colonies on TA agar.]

The arabinose phenotype serves two important purposes in the LTEE. First, we use it to estimate the abundance of competitors in the assays we perform to measure relative fitness. To that end, we typically compete an evolved Ara population sample against the Ara+ ancestor, and vice versa. Second, with respect to the possibility of cross-contamination, we alternate Ara and Ara+ populations during the daily transfers. The idea is that, if an accidental cross-contamination does occur, it will likely involve adjacent populations and lead to cells that have the wrong phenotype (i.e., produce the wrong-colored cells on TA agar) in a population. So we check each population for that phenotype whenever we freeze samples.

When we find one or more cells that produce the wrong-colored colony, we have to figure out what to do. There are various additional checks that we can perform, especially nowadays when DNA sequencing has allowed us to discover many mutations—additional markers—that uniquely identify each population. In particular, these extra markers have, in recent years, let us distinguish between “false alarms” (new mutations that affect colony color on the TA agar) and actual cross-contamination events. In any case, when we’ve had suspected or confirmed cross-contamination events, we restart the invaded population from the previous sample [4]. We then typically monitor that population by plating samples periodically on TA agar, to make sure it didn’t have a low frequency of cross-contaminating invaders even before that earlier sample was frozen. As a consequence of restarting invaded populations, some of the LTEE populations are 500 generations (or multiples thereof) behind the leading edge.

So today’s 10,000th daily transfer applies to some, but not all, of the LTEE populations.

Despite these precautions and procedures, I worried that somehow we had slipped up and there were undetected cross-contamination events. Maybe there had been an especially fun party one Friday night … and on Saturday someone forgot the protocol and transferred all six red Ara populations in a row before moving on to the six white Ara+ populations. In that case, a cross-contamination might occur but not be detected. So I was thrilled when we sequenced hundreds of genomes from different generations of the LTEE populations and there was no evidence of any cross-contamination. Have I mentioned all the terrific people who have worked with me?

One of the unsung heroes of the LTEE is my technician and lab manager, Neerja Hajela. She has worked with me for over 20 years now, and she’s probably done more daily transfers than everyone else combined.

Neerja Hajela 13-Mar-2017

[Neerja Hajela, technician and lab manager extraordinaire.]

By the way, there were not 12, but 15, flasks in the trays while I was doing the transfers. What’s going on with that?

Flasks LTEE day 10,000

[The 15 LTEE flasks in the incubator.]

One of the extras is a blank—a culture without bacteria. If the medium in that flask is turbid the next day, then “Houston, we have a problem.” Another of the extras is a population we’re calling Ara–7. It was spun off population Ara–3 after we discovered—many thousands of generations later—that one lineage in that population had gone extinct for some reason that we do not understand. You can read more about that here. Ara–7 doesn’t count as one of the “real” LTEE populations, but it might prove useful in comparison with Ara–3 at some point in the future.

And the third extra? Remember what I said about cross-contamination? Well, we recently discovered a cross-contamination event in which cells that made red colonies on TA agar were found among the white-colony-forming cells of the Ara+1 population. Postdoc Zachary Blount confirmed they weren’t new mutants that made the wrong-colored colonies in Ara+1; instead, those cells had specific mutations that showed they came from population Ara–1, meaning they were cross-contaminating invaders.

Zachary Blount 13-Mar-2017

[Zachary Blount, aka Dr. Citrate.]

So we restarted Ara+1 from its previous frozen sample, monitored it by plating cells on TA agar, and … alas, up came some more of those red invaders. It’s interesting, in a way, because Ara–1 is one of the most fit LTEE populations, while Ara+1 is the very least fit, which means Ara+1 is especially susceptible to invasion from its Ara–1 neighbor in the daily transfers. Anyhow, we then restarted Ara+1 going back in time 1000 and 1500 generations—hence, the extra flask—and we will monitor those for a while by plating samples on TA agar. If neither of them shows any sign of invaders for several weeks, then we will continue only the one with the fewer “lost” generations and drop the other.

There’s one other little issue related to keeping time in the LTEE. Every day, we remove 0.1 mL from each flask culture and transfer it to 9.9 mL of fresh medium. That 100-fold dilution allows the bacterial population to grow 100-fold before it depletes the available resources. And that 100-fold growth corresponds to log2 100 ≈ 6.64 generations. But we round it up a tad to 6.67 generations, so that every 15 transfers equals 100 generations [5].

In any case, our fielding percentage (baseball jargon for the ratio of plays without errors to total chances on defense) is 10,000 / 10,609 ≈ 0.943. If we exclude the lost days associated with the move from UCI to MSU, then the percentage rises to 0.960. Not bad, not bad at all. Did I mention the terrific people who have worked, and are working, on the LTEE?

This post’s title is a play on the novel A Wrinkle in Time by Madeleine L’Engle.

[1] I first started the LTEE on February 15, 1988, but I then restarted it on February 24, because I got worried that the first arabinose-utilization mutation I had selected, which serves as a neutral marker, wasn’t quite neutral.

[2] So the LTEE experienced a leap day in its very first week!

[3] I was fortunate that three experienced graduate students—Mike Travisano, Paul Turner, and Farida Vasi—moved to MSU even before I did to help set up the lab, and that our research was allowed to continue in my UCI lab—led by technician Sue Simpson and John Mittler, who was finishing his PhD—after I moved in late December, 1991.

[4] To keep all the populations in sync with respect to the freezing cycle, we restart the others at the same time, too. Of course, for the others, we don’t go back in time—we use the latest sample, where the cross-contaminated population was discovered during the quality-control checks associated with the freezing cycle.

[5] In fact, 6.67 generations per day might be a slight underestimate given the possibility of turnover during stationary phase. Moreover, every lineage with a beneficial mutation that sweeps to fixation goes through more than the average number of generations, since each mutant lineage starts as one cell among millions.


Filed under Science, Uncategorized

Who Knows Where the Time Goes

Today is the 29th birthday of the long-term evolution experiment (LTEE). As I wrote on Twitter: “May the cells live long & prosper, both in & out of the -80C freezers.” I hope they—and the rest of the world—will be evolving and improving long after I’m gone.

Anyhow, after my tweet, Luis Zaman asked for a picture of me on my own 29th birthday. (I started the LTEE when I was 31.) Alas, I don’t have one. But I’ve found some pictures from around that time—including just before and after I moved to UC-Irvine to start my first faculty position, and over the next few years up to about the time I started the LTEE.

Summer, 1985: This photo is from Amherst, Massachusetts, where I did my postdoc with the amazing Bruce Levin, who hosted a goodbye party for us. From left to right: Ralph Evans, a brilliant graduate student and dear friend, who died tragically just a few years later of brain cancer. My beautiful wife, Madeleine. Our one-year-old daughter Shoshannah, being held by forever-young Bruce. Yours truly, holding our three-year-old son Daniel. And Miriam Levin, an art historian.


October, 1985: Shoshannah on my shoulders at the San Diego Zoo, a few months after we moved to Irvine.


March, 1986: First-year faculty member burning the midnight oil in our Las Lomas apartment at UCI. Working on a paper? Or getting ready to teach 700 students the next day? (Two sections of Ecology, a required course for Bio Sci majors, with an hour to recuperate in between. It was well worth it, though, because one of the students in one of the many quarters I taught that course was the great Mike Travisano.)


October, 1986: Moving up in the world, we bought a new house on Mendel Court in University Hills. My parents visited, and that’s my mother, Jean, a poet who loved science.


March, 1987: The great Lin Chao came for a visit. We grew pea plants on the trellis below the number 6—after all, it was 6 Mendel Court.


June, 1987: One of the fun events at UCI was Desert X (for extravaganza), hosted by Dick MacMillan, the chair of Ecology and Evolutionary Biology, on his property near Joshua Tree National Park. With Madeleine, who is “holding” our Number 3.


June, 1987: Working Xtra hard at Desert X with close friend and colleague Al Bennett.


September, 1987: With an already smiling one-month-old Natalie.


January, 1989: Time for some snuggles. Meanwhile, the LTEE is not quite a year old.


The title of this post is a song by Fairport Convention, with the hauntingly beautiful voice of the late, great Sandy Denny. You should listen to it.


Filed under Science, Uncategorized

Asking for a Skeptic Friend

I sometimes get email from people asking, in one way or another, whether our long-term evolution experiment (LTEE) with E. coli provides evidence of evolution writ large – new species, new information, or something of that sort. I try to answer these questions by providing some examples of what we’ve seen change, and by putting the LTEE into context. Here’s one such email:

Hi Professor Lenski,

I have a quick question. I’m asking because I am having a discussion with someone who is skeptical of evolution. The question is: Over the 50,000 generations of e-coli has any of the e-coli evolved into something else or is it still e-coli?

I am a non-religious person who likes to think of myself as an adherent to science but I am not sure how to respond to my skeptic-friend.

Thank you!

And here’s my reply:

Hello —-,

50,000 generations, for these bacteria, took place in a matter of ~25 years. They have changed in many (mostly small) ways, and remained the same in many other respects, just as one expects from evolutionary theory. Although these are somewhat technical articles, I have attached 3 PDFs that describe some of the changes that we have seen.

Wiser et al. (2013) document the process of adaptation by natural selection, which has led to the improved competitive fitness of the bacteria relative to their ancestors.

Blount et al. (2012) describe the genetic changes that led one population (out of the 12 in the experiment) to evolve a new capacity to grow on an alternative source of carbon and energy.

Tenaillon et al. (2016) describe changes that have occurred across all 12 populations in their genomes (DNA sequences), which have caused all of them to become more and more dissimilar to their ancestor as time marches on.

Best wishes,


Although these articles were written for other scientists, they make three big points that I hope almost anyone with an open mind can understand.

  • We see organisms adapting to their environment, as evidenced by increased competitiveness relative to their ancestors.
  • Against this backdrop of more or less gradual improvement, we occasionally see much bigger changes.
  • And at the level of their genomes, we see the bacteria becoming more and more different from their ancestors.

In these fundamental respects, evolution in these flasks works in much the same way that evolution works in nature. Of course, the scales of time and space are vastly greater in nature than they are in the lab, and natural environments are far more complex and variable than is the simple one in the LTEE. But the core processes of mutation, drift, and natural selection give rise to evolution in the LTEE, just as they do (along with sex and other forms of gene exchange) in nature.


Filed under Education, Science