I’m happy now to announce what has become of these papers. After positive reviews and revisions, the editor at Science asked the two teams to combine their papers into a single long-format Research Article. Both teams agreed to do so, but it was a lot of work, requiring further reviews and revisions. At last, the combined paper was accepted, and it appears in the latest issue of Science. Hooray!

I think it’s fair to say that the paper provides an unprecedented description and analysis of how the fitness effects of mutations change over time, even under the constant conditions of the LTEE, and even as the overall shape of the DFE was little changed. As a matter of practicality for the high-throughput analyses, the paper relied on generating libraries of insertion mutations. That approach means that the new study focuses on loss-of-function mutations, which are undoubtedly important in the LTEE, but which means that mutations that refine existing functions or even generate new functions are substantially under-represented. Perhaps some future methodology will allow an even more comprehensive analysis, but the new paper provides a great start.

I want to provide a little background, not to the science, but to the interaction among the wonderful people who did the lab work—Alejandro (Alex) Couce and Anurag Limdi along with Melanie Magnan, Siân Owen, and Cristina Herren—and who led the two teams—Michael Baym and Olivier Tenaillon. When I first realized that two teams were preparing papers about the evolution of the DFEs in the LTEE, I had an immediate pang of guilt. When people ask me for LTEE strains, I ask them why they want the samples. That way, I can let them know about potential work that others are doing, which might cause someone’s work to get ‘scooped’ by another team. However, I don’t always remember all the projects, or I might fail to see the connection in the words people use to describe their plans, and people’s projects may also change after they start their work. In any case, thanks to Tanush Jagdish for letting us know that the two teams were pursuing related projects.

After that, the two teams soon began having zoom meetings to openly share and explain their questions, hypotheses, methods, and findings. There was some overlap, of course, but I think that’s valuable in science, because it indicates broad interest in the topic and, moreover, shows that the results are robust. Thus, the decision was made to submit two separate papers as joint submissions to the same journal.  In a thread on social media, Michael nicely described the situation and the resulting comradery (my emphasis added):

As we were working on this we discovered that another group led by Alex Couce and Olivier Tenaillon were working on something very similar.  Instead of competing we decided to talk to one another. While either group could have scooped one another, what we did instead was have some of the most fun scientific calls I’ve ever been part of, in which we discussed our work and our results, and even cross-reviewed each other’s manuscripts. We found that while we were superficially overlapping (TnSeq on the LTEE) the experiments and the questions we asked were complementary. While we’d focused on detrimental mutations after long periods, they looked at early beneficial mutations. Yet we’d found similar trends! In the end, Alex and Olivier’s insights and perspective gave us both a much deeper understanding of the science, and a lot more confidence in the results. As our papers are complementary and speak to one another, we coordinated preprinting and will be co-submitting.

More recently, as we combined the two papers into one, we wanted to leave a ‘footprint’ that would reveal the origins of the integrated study. At the start of the Discussion, we wrote:

This paper began as two separate projects performed by two different teams, using similar but not identical methods. As we discussed our findings together, we discovered that each project reinforced and complemented the other. They reinforce one another by finding the same evolution of the overall form of the DFE; they are complementary because one project delved deeply into the fine-scale genetic changes in the deleterious tail while the other did so for the beneficial tail.

We wondered if the editor might reduce or eliminate this passage, since a paper’s history doesn’t bear directly on its results. We were delighted, therefore, when we saw that the passage remained, because it sheds light on how civility, cooperation, and collaboration can succeed even in the highly competitive world of science.

@article{Couce2024,

title = {Changing fitness effects of mutations through long-term bacterial evolution},

author = {Alejandro Couce and Anurag Limdi and Melanie Magnan and Siân V. Owen and Cristina M. Herren and Richard E. Lenski and Olivier Tenaillon and Michael Baym},

url = {https://www.science.org/doi/10.1126/science.add1417},

doi = {10.1126/science.add1417},

year  = {2024},

date = {2024-01-26},

urldate = {2024-01-26},

journal = {Science},

volume = {383},

number = {6681},

pages = {eadd1417},

abstract = {The distribution of fitness effects of new mutations shapes evolution, but it is challenging to observe how it changes as organisms adapt. Using \textit{Escherichia coli} lineages spanning 50,000 generations of evolution, we quantify the fitness effects of insertion mutations in every gene. Macroscopically, the fraction of deleterious mutations changed little over time whereas the beneficial tail declined sharply, approaching an exponential distribution. Microscopically, changes in individual gene essentiality and deleterious effects often occurred in parallel; altered essentiality is only partly explained by structural variation. The identity and effect sizes of beneficial mutations changed rapidly over time, but many targets of selection remained predictable because of the importance of loss-of-function mutations. Taken together, these results reveal the dynamic—but statistically predictable—nature of mutational fitness effects.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

Why did I use a serial transfer regime in flasks, rather than continuous culture in chemostats? How did I plan to analyze genetic changes before whole-genome sequencing became possible? Why is citrate in the LTEE medium? How do we deal with mistakes that inevitably occur in an experiment that has now run for 35 years? How do we handle requests to share strains? What would I do differently if I was designing the LTEE now?

I always enjoy answering these questions when I give talks, and so I was delighted when Greg Lang and Kerry Geiler-Samerotte asked me to write about these issues and more for a special issue on Experimental Evolution that they are editing for the Journal of Molecular Evolution. 

My paper has now appeared, and you can get a PDF by clicking here. Several other papers for the special issue are also now online, with more still to come. Enjoy the opportunity to learn about old and new experiments, approaches, and discoveries in this exciting field of research.

The image below shows an incubator used in pioneering microbial evolution experiments performed by William Dallinger in the 1880s. It originally appeared in the Journal of the Royal Microscopical Society in 1888.

The work was performed by two outstanding teams:

• Anurag Limdi, Siân Owen, Cristina Herren, and Michael Baym
• Alejandro Couce, Melanie Magnan, and Olivier Tenaillon

Both studies generated high-coverage transposon-insertion libraries in the LTEE ancestor and various evolved isolates. Importantly, the approach allows the identification of the insertion sites of each mutant. They then propagated the mutant libraries in the LTEE environment and tracked the frequencies of all the mutants over time—in essence, bulk fitness assays involving hundreds of thousands of mutations. The figure below, from the paper by Limdi et al., outlines the approach.

Limdi et al. first show that the overall structure of the DFE has hardly changed after 50,000 generations, contrary to some predictions from evolutionary theory.

Limdi et al. also discover that the identity of essential genes has changed substantially, and often in parallel across multiple LTEE lineages.

Couce et al., by contrast, focus their attention on the small tail of the beneficial mutations that drive adaptive evolution. They show that the beneficial tail becomes much smaller over time and approaches an exponential distribution.

Couce et al. also demonstrate rapid turnover in the identity of the genes that harbor potential beneficial mutations.

Together these papers provide an unprecedented description of how the fitness effects of the same mutations can change over time, even under the constant conditions of the LTEE. These changes radically alter the fate of specific mutations, even as the overall genomic and fitness dynamics of the evolving populations follow more predictable trajectories.

Here are links to the two preprints:

• Limdi et al:  https://www.biorxiv.org/content/10.1101/2022.05.17.492023
• Couce et al: https://www.biorxiv.org/content/10.1101/2022.05.17.492360

With the invasion of Ukraine, however, it’s not a day to celebrate. 

The LTEE will move to the capable lab and hands of Jeff Barrick this Spring, after all 12 lines have reached 75,000 generations.

Over the decades, several lines fell behind others due to cross-contamination (or concerns about the possibility), which we detected by examining the alternating Arabinose marker and seeing the resulting colony colors on TA plates. Those lines were then restarted from whole-population samples, but they would be 500 generations behind the others (or a multiple of 500 generations behind in some cases).

The picture above shows red and white colonies growing on TA agar in a Petri dish. The red colonies cannot grow on the sugar arabinose that is part of the TA medium, while the white ones can use arabinose. Half of the LTEE lines started from red colonies (Ara–1 to Ara–6), and half started from white colonies (Ara+1 to Ara+6). We alternate the red and white lines each day during their propagation. That way, if cross-contamination occurs, we can detect it by the presence of bacteria that make colonies that are the wrong color. We check colonies before every periodic freeze of the LTEE. These days, with DNA sequencing, we can also use derived mutations that are unique to each lineage to check whether a putative contamination event is real or not. (Indeed, in some populations, especially those that evolved hypermutability, the colony markers don’t work like they did when the LTEE started.) If we confirm that a cross-contamination event has occurred, we restart the affected population from the last frozen sample of that population.

So today, Devin Lake will propagate the last two lagging populations. Our lab will continue to propagate them until they, too, reach 75,000 generations. The last one should reach that goal in late May.

Zack always adds something about what makes the date special.  He’s an amazing scientist and scholar!

Every day, the E. coli populations in the long-term evolution experiment experience a cycle of renewed resources and growth followed by depletion of their food and then waiting for the next transfer event. 

On a much longer timescale, the LTEE also experiences cycles as it is passed from one scientific generation to the next. This website reflects the beginning of the second scientific generation of the LTEE, as the populations and responsibility for their sustenance will soon pass from my lab to that of the new director, Jeff Barrick.

On this website, you can get an introduction and quick overview of the LTEE including how it works, its goals, some of the key findings, and plans for its future.  You can see a timeline of the experiment with some of the milestones and key events in its history.  You can read, watch, and listen to a few of the news stories about the LTEE.  You can find resources including protocols and links to important datasets.  You can search and find links to the publications that report findings from the LTEE itself as well as descendant experiments that have used the LTEE lines. And last, but not least, you can see the talented people who’ve done and are doing the work behind the LTEE, including propagating the populations, performing analyses, analyzing data, and reporting the findings.

We’ve probably missed some papers, and we know that we’re still missing photos for some participants. We’ve also only scratched the surface of reporting past news.  So please let us know if you find someone or something LTEE-related that you’d like to see included on this website.  For now, enjoy the new beginnings as seasons and generations continue onward!

Link: https://www.nationalgeographic.co.uk/science-and-technology/2021/09/evolving-globs-of-yeast-may-unlock-mysteries-of-multicellular-life

In 1988, Richard Lenski, a thirty-one-year-old biologist at UC Irvine, started an experiment. He divided a population of a common bacterium, E. coli, into twelve flasks. Each flask was kept at thirty-seven degrees Celsius, and contained an identical cocktail of water, glucose, and other nutrients. Each day, as the bacteria replicated, Lenski transferred several drops of each cocktail to a new flask, and every so often he stored samples away in a freezer. His goal was to understand the mechanics of evolution. How quickly, effectively, creatively, and consistently do microorganisms improve their reproductive fitness?

Read more here:

https://www.newyorker.com/science/annals-of-medicine/how-will-the-coronavirus-evolve

Link: https://www.youtube.com/watch?v=w4sLAQvEH-M&t=1s