Paris Sciences et Lettres (PSL) University hosted a winter school centered around Quantitative Viral Dynamics Across Scales. Organized by Prof. Weitz, the school ran for one week (March 21-25, 2022) in Paris, France, and aimed to “bring thought leaders in dialogue with the next generation of early-career scientists to advance the integrative study of quantitative viral dynamics across scales.” Prof. Weitz also served as a lecturer and Scientific Committee Chair. Other guest lecturers include Weitz Group research scientist David Demory, postdoctoral fellow Jacopo Marchi, and several group collaborators.
Plants – even microscopic ones – come in all shapes and sizes. As important players in the Earth’s carbon cycle, marine researchers are particularly interested in what determines the size of phytoplankton, photosynthesizing microorganisms that live near the ocean’s surface. When these microscopic plant-like organisms sink to deeper ocean waters, they sequester the carbon they’ve used for photosynthesis, preventing the carbon from returning to the atmosphere as carbon dioxide. Phytoplankton size can determine whether they sink or stay near the surface to absorb more carbon dioxide, impacting the efficiency with which that carbon sinks to the deep ocean. With an important role in the ocean’s carbon cycle, the factors that determine size distribution in phytoplankton communities is an area of ongoing research.
Prevailing theories of phytoplankton ecology suggest the size distribution of phytoplankton in the ocean is determined largely by two factors: water turbulence and nutrient availability. Such theories predict that water turbulence prevents larger cells from sinking to a depth where they no longer receive sunlight to photosynthesize. Larger phytoplankton also grow faster than their smaller counterparts in the presence of ample nutrients. Therefore, larger species are predicted to comprise most phytoplankton in regions with high turbulence and high nutrients, such as the Southern Ocean by Antarctica. In this paper, Behrenfeld et al. suggest a different way to view phytoplankton abundances. Instead of competition amongst phytoplankton for nutrients as the main determinant of phytoplankton size, Behrenfeld et al. investigate the role predator and prey interactions play in structuring these communities, and how well phytoplankton respond to sudden changes in their environment.
Though phytoplankton are eaten by animals as large as whales, their more common predators are zooplankton, microscopic animals similar to krill that are the intermediate step in the food chain between phytoplankton and larger fish. To understand how zooplankton might determine phytoplankton size, Behrenfeld et al. used a theoretical model to study a food web with phytoplankton (prey) and zooplankton (predators) of different sizes. In this model, smaller zooplankton were limited to eating smaller phytoplankton, but larger zooplankton could eat small to large phytoplankton. Using these assumptions, the authors were able to calculate how many phytoplankton there should be across different sizes. They compared their calculations to global observations of phytoplankton size distributions. While models based on competition for nutrients underestimate the relative proportion of small phytoplankton, Behrenfeld et al.’s model did not.
Though Behrenfeld et al.’s model accurately predicted phytoplankton size under normal conditions, their model did not consider seasonal changes in phytoplankton communities called ‘blooms.’ Blooms occur when changes in ocean mixing provide additional nutrients to the surface, ‘fertilizing’ phytoplankton and allowing them to grow to such a large population size that they can be seen from space. To explain seasonal blooms, the authors adjusted their model to incorporate how quickly different microorganisms respond to rapid changes in nutrient availability. While phytoplankton can respond more or less immediately to nutrient fertilization, predators have a lag time before they start eating more. The authors claim that small predators can respond to changes in food availability quickly while it takes longer for larger predators to respond to their environments. The slow response of large zooplankton gives an advantage to large phytoplankton, who don’t face increased grazing pressure until long after smaller phytoplankton do. By adapting each microorganism’s response time to nutrient changes, Behrenfeld et al.’s model was able to account for seasonal phytoplankton blooms.
Prevailing theories claim competition amongst phytoplankton of different sizes plays a primary role in structuring phytoplankton communities. Veering from traditional models, this study instead suggests zooplankton predation plays an important role in determining the phytoplankton landscape. Their model sufficiently explained real-world observations of phytoplankton communities in global observations, encouraging future research on phytoplankton community assembly to pay increased attention to predation and not just competition for nutrients.
For more information, read the full paper by Behrenfeld et al. (2021) published in ISME Communications.
Throughout the past year, virtual communication has been necessary but challenging. Yet, with the increase in virtual events, there have been unexpected benefits to scientific collaboration and education. These benefits were obvious as over 100 young scientists from Georgia Tech and around the world gathered around their laptops for this year’s virtual iteration of the annual Quantitative Biosciences Modeling Workshop. The hands-on workshop in May of 2021 aimed to teach scientists of all backgrounds and skill levels about how to use computational models to focus on the problem that has been on everyone’s mind for over a year: epidemics.
“The ongoing Covid-19 pandemic made it easy to pick a topic,” says Aaron Pfenning, one of the first-year PhD students in Quantitative Biosciences at Georgia Tech who was responsible for organizing the workshop. Open to graduate students, scientists, and faculty members across the world, the hands-on workshop provided an introduction to how computational modeling is applied to understand and combat epidemics. “The entire workshop was based on a single lab from a class we took in the fall, Foundations in Quantitative Biosciences,” explained Leo Wood, another first-year PhD student involved in organizing the workshop. “But that was enough to fill two days of workshop, and really taught our students a lot.”
The two-day virtual workshop was split into two components: public lectures and hands-on small group sessions. The plenary lectures spurred meaningful discussion on the past, present, and future of using computational models to respond to epidemics. Later, the Applied Bioinformatics Laboratory (ABiL) provided detailed a demonstration on how to create interactive tools dashboards using R Shiny. Between public talks, registered attendees took part in small group sessions. In these sessions, attendees were grouped by computer programming experience, and received hands-on instruction in developing computational models in their preferred programming language from graduate students and postdoctoral volunteers from Georgia Tech.
Yet, planning a workshop is not without its challenges, especially when it is virtual. “It definitely made it harder to enable socializing and networking which is particularly important for people early in their career,” explained Pfenning. Despite these challenges, Pfenning noticed unforeseen benefits to the workshop’s virtual nature. “At the same time, we had a greater outreach, though, as people across different time zones attended.” The workshop, typically attended only by students in the Atlanta area, was open to students across the world. Leo Wood was also taken aback by the international attendance, describing “In my [small group] session alone, I had one person in South Africa and one person in Germany.”
Sabrina Li, a PhD student in the School of Geography & Environment at Oxford University, found the workshop particularly useful for her work. “As someone that currently works on COVID-19 modeling,” Li explained, “I wanted to improve my understanding of the models and assumptions used by many studies in the literature.” Li, who discovered the workshop through Twitter, says she found the public talks engaging, and thought the hands-on sessions served as a great introduction to modeling biomedical data. “Given that I come from a non-epidemiology/biomedical sciences background, I appreciated the introductory content and theory, which helped me to better understand the math and logic used to formulate the models.”
For those who missed the workshop, the materials from the small group sessions, as well as recordings of the public lectures, are available online. Keep an eye out for next year’s Quantitative Biosciences Modeling Workshop – there are many more topics to explore through a modeling lens.
The open ocean is the world’s wettest desert, containing vanishingly small concentrations of nutrients. Yet phytoplankton, a type of water-dwelling microscopic plant, have adapted to grow and survive in this near barren environment. These microorganisms play a key role in the biological pump, an important process that drives the ocean’s carbon cycle and absorbs carbon dioxide from the atmosphere.
Research into phytoplankton’s adaptations to different nutrient levels around the world tells us a lot about the entire marine food web’s response to the changing ocean as climate change continues. Over the years, scientists have made efforts to determine which nutrients set the maximum limit of phytoplankton growth. Unfortunately, experimentally identifying which specific nutrients (usually nitrogen (N), phosphorus (P), or iron (Fe)) limit phytoplankton growth in the ocean is a strenuous process. Through bottle experiments involving incubating phytoplankton in differing nutrient conditions (ex: with added Fe, N, and P), scientists have been able to identify the limiting nutrients in the ocean for different organisms. Still, scientific models have difficulty systematically determining nutrient stress across the global ocean, forcing scientists to continue relying on laborious bottle experiments.
In an effort to create more reliable computational models, recent research by Ustick, et al. (2021) aimed to corroborate results from models of nutrient limitations in different oceans with results from previous bottle experiments. These bottle experiments – the current standard for confirming nutrient limitation – focused on nutrient limitations of Prochlorococcus, a type of cyanobacteria. Prochlorococcus is one of the most abundant phytoplankton in ocean waters and is better equipped to absorb nutrients than most other phytoplankton due to its small size. Therefore, if even Prochlorococcus were experiencing a particular nutrient stress, then we can hypothesize that the entire phytoplankton community might experience the same stress. While it is also possible this could not be the case, as different phytoplankton have different nutrient requirements, understanding the map of nutrient limitation for one major phytoplankton player is a great start and lays groundwork for further study of other species.
Ustick et al. (2021) compared bottle experiment results with compared with two indirect, but larger-scale ways of assessing Prochlorococcus nutrient limitation from the surfaces of the Atlantic, Pacific, and the Indian Ocean. Limiting nutrients for Prochlorococcus were inferred from DNA sequencing samples by looking at the prevalence of genes associated with stress for nitrogen, phosphorus, and iron. The presence of stress genes in different oceans was compared to bottle experiments, as well as the limiting nutrient predicted by a global climate model. It turned out that all three independent methods tended to agree on which regions were primarily iron, phosphorus, or nitrogen stressed.
The biogeochemical cycles in the ocean are affected by which organisms are absorbing what available nutrients. Phytoplankton makes up the base of aquatic food webs, so scientists can use them to represent the states of specific ocean ecosystems. Therefore, understanding how phytoplankton might respond to changing environments is important for understanding the impact on the whole food web, from microorganisms to large fish and marine mammals. Accurately identifying areas of nutrient stress is a key factor in determining ocean physiology and identifying whether different species are adapting to the ocean’s changing environment.
For more information, read the full paper by Ustick et al. (2021) published in Science.
In 1977, scientists in St. Petersburg discovered a virus that was infecting cyanobacteria, a type of photosynthesizing bacteria, living in water on the outskirts of the city. The virus, cyanophage S-2L, had an adaptation that broke traditional rules of DNA construction and may open a whole new world of possibilities for synthetic biology.
Generally, DNA code is determined by the sequence of four different building blocks called nucleotides: adenine (A), thymine (T), guanine (G), and cytosine (C). The two strands of DNA that form the famous double helix are connected by pairs of these nucleotides, A to T and G to C. Yet these bonds are not equal in strength. Because of an additional hydrogen bond between the two, G has a slightly stronger connection to C than A does to T.
However, the cyanophage S-2L virus had all cases of A substituted with a modified version, 2-aminoadenine (Z), which contains one more nitrogen than the traditional A nucleotide. In this virus, Z pairs up with T, a pair that forms three bonds similar to the G:C pair. The additional strength from the Z:T pairings increases the overall stability of the virus’s DNA. But what is the functional benefit of this expansion of the genetic alphabet?
In aquatic environments, where this virus comes from, a virus must drift through the water until it happens to bump into a host. Having more stable DNA helps to prevent UV damage that could destroy a virus in the environment waiting to encounter a host. The increased stability could also protect the virus against host defenses, which often center around special enzymes called nucleases that are designed to break down DNA.
New research published in Science has identified how a combination of host and virus genes create the Z bases and prevent them from getting mixed up with A as virus DNA is assembled. Scientists think this system may have applications in expanding the range of tools used in synthetic biology, including new ‘designer proteins’ that may have therapeutic applications.
Interestingly, Z-DNA may be older than life itself! A 2011 analysis of a carbon-containing meteorite revealed standard and nonstandard DNA structures, including the Z base. If the Z base was formed in space, it may also have been formed in conditions on Earth at the time of the beginning of life.
For more information, read the full study published by Grome and Isaacs in Science.